Issuu

mesozQic^

MESOZOIC GEOLOGY OF THE EASTERN AUSTRALIA PLATE CONFERENCE 23-26 September 1996 Sheraton Brisbane Hotel & Towers Queensland, Australia

DEPARTMENT OF M I N E S AND ENERGY

MINISTER'S FOREWORD Mesozoic 96 highlights the most extensive group of rocks in Queensland. This conference, hosted by the Geological Society of Australia (Queensland Division), aims to focus industry and academic attention on this important, yet understudied, group of rocks. The Department of Mines and Energy is pleased to sponsor the conference because of the importance of Mesozoic hosted resources to Queensland. Events in the Mesozoic Era, 65 to 251 million years ago, led to the present geographic distribution of Australia and its near neighbours. Mesozoic rocks contain oil and gas (Eromanga, Bowen and Surat Basins), coal (Surat Basin, Ipswich, Callide, Tarong, Maryborough and Bathurst Range), oil shale and vanadium (Julia Creek), groundwater (Great Artesian Basin), clays (bentonite and kaolinite), opals (western Queensland), gold (Gympie, Mt Rawdon, Cracow, North Arm), magnetite (Biggenden) and other minerals. Although the energy minerals and groundwater are currently the most valuable for the Queensland economy, there is significant potential for further mineral discovery. This conference is timely for the Department, as many of its GEOMAP 2005 projects are now investigating the eastern areas of the State where potential for mineralisation in Mesozoic rocks is greatest. Its AIRDATA geophysical data acquisition project also covers much of the mineralised and potentially mineralised areas. The Cooper-Eromanga Basin Project will investigate the petroleum potential of the Mesozoic basins strata, Mesozoic 96 provides an excellent opportunity for the Department to obtain feedback to help design more effective projects which optimise our understanding of the Mesozoic rocks and their resources. The organisers are to be congratulated for providing an opportunity for a broad spectrum of geoscientists from academia, industry and government to contribute to better understanding of the Mesozoic era, and for the extensive program prepared.

The Hon Tom Gilmore MLA Queensland Department of Mines & Energy, Brisbane

r i

CONTENTS Presenters listed in alphabetical order Legend • Bold = Keynote speaker Adams, Christopher

A Queensland provenance for New Zealand Permo-Triassic metagreywacke terranes: a review of the age and isotopic evidence

Alder, David

High resolution aeromagnetic and gravity survey used to highlight basin geometry

and structure

Recent investigations of the Great Australian Basin in New South Wales

Allen, Charlotte

The petrogenesis of the Mesozoic rocks, Bowen Region, Queensland, and changes of Tectonic setting from 145 to 100 Ma

Allen, George

Application of sequence stratigraphy to continental successions: implications for Mesozoic cratonic interior basins of eastem Australia

Angus, Maree

The Mount Rawdon gold deposit, southeast Queensland

Ashley, Paul

Metallogenesis related to Triassic magmatism in the central and southem New England Orogen

Baker, Graham

Industrial mineral opportunities in Queensland

147

Barron, Larry

Sub-basaltic alteration Copeton Bingara, NSWwithin diamond-bearing tertiary deep leads, Methane sorption capacity of Upper Cretaceous coals from the Greymouth Coalfield, New Zealand

Beamish, Basil

Combustion behaviour of New Zealand, late cretaceous coals determined by thermogravimetry

Berry, Keith

Eromanga Basin water supply development for Olympic Dam operations

Biggs, Mark

Iron mineralogy of the CalUde coal measures and its effects on coal quality

Birrell, Russell

The application of mobile metal ion surface soil geochemistry to mineral explorations

Black, Philippa

Mesozoic evolution of the Norfolk Ridge System: evidence from New Caledonia and Northem New Zealand

Blevin, Phillip

Intemal evolution and metallogeny of Permo-Triassic high-K granites in the Tenterfield-Stanthorpe region, southem New England Orogen, Australia

Permo-Triassic granite metallogeny of the New England Orogen

101

Boreham, Chris

The significance of Mid-Cretaceous burial and uplift on the maturation and petroleum generation in the Bowen and Surat Basins, eastem Australia

104

Bradshaw, John

The Mesozoic continental margin: Carboniferous-Mesozoic arc terranes in west Antarctica, New Zealand and Australia

114

Bryan, Scott

The Whitsunday Volcanic Province (central Queensland) and the Otway/Gippsland Basins (Victoria): a comparison of Early Cretaceous rifl-related volcano-sedimentary successions

124

Carmichael, David

Bentonites in Queensland

134

Clark, Ian

Interpretation of insitu stress measurements in basin analysis

137

Cooper, Wilson

Industrial mineral opportunities in Queensland

147

Cranfield, Leonard

Geology and mineralisation of the Gympie Province

156

Creenaune, Patrick

The geology and minerahsation of the Mt Cannindah porphyry copper-gold system

157

Cunneen, Ron

Recent developments in the geology of the Gympie Goldfields

162

Eadington, Peter

An outline of palaeoformation water compositions and flow pattems in the Eromanga and Cooper Basins through Mesozoic time

167

Evans, Peter

Fluoride anomalies in aquifers of the Queensland sector of The Great Artesian Basin and their significance

172

Faraj, Basim

Regional clay and carbonate mineralisation of Late Permian coal measures cleats, Bowen Basin, Australia

179

. 180

Fielding, Christopher

Mesozoic sedimentary basins and resources in eastern Australia

Giblin, Angela

An application of groundwater geochemistry to the detection of prospective basement beneath Mesozoic cover

Gleadow, Andrew

Imaging the thermotectonic evolution of eastem Australia during the Mesozoic from fission track dating of apatites

. 195

Golding, Sue

Nature and source of carbonate mineralisation in Bowen Basin coals: implications for the origin of coalseam gases

. 205

Gould, Rod

Cobalt at Mt Manganese, central Queensland

. 213

Green, Peter

Stratigraphic relationships between latest Triassic - Early Cretaceous basins of Queensland

218

186

Gust, David

Magmatism of the NEO: space, time and compositional relationships

224

Habermehl, Rien

Groundwater movement and hydrochemistry of the Great Artesian Basin, Australia

228

Hastie, Lynn

Application for magnetotellurics in petroleum exploration in Queensland and New Guinea

237

Henderson, Robert

Sequence stratigraphic and palaeoenvironmental interpretation of the Bathurst Island Group, Money Shoals Basin

. 245

Hill, Steven

Mesozoic regolith and paleo-landscape features in southeastem Australia

246

Hillier, John

The Great Artesian Basin - management of the water resources after 100 years of development

. 251

Holcombe, Rod

Tectonic evolution of the eastem Gondwanaland Margin during the Late Palaeozoic and Early Mesozoic

. 517

Horn, Anthony

The Mesozoic aquifers of Cape York Peninsula, Queensland

256

Huber, Richie

The contribution of airbome geophysics to exploration for magnetite in the northem New England Fold Belt

264

Johnson, Laurie

Goondicum ilmenite deposits

279

Jones, Michael

Clues to the early Mesozoic structural evolution of southeastem Queensland through integrated interpretation of remote sensing and aeromagnetic data

409

Jorgensen, Peter

Debris flow deposits in the Late Triassic Callide Coal Measures and their impact upon coal mining considerations

283

Kassan, Jochen

Aspects of Triassic sedimentation in the Bowen Basin, Queensland

Kesler, Stephen

Arc evolution and ore deposit models

.. 302

Korsch, Russell

Mesozoic deformational events in eastem Australia and their impact on onshore sedimentary basins

, 308

Lackie, Mark

Thermal events in the Sydney-Bowen Basin as defined by palaeomagnetism

319

Laird, Malcolm

Mid and early Late Cretaceous break-up basins of the South Island, New Zealand

329

Lanzilli, Elio

The Hutton Sandstone to Birkenhead Formation Transition, Gidgealpa southem some: showing an associated unconformity and provenance change from craton-derived .. 337 to volcanic-arc-derived sediment

Li, Jiaorong

Chemical character of groundwater in the Walloon Coal Measures of southeast Queensland

Loutit, Tom

Petroleum systems of the Bowen, Surat and Gunnedah Basins

350

Mackie, Steve

Geological controls on hydrocarbon accumulations in the Eromanga Basin, south west Queensland - a petroleum system perspective

356

McKellar, John

Palynofloral and megafloral indications of palaeoclimate in the Late Triassic, Jurassic and Early Cretaceous of southeastem Queensland 366

McMahon, Gerard

The relationship between groundwater chemical type and Jurassic sedimentary formations: the example of the Sandy Creek catchment, Lockyer, southeast Queensland

374

McNeil, Vivienne

Discrimination of groundwaters in the Great Artesian Basin, Queensland, using the Cl/Mg ratio

383

292

.. 343

MUler, Elizabeth Mortimer, Nicholas

Extensional and compressional regimes of the Cretaceous Circum Pacific implications for the eastern Australia Plate

390

The Mesozoic basement of New Zealand

391

Permian/Triassic sedimentary sequences in north-westem South Island, New Zealand, and correlation with Australian Gondwana sequences

510

Moussavi-Harami, Reza Burial history of the Eromanga Basin in northeast South Australia

400

Murray, Alasdair

Gold mineralisation in the Grahams Creek Volcanics, Maryborough Basin

404

Nichols, Wesley

414

O'Sullivan, Paul

Geophysical faultQueensland location at Callide Coalfields Trap Gully Mine, Callide Basin,trials east for central Late Mesozoic to Early Cenozoic thermotectonic history of the Sydney Basin and the eastern Lachlan Fold Belt, Australia

424

Passmore, Virginia

Australia's premier onshore oil province - the Eromanga Basin

433

Pecover, Simon

Sapphire-producing Cretaceous/Tertiary volcano-fluvial deposits in eastern Australia

442

A new genetic model for the origin of opal in Cretaceous sediments of the Great Australian Basin

450

Pollard, Peter

Granite-hosted, disseminated gold mineralisation at Timbarra, New South Wales

507

Pratt, David

Recent investigations of the Great Australian Basin in New South Wales

Prowse, Geoffrey

Cretaceous aquifers of the Northem Territory's Top End

Raza, Asaf

9 455

Mesozoic denudation and tectonics of the central margin of Austraha: fission track thermochronology

464

Roach, Andrew

Late Triassic volcanism of the Ipswich Basin

476

Robertson, Allan

Diamonds in Queensland

485

Ruxton, Peter Saxby, John

Geology of the Gold Ridge epithermal gold deposit, Guadalcanal Island, Solmon Islands Maturation of Queensland Mesozoic coals: thermal and microscopic analysis of

267

petroleum and gas generation

489

Scott, Steven

Coal seam methane potential of the Walloon coals

496

Senior, Brian

Exploration for precious opal within the Eromanga and Surat Basins, Queensland and New South Wales

501

Sutherland, Frederick Linstead The Rockhanpton plume and its late Mesozoic trace

519

Symonds, Philip

Mesozoic rift basin development off eastem Australia

528

Vasconcelos, Paulo

Geochronological evidence for the preservation of Cretaceous weathering profiles in northwestem Queensland

543

The crystal chemistry and the genesis of chrysoprase

545

Wai, Khin Maung

Geochemistry, formation ophiolitic in Papua implications forenvironment relationshipand of Indian Plateofand eastem rocks Australia Plate New Guinea: Plate tectonic evolution of eastem Papua New Guinea: model of emplacement of ophiolitic rocks in New Guinea

546 547

Ward, Colin

Triassic sedimentation in the Gunnedah and Sydney Basins, New South Wales

272

Weaver, Stephen

Cretaceous magmatism, make-up and break-up of the SW Pacific Gondwana Margin

548

Willey, Edwin

The Woogaroo Subgroup in the HeUdon Hills district, southeast Queensland stratigraphic appraisal and tectonic implications

557

Williams, Neil

The Mesozoic geology of the eastern Australia Plate ~ an overview

564

Worden, John

Major extensional events recorded by Zircon Xenocrystsfi-omthe central Queensland gemfields

569

Yago, Joel

Sedimentology of the Middle Jurassic Walloon Coal Measures in the Great Artesian Basin of eastem Australia

574

A QUEENSLAND PROVENANCE FOR NEW ZEALAND PERMO-TRI ASSIC TORLESSE METAGREYWACKE TERRANES: A REVIEW OF THE AGE AND ISOTOPIC EVIDENCE CJADAMS Institute of Geological & Nuclear Sciences, P 0 Box 31312, Lower Hutt, New Zealand SUMMARY Rubidium-strontium (Rb-Sr) whole-rock isochron dating of Torlesse Supergroup metasediments of the Rakaia terrane of New Zealand yield initial 87Sr/86Sr ratios at the time of metamorphism (latest Triassic-early Jurassic) as low as 0.7065 indicating that at the time of their Permo-Triassic deposition they were derived from a dominantly I-type granitoid terrane (with only subordinate older and more radiogenic country rocks) and in total, having a bulk Rb/Sr ratio 0.9-1.0. 40Ar/39Ar single crystal mineral ages of detrital muscovites from similar Torlesse metagreywackes show an age population dominated by a major (80%) Permian-Triassic (280-205Ma) peak and a minor (15%) mid-Paleozoic (460-410Ma) peak. However, there is a complete absence of peaks in the range 500-460Ma and 410-330Ma, features that would be expected if the source area of the Torlesse included Lachlan or Ross/Adelaide Fold-Belt granitoid terranes of southeastem Australia and Antarctica. Rather, both bulk-rock initial strontium isotope and 40Ar/39Ar detrital mineral age characteristics point uniquely to a provenance in the Hunter-Bowen Fold-Belt and immediately adjacent older terranes of eastem Queensland. A model is proposed for a Torlesse depocentre originating at low-latitude near the Permian-Triassic Queensland Gondwana margin which was subsequently tectonically transported, by strike-slip motion southwards around the Pacific margin of Gondwana, to a mid-latitude New Zealand position by the mid-Cretaceous. INTRODUCTION Widespread, monotonous turbibidite metasedimentary sequences (prehnite-pumpellyite or zeolite facies) of the Torlesse Supergroup (see review in Suggate, 1978) make up much of the main axial mountain ranges of New Zealand (fig.l) and althou^ sparsely fossiliferous (see Campbell and Warren, 1965; Speden, 1976), extend from Carboniferous to Cretaceous, with a major part being Permian-Triassic. They are structurally complex with no formally defined base or top and only rarely have internal stratigraphic boundaries. Earlier litho-stratigraphic definition as Torlesse Supergroup has in recent years given way to a terrane classification (Bishop et al., 1985), in which a major Torlesse Terrane sensu lato (strictly a superterrane) occupies much of the North and South Island Mesozoic basement of Canterbury and Wellington regions (fig.l) and is distinguished from a less well understood, more heterogeneous Waipapa Terrane (Sporli, 1978), mostly occurring in the northem part of the Auckland and Northland regions of the North Island (fig.l). Greenschist facies equivalents of these metasediments extend into Haast Schist (see review in Suggate, 1978; also Mortimer, 1993a,b) in Otago, Alpine, Marlborough and Kaimanawa regions (fig.l). The Torlesse (super)Terrane has been divided into Permo-Triassic, Rakaia and Jurassic-Cretaceous, Pahau, Terranes, separated by a well-recognised melange zone (Esk Head Melange) in the South Island (Bishop et al., 1985) and southem North Island (Rimutuka Melange, informal name). Rakaia Terrane metagreywacke sequences clearly extend into Haast Schist (fig.l) as part of Otago Schist in Otago, Alpine Schist in Canterbury and Marlborough Schist in eastem Marlborough (Mortimer,1993a). The main structural-metamorphic events in the Torlesse terranes (sensu lato) are thought to be primarily Triassic-Jurassic in the Rakaia Terrane (see geochronological studies by Adams et al. (1985), Adams and Graham (1996 in press), Graham and Korsch (1989) and Graham and Adams (1990) and late Jurassic-Cretaceous in the Pahau Terrane (Rangitata Orogeny phases I and II) (Bradshaw et al, 1981; Adams et al., 1985; Adams and Graham, 1996 in press). There are important late structural and to a lesser extent metamorphic events, related to complex imbrication of the metagreywackes and formation of boundary melanges, up to the early-mid Cretaceous (Adams and Graham, 1993; 1996 in press). The Rakaia Terrane metagreywacke suites are uniform turbidite sequences of relatively quartzose character, derived from a dominantly granitoid source area with only very minor volcaniclastic input (MacKinnon, 1983); indeed volcanic horizons are rare in the Torlesse. Pahau Terrane metasediments are similar but with decidely greater volcaniclastic input (MacKinnon, 1983). The position of these two Torlesse terranes is curious however, since both form the easternmost terranes of the New Zealand continental margin and are separated from the southwest Pacific Paleozoic Gondwana margin (the Buller and Takaka Terranes of Nelson (fig.l) and west coast regions of the South Island) and their likely correlates in southeastem Australia and Antarctica, by several other Permian-Triassic-Jurassic terranes of quite different character; basic-intermediate volcanic arc terranes with associated volcaniclastic metasediments (Brook Street, Maitai, Murihiku, Caples/Pelorus Terranes) (see review by Coombs et al. 1976). Therefore, it is not possible to devise a simple scheme to derive Torlesse-type metasediments in the east from their most obvious immediate source at the Gondwana margin to the west. This dilemma regarding the origin of the Torlesse has been the subject of many tectonic solutions attempting to reassemble the New Zealand Paleozoic-Mesozoic tectono-sedimentary terranes to accommodate an appropriate provenance for the Torlesse, either from the Antarctic, Australian or New Zealand Gondwana margin or some

more hypothetical lost "Pacifica" continent (see discussions in Landis and Bishop, 1972; Andrews et al., 1974; Coombs et al., 1976; Kamp, 1980; Bradshaw et al., 1981; MacKinnon, 1983). Rb-Sr and K-Ar geochronological studies of the metasedimentary sequences have attempted to underpin such tectonic models by determining the timing of metamorphism and subsequent cooling history of several terranes and, by comparing these, to work out a timetable for their assembly. A by-product of Rb-Sr whole-rock isochron geochronological studies of the Rakaia Terrane Torlesse metasediments (see for example Graham and Korsch, 1989; Graham and Adams, 1990; Adams and Graham, 1993 and 1996 in press) has been the recognition of distinctive uniform initial 87Sr/86Sr isotopic ratios at the time of metamorphism (examples determined are from upper prehnite-pumpellyite to lowest greenscist facies sequences). These can be distinguished from equivalent data of approximately the same metamorphic age in adjacent Pahau and Waipapa Terranes (Adams and Graham, 1996 in press). More recently, 40Ar/39Ar ages have been determined on detrital muscovites from Torlesse metagreywackes at the lowest metamorphic grade (but from the same localities) in order to assess the influence of various detrital mineral age populations on the bulk-rock initial strontium isotopic variations (Adams and Kelley, 1996 under review). (A parallel study is also being made of detrital zircon age populations in the same samples). INITIAL STRONTIUM ISOTOPE STUDIES IN THE TORLESSE Rb-Sr isochron dating of the Torlesse-type metasediments (this includes types now classified within the adjacent Waipapa and Caples/Pelorus sequences) has been done on whole-rock samples from individual turbidite units from a single locality, ranging from peUte through siltstone to fine greywacke (e.g. Adams and Graham, 1996 in press). The isochron ages are thought to refer to the time of strontium isotopic homogenisation between them, at the onset of burial metamorphism when new mineral reactions involving Ca,Sr,K,Rb-bearing components such as clays, micas, feldspars, calcite, zeolite, epidote are occurring (Korsch and Graham, 1989; Adams and Graham, 1993). This is probably facilitated by the fluid phase during sediment dewatering and/or subsequent hydration/dehydration reactions. In suitable environments this will continue during progressive metamorphism and perhaps even post-metamorphic cooling and/or retrogressive metamorphism. Localities were chosen to cover several such events in the metamorphic history but since they operated on same sediment precursor, it is possible to build up a pattem of increasing initial 87Sr/86Sr ratio (at the time of metamorphism, or more generally, isotopic homogenistion) with decreasing age of metamorphism (or homogenisation). These data plotted on Rb/Sr intersection diagram, i.e. age (t) on x-axis and initial 87Sr/86Sr (i) on the y-axis, demonstrate the isotopic evolution of groups of samples with the same ancestry, in the form of arrays having a positive slope whose gradient is proportional to their bulk (on a km-scale) Rb/Sr ratio (see examples in Adams and Graham, 1993; Adams and Graham, 1996 in press). A summary diagram of this type (fig.2), in which (t)-(i) data for each isochron are enclosed in envelopes to designate separate imbricate packets, is taken from recent work on Rakaia terrane metagreywackes (Torlesse Supergroup in lithostratigraphic terms) on a well-studied cross-section along the south Wellington coast. North Island (Adams and Graham, 1996 in pressi This section is considered to be a stack of several km-scale imbricate packets of Triassic meta-turbidites, abutting a tectonic melange extending into Pahau Terrane metagreywackes. A principal low-grade metamorphism occurred here in the Triassic to early Jurassic (210-185Ma) but there are local ductile zones which record younger metamorphic events in the period latest-Jurassic to early Cretaceous (145-120Ma). For this reason it can be seen that ages from localities in the same imbricate packet range from about 210 Ma to about 120Ma with corresponding initial 87Sr/86Sr isotopic ratios (at time of metamorphism) defining discreet arrays with a slope corresponding to a bulk Rb/Sr ratio about 0.9-1.0 and with minimum initial 87Sr/86Sr ratios 0.7065-0.7070. These are quite different from analogous data for adjacent Pahau and Waipapa sequences in the south Wellington region (fig.2). These Rakaia data thus indicate a source area during Triassic deposition (215-230Ma) which was, on petrographic and geochemical grounds (MacKinnon, 1983), dominantly granitoid, but with a bulk Rb/Sr ratio 0.9-1.0 and initial 87Sr/86Sr ratio <0.7065-0.7070. This impUes a granitoid provenance of I- rather than S-type, early Triassic or older. These critera thus exclude the majority of the extensive mid-Paleozoic granites of the Lachlan Fold-Belt of southeastem Australia and their analogues in the Ross Sea regions of Antarctica (as well as early-Paleozoic granitoid terranes of the Ross/Adelaide Fold-Belt), the majority of which have initial 87Sr/86Sr ratios 0.706-0.715 (representative data are given in Gray,1990; Borg et al., 1987; Vetter and Tessensohn,1987; Adams, 1987). Of course, there are exceptions but they are volumetrically minor. It should be bome in mind that in general terms, the source region of the Torlesse should cover the same area as the metaturbidites themselves (about 1000 x 200km). Several possible source areas around the Pacific margin of Gondwana could be nominated, from southem South America/Antarctic Peninsula/Marie Byrd Land (West Antarctica); North Victoria Land (East Antarctica)/southeastem Austraha (TasmaniaA^ictoria/southem New South Wales) and northeastern Australia (northem New South Wales/Queensland)/ Papua New Guinea, all of which have widespread Paleozoic (and locally Precambrian) granitoid/gneiss basement terranes. However, of these, only the extensive and dominantly Late Paleozoic-early Mesozoic granitoids (mostly I-type) of the Hunter Bowen Fold-Belt of Queensland and northem New South Wales (Richards, 1980; Shaw and Flood, 1981; Hensel et al., 1985; Kent, 1994) satisfy the source area criteria for the Torlesse, the great majority of which have initial 87Sr/86Sr ratios <0.706 (representative data given in Webb and McDougall, 1968) and some much lower, <0.704 (Shaw and Flood, 1981). Significantly, in older terranes adjacent to the west, there is a complex assortment of minor early to mid-Paleozoic granitoids (see review in Richards,

1980) with inliers of Precambrian ortho/paragneiss (see review in Withnall et al., 1980 and representative age data in Richards et al., 1966; Black et al., 1979; Black and McCulloch, 1990) and early Paleozoic metasediments (see review in Henderson 1980). Thus it becomes important to know if the pattem of detrital mineral ages (such as 40Ar/39Ar mica and U-Pb zircon) in Torlesse metagreywackes do, in fact, reflect the complexity of mineral ages seen in a potential Queensland source area. 40AR/39AR AGES OF TORLESSE DETRITAL MUSCOVITES Adams and Kelley (1996 under review) have reported 40Ar/39Ar single crystal ages for representative populations of detrital micas (biotite and muscovite)fromRakaia Terrane Torlesse Supergroup metagreywackes of both North and South Island (and also from the older, early Paleozoic Greenland Group metagreywackes widespread in the Buller Terrane of the west coast. South Island). Of these, only coarse detrital muscovites seem to have withstood the thermal effect of later thermal (metamorphic) overprints and a summary compilation of 40Ar/39Ar age data from several Permo-Triassic Rakaia Terrane localities is shown in fig.3. It is immediately obvious that there are two significant peaks; a major one (80% of total) at 210-280Ma and a minor one (15% of total) at 410-470Ma. Equally significant is the absence of any real peaks at about 330-410Ma, which is a major component of mica age populations (compiled mainly from K-Ar age data sources listed frilly in Adams and Kelley (1996 under review)(fig..3) in the Lachlan Fold-Belt granitoid terranes of southeastem Austraha (and analogues in Antarctica) or at about 460-500Ma, the major mineral age peak of Ross (and Adelaide) Fold-Belt granitoid/gneiss terranes of Antarctica (and South Australia). The Rakaia Terrane detrital muscovite age populations match more comfortably with comparable mica age data from the Hunter-Bowen Fold-Belt of northem New South Wales and Queensland where there are very extensive Permo-Triassic granitoids (220-290Ma) intruding early Paleozoic (and Precambrian) metasediments metamorphosed in mid-Paleozoic times (obliterating much of the older, Precambrian mineral age patterns (fig.3; data sources listed in Adams and Kelley, 1996 under review). It is concluded therefore, that the Queensland sector of the Hunter-Bowen Fold-Belt granitoids and immediately adjacent older terranes could have provided a suitable source, in Permo-Triassic times, for the Torlesse Supergroup greywackes seen in the Rakaia Terrane of New Zealand. ORIGIN OF THE TORLESSE The age and isotopic data described above imply a very large separation, about 2000km, between original source rocks in Queensland and the present location of the Torlesse metagreywackes in New Zealand. It is possible that the Torlesse depocentre originated in a New Zealand position but was fed from the north, over a long distance, by large river/delta/submarine fan systems; an analogy would be the major rivers of northem India and submarine fans (Bengal Fan) of the Indian Ocean. This model is somewhat contradicted by the occasional but persistent evidence of shallower water sedimentation in some of the Torlesse sequences (conglomerates (Smale, 1980; Andrews et al., 1974), deltaic features (Andrews, 1974), plant-bearing beds (summarised in MacKinnon, 1983)). Altematively, it is possible that the Torlesse depocentre was originally located near Queensland in the Permian-Triassic and then subsequently tectonically transported, as a terrane, by strike-slip motion around the Pacific margin of Gondwana during the Jurassic-Cretaceous. This would be an analogous motion to the coastal suspect terranes of westem North America with respect to the the northeastem Pacific Ocean. A lower latitude depocentre for the Torlesse is supported by paleontological evidence indicating a Tethyan affinity for some Permo-Triassic Torlesse biotas, thus originating at a lower latitude (Grant-Mackie, 1985; Aita and Sporli, 1992) than their present mid-latitude position would suggest. TMs subject is treated in a little more detail in Adams and Kelley (1996 under review) but clearly the paleontological and tectonic implications of the model suggested by these present age and isotopic data require much more detailed analysis and research. ACKNOWLEDGEMENTS This review summarises more detailed reported elsewhere, with co-workers Ian Graham (Rb-Sr dating and initial strontium isotope ratios) and Simon Kelley (Ar-Ar dating) and their collaboration is much appreciated. REFERENCES Adams, C.J. and Kelley, S., 1996 (under review). Provenance of Permo-Triassic and Ordovician Metagreywacke Terranes in New Zealand: evidencefirom40Ar/39Ar dating of detrital micas (under review) (paper submitted to Geological Society of America Bulletin). Adams, C.J., 1987. Geochronology of granite terranes in the Ford Ranges, Marie Byrd Land, West Antarctica. New Zealand Joumal of Geology and Geophysics, v.30, p.57-22. Adams, C.J., Bishop, D.G. and Gabites, I.E., 1985. Potassium - argon age studies of a low-grade progressively metamorphosed greywacke sequence, Dansey Pass, South Island, New Zealand. Joumal of the Geological Society of London, v.l42, p.339-349. Adams, C.J. and Graham, I.J. (1995 in press). Geochronological studies of metamorphism and tectonism in a Mesozoic accretionary prism: age/isotopic profiles of Triassic-Jurassic metagreywackes. South Wellington coast. New Zealand. New Zealand Joumal of Geology and Geophysics, v. , p. I

Adams, C J. and Graham, IJ., 1993. KAr and Rb-Sr age studies of the metamorphism and quartz vein Au mineralisation on Terawhiti Hill, near Wellington, New Zealand. Chemical Geology (Isotope Geoscience Section), v. 103, p.235-249. Aita, Y. and Sporli, K.B., 1992. Tectonic and paleobiogeographic significance of radiolarian microfauna in the Permian to Mesozoic basement rocks of the North Island, New Zealand. Palaeogeography, Palaeoclimatology, Palaeoecology, V.96, p.103-125. Andrews, P.B., 1974. Deltaic sediments. Upper Triassic Torlesse Supergroup, Broken River, North Canterbury. New Zealand Journal of Geology and Geophysics, v.l7(2), p.271-299. Andrews, P.B., Speden, I.G. and Bradshaw, J.D., 1976. Lithological and palaeontological content of the Carboniferous Jurassic Canterbury Suite, South Island, New Zealand. New Zealand Joumal of Geology and Geophysics, v. 19(6), p.791-819. Bishop, D.G., Bradshaw, J.D. and Landis, C.A., 1985. Provisional terrane map of South Island, New Zealand. In Howell, D.G. (ed) Tectonostratigraphic Terranes. Circum-Pacific Council for Energy and Mineral Resources Earth Science Series No.l, Houston Black, L.R, Bell, T.H. Rubenach, M.J. and Withnall, I.W., 1979. Geochronology of discrete structural metamorphic events in a multiply deformed Precambrian terrain. Tectonophysics, v.54, pl03-137. Borg, S.G., Stump, E., Chappell, B.W., McCulloch, M.J., Wybom, D., Armstrong, R.L. and Holloway, J.R., 1987. Granitoids of northem Victoria Land, Antarctica: implications of chemical and isotopic variations to regional crustal structure and tectonics. American Joumal of Science, 287,127-169. Campbell, J.D. and Warren, G., 1965. Fossil localities of the Torlesse Group in the South Island. Transactions of the Royal Society of New Zealand, v.3(8), p.99-137. Coombs, D.S., Landis, C.A., Norris, R.J., Sinton, J.M., Boms, D.J. and Craw, D. Hie Dun Mountain Ophiolite Belt, New Zealand, its tectonic setting, constitution, and origin, with special reference to the southem portion. American Joumal of Science, v.276, p.561-603. Graham, LJ. and Adams, C.J., 1990. Rb-Sr and K.-Ar geochronology of turbidites and metavolcanics at Red Rocks, Wellington, New Zealand. New Zealand Joumal of Geology and Geophysics, v.33, p.193-200. Graham, LJ. and Korsch, R.J., 1989, Rb-Sr resetting ages and chemical characterisation of turbidites in an accretionary wedge: Torlesse Complex, Otaki Gorge, New Zealand. Geological Society of America Bulletin, v.lOl, p.355-363. Grant-Mackie, J.A., 1985. New Zealand - New Caledonian Permian - Jurassic faunas, biogeography and terranes. In Cooper, R.A (ed) Homibrook Symposium, New Zealand Geological Survey Record, Department of Scientific and Industrial Research, New Zealand, p.50-52. Gray, C.M., 1990. A strontium isotopic traverse across the granitic rocks of Southeastem Australia: Petrogenetic and tectonic mplications. Austrahan Joumal of Earth Sciences, 37, 331-349. Henderson, R.A., 1980. Structural outline and summary geological history for northeastem Australia. In Henderson, R.A. and Stephenson, P.J. (eds.). The Geology and Geophysics of northeastem Australia. Geological Society of Australia (Queensland Division), 1-26. Kamp, P.J.J., 1980. Pacifica and New Zealand. Proposed elements in Gondwanaland's history. Nature, v.288, p.659-664. Kent, A.J.R., 1994. Geochronology and geochemistry of Palaeozoic intrusive rocks in the Rockvale region, southem New England Origen, New South Wales. Australian Joumal of Earth Sciences, v.41, p.365-379. Landis, CA. and Bishop, D.G., 1972. Plate tectonics and regional stratigraphic-metamorphic relations in the southem part of the New Zealand geosyncline. Geological Society of America Bulletin, v.83, p.2267-2284. Mortimer, N., 1993b, Geology of aago Schist and adjacent rocks. 1:500,000. Institute of Geological and Nuclear Sciences, geological map 7; Institute of Geological and Nuclear Sciences Limited, Lower Hutt, New Zealand. Shaw, S.E. and Flood, R.H. 1981. The New England Batholith, Eastem Austraha. Geochemical variations in time and space. Joumal of Geophysical Research, 86, Bll, 1053-10544. Richards, D.N.G., 1980. Palaeozoic granitoids of northeastem Australia. In Henderson, R.A. and Stephenson, RE (eds.). The geology and geophysics of northeastem Australia. Geological Society of Australia (Queensland Division), p.229-246. Speden, I.G., 1976. Fossil localities in Torlesse Rocks of the North Island, New Zealand. Joumal of the Royal Society of New Zealand, v.6(l), R73-9. Sporli, K.B., 1978. Mesozoic tectonics. North Island, New Zealand. Geological Society of America Bulletin, v.89, p.415-425.

McDougall, I. and Leggo, R, 1965. Isotopic age determinations on granitic rocks from Tasmania. Journal of the Geological Society of Australia, v.l2(2), p.295-332. MacKinnon, T.C., 1983. Origin of the Torlesse terrane and coeval rocks. South Island, New Zealand. Geological Society of America Bulletin, v.94, p.967-985. Mortimer, N., 1993. Metamorphic zones, terranes, and Cenozoic faults in the Marlborough Sdnst, New Zealand. New Zealand Joumal of Geology and Geophysics, v. 36, p.357-368. Speden, I.G., 1976. Fossil localities in Torlesse Rocks of the North Island, New Zealand. Joumal of the Royal Society of New Zealand, v.6(l),p.73-91. Smale, D., 1980. Akatarawa Conglomerate (Permian), Lake Aviemore, South Canterbury. New Zealand Joumal of Geology and Geophysics, v.23, p.279-292. Suggate, R.R (Chief Editor), 1978. The Geology of New Zealand. Government Printer, Wellington (2 volumes), 343 pp. Vetter, U. and Tessensohn, F., 1987. S- and I-type granitoids of North Victoria Land, Antarctica, and their inferred geotectonic setting. Geologische Rundschan, 76/1, 233-243. Webb, A.W. and McDougall, I., 1968. The geochronology of the igneos rocks of Eastern Queensland. Joumal of the Geological Society of Australia, v.l5, p.313-346. Withnall, I.W., Bain, J.H.C. and Rubenach, M.J., 1980. The Precambrian geology of northeastern Queensland. In Henderson, R.A. and Stephenson, P.J. (eds.). The Geology and Geophysics of northeastern Australia, Geological Society of Australia, (Queensland Division), 109-127.

Figure 1. New Zealand Permian-Triassic-Jurassic 'Torlesse" metasedimentary terranes (ligjit shading); including Rakaia and Pahau Terranes of Wellington, Canterbury and Marlborough regions. Also shown is Waipapa Terrane of Northland and Auckland. All of these grade into Haast Schist (dark shading) of Otago and Canterbury, South Island and Kaimanawa Schist of North Island.

Figure 2: Intersection diagram, i.e. Age (Ma) v initial 87Sr/86Sr ratio at time of metamorphism for Torlesse Supergroup metasediments of Rakaia Terrane of Wellington, New Zealand (from Adams & Graham, 1996). Each data box refers to Rb-Sr whole-rock isochron data for localities (arabic numerals) and extural zones (roman numerals) given in reference cited. Data envelopes shaded refer to groups of localities considered to originate in a single imbricate packet: "Terawhiti" and "Eastem" are true. Rakaia-Terrane Torlesse but "Kapiti" shows similarities to Waipapa Terrane analogues (Adams & Graham, unpubUshed data).

T6RBANE **

6 8 8 u(>iMn(Dn.>0Ma

Figure 3: Sketch reconstruction of Pacific Margin of Gondwana at mid-Triassic (225Ma) showing (1) approximate extent of eastem tectono-stratigraphic Permian-Triassic-Jurassic terraties ("Torlesse" and Waipapa) of New Zealand (horizontal ruling) and within this, Rakaia Terrane (crossed ruling); (2) main tectonic elements of Precambrian-Paleozoic Gondwana margin. Insets: histograms (A, B, C) of K-Ar mica ages of granitoid -metamorphic complexes for selected sectors of Gondwana that are potential source areas for Torlesse-type metasediments and actual 40Ar/39Ar detrital muscovite age data (D) for Permian-Triassic, Rakaia metagreywackes of New Zealand. (Diagram modifiedfromAdams & Kelly, 1996).

HIGH RESOLUTION AEROMAGNETIC AND GRAVITY SURVEY USED TO HIGHLIGHT BASIN GEOMETRY AND STRUCTURE. J. D. Alder' and D. A. Pratt^ 1 Department of Mineral Resources 2 Encom Technology Pty Ltd ABSTRACT New aeromagnetic and gravity surveys acquired as part of the Discovery 2000 initiative by the NSW Department of Mineral Resources are reviewed in the context of their importance to future exploration for petroleum in the Surat Basin and the southem edge of Eromanga Basin. The new aeromagnetic survey was flown with a line spacing of 400 metre line spacing and 80 metre terrain clearance. The gravity survey has been undertaken at a 4 km station spacing. Both surveys used differential GPS for precision elevation and position control which provided very accurate results. Comparison with the earlier State aeromagnetic compilation reveals that many new geological features may be interpreted from the high resolution data. Interpretation of these new data combined with recently acquired seismic data and reinterpretation of existing seismic data has indicated new structures within the basins and that the basin shapes themselves have been redefined. These grabens could provide a local petroleum source to charge overlying Jurassic and early Cretaceous sediments. In addition both large and small scale structural features can also be mapped. These data will be used to site the location of future seismic reflection surveys. The level of geological detailed data which can be interpreted from these new high resolution surveys compared to the surveys carried out approximately 10 years ago indicates their great value to exploration projects whether they are mineral, coal or petroleum. INTRODUCTION The 1995 segment of the Discovery 2000 initiative by the New South Wales Department of Mineral Resources (NSWDMR) included a major program of new high resolution airbome geophysical and ground gravity acquisition. This program included areas of importance to both mineral and petroleum exploration. These surveys provide the most significant new regional data sets since AGSO completed the regional coverage of NSW over ten years ago. During 1995 a total of 569,511 kilometres of new airbome magnetic and gamma ray spectrometer data was acquired with 205,000 kilometres specifically over sedimentary basins. Aeromagnetic Surveys Areas over sedimentary basins were flown at line spacings of400 metres compared with 1.5 to 3 kilometres for the earher regional coverage. Ground clearance in the new surveys was lowered to 80 metres compared with 150 to 400 metres for the previous surveys. Positioning accuracy was improved from estimates of approximately 150 metres for the early photo controlled surveys to better than 10 metres using real time GPS positioning. Amplitude accuracy was improved from values ranging from 1 to 5 nanotesla for the earlier surveys to better than 0.1 nanotesla for the high resolution surveys. Sample spacing was improved from the digitising of analogue chart records or from digital recordings made at 50 metre intervals for the old surveys to sampling 5 to 8 metres for the current surveys. All of these improvements combine to provide a sensitivity to geological contrasts which has increased by a factor of approximately 1000. Benefits from high resolution surveys include detection of magnetic anomalies from the sedimentary section and improved characterisation of intrusions and basement related magnetic anomalies. Mapping of features such as faults, joints, fold closures and unconformities is enhanced as a result of the increased resolution. GRAVITY New gravity data have been acquired in the Surat Basin. The area had previously been covered by a combination of regional government surveys using helicopter transport and detailed ground traverses collected as part of petroleum exploration programs. The regional data was mostly acquired at a grid spacing of 11 kilometres with height estimated using barometric procedures. Many of the ground surveys also used barometric methods for height estimation with elevation accuracies between 2 and 10 metres. Height estimation which is required for the processing of gravity data is usually the controlling factor on the accuracy and resolution of land gravity surveys. The new gravity surveys took advantage of the substantial improvement in height determination made available by GPS technology. Using differential GPS procedures heights can be measured to accuracies of better than 0.05 metres. For regional surveys at 4 kilometre sample spacing there can be considerable aliasing of gravity variations due to geological sources. The resolution of the new data resulting from closer measurement spacing and better height control is approximately 50 times higher than that of the existing grid.

Surat Basin Interpretation of aeromagnetic data in the Surat Basin is difiScult due to the complex nature of magnetic basement, nonmagnetic basement and igneous intrusions (Pratt & Rumph, 1995b). The most useful information from a high resolution magnetic survey was predicted to be obtained from the western margin of the basin, beyond the influence of the Carboniferous volcanics and Tertiary intrusives and extrusives. The new survey data within the basin clearly defines the character of the basement making determinations of depth to basement considerably more accurate than could be achieved previously. Without the supporting data such as seismic surveys it is impossible to be absolutely definitive but it is believed that features such as faults with both vertical and horizontal displacement can be mapped from the data. Small circular anomalies which are aligned with these fault and joint features are Tertiary intmsives. Minor anomalies almost certainly related to sedimentary features can also be seen over the Paka Tank Trough anomaly. The gravity data which at the time of writing still remains to be interpreted clearly shows significant variations in basement lithology and basement geometry. CONCLUSIONS New aeromagnetic and gravity data collected during 1995 by the NSWDMR over the Surat and Eromanga Basins provide an excellent opportunity for detailed evaluation of these basins. The level of geological detail which can be interpreted from these surveys indicates their value to exploration projects, and the specifications could be usefully applied to similar studies elsewhere. Explorers can expect to extract new infomiation such as basement depth refinement, possible underlying infrabasins and in conjunction with other data such as seismic turn what was essentially a 2D interpretation into a 3D model. References 1

2 3

PRATT D. A. & RUMPH B. 1995a. Darling, Surat and Eromanga Geophysical Reports In Morton D. J., Alder J. D., Grierson 1. J. & Thomas C. E. (Eds). Proceedings of the 1995 NSW Petroleum Symposium Proceedings, Petroleum Exploration Society of Australia, New South Wales Branch. PRATT D. A. & RUMPH B. 1995b. Surat Basin Geophysical Programme - Reprocessing and assessment. New South Wales for the Department of Mineral Resources by Encom Technology Pty Ltd (Unpubl.) PRATT D. A.. 1995. Discussion and Evaluation of New Geophysical Data In Morton D. J., Alder J. D., Grierson 1. J. & Thomas C. E. (Eds). Proceedings of the 1995 NSW Petroleum Symposium Proceedings, Petroleum Exploration Society of Australia, New South Wales Branch.

RECENT INVESTIGATIONS OF THE GREAT AUSTRALIAN BASIN IN NEW SOUTH WALES J.D. Aiders J. BambenyS DA. Pratt^ and C. Foss' 1 Department of Mineral Resources 2 Encom Technology Pty Ltd Summary The Mesozoic Eromanga and Surat Basins, subdivisions of the larger Great Austrahan Basin, have been relatively unexplored in New South Wales but are major petroleum producing areas in Qld and SA. The NSW Department of Mineral Resources, through the Discovery 2000 program, resorted to innovative techniques and conventional methods to assess the hydrocarbon prospectivity of these basins in New South Wales. An extensive water bore dataset in the basins has been used to constmct depth to basement maps and provide sub-surface regional control. Geophysical logging of deep water bores and geochemical analysis of artesian waters has enabled detailed stratigraphic correlation and estimation of the hydrocarbon potential of the penetrated sections. New seismic reflection data has provided supplementary regional control and new data on known anomalous features. Geophysical modelling indicated several basement deeps infilled with thickened Mesozoic and possibly Devonian and/or Permian sediments beneath parts of the Eromanga and Surat Basins. The potential that thickened sections within these deeps contain mature source rocks adequate for local hydrocarbon generation challenges the long held view that these areas have little intrinsic hydrocarbon potential. Introduction In August 1994, the New South Wales Government provided enhanced funding to a new program called Discovery 2000 to map, evaluate and promote the petroleum prospectivity of sedimentary basins in New South Wales. The Department of Mineral Resources adopted a staged approach for the program which consisted of: Stage 1

Stage 2 Stage 3 Stage 4 Stage 5 Stage 6

Evaluation of existing data a) aeromagnetic and gravity data b) water bore data Acquisition of high resolution aeromagnetic and gravity data a) Geophysical logging of water bores b) Collection of samples from water bores for gas analysis Seismic Surveys Evaluation of new data Stratigraphic drilling

The Department of Mineral Resources is into its third year of the Discovery 2000 program and have completed Stage 4. Althou^ all sedimentary basins in the State were to be reviewed as part of Discovery 2000, the Department decided that the main effort would be concentrated on the Darling Basin and the main components of the Great Australian Basin (Eromanga and Surat Basins). The Great Australian Basin in New South Wales In New South Wales, sediments of the Mesozoic Eromanga and Surat Basins, components of the larger Great Australian Basin, represent the shallower shelfal areas of the main basin, which is located in Queensland and South Austraha (Figure 1). From a petroleum perspective, the New South Wales components are considered distal from the recognised basin centres and main source kitchen areas located in the Queensland and South Austraha portions. The Eromanga Basin in New South Wales is a shallow shelf, which runs along the Queensland border from South Austraha to about longitude 147°E where easterly thickening of sediments across the Cunnamulla Shelf marks a transition to the Surat Basin.

16^

20'

24'

28®

CLARENCE MORETON BASIN

Figure 1. Sub-basins of the Great Australian Basin The Surat Basin in New South Wales is a large embayment on the south eastern comer of the Great Australian Basin and represents a southern extension of the oil and gas productive Surat Basin in Queensland (Figure 1). Previous perceptions of basin shape show a shallow, north-facing saucer overlying flat early Palaeozoic basement. The extreme eastern margin of the basin overlies the Bowen and Gunnedah Basins. The eastern margin, having previously been explored for coal, was considered of lower priority for the Discovery 2000 protocol. Evaluation of Existing Aeromagnetic and Gravity Data. The Department of Mineral Resources engaged Encom Technology Pty Ltd to integrate and interpret existing geophysical data sets covering the Eromanga and Surat Basins. The objective of the study was to investigate geological structure of the basins with particular reference to its hydrocarbon potential. Regional gravity and aeromagnetic data constituted the principal data sets with additional control from water bores, surface mapping, satellite imagery and sparse seismic data. Surat Basin The Surat Basin is located in a complex tectonic setting along the westem margin of the New England Fold Belt. The suture between the New England and the Lachlan Fold Belts lies beneath the sediments of the Surat and Gunnedah Basins, and as such cannot be specifically delineated. Repeated periods of tectonism along this zone are known from their impact on the sediment patterns in the overlying Surat Basin section and displacement of geological markers along the Hunter-Mooki Thrust System. Magnetic data over this area as imaged in Figure 2 are very difficult to interpret due to a complex igneous history (Pratt and Rumph (1)). Regional interpretation is based on the following observations and premises: • outcropping Garrawilla Volcanics cause only weak magnetic field variations against a backdrop of highly magnetic Carboniferous volcanics, beneath the Bowen and Gunnedah Basins, and Tertiary intrusives • magnetic data defines the broad faulted structure of the pre-Carboniferous basin, in which the highly magnetic Late Carboniferous to Early Permian volcanic succession was deposited • the Surat and Gunnedah Basin section does not cause a pronounced gravity low (Figure 3) Tertiary volcanics of the Liverpool and Warrumbungle Ranges, and associated intra-sedimentary intrusives, cause a higher frequency "stippled" character in the magnetic field image of the south eastem part of the area (Figure 2). Higher 10

amplitude anomalous areas occur parallel to the basin's eastem margin, predominantly east of longitude 149oE. These anomalies are interpreted to be due to varying thicknesses of Boggabri Volcanics beneath the Gunnedah and Surat Basins. The base of the volcanics define the Early Carboniferous structural setting and the top of the volcanics defined the shape of the Mesozoic basins. Basement in the western part of the Surat Basin area is largely a northerly extension of the Lachlan Fold Belt and is marked by regional gravity highs (Figure 3). Gravity anomalies in the central part of the basin are interpreted as being caused by intra-basement density contrasts between granitic and metasedimentary basement. Sedimentary troughs that have little density contrast with respect to basement are difficult to detect in this setting. Interpretation of seismic data near Collarenebri and Goodooga has revealed the existence of concealed grabens considered to be filled with Early Permian sediments, probably similar in age and composition to the Reids Dome Beds of the Denison Trough in Queensland (Bamberry and Kouzmina (2)). A regional interpretation of gravity and magnetic images using image enhancement techniques showed a series of NWSE oriented linear features that are interpreted as being possibly Early Permian and are related to offsets along the faulted margin. Movement along these structures may produce broad low amplitude folding, although their significance and influence on sedimentary patterns is still under investigation. Eromanga Basin The magnetic character of the Eromanga basement units differs substantiallyfi-omthat of the Surat Basin (Pratt et al. (3); Figure 4). Broad semi-circular shaped magnetic anomalies are believed to be meta-volcanic units of Lower Devonian age or older and possibly similar to rocks of the Cobar region. High amplitude magnetic anomalies in the north-eastern and central eastem portion of the basin in New South Wales are believed to be associated with Lower Palaeozoic volcanics, similar to those found at the base of the Bancannia Trough near Broken Hill. High amplitude trends on the westem margin are associated with Precambrian rocks of the Wonominta Block. The interpretation has largely focussed on depth to basement beneath an almost ubiquitous blanket cover of Eromanga sediments. The most reliable depth estimates have been derived from analysis of magneticfielddata, whereas gravity data mostly show large scale structures (Figure 5). The interpretation reveals the presence of several previously unrecognised basement deeps. These are probably of Devonian age and include the Urisino Trough, Enngonia Embayment and Brewarrina Trough (Figure 6). The deepest and most conspicuous trough, the Paka Tank Trough, is interpreted to be similar to the Bancannia Trough, which extends beneath Eromanga sediments near the South AustraUan border. Similarly the Morden Trou^ is an extension of the Darling Basin under the Eromanga Basin. The fault-related Caryapundy Trou^ in the north west of the area may contain Pennian or thickened Mesozoic sediments. The fault bounding the Caryapundy Trough is also parallel to the eastem edge of a contemporaneous swampy area known as the Bulloo River Overflow. The lineament and the swampy area suggest the area may still be undergoing active subsidence. Some of the gravity lows are interpreted to reflect the presence of older basins or troughs beneath the Eromanga Basin. Analysis of basement depthfi-omgravity data alone proved unreliable because there are also gravity

Figure 2. Total Magnetic Intensity grey scale image of the Surat Basin 11

SURAT BASIN Bouguer Gravity

Figure 3. Bouguer gravity grey scale image of the Surat Basin

EROMANGA BASIN

TMI Image

100 kilometres

Bancannia Trough

Caiyapundy Trough

1 2

EnngoniaEmt»ayn>ent

4 Louth Block 1 3

Urisifw Trough

Culgoa River Basement High

Myrong Gravity Low 1 4

Brewarnna Trough

Tibooburra Basement High 1 0 1 5

Morden Trough

YantabuDa - War«rT.r,fl Block

1 1

Paka Tar,k Trough

Compton Downs Basement High

Figure 4. Total Magnetic Intensity grey scale image of the Eromanga Basin

EROMANGA BASIN Bouguer Gravity contour interval

2 mgal

800000

Figure 5. Bouguer gravity grey scale image of the Eromanga Basin

CARYAPUNDY TROUGH

ENNGONIA EMBAYMENT

Figure 6. Interpreted thickness of sub-Eromanga sediments derived from magnetic and gravity modelling lows due to intrabasement granites. Discriminating intrabasement granites from basins containing Mesozoic or Palaeozoic sediments has not proved possible with the existing dataset. Accuracy is further reduced because density contrasts between basement and overlying sediments is poorly known. Integrated interpretation of gravity and magnetic data with computer simulation resolved many but not all of these uncertainties. Most of the pre-Eromanga sediments are believed to be of Devonian age, similar to those of the DarUng Basin to the south and the Adavale Basin to the north. The sharpness of the gravity anomalies however suggests that in some cases there is also a contribution from a younger section which could either be thickened Eromanga section or an intermediate Permian section. The association of thickened Mesozoic or Permian above preserved Devonian infra-basins is quite widely observed throughout Eastern Australia. Water bore Data A surprisingly large number of water bores in the Surat and Eromanga Basins were drilled to basement. These bores, combined with bores reaching only shallower acquifers, provide the opportunity for investigation of not only the 13

stratigraphy, but also the type, source and quantity of hydrocarbons in artesian waters. The bores draw water from the Mooga and Pilliga Sandstones in the Surat Basin and the Wyandra and Hooray Sandstones in the Eromanga Basin. Water samples were collected directly from theflumesofflowingbores orfrompressure plugs in the headworks of bores. Other samples, particularly from the Eromanga Basin, were obtained from pumped subartesian bores. Several samples of free flowing gas were also collected. CSIRO Petroleum used gas chromatography to analyse quantity and composition of dissolved hydrocarbons and incorporated mass spectrometry to measure 13C isotope composition of methane and carbon dioxide. Samples collected from the Surat Basin showed that methane is the main gaseous hydrocarbon and is accompanied by varying amounts of carbon dioxide (Table 1). Most samples record small quantities of ethane and a few bores contain traces of propane and butane. Nearly all the analyses record dry gas, however, gases from one bore (Hollywood) approach wet gas composition. Stable isotope analysis showed 13C composition of methane to be outside the compositional range for thermogenic gas but similar to that of shallow dry microbiologically altered gas or coal seam gas. Smith (4) described a process whereby after catagenesis, methane and ethane may be converted to carbon dioxide, which may be subsequently reduced back to methane in an environment, such as an acquifer, leaving dry gas only. The 13C composition of methane is consequently altered to isotopically light signatures (depleted in 13C), mimicing those of biogenic gas. This mechanism can be used to explain the origins of gases observed in bore waters as coal seams may exist in concealed Permian-Triassic troughs described earlier in this report. Many samples taken from the Eromanga Basin also show dry gas composition but several samples show wet gas compositions and, in one sample (Fort Grey), ethane plus butane exceeds methane. The isotopic composition of the Eromanga methane is more enriched in 13C and closer to that of thermogenically derived gases. This complements earlier work by McLaughlin (5) who inferred a thermally mature source for solvent extractable hydrocarbons in bore waters from the southem Eromanga Basin. Interestingly, 13C.CH4 composition from the Hollywood bore in the Surat is similar to that of Eromanga Basin gases. These results indicate possibilities of either long range hydrocarbon migration from recognised source areas, or the presence of previously unrecognised source rocks in concealed basins. Seismic Surveys At the time of writing, seismic acquisition had only just been completed and processing had not commenced. The Department acquired data from two survey lines in the Eromanga Basin and four in the Surat Basin. Field brute stacks from a survey, carried out across the Paka Tank Trough in the Eromanga Basin, confirm the presence of a previously unknown graben underlying Mesozoic sediments. The bulk of the graben shows lowfrequencyreflectors that are interpreted as Devonian sediments. The gross seismic character is similar to that of Devonian Darling Basin sediments. However nested within the graben fill, some high frequency reflectors appear to be discontinuous with the underlying "Devonian" sediments. Sediments higher in the section are interpreted as Permian or early Mesozoic in age. The seismic sections within the Eromanga also show onlapping sediments and drape anticlines of Mesozoic sediment over basement highs. The seismic also indicates the existence of Late Cretaceous-Tertiary structuring resulting in compressional folding and faulting. This age of structuring is the same age as trap formation in existing oilfields such as Tintaburra, which is located immediately to the north in Queensland. Two seismic lines of note were undertaken within the Surat Basin. The first was undertaken along the railway line between Bellata and Moree to investigate depth of sediments within the Bellata trough. The field brute stacks have indicated both Mesozoic and underlying Permian sediments much deeper and more extensive than previously suspected. Onlap of Permian on to basement highs provide opportunity for pinch-out traps, although the stratigraphy of this area is poorly known The second line was undertaken west of Pilliga over interpreted "magnetic and gravity lows". However the field bmte stacks have indicated fairly flat lying Mesozoic sediments. Currently the interpretation of the magnetics and gravity have yet to be resolved but point to a need for an integrated approach. Conclusions The New South Wales portions of the Surat and Eromanga Basins remain undere?q)lored. Large distances from recognised source kitchens and the perceived lack of mature petroleum source rocks has deterred exploration efforts. New reprocessing of geophysical data and newly acquired seismic data have shown the existence of previously unrecognised basement deeps, mostly beneath the Eromanga Basin. The pre-Mesozoic fill of these structures is considered to be mainly Devonian aged, although possibilities for Permian-fill have been identified. The basement deeps may also contain thickened Mesozoic strata, however no stratigraphic data exists for the area. Excellent porosities and permeabilities in the Mooga, Pilliga and Hooray Sandstones has indicated the widespread existence of reservoir rocks in the Mesozoic section. Geochemical analysis of the artesian waters attests to the presence and movement of hydrocarbons in regionally extensive acquifers in both the Surat and Eromanga Basins. Recognition of the troughs underlying the Surat and Eromanga Basins provides new possibilities for the presence of thickened sections containing mature source rocks adequate for local hydrocarbon generation. Identification and 14

delineation of basin deeps beneath the Surat and Eromanga Basins is essential to formulating a regional framework and to understanding basin geometry and structural evolution in New South Wales. The Department recognises that much more work is required to understand the distribution of sediments and hydrocarbons within the Great Australian Basin within New South Wales. This work has answered many questions about these poorly understood areas but has also raised many more. Acknowledgments We thank Dr Ray Shaw for his review of, and suggested changes to this paper. References (1) PRATT D. A. & RUMPH B. 1995. Darling, Surat and Eromanga Geophysical Reports In Morton D. J., Alder J. D., Grierson I. J. & Thomas C. E. (Eds). Proceedings of the 1995 NSW Petroleum Symposium Proceedings, Petroleum Exploration Society of Australia, New South Wales Branch. (2) BAMBERRY W. J. & KOUZMINA G. 1995. Great Australian Basin Project. In Morton D. J., Alder J. D., Grierson I. J. & Thomas C. E. (Eds). Proceedings of the 1995 NSW Petroleum Symposium Proceedings, Petroleum E^qploration Society of Australia, New South Wales Branch. (3) PRATT D. A., RUMPH B. & FOSS C. 1996. Eromanga Basin Geophysical Program - Reprocessing and Assessment, New South Wales, Australia for the Department of Mineral Resources. Encom Technology Pty Ltd. (unpubl. report.). (4) SMITH J. W. 1995. Origin of gases in the Sydney and Bowen Basins. In: Russell, N. J. (Ed) Stable isotope gas analyses for the coal and petroleum industries. A one day workshop. CSIRO Petroleum Program and Abstracts, June 2,1995. 21-2 (5) MCLAUGHLIN R. L 1984. Progress report on PEL 265 for Kells Investments. N S W Geological Survey, Petroleum Geology Report 84/06. methane ethane propane i--butane n-butane CO2 bore name 02+Ar N2 ppm mol% 14 33978 0.50 0.40 0.10 5000 MidkinS 2.50 93.6 71 0.05 0.05 3700 218200 66 0.70 6.80 Carenugal 3.00 1.00 3000 1.50 15.6 825500 343 21.00 Bryanungra 27 3300 93600 WilbyWilby 1.90 88.1 0.34 0.15 0.09 1400 26200 8.3 6.00 91.2 Dolgelly 0.37 0.26 25000 3300 78 11.90 HollywcKMi* 20.50 76.7 43400 58 0.13 Wiirah 5.80 89.4 4300 54000 514000 158 3.30 Salisbury 1* 14.1 29.1 62 trace trace 28000 30.3 589000 10.00 8 Deep 451000 321 3.80 80000 3.8 43.1 Tinaroo * diluted by atmosphere during sampling Table 1 Selected analyses offree-flowinggasfromartesian water bores in the Surat and Eromanga Basin.

THE PETROGENESIS OF THE MESOZOIC ROCKS, BOWEN REGION, QUEENSLAND, AND CHANGES OF TECTONIC SETTING FROM 145 TO 100 MA Charlotte M, Allen and B.W. ChappeU Key Centre for the Geochemistry and Metallogeny of the Continents. Department of Geology Australian National Univerisity, Canberra 0200 INTRODUCTION Cretaceous igneous rocks occur sparsely along the entire length of Australia's eastern seaboard, from as far south as Mount Dromedary to as far north as Cape Upstart. These Cretaceous rocks have been modelled as the product of rifting associated with formation of the Coral and Tasman seas although the oldest documented sea floor off Queensland is -85 Ma (Veevers et aL 1991; Ewan et al, 1992). In the Bowen Region we have identified three main age groups of intrusions: 145-135, 135-125, and 125-100 Ma. We describe here the time-temperature-emplacement level characteristics of the age groups and how these characteristics might relate to tectonic setting. Furthermore, we emphasise the isotopic differences of the Cretaceous (and Triassic) plutons from the Carboniferous-Permian ones that many of them intrude. AGE DISTRIBUTION AND DEFINITION OF SUITES Figure 1 shows geochemistry sample locations divided by age. On this map the structure called the Connors Arch is roughly coincident with the Urannah Suite and the plutons that intrude it. The adjacent Permian and older rocks east of the Arch comprise the Midgeton Block. The oldest group of Cretaceous rocks is restricted to the Don River area and these range in age from 145-135 Ma as determined by K-Ar ages from homblende and biotite (Table 1). Although we call these rocks the Don River suite, the suite includes very diverse rock types including quartz monzodiorite, monzonite, and granodiorite. Besides for age, these rocks share the features of being more mafic than granite, and having calculated Al-in-homblende pressures of 250-215 MPa using the Al-in-homblende method of Anderson and Smith (1995) (Table 1). We have applied the technique strictly to those rocks containing the assemblage Pl+Ksp+Qtz+Hb+Bt+Tnt+Mt (Hollister et aL 1987). The most extensive age group of Cretaceous granites in the Bowen area is 135-125 Ma as determined by K-Ar mineral ages and SHRIMP zircon geochronology (Webb and McDougall, 1968; Allen et aL 1994). We divide this age group into two suites based on location. Those intruding the Permo-Triassic sedimentary' rocks of the Bowen Basin are called the Bowen Suite, and those in the Connors Arch are called the Hecate-Eungella Suite due to their distribution between the Hecate Homestead and Eungella Dam. The Bowen Suite is diverse compositionally (Pattison, 1984). The samples for this study comprise three quartz diorite samples from the eastem part of the basin. These rocks have textures suggesting very shallow levels of intrusion. The Hecate-Eungella Suite includes granodiorite and granite that have many textural and compositional similarities. The granodiorite samples yield Al-in-homblende pressures of 156-76 MPa, with the rocks near Eungella giving the lowest pressures. Several areas of Cretaceous rocks younger than 125 Ma have been identified in the Connors Arch. These include: Mount Pring, Mount Abbot, and Cape Upstart. These rock groups are dominated by granite (s.s.) but also include the quartz syenite at Mount Abbot, and gabbro at Cape Upstart. All granites are very fine grained, vesicular, pink granites that most likely crystallised at very shallow depths. The Cretaceous igneous rocks of the AirUe Beach-Proserpine-Whitsunday area are <125 Ma with important exceptions. There are samples as old as 135 Ma in the eastem Whitsunday Islands. Ewart et al (1992) suggested that the age distribution of granites from the mainlanc to the outer Whitsunday Islands, indicated rifting as igneous rocks <125 Ma occur between areas where 125-135 Ma intrusions are found. Thus the Cretaceous age groups overlap in geographical distribution. GEOCHEMISTRY AND MAGMA TEMPERATURE The age groups defined above have distinct geochemical characteristics, especially if rocks with Si02>65% are examined. We use analyses of the Proserpine Volcanics (Ewart and Parianos, unpubhshed; Parianos, 1993; and two samples from this study) as representative of the overall AirUe-Proserpine-Whitsunday area. The trace element Harker diagrams of Figure 2 show two ways in which the rocks greater or less than 125 Ma differ. First, the older rocks average granodiorite in composition, and include few granites, whereas most of the younger plutonic rocks sampled in this study are granites. Second, the younger rocks contain greater abundances of Zr and Y, and lesser abundances of Sr than the older rocks as emphasised by the diagonal in Figure 2. As fractionation among granitic rocks tends to decrease, not increase, Zr, we attribute the high to very high Zr contents of the <125 Ma granites to higher temperatures at the source after 125 Ma (Table 2). Although the age groups have distinctive trace element characteristics, there is little isotopic contrast across all Cretaceous samples (Figure 3). The entire range of initial Sr ratio is 0.7033-0.7039. This range of values is relatively low for 16

continental marginal rocks, and much lower than those of the Urannah Suite, especially if those ratios are calculated to -120 Ma. Note that Triassic samples have isotopic compositions like the Cretaceous ones. We attribute the relatively high initial Sr ratios of the Bowen Suite to contamination of the intermediate to mafic, shallow intmsions by shallow crust. COMPARISON TO HOST ROCKS Late Carboniferous-Early Permian intrusions probably underUe most of the study area. This conclusion is based on the following: 1) The Urannah Suite, those rocks that make up most of the Connors Arch, are compositionally very similar to the Bulgunonna Suite that occurs west of the northem tip of the Bowen Basin (Black, 1994; Oversby et aL, 1994). Urannah Suite rocks are found in the Midgeton Block near Mackay as well as in the Connors Arch. The Connors Arch is thought to be a basement block uphfled during the westward thrusting the Hunter-Bowen Orogeny at --230 Ma such that the Arch now interrupts the original extent of the Bowen Basin (Fergusson, 1991; Fielding et aL, 1994 and 1996). The Urannah Suite is more radiogenic than the Cretaceous ones with respect to initial Sr, Nd and Pb isotopic composition. We argue elsewhere that the structure of the crust was fundamentally changed during the long period of extension that produced the Late Carbonifeorus-Early Permian Urannah Suite, the dike swarm that cuts the Urannah Suite, the Permian Thunderbolt Granite that intrudes the Urannah Suite, and the Permian volcanic rocks that floor the Bowen Basin (Fig. 1; Allen and Chappell, JGR, provisionally accepted). That period of extension ended at 230 Ma with the Hunter-Bowen Orogeny but not before severe cmstal extension had effectively removed Urannah Suite like rocks or their sources as sources for Cretaceous intrusions. DISCUSSION OF TECTONIC SETTING Cretaceous intrusions ranging in agefrom145 to 103 Ma are found in the Bowen area. Although rocks greater or less than 125 Ma have distinctive trace element signatures, they are indistinguishable with respect to radiogenic isotopes. Furthermore, the age groups overlap in geographic distribution. Ewart et al (1992) have made a strong case for generation of igneous rocks <125 Ma in an extensional environment based on the age distribution of intrusions, and the bimodal distribution of rock types. They argue for emplacement of E-MORB into the crust, crustal melting, and variable mixing between the two. We concur that an extensional setting is likely. What setting generated the older Cretaceous rocks is unclear. Could thefirstevidence of crustal rifling that culminated in the formation of new seafloor at --85 Ma have started 60 Ma before? Falvey and Mutter (1981) have suggested that magmatic activity can predate seafloor formation by as much as 50 Ma. No matter the tectonic setting, it is likely that the sources for all felsic Cretaceous igneous rocks were very similar given the small ranges of isotopic composition. Since t these rocks differ chiefly in trace element abundances, we argue that sources were similar but that the temperature at the source was significantly higher for igneous rocks <125 Ma. It is possible that as rifling progressed during the whole of the Early Cretaceous, progressive heating and isotherms rising in the crust gave rise to granites emplaced at shallow levels in the crust afler 125 Ma. Moreover progressive upUfl of the Connors Arch over the same period can be modelled as uplift on the margin of a rifl zone. ACKNOWLEDGEMENTS Special thanks to Ian McDougall and Joe Wooden for access to their labs. This is publication number ** in the Key Centre for the Geochemical Evolution and Metallogeny of Continents. REFERENCES ALLEN C.M., WOODEN, J.L., CHAPPELL B.W. & WILLIAMS I.S. 1994. Ages, compositions, and sources of the Urannah Batholith. In R.A. Henderson and B.K. Davis (Eds.) Extended Conference Abstracts of New Developments in Geology and Metallogeny: Northem Tasman Orogenic Zone, James Cook University, Townsville, pp. 107-110. ALLEN C.M. & CHAPPELL B.W. provisionally accepted . The Permian-Triassic dike swarm of the Bowen Region, Queensland: the transition from arc front to backarc magmatism. Journal of Geophysical Research, ANDERSON J.L. & SMITH D.R. 1995. The effects of temperature and oxygen fugacity on the Al-in-homblende barometer. American Mineralogist 80, 549-559. BLACK L.P. 1994. U-Pb zircon ion-microprobe ages from the northem Drummond Basin, northeastem Queensland. Austrahan Geological Survey Organisation Record 1994/34. BLUNDY J.D. & HOLLAND T.J.B. 1990. Calci amphibole equilibria and a new amphibole- p 1 a g i o c 1 s e thermometer. Contributions to Mineralogy and Petrology, 104,208-224. EWART A., SCHON R.W. & CHAPPELL B.W. 1992. The Cretaceous volcanic-plutonic province of central Queensland (Australia) coast—a rifl related 'calc-alkaline' province. Transactions of the Royal Society of Edinburgh: Earth Sciences 83, 327-345. FALVEY D.A. & MUTTER J.C. 1981. Regional plate tectonics and the evolution of Australia's passive continental margins. Bureau ofMineral Resources Journal ofAustralian Geology and Geophysics 6,1-29. FERGUSSON C. L. 1991. Thin-skinned thrusting in the northem New England Orogen, central Queensland, Australia. Tectonics 10, 797-806. FIELDING C.R., STEPHENS C.J. & HOLCOMBE R.J. 1996. Permian stratigraphy and Palaeogeography of the northem New England Fold Belt in coastal central Queensland. Australian Joumal of Earth Sciences, in press. 17

FIELDING C.R., HOLCOMBE RJ. & STEPHENS CJ. 1994. A critical evaluation of the Grantleigh Trough, eastcentral Queensland. In: R.J. Holcombe, C.J. Stephens, and C.R. Fielding (Editors), 1994 Field Conference, Capricorn Region, Central Coastal Queensland, Geological Society Australia Inc., Queensland Division, pp. 1730. HOLLISTER L.S., GRISSOM G.C., PETERS E.K., STOWELL H.H. & SISSON, V.B. 1987. Confirmation of the empirical correlation of A1 in homblende with pressure of solidification of calc-alkaline plutons. American Mineralogist 72, 231-239. LEVINGSTON K.R., 1981. Geological evolution and economic geology of the Burdekin River region, Queensland Bureau of Mineral Resources 208. McLENNON T.R 1989. Geological map of Hillalong 1:100,000, preliminary edition. Queensland Department of Mines. OVERSBY B.S., MACKENZIE B.E., MCPHEE J., LAW S.R. & WYBORN D. 1994. The geology of Palaeozoic volcanic and associated rocks in the Burdekin Falls Dam-Conway area, northeast Queensland. Australian Geological Survey Organisation Report 94/21. PARIANOS J.M. 1993. Carboniferous to Tertiary geology of the Airlie Block, northeast Queensland. MS thesis. University of Queensland (unpublished). PATTISON C.L 1984. Igneous intrusions in the Bowen Basin. MS thesis, Queensland University of Technology (unpublished). VEEVERS J.J., POWELL, C.McA. & ROOTS S.R. 1991. Review of seafloor spreading around Australia I: Synthesis of the pattems of spreading. Australian Journal of Earth Sciences 38,373-389. WATSON E. B. & HARRISON T.M. 1983. Zircon saturation revisited: temperature and composition effects in a variety of crustal magma types. Earth Planetary Science Letters 64,295-304. WEBB A.W. & McDOUGALL L 1968. The geochronology of the igneous rocks of eastem Queensland. Journal of the Geological Society of Australia 15, 313-346. Table 1. Ages and aluminum in homblende barometry* Sample

Suite

Age Ma"

P(MPa)

HHO HI19 BW54 BW15 BW52 BW64 BW38

Hecate-Eungella Hecate-Eungella Hecate-Eungella Hecate-Eungella Don River Don River Don River

126(1) 128(1) 129(2) 130(2) 135(3) 142(3) 145(3)

76 99 156 152 216 224 250

Tol-hb^C' 635 630 657 652 705 716 712

Tzire^C'

AT'C

740 746 771 768 781 783 729

105 116 114 116 76 67 17

Aluminum in homblende barometry from Anderson and Smith (1995). Microprobe analyses of 3 to 5 grains each of homblende adjacent quartz and plagioclase adjacent homblende. K-Ar homblende ages from (1) Webb and McDougall (1964), (2) Webb and McDougall (1968), or (3) this study. Plagioclase-homblende thermometry from Blundy and Holland (1990). Zircon saturation temperature from Watson and Harrison (1983). Difference between zircon and pl-hb temperature. Table 2 Zircon thermometry for rocks of granitic composition* Sample

Suite

Age Ma"

CU-232 CU-230 CA-18 BW-14A BW-14B BW-13 BG-118 BG-117 BG-113 BW-42 BW-43 BW-55

Cape Upstart Cape Upstart Proserpine Volcanics, intrusive Mount Pring Mount Pring Mount Pring Mount Abbot Mount Abbot Mount Abbot Hecate-Eungella aplite Hecate-Eungella aplite Hecate-Eungella aplite

103(2) 777 118 (1) 120 (2) 753 788 120 (2) 802 804 130(1) 711 706

W c 800 889 786

Zircon saturation temperature from Watson and Harrison (1983). Ages from (1) Webb and McDougall (1968), or (2) this study.

810

714

V//////A Cretaceous igneous rocks M i ^ ^ T r i a s s i c . Gloucester Granite I 1111111 I Triassic? Airlie Volcanics "2 Permian-Cretaceous dikes k W W W X i Perm.-Tri. Bowen Basin sedimentary rocks Permian volcanic rocks Permian Thundertwit Suite L Carboniferous-E Permian Urannah Suite Devonian-middle Cartx)niferous rocks Cretaceous samples: - 135-145Ma. •125-135 Ma, -100-125 Ma

Figure 1 Bedrock map of study area near the town of Bowen (B). Geochemistry samples are divided by age. Places indicated by squares. C=Collinsville, E=Eungella, H=Hecate homestead, U=Urannah homestead. Triassic rocks are restricted to the coastal area and Whitsunday Islands. Cretaceous volcanic and plutonic rocks are distributedfromthe Bowen Basin to the Whitsunday Islands. The dikes are mostly Permian in age except the ring dikes associated with the Cretaceous Mount Abbot complex. Unpattemed area at "C.H." are Tertiary deposits at Cape Hillsborough.

Figure 2 These trace element Harker Diagrams show the relatively enriched Y and Zr and relatively depleted Sr compositions of the rocks with ages < 125 Ma as compared to the older groups.. Diagonals divide the younger and older samples at Si02>65%. Granite from Ball Bay (18) has Y=112 ppm. Asterisks=Don River Suite (145-135 Ma), dark diamonds=Hecate-Eungella Suite (135-125 Ma), l i ^ t diamonds=Bowen Suite (135-125 Ma), dark stars=Mount Abbot (-120 Ma), light stars=Mount Pring (--120 Ma), dots=Cape Upstart (-100 Ma), cirlces=Proserpine Volcanics (<120 Ma). 20

0.7040

* *

sn Bowen-^% f ) 0 •

•

Cretaceous data 45

i 0 Si02

I Cretaceous I Triassic i Carboniferous

o O O o o o o ^O ^o ^ ^ ^ oN5 O^ oOl oO) oO) Ol Ol O ai o cn o cn Figure 3 Stacked-bar diagram showing numbers of samples of a given initial (Sr^). Mesozoic samples are less radiogenic than Late Carboniferous-Early Permian ones from the Urannah Batholith. Symbols as in Fig. 2. Data sources: this work, Ewart et al. (1992), and Parianos (1993). Inset shows non-systematic relationship of Si02 and Sr^ among Cretaceous rocks. The Bowen Basin samples are more radiogenic than the rest.

APPLICATION OF SEQUENCE STRATIGRAPHY TO CONTINENTAL SUCCESSIONS: IMPLICATIONS FOR MESOZOIC CRATONIC INTERIOR BASINS OF EASTERN AUSTRALIA

Geoi^e Allen, Simon Lang, Oki Musakti and Alfredo Chirinos School of Geology, Queensland University of Technology, Brisbane Summary Sequence stratigraphy has been very successful in advancing the understanding of passive margins and foreland basins. Models developed for these basins are focussed on marme or coastal facies, and therefore they are not directly applicable to cratonic interior basins with thick 'Continental successions that were possibly never connected to the sea. This paper e?q)lores the use of sequence stratigraphic concepts in continental successions, and gives some examples of how they can be applied to Mesozoic cratonic interior basins of Eastem Australia. Introduction Sequence stratigraphic concepts provide a powerful tool to analyse and interpret stratigraphic patterns in terms of changes in sediment supply and relative sea level. During recent years numerous models have emerged that explain the stratal geometries on passive margins and foreland basinswhere short term sea level changes and cyclic stacking patterns are mainly influenced by eustatic fluctuations superimposed over longer-term tectonic subsidence (eg. see numerous papers in Macdonald (1), Macqueen & Leckie (2), Posamentier et al(3). Van Wagoner et al (4), Wilgus et al(5), Weimer & Posamentier, (6). Whilst some workers have been tempted to apply global eustatic sea level curves as a template to rationalise basin stratigraphy, this is an abuse of sequence stratigraphy (see discussion in Weimer & Posamentier, (6)) and is likely to lead to erroneous results in basins where tectonism is a greater influence than eustacy (ie. active margins and interior basins). In cratonic interior basins with no marine connections, the direct influence of eustacy is nil, and this has led to a belief that sequence stratigraphy is not applicable to these settings. The purpose of this paper is to illustrate that the fundamental principles of sequence stratigraphy, based on the concepts of sediment accommodation and sediment supply, are applicable to the continental successions in the vast cratonic interior basins of Eastem Australia (eg., the Jurassic-Cretaceous Surat and Eromanga basins). Principles of Sequence Stratigraphy The basic mechanisms which control sediment patterns in both marine and non marine environments are changes in the space available for sedimentation (accommodation), the rate of sediment influx, and the type of depositional environment; the latter being to a large extent affected by the first two. In fluvial systems, the basic control wMch determines changes in accommodation are modifications of the theoretical equilibrium profile ("graded profile" of geomorphologists), in relation to the rate and type of sediment influx. In an interior basin, the primary factors which will determine changes in the graded profile are tectonic tilting, modifications in fluvial discharge and changes in sediment supply. Although climate and sediment source are important, our studies indicate that tectonic subsidence and tilting of thefluvialprofile have the greatest control on sediment packaging. It is important to emphasise that it is the ratio between the rate of sediment supply (SS) and the rate of subsidence-controlled accommodation (SA) that will determine changes in the style of non-marine sediment stacking pattems. Non-marine sediment stacking pattems Depending on the value of this ratio (SS/SA), two end members of a spectrum of stacking pattern styles emerge (Figure la-d). At one extreme, if the ratio is high (ie. little or nil subsidence, and therefore no significant increase in fluvial accommodation), a lateral amalgamation of thefluvialchannels will result forming a one channel thick, widespread sand sheet (Figure la). At the otiier extreme, if the ratio is very low (ie. very low sediment influx and relatively rapid subsidence), a lacustrine system will result (Figure Id). This latter case is stratigraphically equivalent to trans^ession in a coastal setting in so far as the available accommodation increases faster than the sediment supply can fill it. A rapid decrease in the ratio will result in the development of a distinctive lacustrine "flooding surface" overlyingfluvialdeposits, and when tiie ratio reaches its lowest value, a lacustrine "maximum flooding surface" will result. Maximum flooding surfaces generally form relatively widespread correlation markers. From a practical standpoint, analysing alluvial basins involves recognising stacking pattems based on identifying finingupward versus coarsening-upward trends, and locating key surfaces, because these stacking pattems play a vital role in reservoir connectivity (Figures 1 & 2). Key surfaces Two types of key surfaces need to be identified in sequence analysis of alluvial basins; 1) surfaces related to changes in the rate of sediment fill of the available accommodation; and 2) surfaces related to the creation of negative accommodation. 22

1) Changes in the rate in which accommodation is filled results in changes in the sediment stacking patterns. These changes can be punctuated by key surfaces. The most important of these is the "maximumfloodingsurface" (MPS). This surface represents a change from non-filling tofillingof available accommodation. In the ideal case this surface will be identified by the changefi-omdeepening upward to shallowing upward lacustrine facies. In lacustrine successions the MPS is commonly represented by a shaly interval marked by the hi^est gamma ray peak, and/or a distinctive peak on resistivity/conductivity logs. Depending on local conditions it may be represented by black shale (including oil shale), chemically precipitated limestone or oolitic ironstone. On the coeval alluvial plain, the MPS should lie within the finestgrained interval, and occur below a transitionfrommuddier to sandierfloodplainfacies. For basins Devonian or younger, the MPS may be represented within a significant coaly interval if the climatic conditions were suitable. Although the MPS is a well developed correlation marker in well logs and core, it is however dependant on sediment supply dependant, and therefore it is not strictly a true chronostratigrahic surface, but may approximate one. 2) The creation of negative accommodation occurs when tectonic tilting oversteepens a segment of the fluvial profile. This causes the river to incise downwards. This process results in rejuvenation of the fluvial network, incised valleys and the formation of a stratigraphic unconformity (ie an alluvial "sequence boundary"; SB), The expression of this unconformity in the incised valleys will be coarse-grained channel fills deeply eroding into underlying lacustrine or floodplain dominant shale successions. On the interfluves, the expression of the unconformity may be preserved as an extensive palaeosol. The relative maturity of the palaeosols will give an indication of the length of exposure. The unconformity is a key surface marked by local erosion, and will have a correlative conformity that may lie downstream within a lacustrine prone succession, or an aggrading fluvial succession downstrean from the tectonic hinge line. The succession that lies between two regionally significant unconformities should be recognised as a "sequence". The key surfaces in an alluvial basin are in many ways similar to the key surfaces recognised in a shelfal setting. The sequence boundary is independent of sediment supply, and a function only of accommodation (ie. changes in the equilibrium profile). Maximum flooding surfaces however are dependent on sediment supply. Importance of tectonism on sediment stacking patterns In passive margins, and some foreland basins, sequence boundaries are related to highfrequencyeustatic cycles which create unconformities and accommodation that is filled by onlapping sediments. In cratonic interior alluvial basins develop accommodation trends related to tectonic downwarping and tilting, and the accommodation isfilledby onlapping sediments onto a progressively inclined depositional surface, typically by filling the incised valleys first. In sea-level caused unconformities, the fluvial erosion and incision of the nick points propagates landward from the coast and the effects probably diminish upstream. In contrast, tectonic tilting causes an instantaneous effect over the entire profile. One result of this is to bring about very rapid local changes in sediment pattems throughout the basin. It is therefore through careful correlation of the key surfaces, that the gross sequence stacking geometric pattems emerge that enable the effects of tectonism to be identified. Implications for Mesozoic cratonic interior basins of Eastern Australia The Eromanga and Surat Basins form the greater part of a vast Mesozoic interior basin system covering a large part of Central and Eastern Austraha. Along with the inter-connected Clarence-Moreton Basin and Nambour Basin to the east, and the Carpentaria Basin to the north, all these basins share a similar depositional history. The basins formed within a huge intra-cratonic downwarp that developed to the west of an active volcanic arc system that lay offshore of present coastline, along the Eastern Australia plate boundary through the Mesozoic (Veevers, 7). In Jurassic to Early Cretaceous, these basins accumulated entirely continental successions, with no evidence of a sea connection. Sediment supply altemated between mature quartzose sediments derived from the craton to the south and west, and volcanicderived sedimentfromthe east dependant on arc activity (Veevers, 7). Subtle uplift along basement highs between these basins during the Mesozoic may also have played a role in altering palaeoflow directions and sediment supply. Marine conditions eventually inundated the basins, but not until near the end of Early Cretaceous (late Neocomian), coinciding with a global marine transgression. It has been recognised by several workers that these basins contain a distinct cyclicity, and this has been related to tectonoeustatic cyclicity (Veevers, 7) and global eustatic cyclicity (Exon & Burger (8), Burger (9), and Totterdell et al (10). Others however have suggested that changes in subsidence and sediment supply pattems may have played are more important role in controlling the apparent cyclicity (Green et al, (11), Fielding et aL (12), Green & McKellar, this volume). In the Eromanga Basin, several large-scale cycles offluvialto fluvio-lacustrine successions can be identified covering a vast area (Green et al, (11)). The fluvial units are dominated by widespread and amalgamated sandstones that are important water and petroleum reservoirs (eg. Hutton, Adori, Hooray Formations and Wyandra Member). The fluviolacustrine units typically have finer-grained sandstones, and dominated by siltstones and mudstones (e.g. Birkhead, Westboume and lower part of Cadna Owie Formations). The boundary between lacustrine and the overlying fluvial facies is mostly sharp and correlable over the whole basin, indicating a regional significance (Figure 3). Some sequence boundaries appear to coincide with the lithostratigraphic unit boundaries (eg. base of the Hutton, Adori, and Hooray Sandstones), but it is likely that other sequence boundaries 23

are more cryptic (eg. within the Hooray to Cadna-Owie transition). Detailed well-log correlations approached from a sequence stratigraphic rather than a lithostratigraphic perspective has resulted in the recognition of important sequence boundaries within formations. These sequence boundaries separate isopach successions from distinctly wedge-shaped successions, clearly indicating a role for tectonic subsidence controlling sediment stacking pattems. Careful subsurface correlation has shown that tilting effects are important not only for influencing changes in accommodation, but also for releasing sediment from reworking the uptilted portions of the underlying successions. Despite these effects, maximum flooding surfaces within the thick lacustrine succession are able to be correlated regionally, and imply creation of accommodation at a greater rate than sediment supply almost uniformly across the basin. Whilst it is tempting to match these "maximum transgressive" events with global eustacy, a tectonic control influencing regional subsidence is an equally feasible explanation. In the Surat Basin, like the Eromanga Basin, a similar pattem of large scale cyclicity of fluvial to fluvio-lacustrine successions has be identified (Totterdell et ah (11), Fielding et al (12)). The Surat Basin appears to have a more complete stratigraphic record than the Eromanga Basin (Breen et al this volume). Sequence boundaries at the base of the Precipice, Hutton, and Springbok (df. Adori) Sandstones occur earlier than counterparts in the Eromanga Basin, suggesting topographic, and therefore probably tectonic controls on the nature of the unconformity surfaces. As in the Eromanga Basin, detailed correlations appear to indicate that cryptic sequence boundaries may occur within the lithostratigraphic units, rather than coinciding exactly with them. Of particular interest are the fluvial to lacustrine transitions that occur within several intervals, but especially the Precipice Sandstone-Evergreen Formation interval. The lacustrine facies in the Evergreen Formation (including the Boxvale Sandstone Member and Westgrove Ironstone Member) contain excellent examples of coarsening-upward mouth bars, flood and/or storm-generated density flow and hummocky deposits, rythmically-bedded lacustrine fines, and oolitic ironstones indicative of sediment starvation. Detailed correlation within the upper Precipice-Evergreen interval shows several key surfaces that can be correlated around the basin, and our work indicates that tectonic rather than eustatic controls may be more likely a influence on sediment stacking pattems. Conclusions Sequence stratigraphy is applicable to Eastern Australia's vast Mesozoic cratonic interior provided the fundamental stratigraphic principles of subsidence versus sediment supply are understood. Detailed stratigraphic correlation using key surfaces (fluvial sequence boundaries and lacustrine maximum flooding surfaces) in continental successions appears to reveal an important tectonic control on sediment stacking pattems, possibly more important than the indirect influences of eustatic sea-level fluctuations far from the basin. Acknowledgments The logistic and financial support to undertake studies in Queensland provided by the Geological Survey, Department of Mines and Energy, Queensland, and Santos Queensland & NT is gratefiilly acknowledged. References 1. 2. 3.

4. 5. 6. 7. 8. 9.

10.

11.

Macdonald, D.I.M., (Editor): Sedimentation, Tectonics and Eustacy, Special Publication No. 12, Intemational Association of Sedimentologists. Blackwell Scientific Publications, Oxford, 1991, 518p. Macqueen, R.W. & Leckie, D. A. (Editors): Foreland Basins and Fold Belts. AAPG Memoir 55,1992,460p. Posamentier, H.W., Summerhayes, C.P., Haq, B.U., & Allen, G.P (Editors): Sequence Stratigraphy and Facies Associations. Special Publication No. 18, Intemational Association of Sedimentologists. Blackwell Scientific Publications, Oxford, 1993, 644p. Van Wagoner , J.C.., Mitchum, R.M., Campion, K.M., and Rahmanian, V.D.: Siliciclastic Sequence Stratigraphy in Well Logs, Cores, and Outcrops. AAPG Methods in Exploration Series, No. 7., 1990,55p. Wilgus, C.K., Hastings, B.S., Ross, C.A., Posamentier, H.W., Van Wagoner, J.C., and Kendall, C.G.St.C. (Editors): Sea-level changes: an integrated approach: SEPM Special Publication No. 42,1988. Weimer, P. & Posamentier, H,W. (Editors), Siliciclastic Sequence Stratigraphy. Recent Developments and Applications. AAPG Memoir 58, 1993,492p. Veevers, J.J. (Editor): Phanerzoic Earth history of Australia. Oxford University Press, New York, 418p., 1984. Exon, N.F., & Burger, D.: Sedimentary cycles in the Surat Basin, and global changes in sea level. BMR Journal of Australian Geology and Geophysics, 6, 1981, 153-159. Burger, D.: Palynology, cyclic sedimentation, and palaeoenvironments in the Late Mezozoic of the Eromanga Basin. In Gravestock, D.I., Mooore., RS., & Pitt, G.M., (Editors): Contributions to the geology and hydrocarbon potential of the Eromanga Basin. Geological Society ofAustralia, Special Publication, 12, 1986, 53-70. Totterdell, J.M., Brakel, A.T., Wells, A.T & Hofi&nann, K.L.: Basin phases and sequence stratigraphy of the Bowen Basin. In FoUington, I.L., Beeston, J.W., & Hamilton, L.H. (Editors): Proceedings of the Bowen Basin Symposium 1995. 1-3 October, Mackay, Qld. Geological Society of Austraha Coal Geology Group, Brisbane, 1995, 247-256. Green, P.M., Brain, T.J., John, B.H., and Almond, C.S.: Subsurface correlation of the Jurassic rocks of the Southeastern Eromanga Basin, Queensland. Geological Survey of Queensland Record 1981/1, 1989, 26p. 24

12.

Fielding, C.R., Kassan, J., & Draper, JJ.: Geology of the Bowen and Surat Basins, Eastern Queensland. Australasian Sedimentologists Group Field Guide Series l^o. 8,1993,126p.

No Subsidence

Amalgamated Channel Deposits

Slow Subsidence

Clustered Channel Deposits

Rapid Subsidence

Isolated Channel Deposits

D Lacustrine Delta

Very Rapid Subsidence

Floodplain Shale Channel-fill Sand

Lacustrine Shale

Fig. 1 - Illustration of the effects of increasing the rate of fluvial accommodation on alluvial stacking patterns Sediment supply ,s assumed to be constant and fluvial accommodation is caused by tectonic subsidence. The higher the r a ^ o llrZ^Z.T'' the channels withm floodplam shales. When the rate of increase of accomodation is much higher than the rate of sediment supply, a lacustrine system is formed on the alluvial plain m ) These concepts are based on the work of Allen (1974) and Leeder (1978).

Decelerating accommodation Rate = increasing channel

dusteiing

Lacustrine Delta - -

Lacustrine MFS Coal

Accelerating accommodation rate = decreasing channel clustering

Fig. 2 - Hypothetical example of the effects of changes in the rates of increase in fluvial accommodation on an alluvial succession. Intervals of low accommodation rates are characterised by amalgamated channels and high reservoir connectivity; intervals of rapid increase in accommodation are characterised by isolated channels and low reservoir connectivity.

WALLUMBILLA WYANHRA CADNA - OWIE

Marine Shelf to Coastal ~ Braided Ruvial Ucustrlne

HCX)RAY

Braided and Meandering Ruvial

WESTBORNE

Floodplain to Lacustrine

ADORI

Braided Ruvial

BIRKHEAD

HUTTON

Roodplain and aandering Channels

-SB-

-MFS

MFS

SB— %

MFS

Braided Ruvial

100 m

-400 kmFig. 3 - Example of a regional correlation in the Eromanga Basin in eastern Australia. This basin has formed in an intracratonic setting and as illustrated, the large-scale cycles in alluvial stacking patterns and key surfaces (MFS, SB) can be correlated over hundred's of km. (adapted from Green etai 1989).

THE MT RAWDON GOLD DEPOSIT, SOUTHEAST QUEENSLAND

Maree Angus Senior Project Geologist Placer Exploration Limited, Brisbane Summary The mineralisation and host rocks at Mt Rawdon are Mid to Upper Triassic in age and lies adjacent to the intersection of the Swindon Fault and a north-northwest trending structure, parallel to the Perry Fault. The area has been extensively explored by various companies since the late 1970's and at the time of writing, Placer Pacific is nearing completion of a full feasibility study on the deposit. The origin of the geologically distinct rocks hosting the mineralisation is still the subject of some debate. Introduction The Mt Rawdon gold deposit is situated at latitude 25°16.5'S, longitude 15°46'E, approximately 270 kilometres northnorthwest of Brisbane and 70 kilometres west-southwest of Bundaberg. The deposit lies within the Gympie Province of the New England Orogen (Figure 1). The terrain around Mt Rawdon is dissected hilly to mountainous terrain with the most prominent ranges trending northnorthwest. The ranges are dominated by Lower Triassic granitic intmsives and pyroclastics of the Aranbanga Volcanic Group. Mt Rawdon rises to 256 metres ASL, approximately 150 metres above the Perry River. The most recent resource estimate defined a measured and indicated geological resource of 26.4 Mt at 1.15 g/t Au (932000 contained ounces) and 4.4 g/t Ag. At the time of writing, the mining reserve had not been determined. Deposit History Alluvial gold was discovered at Mt Rawdon in 1946. A small gold rush ensued and the gullies surrounding Mt Rawdon were intensively worked for alluvial gold. A ten-head stamp battery was erected in 1950 and treated ore from small, shallow pits, shafts and open cuts predominantly on the southem slopes of Mt Rawdon. Mining was discontinued in 1953. Total recorded production was 5972g of gold from 758 tonnes of ore (not including alluvial production). A number of mining companies have explored the deposit area including Noranda, Newmont, Samantha, Getty Oil and BHP. The first resource estimate by Placer Exploration was carried out in 1981 (based on 43 percussion holes) and defined a resource of approximately 19 Mt at 1.4 g/t (0.7 g/t cut). Since then drilling totalling 20000 metres has increased the estimate by approximately 76700 ounces. Regional Geology Cranfield (1) identified and mapped the rocks at Mt Rawdon on the Maryborough 1:250 000 sheet as part of the Aranbanga Volcanic Group (AVG) which unconformably overlies rocks of the Curtis Island Group (CIG) and is intmded by Permo-Triassic granitoids. The AVG, previously referred to as the Aranbanga Beds (2), lies within the Coastal Block, one of three major stratigraphic blocks recognised in the region. These blocks young eastward from the Carboniferous Coastal Block, to the Late Carboniferous Goodnight Block and the Permian Gympie Block. Mt Rawdon lies within the eastem Gympie Province of the New England Orogen (NEO). The NEO is interpreted to be a complex tectonic collage of terranes which accreted onto the eastem margin of Australia in the Late Paleozoic (3). It forms the eastem part of the Paleozoic - Early Mesozoic Tasman Fold Belt System described by Day et al. (4) and is subdivided into three provinces; Yarrol in the north, Gympie in the east and New England in the south (3).

Rgurel. Regional Tectonic Setting (modified after Murray, !988)

Cranfield (5) proposes two theories for the evolution of the Gympie Province. Thefirstsuggests that the Gympie Province was an exotic terrane which collided with the AustraUan continent during the Middle to Late Triassic and the second that the Gympie Province formed in place in an extensional basin partly above the Carboniferous accretionary wedge complex. Two major lineaments in the region are the ENE trending Swindon Fault zone and the NNW trending Perry Fault zone. The Coastal Block sediments have a layering and a foliation that reflects these trends. The Swindon Fault zone marks the contact between the AVG and the CIG metasediments to the south. Regionally, the Perry Fault zone is a series of upright faults which are aligned with the eastem margin of the Middle Triassic Esk Trough in the south (6). The dominant movement along these faults is suggested to be sinistral for the Perry Fault zone and dextral along the Swindon Fault zone Multiple phases of volcanic and intrusive activity occurred during the Middle to Late Triassic. Cranfield (5) identifies three main phases. The first at 235-230 Ma includes granodiorite and diorite intrusives, followed by the eruption of olivine basalt, pyroxene andesites and rhyolitic ignimbrite of the AVG at 228-223 Ma. Thefinalphase at 218-215 Ma is associated with cauldron collapse and ringfiracturing,resurgent doming and subvolcanic felsic intmsions. Stephens (6) suggests that the AVG is derived from the partial melting of calcalkaline andesitic volcanics of the Esk Trough. The trace element signatures of calcalkaline and undepleted subcontinental mantle for the AVG suggests they formed during extensional tectonism in an intracratonic rifl setting (6). 28

Molybdenum, copper, copper-gold and epithermal gold mineralisation in the region (Maryborough 1:250 000 sheet) are closely associated with Triassic magmatism (5), Deposit Geology The Mt Rawdon Complex is the local name for the fault bound package of volcanic and intrusive rocks which host mineralisation at Mt Rawdon. While many of the rocks present at Mt Rawdon are peculiar to the Mt Rawdon Complex (MRC), lithologies such as the quartz trachyte dykes are more regional in their extent. The local geology at Mt Rawdon is dominated by dacite intmsives and rhyodacitic volcaniclastics (Figure 2). The two are considered genetically linked, and have been intruded by a sequence of acidic to basic dykes and plugs. The volcaniclastic sequence which hosts the bulk of the mineralisation at Mt Rawdon was deposited during a period of extensional tectonism as grabenfillmaterial. Generally the sequence strikes approximately north west and dips shallowly to moderately south west. The volcaniclastics are typically massive and poorly sorted, with a supporting matrix composed of varying proportions of fine grained lithics, glass, quartz and feldspar crystals. The lithic clasts are generally sub-rounded to rounded and less than 25 millimetres across, though clasts up to 4 metres across have been observed. Continuous marker horizons are absent, but thin ash-rich layers high in the sequence are in places correlateable over 50 to 100 metres. Previously the volcaniclastics have been divided into a lower andesitic to dacite rich unit with an often welded dacite ash matrix, and an upper rhyodacitic lapilli tuff unit with a non-welded sandy matrix. This apparent difference can also be attributed to a decrease in sericite alteration of the volcaniclastics with depth. Clast types include quartzo-feldspathic sandstone and lesser schist of basement origin, rare fine grained black material (possibly basalt) and welded tuffaceous material. By far the most common clast type is of dacitic composition, similar in texture and mineralogy to the dacitic intrusions to the south east of the main minerahsed zone. M. Tons (pers. comm.) has suggested similarities to rocks in Fiji and Papua New Guinea and considers the volcaniclastics to be dominantly a pepperite or hyaloclastite breccia formed by the intrusion of dacite into a wet sub-marine volcanic pile. The main dacite intrusion occupies an area of 4800 m^ and lies south of the main mineralised zone and east of the southwest minerahsed zone. The margins of the main dacite intrusion are steep and often brecciated and mineralised. Depending upon the degree and type of alteration, the dacite ranges in colour from pale cream to dark green. Plagioclase phenocrysts (10-15%) are often replaced by chlorite and carbonate, biotite (1-3%) by opaque oxides and sericite, and amphiboles (0.1%) by chlorite. Disseminated pyrite (1-3%) occurs in the ground mass of quartz (10-25%) and feldspar (20-30%). The dacite dykes at Mt Rawdon are mineralogically and texturally similar to the main dacite intrusion and are thought to be genetically linked. The dykes strike parallel to the northem margin of the main dacite mass and dip at 50-70- towards the south-southeast ie, they dip toward the main dacite intrusion. The dacite dykes are generally well mineralised, especially at their margins, and form a swarm through the main zone of mineralisation. Their exact geometry is hard to predict as they thicken and thin apparently at random and are often difficult to correlate between drill holes. Brooker (7) suggests that there are two phases of dacite dykes ie. the early cream coloured, phyllic altered dacite dykes which are synchronous with the main dacite body, and younger propylitically altered dacites.

South - North Section 200 RL

- 100 RL

- ORL

Legend Andesite/Trachyandesite Andesitic Volcaniciastics Basement Metamorphics Hydrothermal Breccia Dacite Porphyry

Fragmerrtal Dacite Qtz-Feld Porphyry Qtz-Feld-Bt Porphyry Dyke Rhyodacitic Volcaniciastics Trachyte Dyke

^ A

Breccia Mineralised Envelope inferred Faults Pit Outline

Figure 2. Mt Rawdon Local Geology

The smaller fragmental dacite intrusion lies at the northwest comer of the main dacite body, to which it is compositionally equivalent. The exact relationship between the two is not known but the fragmental is interpreted as being slightly younger than the main dacite intrusion. The fragmental dacite intrusion is host to significant mineralisation only where it has been well fractured or brecciated. The margins of the intrusion are steep and are locally flow banded. The fragmental dacite is commonly crackled and its margins often show a gradation from crackled dacite to dacite clast rich 30

- dacite supported breccia, through dacite clast rich-volcaniclastic supported breccia, to volcaniclastic. This observation is consistent with the interpretation that much of the volcaniclastic sequence close to the dacite intrusion is a hyaloclastite breccia. Both mafic-poor and mafic-rich trachyandesites are present at Mt Rawdon. The two types are texturally and mineralogically similar and are considered to be differentiates of the same magma. The trachyandesites generally occur as dykes 2 - 3 metres in width, dipping 60° - 70° toward the main dacite intrusive (ie. similar orientation to the dacite dyke swarm). Like the dacite dykes, the trachyandesite dykes occur more commonly in the prospect area than is obvious at surface. The trachyandesite intrusions host significant mineralisation only where they have been well fractured or brecciated. A series of four quartz trachyte dykes intrude the mdneralised zone at Mt Rawdon. These dykes are not confined to the prospect area, but intrude both the basement CIG and the AVG. The Geological Survey of Queensland regards the trachytes as an intrusive phase of the AVG. Throughout the main mineralised zone the dykes strike approximately eastwest and dip to the north at approximately 30-. The trachyte dykes are without exception only very weakly mineralised. In places they are moderately sericite altered and are cut by stylolites, suggesting that they were intruded just prior to or syn mineralisation, but were unsuitable hosts. The dykes generally have chilled andflowbanded margins and are sparsely porphyritic quartz trachytes with small K-feldspar phenocrysts (4-5%), 1-2% albite phenocrysts, and rare small quartz phenocrysts (0.5-2%). The groundmass is generally 75% feldspar with 5-8% deformed quartz grains (7). The margins of these dykes can be very fractured and in places appear sheared. A porphyritic quartz-feldspar-biotite dyke strikes and dips subparallel to the quartz trachyte dykes, cutting one of the quartz trachyte dykes to the southeast of the main mineralised area. The dyke is coarsely porphyritic with 3-5% quartz eyes, 5-10% plagioclase and 3-5% biotite phenocrysts. The ground mass is dominantly K-feldspar with 30-35% plagioclase. Petrographic work by Brooker (7) categorises the quartz-feldspar-biotite porphyry dyke as a trachyandesite (based on feldspar proportions and composition). Like the quartz trachytes the quartz-feldspar-biotite porphyry dyke is only weakly mineralised throu^out the deposit. Narrow, late andesitic intrusions are common throughout the deposit area. Some carbonate alteration is present , but otherwise the dykes are dark green - black and relatively fresh, suggesting that they are post mineralisation and most of the alteration. The ?Late Devonian to Early Carboniferous CIG which surrounds the MRC comprises mica schist, gneiss, amphibolite and quartzite. The rocks have (prior to the formation of the MRC) been metamorphosed and tightly folded. The sequence is interpreted to have a thickness of ^3000m. The Late Triassic AVG comprises andesitic to rhyolitic flows and pyroclastics, minor polymictic conglomerate and volcanic breccia. The sequence is in excess of 500m thick and represents subaerial flows and pyroclastics derived from an area south east of Mt Perry township (1,8). Alteration & Mineralisation The alteration and mineralisation at Mt Rawdon are multi-stage events which overprint the volcaniclastic sequence, the dacite suite and to a lesser extent the trachyandesite suite. Permeability is the main control on the distribution of the mineralisation. The volcaniclastics and the dacite suite are the most altered rocks with early pervasive propyllitic alteration having been over printed by pervasive sericite alteration prior to the mineralisation. The sericite alteration flanks the westem margin of the main dacite intrusion and continues through and around the fragmental dacite to outline a bowl shaped area adjacent to the north margin of the main dacite. The main dacite is itself pervasively sericite altered and weakly silicified, though the degree of alteration decreases rapidly inwardfromthe margins. Sericite alteration within the mineralised units is accompanied by lesser amounts of illite-smectite, kaolinite (supergene?), quartz and pyrite alteration. The surface extent of mineralisation (>0.7 g/t Au) at Mt Rawdon forms a crudely ovoid zone of dimensions 200m x 300m (Figure 1). The majority of the mineralisation takes the form of pyrite disseminations (generally <3mm diameter) in the matrix of the volcaniclastics within the sericite altered zone. The percentage of disseminated pyrite present rarely exceeds 5%. Veinlets are a significant host to minerahsation and carry variable amounts of pyrite, galena, chalcopyrite, sphalerite, arsenopyrite and free gold. Their width rarely exceeds 5mm (average 1mm) and in areas of average grade (1 g/t) occur at intervals of approximately 0.5m. Veinlets can be planar or quite irregular (stylolitic) and discontinuous. Within any given area at Mt Rawdon the gold grade generally increases as pyrite alteration and sulphide veining increases. An internal study shows that whilst the majority of the gold is hosted by pyrite and base metal veins, there is still a significant proportion of gold contained in the disseminated sulphide mineralisation. The gangue comprises chlorite, carbonate, tremolite and epidote. Other sulphide minerals present include chalcopyrite, sphalerite and galena which occur as discrete grains and as fracture fill in pyrite. Matildite (AgBiS2) , bismuthiniteaikenite, native bismuth, sulfosalts of the (AgBi-Pb2)S2 series, arsenopyrite and hessite (Ag2Te) occur in lesser amounts. There appear to be at least three phases of pyrite alteration present, which comprise an early disseminated replacement phase, then sulphide veining, followed by the breccia veining and associated silica-pyrite alteration. Silica-pyrite 31

alteration associated with breccia dykes forms the other major contributor to gold distribution at Mt Rawdon. The breccia dykes can be up to 1 metre thick and generally dip steeply toward the main dacite body. Gold and electrum occur as free grains, inclusions within base metal sulphides and grains filling fractures or forming inclusions within pyrite (7). The gold observed at Mt Rawdon is closely associated with base metals, bismuth and sulfosalts. Close inspection of hand specimens grading > 5 g/t Au generally reveals visible gold. In thin section, gold is apparent in most intervals > 0.2 g/t Au (7), Gold fineness ranges from 348 - 881 and most grains are between 10 and 30 microns diameter. Microprobe analyses by Brooker (7) revealed two populations of gold grains within this fineness range possibly indicating two phases of gold deposition. Structure The minerahsation at Mt Rawdon lies adjacent to the intersection of the Swindon Fault and the Perry Fault parallel Rawdon Fault. The mineralised envelope plunges south-west at approximately 40° toward this intersection. The best hole in the deposit is located at the north-west comer of the main dacite body within intensely altered and fractured dacite. The northern and western boundaries of the main dacite body strike parallel to the Swindon Fault and Rawdon Fault respectively. Detailed surface mapping of the deposit area has resulted in an extensive surface structure database being developed. The orientation of the pre mineralisation intrusions at Mt Rawdon suggests a strong influence on their emplacement by features in several regionally important orientations including Swindon Fauh and Perry Fault orientations. Zones of brecciation and breccia veining also appear to strike sub parallel to regional trends. Since 1994, approximately 50 oriented drill holes have been completed at Mt Rawdon. Analysis of this data shows changing directions of veining as trends in dyke orientation change locally. This suggests that mineralisation and intrusive activity are closely linked in time and space. Narrow zones of brecciation and faulting occur througfiout the deposit. Faulted intersections are generally rubbly and very few of them have been successfiilly oriented. Attempts to correlate these zones between drill holes and map them on the surface have resulted in only limited success. As such all faults shown on the deposit geology map are still regarded as inferred. The volcanic breccia which approximates the western contact of the main dacite body is the most continuous breccia zone within the deposit and is well mineralised. There is also a well developed hydrothermal breccia developed around the fragmental dacite body. Deposit Genesis As a result of over two decades of exploration at Mt Rawdon there are several schools of thought as to the origins of the host rocks and the mineralisation. In the current exploration model the volcaniclastic sequence is deposited adjacent to a syndepositional dacite dome, in a pull apart basin that develops at the intersection of the Swindon and Perry Faults. Disseminated mineralisation may have been introduced at this stage, but it is not until the stress regime changed to allow intrusion of the trachyte dykes that much of the vein mineralisation developed Conclusion The intersection of two regional structures is a common focus for mineralisation. The Perry Fault and Swindon Fault trends are important in localising structural complexity and/or mineraUsation at Mt Rawdon and at other deposits/prospects in the region. There are three main factors that have contributed to the locahsation of mineralisation at Mt Rawdon. 1. The volcaniclastics and breccias are generally permeable and a good host to disseminated mineralisation. 2. The presence of multiple intrusive phases within the Mt Rawdon Complex coincident with changes in stress regimes has resulted in intrusive contacts that now localise significant mineralisation. 3. Some of the intrusive phases appear to be contemporaneous with mineraUsation and are a source of ore fluids. The age of the intrusive activity is similar to the age of the Hogback Granite to the north. The link between mineralisation and Mid to Late Triassic magmatism has long been inferred in the region and research to begin in 1997 will investigate this link at Mt Rawdon and confirm the origin of the volcaniclastic sequence. References 1.

CRANFIELD, L.C., 1989. New and Revised Stratigraphic Units in the Maryborough 1:250000 sheet area. Southeast Queensland. Qld Govt. Mining Journal, 90, 163-174. 32

ELLIS, P.L., 1968. Geology of the Maryborough 1:250000 sheet area. Geological Survey of Queensland. Report

FLOOD, P.G., 1988. New England Orogen - Geosyncline, Mobile Belt and Terranes, in New England Orogen Tectonics and Metallogenesis, (ID. Kleeman ed.) DAY, R.W. et. AL, 1978. The Eastern Part of the Tasman Orogenic Zone. Tectonophysics, 48, 327-364. CRANFIELD, L.C., 1994: 1:250000 Geological Series - Explanatory Notes - Maryborough, Queensland. Geological Survey of Queensland, Department ofMinerals and Energy. STEPHENS, C.J., 1991: The Mungore Cauldron and Gayndah Centre-Late Triassic large scale silicic volcanism in the New England Fold Belt near Gayndah, southeast Queensland. Unpubl PhD thesis, University of Queensland. BROOKER, M.R., 1991: Geology, Alteration and Mineralisation of the Mt Rawdon Diatreme-hosted Gold Deposit. Unpubl MSc thesis, James Cook University ofNorth Queensland. CRANFIELD, L.C., 1986. The Geology of the South Burnett District, in 1986 Field Conference, South Burnett District (W.F. Wilmott ed).

4. 5. 6. 7. 8.

26.

METALLOGENESIS RELATED TO TRIASSIC MAGMATISM IN THE NEW ENGLAND OROGEN RM. AshleyS R.G. Banies% S.D. Golding' and C J. Stephens' 1 Department of Geology and Geophysics, University of New England, Armidale, NSW 2351 2 Geological Survey of NSW, Department of Mineral Resources, 97 Faulkner St., Armidale, NSW 2350 3 Department of Earth Sciences, University of Queensland, Brisbane, Queensland 4072 4 Central Norseman Gold Corporation Limited, Norseman, WA 6443 Summary Triassic magmatism in the New England Orogen is directly and indirectly responsible for the majority of metal deposits. Granitoid emplacement and associated volcanism span much of the Triassic, with an initial magmatic arc setting being followed by periods of extension and rift-related magmatism. Mineral deposits are numerous and are classified into six major types, all of which have probably formed at high crustal levels: granitoid-related deposits, porphyry and breccia-hosted deposits, skams, epithermal systems, layered intrusions and mesothermal vein systems. Apart from deposits in layered intrusions, all others are of hydrothermal origin, commonly with a magmatic fluid source, but in places involving extemal fluids. Currently operating mines include Gympie, Shamrock and Manumbar (Au), Hillgrove (AuSb) and Biggenden (magnetite), but large resources of Au are outlined at Mt Rawdon and Timbarra. Mineral exploration potential is assessed as moderate to high for future discoveries related to Triassic magmatism in the Orogen. Introduction In the Tasman Fold Belt System of eastern Australia, there is evidence for several episodes of magmatism and related metallogenesis (1-10). A major episode occurred in the Triassic, affecting the New England Orogen (NEO) in northeastem NSW and eastem Queensland (2,3,7). The Triassic magmatic and metallogenic episode in the NEO is regionally extensive and historically important for the development of the region, with significant discoveries of Au, Sn, Mo, Cu and Sb being made in the latter half of the 19th century. Although mineral deposits related to Triassic magmatism are exceedingly numerous (several thousand recorded), in the main, production has been relatively small and to date there have been few "world class" discoveries. Gold production exceeding 1101 from Gympie far exceeds that from any other Triassic deposit in the Orogen (being fourth largest in the Tasman Fold Belt System); other significant Au producers have included Hillgrove (231), Nundle, Eidsvold, Drake and Uralla-Rocky River (11). In addition to Au, Triassic magmatism in the NEO was responsible for the generation of three major Sn fields, at Stanthorpe, Emmaville-Torrington and Tingha-Elsmore, with production of approximately 65 000,90 000 and 70 0001 of cassiterite concentrates from the respective fields (12). Lastly, the Hillgrove deposits have been Australia's premier source of Sb, with production exceeding 48 000 t of stibnite concentrates (13). Current mining of deposits related directly and indirectly to Triassic magmatism in the NEO are restricted to Au at Gympie, Manumbar and Shamrock, AuSb at Hillgrove, and magnetite at Biggenden. Major Au resources occur at Mt Rawdon (14; Angus, this volume) and Timbarra (Simmons and Pollard, this volume) and may be subject to development in the near future. Significant resources of AuAg remain at Drake, SbAu at Brackins Spur and Au at Enmore-Melrose. A large low-grade Sn resource occurs at Taronga (12) and together with topaz (silexite) deposits at Torrington, kaolinite at Elsmore, Sb at Bielsdown, Zn at Ban Ban and ilmenite at Goondicum, offer potential for fiiture mining operations. For the purposes of this paper, the Triassic spans the period from 251 Ma to 205 Ma (15). Consequently, we include discussion of magmatic activity and associated mineralisation which probably extended across the Permian-Triassic boundary at around 250 Ma. This event appears to have been more significant in the southem part of the NEO (SNEO) compared to the northem part (NNEO) (i.e. the respective parts of the Orogen south and north of the Clarence-Moreton and Surat Basins). Triassic tectonics and magmatism in the New England Orogen Models for the Triassic development of the NEO and associated magmatism have been proposed, inter alia, by Murray (2, 3), Collins et al. (16), Gust et al. (17), Veevers et al. (18) and Holcombe et al. (19). Geochronological data for this paper'are mostly derived from the compilations of Shaw and Flood (20) and Veevers et al. (18). In the NEO, the late Permian deformation of the Hunter-Bowen Orogeny (-265-260 Ma) caused folding, faulting and uplift of the DevonianCarboniferous arc-forearc and accretion-subduction assemblages, late Carboniferous to early Permian S-type granites and early Permian volcanic and sedimentary rocks. West-directed thrusting occurred in the westem part of the NEO. A major period of dominantly calc-alkaline magma emplacement to high crustal levels followed, commencing at about 255 Ma, and continuing throu^ the Triassic to about 210 Ma, with intrusion of several hundred plutons and eruption of com a ^ a t i c volcanic rocks. Continued uplift was probably occurring in the early part of this period and subsequently led to a major erosional cycle with deposition of magmatic arc-derived sedimentary rocks into adjacent basins.

The tectonic setting in the NEO may have changed from magmatic arc (in the period from --255 Ma to -240 Ma; residual from subduction that may have been occurring in the late Permian) to events of extension and rift-related volcanism and intrusion. In the NNEO, the Esk Trough records an early period of extension and magmatism (-241-233 Ma), whereas a later period of extension and magmatism occurred from '-236-~210 Ma in both the SNEO and NNEO. In both portions of the NEO, there was co-magmatic volcanism accompanying emplacement of plutons. Indeed, magmatic emplacement was more-or-less continuous in the NEO through the Triassic, wi& a major peak between 251-238 Ma and a subsidiary peak between 228-224 Ma. This pattem is clearer in the SNEO, but its more diffuse nature in the NNEO may be due to insufficient geochronological data. In the SNEO, differing suites of granitoids and associated volcanic rocks have long been recognised (e.g. 21-23) and are associated with specific hydrothermal mineralisation styles (4). The Clarence River Suite has relatively primitive I-type (to M-type) geochemical characteristics and an age range of 255-248 Ma (20, 23-25). It has strong geochemical affinities with the spatially associated Drake Volcanics and is regarded as a likely plutonic equivalent. Several CuAu vein type and magnetite skam deposits are related to the Clarence River Suite and there are also weak alteration/minerahsation zones with porphyry-type affinities. The Drake Volcanics host numerous epithermal precious metal-base metal deposits. The Moonbi, Uralla and Nundle Suites are I-type, but the Moonbi Suite has higJi-K affinities. They were emplaced between 251-243 Ma and may have eruptive equivalents in the extensive Wandsworth Volcanic Group (including the cauldron-filling Dundee Rhyodacite). The Moonbi and Uralla Suites display a wide range in fractionation with the Moonbi Suite being rather oxidised. More mafic members of this suite are associated with CuAuMoW skams and veins whereas more felsic members are associated with major Mo (-Bi) and possibly Au occurrences. Despite their widespread preservation, no high-level epithermal or breccia-hosted mineralisation is known from the Wandsworth Volcanic Group; most mineralisation is of mesothermal vein type, related to the Moonbi and leucogranite suites. Widespread mesothermal vein Au, grading to vein Sb deposits occur in the SNEO. Although there is no strong spatial link between these deposits and granitoids, there are temporal links between the Uralla and Moonbi Suites and lamprophyre dykes which commonly occur in the same structures hosting Au and Sb deposits. The dykes have geochemical affinities with the Moonbi Suite (13,26). Several plutons of leucogranite with an age of 244-236 Ma occur in the SNEO. They are of fractionated I-type affinity and have near-minimum melt compositions. They host numerous and diverse mineral deposits, including veins, pipes, greisen systems, disseminations and breccia masses and have produced most of the Sn, W, Mo, As and some of the Au, Ag, Pb and Zn in the SNEO (4). From metal association and oxidation state considerations, it is probable that two leucogranite suites exist in the SNEO: a westerly, more reduced suite with Sn-dominant deposits and polymetallic veins, and an easterly, more oxidised suite with Mo and Au deposits (cf. 27). Volcanic equivalents of the leucogranites do not appear to have been recognised. The Gundle Suite of I-type and minor A-type granitoids with an age range of 236-222 Ma occurs on the eastem side of the SN'EO (20, 28). Manifestations of this magmatism could also include bimodal volcanics in the Lome Basin, the Werrikimbe Ignimbrite, bimodal volcanics at the southem part of Clarence-Moreton Basin, felsic volcanics in the Ipswich Basin, and 210-206 Ma granitoids intrusive into the Lome Basin. Like the Moonbi Suite, there is a large range in fractionation and possibly differences in oxidation state in the Gundle Suite granitoids. Some plutons contain Mo vein deposits, others are associated with Sn, Au and polymetallic vein systems. In the NNEO, magmatism was relatively continuous from 251-220 Ma and included an initial period of magmatic arc development, followed by formation of the volcano-sedimentary rift of the Esk Trough (at about 241-233 Ma), with emplacement of the andesite-dominated Toogoolawah Group. Intrusions prior to 233 Ma include most of the layered mafic to ultramafic bodies and I-type granitoids showing a wide compositional range. In the late Triassic, much magmatic activity was concentrated between 228-220 Ma and included the rift-related bimodal Aranbanga Volcanic Group, emplaced at about 221 Ma, together with the Muncon, Agnes Water, Ooramera, North Arm and Chillingham Volcanics, and the Brisbane Tuff (e.g. 29-31). In places, these were accompanied by intrusion of I-type granitoids (commonly rather leucocratic and locally in volcanic cauldron complexes, e.g. 32, 33) and by rare A-type granites; however, certain of the late Triassic volcanic suites do not appear to have exposed plutonic equivalents. The Gympie Province, on the eastem margin of the NNEO, must have been deformed by 235-240 Ma (31), as it has been intruded by I-type plutons of this age. It was subsequently overlain by the little-deformed late Triassic North Arm Volcanics. In the NNEO, the layered mafic to ultramafic intrusions contain cumulate concentrations of magnetite and ilmenite (e.g. 34-36) and rarely, sulphides associated with minor PGE values (37). The relatively mafic I-type granitoids and porphyry bodies emplaced prior to 233 Ma appear to have been responsible for widespread mesothermal Cu-dominant and Au veins, porphyry Cu (MoAu) and porphyry Mo deposits and some skams and Au-bearing breccia masses. The Manumbar epithermal precious metal system and Hg deposits in the Kilkivan region (38) may be related to andesitic volcanism in the Toogoolawah Group. The timing of formation of mesothermal Au and Sb vein systems and their relation to magmatism in the NNEO (e.g. in the Gympie Province) remains equivocal, but it could be related to the deformational event in the Gympie Province in the early to middle Triassic. Mineral deposits related to the Late Triassic magmatic activity include a variety of skam, vein and breccia systems, commonly containing Au, but with base metals and Bi, associated with granitoids, and epithermal, low-sulphidation precious metal systems associated with felsic volcanics.

Leucogranites of the type associated with major Sn (W) deposits and polymetallic veins in the SNEO appear to be absent in the NNEO. Equivalents of the high-K Moonbi Plutonic Suite (with an extended fractionation range leading to the formation of significant Mo and Au deposits) may also be uncommon in the NNEO. It is possible that equivalents of the Clarence River Plutonic Suite and Gundle Suite constitute many of the plutons in the NNEO. There are however, major differences in the geochemical nature and the ages of the volcanic assemblages in the two parts of the Orogen (e.g. there appear to be no direct equivalents of the Wandsworth Volcanic Group or the Drake Volcanics in the NNEO and Ukewise, temporal/geochemical correlatives of the Toogoolawah Group and Aranbanga Volcanics are uncommon in the SNEO). Clearly there has been significantly different levels of erosion (and by inference, uplift) in the NEO. This has influenced the distribution of onlapping Mesozoic sedimentary basins, the preservation of Triassic volcanics and high-level hydrothermal systems, such as porphyry, breccia-hosted and epithemial deposits, and led to variable levels of exposure of the granitoids. In places, high level components of granitoids, including hydrothermal mineral deposits, such as porphyry and greisen systems, have been eroded, leading to the formation of placer deposits, especially Sn and Au. Placer deposits could represent continued recycling from the early Mesozoic degradatior of prior high-standing magmatic arcs, and from the Tertiary rejuvenation leading to the formation of the eastem highlai : as of Australia. Mineral Deposits Six major classes of mineral deposit have been identified as being directly or indkectly related to Triassic magmatism in the New England Orogen (Table 1). Data for this compilation are from Weber (39 j, Murray (2, 3), Bames et al. (4) and GilUgan and Bames (5), augmented by more recent sources. Granitoid-related deposits

This type of mineralisation is hosted within granitoids or located in the immediately adjacent country rocks. The historically important Sn deposits (cassiterite-bearing disseminated greisen, simple greisen-bordered vein, sheeted vein and pegmatite) are of this type and are associated with SNEO leucogranites (Ruby Creek, Mole, Gilgai, Elsmore Granites) in the Stanthorpe, Emmaville-Torrington and Tingha-Elsmore regions (4,12,39). Other examples occur in the Gundle Suite granitoids. Granites are typically relatively reduced in character and in places show conspicuous enrichment in F (and rarely, B). Deposits typically contain cassiterite, with local wolframite, molybdenite, arsenopyrite and later base metal sulphides. Hydrothermal alteration of microgranite to form quartz-topaz rock (silexite) is characteristic of the roof zone of the Mole Granite. Some silexite deposits contain low grade disseminated wolframite as well as Li and Be minerals (40) and have been investigated as a source of industrial topaz. Recent Ar-Ar age determinations on greisen alteration from deposits associated with the Mole Granite have indicated hydrothermal formation at -241-246 Ma (41). Molybdenum deposits are commonly associated with relatively oxidised granitoids in the SNEO (probably mostly members of the Moonbi Suite) which range from mesocratic to leucocratic. Deposits are of pipe, vein and disseminated form, in places associated with local strong sericitic and silicic alteration, adjacent to the margins (roof) of the intrusions. Some deposits have also been exploited for Bi (e.g. Kingsgate) and may contain minor Sn, W, Au, base metals and U (42). It is proposed that there is a link between the Mo-bearing granites in the SNEO and the occurrence of certain Au deposits. The Timbarra Au deposit contains disseminated Au hosted in weakly altered (sericite-chlorite-carbonate) leucocratic granite (Stanthorpe Granite), concentrated near the lower margins of dipping microgranite sills and dykes. Anomalous Mo and As values are associated with Au mineralisation. Several placer Au deposits in the eastem part of the SNEO can be inferred to have been derived by the erosion of similar granite-hosted disseminated Au occurrences. Other Au vein systems are hosted in granitoids and adjacent country rocks in the NEO, but contain base metal sulphides, arsenopyrite, pyrrhotite and Bi minerals (e.g. Bouldercombe, Eidsvold, Mt Steadman, Comet) and have typical mesothermal vein characteristics.

Table 1. Mineral deposit types related to Triassic magmatism in the New England Orogen Granitoid-related deposits Sn (W) deposits (veins, sheeted veins, stockworks, disseminations, pegmatites) Topaz-W (LiBe) deposits (silexite alteration zones in microgranite) Mo (BiSnWAu) deposits (pipes, veins, sheeted veins, disseminations) Au deposits (disseminations and veins) Cu-dominant mesothermal veins (commonly with AuMoWZnPbAg association) Polymetallic veins (commonly with AsCuZnPbAgSn association) Hyi:othermal kaohnite Porphyry and breccia-hosted deposits Cu (MoAu) porphyry systems Mo (SnW) porphyry and breccia systems Au (Ag CuPbZnBi) breccia masses Skams Cu (Au) skams Zn skams Magnetite (AuCuBi) skams Sn (CuAgAs) skams W (Mo) skams Epithermal veins and breccias Low-sulphidation, precious metal and base metal types Low-sulphidation, precious metal types Layered mafic to ultramafic intrusions Magnetite-ilmenite cumulates Sulphide-bearing PGE occurrences Mesothermal AuSbWHg veins (indirect association with igneous rocks) Au veins AuSb (W) veins Sb veins Hg veins Numerous granite-related mesothermal vein systems, dominated by Cu and Au, have been historically worked in the NEO, commonly with Au production being derived from the weathered and supergene-enriched portions of the deposits. Associated granitoids are typically rather mafic I-types and porphyry dykes are also commonly found with this type of deposit, which show zoned alteration assemblages, locally involving potassic alteration, grading outwards into sericitic and propylitic alteration. Steeply dipping vein systems are commonly rich in quartz, pyrite and chalcopyrite, but may also contain appreciable amounts of sphalerite, galena, molybdenite and scheelite. Some deposits have been emplaced into country rocks, including those at Mt Mudlo, Ban Ban Au mine, Warroo and Cangai (into metasedimentary rocks) and Mt Perry (into older granite). The operating Shamrock Au mine near Kilkivan is viewed as an example of this type of vein deposit, although it may have characterisitics transitional into a porphyry ore system. Another class of vein system, characterised by a polymetallic assemblage, is associated with SNEO leucogranites. Deposits are located within the causative intmsions (e.g. at Conrad, Webbs Consols, Rivertree) or within adjacent metasediments/felsic volcanics (Collisons, Tangoa, Boorook) or older granitoids (e.g. Rockvale area). Strong sericite alteration is typical about steeply dipping lode stmctures. Several deposits have been worked for Ag (e.g. Collisons, Webbs Consols, Conrad) or As (e.g. Ottery, Mole River, Jibbinbar), but ores were metallurgically complex, with relatively high values of Cu, Pb, Zn, Sn, and locally Sb and E Sulphide-rich vein material may be associated with quartz, carbonates and fluorite. Polymetallic veins are generally viewed as more distal or temporally later manifestations of the leucogranite mineralisation spectrum ranging from Sn-W-Mo to As-Cu to Zn-Pb-Ag-Sb. A high grade kaolinite resource occurs at Elsmore in the SNEO (43), where leucogranite has been altered to a kaolinite-quartz assemblage in the roof zone of the intmsion. It is possible that the resource is due to low-temperature retrograde alteration following the formation of an associated greisen-bordered Sn (MoW) sheeted vein system. Porphyry and breccia-hosted deposits Numerous large-scale hydrothermal systems with fracture-controlled and disseminated Cu (MoAu) mineralisation of the type characteristic of porphyry ore deposits occiir in the NNEO (1-3) but no deposit to date has proven to be economically 37

viable. Most are of Triassic age and are associated with diorite, tonalite and granodiorite porphyries (commonly in complex, multiphase intrusions) with zoned alteration of potassic, phyllic and propylitic types. Emplacement of porphyry intrusions has been influenced by regional-scale lineaments. Mineralisation occurs within the intrusives and in associated breccia masses and fractured country rock. Fracture-controlled and disseminated pyrite is typical and may be accompanied by chalcopyrite, bomite, molybdenite, magnetite and hematite. Hypogene grades at several deposits are 0.25-0.3% Cu (e.g. Moonmera, Riverhead, Coalstoun) and supergene processes have locally enhanced Cu values. To date, no deposit with strong porphyry Cu system alteration-mineralisation characteristics has been recognised in the SNEO. This may be a function of a generally deeper level of erosion of intmsions in some areas, or altematively, not deep enough erosion (e.g. in the Drake Volcanics). Fracture-controlled and disseminated Mo (SnW) mineraUsation occurs in several places in the NEO and has certain characteristics of porphyry Mo systems (e.g. Anduramba, Glen Eden). Again, however, no deposit has proven to be of ore grade. Deposits are associated with leucogranite, microgranite and quartz-feldspar porphyry, with associated strong silicification and greisenisation, grading into sericitic and argillic alteration. Minerali^tion consists of molybdenite, with minor cassiterite, wolframite, topaz, beryl and base metal sulphides. The Glen Eden MoSnW system in the SNEO displays a central greisen breccia zone superimposed on felsic volcanics, with afSnities to vein and pipe Mo occurrences in the region. Breccia-hosted AuAg (-base metal-Bi) mineralisation is associated with Triassic felsic to intermediate porphyritic intrusives at Mt Rawdon and Mt Shamrock-Mt Ophir in the NNEO. There is a spatial association between these occurrences and felsic to intermediate volcanics of the Aranbanga Volcanic Group. The Mt Rawdon mass has been interpreted as a phreatomagmatic diatreme by Brooker and Jaireth (14) and displays argillic, propylitic and phyllic styles of alteration. Significant Au mineralisation is largely associated with sericite alteration and with pyrite, base metal sulphides, chlorite and carbonate. Rare Bi minerals occur at Mt Rawdon and are notable trace constituents in several Triassic Au- and base metal-bearing hydrothermal systems in the Biggenden region, including the Mt Shamrock breccia pipe. Skams Several skam deposits are located adjacent to Triassic granitoids in the NEO, but are not abundant, mainly due to the paucity of suitably reactive host rocks (e.g. carbonates). Cu (Au) skams are located adjacent to mafic granitoids (e.g. at Glassford Creek, Attunga) and display rather oxidised mineralisation assemblages (andradite, magnetite, pyrite, chalcopyrite, bomite) with associated minor Bi, W and Mo minerals. Endoskam alteration of the intrusives occurs locally. The Ban Ban Zn deposit lies adjacent to leucocratic, greisen-altered granite and replacement of marble has led to the formation of stratabound, sphalerite-rich gamet-dominated skam, with minor Cu, Pb, Sn, Bi and Ag. The Biggenden magnetite (CuBiAu) deposit, a current major source of magnetite for coal washery purposes, lies adjacent to a relatively unaltered, late Triassic granite. Magnetite-gamet-calcite skam with associated pegmatitic calcite masses have replaced marble and metasedimentary and metavolcanic homfelses. Sulphide-rich veins, patches and disseminations are dominated by pyrite and chalcopyrite, but are locally rich in Bi minerals, with minor molybdenite, arsenopyrite and cobaltite. Sulphides are associated with retrogression of the prograde skam assemblages. Gold was formerly produced from Biggenden, mainly from the supergene zone. The low grade Sn (CuAg) skam at Willi Willi has formed by gamethedenbergite replacement of carbonate-bearing strata adjacent to a granitoid member of the Gundle Suite. Tin occurs as cassiterite, stannian andradite and malayaite, and is associated with Fe, As, Cu, Zn, Pb, Sn and Bi sulphides. The scheelite-molybdenite-bearing andradite skam at Attunga lies adjacent to a high-K intermediate intrusion, with major exoskam replacement of marble and calc-silicate homfelses and minor endoskam replacement of the quartz monzonite intmsive. In general, the relationship between types of skam mineralisation and the chemistry and oxidation state of the implied causative intmsions in the NEO conform to models for skam formation of Einaudi et al. (44) and Meinert (45). There are, however, potential inconsistencies at certain locations (e.g. Biggenden) and the need for additional research on these hydrothermal systems remains. Epithermal veins and breccias

Mineral deposits and other hydrothermal systems with epithermal vein and breccia characteristics are uncommon in tiie Triassic of the NEO. In part, this may be a function of the current level of exposure and the relatively low preservation potential of these types of systems. It is likely that the recognised epithermal systems are all variants of the "lowsulphidation" type of White and Hedenquist (46). Several epithermal AuAg-base metal deposits occur in the Drake Volcanics in the SNEO and are locally associated with porphyritic andesitic, dacitic and rhyolitic intmsives, and andesitic lavas and fragmentals (47-51). Intense sericite-quartz-carbonate alteration and silicification associated with vein, stockwork, breccia and disseminated mineraUsation grades outwards into propylitic assemblages. Electrum is commonly found with Ag sulphosalts, base metal sulphides, tetrahedrite and pyrite. Although recent mining at Drake (1989-90) extracted only Au and Ag (in part from supergene-enriched deposits), several other deposits have been previously worked for Cu; some of these have characteristics gradational into mesothermal Cu (Au) veins. In the NNEO, epithermal precious metal mineralisation with low base metal values occurs at Manumbar (operating mine exploiting carbonate-rich veins hosted in andesitic volcanics of the Toogoolawah Group) and North Arm. At the latter, quartz vein, stockwork and breccia-hosted AuAg mineralisation occurs in rhyolitic and dacitic rocks of the North Arm Volcanics showing local Kfeldspar and widespread phyllic, argillic and propylitic alteration. The nearby late Triassic alteration zone at Mt Ninderry 38

(52) is interpreted to represent an acid sulphate cap over a potential boiling zone containing low-sulphidation style epithermal mineralisation. Layered mafic to ultramafic intrusions Several late Permian to Triassic layered intrusions in central and southern Queensland have been prospected for magnetite. Small resources of medium grade Fe occur at Eulogie Park and larger, lower grade deposits containing magnetite and ilmenite occur at Hawkwood and Wateranga (2). Detrital ilmenite occurrences (due to weathering of the intrusions) have also been investigated at Eulogie Park, Goondicum and Wateranga. A major problem with potential exploitation of magnetite resources in these intrusions is their low grade and intergrowth of magnetite with Ti-bearing phases and apatite. E^loration for PGEOs has also been focussed on the layered intrusions, but only the Bucknalla Complex has provided encouragement to date, with sporadic PGE values being found with late magmatic CuFe sulphides (37). MesothermalAuSbWHg veins This diverse category of mineral deposit occurs throughout the NEO, and although examples are far more numerous in the SNEO, the Gympie Au deposit in the NNEO has by far the largest recorded production. Deposits have a common thread of structural control and a mineralogical constitution which is generally simple, i.e. quartz-rich veins, commonly with minor pyrite, arsenopyrite and carbonate, with variants grading into stibnite-rich and locally scheelite- and base metal sulphide-bearing. A distinct type, probably formed at shallower crustal depths, is the uncommon, epithermal-like, cinnabar-carbonate-qxiartz veins, which, in places, appear to be superimposed on earlier Cu (Au) vein systems (e.g. in the Kilkivan and Baryulgil areas). As a whole, this deposit class is found as veins, stockworks and breccia masses cutting diverse rock types, consequently with quite variable alteration haloes which are dependent on host rock permeability and reactivity. Alteration minerals include carbonates, sericite, chlorite, quartz, pyrite and arsenopyrite; fuchsite is conspicuous in ultramafic to mafic host rocks. Intermediate to mafic dykes, emplaced along the same minerahsed structures and at similar times to hydrothermal activity are locally common; some dykes in the SNEO are known to be lamprophyre (13,26, 53), but others (e.g. at Gympie) are termed OmicrodioriteO or OdoleriteO. Due to the spatial and temporal relationships between dykes and mineralisation, more investigations of these dykes need to be published. In places, mineralisation occupies brittle structures superimposed on earlier mylonitic fabrics, implying re-activation of structures at high crustal levels. Apart from the restricted occurrences of dykes in the lode channels, there is no commonality for the occurrence of igneous rocks which could be genetically implicated in the formation of this deposit type. In the NEO, many groups of mesothermal Au deposits occur remote from similar-age intrusions, being found in metasedimentary and metavolcanic rocks (e.g. Gympie, Warwick area, Kingston, Coramba-Orara, Dalmorton, Hillgrove, Enmore-Melrose, Weabonga), in older, commonly deformed granitoids (e.g. Kookabookra, Hillgrove, Enmore-Melrose, ?Uralla-Rocky River) and in serpentinite (e.g. Canoona, Mt Wheeler-Cawarral, Bingara, Nundle). In parts of the SNEO, it can be implied that many of the mesothermal Au and Sb vein systems must have formed in the late Permian to early Triassic, based on geochronology (257-243 Ma) and field relations (13, 54). This period overlaps with the emplacement of the voluminous I-type plutons into the SNEO (20). However, geochronological data on certain groups of metasediment-hosted deposits with spatial links to the middle Triassic Gundle Suite granitoids (e.g. Munga Creek, Taylors Arm and Bielsdown Sb deposits, Coramba-Orara Au deposits) are lacking. In the NNEO, the timing of mesothermal Au and Sb mineralisation remains unclear, but in the Gympie Province it must have developed during the early to middle Triassic deformation. In the Kilkivan district, at least some of the Hg vein deposits occur in andesitic rocks of the early to middle Triassic Toogoolawah Group, thus providing a maximum age limit for their formation. It is proposed that mesothermal Au and Sb mineralisation in the NEO may have formed episodically, with an earlier phase in the late Permian to early Triassic, mostly manifest in the SNEO, and a later phase in the early to middle Triassic, mostly in the NNEO, but possible including extensive Au and Sb deposits on the eastem side of the SNEO. These two periods correspond with (1) granitoid intrusion and contact metamorphism, uphft and regional relaxation (indicated by extensional veining), mainly in the westem part of the SNEO in the interval from -255-243 Ma, and (2) granitoid intrusion and contact metamorphism, deformation of the Gympie Province, and uplift, perhaps affecting the eastem parts of the Orogen in the interval from --235-225 Ma. Deep crustal processes involving devolatilisation, perhaps due to high heat flow and magma emplacement, may be implicated in the generation of this class of mineral deposit; stable and Pb isotopic data are not definitive of a direct magmatic fluid contribution. Fluid and metal sources Although most of the Triassic metallogeny of the NEO can be related directly or indirectly to magmatic processes, it is likely that in the hydrothermal deposits (which comprise all deposits apart from those in layered intrusions), the sources of fluids and metals may be diverse. Limited stable isotope and fluid inclusion studies (14, 55-58) of deposits with a strong link to the crystallisation of granitoids and related porphyry intrusions (i.e. granitoid-related deposits, porphyry and breccia-hosted deposits and skams) have indicated that high-temperature magmatic fluids are involved in initial mineralisation and that mineral deposition has occurred due to a variety of mechanisms including rock reaction, 39

temperature decrease and fluid boiling. Consequently, it is likely that the granitoids were the major source of ore components. In some deposits there is strong evidence for the later incursion of extemal (e.g. meteoric) fluids causing retrograde alteration and lower temperature mineralisation. Epithermal mineralisation at Red Rock (Drake area) may have a predominantly marine fluid source (49), whereas at North Arm and Mt Ninderry, Ashley and Andrew (52) interpreted that Triassic meteoric waters were dominantly responsible for alteration and mineralisation. In each case, the source of metals could be inferred to be from leaching of the rock sequences. Fluid and metal sources for mesothermal Au and Sb deposits in the NEO are more equivocal as stable isotope results are consistent with either a homogenised crustal or magmatic origin (e.g. 13, 59). A metamorphogenic source is favoured, perhaps derived from devolatilisation reactions during crustal emplacement of major volumes of granitoids, but direct contribution of fluids and metals from crystaUising granitoids is also possible. Discovery Potential Based on historical discoveries, re-interpretations of the nature and genesis of certain deposit types, and advances in exploration techniques, it is predicted that the NEO retains a moderate to high potential for the discovery of additional ore deposits and that these will be mostly related to Triassic magmatism. In less eroded portions of the Orogen, epithermal and breccia-hosted AuAg and porphyry CuAu remain prospective at sites of regional structural control; clearly the NNEO retains considerable potential, but in the SNEO, the Drake Volcanics offer the best encouragement. Porphyry Mo and Mo-bearing breccia masses may constitute targets in the SNEO, in association with the more fractionated, oxidised I-type granitoids, in areas variably covered by felsic volcanic sequences. The unusual (at present) granite-hosted deposit at Timbarra presents inspiration for additional Au discovery infractionated,oxidised granites. An appreciation of subtle alteration, pluton roof geometry, late-stage dykes/sills and the presence of alluvial/eluvial Au deposits is necessary. The Orogen has limited potential for the generation of skam deposits as potentially reactive host rocks are not common in sites proximal to intrusives. The Attunga region hosts potential for CuAuWMo skams and the Biggenden-Kilkivan and Many Peaks-Calliope regions potential for CuAu and magnetite skams. More distal replacements of carbonate and organic-bearing sequences to form the subtle "Carhn-type" Au deposits are possible in these regions. The SNEO retains considerable potential for occurrence of sheeted vein, disseminated greisen and replacement Sn (W) deposits, but these are likely to be under cover and related to cupolas in the roof zones of leucogranites. Finally, both portions of the NEO offer fiirther potential for discovery of mesothermal Au and AuSb vein systems in suitable structural settings, perhaps along strike from, or adjacent to, historically productive mines such as at Gympie, Hillgrove and Enmore-Melrose. Acknowledgements For the acquisition of valuable background knowledge on the metallogenesis and geological development of the New England Orogen, we wish to thank Len Cranfield, Peter Flood, Cec Murray and Jim Stroud. RGB publishes with the pennisssion of the Director-General of the NSW Department of Mineral Resources. References 1.

Horton D.J. 1978. Porphyry-type copper-molybdenum belts in eastem Queensland. Economic Geology 73, 904921. Murray C.G. 1986. Metallogeny and tectonic development of the Tasman Fold Belt System in Queensland. Ore Geology Reviews 1, 315^00. Murray C.G. 1990. Tasman Fold Belt in Queensland. In: Hughes F.E. (ed.) Geology of the mineral deposits of Australia and Papua New Guinea. Australasian Institute of Mining and Metallurgy, Melboume, pp. 1431-1450. Bames R.G., Brown R.E., Brownlow, J.W., Gilligan L.B., Krynen J. and WiUis LL. 1988. A review of the mineral deposits of the New England Orogen in New South Wales. In: Kleeman J.D. (ed.) New England Orogen tectonics and metallogenesis. Department of Geology and Geophysics, University of New England, Armidale, pp. 211-227. Gilligan L.B. and Bames R.G. 1990. New England Fold Belt, New South Wales - regional geology and mineralisation. In: Hughes F.E. (ed.) Geology of the mineral deposits of Australia and Papua New Guinea. Australasian Institute of Mining and Metallurgy, Melboume, pp. 1417-1423. Suppel D.W. and Scheibner E. 1990. Lachlan Fold Belt in New South Wales - regional geology and mineral deposits. In: Hughes F.E. (ed.) Geology of the mineral deposits of Australia and Papua New Guinea. Australasian Institute of Mining and Metallurgy, Melboume, pp. 1321-1327. Golding S.D., Stephens C.J., Vasconcelos PV., Holcombe R.J. and Fielding C.R. 1994. Metallogeny of the Northem New England Fold Belt. In: Holcombe R.J., Stephens C.R. and Fielding C.R. (eds.) 1994 Field Conference, Capricorn region, central Queensland. Geological Society of Australia, Queensland Division, Brisbane, pp. 137-145. Perkins C., Walshe J.L. and Morrison G. 1995. Metallogenic episodes of the Tasman Fold Belt System, eastem Australia. Economic Geology, 90,1443-1466. Walshe J.L., Heithersay PS. and Morrison G.W. 1995. Toward an understanding of the metallogeny of the Tasman Fold Belt System. Economic Geology, 90,1382-1401. Suppel D.W., Bames R.G. and Scheibner E. in press. The Palaeozoic in New South Wales - geology and mineral resources. AGSO Journal of Australian Geology and Geophysics.

2. 3. 4.

8. 9. 10. 40

11. 12.

13. 14. 15. 16.

17.

18.

19.

20.

21. 22. 23. 24.

25.

26. 27.

28. 29. 30.

31. 32.

33.

Woodall R. 1979. Gold - Australia and the world. In: Glover J.E. and Groves D.L (eds.) Gold mineralization. Geology Department and Extension Service, University of Westem Australia, Perth, pp. 1-34. Kleeman J.D. 1990. Tin-tungsten in New England. In: Glasson K.R. and Rattigan J.H. (eds.) Geological aspects of the discovery of some important mineral deposits in Australia. Australasian Institute of Mining and Metallurgy, Melbourne, pp. 243-249. Ashley RM., Cook N.D.J., Hill, R.L. and Kent A.J.R. 1994. Shoshonitic lamprophyre dykes and their relation to mesothermal Au-Sb veins at Hillgrove, New South Wales, AustraHa. Lithos, 32, 249-272. Brooker M. and Jaireth S. 1995. Mount Rawdon, southeast Queensland, Austraha - a diatreme-hosted gold-silver deposit. Economic Geology, 90,1799-1817. Balme, B.E. and Foster C.B. 1996. Triassic. In: Young G.C. and Laurie J.R. (eds.) An Australian Phanerozoic timescale, pp. 136-147. Oxford University Press, Melbourne. Collins W.J., Offler R., Farrell T.R. and Landenberger B. 1993. A revised late Palaeozoic-early Mesozoic tectonic history for the southem New England Fold Belt. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Australia. Department of Geology and Geophysics, University of New England, Armidale, pp. 69-84. Gust D.A., Stephens C.J. and Grenfell A.T. 1993. Granitoids of the northem NEO: their distribution in time and space and their tectonic implications. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Australia. Department of Geology and Geophysics, University of New England, Armidale, pp. 565-571. Veevers J.J., Conaghan P.J., Powell C.McA., Cowan E.J., McDonnell K.L. and Shaw S.E. 1994. Eastem Australia. In: Veevers J.J. and Powell C.McA. (eds.) Permian-Triassic Pangean basins and foldbelts along the Panthalassan margin of Gondwanaland. Geological Society of America Memoir, 184,11-171. Holcombe R.J., Stephens C.J., Fielding C.R., Gust D., Little T.A., Sliwa R., Kassan J., McPhie J. and Ewart A. in press. Tectonic evolution of the northem New England Fold Beh: the Permian-Triassic Hunter-Bowen event. AustraUan Joumal of Earth Sciences. Shaw S.E. and Flood RJi. 1993. A compilation of late Permian and Triassic biotite Rb-Sr data from the New England Bathohth and areas to the southeast. In: Carr PF. Centre for Isotope Studies Research Report 1991-92. CSIRO, North Ryde, pp. 151-155. Chappell B.W. 1978. Granitoids from the Moonbi district. New England Batholith, eastem Australia. Joumal of the Geological Society of Australia, 25,267-283. Chappell B.W. 1994. Lachlan and New England: fold belts of contrasting magmatic and tectonic development. Joumal and Proceedings Royal Society of New South Wales, 127,47-59. Shaw S.E. and Flood R.H. 1981. The New England Batholith, eastem Austraha: geochemical variations in time and space. Joumal of Geophysical Research, B86,10530-10544. Bryant C.J. and Arculus R.J. 1993. Geochemistry of the Clarence River Suite granitoids. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Australia. Department of Geology and Geophysics, University of New England, Armidale, pp. 349-352. Bryant C.J., Cosca M.A. and Arculus R.J. in press. 40Ar/39Ar ages of Clarence River Supersuite intmsions from the northem portion of the New England Batholith, southem New England Orogen. Australian Joumal of Earth Sciences. Kent A.J.R. 1994. Geochronology and geochemistry of Palaeozoic intrusive rocks in the Rockvale region, southem New England Orogen, New South Wales. Australian Joumal of Earth Sciences, 41, 365-379. Blevin P.L. and Chappell B.W. 1993. The influence of fractionation and magma redox on the distribution of mineralisation associated with the New England Batholith. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Australia. Department of Geology and Geophysics, University of New England, Armidale, pp. 423429. Landenberger B. and Collins W.J. 1996. Derivation of I-type granites from a dehydrated chamockitic lower crust: evidence from the Chaelundi Complex, eastem Austraha. Joumal of Petrology, 37,145-170. Day R.W., Whitaker W.G., Murray C.G., Wilson LH. and Grimes K.G. 1983. Queensland geology. Geological Survey of Queensland Publication, 383. Roach A. 1993. Felsic volcanism in the Ipswich Basin. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Austraha. Department of Geology and Geophysics, University of New England, Armidale, pp. 643-648. Cranfield L.C. 1994. Maryborough 1:250 000 geological series explanatory notes. Geological Survey of Queensland, Brisbane. Stephens C.J. 1992. The Mungore Cauldron and Gayndah Centre: large scale silicic volcanism in the New England Fold Belt near Gayndah, southeast Queensland. PhD thesis. University of Queensland, Brisbane (unpublished). Stephens C.J., Schsn R.W. and Ewart A. 1993. Mesozoic cmstal extension in the northem New England Orogen: geochemical and isotopic evidence from large scale silicic magmatism. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Australia. Department of Geology and Geophysics, University of New England, Armidale, pp. 637-642.

34.

Wilson M.M. and Mathison CI. 1968. The Eulogie Park Gabbro, a layered basic intrusion from eastern Queensland. Journal of the Geological Society of Australia, 15,139-158. Whitaker W.G., Murphy P.R. and Rollason R. 1974. Geology of the Mundubbera 1:250 000 sheet area. Geological Survey of Queensland Report, 84. 36. Ambler, E.P. and Ashley, RM. 1977. Vermicular orthopyroxene-magnetite symplectites from the Wateranga layered mafic intrusion, Queensland, Australia. Lithos, 10,163-172. 37. Reeves S.J. and Keays R.R. 1995. The platinum-group element geochemistry of the Bucknalla Layered Complex, central Queensland. Australian Joumal of Earth Sciences, 42,187-201. 38. Brooks J.H., Syvret J.N. and Sawers J.D. 1974. Mineral resources of the Kilkivan district. Geological Survey of Queensland Report, 60. 39. Weber C.R. 1974. Woolomin-Texas Block, plutonic rocks and intruded sediments. In: Markham N.L. and Basden H. (eds.) The mineral deposits of New South Wales. Geological Survey of New South Wales, Sydney, pp. 350391. 40. Plimer I.R. and Kleeman J.D. 1985. Mineralization associated with the Mole Granite, Australia. In: H i ^ heat production (HHP) granites. Institution of Mining and Metallurgy, London, pp. 563-570. 41. Kleeman J.D., Plimer LR., Lu J., Foster DA. and Davidson R. in press. Timing of thermal and mineralization events with the Mole Granite, Austraha. Australian Joumal of Earth Sciences. 42. Weber C.R., Paterson I.B.L. and Townsend D.J. 1978. Molybdenum in New South Wales. Geological Survey of New Mineral 43. deposits of the Inverell 1:100 000 sheet. Geological Survey of New 43. BrownSouth R.E.Wales, and Stroud W.J.Resources, 1993. Mineral South Wales Quarterly Notes, 91,11-33. 44. Einaudi M.T., Meinert L.D. and Newberry R.J. 1981. Skam deposits. Economic Geology 75th Anniversary Volume, 317-391. 45. Meinert L.D. 1992. Skams and skam deposits. Geoscience Canada, 19,145-162. 46. White N.C. and Hedenquist J.W. 1990. Epithermal environments and styles of mineralization: variations and their causes, and guidelines for exploration. Joumal of Geochemical Exploration, 36,445-474. 47. Bottomer L.R. 1986. Epithermal silver-gold minerahzation in the Drake area, northeastem NSW. Austrahan Joumal of Earth Sciences, 33,457-473. 48. Perkins C. 1987. The Red Rock deposit: a late Permian submarine epithermal precious metal system in northeastem New South Wales. Proceedings of Pacific Rim Congress 87, pp. 895-898. Australasian Institute of Mining and Metallurgy, Melbourne. 49. Perkins C. 1988. The Red Rock deposit: late Permian submarine epithermal Ag-Au mineralization in the Drake Volcanics. In: Kleeman J.D. (ed.) New England Orogen tectonics and metallogenesis. Department of Geology and Geophysics, University of New England, Armidale, pp. 275-289. 50. Smith S.G. 1989. Geology and geochemistry of Permian epithermal mineral deposits at Drake. BSc Honours thesis. University of New England, Armidale (unpubhshed). 51. Houston M.J. 1993. The geology and mineralisation of the Drake mine area, northern NSW, Australia. In: Rood RG. and Aitchison J.C. (eds.) New England Orogen, eastern Australia. Department of Geology and Geophysics, University of New England, Armidale, pp. 337-348. 52. Ashley P.M. and Andrew A.S. 1992. The Mt Ninderry acid sulphate alteration zone and its relation to epithermal mineralization in the North Arm Volcanics, southeast Queensland. Australian Joumal of Earth Sciences, 39, 79-98. 53. Henley H.F. 1991. Lamprophyres and gold in the New England Fold Belt. Geological Survey of New South Wales Quarterly Notes, 83, 7-24. 54. Ashley P.M. and Brownlow J.W. 1993. Silica-carbonate alteration zones in the Great Serpentinite Belt, southern New England Orogen: their nature and significance. In: Flood P.G. and Aitchison J.C. (eds.) New England Orogen, eastem Austraha. Department of Geology and Geophysics, University of New England, Armidale, pp. 197-214. 55. Eadington P.J. 1982. A brief survey of fluid inclusions and their significance in the base metal ores at the Conrad lodes and Webb's Consols deposit. In: Flood RG. and Runnegar B. (eds.) New England geology. Department of Geology, University of New England and AHV Club, Armidale, pp. 269-276. 56. Eadington P.J. 1983. A fluid inclusion investigation of ore formation in a tin-mineralized granite. New England, New South Wales. Economic Geology, 78,1204-1221. 57. Sun S.-S. and Eadington P.J. 1987. Oxygen isotope evidence for the mixing of magmatic and meteoric waters during tin mineralization in the Mole Granite, New South Wales, Austraha. Economic Geology 82,43-52. 58. Plimer LR., Lu J., Foster D. and Kleeman J.D. 1995. Ar-Ar dating of multiphase mineralisation associated with the Mole Granite, Australia. In: Pasava, J., Kr'bek, B. and ZJk, K. (eds) Mineral deposits: from their origin to their environmental impacts. Proceedings of the Third Biennial Meeting of the Society for Geology Applied to Mineral Deposits, Prague, pp. 497-500. 59. Golding S.D., Wilson A.E, Scott M., Anderson P.K., Waring C.L., Flitcroft M. and Rypkema HA. 1987. Isotopic evidence for the diverse origins of gold mineralization in Queensland. Papers of the Department of Geology, University of Queensland, 12, 65-83.

35.

SUB-BASALTIC ALTERATION WITHIN DIAMOND-BEARING TERTIARY DEEP LEADS, COPETON BINGARA, NSW L. M. Barron^ and B. J Barron^ 1 New South Wales Department of Mineral Resources, 29-57 Christie Street, R 0. Box 536, St Leonards, NSW 2065, Australia. 2 Consulting Petrologist, 7 Fairview Avenue, St Ives NSW 2075, Australia. Summary The Copeton-Bingara Tertiary deep leads produced up to 500,000 carats of diamonds, and considering that the target horizons are predominantly clayey, the hundred year old adits and extensive but mostly unsupported mine workings have survived remarkably well. Field evidence shows that clay alteration occurred rapidly after capping by basalt, at constant bulk volume. The sub-basaltic alteration process is driven by heat from the capping basalt, and acts on both the deep lead materials (producing oxidised clays) and the capping basalt (reduced clays). The clays may have enhanced commercial value due to the process of formation. Introduction The Copeton-Bingara Tertiary deep leads have been mined over the last 100 years with the production of up to two million diamonds averaging about 0.25 carats. These diamonds are unusual in many ways. They occur in a Phanerozoic setting more than 1500 km from the nearest craton. Each deposit exhibits a strong tribal aspect in terms of the diamond characteristics while a high percentage of the diamonds are gem quality (MacNevin 1). The diamonds show high variable nitrogen contents (Meyer et. al. 2), a lack of abrasion (MacNevin 1), and extreme resistant to high temperature ablation (Joris 3). They have dominantly rounded growth crystal shapes with abundant large euhedral indentations (Barron et. al. 4) and growth twins. They contain very unusual inclusions of coesite and grossular and are composed of isotopically heavy carbon (Sobolev, 5; Meyer et. al. 2). The subduction diamond model (Barron et. al., 4) was developed to explain the presence of these and other diamonds in the Phanerozoic of eastem Australia. During subduction, if the slab is cool enough and subducts at an optimal rate, diamonds will form at about 100km depth within a low temperature window in the upper portion of the descending slab. Termination of subduction traps the diamond-bearing portion of the slab within the base of the lithosphere. The diamonds are sampled by obduction or by later intrusions of basanite-nephelinite which form diatremes at the surface, and diamonds are shed into the drainage through erosion. The most productive diamond-forming event in eastem Australia appears to be the Late Carboniferous termination of subduction which affected New England. Sub-Basaltic Clay Alteration The adits and extensive mine workings in the Copeton-Bingara deep leads are mostly unsupported, but have survived remarkably well, especially considering that the deep lead materials are now composed of minor quartz and dominant clay. The study of this intense clay-alteration process involved on site investigations, thin section examination, coupled with interpretation of photographs of the mine workings (Barron, 6). Deposition in the deep leads was through a range of competing weathering, alluvial/mechanical and pyroclastic/volcaniclastic processes. Source materials comprised quartzite, basalt and dolerite (Bingara, Copeton), granite (Copeton), basement lithics and quartz sand and grit. Small scale Tertiary(?) growth faulting was involved at many localities, while intrusive pyroclastic activity combined with contemporaneous alluvial activity at Craddocks and Monte Cristo. The mdne workings reveal that the across-channel topography ranged from gentle (Copeton) to steep (Bingara), but the packing of clasts is exceptionally tight in both settings, reflecting high energy deposition. Only a few mine workings (e.g. Wonderland) show low energy depositional settings, but such environments are well represented in drill core from both Copeton and Bingara as laterally extensive intersections of peaty and lignitic horizons intermixed with silts and muds/clays. There is no sign of plastic flattening at the base of large (2-4m) boulders of clay-altered granite embedded in the Copeton deep leads. Thus their alteration post-dated the actual capping by basalt, but must have occurred in one rapid stage as part of a broader process that operated on the whole of the deep lead material and the capping basalt itself There is remarkable preservation of large, medium and small scale undeformed textures, including those that result from primary and secondary igneous and sedimentary processes. This preservation indicates that the clay-alteration is a constant volume process overall. Theoretical conversion of the reactants to the end-stage products results in a strong theorectical volume contrast. This theoretical contrast involves up to a 50% shrinkage for derivation from granitic material, but less (10-20%) from basaltic material. Furthermore, the theoretical negative bulk volume contrast should be increased because of local passive clay-infilling of interclast porosity. In the deep lead materials, the major components of Mg, Ca, Fe, and alkalies are preferentially removed by a strong 'metasomatic' movement of soluble major components liberated during the alteration. The new clay materials have a density significantly lower than clay mineral densities, so that some of the theoretical volume reduction is taken up by production of clay with a significant micro-porosity.

The complete clay-alteration of deep lead material is not due to subsurface weathering, but is due to interaction between the capping basalt, a wet paleodrainage, a trapped permeable alluvial deposit, and thermal reactivation of the groundwater moving through the deep lead, all resulting in a process called sub-basaltic alteration. The process is driven by the heat provided by the capping basalt as it solidifies, which forces convection of groundwater within the deep lead mass. The net result is that a white kaolinitic claystone replaces the deep lead materials (including metasediments, acid and basic rock types). An analogous process also alters the capping basalt, but produces black to green swelling claystones. Thus the intense clay alteration occurs under oxidising conditions within the deep lead materials, but under reducing conditions within the basalt lavas and dykes, reflecting the oxidation state of the original bulk materials. The clay materials produced in the deep leads by sub-basaltic alteration are expected to have enhanced mechanical and ceramic properties because they are formed at a low confining pressure and moderately elevated temperatures. There are significant parallels between this sub-basaltic alteration process and the texture-preserving intense clay alteration that is typical of basaltic diatremes and their basalt-capped widespread but thin volcaniclastic apron deposits, for instance in New England. These clay materials are usually dark khaki-coloured with high iron levels reflecting overprinting of altered basaltic material by fine grained iron carbonates and smectites. Within the apron deposits, whitecoloured clayey deposits may also be present (e.g. at Elsmore and Kings Plains). This white colour may indicate a combination of some sort of intermediate alluvial processing (pre-capping) plus simple oxidation/removal of iron in the basalt-capped groundwater environment. There are many localities where new basaltic plugs and overriding lavas have converted these clay-altered materials to fine grained emery. Investigations show that this emery has perfect preservation of the primary relict textures of bedding, fi-agment outlines, phenocrysts, vesicles and microlites, and the secondary relict features of differential iron content derived by oxidation of differential Fe-carbonate abundances (both primary and secondary). This homfelsing effect does not seem to occur in the deep lead materials, suggesting that the clay-altered materials have to be completely dry before the homfelsing may occur. The presence of emery cobbles (with relict basaltic breccia textures) in a stream probably indicates the presence of an upstream diatreme. Conclusions Over the last one hundred years a significant number of diamonds have been mined from extensive workings in the Copeton-Bingara deep leads which are capped by Tertiary alkali basalt. Adits and largely unsupported mine workings survive in remarkably good condition, despite the dominantly clayey nature of the target horizons. These materials show complete preservation of undeformed relict textures at all scales, indicating that the clay alteration process was a rapid constant volume process that occurred after capping by basalt. This bulk volume contraint contradicts material balance arguments which indicate that clay alteration should proceed with a dramatic loss in volume. Some of the theoretical volume loss is taken up by formation of a micro-porosity which reduces the density of the clayey materials significantly below that of the ideal clay mineral. The sub-basaltic clay-alteration process occurs at low confining pressure, moderate temperature, with a high water to rock ratio due to groundwater migration and reactivation. The process produces clay with a high micro-porosity, h i ^ crystallinity and coarse uniform grain size. Acknowledgements It is a pleasure to acknowledge the input of co-workers of the New South Wales Geological Survey. This includes discussions with Steve Lishmund and Geoff Oakes which were formative. This investigation would not have been possible without the assistance of David Bames who took high quality photographs within the deep lead workings, under the direction of the lead author and Bob Brown. References 1 MacNevin A. A. Diamonds In New South Wales. Geological Survey ofiVew South Wales Mineral Resources 42. 2 Meyer H. O. A., Milledge H. J. & Sutherland F. L. Unusual diamonds and unique inclusions from New South Wales, Australia. Abstracts 6th Intemational Kimberlite Conference, Novo Sibirsk (1995), 379-381. 3 Joris A. A Destiny in Diamonds. Boolarong Publications, Brisbane 1986. 4 Barron L. M.,Lishmund S. R., Oakes G. M., Barron B. J., & Sutherland F. L. Subduction model for the origin of some diamonds in the Phanerozoic of eastem New South Wales. Australian Journal ofEarth Sciences (1996) 43, 257-267. 5 Sobolev N. V. Crystalline inclusions in diamondsfromNew South Wales, Australia. In Glover J. E. and Harris P. G. eds. Kimberlite occurrences and origin: A basis for conceptual models in exploration. Geology Department and University extension. University of Westem Australia, Publication(1984) 8,213-226. 6 Barron L.M. Interpretation of photographs taken within Copeton Bingara deep lead diamond mines. Unpublished report of the Geological Survey of New South Wales (1994), GS1994/102.

METHANE SORPTION CAPACITY OF UPPER CRETACEOUS COALS FROM THE GREYMOUTH COALFIELD, NEW ZEALAND B. Basil Beamish', Peter J. Crosdale' and Tim A. Moore' 1. Department of Geology, The University ofAuckland, Private Bag 92019, Auckland, New Zealand 2. Coalseam Gas Research Institute, Department of Earth Sciences, James Cook University of North Queensland, Townsville, Queensland, 4811, Australia 3. Coal Research Limited, PO Box 31 244, Lower Hutt, New Zealand Summary Coal seams in the Greymouth Coalfield are gassy, but little is known about their gas sorption properties. Preliminary results of a current research programme are presented, which provide some answers to the role of rank, moisture and coal type in methane sorption capacity of these coals. Implications of these effects are also discussed in relation to mining, outburst potential and methane recovery. Introduction Greymouth Coalfield is situated on the west coast of the South Island of New Zealand (Figure 1). Economic coal seams are contained in the Upper Cretaceous Paparoa Coal Measures (Barry et al. (1)) and are mined by both opencut and underground methods. Several major explosions initiated by ignitions of methane have taken place in the Greymouth Coalfield: Brunner (1896); Dobson (1926); Kaye's (1940); and Strongman (1967). High gas emissions are also reported in mines department records (Patterson et al. (2)). Under certain combined conditions of high gas content gradient, geologically disturbed areas and lower coal strengths, the phenomenon of instantaneous outbursts of coal and gas may occur. These have not been experienced in New Zealand presumably due to the slower face advance rates compared to the Bowen or Sydney Basins of Austraha. Interest is now being shown in recovering the methane firom the Greymouth Coalfield (Cave and Newman(3)) and as several mdnes are now looking at accessing deeper, gassier coal resources there is a need to understand the fiindamentals of gas sorption behaviour of New Zealand coals. One of the key parameters needed to assess gas behaviour of coal is the sorption isotherm. This is obtained under controlled laboratory conditions by measuring the amount of gas adsorbed by the coal at given pressures and a fixed temperature. Effects of coal rank, moisture and ash content on sorption capacity are well documented (Schwarzer and Byrer(4)), although these authors highUght a general misconception of sorption capacity increasing with rank per se. Opinions on the effects of coal type on sorption capacity are divided. Ettinger et al. (5), foundfiisinite-richcoal to sorb more methane than vitrinite-rich coal, while Beamish et al.(6) and Lamberson and Bustin (7) have observed the opposite. Maceral composition has also been observed to have little effect (Faiz et al. (8)). As no published sorption data exists for New Zealand coals, a collaborative study of coals from the Greymouth Coalfield, between the Department of Geology at The University of Auckland, the Coalseam Gas Research Institute and Coal Research Limited, is in progress to address this deficiency. Methane sorption testing has been performed using a high pressure microbalance. The results from this new study will enable these coals to be placed into perspective with respect to gas emission and outburst hazard potential. In addition the information will enable informed decisions to be made on gas recovery options such as vertical well degasification and in-seam drilling and drainage.

Greymouth Coalfield

100 STEWART ISLAND

Experimental

200

Figure !• Location of the Upper Cretaceous Greymouth Coalfield, New Zealand

Coal Samples and Analytical Data

Two sets of coal samples from the Coal Research Limited sample bank have been used for analysis. The first set of coals cover a rank range from high volatile bituminous C to low volatile bituminous (Table 1). A second setfromMoody Creek Mine and the Rapahoe sector has been analysed to Ic&k at coal type effects. The second suite was designed to be isorank, however, the Rapahoe sector samples are of sHghtly lower rank and are borderline hvC / hvB coals. Proximate analysis (Table 1) has been performed using a thermogravimetric technique developed at The University of Auckland (Beamish (9)). Coal types have been expressed in temis of banding variation. Banding was quantified on core and mine faces by point co unting vitrain bands, which are preserved pieces of wood. Non-vitrain band areas (matrix) are comprised of broken down plant pieces, including small fragments and humic gels, plus spores/pollen, resins, fungal remains, waxes and charcoal (i.e. inertinite). Methane Sorption Testing

Sorption isotherms are generated under controlled laboratory conditions by measuring the amount of gas adsorbed by the coal at given pressures and afixedtemperature. They have four primary uses:

Table 1 Analytical and sorption data for Greymouth Coalfield samples Colliery/ Langmuir Rank Sample Proximate analysis Sector volxime Moisture Ash Volatile (cc/g, daf) matter (%,ar) (%,db) (%, daf) 35.42 Kiwi 5.7 8.8 44.9 54/920 4.4 54/921 Strongman 40.6 32.26 2.8 Roa 1.0 8.8 21.0 21.13 54/923 Rapahoe 1.2 46.7 25.50 752-5 6.0 38.12 55/382 Rapahoe 5.9 1.3 43.0 3.4 39.57 55/381 Rapahoe 6.6 40.3 28.17 Moody Creek 3.3 2.1 38.9 55/508 1.4 37.2 35.95 752-8 Rapahoe 7.8 1. 2. 3. 4.

Type

hvCb hvBb Ivb hvBb hvBb hvBb hvBb hvBb

Bright Non-banded Bright <25% vitrain Bright 25-30% vitrain Bright >30% vitrain Pure vitrain

To describe the gas sorption capacity of coal as a function of pressure; To estimate the gas content of a coal in-situ; To determine the pressure at which gas will be released from a coal seam; and To calculate the likely amount of gas which can be recovered based on the difference in seam gas pressure and the final completion pressure of desorption.

To assist in computer modelling of gas flow from coal seams it is normal practice to describe the isotherms by an equation. The most common form of isotherm equation used to describe coal data is the Langmuir isotherm (Langmuir (10)). It has the general form : VlP V=PL + P

(1)

= volume adsorbed V P = absolute gas pressure in atmospheres = maximum sorptive capacity of coal (Langmuir volume) VI = pressure at half the Langmuir volume Pl To test the validity of a Langmuir fit, PA^ is plotted against P, which should yield a straight line. where,

Methane sorption testing was performed using a microgravimetric technique (Crosdale and Beamish (11); Levine et al. (12)). Dry coal isotherms were obtained for all samples using -212|im particle size at 0.5-2.0 MPa pressure steps up to a maximum pressure of between 7 and 10 MPa. Moist coal isotherms were obtained for the Kiwi and Strongman coals for comparison of moisture effects on sorption capacity. Methane Content Calculation

Weigiit recorded by the microbalance must be corrected for buoyancy effects. A buoyant force applied to all balance components, including the sample, equals the weight of gas displaced. The weight of displaced gas is estimated from the volume of the balance, including the sample, and the gas density at each pressure, derived from the real gas equation. Volume of the microbalance components, including weighing arms, sample pans and counterweights, was experimentally evaluated. Initial sample volume was calculated from its helium density as determined in the microbalance prior to methane adsorption. Knowing the density of sorbed methane on coal (= 618.9g/l; van der Sommen et al. (13)), the weight of sorbed methane can be calculated (Crosdale (14); Levine et al. (12)): Ws = (Wmb - PgCVmb - Vc))/(1 - Pg/Ps) where,

(2)

Wg = weight of sorbed methane Wjnb = microbalance output weight Vjnb ~ microbalance volume V,. = volume of coal 'c = gas density Pg Pg = sorbate density

Results and Discussion Coal Rank Effect on Sorption Capacity

Adsorption isotherms for the three run of mine samples are contained in Figure 2. At low pressures ( 0 - 3 MPa) the methane content of the Kiwi and Strongman coals are similar, but the Roa coal has a slightly higher methane content in this range. At pressures above 3 MPa, the methane content of the three coals rapidly diverges. The Langmuir volumes foi 47

these three coals (Table 1) displays a rank progression of hvCb>hvBb>lvb, indicating that the lower rank coal has a higher internal surface area available for sorption in a dry coal state. Adsorption isotherm tests on other New Zealand coals (Crosdale et al (15)) suggest that the maximum sorption capacity passes through a minimum in the medium volatile bituminous ranL These findings are consistent with those of Moffat and Weale (16) and Botham (17). Moisture Effect on Sorption Capacity The presence of moisture in coal reduces its methane sorption capacity, as the moisture competes with methane for sorption sites. This is illustrated by moist versus dry coal isotherms for the Kiwi and Strongman samples (Figure 3). The effect is more noticeable for the Kiwi sample which was run at a moisture content of 5.3%, compared with 4.4% for the Strongman sample. A rank reversal is now apparent in the methane content of the two coals under moist conditions. Consequently, in lower rank coals which contain more moisture than high rank coals, the methane sorption capacity would be less. It is this feature of the coals which often leads to the generaUsation of increasing sorption capacity with rank. A major implication from the shape of the moist coal isotherms is that for a given methane content, lower rank coals will have higher seam gas pressures. Beamish and Vance (18), found this to be the case in a subbituminous coal seam which had a maximum measured gas content of 2.2 m3/t and a measured seam gas pressure of 0.5 MPa. Maximum gas contents of 4-6 m3/t have been quoted for Greymouth coals (Vance and Cave (19)), which according to the moist coal isotherms presented here would equate to seam gas pressures close to 1 MPa. Australian coal mines have experienced outbursts under these seam gas pressure conditions. METHANE ADSORPTION ISOTHERMS

Pressure (MPa) »Kiwi

••Strongman -•-Roa

Figure 2. Methane adsorption isotherms of dry coals from Greymouth coalfield collieries

METHANE ADSORPTION ISOTHERMS

4 6 Pressure (MPa) s Kiwi (0% moist)

^ Strongman (0% moist)

* Kiwi (5.3% moist)

• Strongman (4.4% moist)

Figure 3. Effect of moisture content on methane absorption isotherms from Kiwi and Strongman Collieries Coal Type Effect on Sorption Capacity There is a pronounced trend in the analytical data of the coal types used in this study (Figure 4). As the vitrain band content of the coal types increases, the volatile matter content decreases. However, the Moody Creek sample is of sHghtly higher rank than other samples and is from a different seam. Distinct differences exist in the methane sorption capacity of these coals (Figure 5). Langmuir volumes rangefrom25.50 to 39.57 cc/g, indicating the large gas content variability that can exist in these seams. For the Rapahoe sector samples, a simple trend would be expected of increasing methane content with increasing vitrain content. As expected, the nonbanded coal type has the lowest sorption capacity. No clear distinction is observed between the banded types. The amount of structured vitrinite in the coals is only indicated by the lithotype analysis and type-related trends will be clarified as the micropetrographic analyses become available. The Moody Creek sample (55/508) is comparable in sorption characteristics to the bright, non-banded coal from the Rapahoe Sector. Deviationfromthe expected intermediate isothemi is attributed to its slightly hi^er rank and inter-seam variability. An interesting feature of the Rapahoe isotherms is that a composite of the coal types would produce an isotherm between the Kiwi and Strongman run of mine coal, which is not unreasonable as the samples comefroman area adjacent to these mines. The variation observed in sorption capacity with coal type suggests gas content will vary through the seam profile. This has imphcations for in-seam drilling and drainage, whereby uniform drainage may not occur if gassier parts of the seam are not intersected. Similarly, for targeting seams in vertical well degasification programmes the results suggest that seams high in bright non-banded coal types would be poor targets as they would have less contained gas. Conclusions and Recommendations 1. The Langmuir volume or maximum methane sorption capacity for coalsfromthe Greymouth Coalfield decreases with rank as follows: hvCb>hvBb>lvb. Sorption testing of other New Zealand coals suggest that a minimum occurs in medium volatile bituminous coals. 2. Methane sorption capacity rapidly decreases with moisture content. Hence low rank coals with high moisture contents have a lower sorption capacity than high rank coals with low moisture contents. This is due to the competing effect between moisture and methane for sorption sites on the intemal surfaces of the coal pores. At a sorption pressure of approximately 7 MPa the methane content of a high volatile bituminous C coal sample is 24.47 cc/g in a dty state and 17.34 cc/g at a moisture content of 5.3%. At a similar sorption pressure, the methane content of a high volatile bituminous B coal is 23.29 cc/g in a dry state and 21.35 cc/g at a moisture content of 4.4%. The level of moisture reduction at different pressures is not uniform. 49

3. Coal type has a significant impact on the methane sorption capacity of Greymouth coals. Bright non-banded coal types have i e lowest methane sorption capacity, and bri^t coals with 25-30% vitrain bands have the greatest methane sorption capacity. At pressures of 2 MPa the difference between the two coal types can be as much as 5 cc/g. 4. The differences in methane sorption capacity of the coal types is attributable to differences in their physical nature (pore sizes and distribution). Pore studies of coals from the Bowen Basin have identified major effects of maceral composition. Future work on Greymouth coal types will look at the maceral composition of the coals and lump samples need to be tested to assess the sorption rate behaviour of the coals as this will be important for assessing gas emission problems in underground workings. Acknowledgments The authors would like to thank Greymouth Coal Limited and Moody Creek Coal Mining Co. Ltd. for access to coal cores and mine samples used for coal type analysis. 48

• I. 752-5

55/382

55/381

55/508

752-8

Figure 4. Variation in volatile matter related to coal type, rapahoe sector and Moody Creek. Rapahoe sector samples are isorank (borderline hvCVhvB) while the Moody Creek sanple (55/508) is slightly higher in rank (hvB). METHANE ADSORPTION ISOTHERMS

752-5

4 6 Pressure (MPa)

-^55/382 ^55/381 • 5 5 / 5 0 8 * 752-8

Figure 5. Methane absorption isotherms of dry coalsfromthe Rapahoe sector and Moody Creek. Differences are in part attributed to variations in coal type. Rapahoe sector samples are isorank (borderline hvC/hvB) while the Moody Creek sample (55/508) is sUghtly higher in rank (hvB).

References 1. 2.

5. 6.

7. 8.

9. 10. 11.

12.

13. 14. 15. 16. 17. 18. 19.

Barry, J.M., Duff, S.W. and MacFarlan, D.A.B, 1994. Coal Resources of New Zealand. Resource Information Report 16. Energy and Resources Division, Ministry of Commerce, New Zealand, 73p. Patterson, J.K., Elliott, W., Eyeington, S.R., McGhie, T. and Prendiville, A.V., 1967. Report of commission to inquire into disaster at Strongmian Mine. Appendix to the joumals of the House of Representatives of New Zealand C-4. Cave, M. and Newman, J., 1995. Vitrinite reflectance variations between Brunner and Paparoa coal measure sequences in the Greymouth Coalfield and the implications for petroleum source rock maturation; some initial results, in Proceedings of the Sixth New Zealand Coal Conference, Wellington, New Zealand, October 1995,1, 3-14. Schwarzer, R.R. and Byrer, C.W., 1983. Variation in the quantity of methane adsorbed by selected coals as a function of coal petrology and chemistry. US Department of Energy Report, Contract No. DE-AC2180MC14219. Ettinger, I., Eremin, L, Zimakov, B. and Yanovskaya, M., 1966. Natural factors influencing coal sorption properties - 1 - Petrography and the sorption properties of coals. Fuel, 45,267-275. Beamish, B.B., Crosdale, P. J. and Gamson, RD. 1993 Chracterising the methane sorption behaviour of banded coals in the Bowen Basin, Australia. In Proc. 1993 Intemat. Coalbed Methane Symp. The University of Alabama, Tuscaloosa, 145-150. Lamberson, M.N. and Bustin, RM., 1993. Coalbed methane characteristics of Gates Formation coals, Northeastem British Columbia: Effect of maceral composition. AAPG Bulletin, 77, 12, 2062-2076. Faiz, M.M., Aziz, N.I., Hutton, A.C. and Jones, B.C., 1992. Porosity and gas sorption capacity of some eastem Australian coals in relation to coal rank and composition, in Proceedings of Symposium on Coalbed Methane Research and Development in Australia, Townsville, Queensland, November 1992, 4, 9-20. Beamish, 1994. Proximate analysis of New Zealand and Australian coals by thermogravimetry. New Zealand Journal of Geology and Geophysics, 37, 387-392,1994. Langmuir, 1.1918. Adsorption of gases on glass, mica and platinum. 1 Am, Chem. Soc., 40, 1361. Crosdale, P. and Beamish, B. 1993. Maceral effects on methane sorption by coal. In Beeston, J.W. (Ed) New Developments in Coal Geology : A Symposium. Coal Geology Group, Geological Society of Australia, Brisbane, 95-98. Levine, J.R., Johnson, P.W. and Beamish, B.B, 1993. High pressure microbalance sorption studies, in Proceedings of the 1993 International Coalbed Methane Symposium, The University of Alabama/Tuscaloosa, Birmin^am, Alabama, USA, May 1993,187-196. Sommen, J. van der, Zwietering, P., Eillebracht, B.J.M. and Krevelen, D.W. van, 1955. Chemical structure and properties of coal. 12. Sorption capacity for methane. Fuel, 34,444-448. Crosdale, P.J. 1993. High Pressure Microbalance Analysis. Coalseam Gas Research Institute, Technical Report, CGRI TR93/4b. Department of Geology, James Cook University, Townsville. 20pp. Crosdale, P.J., Beamish, B.B. and Valix, M. under review. Methane sorption related to coal composition. International Journal of Coal Geology, Moffat, D.H. and Weale, K.E., 1955. Sorption by coal of methane at higji pressures. Fuel, 34,449^62. Botham, J.C., 1958. Association of gases with coal - Permeability tests. TM 62/58 CG, Mines Branch, Ottawa, Canada. Beamish, B.B. and Vance, W.E., 1990. Gas research at Huntly West Mine, New Zealand. Department of Mining Engineering, University of Auckland, New Zealand, November 1990, 32p. Vance, W.E. and Cave, M.P., 1992. Coalbed gas in New Zealand: A potential energy source, in Proceedings of Symposium on Coalbed Methane Research and Development in Australia, Townsville, Australia, 1, November 1992,31-43.

COMBUSTION BEHAVIOUR OF NEW ZEALAND, LATE CRETACEOUS COALS DETERMINED BY THERMOGRAVIMETRY.

Kathy E. Benfell*, B. Basil Beamish and KA. Rodgers Department of Geology, The University ofAuckland, Private Bag 92019, Auckland, New Zealand * Address for correspondence: Department of Geology, The University ofNewcastle, Callaghan, Newcastle, NSW 2308, Australia. Twenty New Zealand South Island Late Cretaceous coals were subjected to thermogravimetric analysis (TG) and derivative thermogravimetric analysis (DTG) to evaluate the applicability of these techniques to these coals. The key characteristics proved to be: the maximum rate of combustion (R^) where the rate of weight loss of the sample is greatest, the peak temperature (Tg) at which this rate occurs, and the temperature of char bumout (Tg). Generally, New Zealand lignites and sub-bituminous coals have lower bumout temperatures (420.7-545.1°C) than bituminous coals (543.8-646.6°C). There is a sizeable variation in Tg (420.7°C-636.5°C) in coals with volatile matter contents above 40%, whereas a coal with a lower volatile content has a Tg value of 646.6°C. The temperature of char bumout gives a better indication of combustion efficiency than rank or volatile matter content alone, providing industrial users with a clear indication of a coal's bumout performance and its suitability for a particular end usage. INTRODUCTION New Zealand has large coal reserves well suited as an industrial energy resource. However, traditional coal classification methods and analytical techniques do not ideally reflect the thermal behaviour of New Zealand coals as shown by Gunn et al (1). An altemative procedure for analysing coal combustion behaviour is provided by thermogravimetric (TG) and derivative thermogravimetric (DTG) procedures (e.g. Wagoner and Winegartner (2)). Wame (3) demonstrated that thermogravimetry provides an efficient and cost-effective way of monitoring the effects on combustion of coal rank, maceral composition and physical properties such as surface area, and Carpenter and Skompska (4) stressed the versatility of the method in rating coals and predicting their combustion behaviour. Nonetheless New Zealand Late Cretaceous coals typically have volatile matters greater than any previously analysed by thermogravimetry and it has been necessary to demonstrate the applicability of the procedures to coals of this type. To this end, twenty New Zealand South Island Late Cretaceous coals, ranging from lignite to low volatile bituminous, were subjected to TG and DTG, the results of which are summarised here. The samples used, along with their corresponding analytical data, were suppUed by the Coal Research Association of New Zealand from its industrial coal database. Figure 1 shows the location of the regions where these samples came from. The apparatus used was a Polymer Laboratories Rheometric Scientific (formerly Stanton Redcroft) Simultaneous Thermal Analyser STA 1500 capable of simultaneous determination of the DTA and TG profiles of samples. Cylindrical alumina cmcibles with 0.5 mm thick walls were placed directly on the type R (Pt-13%Rh/Pt) thermocouple plates. Accompanying Polymer Laboratories software derived DTG data from the recorded TG signal. Experimental parameters and conditions were constrained to permit direct comparisons of different samples and to ensure repeatability (Benfell et al, (5)), and were similar to those of Morgan et al (6) and Crelling et al (7). The purge gas used was compressed dry air at aflowrate of 50 ml min-\ A heating rate of 15°C min"' was applied from room temperature up to a maximum of 900°C which allowed complete combustion of the samples. The humidity of the laboratory was stable, seldom varying beyond the confidence limit of the hygrometer. The systematic error in the temperature scale of the STA 1500 is ±1°C, as determined by calibration with high-purity indium, tin, lead, zinc and gold standards.

LEGEND

Coaindds

Main citics/towns Weslporl

Pacific Ocean

Figure 1. Map of the South Island of New Zealand showing the different coal regions and the location of the coalfields within them after Barry et al. (8). CHARACTERISTICS OF TG/DTG PLOTS Thermogravimetric analysis (TG) measures change in weight of a sample heated in a given atmosphere as a function of time (isothermal mode) or of temperature (non-isothermal mode). Where the fiimace atmosphere is air. Gumming and McLaughlin (9) noted a burning profile is obtained for coals. Derivative thermogravimetric analysis (DIG) shows rate of wei^t loss with time as a function of temperature. The effects on combustion of coal rank, maceral composition, mineral matter and physical properties such as surface area can be established by measuring the temperature at which various events occur. Characteristic temperatures of the combustion process indicated in Figure 2 are those defined by Crelling et al (7). Of these, Benfell et al (5) found the temperature of char bumout, Tg, the peak temperature, Tg, and the maximum rate of combustion, R^, to be the most suitable to characterise the combustion of coals. For example, Morgan et al (6) have demonstrated that Tg can be used to assess the furnace residence time to minimise unbumt carbon loss and hence provides a measure of the bumout performance of a coal. Superior bumout performance is observed in coals with lower bumout temperatures; coals with a higher Tg are harder to bum, requiring longer fumace residence times or higher temperatures to achieve complete combustion. When industrial boilers changed from South African coals to Botswana coals combustion efficiency decreased. The Botswana coal had a bumout time almost 50% higher than that of the South African coal (Carpenter and Skorupska (4)). 53

- 0

ta %

- -5 -i'c

- -10 .2

(D a£ 2 00

0-J

TG Curve DIG Curve T6 V 1 1 1 1 1 1 1 1 1 1 I I

Bo o - -15 ^o ^ r-^

100 200 300 400 500 600 700 800 900 Temperature (°C)

-20

Figure 2. TG and DIG curves of san^le 54/794fromReefton showing characteristic temperatures after Crelling et al. (7). In order to confirm the quality of the data obtainedfromthe thermal analyser for a typical New Zealand coal, 10 separate runs were made of sample CR54/804, a high volatile B bituminous, vitrinite-rich coalfromthe BuUer region (vitrinite ranges between 88-96%, Black (10)). The largest standard deviation was found for the temperature of the maximum rate of combustion, Tg: 1.9°C for a range of 523 - 528°C with a mean of 525.9°C (Table 1). In contrast, the tenperature of char burnout, Tg, showed a standard deviation of 0.6°C (range 605 - 606°C, mean 605.7°C). The overall repeatabiUty of the temperature measurements of <±2°C is a considerable improvement on the repeatability of ±6°C achieved by Morgan et al. (6) and ±4°C by Crelling et al (7). PEAK RUN T6 Tg RATE (°C) (%wt min-0 (°C) # 13.67 604.8 748 523.4 13.40 606.1 749 528.4 13.72 606.1 750 525.9 13.80 606.1 751 527.2 13.37 604.8 752 524.6 13.37 606.1 753 523.4 13.52 606.1 754 527.2 13.71 606.1 755 525.9 14.03 606.1 756 524.6 13.74 604.8 757 528.4 13.63 605.7 mean 525.9 0.2 0.6 1.9 a COMBUSTION BEHAVIOUR OF NEW ZEALAND COALS The New Zealand coals studied ranged in ASTM rankfromlignite to low volatile bituminous. Black (10) showed that New Zealand coals have distinct differencesfromthe older Carboniferous and Permian to Mesozoic coals of Europe, North America and Australia, including:

a strongly detrital nature; high vitrinite and very low inertinite contents, compared to the high inertinite contents and plentiful fusinite of both Carboniferous and Permo-Mesozoic coals; and 111. the morphologic appearance, and having some macerals characteristic of low-rank coals, but the rank (i.e. reflectance) of hard, higher rank coals. The lowest rank coals are lignitesfromthe Otago Coal Region. These are typically vitrinite-rich (range of 85-88%), with about 6 % inertinite and 4-9 % liptinite. The Canterbury samplefromMt Somers is also in the lignite-subbituminous rank range. This coal has 71-72% vitrinite, 14—16% inertinite and 3-6% liptinite. Greymouth, Reefton and Duller West Coast coals contain less than 6% inertinite. Most Greymouth and Reefton coals have vitrinite contents between 82 and 92%, and liptinite contents between 5 and 18%. Greymouth coals have more liptinite and vitrinite, and no pyrite compared to the Buller coals. Buming profiles of six samples of different rank are plotted in Figure 3. Both Tg and Tg increase with increasing coal rank, as found by Morgan et al (6) and Crelling et al (7), suggesting higher rank coals may show inferior bumout performance. In Figure 4 the volatile matter contents, dry ash free, of the New Zealand samples' are plotted against their Tg values. The Greymouth sample with a volatile matter content less than 40% has a high Tg values of 646.6°C. In contrast, coals with volatile matter contents greater than 40% have a wide range of values for Tgfrom420°C to 636°C. Generally, New Zealand lignites have lower bumout temperatures than bituminous coals. 1.

ii.

O 'Ocd

/<—\\b (54/923)

hvBb (54/794)

'aS

E -10

a -apo Xi CO

subA (54/798) hvCb (54/783)

-20

subB (54/789)

lignite (53/830)

-30-

100

—T" 200

—r-

300

400

I 500

600

—1— 700

Temperature (C) Figure 3. Buming profiles of six New Zealand samples of differing ranks.

—r— 800

900

650 n • C

C Co

o oo

600 -

I 54/814 54/926

oc 550 Urn JD ca jc. o o

£<uQ. 500 £

• o

450 -

Grevmouch Reefton CanterbunOtago Coking coal r 53/830

1 1 1 r "T" 60 40 50 Volatile matter (%, daf) Figure 4. Plot of volatile matter versus temperature of char bumout of New Zealand Late Cretaceous coals. Regionally, the following differences are apparent: 1. The Otago coal 53/830 has the lowest Tg value; 2. sub-bituminous and high volatile bituminous C Reeflon coals form a grouping with Tg values of 544°C to 574°C. There is a conspicuous gap of more than 100°C between the Reeflon grouping and the Otago sample; 3. Greymouth h i ^ volatile bituminous coals, and the coking coals have the highest bumout temperatures, higher than 584°C. The Otago sample 54/814 from Kai Point and the Canterbury sample 54/926 from Mt Somers have bumout temperatures of about 565°C, similar to lower-ranked Reeflon coals. Local differences exist, with some sub-bituminous coals showing higher char bumout temperatures than migjit otherwise be expected from their Tg values and rank. Regional differences are also apparent when samples are rated according to their peak temperatures, T6 (Figure 5): L Otago coal 53/830 has the lowest value for Tg, 393.8°C; 2. the Reeflon samples show a rapid decrease in peak temperature at volatile matter contents above 45%; 3. coals from Greymouth have the highest peak temperatures, but show a slight decrease in Tg as volatile matter content increases. Regional differences are less apparent in plots of volatile matter versus maximum rate of combustion (Figure 6): L the Greymouth samples have lower maximum rates of combustion (13%wt min"' to 18%wt min'O, with the samples clustered at the high-volatile side of the plot; 400

-1

above 45% volatile matter content, the Reefton and Otago samples show a wide range of maximum rates of combustion (13%wt min"' to 3 l%wt min"^) with Otago sample 53/830 having the highest maximum rate; 3. most of the Reefton coals fall at rates of 17%wt min"' and lower. CONCLUSIONS The following conclusions can be madefromthis work: 1. the key parameters of a coal's burning profile for characterising combustion behaviour are the temperature of char burnout, Tg; the peak temperature, T6 and the maximum rate of combustion, R^^; 2. the results obtained by thermogravimetric and derivative thermogravimetric techniques are highly repeatable; 3. TG and DTG can reveal the differences between coals of different ranks, and also between coals of the same rank. For example, New Zealand high volatile coals have a wide range of Tg values (from 420°C to 636°C) above volatile matter contents of 40%. 550 n • C 525 500 -

o ocP

I£ i

475 o

H 450 05 £ 425 400 375

54/926 54/814

Greymouth

O ijji X C

Reefton Canterbun' Otago Coking coal ,53/830

40 50 60 Volatile matter (%, daf) Figure 5. Plot of volatile matter versus peak temperature of New Zealand Late Cretaceous coals.

32 -

.53/830

28 ^ Occ •T3 T^ c £ 5 24 co

Q O cja

Gre>mouth Reefion Canterbury

Otago

Coking coal

£o o U. 20 E£ cc

o 16

•

Cn§ OQ

54/926 ^54/814

"T" -1 60 40 50 Volatile matter (%, daf) Figure 6. Plot of volatile matter versus maximum rate of combustion of New Zealand Late Cretaceous coals. Most of the New Zealand coals tested in this study have volatile matters greater than those previously analysed by thermogravimetry (for example Morgan et al (6) and Gelling et al (7)). The extension of the rank dependence of combustion found in these studies does not appear to be directly transferable. TG and DIG show that there is a substantial variation in T8 in coals with volatile matter contents above 40%, whereas a coal with a lower volatile content has a T8 value of 646.6 °C. Some sub-bituminous coals show higjier char burnout temperatures than coals of similar rank and T6 values. Other coals have lower T8 values than may otherwise have been expected from their T6 values and rank. The value of T8 gives a better indication of combustion efficiency than rank or volatile matter content alone. This technique may be used by industrial operators to give an indication of bumout performance before a particular coal is purchased, aiding the evaluation of the coal's suitability for the proposed usage. ACKNOWLEDGMENTS Financial assistance for analytical costs of this project has been provided by the Coal Research Association's Programme Development Group. Dr R.J. Sims provided invaluable assistance with the STA 1500. REFERENCES (1) Gunn P.R., Beamish B.B. and DePetris T. New Zealand coal qualities - How do they rate in world markets? Proceedings of the 28th Annual Conference 1994, New Zealand Branch of the AusIMM, Taupo, August, 1994, pp. 83-99. (2) Wagoner, C.L. and Winegartner, E.C. Further development of the buming profile. Joumal of Engineering for Power, 1973,95: 119423. 12

(3) (4) (5) (6) (7) (8) (9) (10)

Wame, S.StJ. Proximate analysis of coal, oil shale, low quality fossil fuels and related materials by thermogravimetry. Trends in analytical chemistry, 1991,10: 195-199. Carpenter, A.M. and Skompska, N.M. Coal combustion - analysis and testing. IEACR/64, lEA Coal Research,. London, 1993, 97 pp. Benfell, K.E., Beamish, B.B. and Rodgers, K.A. Thermogravimetric analytical procedures for characterising New Zealand and Eastem Australian coals. Thermochimica Acta, (in press). Morgan, RA., Robertson, S.D. and Unsworth, J.F. Combustion studies by thermogravimetric analysis. 1. Coal oxidation. Fuel, 1986, 65: 1546-1551. Crelling, J.C., Hippo, E.J., Woemer, B.A. and West, D.R Combustion characteristics of selected whole coals and macerals. Fuel, 1992, 71:151-158. Barry, J.M., Duff, S.W. and MacFarlan, D.A.B. Coal Resources of New Zealand. Resource Information Report 16. Energy and Resources Division, Ministry of Commerce, New Zealand, 1994, 73pp. Cumming, J.W. and McLaughlin, J. The thermogravimetric behaviour of coal. Hiermochimica Acta, 1982, 57: 253-272. Black, RM. A Reconnaissance Survey of the Petrology of New Zealand Coals. Report No. 51, New Zealand Energy Research and Development Committee, University of Auckland, Auckland, New Zealand, 1980,49pp.

EROMANGA BASIN WATER SUPPLY DEVELOPMENT FOR OLYMPIC DAM OPERATIONS

KJL. Berry' and D. Armstrong' 1 Senior Hydrogeologist, Exploration Division , WMC Resources, Belmont W.A. 2 Principal Hydrogeologist, Lisdon Associates, Glenalta S.A Summary Mine and prcx^essing water supplies are obtained from Borefield A, located near the southwest comer of the Eromanga Basin, 100km north of the site. The borefield comprises nine production bores, tapping the Jurassic Algebuckina Sandstone at depths of 100-200m. Abstraction has resulted in depressurization of the artesian aquifer, and has affected flows from some nearby springs. The effects have been judged environmentally significant due to fauna and flora supported by the springs. Expansion of the Olympic Dam Operations is dependent on increased water supphes. Borefield B has been located in a deeper, more productive part of the basin, 90km from Borefield. The new borefield will result in reduced potential for impacts in sensitive areas near the Basin margin.

STUDY AREA

HYD-236

Introduction WMC Resources Ltd owns the Olympic Dam mine which is situated 550 km north of Adelaide. The Operation currently processes about 3.0Mt/a of ore for the recovery of 85,0001 copper, 1,5001 uranium oxide, 850 kg of gold and 13,000 kg of silver. Groundwater for plant and domestic use is currently drawn at an average rate of 16ML/d. Water supplies for Olmpic Dam are currently obtained from Borefield A, located in a small southem embayment of the Eromanga Basin - its closest point to the mine site (Figure 1: HYD-236, Potentiometric contours after Habermehl, 1980). The Eromanga Basin is roughly synonymous with the broader hydrogeological entity, the Great Artesian Basin, within the study area. The southem margin is a major groundwater discharge area for the Great Artesian Basin. Discharge from the Jurassic Algebuckina Sandstone aquifer occurs under natural artesian pressure by spring flow, associated with faults or outcrops and by diffuse leakage through thinner confining shales. Discharge from bores has also been focussed on the marginal areas, since the aquifer is shallower and cheaper to develop, and pastoral activities are more intensive. AbstractionfromBorefield A is regulated by drawdown limits at the boundary of "Special Water License 1". As these limits are approached, there is concern over the ecological effects of declining flow from spring complexes in the vicinity of the Borefield. The current proposed (Kinhill, 1995) expansion at Olympic Dam Operations would increase the water requirement to a maximum of 42ML/d. Criteria for the expanded supply system were to allow reduced abstraction at Borefield A and to minimize impacts on natural spring discharge and pastoral bores. The adopted site for Borefield B is 100km fiirther into 60

the Basin, comprising 3 production bores and twelve observation bores at a distance of up to 45km. The latter will be used to assess compliance with a prescribed maximum drawdown at the boundaries of Special Water License 2. Geology

The stratigraphic section comprises sediments of the Tertiary Lake Eyre Basin and the Mesozoic Eromanga Basin. Hie Cooper and Simpson Desert Basins underlie the Eromanga Basin towards the margins of the study area but are considered equivalent to Proterozoic metasediments in forming a low permeability hydrogeologic basement. BOREFIELD A - BOREFIELD B GEOLOGICAL CROSS SECTION

EXPLORATION DIVISION

BOREFIELD A

MULOORINA BORE A85CU03 A85MA07

BOREFIELD B GAB 52 82RBF (saismid

30 kilometres VERTICAL EXAGGERATION = 40 X

June 1996

HYD-234

Mixed lithologies of the non-marine Lake Eyre Basin contain only highly saline groundwater. The upper Cretaceous Winton Formation has been developed as an aquifer on the eastem side of the Basin. In South Australia the Winton is dominated by siltstone with lesser fme-grained lithic sands of low permeability. The complete Lower Cretaceous sequence, including Mackunda, Oodnadatta, Coorikianna, Bulldog Shale and Cadna-Owie Formations is dominated by siltstone and clay stone forming a classic confining layer ranging in thickness from 500m in the south to 700m in deeper troughs to the north. Useful permeability is limited to the basal Algebuckina Sandstone, of upper Jurassic age which comprises mostly coarse fluvial sands. Most of the stratigraphic units are fairly uniform in thickness across the area. Increase in Basin depth to the north results from irregular thickening of the Algebuckina Sandstone, and regular thickening of the Winton Formation and Lake Eyre Basin sediments. Thickness variations of the Algebuckina Formation are especially marked near the Basin margin. Controls include a significant pre-existing topography and syn to postdepositional extensional faulting which gave way to broad scale folding after deposition of the Cadna-Owie Fomiation. Near the basin margin structural and palaeotopographical variations in the Algebuckina Sandstone are large compared to the aquifer thickness. In these areas the structure controls groundwater flow by isolating the aquifer into a series of grabens/palaeovalleys (Figure 2: HYD234). The aquifer thickness throughout much of the study area can be estimatedfromseismic coverage, as the interval between the seismic C and Z horizons, less the thickness of the Cadna-Owie Formation which is reasonably uniform at about 40m. To the north the aquifer thickens markedly, reaching in excess of 750m where it is underlain by the generally fine-grained Poolowanna Beds and Simpson Desert Basin sediments. To the northeast the low permeability Cooper Basin sediments underlie the aquifer sequence which is split by the Birkhead Formation into the upper Mooga Formation and lower Hutton Formation. Thickening of the Birkhead accompanies a trend of reduced permeability to the northeast, in the sand dominated formations. Hydrogeology

Hydrogeologic background is provided by the results of the Great Artesian Basin study, commenced in 1971 and covered in (amongst other reports) Audibert (1976), Seidel (1978) and Habermehl (1980). The main aquifer can be considered z 61

pressurized leaky container with pressure levels maintained by rainfall recharge in up-lying areas of aquifer outcrop (Northern Territory and Queensland). The southern margin of the Basin is a major area of natural discharge. Natural discharge is via springs and diffuse vertical leakage through confining layers, especially where they thin near the basin margin. Springflowsoccur where the aquifer outcrops at the down-gradient end of the flow system, or where major structures have provided flow paths to the surface. Many of the springs in the study area occur near outcrop, with small fault displacements raising the aquifer and allowing a preferred path to a topographically depressed area. Development of the Basin by drilling of water bores since 1880, resulted in boreflowbecoming the major component of groundwater discharge. Pressure and natural discharge is thought to have stabilized at lower levels with greater recharge and throughflow matching discharge from bores (Habermehl, 1980). Borefield A Borefleld A is located in a NNE trending halfgraben. The Proterozoic bedrock surface in the area is well described by a series of seismic refraction traverses (AGC, Sept 1987) showing a palaeotopographic depression which bifurcates to the south, corresponding closely to existing creek locations. Drillhole data shows the Algebuckina Sandstone aquifer pinches out to the west against a steeply rising bedrock surface. The aquifer is truncated to the east against a basement horst including the Hermit Hill Proterozoic inlier. Faulting along the Norwest fault zone has subdivided the area into two parallel (half)grabens: the Wellfield and Northeast Sub-Basins (Figure 3: HYD-232). The structural setting imposes hydrauhc barrier boundaries on three sides of the Borefield. Numerical simulations for prediction of the aquifer response, presented in the original EIS, indicated that production could be sustained by steadystate inflow to the grabenfiromthe north (Kinhill-Steams Rodger, 1982). Significant reductions in spring and bore flow rate were predicted to be localized. These predictions have proven to be substantially accurate.

Of particular recent concem has been declining flow rates from springs at Bopeechee. This is due to hydraulic communication between the Wellfield and Northeast SubBasins across and around the Norwest Fault Zone. The main strategy for limiting abstraction impacts near Borefield A is the development of Borefield B. Based on the ten year record of monthly water production, aquifer pressure and springflowdata, a fairly rapid recovery of springflowrate is expected upon reduction of the production ratefiromBorefield A. As an interim measure a system has been established whereby up to 0.5ML/d of water takenfi-omBorefield A is pumped across the Norwest Fault Zone hydraulic barrier to bore GAB20 where it is reinjected into the aquifer, thereby maintaining aquifer pressure in the southem part of the Northeast Sub-Basin. By this means pressure levels in the area have been stabilized over a period of increasing average abstraction. 62

Borefield B The general concept of a second borefield was considered in the original EIS presented in 1982. Planning for the construction of Borefield B commenced in 1992. Initial investigations, including 120km of reflection seismic and 8 drillholes were focused on an area 50km northeast of Borefield A. The aquifer thickness in this area was found to be highly variable, with complete pinchout over structural highs. A numerical groundwater flow model used to simulate the effects abstraction at the site, showed that drawdown would be localized by the structural constraints, and that long term impacts would be significant at the Basin margin. Borefield B has been constructed a further 50km to the northeast of the original site. Pre-existing exploration seismic data was used for targeting. The borefield is located in a broad structural low across which the aquifer thickens to 120m. Geophysical data shows aquifer continuity and further, more gradual thickening to the north. To the south, drawdown is limited by partial hydraulic discontinuity across structural highs. The borefield comprises three bores to a total depth of 750m. The Algebuckina Sandstone comprising weakly or uncemented clean coarse sands. Testing results gave an average hydraulic conductivity for the section of 10 m/day. The maximum artesian flow rate from each of the bores was up to 17.5 ML/d (200L/sec) at temperatures in excess of 60 degrees centigrade. A network of observation bores encircles the borefield. Regular pressure measurements at these bores will be used to assess compliance with the allowable drawdown at the boundary of the Special Water Licence. An extensive gravity survey was undertaken to allow interpolation of the structures effecting Mesozoic structures between seismic lines, and to aid in targeting of the observation bores. Groundwater Flow Model As part of the environmental impact assessment, it was necessary to predict the drawdown effects of the proposed abstraction schedule and to demonstrate how this drawdown would effect existing discharge from bores and springs. For this purpose a numerical model of the hydrogeological system was constructed, and is being updated based on the results of the Borefield B construction programme. The model uses upgradient constant head boundaries in the Algebuckina Sandstone to simulate inflow from recharge. Direct discharge from the aquifer occurs via bore and spring flow . Diffuse leakage occurs via confining layers at a rate proportional to their thickness. The conceptual model and hydrogeological layers are shown in Figure 4 (HYD-233). DIFFUSE LEAKAGE LAYER 2 LEAKAGE —7— LAYER 1 LEAKAGE SURFACE SPRINGS LAYER 3 LEAKAGE

PASTORAL BORES LAYER 3 LEAKAGE

CONCEPTUAL FLOW MODEL LAKE EYRE BASIM

LAYER 3 CONSTANT HEADS

EROMANGA BASIN LAYER 1

CEMOZOIC TO

- WINTON FM

AOUITARO

LOWER CRETACEOUS

-MACKUNDAFM . OCONADATTA FM - COORIKIANNA FM

LAYER 2 AOUITARO

LOWER CRETACEOUS

• BULLDOG SHALE -CAONA -OWLE FM - M O O G A FM

LAYER 3

LOWER CRETACEOUS

AQUIFER

MID JURASSIC

. BIRKHEAD FM - HUTTON F M • ALGEBUCKINA SST - P O O L O W A N N A BEDS

LAYER <

LOWER JURASSIC

SIMPSON DESERT BASIN

AQUITARD

PR0TER020IC

COOPER BASIN PROTEROZOIC METASEDS

June 1995

HYD-233

Hydrogeological layer positions determined from surveying, drilling and seismic data were interpolated onto a finite difference grid. Input data for the model, and initial conditions, were obtained from the following sources: • Aquifer pressure gradient from shut in pressure at bores • Aquifer parameters by flow tests • Long term discharge rates from pastoral bores • Estimates of spring discharge rates • Leakage through overlying shales determined by chemical profiles (Woods,1990) 63

• Response of shallow observation bores at Borefield A which indicated that storage in the confining beds is an important source of water. Parameters subject to the above constraints were adjusted such that the simulated pressure field matched observed and assumed pressures, following the conclusions of Habermehl (1980), that the Basin is presently in approximately steady state equilibrium. Reasonable calibration against observation bore hydrographs in the Borefield A area was achieved before attempting to simulate the performance of Borefield B. (Figure 5) Groundwater Flow Mode! C a l i b r a t i o n at B o r e f i e l d

GAB

1000 Days

2000 3000 from Commencement

^-Measured

He-j-dS i m u I a t e d

4000 P r o d u c t i o n

Head

Figure 5 Because of the immense size of the aquifer system, drawdown over periods up to several decades is primarily controlled by the geometry, pemieability and storage characteristics of the aquifer, which are fairly well constrained. For longer simulations, model boundary conditions which determine the groundwater throughflow rate become important. Simulation of the operation of Borefields A and B at a combined rate of 42ML/d for 20 years, showed that the induced cone of drawdown (Figure 6: HYD-235) has a relatively minor impact on spring discharge at the Basin margin. The operation of some nearby pastoral bores, which at existing artesian pressures drive hydroelectric turbines and/or extensive pipelines or bore drains, may be affected by the drawdown. Under the provisions of the Special Water Licence existing usage will be maintained to the extent that it is effected by the development. Simulated Regional Water Balance Changes in the regional water balance within the model domain during the long term simulation are shown in Figure 7. Increases in borefield abstraction are matched by initial storage depletion and minor reduction in pastoral bore and spring flow rates, with an accompanying reduction in vertical leakage losses due to reduced heads within the cone of drawdown. After the borefield abstraction rate reaches its maximum there is a gradual reduction in the rate of storage depletion which is largely compensated for by greater inflow to the northem boundary of the model. The constant head boundary at the northem limit of the model simulates the effect of confined and unconfined beyond the model boundary and the potential for increased recharge under larger hydraulic gradients.

MODEL SIMULATED DRAWDOWN AFTER 20 YEARS AT 42 ML/d

100 JULY 1995

kilometres

HYD-235

The Eastern boundary of the model domain is conservatively represented by a no-flow boundary v^hich isolates itfromany influences which could extend west from the Cooper/Eromanga Basin hydrocarbon production wellfields, where significant net groundwater abstraction occurs (Armstrong, 1990). The Westem boundary of the model domain is also a no-flow boundary. The effect of the two lateral no-flow boundaries is to source all inflow to the Algebuckina Sandstone within the model domain from the northem, constant head boundary. The water balance demonstrates that even under these conservative constraints, the discharge from the borefields does not dramatically increase the boundary inflow over the modelled period (42 years).

FLOW

MODEL

WATER

2001

2 0 11

BALANCE

125

100 75 50 25

0 -25 -50 -75

100 125

19 B 1

1 99 1

YEAR

^INFLOW

^ 0 0 0

^STORAGE

BORES

|PASTORAL

202 1

203 1

204 1

BOREiaaiEAKAGE

Condusions Concerns over the potential impacts of Borefield A on nearby spring complexes, and e?q)ansion of the Olympic Dam Operations water supply requirement, will be satisfactorily addressed by the Borefield B development. Careful siting of the Borefield along with a comprehensive monitoring and reporting programme, and regulation of operation by prescribed maximum impacts will ensure satisfactory long term operation of the water supply system and minimise impacts on the sensitive springs. References A.G.C., 1987. Wellfield A Construction. Unpubhshed Report Armstrong D.& Rowan I.,1986. The development of a management strategy for the Olympic Dam Project water supply scheme. Proceedings of the Intemational Conference on Groundwater Systems Under Stress. AWRC Series No. 13. Amistrong D. 1990. In: Natural History of the North East Deserts. Royal Society of South AustraUa. Adelaide 1990. Audibert, 1976. Report on the Great Artesian Basin Hydrogeological Study 1972-1974. BMR Record 1976/5 Habermehl,1980. The Great Artesian Basin, Australia. BMR Jour. Geol and Geophysics 5. Kinhill, 1995 Olympic Dam Operations Survey and Assessment Report. Supplementary Environmental Studies. Borefield B Development. Kinhill-Steams,Rodger,1982. Olympic Dam Project. Draft Environmental Impact Statement. Seidel, 1978. Hydrauhc Calibration of the GABHYD model of the Great Artesian Basin. BMR Record 1978/12. Woods, 1990. Estimating Groundwater Discharge at the Southem Margin of the Great Artesian Basin Near Lake Eyre, South Australia. Proc. Int. Conf. on Groundwater in Large Sedimentary^ Basins. AWRC series No. 20

IRON MINERALOGY OF THE CALLIDE COAL MEASURES AND ITS EFFECTS ON COAL QUALITY Mark Biggs, Senior Geologist, Callide Coalfields Pty. Limited Keywords: Callide Coal Measuresy iron mineralogy in coal SIROLOG, effects on coal quality prediction. Summary This paper presents some results of research undertaken by the author as part of a Master of AppUed Science through the Queensland University of Technology. CalUde Coalfields produces a sub-bituminous, dull, sub-hydrous, steaming coal, primarily for domestic power generation from four deposits within the Callide Basin. The majority of coal is won from the Callide Seam Member. This coals' most striking feature is a high inherent mineral matter content, and more specifically, significant concentrations of iron minerals are present. These minerals subsequently contribute to high levels of iron oxide in the coal ash. The spatial variability of these iron minerals is high, which in tum, effects coal quality parameters during the coals' assessment and utilisation. This study highlighted the need to fiilly characterise the coal. This was achieved by using a combination of more conventional techniques such as reflected light petrography, X-Ray Diffraction and SEM techniques that examined the mineralogy of iron species; and interpretation of SIROLOG downhole geophysical log data. The study found that upwards of 90% of the iron minerals in the Callide Seam Member are represented by siderite/goethite and minor pyrite. The remainder is represented by primary iron oxides (haematite) and other secondary hydrated iron oxides. Some of the iron oxides are weakly magnetic. High concentrations of these iron minerals, specifically siderite, in Callide Seam Member coals significantly influence a number of the coals' physical and geochemical properties, as summarised below:• as the siderite is observable in hand specimens it imparts a brownish hue to the coal; • interpretation of downhole geophysical logs due to the effect of high siderite concentrations effecting resistivity, density, SIROLOG coal ash, and magnetic susceptibility logs; • coal relative density is somewhat hi^er than expected; • makes determining mineral matter, water of hydration, organic sulphur, and even total iron difiScult in some cases. The standard formulas for estimating mineral matter do not apply at Callide (mineral matterrraw ash ratio is 1.5:1). High siderite contents also contribute significantly to the total volatile matter of the coal; • combustion behaviour is unpredictable on occasions, and is dependant upon the ratio and association of siderite to the other minerals present. • plays a role in improving the coals' carbon char reactivity potential. The coal characterisation study completed is seen as an integral part of ongoing investigations aimed at determining lithological and structural controls to the iron distribution in the coal. Eventually, a methodology for predicting iron mineral concentrations in advance of mining and reconciling "as-sold" coal quality to that predicted will be developed. Predictions are currently being assessed at the coal face as mining progresses through high-iron zones. Introduction The Callide Basin is located approximately 120 kilometres south-west of Gladstone and 450 kilometres north-northwest of Brisbane in eastem Central (^eensland. The coalfield covers an area of about 18,000 hectares and is located within a north-west to south-east trending synclinal basin 22.5 kilometres long by 8 kilometres wide. Present open-cut mining operations are located in the Dunn Creek, Trap Gully and The Hut areas in the south as part of the Callide Mine, and at Boundary Hill in the north-west of the basin (Figure 1).

Kilometres

Power Station

Most of the coal is won from the CaUide Seam Member of the Callide Coal Measures. This unit is of Late Triassic age and contains alternating sequences of conglomerate, sandstone, siltstone, mudstone, carbonaceous shale, and subbituminous durainous coal of varying hardness and thickness. A detailed description of the geology of the area is given by Biggs et al (1). Current mining seam thicknesses are 16-21 metres, althou^ seam thicknesses down to, and including 0.3m can be encountered . Figure 2 outlines a typical composite stratigraphic column of the various coal seam horizons and their total iron contents. Coal quality indicators range widely vertically and laterally, eg. Total Moisture (16-20%), Raw ash (8-28%), Iron oxide in coal ash (0.5-72.0%), and Initial Deformation Temperature, reducing atmosphere (10801600+°C).

Figure 2 CALUDE COAL MEASURES TOTAL IRON CONTENT TYPICAL STRATIGRAPHY ALSO SHOWING . 10 TOTAL IRON CONCENTRATIONS

195 Ma

U i<

^AVERAGE

PRECIPICE SANDSTONE

3.0 SANDSTONES

1.0 204 Ma

150 TRIASSIC-JURASSIC DISCONFORMTTY

MARKER SEAM HORIZONS

0 0.5 COAL

0.2 10.0

0.2 COAL

125

2.9

CALLEDE SEAM MEMBER

63.0 PARTINGS

10.0 0 •

100

I SANDSTONES

0 1.0

CALLEDE

SAWMEX SEAM HORIZONS

COAL MEASURES 210 Ma

3.0

[ ]

COAL

0.2

DISCONFORMTTY

220 iCb

75 0 0.3 BOTTOM SEAM HORIZONS

COAL

0.1

UNCONFORMTTY

230 Ma 242 Ma

BASAL CONGLOMERATE MEMBER

Insufficient data

01- Jvu; Scalein metres

: C/3 < <

OJ C^ H

1.0

UNASSIGNED TRIASSIC VOLCANICS

27.0 i VOLCANICS) >

8.0

250 Ms

Methods of Investigation The work program was performed by the author with assistance from Technical Services staff at the mine and involved the collection of field mapping and drilling data. This enabled a network of cored holes to be established at the major deposits from which lithological strip logs can be drawn and cross-section or fence diagrams generated. Prior to this study, Callide's existing geological mine planning databases contained mostly lithological, geotechnical and base analytical (proximates, ultimates, ash fusion) data only. The author used Callide's existing work-station based mine planning software (Exploration Computer Science (ECS) Minex) to create a series of new data types and variables within the existing framework, so as to enable grid models to be generated for:• coal petrographic analyses ; • coal and parting mineral matter by low temperature ashing. Mineral identification by XRD and/or SEM; • all available coal ash and trace element ply data by AAS or XRF techniques; 69

• estimates of ash, iron, and silica from downhole geophysical data , eg. SIROLOG; • all airphoto and Landsat linear data; • all groundwater data. Considerable use was made of existing downhole geophysical logging that had been collected in conjunction with exploration and in-pit quality drilling at Callide since 1981, and held within Callide's geological modelling databases. From 1981 - 1989 this wholly consisted of BPB coal combination sonde data (caliper, natural gamma, long and short spaced density), with a lesser amount of focussed electric (resistivity), neutron-neutron, and sonic logging. While providing a wealth of information on coal quality and thickness, data on iron distribution from these log responses was found to be limited. Since 1989 all exploration and mine planning logging has been conducted using the SIROLOG sonde. This logging technology employs the prompt neutron-gamma method (neutron-capture), for the determination of ash, total iron, total sihca, and ash fusion temperatures in coal seams and partings (Charbucinski et al, (2); Biggs, (3); Borsaru et al, (4)). This information enabled updating of the quality databases which in tum allowed rigorous statistical analyses of available values (by seam) and a trial of geostatistical (ordinary kriging) methods for selected seams to be undertaken. Visual Recognition Iron minerals are visible throughout the CalUde Coal Measures in exposures artificially created through the mining process (eg, highwalls) and can generally be grouped into three classes:• haematite ironstones in layers or large nodules at or near the boundary with the Precipice Sandstone • siderite/goethite-rich layers in coals, or as disseminated nodules, • sphaerosiderite and siderite-rich claystone layers, most commonly as intra-seam partings, and sometimes as thin layers at the base of the Callide Seam. Iron minerals present are a distinctive feature of the Callide Seam Member being visible in hand specimen. Megascopic mineral components include discrete bands and laminae of siderite-rich clay, lenticles, isolated nodules, and more closely packed aggregates forming well defined concretions. A significant presence of iron minerals gives the CalUde coal a distinct brownish hue upon field examination. Detailed mapping of all the high-iron zones in highwall exposures is not practicable because of access and safety restrictions. A number of techniques have been trialed and/or proposed recently to enable iron minerals, particularly siderite, to be identified in the coal face objectively efficiently, and cheaply. Several of the methods rely on the absorption by coal of infra-red radiation or X-ray-radiography, but are not discussed fiirther here. Characterisation Of Mineral Matter Previous work on mineral matter in coal from several deposits in the basin has been undertaken by Ward (5) who subjected coal samples to low temperature radio-frequency ashing and XRD techniques. At Boundary Hill, Faraj (6) and Patterson and Marvig (7), conducted scanning electron microscopy and microprobe analyses of coal seam composites. The author conducted complementary studies on new coal samples, including intra-seam partings, and increased the coverage to include all known coal deposits/ prospects within the basin. The results of this work are summarised in Table 1, below:-

Table 1. Mineral Matter in Callide Coal Significant (>3 %)

Minor (1-3 %)

Mineral Species

All Kaolinite Siderite Quartz Goethite Haematite Gypsum Smectite-Kaolinite niite Pyrite Montmorillionite Illite- smectite Calcite Ilmenite Nacrite Fluorapatite Dickite

% Total Mineral Matter

59.1 16.8 10.0 5.0 3.0 1.0 1.0 0.6 0.5 0.5 0.3 0.3 0.2 0.2 0.2 0.1

% in Coal

16.02 4.55 2.71 1.36 0.81 0.27 0.27 0.16 0.14 0.14 0.08 0.08 0.05 0.05 0.05 0.03

Others

Trace (<0.1 %) Anorthite Gibbsite Bilxbyite Groutite K-feldspar Strontianite Muscovite Ca- gorceixite Berliaite Khademite Dawsonite Halloysite Monazite Zircon Rutile Sphalerite Halotrichite

uncombusted organics 1.2 Coquimbite TOTAL 100.0 25.0 The important minerals detected in Callide Basin coals were in decreasing order, kaolinite, siderite, quartz, goethite, haematite, gypsum, smectite-kaolinite, illite, and pyrite. Trace minerals detected were titanium oxide, ilmenite, pyrrhotite, sphalerite, zircon and chalcopyrite. Calcium and magnesium appear to be organically bound in the coal. Three iron containing minerals, siderite, goethite and pyrite account for the bulk of iron present in the coal samples examined. The siderite has been found to be partially oxidised to an iron hydroxide phase, suggested to be cryptocrystalline goethite (aFeO.OH). Relative amounts of siderite and associated goethite are hard to predict, but on average, about 1/7 of the siderite is so effected. Additionally, two main forms of siderite with essentially identical chemical composition were observed by Faraj (6) in variable amounts. Siderite nodules formed early in diagenesis and cleat/vein and other infillings in coal, presumably formed much later after consolidation of the coal. The chemical composition of siderite was essentially the same regardless of form, locality or stratigraphic depth and therefore is not a useful indicator of changes in depositional environment or different stages of diagenesis. The average formula determined by the current study is FCQ 95 MgQ QI ^ 0 . 0 1 Cao.oi CO3. Attempts were made to investigate the correlation of SIROLOG iron and silica data with mineral matter determination, however, of the major iron mineral contents could be made from chemical analyses for carbonate carbon and pyritic sulphur in the coal. The balance of total iron could then be expressed as goethite. Iron Mineral Distributions The distribution of the iron minerals was investigated by combining outcrop observations with detailed chemical analyses of core samples, SIROLOG iron trace fence diagrams, and a more extensive set of SIROLOG iron estimates on a seam by seam basis. Contouring of iron in ash was initially conducted on a deposit or seam basis. Both outcrop observations and quantitative iron data suggest, however, that the distribution of iron is too complex to be adequately expressed in seam by seam maps. Rather, iron minerals occur as an overlapping network of laterally discontinuous patches and zones throughout the coal body, across the mine area. To overcome this, a working section composited iron colour shaded contour map was generated for the entire basin (Figure 3). Some association with structural features is evident on the plan.

In conjunction with the mineralogy study, a detaHed petrographic study was undertaken to determine whether high iron occurrence could be related to coal facies distributions within the seams. The study examined the vertical and lateral variability in petrographic composition and iron content. Siderite showed no particular affinity for any coal maceral observed, however was marginally more abundant as simple spherules (average diameter 250 (m), complex nodules, or as cleat infillings in attrital inertinites. This study, in agreement with previous workers, found that much of the iron is present as siderite, but in two forms. The first phase of siderite formed during the accumulation of shallow burial of the formative peat. A second, more widely distributed phase, formed much later at elevated temperatures and associated with circulating groundwaters. The second phase overlaps considerably with the first, but is concentrated along thin beds of powdered coal ("sooty" to touch), adjacent to clastic partings; along stmctural discontinuities; and near mafic igneous intrusions (Fielding (8)). Iron estimates from the SIROLOG probe are only generalising the iron occurrence and could not differentiate between these different phases. Further field studies are focussing on aerial mapping of zones of structural disturbance; isopach mapping of clastic partings; and delineation of areas and seams affected by structuring and adjacent to thick clastic partings to further define iron mineral distribution. Geochemical Recognition The author investigated whether the high levels of iron mineralisation in the Calhde Seam Member have had an effect on trace element distributions, through scavenging processes. Certain trace elements are easily absorbed through ion exchange onto colloidal particles. Iron and manganese oxides and hydroxides, organic matter, clays, and silica are the most common natural materials occurring as colloidal particles. Under favourable conditions small amounts of colloidal materials can scavenge important amounts of dissolved elements from solution (Rose et al.(9); Butt et al. (10)). The author has studied the trace elements in 25 coal and parting samples using ICP-AES techniques. Apart from finding that Sr and Ba increase as the iron content does (probably substituted within the siderite), no consistent increase in trace element concentrations can be seen for an increasing iron mineral concentrations in the coal or partings. Geophysical Log Interpretation High concentrations of iron minerals within the coal and intra seam partings do effect interpretation of geophysical logs, because of their unique properties, as shown in Table 2.

TABLE 2. GEOPHYSICAL LOG CHARACTERISTICS FOR COAL AND ASSOCIATED STRATA. Density Lithology Velocity

Gamma

(g cm.3)

Neutron (c.p.s.)

ray api

Sonic

Porosity

Count

Resistivity

Travel time

units

Rate

jisecfr'

High

Low

Med*

ms-^ Shale Med* Sandstone High Coal Low* Cindered coal Med. Dolerite High Siderite & High pyrite in coal

High

Low

High

Low

High

Intermed

Low

High

Low

High

High*

Low

High

Low

Intermed to Low

Med.

Low

High

Low

High

Low

High

Low

High

Low

High

Low

* Variable. (after Renwick (11)). Previous mine studies have attempted to use BPB long-spaced density logs to give estimated values of ash, relative density and specific energy, but they did not study the effects of siderite on the density logs. The author has found only limited success in using long-spaced density logs to estimate coal quality parameters, as accuracies determined from ply sample results, for an 8% to 60% ash range were only marginal (±3-6% absolute for ash content). Others have found that high concentrations of siderite minerals can affect other logs. Tittle (12) found that siderite was affecting interpretation of the compensated neutron log during well logging in petroleum exploration. Murphy et al. (13) found that high concentrations of siderite (50-60%) were affecting interpretation of detailed resistivity logs in a similar situation. High concentrations of iron minerals have found to also effect interpretation of SIROLOG density traces as shown in Figure 4 which plot assay histograms and geophysical traces against lithology for a borehole, C2866, from Boundary Hill. The geophysical logs on the left of the lithology column are LSDEN (blue), LEDEN (red), Fe203 in ash% (green) respectively, and then on the right are FERAT (red) calorific value (pink) and ASHRAT (black) respectively. Note how the Al seam has a slightly higher composite iron in ash value. However the FERAT trace more clearly defines that is the top two metres of the seam that is high in iron. This is affecting the response of both the LEDEN (5-5 probe, (7.5-30 Kev) and the ASHRAT trace (n-^, the ration of 2.6-5.18Mev/0.38-0.60Mev), both of which are used routinely for seam thickness and parting delineation. Any interpretation using these logs alone would incorrectly classify the top 2 metres of the seam as inferior coal. Observation of the calorific values though would suggest that this is still reasonable quality coal.

S s

SlrtOuOC isd.eq'

LEGEND

Lithology F ^

Reference

MUDSTONE I SANDSTONE 1 COAL INFERIOR

^ •

SILTSTONE COAL ( S U F F I X E D ) CORE LOSS

g ^ i CLAYSTONE

FIGURE 4

muu

LiTMotCXc

SHOsoii^ff

High iron mineral concentrations appear to be also affecting interpretation of the focussed electric (resistivity) logs. Recent re-analysis of BPB focussed electric (apparent resistivity) log at the Kilbumie deposit has shown that as the iron values decrease downhole, the resistivity values increase in a general sense. Additionally the resistivity log appears to be discriminating high iron, but thin (<0.2m), intra-seam clastic partings that do not show up on the long-spaced density logs. It would appear then, that some differentiation of iron minerals from surrounding coal and clastic lithologies is possible using the resistivity method. The author sougjit to map the abundance of iron minerals in coal seams mapped during highwall inspections using magnetic properties. Magnetic susceptibility is largely govemed by the magnetic mineralfraction(nonnally magnetic %) of a rock, which accounts for most of the susceptibility observed. This value depends upon several factors, such as the 74

intensity of the magnetising field, chemical composition and grain size of the magnetite. The magnetic susceptibility was measured on core material from borehole C6014 (Central Reserve) on a 5cm increment basis. Two methods were employed, a portable meter (GMS-2) and a downhole logging tool. Analysis of core measurements from C6014 using the meter, compared to downhole measurements, and SIROLOG data showed good agreement where nodular ironstone bands were intersected between seams (probably mostly haematite). However the response in coal seams was variable, with not all high iron (high siderite?) zones recording correspondingly high susceptibilities. In fact some high iron zones gave no magnetic reading at all. Effect on Coal Quality Analysis The higji iron mineral contents contained in Calhde Coal appear to impact either directly and indirectly on the coals' classification (Sanders (14)). Problems have been observed periodically in determining some standard laboratory analysis; mineral matter and water of hydration Sanders (14); pyritic sulphur; iron oxide in ash. As Howse (15) reported, a commercial laboratory experienced problems analysing exploration samples for forms of sulphur. A number of the results had pyritic and sulphate values which when added exceeded the total sulphur result and hence generated a negative organic sulphur (range -0.03 to -0.47%). Replicate analyses at other NATA registered laboratories experienced similar problems. All the problem coal contained a higher than usual iron content. It was subsequently found that even with extended washings the iron could not be removed and was artificially increasing the pyritic result. Samples that exhibited this problem were re-analysed according to the ISO Standard with little improvement. It would appear that the Australian Standard for Forms of Sulphur is not suitable for analysis of high iron coals. A mention of this inadequacy does appear in the current standard (AS1038.11-1993) but no altemative method for analysis is offered. During the course of analysis some of the coal ash samples appeared to contain magnetic iron minerals of a higher oxidation state than the normally reported, Fe203. This causes the reported analysis totals of such samples to be much lower than expected (T. Krapkat, pers. comm.). The phenomenon does not appear to be related to ashing of the samples in air, have comparable iron contents but contain little (if any) magnetic component, and their totals are normal. Therefore it is likely that these minerals are inherent in the ash. One such example was L1906:16 (MB59), which reported 39.4% Fe^-" and 8.1% Fe^. Analysis of this sample revealed mineral matter dominated by haematite and goethite. Measurements of magnetic susceptibility recorded 1.4x10"^ (9.3% magnetite). Meaningful comparison of the organic component of different coals is only possible if allowance is made for dilution by inorganic mineral matter. Determination of mineral matter content from the amount of ash, sulphur, and other inorganic compounds is based on assumptions regarding the occurrence of these materials, and the validity of these assumptions in each case depends on the assemblage of minerals actually involved. Radio frequency oxidation of Callide Coal (Ward (5); Steer and Ward (16)) has shown that traditional methods of estimating mineral matter are inadequate, as shown in Tables 3a,b below:TABLE 3a COMPARISON OF ACTUAL VERSUS ESTIMATED MINERAL MATTER AND WATER OF HYDRATION FOR THREE CALLIDE SAMPLES SAMPLE

RAW ASH (% ad)

Mineral Matter

Mineral Matter RFO %ad.

Mineral 1 Matter (petrology)

W.hyd % ACTUAL KMC

W.hyd % 2 FORMULA

BH:BLK4-1983 TG-lA-1983 D2U-1983

10.4 12.9 16.4

13.7 16.6 22.0

6.0 7.0 11.0

12.0 15.3 20.4

1.0 1.8 1.9

1.0 1.3 1.6

1 King-Maries Crossley formula MM = 1.13ASH + 0.8C02 + 0.5Spyr + 2.85SS04- -2.85Sash + 0.5cl. 2 Australian Standard AS1038.16-1995 (WJiyd = O.IASH) TABLE 3b LOCATION

# SAMPLES

RATIO RFO MM/ASH

Dunn Creek Trap Gully The Hut Boundary Hill

63 47 21 14

1.45 1.59 1.54 1.46

Other methods of estimating mineral matter (ie BCURA, Parr) were also found to be inadequate. Classifying Callide Coal has been difficult in the past (Glikson and Fielding (17), Esterle (18)) as it does not fit well into traditional classification charts ie Seylers Classification; Hardgrove Grindibility Index vs RoMAX (vitrinite); Volatile Matter vs Carbon. Points raised are:• the volatile matter is too low for the carbon content (dmmf basis); 75

• hydrogen (dmmf basis) is very low (sub-hydrous); • vitrinite reflectance appears to be too low for the carbon contents and hardgrove grindability index (Chalcott and others, (19); Spero (20). Further investigations of these matters are being pursued. Storage and Handling Iron minerals in Calhde Coal can be associated with a number of problems in the mining, preparation and storage of coal and coal mine products. Nodular siderite "sheets" (sphaleosiderite), and siderite-rich claystones, commonly found as lenticles, bands or concretionary masses in the coal seams and intra-seam partings, can give rise to undue abrasion of the cutting surfaces on mining machinery when encountered during mining. An increase of iron minerals does appear to influence the grindability of coal. For coal of a given rank, Spero (19) rates the hardgrove grindability index of macerals in the following general order:micrinite<liptinite<vitrinites<semi-fusinite<fusinite Chalcott and others (20) and Spero (19) have shown that Callide Coal has an anomalously high HGI for the rank. Probable reasons are high semi-fusinite contents and porous non-cemented nature of these macerals (J. Esterle pers. comm.). Analyses of several samples of sphaleosiderite showed a range of properties as shown below:20-70 % Siderite Content Relative Density Hardgrove Grindability Index Uniaxial Compressive Strength (MPa)

2.58-3.31 44-62 25-150

Large masses of hard minerals, such as quartz (H=7) or possibly quartz (H=6) rather than siderite (H=5.5-6), however, are more likely to give rise to excessive abrasion or wear on the grinding surfaces, resulting in reduced mill life. Spero (19) surmised from this data that recirculated material in the grinding zone of the mill would be significantly more abrasive than the product coal, because of the greater concentration of quartz and/or siderite in the higher density fractions of the coal. However, comparisons carried out by the author (Table 4 below) could not link abrasion index, hardgrove grindability, and iron oxide in ash contents for several frill seam composites taken from boreholes at The Hut, Dunn Creek, Trap Gully and Boundary Hill Deposits. TABLE 4. VARIATIONS IN GRINDABILITY WITH IRON CONTENT at a Total Abrasion Iron Oxide in Sample LD. Moisture Index Ash (%db) (%ar) (mg steel/kg) C2650 F2463 C2467 C2468 C1771 C2867 C2466 F2464

8.0 —

10.0 10.0 13.0 17.0 20.5 21.8

7 7 25 18 30 9 15 20

13.5 10.5 11.1 11.1 12.5 14.4 11.2 10.8

Hardgrove Grindability Index

Slagging Index (red ®C)

75 76 64 65 73 73 71 71

1350 1440 1370 1370 1390 1335 1305 1220

Slagging Index = 0.8x initial def. temp (red) + 0.2 x hemisphere def. temp (red) Relative Density An in-house study of relative density found a great variability of relative density between seams and deposits. Great difficulty was found correlating in-situ linear density, measured from calibrated BPB downhole geophysical logging and densities converted to an "as received" basis from laboratory supplied air dried results (Australian Standard 1038.21.1:1994). Increases in mineral matter in coal will affect the density of the coal. However the relative densities of Callide coal are anomalously high at an average of 1.50 (air dried). This has previously been attributed to the high semifiisinite and interodetrinite contents (Stevens, (21)). Figure 5 shows for a large diameter borehole (Dunn Creek L1906) relative density vs iron oxide in ash. The author believes that the elevated relative densities of Callide Coal are partly due to high inertinite contents and partly due to the high siderite content (on average 16.9% of the mineral matter). Combustion and Slagging Callide coal has shown to be dominated by kaolinite siderite, quartz, iron oxide in its mineral assemblage. Those seams and deposits dominated by kaolinite are generally non-slagging, and since kaolinite does not ftise until temperatures of around 1700°C are attained. The resultant ash is generally quite refractory under fumace operating conditions. The very subordinate amount of coal that contains minor amounts of mixed-layer clays also appears to have a refactory ash, but 76

some coalfromthe Hut deposit containing abundant montmorillonite, or illite generally has initial deformation ash fusion temperatures, between 1000 and 1300°C. It appears that those CaUide coals which contain a significant proportion of siderite, pyrite, sulphates or iron oxides can also produce low ash fusion temperatures, even thou^ the associated clay may be quite refractory in nature. The interplay of processes that control ash behaviour is very complex, and depends on factors such as the atmosphere, temperatures and coal chemistry involved. With pulverised fuel, it also depends on the distribution of minerals and other inorganic constituents in the coal, and the opportunities that the different components have to react with each other in the combustion chamber. This situation is currently being investigated by the author, whereby ratios of siderite, kaolinite and quartz are being examined as to their effects on combustion behaviour.

<C/3 S O c o

80.00 60.00 40.00

• Fe203 %

20.00 0.00 1.00

1.50

2.00

2.50

3.00

Relative Density (ad) Figure 5. Dunn Creek Li906 and L1907 Carbon Char Reactivity Iron and steel can be produced an altemative process using gases such as hydrogen and carbon monoxide in a kihi to achieve direct reduction of the oxide ore. The gases in many plants of this type are derived from natural gas or petroleum products, but a number of operations also exists, or are under investigation, where they are produced by partial combustion of fuel coals. Low rank coals produce chars of highest reactivity, and the use of lignites and sub-bituminous coals result in lower process temperatures, improved productivity and better plant operation. Considerable experimental work has been conducted on Boundary Hill coal (eg. Stephans (21)) which has carbon char reactivities suitable for this process, only slightly inferior to New Zealand coals currently used in the process. Boundary Hill Coal, specifically seams A2 and A3 have 3 fold higher reactivities than the rest of the basin. Coal rank affects char reactivity as does the presence of catalytic elements in the ash, such as sodium, potassium and iron. Research to date has shown that the following factors influence the carbon char reactivity at Callide, and improvements relate to coals with:• higher high liptinite contents; • increased porosity; • higher P and K20 in mineral matter; • increased percentage of inertodetrinite (broken macerals) • significant variations in vitrinite and siderite concentrations. Summary And Conclusions A range of coal characterisation techniques have been used to investigate the distribution and mineralogy of mineral matter, and specifically iron minerals, within Callide Coal Seam Member. CXher information gained from the SIROLOG probe which measures in-situ raw ash, siUca and iron provided a major adjunct to existing laboratory analyses. Examination of downhole traces complemented mapping of higjiwall closures, and indeed did highUght that the rates of spatial variability in iron mineral content throughout the Callide basin are high. Other techniques were used to identify two phases of emplacement of the major iron mineral - siderite. One phase is thought to occur syn- to epigenetic with peat accumulation and early burial history, the other post-coalification and associated with intrusive volcanics and groundwater movement. This variability creates difiBculties in predicting zones of high iron mineral matter contents in advance of mining. Field mapping at some locations within the Callide Basin suggests that much of the iron mineralisation is secondary and distributed relative to the structural grain of the geology, rather than the sedimentological features. The proportions of each phase are as yet unknown. SIROLOG is a unsuitable tool in this regard, being only capable of measuring total element (eg. iron) concentrations At the highwall scale, iron mineralisation decreases rapidly (within a few hundred metres) away from dykes, faults and joints. It is suggested that in addition to following vertical features such as faults/joints and cleat, secondary iron mineralisation in these areas of high iron have followed horizontal discontinuities within seams. These discontinuities were created by clastic partings, boundaries between petrographic cycles, small scale erosive features and fracture zones caused by compressive deformation.

These high concentrations of iron minerals, specifically siderite, in Callide Seam Member coals significantly influence a number of the coals' physical and geochemical properties which in tum effect its utiUsation. Adverse and unpredicted combustion behaviour has been experienced by some of the coals end users. Work is ongoing to develop a methodology for predicting the coals behaviour prior to mining. This is only possible through conducting a thorough coal characterisation, with particular emphasis needed fully identifying all mineral matter in the coal. REFERENCES 1. Biggs, M.S., (1995), Burgess, A.W. and Patrick, R.B. Callide Basin, in Ward, C.R., Harrington, H.J., Mallet, C.W., and Beeston, J.W.. eds Geology of Australian Coal Basins, Geological Society ofAustralia Coal Geology Group Special Publication, 1, pp 471-488. 2. Charbucinski, J.,(1986) Youl, S.F., Eisler, P.L. Borsaru, M. Prompt Neutron Gamma Logging for Coal Ash in Water Filled Boreholes. Geophysics, 51, 5, pp 1110-1118. 3. Biggs, M.S. (1991) The Application of Neutron-Gamma Sirolog to Estimate Iron Content in Coal and Implication for Estimating Ash Fusion Characteristics, in Griffiths P., ed, Queensland Coal Symposium, Aust. Inst. Min. Met., Brisbane, Qld, pp 187-198. 4. Borsaru, M.,(1993a), Biggs, M.S. & Nichols, W.J.F. Neutron-Gamma Logging for Iron in Coal and Implications for Estimating the Ash Fusion characteristics at Callide Mine. Nucl Geophys., 7(4): pp539-545. 5. Ward, C. R. (1990) Mineral Matter Analysis of Coal Samples from Callide, Queensland, Report R685 by Unisearch for Callide Coalfields Pty Limited, University of New South Wales, unpubUshed report pp 1-16. 6. Faraj, B.S.M. (1993) Investigations into Iron Mineralisation in Boundary Hill Coal, Dep of Earth Sciences, University of Queensland, unpublished report, pp 1-3 (25 May 1993). 7. Patterson, J.H.,(1993), & Marvig, P. Characterisation of Iron Contaiaing Minerals in Callide Basin Coals. Report CET/IR184R, CSIRO - Division of Coal and Energy Technology, unpublished report, pp 1-10 (December 1993). 8. Fielding, C. R (1993) Origin and Distribution of Iron-Bearing Minerals in Coal Seams at the Boundary Hill Mine. Dept of Earth Sciences, University of Queensland, unpublished report, pp 1-12 (Nov 1993). 9. Rose, A.W.,(1979), Hawkes, H.E. & Webb, J.S. Geochemistry in Mineral Ejq}loration Second Edition, Academic Press, London, UK, 657pp. 10. Butt, C., (1995), Gray, D., Douglas, G. & Fulwood, K. Geochemistry & Mineralogy of Lignites in the Ambassador Deposit, Mulga Rock, CSIRO Exploration & Mining Research News, 4, July 95:10-12. 11. Renwick, R J. (1982) Kilbumie Preliminary Petrophysical Study, The Shell Co. of Aust. Pty Ltd - Coal Division, unpubUshed report CEPR 6/82:1-9. 12. Tittle, C.W. (1989) Effect of Pyrite and Siderite on the Response of the Compensated Neutron Log. Nuclear Geophysics, Vol 3, No. 4: pp 335-337. 13. Murphy. W.F.,(1992), Auzerais, F.M., Luling, M.G., Anderson, B.L, tomanic, L, Bonner, S.D., Sakurai, Sinichi, & Wolcott, D.S. Proceedings of the 1992 SPE Annual Technical Conference and Exhibition. Washington, DC, USA. Oct. 4-7 1992.PP 155-170. 14. Sanders, RH, (1983) Evaluation of the Circumstances Contributing to the High Total Moisture Situation at the Boundary Hill Mine of the Callide Coalfield. Australian Coal Industry Research Laboratory Ltd, Report 02-8436, unpubUshed report. 15. Howse, R., (1992) Callide Coal: Forms of Sulphur , Carbon Consulting Intemational, unpubUshed memo to Callide Coalfields Pty. Ltd, July 1992: 1-2. 16. Steer, J., (1995) & Ward, C.R. Mineralogical Analysis of Coals of 10 Callide Coalfield Samples. University of New South Wales (Unisearch Ltd.), unpublished report 24185-01, June 1995: 1-5. 17. Glikson, M., (1991a) & Feilding, C.R. The Late Triassic Callide Coalfields Measures, Queensland, Australia: Coal Petrology & Depositional Environment. IntemationalJoumal of Coal Geology, 17: 313-332. 18. Esterle, J. S. (1992)Maceral Reflectance Variability and Recognition in Callide Coals. Report IR 174. CSIRO, Division of Geomechanics, unpublished report, pp 1-30 (December 1992). 19. Callcott, T.G., (1990), Callcott, R & Quinn, G.W. Review of Coal Characteristics of the Bowen Basin in Beestow, J. W (editor) Proceedings Bowen Basin Symposium 1990, Geological Society of AustraUa, Queensland Division, Brisbane, Sep 1990: 47-53. 20. Spero, C., (1991) Developments in Laboratory Testing & Evaluation of Coal Grindability, in Griffiths, P., ed, Queensland Coal Symposium, Australasian Institute of Mining & Metallurgy, Brisbane, Queensland, 73-82. 21. Stevens, J., (1978) A Palynological & Sedimentological Study of the Callide Basin, East Central Queensland. University of Queensland Honours Thesis (unpublished). 22. Esterle, J. S. (1994)Petrographic Variation in Coal Seams at Boundary Hill Mine, CalUde Coalfields, and it's relation to Iron Content, Report 5A, CSIRO -Division of Exploration and Mining, unpublished report pp 1-45 (Jan 1994).

THE APPLICATION OF MOBILE METAL ION SURFACE SOIL GEOCHEMISTRY TO MINERAL EXPLORATION R.D.Birrell, A.W. Mann, J.L. Perdrix, A.T. Mann and D.B. Humphreys MMI Technology, PO Box 822, West Perth, Western Australia 6872 Ph : 619 321 8999, Fax : 619 321 8103, Email: mmi@ozemaiLcom,au Summary 'Mobile Metal Ions' is a term used to describe metal ions that have been released from mineralization at depth, which have moved towards the surface, usually vertically, and which are loosely attached to surface soil particles. Six years of research and development both in the field and the laboratory have been undertaken. The process involves collection of soil samples and mixing with an optimum leachant designed specifically for commodity elements. After settling, the liquor containing the weakly- attached metal ions is analysed directly by ICP-MS. Mobil Metal Ion geochemical anomalies can be coincident with broader conventional geochemical responses, however, they are sharper and appear to precisely define primary or source mdneralization at depth. Of significance to exploration, this technique has also identified buried mineralziation covered by transported regolith units. MMI geochemistry has the potential to significantly reduce exploration costs particularly at the discovery phase of drilling programmes. Introduction Work commenced by undertaking 13 case studies predominantly within Australia to test the concept or phenomenon that has been previously termed the 'geo-gas phenomenon'. A number of exploration techniques including the Russian CHIM surveys have all endeavoured to measure these metal ion responses in surface soils. The initial case studies were undertaken in conjunction with mining companies over gold, nickel, and base metal deposits. All study areas had previously failed to respond to conventional geochemical surveys, or had given broad indistinct pattems that could not be related to buried mineralization. Results of three surveys are shown in Figure 1. These surveys were all completed over economic mineralization that had, are or were mined subsequently. Since the original work, in excess of 150 exploration studies from Australia, Africa, South America, Europe, and USA have been carried out in order to assess the technology for various commodity types, geological settings, regolith and landform situations, and climatic regimes. As a result of these orientation surveys and on-going research, it is now beHeved that Mobile Metal Ions are loosely attached or interstitial within the near surface soil particles. Soil particles above mineralization, may contain an element either in its metallic form, bound within iron-oxy-hydroxides, i.e. laterite, or associated with manganese oxides, carbonates, or silica. In addition, but at significantly lower concentrations, loosely attached to the soil particles of mobile metal ions have been identified. MobUe Metal Ions in Soils Evidence for the existence of Mobile Metal Ions is available. Soil samples identified as anomalous using MMI extractants were added to de-ionized water. As shown in Table 1, significant levels of commodity elements, including up to 15 parts per billion gold, can be simply "washed"firomthe soil into the water. However when de-ionized water is used as an extractant, poor repeatability usually results. Ions will readily come out of solution and attach themselves to organics, plastics, and glass containers. This produces particularly erratic results when water is used as -an extractant for exploration. Table 1 Metal Ions in De-ionized Water Zn Cd M

up to 15000 ppb up to 30 ppb up to 1000 ppb

Pb Cu Au

up to 200 ppb up to 50 ppb up to 15 ppb

The principle of Mobile Metal Ion Geochemistry is to place the soil into an extractant solution that allows the looselyattached metal ions to be taken up into solution. The extractants are weak, and as such do not selectively attack any specific substrate to access metals bound in iron or manganese components of the soils. This is essentially the major difference between Mobile Metal Ion Geochemistry and other partial digest techniques. Mobile Metal Ion Geochemistry is designed to provide a medium in which the metal ions can be collected and held. Digestants are specifically designed to prevent ions re-attaching themselves onto organics or hardware used in the laboratory. To date work has focused on 9 commodity elements, which included: Cu, Pb, Zn, Cd, Ni, Au, Ag, Pd and Co. Each metal has an optimum leachant. However, economic considerations, dictate that some grouping of the metals should occur for each leachant. Currently, two leachant solutions are used, the first to extract copper, lead, zinc and cadmium, and the second for nickel, gold, silver, palladium and cobalt. Work is continuing to expand the element suite. 79

MobOe Metal Ion Anomalies The main characteristics of Mobile Metal Ion Anomalies are: (1) (2) (3) (4) (5) (6) (7)

The anomalies are usually very sharp and can be contained within broader conventional anomaUes; They are directly above, or can be up-dip from primary mineralization as leakage anomalies; Commodity element responses are by far the most common; Generally, there are no significant path finders with the exception of cadmium which is sympathetic with zinc and cobalt with nickel; There is a very low per percentage of false anomalies, that is to say, Mobile Metal Ion Geochemistry generally does not give noisy responses; The anomaUes appear in soils, over weathered and transported overburden; and They are reproducible over time.

Other useful attributes in relation to the distribution of the metal ions within soils is the absence of any nugget effect and responses from around a sample site are always repeatable, even after rains. These characteristics are shown in the results given in Figure 2 which show the gold analyses for two samples composited and subsequently divided into four aliquots Application of Mobile Metal Ion Geochemistry Samples are collected at the surface and sieved to about minus five millimetre (-5mm). Where freight is not a major expense, generally 500 grams of each sample is collected. However, when freight costs are prohibitive, 100 grams of material'is required for each extraction. Samples are collected into plastic bags, and upon reaching the laboratory an aliquot is added to the relevant extractant, shaken and left to settle. After 24 hours an aliquot of the liquor is extracted and analysed using ICP-MS. An important step in using MMI data is to remember that it is geochemical data and as such quality control is of paramount importance given the low levels of detection that are significant. Check and duplicate samples are inserted to ensure that quality control is maintained. Once the data have been accepted, a background for each element population is calculated and the data normalized to that background by dividing each individual sample analysis by the chosen background value. The units used for Mobile Metal Ion Geochemistry are termed Response Ratios Discovery of The Golden Web Gold Deposit move to production, however MMI surveys have identified previously-undiscovered (but un-economic) gold, base metal, and nickel mineraUzation. The initial sampling at Golden Web was undertaken in January 1995. The mine area originally had an extensive, well-developed lateritic duricrust with a deep weathering profile to around 70 metres. This profile has been eroded, leaving a duricrust remnant to the north of the section and exposing the clay saprolite beneath. Continuing erosion formed a colluvial scree burying the previously-exposed paleo-surface. Subsequent rejuvenation of the drainage within the area has seen the encroachment of an alluvial plain from the south of the section (see Figure 3). A previous conventional sampling programme collected surface soils and analysed them for gold using fire assay. The remnant lateritic duricrust has elevated levels of gold, typical of iron-rich material near mineralization (see Figure 4). Moving along the traverse, responses from the source have been masked by the colluvial scree and alluvium encroaching from the south. There is, nevertheless, a conventional gold response within this area because of the mechanical transport of gold in lateritic material down-slope forming the colluvial fan. Previous companies had drilled this area but focused their attention on the elevated gold responses in the remnant laterite. An MMI survey was undertaken across this area, that had been identified as a priority area because of its prospective geology, structure, and magnetic trends. The area straddles the edge of a large granitic dome and its contact with a mafic/ultramafic sequence. The MMI survey, while having some response from the degradation of the lateritic duricrust further to the north, clearly responds to source primary mineralization at depth (see Figure 5). St. Francis Mining identified a response ratio level of greater than thirty (>30) times background to control its initial drilling programme. Eight (8) holes drilled to 40 metres were completed within the >30 response ratio contour. A summary of the intercepts greater than two metres (2m) with a grade above one gram per tonne (>1.0 g/t) is shown in Table 2. Of these 8 holes better than 60% of the holes achieved these parameters. It is interesting to note, that a fijrther 28 holes were drilled at a later stage between response ratio intervals of 5 and 15 times background, and that of these 28 holes, only one retumed an intercept of 8 metres at 4.4 g/t.

Table 2. Significant Intercepts from Golden Web - Initial Drilling Programme Hole Number

Intersection Width

Grade g/t Au

GWB 1 GWB2 GWB 5 GWB 7 GWB 8

12 m 4m 2m 12m 8m

4.9 12.5 2.7 3.9 12.0

Nepean Nickel Mine Metals Exploration Ltd, in joint venture with Freeport, held the Nepean area and discovered the Nepean nickel deposit located 26km south-southwest of Coolgardie in March 1968. No gossans were apparent at the site of the discovery and the successful diamond drill hole which intersected nickel sulphides was located on the basis of magnetic and induced polarisation geophysical anomalies. These strong anomalies were not related to the host rock or the nickel sulphides, but represented a broad complex ultramafic unit and sulphide-bearing metasedimentary rocks on the footwall of this unit. The intersection of the host rock and discovery of the nickel sulphides resulted fi-om a decision to continue the drill hole beyond the apparent target. The pre-mining ore reserve at Nepean was estimated at 1.2 million tonnes at 4.0% Ni with an average Ni/Cu ratio of 15. Nickel concentrate was produced between January 1970 and 1983 and from late 1985 to May 1987 and contained a total of 34,861 tonnes ofNi. During the early phases of coloration, surface soil sampling was undertaken. The presence of nickeUferous ultramafics and a lateritic duricrust 200m west of the mineralization gave an indistinct pattern for the Nepean deposit. The area was characterised by a broad anomalous response in Ni and Cu with elevated values reflecting lateritic remnants. The regolith within the area was further complicated by encroachment of alluvium and colluvium that masked the position of the mineralization at the surface. As a result, surface geochemistry failed to identify the minerahzation at Nepean. Recent exploration of the Nepean Nickel Mine area occurred been between March and July 1993, including MMI geochemical orientation studies carried out over the mine area and ultramafic units to the south and east. The survey involved a field inspection, sample collection, analyses and interpretation of results, which successfully identified the massive nickel mineralization (see Figure 6). Also outlined is the MMI survey nickel anomahes to the east and the southeast of the Nepean Mine Shaft in the interpreted Unit 4 and 5 positions - awayfi-omthe known ore zones. These positions do not appear to have been drill tested by the prior operators. At Nepean, MMI Geochemistry has demonstrated its ability, when compared to conventional geochemistry, to more precisely define Ni Mineralization at depth. Recent reconnaissance drilling of an anomaly east of the Nepean shaft (Figure 6) has reported significant Ni responses and confirmed the existence of prospective host lithologies. Application to Exploration To date, the main application of Mobile Metal Ion Geochemistry has been for more precise drill target definition, reducing or eliminating broader drilling programmes and allowing re-direction of drilling dollars toward reserve definition. Because of the distribution of metal ions within the soils, at a regional scale, companies are now compositing individual sample sites for up to 400 metres (4 x 100m sample sites), and retaining an archived sample from each site (see Figure 2). When the composite sample retums an anomalous result the individual archives can then be retrieved quickly and analysed to provide a more distinct geochemical pattem. Conclusions While Mobile Metal Ion Geochemistry has demonstrated some remarkable capabilities, it should not be regarded as a panacea for exploration. MMI geochemistry does not work in all regolith situations, but it has demonstrated its effectiveness in terrains where conventional geochemistry has failed. Where regolith profiles are mature with residual soil components, both conventional and MMI anomalies can be detected; however, MMI responses are more tightly constrained and occur directly over mineralization (see Figure 7). Where there is an influx of transported or depositional units, the profile has an immature regolith that masks the previous geochemical signatures. In time, the geochemical anomalies begin to re-establish themselves in the soil profile. The time taken for metal ions to penetrate the recent, near surface material and start accumulating is short when compared with the time required for pedogenic processes to re-establish conventional geochemical anomalies. This may provide some explanation as to the success of MMI in depositional terrains where conventional geochemistry has failed. Research work is continuing at the Geochemistry Research Centre (a private enterprise research facility) supported by Industry and the West Australian State Government. The research is addressing important aspects that have been

identified by previous projects and aims to increase the level of understanding on the release, transport, and integration of metal ions in the soil profile. Armed with a better understanding, the application of Mobile Metal Ion geochemical surveys for mineral exploration may by expanded into even more difiScult depositional terrains and downhole. Even so, the technology is already having a positive impact on the exploration industry.

Base Metals q 100

5a 8060

11 uverburden Depth 11 >300m 1 (Transported 1 - 4m) P 1*3

40 Pi 20 0

'r Metres

Gold 50

1 40 Overburden Depth sC8 30

20 s* in 10

ifi

(Transported 1 - 3m)

^^^ ^^^

^^^ #

^^^ ^^^

Metres

Nickel o

35 30 25

09 20 O 15 So

s- 10 5

A 1 2 S1 ^ ta 1

"I"

^ ^ # Metres #

Overburden Depth A iOfkn B"250ffi (Transported l--3m> -i

1 #

Mobile Metal Ion Distribution

Sample 1 17.6 ppb Au

Sample 2 46.8 ppb Au

1 3ub-sam 1

1 iiiip

i i i i>-sam§^

31.8 ppb

32.6 ppb

31.2 ppb

>ub-5 :am%

\ ^

32.9 ppb

Golden Web Prospect Coolgardie, Western Australia

Northing

Golden Web Prospect Conventional Geochemistry (Fire Assay)

0.5 0.4 E 0.3 Q. Q. < 0.2 0.1

Northing

^ ^ Source of Gold Mineralisation

Golden Web Prospect MMI Geochemistry 80

Lateritic Duricrust

Lateritic Colluvial Scree

'•S2

c o Q. CO a>

Northing 00

< *

•f & f ^ Source of Gold Mineralisation

MMI NEPEAN ORIENTATION SURVEY 5100N

Drill indicated limits of significant Nickel mineralization (>1m width >2%) between 50-450m depth.

4900N

MMI Anomaly 4750N

H 4550N

4300N

Scale

400m

Nepean Shaft

MMI Surface Geochemistry Residual soil profile MMI Partial Digestion Precise Anomalies vertically above mineralization

Conventional Geochemistry Broarder anomalies often displaced awav from mineralization

MESOZOIC EVOLUTION OF THE NORFOLK RIDGE SYSTEM: EVIDENCE FROM NEW CALEDONIA AND NORTHERN NEW ZEALAND Philippa Black Department of Geology, University of Auckland, Private Bag 92019, Auckland New Zealand Summary Pre-Cretaceous basement terranes exposed in the Norfolk Ridge System are described and where possible correlated with the terrane sequence of the South Island of New Zealand. Two groups of terranes are recognised in the Norfolk Ridge : Permian - Jurassic volcanic arc related terranes (Brook Street and Murihiku) and terranes in an accretionary complex (Waipapa and Caples). The Torlesse terrane, which outcrops extensively in the South Island wedges out in the North Island of New Zealand, and is not a component of the Norfolk Ridge basement. In New Caledonia a poorly understood terrane in the accretionary complex, and known locally as "ante-Permian", may possibly be correlated with one of the terranes in the mid to late Paleozoic complexes of the New England fold belt. The time of suturing of the arc-related and accretionary wedge terranes may vary along the ridge system and in places could be as late as early Tertiary. Introduction The Norfolk Ridge System is an assemblage of well defined terranes accreted to the eastem rim of the Australian plate during the Permian through to early Cretaceous, and fragmented and transported eastward by the opening of the Tasman and New Caledonian basins during the late Cretaceous. While the major part of the Norfolk Ridge is known only from oceanographic and particularly geophysical studies (1) the northem and southem parts of the ridge are well e^qposed in New Caledonia and northem New Zealand. The terrane classification of New Zealand Paleozoic - late Mesozoic rocks, based on the tectonostratigraphic associations of pre-Cretaceous rocks of the South Island, is now well estabhshed and accepted (Figure 1). A Western (or Foreland) Province of Gondwana terranes and granitic complexes is sutured along a Median Tectonic Line to an Eastem Province dominated by volcanic arc and subduction complex terranes. However the South Island, a fragment of Australian plate, and lying on the extension of the Lord Howe Rise, is not part of the Norfolk Ridge and the assemblage of the basement terranes recognised along the Norfolk Ridge is in some important respects quite different to the assemblage recognised in the South Island. While it is generally accepted that the Murihiku terrane extends north into New Caledonia other terranes do not and some South Island Eastem Province terranes do not have counterparts even in the northem North Island. Basement rocks exposed along the Norfolk Ridge can be divided into : (i) (ii)

Terranes representing Permian - Jurassic volcanic arcs and arc-related sediments. These rocks are fossiliferous, contain thick coherent sequences, are well known stratigraphically, and weakly or unmetamorphosed. Accretionary complexes (subduction wedges) of mainly terrigenous volcaniclastic sediments, locally extensively deformed and melange-hke, containing slices of material from diverse sources (including ocean floor material) and of varying ages. All of these rocks are metamorphosed, some polymetamorphosed, and locally high pressure (lawsonite-bearing) assemblages are preserved.

The volcanic-arc related sequences and accretionary complexes are sutured together along well recognised fault zones which have remained active until the mid-Tertiary. The accretionary complexes have been overprinted by a late metamorphism which decreases in grade away from the suture zone (3,4). Although the similarities are striking there are also significant differences in the basement terranes assemblage of the southem and northem portions of the Norfolk Ridge. Southern Norfolk Ridge - Northem North Island Tuflfaceous and volcaniclastic sediments of the Murihiku terrane outcrop along the west coast of the North Island and have been found in drill holes in Northland and the Auckland area. Brook Street terrane is not exposed but is believed to exist beneath the voluminous Tertiary deposits westward of the Murihiku and beneath the Taranaki basin. In the South Island the Murihiku terrane is juxtaposed on its eastem margin against the Dun Mountain terrane (an obducted ophiolite sequence) which in turn is separated by a suture zone (Livingston Suture) from the accretionary wedge (Caples and Torlesse terranes). In the North Island a magnetic anomaly (Stokes Magnetic Anomaly) coinciding with the boundary between the Murihiku and Waipapa terranes and rare small serpentinite outcrops suggests that the Dun Mountain Ophiolite Belt may be present although on a very reduced scale.

In the accretionary wedge, some Caples terrane has been recognised by Black (5) but the bulk of the material is Waipapa terrane. The Torlesse terrane, composed of remarkably uniform arkosic sediments and their metamorphosed equivalents within the Haast and Alpine schists, which outcrop extensively in the South Island and southeastem North Island are tectonically interleaved with Waipapa terrane sediments in the central North Island, and not recognised with any certainty in the northern North Island. Geochemical, petrologic, lithologic associations and age relations suggest that the Waipapa terrane contains two facies (subterranes) - the "Hunua facies" containing ocean floor material as well as the volcaniclastic sediments, and a coarser-grained, younger and less deformed "Morrinsville facies", lacking ocean floor material (6). The sequence in the North Island has been described by Black (5). Much of the "Hunua facies" is extensively deformed and lawsonite has been found in shales from a melange zone in Northland (6). While some older Torlesse may occur interleaved with melanges that typify the "Hunua facies" the majority of sediments are volcaniclastic. The younger Jurassic-Cretaceous accretionary complex possibly correlates with the Pahau-Mata terrane. While the Bradshaw's subdivision (2) of subduction wedges into Permian - Triassic and Jurassic ~ Cretaceous accretionary complexes is broadly correct, there is no clear break and accretion was probably generally continuous from the Permian to the early Cretaceous. The timing of the association of the Murihiku and the Waipapa terranes is uncertain. In the North Island, sedimentation in the Murihiku continues until the uppermost Jurassic. There are no Cretaceous sediments which are transgressive on either the Murihiku or the Waipapa terranes. The first sediments known to be directly deposited on the northem North Island Mesozoic basement terranes are the mid-Eocene to Oligocene Te Kuiti Group sediments which include the Northland and Waikato coal measure basins developed in large part along the sutures zone separating the two terranes. It seems possible that the terranes may not have been juxtaposed until the early Tertiary. Northern Norfolk Ridge - New Caledonia Equivalents of the Murihiku terrane have long been recognised in westem New Caledonia. Also recognised is a succession of intermediate arc volcanics and fossiliferous Permian epiclastic sediments. The Permian sequence has been correlated with the Brook Street terrane on the basis of tectonic position and general lithologic and stratigraphic similarities with similar rocks in the South Island. No rocks similar to the Dun Mountain terrane have been recognised in New Caledonia. The two arc-related terranes outcropping along the southwest coast are separated by a distinct tectonic boundary (suture) from the poorly known "greywacke" terranes of the Central Chain of New Caledonia. The New Caledonian central chain "greywackes" are closely similar petrographically to the sediments of the Waipapa terrane and particularly to the "Morrinsville facies". They are tectonically compHcated, containing different tectonostratigraphic packets of rocks. The Caples terrane has not been recognised although Permian volcanics and associated sediments in the accretionary complex in the southem part of the Central Chain may be their equivalent. The Torlesse terrane is unknown in New Caledonia. Within the Central Chain greywacke complex Paris (3) recognised a generally age related distribution of volcaniclastic sediments away from the suture. Permian material is known interleaved with Triassic volcaniclastic sediments in the south while in the north the sediments are Jurassic in age. Tectonic units known as the "ante-Permian" (3) occupy the tectonic position of the Caples terrane but the lithologic similarities to rocks described as Caples in the South Island of New Zealand is not strong and the "ante-Permian" is probably better correlated with one of the terranes of the mid to late Paleozoic complexes of the New England fold belt. Relics of blueschist metamorphism are common in the accretionary complex. In New Caledonia Murihiku sedimentation ceased in the mid Jurassic although in sedimentation in the Central Chain terranes continued into the early Cretaceous. A late Cretaceous extensional regime related to the fragmentation of the Australian plate boundary resulted in formation of extensional basins in which carbonaceous sediments were deposited directly on both the Murihiku and the Central Chain greywackes. Consequently in New Caledonia the two groups of terranes were juxtaposed before the late Cretaceous. Discussion While there are differences in the nature of the basement terranes, and possibly also in the timing of their accretion, the general similarities between the terrane associations along the Norfolk Ridge are striking and they are clearly distinct from those recognised in the South Island. Neither the granite-derived Torlesse nor the obducted Dun Mountain (ophiolite belt) terrane are known other than as a minor component in the southernmost portion of the Norfolk Ridge. The material of the subducted wedge component of the Norfolk Ridge is largely volcaniclastic and oceanic derived, although of mixed sources and correlated with the Waipapa terrane. The timing of the accretion of the individual packets of material within the subduction complex is uncertain and will remain so until there is better stratigraphic control on individual units. In the case of the arc-related terranes, geochemical evidence from the Brook Street and Murihiku terranes of the South Island indicates no similarities so the two terranes cannot have been associated until after Murihiku sedimentation ceased (in the late Jurassic in the North Island). The time of suturing of the volcanic arc - related terranes and the subduction wedge is of particular interest. The Brook Street and Murihiku terranes are very weakly metamorphosed. On the other hand the terranes that form the accretionary wedge are metamorphosed (up to greenschist facies) and discrete packages with different grades of metamorphism are 91

recognised indicating that metamoiphosed occured prior to accretion. A later metamorphism has been superimposed on the whole accretionary wedge and this metamorphism clearly decreases in grade to the east and awayfromthe suture (3, 4). In New Caledonia the metamorphism and suturing must have been mid Cretaceous but in northem New Zealand it could have been later and possibly even in the early Tertiary. Acknowledgments I am indebted to Jack Grant-Mackie for many useful discussions. References 1. Eade, J.V. The Norfolk Ridge system and its margins. In Naim, A.E.M., Stehli, EG. and Uyeda, S. (Eds) The Ocean Basins and Margins. Plenum Press, New York, 1988, 303-324. 2. Bradshaw, J.D. Cretaceous geotectonic pattems in the New Zealand region. Tectonics 8,1989, pp 803-820. 3. Paris, J-P. Geology de la Nouvelle-Caledonie. Memoire du Bureau de Recherches Geologiques et Minieres 113, 198l', 278pp. 4. Black, P.M., Clark, A.S.B and Hawke, A. Diagenesis and very low-grade metamorphism of volcaniclastic sandstones from contrasting geodynamic environments. North Island Murihiku and Waipapa terranes. Joumal of metamorphic geology 11,1993, pp 429-435. 5. Black, P.M. The Waipapa Terrane, North Island, New Zealand : subdivision and correlation Geoscience reports of Shizuoka University 20, March 1994, pp 55-62. 6. Kear, D. Basement rock facies - Northem North Island. New Zealand Joumal of Geology and Geophysics 14, 197l',pp 275-283. 7. Swain, A. Structure in Waipapa and Waitemata Group rocks near Leigh. Unpublished MSc thesis. University of Auckland Library, 1993.

Figure 1. Simplified Pre-Cretaceous terrane map with main morphological features and stmctural features oflhe region adjacent to the Norfolk Ridge system. Dotted heavy lines along the Norfolk Ridge are magnetic anomalies after Eade (1). Terranes in the New Zealand region after Bradshaw (2) and in New Caledonia derived from the geological synthesis of Paris (3). Lighter dashed lines are terrane boundaries. MTL = Median Tectonic Line; B = Brook Street terrane; M = Murihiku terrane; DMOB = Dun Mountain terrane (the Livingston Suture is the eastern boundary, between the Dun Mountain and Caples terranes); C= Caples terrane; T = Torlesse terrane; P = Pahau and Mata terranes (known sometimes as younger Torlesse), Wj^ = Waipapa terrane ("Hunua facies"); Wj^ = Waipapa terrane ("Morrinsville fades"); AP = "ante-Permian" of Paris (3).

INTERNAL EVOLUTION AND METALLOGENY OF PERMOTRIASSIC HIGH-K GRANITES IN THE TENTERFIELDSTANTHORPE REGION, SOUTHERN NEW ENGLAND OROGEN, AUSTRALIA.

Phillip L. Blevin and Bruce W. Chappell National Key Centre for Geochemical Evolution and Metallogeny of Continents, Geology Department, Australian National University, Canberra ACT 0200 Summary The high-K granites of the Tenterfield-Stanthorpe region in the southern portion of the New England Orogen form a distinct group within the larger I-type Moonbi Supersuite. The granites comprise a textural and compositional continuum from coarse grained, sphene bearing monzogranites to the highly evolved Ruby Creek Granite and its equivalents. Compositional variation in the more felsic members was controlled by fractional crystallisation mechanisms. Textural and chemical evidence suggests that these magmas may have also been volatile saturated at the time of emplacement and crystallisation. Mineralisation ranges from Mo and base metal dominated to Sn dominated systems in the morefractionatedrocks. The association of Sn mineralisation with oxidised I-type magmas is not inconsistent with fractionation-redox controls on ore element behaviour in magmatic systems provided the crystallisation of potential Sn sequestering phases is suppressed or delayed relative to the onset of volatile saturation. Introduction The granites (sensu lato) of the southem New England Orogen (sNEO) were subdivided by Shaw and Flood (1) into several plutonic suites: three I-type (Clarence River, Uralla and Moonbi), and two S-type (Hillgrove and Bundarra). Several economically important and voluminous granites however were not classified into specific suites because of their fractionated and felsic compositions. These granites were assigned to a separate "leucoadamellite" grouping. Chemical and petrographic studies on the granites of the this group in the Tenterfield-Stanthorpe region, including the Stanthorpe Granite, Ruby Creek Granite and various felsic granites in the Bolivia-Dundee area show them to be chemically very similar to the Moonbi Plutonic Suite of Shaw and Flood (1), and have subsequently been included into the Moonbi Supersuite of Chappell (2). These granites, while compositionally similar to those in the Moonbi area to the south are distinct in several ways (see below). For this reason the Moonbi Supersuite is herein informally subdivided on spatial grounds into two groups: the Moonbi Granite Group (MGG), comprising the Moonbi Supersuite granites in the Tamworth region, and the Stanthorpe Granite Group (SGG) comprising the Moonbi Supersuite granites in the Tenterfield-Stanthorpe region (Figure 1). Other members of the Moonbi Supersuite to the east of Glen Innes appear to be similar to those of the SGG. The Stanthorpe Granite Group The SGG can be subdivided on petrographic and chemical criteria into three main "types" and one subtype (Figure 1 and Table 1): Bungulla Type: This comprises coarse grained, sphene-bearing granites characterised by prominent rectangular pink (rarely white) K-feldspar megacrysts. Mafic microgranular enclaves of dioritic composition are often abundant. Hornblende is present in addition to biotite. Clinopyroxenes may be present as inclusions within enclave plagioclases. Kfeldspar megacrysts may range from 5 to 30 percent of the rock and vary in length from less than 2cm to 10cm. Named and mapped units included in this type are: Undercliffe, Bungulla, Maryland and Bookookoorara. Unnamed exposures in the Stanthorpe Granite to the south east of the Undercliffe Falls unit are also included. Stanthorpe Type: This is the most extensive type and is characterised by coarse to medium grained equigranular to mildly porphyritic granites with biotite » homblende. K-feldspars are typically pink, though locally can be white. Flow alignment of K-feldspar megacrysts are present near extemal and internal contact zones. The Stanthorpe Type is texturally transitional to the Bungulla Type in places. Homblende is restricted to inclusions within plagioclase aggregates. The Stanthorpe Type may be texturally variable over small distances with a variety of textural types commonly present within any one area. The Bolivia Subtype is very similar to the Stanthorpe Type and occurs as discrete mapped units in the south of the study area (Figure 1). This type is typically finer grained, may be slightly more porphyritic, and has a higher biotite/homblende ratio than the "average" Stanthorpe Type.

PhitonVkx:atk)ns mentioiied m the text 1 Bolivia Range 2 Mount Jonblee 3 Peiiics Sugarloaf 4 Pycs Creek 5 Red Rock 6 Kilminster 7 Ruby Creek 8 Undercliffe Falls 9 Jibbinbar N

Chemcal-petiograpiiic subdivision of the StandKxpe Granite Groiq> Associaied Min^alisadon *

Sn (± W, polymetallic)

Mo(±WBiAu)

ixriymetallic (As Cu Ag Pb Zn Sn Mo cic)

Ruby Creek Type Bolivia Subtype Standiwpe Type BunguIlaType

Figure 1. Map showing the distribution of granite types within the Stanthorpe Granite Group of the Moonbi Supersuite. Ruby Creek Type: This comprises medium to fine grained equigranular pale pink to white leucocratic granite, as well as texturally diverse porphyritic, aplitic, microgranitic and pegmatitic phases. Modal biotite contents are usually percent or less and homblende is absent. A wide variety of accessory minerals are present in thin section: fluorite, chlorite, muscovite, carbonate, zircon, tourmaline, magnetite, ilmenite, cassiterite, wolframite and various sulfides (pyrite, molybdenite, chalcopyrite, arsenopyrite). Sulfides (chiefly pyrite and molybdenite) often occur as disseminated spots throughout the rock with little or no macroscopic evidence for hydrothermal alteration. Miarolitic variants of the Ruby Creek Type are common. These rocks often have a two domain texture comprising a coarser anastomosing and extemally nucleated domain surrounding enclosed, intemally nucleated aplitic domains. This texture was termed an "interconnected miarolitic texture" (IMT) by Candela and Blevin (3) and is interpreted as evidence for magmatic volatile phase permeability during crystallisation. These phases are often associated with sulfide minerals occurring as disseminations, cavity infill, and in cross cutting quartz veins.

Table 1: Internal subdivision of the Stanthorpe Granite Group Rb/Sr CeNA^ Apsat bi/(bi +hb) Si02-% (av, Is) av (range) 1.0 787,17 1.6 35 Ruby Creek 76.6 ±0.3 (20-100) Type

Bolivia Subtype

75.1 ±1.3

3.7

Stanthorpe 74.6 ±1.9 Bungulla Type

68.4 ±2.4

0.6 (0.3-1)

6.0

Plutons

(TJZ, Is)

mineralisation

Sn; W-Sn; SnRuby Creek W-Mo-As-Bi; AgRed Range As-base metal Jibbinbar unnamed dykes and masses pipe and vein Mt Jonblee Wo-W-Bi-Au Petries Sugarloaf Bolivia Range Nonnington Pyes Creek Stanthorpe 875,40 0.75-1.0 disseminated Mo Billyrimba Clive Mt Makenzie none <0.75 Underclifife Falls 939, 26 Bungulla Bookookoorara Maryland

Spatial and age relationships within the SGG Dating data on the New Engalnd Granites has recently been summarised by Shaw (4). The Bungulla Type has been dated at around 245 Ma while those of the Ruby Creek and Stanthorpe Types (including the Bolivia Range, Mount Jonblee, Mackenzie and Nonnington plutons) are indistinguishable from each other at around 238-244 Ma (average 240 Ma). These ages are Rb-Sr biotite ages and assume a uniform whole rock initial Sr ratio for all units of 0.705. The dates are consistent with the essentially coeval nature of all three types and conform to intmsive relationships where they are exposed. Granites of the MGG are somewhat older (--250 Ma). The Bungulla and Stanthorpe Types usually occur with the foraier located in topographic lows and the later in topographic highs and/or in elevated areas. Textural gradations between the two may occur particularly where significant vertical exposure is present (eg. the sides of the Rocky River gorge in the Timbarra area, south east of Tenterfield). The Bolivia Subtype forms topographic highs relative to the Bungulla Type in the southem portion of the SGG. The presence of intemal contacts, and its areal dimensions suggests that the Stanthorpe Granite as currently mapped represents a series of intrusive pulses. Magnetic and chemical variations suggest that the main mass of Stanthorpe Granite around the town of Stanthorpe, and the mass to the east of the "screen" formed by sediments and the UndercUfife Falls and Maryland intrusions may be compositionally slightly different. The Ruby Creek Type occurs mostly within the Stanthorpe Type as small stock-like intrusions, dykes, and segregations with gradational contacts. Intrusions of the Ruby Creek Type occurs into the overlying sediments as small mineralised plugs (Red Rock, Jibbinbar) or dykes (Kilminster) to the west of Stanthorpe. Some units mapped as Ruby Creek Granite, particularly those west and north west of Stanthorpe, may represent contact zone variants of Stanthorpe Type intrusions, indicating that the roof zones of these bodies are only just being exposed in these areas. Petrographic and chemical aspects The SGG shows an intemal continuum of decreasing homblende/biotite ratios, opaque and sphene contents with increasing Si02. Amphibole is often present within the Stanthorpe Type only where it occurs as inclusions within plagioclase. These rocks are generally too felsic (-75 percent Si02) for amphibole to have been a stable magmatic phase, and the amphibole and their plagioclase host crystals must represent earher crystallised material brou^t in or mixed with Stanthorpe Type magmas during emplacement. Primary sphene decreases rapidly with increasing Si02 to be absent in rocks with Si02 greater than -11 percent. This reflects decreases in the Ti02 and CaO content to very small amounts in these rocks rather than to changes in oxidation state. The compositional variations between the Bungulla, Stanthorpe and Ruby Creek Types is consistent with them being part of a differentiating magma series, with compositional variations in the more felsic members being controlled dominantly by feldsparfractionation(Figure 2). Slight differences between the Bungulla and the Stanthorpe Types may represent the presence of earlier crystallised and/or cumulate material in the former. The Bolivia Range subtype is compositionally indistinguishable from the Stanthorpe Type. The Ruby Creek Type has a mineralogy consistent with crystallisation from 96

near minimum melt composition and is distinctly more fractionated than the Stanthorpe Type (Figure 2), although the overall compositional continuity is apparent.

100

• + O A •

Bungulla Type Stanthorpe Type Bolivia Subtype Ruby Ck Type Kilminster dykes

C/D

Si02 Figure 2. Plot of Rb/Sr versus Si02 for the Stanthorpe Granite Group. Note the compositional continuum. The Ruby Creek Type is anomalously enriched in some ore elements and S. Some of these are explicable in terms of fractional crystallisation and the retention of certain ore element (eg. Sn, Pb) in the residual melt fraction. Others (Zn, Cu, Pb, S) are negatively correlated with Si02 in the Moonbi Supersuite overall, but occur in elevated amounts in half of the Ruby Creek Type samples (see analyses of Ruby Creek Type in Table 2 with S greater than 0.02 weight percent). These trends are not consistent with fractional crystallisation, and their association with texturally variable and IMT bearing rocks indicate that these elements have been introduced hydrothermally at the magmatic stage by a discrete magmatic volatile phase. Comparison with the MGG: The SGG and MGG differ by the former having higher sphene and opaque contents and biotite/homblende ratios for any given composition. Lower ferromagnesian contents for the SGG is in accord with a higher proportion of Fe (and Ti) present in magnetite, ilmenite and sphene relative to the MGG and is consistent with the higher magnetic susceptibilities and whole rock Fe203/Fe0 values of the SGG relative to the MGG (Table 2). The MGG differs compositionally from the SGG in being as a group less felsic, with strongly fractionated rocks being absent. The MGG also has higher K2O, MgO, MgO/FeO*, Ba, Sr, Pb, Th, Ni and As, and lower Ti, Na20, Y than the SGG (Table 2). The Moonbi Supersuite as a whole shows a marked decrease in LREE/HREE (as proxied by CeAO ratios with increasing Si02. The MGG in many ways represents exaggerations of those compositional traits that set the Moonbi Supersuite apart from other I-type supersuites in the New England Orogen. The higher MgO/FeO* ratio and lower JO2 of the MGG relative to the SGG is probably responsible for the lower biotite/homblende ratios and lower sphene and magnetite contents in the former.

Table 2: Average chemical analyses of Stanthorpe and Moonbi Group Granites compared. STANTHORPE GROUP MOONBI GRANITES Ruby Oc Stanthorpe Ruby Ck Bungulla Moonbi Type Type Type Type Group1 (S < 0.02%) (S > 0.02%) 5 5 70 22 48 No analyses 76.56 76.57 74.62 68.43 68.73 Si02 0.06 0.07 0.19 0.49 0.44 Ti02 12.30 12.47 12.96 14.56 14.59 A1203 0.44 0.60 0.71 1.73 0.92 Fe203 0.39 0.28 0.71 1.45 2.09 FeO 0.03 0.04 0.03 0.08 0.08 MnO 0.05 0.06 0.29 1.25 1.41 MgO 0.42 0.45 0.93 2.82 2.85 CaO 3.46 3.85 3.60 3.51 3.38 Na20 5.08 4.63 4.68 4.01 4.36 K20 0.01 0.01 0.05 0.17 0.17 P205 0.05 <0.02 0.01 0.02 0.00 S 27.8 44.0 225.3 525.9 815.0 Ba 335.2 324.2 259.4 174.3 214.5 Rb 9.6 13.0 82.2 333.5 363.0 Sr 34.2 31.2 24.4 20.2 43.3 Pb 42.9 42.7 35.3 27.3 35.9 Th 13.3 12.3 9.0 6.3 8.4 U 116.6 112.2 141.6 182.8 180.4 Zr 13.7 15.9 11.8 8.6 10.5 Nb 65.4 67.4 25.4 47.9 18.8 Y 14.8 18.8 29.3 29.6 30.2 La 42.6 51.6 68.9 63.0 66.2 Ce 3.0 4.2 4.8 9.9 7.6 Sc 0.4 0.6 10.5 50.0 49.8 V 0.0 0.0 1.7 20.7 29.5 Cr 228.0 242.0 310.2 593.4 640.8 Mn 1.8 1.8 3.3 7.8 9.6 Co 0.0 0.0 0.7 5.6 10.1 Ni 18.0 2.6 1.2 4.8 8.9 Cu 38.4 33.8 29.2 49.9 49.3 Zn 17.6 17.7 16.3 15.9 16.2 Ga 3.6 0.3 3.0 2.9 5.9 As 13.8 11.0 6.1 4.1 3.6 Sn 2.1 1.0 1.2 0.4 Fe203/Fe02 34.9 24.9 3.16 0.52 0.59 l.lRb/Sr2 Notes: 1. Inlet Monzonite not included; 2. ratios calculated from average values. Level and temperatures of emplacement of the SGG Several lines of evidence suggest that the SGG has been emplaced to very shallow crustal levels. The SGG is spatially associated with contemporaneous volcanic rocks, into which the SGG intrudes. Candela and Blevin (3) have estimated that miarolitic textures in granitic rocks require pressures of around 2 kb or less. Both the Ruby Creek Type and aplites within the Stanthorpe Type have normative compositions that lie between the 500- and 1000-bar minima of the haplogranite system of Tuttle and Bowen (5). Lack of significant shifts in bulk composition towards higher normative Qz or Ab reflect the lack of significant magmatic enrichments of F, B or CI. Apatite and zircon saturation temperatures of the SGG strongly converge to around 760°C to 780°C for both the Ruby Creek Type and for aplites in the Stanthorpe Type. These temperatures correspond closely to the expected temperature range of crystallisation for melts of this composition at low pressures. Amphibole barometry in the Stanthorpe and Bungulla Types reveal two populations. One group, occurring as inclusions in plagioclase within the Stanthorpe Type and as ragged crystals in the Bungulla Type have A1 contents (SAl per 13 cations) of 1.3 -1.5. The second group comprises small euhedral crystals within the Bungulla Type and have A1 contents of 0.8 - 0.9. Al-in-homblende barometry calculated using the equations of Anderson and Smith (6) indicates calculated 98

pressures for these values of 2.8 - 3.4 kb and kb respectively, assuming afixedtemperature of 700°C. If the higher A1 group of amphiboles crystallised at higher temperatures, calculated pressures would be lower (eg. kb at 800°C). Metallogeny of the SGG. Sn deposits: Despite significant alluvial Sn production (-'50,000t) in the Stanthorpe region shed from the Ruby Creek Type, hardrock workings for Sn have been trivial and mainly focussed around Sundown (Red Rock) where subeconomic Sn and base metal mineraUsation is associated with a small boss of Ruby Creek Type granite. Hard rock tin mineralisation in the Ruby Creek Granite occurs as quartz vein stockworks and disseminations associated with wolframite and arsenopyrite. Mo-deposits: Molybdenum mineralisation occurs as pipe-Uke deposits (such as those associated with the Mount Jonblee Leucoadamellite) and in quartz veins and disseminations in granite, pegmatites and aplites. MineraUsation of this latter type is widespread in the Stanthorpe Granite and the Bolivia Range, Pyes Creek and Nonnington Leucoadamellites. The larger pipe and greisen hosted deposits were producers of Mo and minor Bi and Ag. Associated elements and minerals included molybdenite, native bismuth, various bismuth tellurides and sulphides, pyrite, chalcopyrite, covellite, silver minerals, sphalerite, galena and cassiterite. Gold was recorded in several of these deposits by Jones (7), and was probably mainly associated with Bi minerals. The form of these deposits and their mineralogies are very similar to those associated with the Kingsgate and Red Range Granites, both of which are probably members of the Moonbi Supersuite. A feature of the Stanthorpe Granite is the presence of molybdenite flakes alongfiracturesof the granite, over large areas. Molybdenite along these fi^ctures may be associated with sulfide minerals and can fomi very elaborate radiating crystal groups. Molybdenite and sulfide also fills miarolitic cavities in the Ruby Creek Granite along with carbonates, fluorite, quartz, chlorite etc. Quartz precipitation is often absent along thefinerfi-acturesin otherwise "fi-esh granite" although the quartz adjacent thesefiracturesmay be smoky, indicating the hydrothemial transport and deposition of radioactive elements. Gold and base metals: Alluvial mining for Au has occurred in the Surface Hill area east south east of Tenterfield and is associated with molybdenite occurrences. Gold was also reported from alluvial Sn workings sp)atially associated with exposures of Ruby Creek Type granites in the Stanthorpe-Ruby Creek areas by Saint Smith (8). Polymetallic (Pb-Zn-CuAs-Ag-Bi-Mo) sulfide minerahsation, often associated with cassiterite and wolframite, is associated with dykes of Stanthorpe Granite aflSnity in the Kihninster area to the west of Stanthorpe. In this general area molybdenite, cassiterite and wolfi-amite also occur as disseminations and quartz veins along the westem contact zone of the Stanthorpe Type usually in marginal variants. Arsenic has also been producedfromthe Jibbinbar area. Discussion Studies, for example by Blevin and Chappell (9), that subdivide igneous rock types on the basis of compositional parameters such as the degree of fractionation and oxidation state recognise a fundamental subdivision of ore metals associated with oxidised igneous rocks (Cu, Au and Mo±W) and those associated with reduced rocks (Sn, W). The MGG is associated with ore elements typical of relatively oxidised igneous rocks only (Cu-MoW-Au). An apparent contradiction is present in the metallogeny of the SGG where both Mo and Sn minerahsation is present. During fractional crystallisation of magmas compatible elements are removed from the melt fraction via substitution into crystalhsing mineral phases. In oxidised magmas the Sn^VSn^^ ratio is relatively high. Sn^ is able to substitute for Ti and Fe'^ in sphene and ferromagnesian sihcate minerals. In reduced magmas the Sn^VSn^" ratio is lower, and consequently Sn behaves more incompatibly and as a consequence is preferentially concentrated in the meh fraction. Mo in oxidised magmas is relatively incompatible while in reduced magmas Mo has lower average oxidation state (--4+) and is able to substitute for Ti and Fe in various Fe and/or Ti bearing phases. Thus in relatively oxidised magmas such as those of the Moonbi Supersuite, Mo should be incompatible relative to Sn and be available for partitioning from residual melts into exsolving fluid phases. Such mechanisms of ore element enrichment or depletion through selective melt-crystal partitioning requires the presence of early formed appropriate crystalline phases. Mo dominant mineralisation is associated with the Stanthorpe Type and Bohvia Subtype, while Sn (±W, Bi, Mo, As) mineralisation is only dominant in the more fractionated Ruby Creek Type. In the Bungulla and Stanthorpe Types the presence of early crystallising sphene, biotite and amphibole was capable of removing Sn but allowing Mo to build up sufficiently to generate Mo dominated mineralisation in hydrothermal systems associated with the Stanthorpe and Bolivia Range Types. In the Ruby Creek Type, primary sphene is not present, and biotite and Fe-Ti oxides are late crystallising. This, in addition to significant degrees of fractional crystallisation, causes most ore elements to behave incompatibly as there are no appropriate hosts (apart from quartz and feldspar) into which they can be sequestered. Tin mineralisation is absent from the MGG because highly fractionated members of the MGG are not present. A model for the chemical, petrographic and metaUogenic evolution of the SGG. The abundance of enclaves within the Bungulla Type, and the chemical association of the Stanthorpe Group with the less evolved Moonbi Supersuite, indicate that the Stanthorpe Group was ultimately derivedfroma more mafic, high-K source. This source was also isotopically unevolved and relatively oxidised. Compositional variations within the felsic members 99

of the supersuite were controlled dominantly through fractional crystaUisation mechanisms. Early phases (Bungulla Type) were the compositionally least evolved although a textural and compositional spectrum between the Bungulla and Stanthorpe Types is apparent. The intrusions were emplaced to high, probably subvolcanic levels in the crust. Discrete intrusions with relatively small cross sectional areas relative to their vertical extent (ie. those with a relatively high aspect ratio) provide a good focus in their apical regions for exsolved hydrothermal fluids. These conditions appear to have been met in the Bolivia Subtype intrusions where pipe-like Mo-dominant mineralisation occurs in the upper margins and apical portions of plutons such as Mount Jonblee and Bolivia Range. The plutons comprising the Stanthorpe Granite Type in the Stanthorpe region however were emplaced as areally extensive bodies (low aspect ratios) which do not provide a good focus for exsolving hydrothermal fluids. Consequently mineralised fluids were uselessly dispersed throughout the intrusions as disseminations and along fractures and joints. Extended fractionation within the Stanthorpe Type magma chambers lead to the generation of the more highly evolved Ruby Creek Type magmas. Some underwent volatile exsolution becoming buoyant and rose as volatile saturated plumes into the upper portions of the Stanthorpe Type chambers ponding near the roof as segregations, or intruding into crystallised portions of the chambers as stocks and dykes, or even into the the overlying rocks (as at Red Rock, Kilminster and Jibbinbar). These intrusions were the focus of hydrothermal fluid transport of volatiles and metals not only from within the Ruby Creek Type magmas themselves but probably also from potentially large volumes of crystallising magma at depth, hence their enrichment in otherwise relatively compatible ore elements (Cu, Zn, Au etc). Conclusions The Stanthorpe Group of high-K granites is part of the I-type Moonbi Supersuite and formed from a similar protolith by progressive differentiation. The Moonbi Supersuite provides an important illustration of the relationships between igneous chemistry, level of emplacement and the nature and ore element ratios of associated mineralisation. It demonstrates that Sn mineralisation can be generated in oxidised igneous rocks provided they are felsic and the crystallisation of potential sequestering phases is late or absent. References 1 2 3 4

5 6 7 8 9

Shaw, S. E. and Flood, R. H., 1981. The New England Batholith, eastem Australia: geochemical variations in space and time. Journal of geophysical research, v.86, p. 10530-44 Chappell, B. W., 1994. Lachlan and New England: fold belts of contrasting magmatic and tectonic development. Journal of the Royal Society of New South Wales, v. 127, p. 47-59. Candela, R A. and Blevin, R L., 1995. Do some miarolitic granites preserve evidence of magmatic volatile phase permeability? Economic Geology, v. 90, p. 2310-2316. Shaw, S. E., 1994. Ages of Granitoids. In Veevers, J. J. and Powell C. McA. eds. Permian-Triassic Pangean basins and foldbelts along the Panthalassan Margin of Gondwanaland, pp. 156-159. Geological Society of America Memoir 184. Tuttle, O. F. and Bowen, N. L., 1958. Origin of granite in the light of e?q)erimenatl studies in the system NaAlSi308-KAlSi308-Si02-H20. Geological Society of America Memoir 74,153p. Anderson, L. J. and Smith, D. R., ms. The effects of temperature and oxygen fugacity on the Al-in-homblende barometer. Submitted to American Mineralogist. Jones, D. A., 1976. The geology of the molybdenum deposits of New England. B.Sc Thesis (Unpubl). University of New England, Armidale. 190p. Saint Smith, E. C., 1914. Geology and mineral resources of the Stanthorpe, Ballandean and Wallangarra districts. Queensland Geological Survey Publication 243,165p. Blevin, P. L. and Chappell, B. W., 1992. The role of magma sources, oxidation states and fractionation in determining the granite metallogeny of eastem Australia. Trans R Soc Edinburgh Earth Sci 83,305-316.

100

PERMO-TRIASSIC GRANITE METALLOGENY OF THE NEW ENGLAND OROGEN Phillip L. Blevin and Bruce W. ChappeU National Key Centre for Geochemical Evolution and Metallogeny of Continents, Geology Department, Australian National University, Canberra ACT 0200 Introduction The New England Orogen can be broken into two principal metallogenic domains. The northern NEO (nNEO) is principally a Cu-Mo-Au province while the southern NEO (sNEO) is principally a Sn-Mo-Wpolymetallic province. Importantly, these metallogenic associations are replicated throu^ time within each province. Thus Cu-Mo-Au mineralisation occurs within the nNEO associated with magmatism during the Devonian, Permo-Carboniferous, Triassic and Cretaceous, while in the sNEO Sn mineralisation is associated with magmatism during the Carboniferous, Permo-Triassic and Triassic. Magmatic framework of New England Permian to Triassic I-type magmatism dominates the New England Batholith of the sNEO. Mineralisation is mainly associated with the h i ^ - K Moonbi Supersuite and with two highly fractionated leucomonzogranites (the Gilgai and Mole plutons. Granites of the Moonbi Supersuite are K-rich and have distinctive trace element compositions, most notably high Sr, Ba, Pb, Th and Cs and relatively low Y. Both I- and S-type granites within the sNEO are isotopically primitive despite their generally felsic nature, indicating that they were derived from only slightly older source rocks. Granites of the Clarence River Supersuite have distinctly lower K2O than the Uralla and Moonbi Supersuites. This may also be applicable to the Gundle, Carrai, Daisy Plain and Round Mountain plutons. A geographic separation of the I-types in the sNEO can therefore be made into a westem higli-K association and a low-K I-type association to the east of the Hillgrove Supersuite. The nNEO was the site of extensive plutonism in the late Carboniferous to early Permian and the early Triassic, extending down along the central NEO into the sNEO. These granites are typically low- to medium-K diorites, tonalites and granodiorites, with chemical and isotopic signatures indicative of continental margin afiSnity. The limit of the northerly Permo-Carboniferous province is unclear but probably does not necessarily coincide with the northerly limit of the NEO as currently determined from near-surface geology. Igneous metallogenic relationships The NEO was divided into several intrusive metallogenic provinces by Blevin et al. (1) spanning the Devonian to the Cretaceous (Figure 1). The two provinces of Permo-Triassic age are the Moonbi-Bathurst Intrusive Metallogenic Province in the sNEO, comprising oxidised high-K granites, and the Central NEO Intrusive Metallogenic Province, comprising Permo-Triassic intrusions extending from the north coast of New South Wales to the Rockhampton area in Queensland. Further to the north magmatism is dominantly either Permo-Carboniferous (Urannah Batholith) or Cretaceous. Devonian igneous rocks are developed in the Rockhampton region extending to the north along the coast. Moonbi-Bathurst Intrusive Metallogenic Province: In the sNEO most mineraUsation is related to the Moonbi Supersuite, the Mole Granite and the Gilgai Granite. The high-K association extends from Tamworth to Warwick. It is probably more extensive to the north. The Carboniferous granites of the north eastem Lachlan Fold Belt are most probably related. Indeed, similar Mo-W skams, minor Cu and Au mineralisation are associated with these granites and with the granites in the Moonbi-Attunga region of the sNEO. To the north within the sNEO, granites of the Moonbi Supersuite become more felsic and fractionated and are associated with Sn, Mo and W (Blevin and Chappell, this volume). It also becomes more shallowly exposed. Gold is an accessory in many deposits. The majority of Sn mined was alluvial, shed from low grade disseminated and sheeted vein or stockwork deposits. Sn mineralised granites were polymetallic, often associated with a wide range of metals (Sn, W, Mo, Bi, As, Ag, Pb etc). Complex metal zonation pattems is also present around the plutons. The region is characterised by very large numbers of small deposits, only a few of which approach the tonnage and grade of viable hard rock mines. Central NEO Intrusive Metallogenic Province: Very little published data is available on the chemistry of intrusive rocks associated with mineralisation in the nNEO. Cu dominant systems such as Coalstoun and Moonmera are associated with granodiorite porphyries, andesitic tuffisites and related dykes inferred to be later stage differentiates of the Bouldercombe Complex. Regionally, the Bouldercombe Complex is also associated with porphyry Cu-Mo style mineralisation at Sandy Creek-Gordon, Struck Oil and (?) Moongan. The Coalstoun deposit is associated with shallow porphyritic tonalitic to quartz dioritic intrusives. Intrusive rocks associated with Cu deposits in the nNEO are oxidised and intermediate to felsic, and in broad compositional temis are not dissimilar to intrusive rocks associated with porphyry style Cu deposits in other continental marginal settings. Mo dominant systems include Mo-Cu pipes in the Mount Perry area which are associated with medium-K series granites unlike similar pipe-like deposits associated with more fractionated high-K granites in north Queensland and in the sNEO. 101

The Mount Perry pipes are however Mo-Cu rather than being associated with elevated W-Bi as elsewhere. Further to the south east porphyry Mo mineralisation occurs at Anduramba. While the rhyolitic portions of the deposit have textural characteristics similar to Climax style porphyry Mo deposits, the Anduramba intrusions differ in that they are significantly less fractionated. Discussion and conclusions Several unusual features are apparent in the metallogeny of the NEO. The NEO has distinct zones of contrasting metallogenic that correlate, in Ae first pass, with broad scale variations in granite type. The associations are not time specific. That is, there are no "metal specific" metallogenic epochs in the NEO, rather metallogenic epochs mirror the major magmatic episodes that have occurred in the orogen. The NEO is relatively easily accommodated into a continental margin model, particularly when compared with the Lachlan Fold Belt. The role of plate tectonics as a dominant control on ore element ratios in igneous related mineralisation has long held currency in the economic geology and tectonic literature. Porphyry Cu deposits occur overwhelmingly along linear calcalkaline subduction-related volcanoplutonic arcs. Plutonic rocks associated with porphyry copper deposits within island arc settings are dominantly homblende and homblende-biotite diorites and quartz diorites while those in continental marginal settings are more typically granodiorites and quartz monzonite. Both are magnetite bearing. Sr isotope data indicates that intrusives located in continental margin type situations are more isotopically evolved than those in island arc settings. Existing evidence strongly indicates that there is no one "source" for potential porphyry Cu producing magmas. Considerable metallogenic diversity can be present in any single igneous supersxiite. An example is the Moonbi Supersuite which is associated with Cu-Mo-Au at the less evolved end, through Mo to Sn at the most evolved (fractionated) end. The relationships indicate that mineralisation is a product of magmatic and hydrothermal processes and is not a product of anomalous metal contents inherited by magmas. Relationships within the NEO are also consistent with broad redox-fractionation controls on ore metal-igneous associations. In the nNEO many uneconomic porphyry style Cu-Mo-Au systems occur, all with relatively low grades. From what little is known from &e related igneous rocks, there is no a priori reason why these rocks should be considered poor candidates for producing substantial mineralisation. References 1

Blevin, P. L. Chappell, B. W. and Allen, C. M. 1996. intrusive metallogenic provinces in eastem Australia based on granite source and composition. Transactions of the Royal Society of Edinburgh: Earth Sciences. (In press)

102

N Central Queensland Province ( Cretaceous)

500 km

Northern NEO Province (Permo-Carb)

NEW ENGLAND OROGEN

Mount Morgan Calliope Province (Devonian) Central NEO Province

Bathurst-Moonbi Province

LACHLAN FOLD BELT

Figure 1. Map showing distribution of intrusive metallogenic provinces in the New England Qrogen. Non PermoTriassic Provinces are italicised.

103

THE SIGNIFICANCE OF MID-CRETACEOUS BURIAL AND UPLIFT ON THE MATURATION AND PETROLEUM GENERATION IN THE BOWEN AND SURAT BASINS, EASTERN AUSTRALIA C J. Boreham' R.J. Korsch' & D.C. CarmichaeP 1 Marine, Petroleum & Sedimentary Resources Program, Australian Geological Survey Organisation, GPO Box 378, Canberra, ACT, 2601 2 Geological Survey of Queensland, Department of Resources and Energy GPO Box 194, Brisbane, Queensland, 4001 Summary In the Bowen and Surat Basins, burial and maturation of the potential source rocks, petroleum generation and migration of the hydrocarbons are all controlled by depositional and deformational events in the Mesozoic, and particularly in the Middle to Late Cretaceous. Maximum palaeotemperatures were experienced during the early Late Cretaceous on the western margins and the southem extremities of the Taroom Trough in response to both maximum burial and heat flow. Petroleum generation was effectively quenched from 90Ma to present following basin wide uplift and erosion. In the northem and central Taroom Trough, high heat flow in the Permian and thick Triassic sedimentation, effectively exhausted oil generation by the Middle Triassic with consequential residual gas generation emulating from the Permian source rocks during the Cretaceous. Oil-prone source rocks are more prevalent in the Late Pennian sediments whereas gas-prone source rocks dominate the early Late and late Early Permian. Fortunately, the chemical kinetics of oil and gas generation dictate that oil generation from the Late Permian sediments is synchronous with gas generation within the more mature and older source rocks. Thus, gas mobilisation of Hquids within the secondary migration pathway is an effective medium to support long range migration from the source rocks to the reservoirs primarily perceived as the basin margins. Introduction The Permo-Triassic Bowen Basin and overlying Jurassic-Cretaceous Surat Basin covering an area of approximately 200,000km'^ are situated in east-central Queensland and extend south into northem New South Wales. The Surat Basin has supported the first commercial oil production in onshore Austraha at Moonie. This oil field has accounted for over 90% of the recoverable oil reserves in the region during the last four decades. During this time the majority of the recoverable oil reserves in the basins have already been produced. Therefore, it is the challenge for Government and Industry to provide new and refined concepts in order to decrease the exploration risk in the search for new reserves. It is the intention of this study to meet this challenge by addressing the important issue of better understanding the timing of hydrocarbon generation and fluid flow in the basins. The results presented here are derived from a regional burial and thermal history modelling study from the Queensland sector of the basins and utilise the extensive geochemical database for the petroleum and potential source rocks (Boreham, 1994,1995; Carmichael and Boreham, 1996) and the geological framework (Fig. 1) derived from the National Geoscience Mapping Accord (NOMA) project Sedimentary Basins of Eastern Australia (Korsch and Totterdell, 1995; this volume). Geohistory modeUing An important component of the study is the use of geohistory and maturation modelling in order to reconstruct the basin architecture and to understand when and how much and when the fluids (oil and gas) were introduced into the basins. Described below are the values used for the most critical parameters in basin modelling. Modelling of burial history, maturation and petroleum generation was undertaken for 46 wells (Table 1; 36 petroleum exploration wells and 10 synthetic wells constructed from seismic data) using WinBury'^^ Version 2 software (Paltech Pty Ltd, Australia). Stratigraphy The stratigraphy shown in Figure 1, has resulted from a combination of the stratigraphic sequences mapped in seismic data from the southem Bowen and Surat Basins by the Australian Geological Survey Organisation (AGSO) and the Geological Survey of Queensland (GSQ), and a recent reinterpretation by the GSQ of the lithostratigraphy based on geophysical well logs. The latter has led to a revision of the Permian stratigraphy in the southem Bowen Basin (Beeston et al., 1995; Beeston & Green, 1995). The principal difference between the sequence stratigraphic and the lithostratigraphic interpretations concerns the southem Taroom Trough in that we consider that Beeston et al.'s upper coal unit of the Burunga Formation is the equivalent of the Bandanna Formation/Baralaba Coal Measures to the north. Hence for the purposes of this paper we combine the Burunga coals with the other Late Permian coal measures units. Missing section and unconformities The unconformity between the Buffel Formation and the Oxtrack Formation is, in part, equivalent to the intra-Aldebaran Sandstone unconformity. The unconformity is recognised in outcrops in the north and eastern Taroom Trough. It has not been mapped seismically, except in the Denison Trough and in the Gunnedah Basin in New South Wales. 104

Age

100-

Palynological zones

Surat B a s i n

Sequence Tectonic Basin boundaries events phas

Queensland

APK7

=i><}=

Section removed Griman Cr Fm

APKe APK5

Surat Sit

APK4

if,

Coreena Mbr Doncaster Mbr

APK3

Not mapped seismically

Bungil F m APK2

M o o g a Sst

APK1

Orallo F m

-S50

Gubberamunda Sst

150APJ6

Westboume Formation

Springbok Sst

APJ5

-340

Walloon C M APJ4 Hutton S a n d s t o n e

-835830-

12/A/167

APJ3

APJ2

Evergreen

Formation

Precipice

Sandstone

—820

-810-

Figure 1. Permian to Cretaceous correlation chart of the southern Taroom Trough region, Bowen and Surat Basins showing the relationship between lithostratigraphy, sequences and the main tectonic events (modified from Korsch & Totterdell, this volume). There is a significant unconformity at the base of the Early Triassic Clematis Group (Elliott, 1993). This is best observed in the westem part of the study area where sediments of the upper Clematis Group (Showgrounds Sandstone) sit directly on Rewan Group sediments (e.g. Apple Tree 1 and Inglestone 1). Farther west, the Showgrounds Sandstone sits on the Late Permian coal measures unit (e.g. Redbank 1 and Teelba 1). There is a major unconformity at the base of the Surat Basin succession, particularly in the eastem part of the basin where up to 4900m (extrapolated from seismic data) of Triassic section has been eroded (Table 1). In the south-west, there appears to be little evidence of major erosion and the Surat Basin succession lies parallel to the Moolayember Formation resulting in a major hiatus. From modelling we estimate about 200-300 m of sediment was deposited above the currently preserved top Moolayember Formation prior to being eroded (Table 1). As recorded by apatitefissiontrack datafi-omseveral wells in the basin (Raza et al., 1995) there has been a rapid uphft in the interval 95-90 Ma, The amount eroded following this upUfl is a combination of the amount added in post-Griman Creek time (see above), and an estimate of the amount of the known Surat Basin succession that has been removed down to the present day surface exposures. The amount of Late Cretaceous to present day erosion that has occurred has been modelled as a constant 250 m. Further, it was necessary to include a considerable thickness of additional Cretaceous sediment (up to 1200 m; Table 1) above thefinalunit of the Surat Basin succession, the Griman Creek Formation. This amount is needed by the burial history analysis to enable the maturity levels calculated by WinBury to match the measured vitrinite reflectance valuesfiromboth the Bowen and Surat Basins.

105

Maturity parameters The primary source of vitrinite reflectance (Ro) data are two-fold; open file GSQ data and a significant number of new measurements (Beeston et al., 1996). The latter, which includes data for most of the e}q)loration wells in this study, provides a very consistent vitrinite reflectance data set by which comparisons between wells can be confidently made. Calculated vitrinite reflectance are determined using the kinetic model for vitrinite reflectance (Sweeney and Bumham, 1990). The aim of the maturation modelling process is to make iterative adjustments to the palaeoheatflow history and/or amounts of sediment removed by erosion in order to obtain a best-fit profile between observed and calculated vitrinite reflectance whilst retaining geologically plausible input parameters. Source rock quality and effective source rock thickness Boreham (1995) and Carmichael and Boreham (1996) have shown that the Permian coals and associated sediments are the major source of oil and gas in the region. A small contribution from Triassic sediments has also been recognised at the Rednook and Roswin North fields (Boreham, 1995; Al Arouri, 1996). In order to provide a regional database for source quantity, quality and maturity, present-day total organic carbon (TOC) and Rock Eval data were obtained for the Moolayember Formation and the total Permian succession for 29 e:q)loration wells (Carmichael and Boreham, 1996). In order to limit the number of analyses, 50m composites were used in order to generate the continuous downhole geochemical profile. This interval thickness also constrains the temperature to within 1°C between the top and bottom of the interval and within acceptable temperature errors for the kinetic treatment. Coals and mudstones/siltstones were handpicked from each interval and analysed separately for TOC and Rock Eval. The thickness of coal and mudstone/siltstone was determined from the lithological composition of the 50m interval. The sum of individual interval thickness of coal and mudstone/siltstone gave the total rock thicknesses for the package. Where Rock Eval indicated that the sediments had reached the mature or overmature stage, initial TOC and Hydrogen Index (HI) were estimated from regional maturation profiles (Carmichael and Boreham, 1996). For the synthetic wells, or successions with no experimental data, extrapolated geochemistry data were obtained from contours of weighted HI and TOC for the analysed wells. Here, the effective source rock thickness was calculated from the global lithological composition of individual Uthostratigraphic successions. The Moolayember Formation and Permian successions analysed have been "grouped" into six source rock 'packages': (1) Moolayember Fomiation; (2) Bandanna Fonnation/Baralaba Coal Measures; (3) Burunga Fomiation/Scotia Coal Member/Tinowan Formation; (4) Banana/Muggleton Formations; (5) FlatTop/Barfield/Oxtrack/Buffel Formations; (6) Reids Dome beds and equivalent units. The groupings reflect major changes in depositional conditions and regional sequence boundaries (Fig. 1). Further, the larger scale groupings were necessary to fit within the constraints of WinBury where only a finite number of 'source units' were allowed in a well. Kinetic Treatment Separate chemical kinetics were used to define potential source rocks (coals and mudstones/siltstones) containing oilprone (HI>200), wet gas- and condensate-prone (150<HI<200) and gas-prone (HI<150) kerogen (Carmichael and Borehami, 1996^ For the Moolayember Formation, kinetic data of Al Arouri (1996) were used to define an oil-prone and gas-prone kerogen. Kinetic parameters were determined individually for both gas (C1-C4) and oil (C5+) (compositional kinetics; Boreham et al., 1996) for a Permian source rock and used to define three basic oil- to gas-prone kerogen types for the Permian sediments according to their initial HI values (see above). The gas and oil generation kinetics were determined using multiple parallel first-order reactions to yield a single frequency factor and a set of discrete activation energies which increment, typically, by 1 Kcal/mole. A 'reaction potential' for each individual activation energy was thus determined whereby the sum of the reaction potentials for both oil and gas was equivalent to the initial HI value for the kerogen. Using this approach, the influence of maturation on the gas-to-oil ratio (GOR) can be directly determined and was seen to change from low to high GOR over the maturity range of petroleum generation (Boreham et al., 1996). The chemical kinetics for oil-to-gas cracking of light oils (Horsfield et al., 1992) were used to describe the cracking process for residual oil (deemed to have a similar composition as the reservoired li^t oils) retained within source rocks. For the determination of masses of hydrocarbons generated within the 'package', the rock types were layered, from top to bottom, in the order, mudstone/siltstone, coal, and remaining sediment. For those wells with experimental data for 50m intervals, this layering order was used within each interval. However, with extrapolated data, the total effective thickness of each of the rock types was used. Although this is not the ideal arrangement of the effective source intervals within the source unit it is what is allowed within the current constraints of the software. An altemative approach whereby the ordering was reversed such that remaining sediment, mudstone/siltstone, and, at the base, coal can result in up to a 50% increase in masses of petroleum expelled from the coal. The ejqjulsion of gas and oil has been defined to occur when the volume of generated hydrocarbons exceeds a specified pore-saturation threshold. In WinBury this has been 'hardcoded' at 60% of total pore volume for oil while gas expulsion is assumed to be quantitative. We consider, however, that the pore saturation level for oil will be higher for coals («80%, see Forbes et al., 1991) and possible lower for mudrocks (40-50%), while some gas retention will be evident although quantitatively not as much as for the liquid hydrocarbons. This inflexibility in modification of saturation threshold will 106

result in a slightly earlier timing of oil expulsion from the coals, a slight over and underestimation in the quantities of oil e:?q)elled from coal and argillaceous sediments, respectively, and an overestimation in the gas yieldsfromboth rock types. Other Parameters Apatite Fission Track data: In conjunction with vitrinite reflectance,fissiontrack modelling was extensively used, mainly along the eastem margin to further constrain the heat-flow history (see also Raza et al., 1995; this volume). Chronostratigraphy: We follow the AGSO Timescale (Young and Laurie, 1996) for the Mesozoic, with the one exception being to place the Permian-Triassic boundary at 248 Ma. For the Permian, C.B. Foster and R.J. Korsch (unpublished data) have assigned numerical ages to the stratigraphy in the Bowen Basin on the basis of palynological and foraminiferal zones and, as yet unpublished, SHRIMP zircon agesfromtuffs. Lithology: An estimate of the approximate proportions of the main lithologies has been made for each of the formations. In addition, convective sandstone was included for important reservoir or groundwater-bearing formations. For modelling purposes, the same proportions of lithologies within a formation was used across the basin. Sea level: The eustatic sea level used here was the Exxon 1987 curve (Haq et al., 1987) for the Mesozoic section, in combination with the sea level curve for the Permian compiled by Ross and Ross (1987). The curves have been "stretched" and "squeezed" between common tie points to dovetail mesh with the timescale used here. Surface temperature: The present day surface temperature is taken as 8°C for all wells. Water depth: Estimates of palaeo-water depth are based on lithofacies and palaeoenvironmental indicators such as marine faunas, palaeosols and coal. Bottom hole temperatures: The bottom hole temperatures are those available from the open file well completion reports housed at the Geological Survey of Queensland. Porosity: porosity data were extracted from the PORPERM database (Miyazaki et al., 1990) which contains porosity measurements on core plugs carried out by the Bureau of Mineral Resources. Defaults: Where it has not been possible to obtain parameters from the studied wells the default parameters in WinBury have been used (e.g. for compaction factors, densities of lithologies, conductivities etc.),. Time (Ma) 150 100 50 L

120-

Heat flow -a- Tectonic subsidence + (Min/max range)

100 -

Present heatflow = 65.000 (mW m-2)

E £ 60H 0

40 -

0 -J

PERMIAN

TRIASSIC

JURASSIC

CRETACEOUS

TERTIARY

1-3.0

Figure 2. Heat flow and tectonic subsidence plot for 82-50 synthetic well (for location see Fig. 4 and Table 1).

107

Results and Discussion Heat flow model Vitrinite reflectance andfissiontrack data are the primary input parameters used to constrain the maturation history and heatflow in the southern Taroom Trough. The main features of the heatflowmodel, shown in Figure 2 for the synthetic well 82-50, incorporate a high flow during basin extension in the Late to Middle Permian, a slow decay during the Late Permian thermal relaxation phase followed by a period of low heat flow during foreland loading throughout the Early Triassic. Rapid uphft and erosion during the Middle to Late Triassic have resulted in higher heatflowvalues followed by a further decay during the Jurassic and Early Cretaceous thermal subsidence. Continuing subsidence and crustal flexure during the Middle to Late Cretaceous was accompanied by a steady rise in heat flow mainly due to episodic magmatic intrusions from approximately 130Ma (Golding et aL, this volume) cukninating in a maximum heat flow at approximately 95Ma in the Late Cretaceous. Well to well perturbations in the general heat flow model result from proximity to basement giving an offset to higher heatflow,a trend to higher geothermal gradients from east to west, and the delay from east to west propagation of foreland loading induced subsidence. The heat 'pulse' maximising in the early Late Cretaceous resulted in maximum palaeotemperatures over the majority of the basin except in the north where maximum paleotemperatures coincided with the deposition of the thickest Triassic successions along the depositional axis of the Taroom Trough. The timing of maximum palaeotemperatures during the early Late Cretaceous is consistent with a continuous vitrinite reflectance-depth profiles observed across major regional unconformities (Beeston et al., 1996). Leading up to that time, apatitefissiontrack lengths have been annealed and only begin to grow again in response to cooling associated with rapid uplift and erosion in the interval 95-90Ma. Although heat flow had maximised by the early Late Cretaceous, up to an additional 1200m (Table 1) of Late Cretaceous sediment (99-95Ma) was still required above the youngest preserved unit in the Surat Basin, the Griman Creek Formation (Fig. 1). This extra thickness alleviates the need to increase heatflow to unrealistic levels and has credence in major erosional events further east (Korsch and Totterdell, this volume). Further it is consistent with apatitefissiontrack analysis (Raza et al., 1995; this volume). A thick Late Cretaceous succession has also been deemed as critical for effective maturation modelling (Al Arouri, 1996). 2.0 •

1.8 -

1.6 -

• Top Baralaba Coal Measures O Top Burunga Formation

— 1.4 1.2 -

: 1.0 -

•s

0.8 0.6 -

O O

Gas maturity 1.05<Ro%<1.4

Oil/condensate maturity 0.65<Ro%<1.05

P ^ ^ G a s Window

0.4Temperature (°C) 12/A/16! Figure 3. Modelled vitrinite reflectance verses palaeotemperature at 95Ma for the tops of the Baralaba CM and Burunga Formation (data points correspond to individual wells). Superimposed is the temperature range over all wells atTR=0.1 and 0.5.

108

148" a )

148° b)

Figure 4. Present day depth (m below sea level) contours for modelled vitrinite reflectances of a) 0.65% b) 1.05% and c) 1.4% corresponding to the main oil window between a) and b) and the main gas window between b) and c). Note actual depths placed against the well symbol. Timing of hydrocarbon generation and quantities expelled The process of petroleum generation is govemed by laws of chemical kinetics which require both time and temperature to be accurately known. This requirement has been fulfilled once the burial and thermal history has been reconstructed. Although there are obvious uncertainties and approximations in the derivation of the geohistory model, a true test for the appropriateness of the model rests with the 'fit' between model predictions and field data. To a large extent this has been done on a well by well basis when the modelled datasets are 'fitted' to the experimental data (eg. observed verses modelled maturity). We have extended this approach and have used all the wells to predict basin-wide behaviour for independent datasets. Figure 3 shows the modelled maximum vitrinite reflectance at the tops of the Baralaba Coal Measures and Burunga Formation, (corresponding to the sequence boundaries B65 and B70; Fig. 1) plotted against palaeotemperature at 95 Ma for those wells where the maximum palaeotemperature occurs in the early Late Cretaceous. Superimposed on Figure 3 is the temperature range over all the wells where a source rock placed at the tops of the 109

Baralaba CM and Burunga Formation attains a transformation ratio (TR) of 0.1 and 0.5 (determined from the kinetic parameters). The former TR is considered to represent the onset of oil generation and beginning of primary migration, while the latter value corresponds to the end of the main phase of oil generation and the beginning of gas generation (the end of gas generation will occur at higher TR but was not estimated here). The beginning of primary migration of oil and gas correspond to a projected vitrinite reflectance of-0.7% and 1.1%, respectively (Figure 3), in good agreement with the observed maturity levels of reservoired oil and condensate of 0.65-1.05% Ro and natural gas maturity between 1.05 and 1.4% Ro (Boreham, 1995). Present-day burial depths to the tops of these three important vitrinite reflectance surfaces are shown in Figure 4. South of 27°S the onset of oil generation occurs at ~1600m (below sea level) and progressively shallows in a northerly direction to near surface in the most northerly wells. Further, the 0.65% isoreflectance value occurs at greater depths along the axis of the Taroom Trough compared with the margins, primarily in response to higher heat flow from shallow basement rocks in the west and greater proportion of the sediment lost to erosion at the eastem margins. The other two critical stages for petroleum generation at 1.05% and 1.4% Ro show similar regional trends compared with the onset of oil generation, with the greatest depths to the end of oil and gas generation occurring to the south at 3600m and over 4000m, respectively. Over most of the study area, the timing of the main phase of petroleum generation and e^qjulsion coincided with events in the Mid to early Late Cretaceous (Figs. 5a and 5c for example well). Although various sedimentary successions throughout the southem Taroom Trough undoubtedly attain appropriate maturities for the conversion of kerogen intofreehydrocarbons, the quantity and quahty of the organic matter (Carmichael and Boreham, 1996) is also instrumental in determining whether sufficient petroleum has been generated to support primary migrationfromthe source rock. Table 1 lists the quantities of oil and gas generated from both coal and mudrocks for the three main productive 'packages'. The majority of the oil in the region is generated and e^elled from sediments between B55 and B70 sequence boundaries (Fig. 1) whereas major gas production occurred mainly from the older sediments. Further, the combination of high heat flow and extra sediment thickness during the Late Cretaceous accounted for the majority of this petroleum generation and allowed the bulk of the Permian source rocks, and Triassic Moolayember Formation to the nor&, to actively expel oil and gas for the first time. However, this regional study appears to have insufficient resolution at present to resolve more localised issues, for example the small Triassic source contribution on the southwest margin (Boreham, 1995; A1 Arouri, 1996). Although the timing of generation and expulsion of hydrocarbons is strongly linked to Cretaceous events it should not be overlooked that chemical kinetics predetermine this association. Figures 5b and 5c illustrates the disparate conclusions that can be obtained when different kinetic parameters are applied to identical burial and maturation histories (Fig. 5a). In this illustrationfromthe 82-50 synthetic well, the effective source rock has a HI~250 mg hydrocarbons/gTOC and the chemical kinetic parameters are either a mixed kerogen type containing a 2:3 mixture of the Winbury default Type II and Type in kerogens (the same defaults are used other commercially available modelling software and is commonly used approach using admixtures) or the experimentally measured compositional kinetics for this Typell/ni organic matter (Boreham et al., 1996). For the former mixed kerogen type, all the oil and the majority of the gas generation have been completed during the Late Triassic due to hydrocarbon generation at lower temperatures in response to a lower activation energy distribution. Using the measured chemical kinetics, the bulk of Uquid and gas generation occurs at higher temperatures only attained during the Late Cretaceous. Conclusion Burial history modelling for 46 wells in the Bowen and Surat Basins in eastem Australia has enabled the reassessment of the timing of petroleum generation and migration. Over most of these basins, a period of extensive generation and mobilisation of oil and gas occurred during the time of both maximum burial and h i ^ heat flow in the Early Cretaceous (c. 95 Ma). It is only in the deepest parts of the basin, particularly along the axis of the Taroom Trough, where there was extensive burial in the Late Permian to Middle Triassic, that witnessed earlier generation, starting in the Late Permian but mainly within the Triassic. On the western margins of the basins in Queensland, effective trap formation is also likely to be coincident with the high rate of hydrocarbon generation and expulsion in the Early Cretaceous. Maximum palaeotemperatures were evident at this time and represent the primary control on petroleum generation.

110

DONCASTER Mbr Undiff WALLOON CM HUTTON Sst MOOLAYEMBER Fm CLEMATIS Gp REWAN Gp BARALABA CM BURUNGA Fm SCOTIA Mbr BANANA Fm — •

OXTRACK Fm

Sea level Sediment interface ISO-Ro

200

Time (Ma)

150

250

200

TRIASSIC

-J

Time (Ma)

150

JURASSIC

Oil expelled

•30

-25

Gas expelled

-20 % DC . . I

O-

s<D

L-10 <aE >O

-5

Figure 5. a) geohistory plot for 82-50 synthetic well b) rates of hydrocarbon generation using default kinetics c) rates of hydrocarbon generation using experimentally determined compositional kinetics. Another important factor that interplays with temperature is source character and type, expressed in modelling studies by unique chemical kinetics. Geochemical analyses have demonstrated an almost exclusive Permian source for the oil and gas. Late Permian sediments are the main source for oil wherein the coals are more oil-prone than rocks containing dispersed organic matter within mudstones and siltstones. Contrarily, the mudstones and siltstones lower in the section probably provide the bulk of the gas in the region, generation of which is considered important for efficient secondary migration of liquids to the reservoirs. On the westem side of the basins, a local source for petroleum is considered unlikely due to insufficient maturity for the organic matter to generate and expel hydrocarbons. Tlie discounting of local sources implies that long migration distances, sometimes up to 100 km, is required to charge the main basin margin reservoirs.

111

Acknowledgements CJB and RJK publish with the permission of the Executive Director, AGSO. Ian Deighton and M. Nicoll are thanked for assistance with the geohistory modelling and map generation using Petroseis, respectively. Ray Shaw is also thanked for his thorough review. References A1 Arouri, K., 1996. Ph.D. thesis. Dept. of Geology and Geophysics, Uni. of Adelaide, Australia. Beeston J.W. & Green, P.M., 1995. New stratigraphic names in the southem Taroom Trougji, Queensland. Queensland Govemment Mining Joumal, 96(1120), 23-28. Beeston, J.W., Dixon, O. & Green, P.M., 1995. Depositional history of the southem Taroom Trough, Queensland. APEA Joumal, 35, 344-357. Beeston, J.W., Smith, R.J., McKillop, M.D., Craig, C.H. and Newsome, R.W., 1996. Vitrinite reflectance data and maturation in the Surat Basin and southem Bowen Basin. Queensland Department of Mines Special Issue, in press. Boreham, C.J., 1994. Origin of Petroleum in the Bowen and Surat Basins: implications for source, maturity and migration. Australian Geological Survey Organisation, Record, 1994/42,106pp. Boreham, C.J., 1995. Origin of petroleum in the Bowen and Surat Basins: geochemistry revisited. APEA Joumal, 35, 579-612. Boreham, C.J., Horsfield, B., Schenk, H.J., Leistner, E, Horvath, Z. and Atkinson, L, 1996. Predicting the quantities of oil and gas generated from Permian coals using pyrolytic methods: results of kinetic modelling. Department of Industry, Science and Technology, Canberra, Intemational Science and Technology Program, Report C94/06065. Carmichael, D.M. & Boreham, C.J., 1996. Source-rock evaluation of the southem Taroom Trough. Queensland Department of Mines Special Issue, in press. Elliott, L.G., 1993. Post-Carboniferous tectonic evolution of eastern Australia. APEA Joumal, 33, 215-236. Forbes, P.L., Ungerer, P.M., Kuhfuss, A.B., Riis, E and Eggen, S., 1991. Compositional modelling of petroleum generation and expulsion: trial appUcation to a local mass balance in the Smorbukk Sorfield,Haltenbanken area, Norway. AAPG Bulletin, 75, 873-893. Haq, B.U., Hardenbol, J. & Vail, P.R., 1987. Chronology offluctuatingsea levels since the Triassic. Science, 235,11561167. Horsfield, B., Schenk, H.J., Mills, N. and Welte, D.H., 1992. An investigation of the in-reservoir conversion of oil to gas: compositional and kinetic findings from closed-system programmed-temperature pyrolysis. Organic Geochemistry, 19,191-204. Korsch, R.J. & Totterdell, J.M., 1995. Structural events and deformational styles in the Bowen Basin. In: Follington, I.W., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995....150 years on.... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 27-35. Miyazaki, S., Le Poidevin, S.R. & Wright, D.J., 1990. PORPERM: porosity and permeability database. Bureau of Mineral Resources, Geology & Geophysics, Record, 1990/88,286 pp. Raza, A., Hill, K.C. & Korsch, R.J., 1995. Mid-Cretaceous regional uplift and denudation of the Bowen-Surat Basins, Queensland and its relation to Tasman Sea rifling. In: Follington, I.W., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995....150 years on.... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 1-8. Ross, C.A. & Ross, J.ILP., 1987. Late Paleozoic sea levels and depositional sequences. In: Ross, C.A. and Haman, D. (editors) Timing and depositional history of eustatic sequences: constraints on seismic stratigraphy. Cushman Foundation for Foraminiferal research. Special Publication, 24,137-149. Sweeney, J.J. & Bumham, A.K., 1990. Evaluation of a simple model of vitrinite reflectance based on chemical kinetics. AAPG Bulletin, 74,1559-1570. Young, G.C. & Laurie, J.R., 1996. An AustraUan Phanerozoic Timescale. Oxford University Press, Melbourne, 279pp.

112

LATITUDELONGrrUDEUtcTriassic 95-90Ma Packagc Oil' Gas Oil Ga.s Oil Oaa Oil Gas aosion (m) erosion (m) 0-141 Ma 0-141 Ma 141-205 Ma 141-205 Ma 205-249 Ma 205-249 Ma 249-285 Ma 249-285 Ma -28.0381 147.8206 200 700 -27.4719 149.8889 400 2 257.26 80.71 22.71 0.27 0.00 0.00 0.00 0.00 3 241.02 105.41 42.24 4.99 5.49 0.04 0.00 0.00 82-50 sp268 -26.5817 149.7764 600 600 2 129.32 226.74 0.00 1.37 0.00 0.00 0.00 0.00 3 72.29 191.44 14.99 7.67 23.83 3.24 0.71 0.01 4+5 0.00 101.61 0.00 5.15 0.00 10.37 0.00 0.79 82-7.spn40 -27.7575 149.9136 400 900 2 49.94 38.49 0.00 0.00 0.00 0.00 0.00 0.00 3 201.05 84.63 8.46 0.12 0.00 0.00 0.00 0.00 84-03 sp689 -26.1983 149.6492 300 1000 2 124.98 127.38 0.00 U.15 0.00 0.09 0.00 0.00 3 31.80 92.33 3.34 2.07 29.63 9.91 0.41 0.02 4+5 0.00 102.12 0.00 2.21 0.00 37.31 0.00 1.32 86-P38 sp 836 -26.7506 148.5986 200 700 ALICK CREEK 01 -27.3136 150.5322 2000 650 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3 11.48 0.73 4.59 0.36 23.35 0.37 0.00 0.00 20.64 4+5 7.59 1.83 2.50 45.16 5.64 1.25 O.OI APPLE n i E E 01 -26.9 149.2697 2 4.07 1.13 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.58 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.00 0.08 0.00 0.00 0.00 0.00 0.00 0.00 ARBROTHOl -26.6833 148.4667 200 700 BALLYMENAOl -28.0264 150.2425 1500 1000 2 3.29 0.61 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 2.05 0.00 0.00 0.00 0.00 0.00 0.00 BELLBIRDOl -27.8481 149.495 300 1000 2 0.18 0.48 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.33 0.00 0.00 0.00 0.00 0.00 0.00 BENGALLAOl -26.6847 149.2306 300 700 2 7.69 0.88 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.40 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.00 0.16 0.00 0.00 0.00 0.00 0.00 0.00 BOGONGOl -28.1331 147.7886 200 700 BOOBERANNAOl -28.465 149.8081 410 1000 2 0.02 0.04 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.10 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.00 BURUNGAOl -26 150.0786 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3 19.68 7.16 1.22 0.95 26.49 1.71 0.28 0.00 4+5 3.83 5.18 0.37 2.58 30.91 1.84 14.28 0.19 C82-T-52/.SP 156 -25.7225 149.6731 2 0.00 93.89 0.01 41.43 92.99 108.58 0.00 0.08 3 0.00 13.06 0.00 27.33 156.59 163.40 10.96 3.35 4+5 0.00 0.00 0.00 0.00 0.00 208.98 0.00 13.38 CAB A WIN 01 -27.4961 150.1894 2 11.72 2.62 0.00 0.00 0.00 0.00 0.00 0.00 3 25.89 5.60 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.00 0.66 0.00 0.00 0.00 0.00 0.00 0.00 CANAAN 01 -26.4601 150.7131 1200 200 COALBAH 01 -27.2244 149.6492 300 1000 2 297.35 94.14 11.49 0.14 0.00 0.00 0.00 0.00 3 259.67 114.47 0.94 34.62 2.13 0.01 0.00 0.00 2 COCKATOO CREEK 0 -25.5728 150.1031 1.16 0.95 0.68 1.54 24.52 2.22 0.00 0.00 3 0.82 0.79 0.10 0.89 3.93 2.41 0.00 0.00 4+5 15.05 16.00 0.98 9.00 87.89 84.69 0.88 0.32 COONARDOOOl -26.6 148.9903 2 0.00 0.32 0.00 0.00 0.00 0.00 0.00 0.00 3 1.32 0.78 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.00 O.IO 0.00 0.00 0.00 0.00 0.00 0.00 DULACCAOl 149.8761 2 26.75 49.24 0.00 0.06 0.00 0.00 0.00 0.00 3 144.46 99.93 11.19 1.42 22.45 0.68 0.12 0.00 4+5 0.00 106.94 0.00 4.82 0.00 9.92 0.00 0.60 EBONY 02 -26.9819 148.9681 200 600 2 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 FLINTONOl -27.9144 149.6694 300 1050 2 11.58 1.05 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.00 GLENHAUGHTONOl -25.2083 149.1292 2 17.95 19.82 2.29 0.64 0.80 0.05 0.00 0.00 3 25.20 80.82 3.91 3.37 2.22 0.57 0.00 0.00 4+5 0.00 30.10 0.00 0.46 0.00 1.01 0.00 0.00 GRAFTON RANGE 01 -26.3986 148.9486 200 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 INGLESTONEOl -27.55 149.6608 300 2 51.70 16.04 0.60 0.00 0.00 0.00 0.00 0.00 3 73.38 20.80 0.54 0.00 0.00 0.00 0.00 0.00 LETHBRIDGEOl -27.0972 148.9275 200 600 2 1.11 0.01 0.00 0.00 0.00 0.00 0.00 0.00 MACINTYREOl -28.6194 400 149.85 800 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 MAXIMA MAX 01 -28.2603 150.1525 1000 600 2 0.00 0.20 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.55 0.00 0.00 0.00 0.00 0.00 0.00 MEANDARRAOl -27.2181 149.8767 450 700 2 166.90 46.75 3.17 0.02 0.00 0.00 0.00 0.00 % 154.94 68.48 3 14.79 0.2X 0.00 0.00 0.00 0.00 MEELEEBEEOl -26.1833 300 149.2 800 2 0.27 3.13 0.00 0.(X) 0.00 0.00 0.00 0.00 3 0.00 0.37 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.00 0.29 0.(X) 0.00 0.00 0.00 0.00 0.00 MOONIEOI -27.7456 150.2569 2500 4+5 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 MUGGLETXDNOl -26.4056 149.3058 300 2 2.29 2.55 0.00 0.00 0.00 0.00 0.00 0.00 3 9.16 3.19 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.21 2.09 0.00 0.00 0.00 0.00 0.00 0.00 p81-113spl58 400 2 0.00 0.29 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 1.59 0.00 0.00 0.00 0.00 0.00 0.00 REDBANKOl -28.4075 149.4247 300 1050 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 REDNOOKOl -27.3194 149.2625 300 1000 2 29.47 1.63 0.00 0.00 0.00 0.00 0.00 0.00 ROCKWOODOl -26.9858 150.3708 2500 400 S78-6sp280 600 -26.5825 150.0322 1200 2 169.84 199.30 0.00 0.00 0.00 0.00 0.00 0.00 3 153.46 228.21 9.91 3.88 34.54 1.74 1.77 0.00 4+5 0.00 202.34 0.00 3.07 0.00 33.99 0.00 2.91 SUSSEX DOWNS 01 -27.7747 150.1403 1500 800 4.24 2 2.% 0.00 0.00 0.00 0.00 0.00 0.00 6.54 3 5.63 0.00 0.00 0.00 0.00 0.00 0.00 4+5 0.00 1.92 0.00 0.(X) 0.00 0.00 0.00 0.00 T82.L-102/.';p 131 -25.4419 149.7528 300 1000 2 0.00 0.00 0.00 0.00 154.25 268.32 8.07 8.56 3 0.00 0.00 0.00 0.(X) 65.81 90.32 4.98 5.17 4+5 0.00 0.00 0.00 ().(X) 0.00 0.00 0.00 4.50 2 TEELBAOl -28.0539 149.3578 300 1000 0.14 0.27 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 TIGGRIGIE CREEK 01 -25.7722 149.3572 300 1000 2 49.08 19.63 3.04 0.03 0.37 0.00 0.00 0.00 3 0.00 144.38 0.(X) 3.62 0.00 0.00 0.00 0.00 UNDULLAOl -27.2278 150.2678 1500 700 2 1.% 0.08 0.00 0.00 0.00 0.00 0.00 0.00 3 143.35 11.32 1.14 0.01 0.45 0.00 0.00 0.00 4+5 7.31 2.85 0.11 0.00 0.07 0.00 0.00 0.00 WANDOANOl -26.175 149.9 2 401.20 113.04 14.57 0.16 8.49 0.02 0.00 0.00 3 88.73 138.83 12.24 5.49 26.64 3.48 1.63 0.04 0.00 51.77 4+5 0.00 1.60 0.00 7.30 0.00 0.46 -27.5675 149.0608 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 80Fsp 184 82-12 spl72

• unitsa barrcls/m2 (volume of oil or gas expelled from a Im x Im column between the top and bottom of the tola! .sourec nKk interval within the ' packagc')

Table 1. Eroded section and volumes of oil and gas generated for various potential source rock intervals.

113

THE MESOZOIC CONTINENTAL MARGIN: CARBONIFEROUSMESOZOIC ARC TERRANES IN WEST ANTARCTICA, NEW ZEALAND AND AUSTRALIA J. D. BradshawS R.J. PankhurstS S.D. Weaver^, B.C. StoreyS R. J. Muir^ and T.R.Ireland^ 1 Department of Geological Sciences, University of Canterbury, Christchurch, N.Z. 2 British Antarctic Survey, High Cross, Cambridge CBS OET, U.K. 3 Reseach School of Earth Sciences, Australian National University, Canberra, Australia Summary Recent research in West Antarctica suggests that Marie Byrd Land is composed of two geological provinces, a western Ross Province and an eastern Amundsen Province. These correlate geologically with the Western Province of New Zealand and the Median Tectonic Zone. In New Zealand the latter is tectonically attenuated. The western units correlate with the Lachlan Fold Belt, and the Amundsen Province and MTZ relate best to the New England Fold Belt and eastem Queensland magmatic rocks. Introduction The Paleozoic to mid-Cretaceous 'basement' rocks of New Zealand are best understood in terms of eigjit major tectonostratigraphic terranes and a number of smaller tectonic entities (Bishop et al. 1; Bradshaw, 2). These in turn can be grouped into three superterranes, which for historic reasons, are known as the Westem Province, the Median Tectonic Zone, and the Eastem Province. The Westem Province comprises mainly early Paleozoic rocks cut by DevonianCarboniferous and Cretaceous granitoids (Cooper & Tulloch, 3), the Median Tectonic Zone is composed of the remains of a number of Late Paleozoic and Mesozoic magmatic arcs (Bradshaw, 4; Kimbrough et al. 5), and the Eastem Province comprises arc, arc-derived and accretionary complex rocks of Permian to Cretaceous age (Bradshaw, 2), the products of plate convergence and subduction. Research by the SPRITE expeditions to Marie Byrd Land (Bradshaw et al. 6) has shown that Marie Byrd Land comprises two crustal units or provinces (DiVenere et al. 7; Bradshaw et al. 8; Pankhurst et al. 9) that correlate geologically with the Westem Province and Median Tectonic Zone of New Zealand. The whole pattern is coherent when plotted on the tight fit for the Ross Sea - Marie Byrd Land - New Zealand region that has been proposed by Lawver and Gahagan (10). Correlatives of the Eastem Province of New Zealand do not occur in Marie Byrd Land. Events in these provinces or superterranes characterize at least 2000 km of the Late Paleozoic-Mesozoic margin. From westem New Zealand they can be projected on to the Challenger Plateau and Lord Howe Rise towards eastem Australia. Although some of the same elements clearly occur in eastem Australia as the Lachlan Fold Belt, the New England Fold Belt, and Mesozoic magmatic rocks in eastem Queensland, in Australia they appear to be parts of a more coherent and mainly autochthonous tectonic assemblage. West Antarctica The new data presented here come mainly from Marie Byrd Land, where pre-Cenozoic rocks are confined to a relatively a narrow zone of mountain ranges close to the coast, and the islands of Pine Island Bay. Thurston Island is geologically related to this region. Paleomagnetic studies by DiVenere et al (7) have suggested that Marie Byrd Land comprises two units, a westem block that lay close to East Antarctica (and particularly northern Victoria Land) until the Late Cretaceous, and an eastem block that originated further to the east and only became welded to the westem block between 117 and 100 Ma. The shape of these two blocks is poorly constrained by paleomagnetism but additional geochemical, geochronological, and isotopic data, suggest that the eastem unit is bounded by a line through the mouth of the Land Glacier that passes east of Mt McCoy and then runs approximately eastwards, sub-parallel to the coast, to he between the Kohler Range and Mt Murphy (Fig.l) (Bradshaw et al. 8). The names 'Ross Province' (west) and 'Amundsen Province' (east) have been suggested by Pankhurst et al. ( 9).

114

Figure 1. Distribution of significant localities in Marie Byrd Land. The dashed line marks the boundary between the Ross Province (southwest) and Amundsen Province (north). The Ross Province is typified by two rock types, the Paleozoic Swanson Formation and the Ford Granodiorite Suite. The Swanson Formation (Bradshaw et al. 11) is a thick monotonous sequence of sandstone dominated quartzose turbidites disposed in large upright variably plunging folds. Where not influenced by later thermal metamorphism, it yields K-Ar ages between 400 and 450 Ma, which are believed to reflect progressive cooling after defomiation and cleavage formation (Adams, 12). More recently, Rb-Sr isochrons for Swanson Formation in Edward VE Peninsula point to metamorphism between 421 and 432 Ma (Adams et al. 13). The Swanson Formation is cut by discrete plutons of the Ford Granodiorite, a granodiorite-monzogranite suite of calcalkaline T type granitoids with low initial ''Srf'Sr ratios of 0.704-0.706 (Weaver et al. 14). Rb-Sr geochronological results on these granitoids by Adams (15) and Pankhurst et al. 9), point to cooling ages between 345 and 380 Ma. A granitoid with 'S' type mineralogy from Bruner Hill near Mt Shirley, west of the mouth of the Land Glacier (Fig. 4), has an mid-Carboniferous U-Pb zircon age of 322±9 Ma (Mukasa,16), but the strontium initial ratio is low. Richard et al.'(17) record a Permian granite at Neptune Nunataks on the south side of the Fosdick Mountains and Adams et al. (13) have found Permo-Triassic minor intrusions in the Edward VE Penisula. Cretaceous granitoids are widespread in Westem Marie Byrd Land. The Fosdick Mountains are made of a migmatite complex with a significant component of Cretaceous magmatic material plus anatectic material, uplifted during crustal extension between 100-94 Ma (Richard et al. 17). MidCretaceous 'A' type granites have been described from the Edward VII Peninsula by Weaver et al. (18,19), and these and other extension related plutons are discussed in a companion paper by Weaver et al. (herein) West of the Ross Sea, in northem Victoria Land (Fig 1), the Robertson Bay terrane (RBT) resembles the Paleozoic geology of the Ross Province. Tlie main sedimentary unit, the Robertson Bay Group, closely resembles the Swanson Formation in lithology and deformation style. K-Ar ages for post-metamorphic cooling appear to be slightly older than in Marie Byrd Land but are diachronous, decreasingfrom500 Ma in the west to 460 Ma in the east (Adams & Kreuzer, 20; Dallmeyer & Wright, 21). The Robertson Bay Group is host to a suite of granitoids called the Admiralty Intrusives (Stump, 22 and references therein). The Admiralty Intrusives range from monzogranite to diorite and fomi an T type suite showing increasing crustal influence to the northeast. Available geochronology (Stump, 22) points to intrusion in a short period between 380 and 390 Ma. As in Marie Byrd Land, a subsidiary phase of Carboniferous granitic intrusion occurs as the Salamander Granite complex for which there are Rb-Sr isochron ages of 319±5 and 319±26 Ma (Stumn 22). Palaeozoic correlations between Australia (particularly Victoria and Tasmania) and northem Victoria Land have been widely discussed (See Stump et al. 23). Rock exposures in Marie Byrd land east of the Land Glacier (Amundsen Province) are very limited and are concentrated in four areas: the Ruppert Coast and Hobbs Coast as far east as the Demas Range, the Kohler Range and the Bear and Martin peninsulas, the islets and coast of Pine Island Bay, and Thurston Island (Fig.l). Basement rocks are confined to the coastal region and are not exposed in the southem part of Marie Byrd Land. It seems likely that older continental crust forms the southem region as indicated by exposures of old rocks in the pedestal of the Cenozoic volcano Mt Murphy, where metamorphic rocks include biotite paragneiss, calc-silicate gneiss with bands of marble and amphibolite' Granodioritic and dioritic orthogneiss sheets cut the metasediments and Storey et al (24) report U-Pb SHRIMP zircon 115

ages of 504±8 Ma suggesting that the host is no younger than Cambrian and a possible link with Granite Harbour magmatism in the Transantarctic Mountains. The occurrence of Gondwana type basement to the south is supported by outcrops in the English Coast (Fig.l), southeast of Hiurston Island, that include quartzites of possible pre-Permian age and lithic sandstones with Glossopteris (Laudon, 25). Mudstone erratics at Milan Rock, east of the Land Glacier, have a Mid-Late Devonian flora (Grindley and Mildenhall, 26) and suggest Devonian sedimentary sequence nearby. However, as this age is almost identical to the Ford Granodiorite suite that occurs close by in Mt McCoy, the sediments may point to tectonic juxtaposition. The Amundsen Province (Pankhurst et al. 9) is dominated by plutonic and subsidiary volcanic rocks that range in age from Carboniferous to Late Cretaceous. Overall compositions suggest the area was the site of long-sustained continental margin arc magmatism and that the rocks are relics of a succession of arcs that developed along the Gondwana margin. Well within the main magmatic belt older granitoids occur, and though it is not possible to determine whether they are basement 'inliers' or tectonic slices, their occurrence is consistent with the continental margin arc character of the belt as a whole. In the Clark Islands (Fig. 3) deformed and mylonitic granitoids have a whole rock Rb-Sr isochron age of446±16 Ma, and slightly younger gneissic granitoids at Slater rocks in the Kohler Range are 419+20 (Storey et al. 24) by the same method. Carboniferous orthogneiss on Thurston Island (Fig. 2) appears to be a typical T type granitoid and has an RbSr isochron age of 309±5 (Pankhurst et al. 27 Leat et al. 28) Texturally distinctive Early Permian T type granitoids are well developed in the Kohler Range and give a well defined Rb-Sr isochron age of 276±2 Ma (Pankhurst et al. 9). Orthogneiss and mylonitic granite in the Bear Peninsula may also be Late Paleozoic but at present the geochronological interpretation is not well constrained. Schists at Mt Petras and paragneiss within Cretaceous granitoids of the Demas Range appear to be derived from late Paleozoic or early Mesozoic sediments with detrital zircons ages in the range 310-410 Ma. Both show a marked abundance peak at 350-360 Ma, suggesting a source dominated by Late Devonian and Early Carboniferous magmatic rocks (Storey et al. 24). The age of the Petras Schists is similar to the age of protolith of metamorphic rocks in Edward Vn Peninsula and the province boundary may he to the north of this point. Early Mesozoic granitoids occur at Kinsey Ridge in the westem part of the Province and have a Rb/Sr isochron age of 239±4 and a low initial ®'Sr/^Sr ratio of 0.7035 (Pankhurst et al. 9). This is broadly comparable with a U-Pb zircon age of 253+4 (upper intercept) obtained by Palais et al. (29). Triassic granitoids are also well developed on Thurston Island (Pankhurst et al. 27). Mid-Jurassic rocks are widespread in the Pine Island Bay area and are typical T type suite of quartz diorites and granodiorites with rare gabbros. Ages rangefrom160-180 Ma (Pankhurst, unpublished). Late Jurassic T type granitoids are also widely developed in Thurston Island (Pankhurst et al. 27) aad range in agefrom152 to 142 Ma. Sli^tly younger Early Cretaceous rocks occur in the Demas Range in the eastem Hobbs Coast, with ages of 134±13 and 132±8 Ma (Mukasa, 16). These rocks are complex granitoids and rangefromearly, megacrystic and banded mafic phases to uniform massive felsic granitoids. Contemporaneous mutual intrusion and interaction between mafic dikes and the host is characteristic. These rocks are followed throughout Marie Byrd Land by a more uniform suite of calc-alkaline T type granodiorites and related volcanics in the period 124-108 Ma (Weaver et al. 19). Very similar episodes of magmatism occur in Thurston Island and extend from 125 to 110 Ma (Pankhurst et al. 27, Leat et al. 28) and in Pine Island Bay (Pankhurst, unpubUshed). In both areas however younger Cretaceous activity continues to be calc-alkaline in character until at least 96 Ma (Mukasa et al. 16) and the anorogenic "break-up" magmatism (Weaver et al. 19) is not represented.

116

Thurston ca300| ^ ' [i65±i1 Island 1S2±2| |-I45±2| |145±2|237±6l

•^-^nSw

72°-\ Morgan IsMts] |286±6| Inlet

rju./I

Figure 2a. Pine Island Bay - Thurston Island area. The headland and islands of Pine Island Bay, with the exception of Clark Islands are Jurassic and Cretaceous granitoids (Mukasa etal. 16; Pankhurst, unpub). Selected ages from Thurston island from Pankhurst et al. (26).

117

Figure 2b. SHRIMP U-Pb geochronology of Paleozoic and Mesozoic granitoids in the Median Tectonic Zone, southwest South Island (Muir et al. 36 and unpubUshed data). New Zealand: the Western Province and the Median Tectonic Zone The Western province comprises the Buller terrane and the Takaka terranes, separated along the Anatoki Thrust (Cooper & Tulloch, 3). The Takaka terrane is very diverse and includes two contrasted facies associations. The older one comprises Cambrian volcanic rocks, limestone, conglomerate and turbidites, partly seen as components of a melange. In the younger association, includes a thick sandstone-mudstone succession of Ordovician age, with major lenses of marble, the Silurian is largely quartzitic, and over a kilometre of mudstone and fine sandstone is developed in the Devonian. The Buller terrane lies to the west and is far more extensive. It is also thought to underlie the southern half of the Campbell Plateau and has been encountered in five petroleum exploration holes in the Great South Basin. Two sedimentary facies are widely developed, a monotonous turbidite succession of probable Cambrian-Early Ordovician age, and a better differentiated succession of quartzitic sandstone and graptolitic black shales of Lower and Middle Ordovician age. The latter is only seen east of the Karamea Batholith. K-Ar geochronology suggests initial metamorphism and cleavage formation in the latest Ordovician or Silurian with a maximum age of 438±7 Ma. The western part of the Buller terrane, specifically the Greenland Group, closely resembles the Swanson Formation in sedimentology, deformation style, composition and details of the K-Ar and Rb-Sr isotope systematics (Bradshaw et al. 11; Adams, 12). Grey turbiditic schist on Campbell Island also has a K-Ar age of 443±6 Ma (Adams et al. 30) and provides a useful link on the Campbell Plateau, midway between Marie Byrd Land and westem New Zealand. 118

The Buller terrane is cut by granitoids of Late Devonian and Late Carboniferous age (Muir et al. 31,32). The Devonian granitoids, the Karamea suite (Fig. 3) includes both, 'S' and T types (Muir et al. 33) aad the Takaka terrane has mafic rocks of similar age and comparable geochemistry. Carboniferous granitoids of-325-330 Ma also occur (Muir et al, 32). The combination of Swanson Formation, Late Devonian Ford Granodiorite and mid-Carboniferous granitoids in Marie Byrd Land shows a compelling similarity to the Buller Terrane of the Westem Province. The Buller terrane is overlain by one small outlier of volcaniclastic Triassic sandstone, cut by a dolerite sill that closely resembles the Jurassic Ferrar Dolerite of the Transantarctic Mountains and the Tasmanian Dolerites (Mortimer et al. 33). Both the Buller and Takaka terranes are cut by Late Mesozoic granitoids, the most important of which is the Separation Point Batholith (Muir et al. 34), which cuts both the Takaka terrane and the magmatic rocks of the Median Tectonic Zone, stitching them together by 118±4 Ma. (Bradshaw, 4; Muir et al. 34). A slightly younger suite of granitoids with T type affinity were enq)laced ~ 110 Ma and are related to or closely followed by the onset of crustal extension. The Median Tectonic Zone is a narrow belt of basic to acid plutonic rocks with subsidiary units of volcanic rocks and volcanoclastic sandstones. Many contacts are faulted and discrete terranes have been identified within the Median Tectonic Zone. Conventional U-Pb dating (Kimbrough et al. 35) and U-Pb SHRIMP ages (Muir et al. 36 and unpublished ages) show that magmatism extended from the Early Carboniferous to Early Cretaceous, with notable gaps in the Permian and Jurassic (Fig. 3). Inception of magmatism at least as early as the Carboniferous (Fig. 2b) is shown by the Lake Roxburgh Tonalite (344±4 Ma) and the Pomona Island Granite (303±5 Ma) both dated by the U-Pb SHRIMP method (Muir et al. 36). In the northem South Island, the Echinus Granite cuts a genetically related series of mylonitic and submylonitic gneissic granitoids (Beresford et al. 37) with a U-Pb age of 310±4 (Kimbrough et al. 35). This is remarkably close to the Morgan Inlet Gneiss of Thurston Island (309±5 Ma). U-Pb geochronology by conventional mass spectrometry and by SHRIMP show some unresolved differences in the ages of Mesozoic activity, in particular the SHRIMP data suggest important magmatism around 160 Ma. This may however simply reflect sampled localities and work is in progress on this problem. All the rocks are calc-alkaline and resemble continental margin arc rocks of Andean type. Volcanoclastic sediments are typically plant-bearing and are probably mainly sub-areal. (Bradshaw, 4 and refs). Magmatism in the northem South Island appears to have concluded about 18 million years before the intrusion of the alkali-calcic Separation Point Batholith. In southem New Zealand the distinctive Separation Point type magmaism occurs at deep levels in the Westem Province (Westem Fiordland Orthogneiss) and at shallow levels cutting the Median Tectonic Zone. In the south this magmatism starts 5-7 miUion years earlier than in the north (Muir et al. 36). The strong compositional contrasts in the rocks of the Median Tectonic Zone are probably responsible for the development of strong magnetic anomahes (800-1100 nT, total force) over the zone. Although some parts of other terranes, particularly the Brook Street terrane and the Dun Mountain Ophiolite of the Maitai terrane, also have marked magnetic anomahes, these are rarely as strong or as extensive as those associated with the MTZ (Woodward & Hatherton, 38). Much flatter fields characterise the Westem Province and the majority of the Eastem Province. Strong magnetic anomalies in the northem and central Campbell Plateau, the Campbell Magnetic Anomaly System, probably indicate the offshore extension of the zone (Fig. 2b). A recently pubhshed refinement of magnetic anomalies by Davy (39) shows that the zone of strong anomahes sweeps north to include the Bounty Islands (Fig. 3) which are made up of granitic rocks of Early Jurassic age (194±3 Ma, U-Pb SHRIMP). The southem margin of the MTZ is not well constrained on the Campbell Plateau, but projection of the Land Glacier boundary on to the reconstruction of Lawver and Gahagan (10) suggests that it reaches the edge of the plateau just to the south of Antipodes Island, close to the southem limit of the Campbell Magnetic anomaly System as indicated by Grindley and Davey (40). Similar zones of strong anomahes continue northwest of New Zealand along the Challenger Plateau towards the Lord Howe Rise.

119

W. Marie

• El

Byrd L. •

E. Marie Byrd L.

ES3 M

••FD

casa

C2S3

Figure 3. Phases of significant magmatism in the Western province and Median tectonic zone of New Zealand compared with those of northern Victoria Land, Marie Byrd land and Thurston Island. The blank squares at the top of the figure are magmatic events related to extension and break-up rather than subduction. FD=Ferrar Dolerite 3b Trends of strong magnetic anomalies across the Campbell Plateau between the MTZ and the zone of similar rocks in Marie Byrd Land on a continentalfitbased on Lawver and Gahagen (10). Large amplitude magnetic anomalies also occur in Thurston Island and continuously along the westem side of the Antarctic Peninsula. Interpretation The Westem Province-Ross Province and the Median Tectonic Zone-Amundsen Province are juxtaposed regions with contrasted geological history and development. In New Zealand they are in tectonic contact or stitched by younger intrusive rocks of Cretaceous age (Bradshaw, 4). In Antarctica the nature of the contact cannot be determined but paleomagnetism also supports Cretaceous amalgamation (DiVenere et al. 7). In addition to the geological evidence, Nd model ages for the two Antarctic provinces differ. In the Ross Province model ages of 1300 to 1500 Ma are typical while to the northeast in the Amundsen Province younger ages in the range 1000-1300 Ma prevail (Pankhurst et al. 9). The Median Tectonic Zone of New Zealand probably extends across the northem part of the Campbell Plateau to include the Bounty Islands as a zone not less than 300 km wide near the eastem margin. In Marie Byrd Land, the minimum width is 400 km from Mt Murphy to Thurston Island. In the mainland of New Zealand, it is only one tenth of this width or less. In places along its eastem margin it can be seen to have been overthmst by the Brook Street terrane after --140 Ma (Johnston et al. 41) and the westem margin is tectonic or cut by the younger Separation Point Batholith (Bradshaw, 4). Internal contacts may also be tectonic. The Median Tectonic Zone of New Zealand therefore appears to be an attenuated representation of a major zone of arc magmatism of Carboniferous to Cretaceous age on the Gondwana margin. The recognition of the zone and the scale of its development is particularly important because of the enormous development of Permian and Mesozoic sedimentary terranes in New Zealand that appear to have been derived from magmatic arc sources. Similarly the persistent presence of volcanic detritus in Permian and Triassic Beacon Supergroup sandstones in the Transantarctic Mountains may also be satisfactorily explained. Thurston Island has been linked with the Antarctic Peninsula arc (Pankhurst et al. 26), in turn linking Marie Byrd Land and New Zealand in a quasi-continuous zone of Carboniferous to Cretaceous magmatic arcs around the South Pacific. Australia The marked similarities between Australia, Northem Victoria Land and New Zealand have been commented on by several authors. This similarity extends to the Ross Province of westem Marie Byrd Land and the 'Lachlan-Cape York Granite Belt' of Chappell (42) probably continues below the Ross Sea into the southwestern part of West Antarctica. The Devonian granitoids of Antarctica and New Zealand are younger than the majority of the Lachlan plutons but correspond closely with those of Tasmania. The subsidiary suite of Carboniferous plutons ranging from 340-312 Ma, with a peak at 325-320 Ma are close in age to the subsidiary suite in Antarctica-New Zealand. Predominantly westward directed subduction has been proposed as the ultimate driving force. The eastem limit of the Lachlan Fold Belt is concealed beneath the Sydney Basin and in both New Zealand and Antarctica it appears to be tectonic probably marking a line of truncation. Comparisons between the New England Fold Belt and the Amundsen Province-Median Tectonic zone are much more diflScult to draw, largely because of the contrasted nature of the exposure. The New England Fold Belt contains a great 120

diversity of sedimentary and volcanic rocks cut by Carboniferous and Permian granitoids. In the other two regions magmatic rocks, mainly plutonic, are almost exclusively exposed, although in Marie Byrd Land the ice covered regions could conceal as much sedimentary and volcanic rock as occurs in New England. The New England Fold Belt was generated by Late Paleozoic plate convergence and has a complex tectonic history which concluded with the intrusion of the granitoid suites between --310 and -260 Ma (Chappell, 42) that overlap with some of the ages in the other provinces. What distinguishes the Median Tectonic Zone and the Amundsen Province is voluminous episodic plutonism continuing through much of the Mesozoic. The onshore New England Fold Belt is considered incomplete, with the outer part lying in the Lord Howe Rise. Scattered occurrences of volcanic and intusive rocks, and wedges of volcanogenic sediment in sedimentary basins lying on the New England Fold Belt point to persistent arc magmatism near or to the east of the present Queensland coastline (Veevers, 43) until the mid-Cretaceous. We have been unable to find any details of the composition of these volcanic sediments that would point to chemical changes within the arc sources. Comparison with Cenozoic convergent margin arcs suggests that convergent margins evolve and volcanic compositions reflect changes in geotectonic parameters and subduction zone character. In general, it appears that during the Mesozoic the development of northeastem Australia, New Zealand and West Antarctica was driven by the subduction of the Phoenix plate (Coney, 44). The Eastem Province of New Zealand and correlative rocks in New Caledonia are major new bodies of continental crust largely produced by this convergence but these outboard terranes do not appear in Australia or West Antarctica except for the Antarctic Peninsula (particularly Alexander Island, Fig.l). In the last area, the Fossil Bluff Group and LeMay Group are closely comparable Mesozoic forearc and accretionary-complex rocks. The Mesozoic history of east Australia in intimately involved in the broader picture of the interaction of the Gondwana margin and the Phoenix plate. The Australia coiiponents appear to be largely autochthonous and may provide a valuable reference point in attempts to resolve the origin of displaced terranes. Conclusions Magmatic rocks developed as a consequence of plate convergence are a key to understanding the interaction of the Phoenix plate and the Gondwana margin during the Mesozoic. A Paleozoic hinterland of Lachlan Fold Belt type is common to Australia, New Zealand and West Antarctica. On the outboard side is developed a succession of Carboniferous to Cretaceous magmatic arcs, and more locally, forearc and accretionary complexes. Geochemical and geochronological study of these arcs, both as igneous rocks and as igneous clasts in pene-contemporaneous sedimentary deposits will indicate the architecture and petrogenetic evolution of the arc systems and point to major events along the plate boundary. Acknowledgements The paper includes some of the findings and laboratory results from the joint New Zealand, USA, UK expeditions to Marie Byrd Land in 1990-2 (South Pacific Rim Intemational Tectonics Expedition) and the New Zealand Granites and Crustal Evolution programme of the University of Canterbury (FRST contract UOC 313). We acknowledge the valuable contributions of the other participants, Ian Dalziel, Sam Mukasa, Vic DiVenere, Anne Grunow, and David Palais. Logistic support in Antarctica was provided by National Science Foundation, British Antarctic Survey, and New Zealand Antarctic Programme. We thank the crews of VXE 6, USCG Polar Sea and the BAS air unit for their willing support, and Peter Cleary, Damion O'Carol, and John Roberts for their help in thefield.In particular we wish to recognise the contribution to the SPRITEfieldprogamme of Andy Harris, who died descending from Mt Everest in May 1996. References 1. Bishop, D.G., Bradshaw, ID. & Landis, C A. 1985. Provisional terrane map of South Island, New Zealand. In: Howell, D. G. (ed.), Tectonostratigraphic terranes of the Circum-Pacific region, Circum Pacific Council for Minerals and Energy, Houston, 515-521. 2. Bradshaw, J.D., 1989. Cretaceous geotectonic pattems in the New Zealand Region. Tectonics., 8, 803-820. 3. Cooper, R. A. & Tulloch, A.J., 1992. Early Paleozoic terranes in New Zealand and their relationship to the Lachlan Fold Belt. Tectonophysics, 214, 129-144. 4. Bradshaw, J.D., 1993. A review of the Median Tectonic Zone: terrane boundaries and terrane amalgamation near the Median Tectonic Line, New Zealand Journal of Geology and Geophysics., 36, 117-125. 5. Kimbrough, D.L., Tulloch, A.J., Coombs, D.S., Landis, C.A., Johnston, M.R. & Mattinson, J.M. 1994. Uraniumlead zircon ages from the Median Tectonic Zone, New Zealand. New Zealand Journal of Geology and Geophysics., 37, 393-419. 6. Bradshaw J. D., Dalziel, L W.D., DiVenere, V., Mukasa, S.D., Pankhurst, R.J., Storey, D.C. & Weaver, S. D. 1991. The southern rim of the Pacific: new work in the Pre-Cenozoic rocks of Marie Byrd Land. 6th Intemational Symposium on Antarctic earth sciencs. National Institute for Polar Research, Tokyo. 7. Di Venere, V. J., Kent, D.V. & Dalziel, I.W.D., 1995. Early Cretaceous paleomagnetic results from Marie Byrd Land, West Antarctica: Implications for the Weddellia collage of crustal blocks. Journal of Geophysical Research, 70ft B5, 8133-8151. 8. Bradshaw, J. D., Pankhurst, R. J., Weaver, S.D., Storey, B.C., Muir, R.J. & Ireland, T.R. in press. New Zealand superterranes recognised in Marie Byrd land and Thurston Island. Terra Antartica. 121

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

Pankhurst, RJ., Weaver, S.D., Bradshaw, J. D., Storey, B. C. and Ireland. The pre-Mesozoic geology of Marie Byrd Land, Antarctica. Submitted, Journal of Geophysical research. Lawver, L. A. & Gahagan, L. M., 1994. Constraints on timing of extension in the Ross Sea region. Terra Antarctica. 1 (3), 545-552. Bradshaw, J. D., Andrews, R B, & Field, D. B., 1983. Swanson Fomaation and related rocks of Marie Byrd Land and comparison with the Robertson Bay Group of north Victoria Land. In: R. L. Oliver, R R. James & J. B. Jago, (Qds.), Antarctic Earth Science. Australian Academy of Science, Canberra, 274-79. Adams, C. J. 1986. Geochronological studies of the Swanson Formation of Marie Byrd Land, West Antarctica, and its correlation with northem Victoria Land, East Antarctica and South Island, New Zealand. New Zealand Journal of Geology and Geophysics., 29, 345-358. Adams, C.L, Seward, D. & Weaver, S.D., 1995. Geochronology of Cretaceous granites and metasedimentary basement on Edward VII Peninsula, Marie Byrd Land, West Antarctica. Antarctic Science, 7, 265-277. Weaver, S. D., Bradshaw, J. D. & Adams, C.J., 1991. Granitoids of the Ford Ranges, Marie Byrd Land Antarctica. In: Thomson, M. R. A., Crame, J. A. & Thomson, J.W. (Eds.), Geological Evolution ofAntarctica, Cambridge University Press, 345-351. Adams, C. J., 1987. Geochronology of granite terranes in the Ford Ranges, Marie Byrd Land, West Antarctica: New Zealand Journal of Geology and Geophysics, 30, 345-358. Mukasa, S. B., Dalziel, I .W. D. & Pankhurst, R. J., 1994. U-Pb and ^Ar/^'Ar age constraints on the development and subsequent fragmentation of Gondwanaland's Pacific margin, Marie Byrd Land, Antarctica. EOS, Transactions american Geophysical Union., 75, #44, 692. Richard, S.M., Smith, C.H., Kimbrough, D.L., Fitzgerald, RG. & Luyendyk, B.R 1994. Cooling history of the northem Ford Ranges. Tectonics 13, 837-857. Weaver, S. D., Adams, C. J., Pankhurst, R. L and Gibson, 1. L., 1992. Granites of Edward VII Peninsula, Marie Byrd Land: anorogenic magmatism related to Antarctic-New Zealand rifting. Geological Society of America, Special Paper, 272, 281-290. Weaver, S. D., Storey, B. C., Pankhurst, R. J., Mukasa, S. B. DiVenere, V. L & Bradshaw, L D., 1994. AntarctciaNew Zealand rifting and Marie Byrd Land lithospheric magmatism linked to ridge subduction and mantle plume activity. Geology, 22, 811-814. Adams, C. J. & Kreuser, F., 1984. Potassium-argon age studies of slates and plyllites from the Bowers and Robertson Bay terranes, north Victoria Land, Antarctica. Geologisches Jahrbuch, B60, 265-288. Dallmeyer, R. D. & Wright, T. 0., 1992. Diachronous cleavage development in the Robertson Bay Terrane, northem Victoria Land, Antarctica: tectonic implications. Tectonics, 11,437-448. Stump, E., 1995. The Ross Orogen of the Transantarctic Mountains. Cambridge University Press. Stump , E., White, A. J. R. & Borg, S. G., 1986. Reconstructions of Australia and Antarctica: evidence from granites and recent mapping. Earth and Planarary Science Letters, 97, 348 Storey, B. C., Bradshaw, J. D., Ireland, T.R., Pankhurst, R. J. and Weaver, S. D., 1995. Paleozoic basement rocks in Marie Byrd Land. Abstracts VIIInternational symposium on Antarctic Earth Sciences, Siena, 1995, p. 362. Laudon, T. S. 1991. Petrology of sedimentary rocks from the English Coast, eastem Ellesworth Land. In: Thomson, M. R. A., Crame, J. A. & Thomson, J.W. (Eds.), Geological Evolution of Antarctica, Cambridge University Press, 455-465. Grindley, G. W. & Mildenhall, D. C., 1980. A mid-late Devonian Florafromthe Ruppert Coast, Marie Byrd Land, West Antarctica. Journal of the Royal Society ofNew Zealand, iO, 271-285. Pankhurst, R. L, Millar, 1. L., Grunow, A. M. & Storey, B. C., 1993. The Pre-Cenozoic magmatic history of the Thurston Island crustal block. West Antarctica. Journal of Geophysical Research, 98, B7,11835-11849. Leat, P. T., Storey, B. C & Pankhurst, R. J. 1993. Geochemistry of Palaeozoic-Mesozoic Pacific rim orogenic magmatism, Thurston Island, West Antarctica. Antarctic Science, 5, 281-296. Pallais, D.G., Mukasa, S.B., & Weaver, S. D. 1993. U-Pb and ^^Ar/^'Ar geochronology for plutons along the Ruppert Coast and Hobbs Coast, Marie Byrd Land. Abstracts, American Geophysical Union, Spring Meeting, 1993,1.123. Adams, C. J., Morris, P. A. & Beggs, J. M., 1979. Age and correlation of volcanic rocks of Campbell Island and metamorphic basement of the Campbell Plateau, Southwest Pacific. New Zealand Journal of Geology and Geophysics, 22, 679-691. Muir, R .J., Ireland, T. R., Weaver, S. D. & Bradshaw, J. D., 1994. Ion microprobe U-Pb zircon geochronology of granitic miagmatism in the Westem Province of the South Island, New Zealand. Chemical Geology, 113, 171-189. Muir, R. J., Ireland, T.R., Weaver, S. D. and Bradshaw, J. D., 1996. Ion microprobe dating of Paleozoic granitoids: Devonian magmatism in New Zealand and its correlation with Australia and Antarctica. Chemical Geology, 127, 191-210. Mortimer, N. Parkinson, D., Raine, J. I. Adams, C. J.,Graham, 1.1, Oliver, P. J. and Palmer, K., 1995. Ferrar magmatic province rocks discovered in New Zealand: implications for Mesozoic Gondwana geology. Geology, 23, 185-188. 122

34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.

.Muir, R.J., Weaver, S.D., Bradshaw, J.D., Eby, G.N. & Evans, J.A., 1995 Geochemistry of the Cretaceous Separation Point Batholith, New Zealand: granitoid magmas formed by melting of mafic lithospere. Journal ofthe Geological Society ofLondon, 152, 89-701. Kimbrough, D.L., Tulloch, A.J., Geary, E., Coombs, D.S. & Landis, CA., 1993. Isotopic ages from the Nelson region of South Island, New Zealand: crustal structure and the definition of the Median Tectonic Zone, Tectonophysics, 225, 433-488. Muir, R. J., Ireland, T. R., Weaver, S. D., Bradshaw, J. D. & Shelley, D., 1994. Geochronology of Eastem Fiordland, South Island, New Zealand. Geological Society of New Zealand Miscellaneous Publication, 80A, 140. Beresford, S. W., Bradshaw, J.D., Weaver, S.W. & Muir, R.J. in press. Echinus Granite and Pepin Group of Pepin Island, northeast Nelson, New Zealand: Drumduan terrane basement or exotic fragment in the Median Tectonic Zone. New Zealand Journal of Geology and Geophysics. Woodward , D.J. & Hatherton, T. 1975. Magnetic anomalies over southem New Zealand. New Zealand Journal of Geology and Geophysics, 18, 65-82. Davy, B., 1994. The Bounty Trough -basement structure influences on sedimentary basin evolution. In: Ballance, RF. (ed.), South Pacific Sedimentary Basins, Elsevier, Amsterdam, 69-92. Grindley G.W. & Davey F.J., 1982, The reconstruction of New Zealand, Australia and Antarctica. In: Craddock, C. (ed.), Antarctic Geoscience, University of Wisconsin Press, Madison, 15-29. Johnston, M. R., Raine, J. 1. & Watters, W. A., 1987. Drumduan Group of East Nelson, New Zealand: plant bearing Jurassic arc rocks metamorphosed during terrane interaction. Journal of the Royal Society of New Zealand ,17, 275-301. Chappell, B. W. 1994. Lachlan and New England: Fold belts of contrastsing magmatic and tectonic development. Journal of the Royal Society ofNew South Wales, 127, 47-59. Veevers, J.J. 1984. Phanerozoic earth history of AustraUa. Clarendon Press, Oxford, 418p. Coney, P.J. 1992. The Lachlan belt of eastem Austraha and circum-Pacific tectonic eveolution. Tectonophysics, 214, 1-25.

123

THE WHITSUNDAY VOLCANIC PROVINCE (CENTRAL QUEENSLAND) AND THE GIPPSLAND/OTWAY BASINS (VICTORIA): A COMPARISON OF EARLY CRETACEOUS RIFTRELATED VOLCANO-SEDIMENTARY SUCCESSIONS. Scott Bryan^, Andrew ConstantineS Chris Stephens'S Tony Ewarf, Richard Sch6n% John Parianos® 1 Department of Earth Sciences, Monash University, Clayton, Victoria 3168 Australia 2 Centre for Microscopy and Microanalysis, University of Queensland, St Lucia, Queensland 4072 Australia 3 Department of Earth Sciences, University of Queensland, St Lucia, Queensland 4072 Australia 4 Central Norseman Gold Corporation Ltd, Norseman, Western Australia 6443 Australia 5 QNILimited, PO Box 7879 Waterfront Place, Brisbane, Queensland 4001 Australia email: ST-bryan@artemis.earth.monash.edu.au Summary We report on two large volume Early Cretaceous provinces located at the northeastern and southeastern margins of the Australian continent. The Whitsunday Volcanic Province is part of an intermediate to silicic, high-K calc-alkaline volcanic belt extending along the central and southem Queensland coast. Hie Otway/Gippsland basins were initiated by the breakup of Antarctica and Australia, but contain up to 3km thickness of extrabasinal volcanogenic sediment supplied from the east. These two provinces are significant for: 1) the accumulation of large volumes of primary volcanic and volcanically-derived material; 2) the compositional similarity between phenocryst phases and detrital mineral grains; and 3) radiometric age dating indicating a major volcanic episode between 125-95 Ma. A causal relationship between volcanism (Whitsunday Volcanic Province) and volcaniclastic sedimentation in the Otway/Gippsland basins and Great Artesian Basin system is suggested. We propose these provinces record the breakup of eastem Gondwana and the formation of the eastem Austrahan passive margin.

Figure 1. Map showing locations of the Whitsunday volcanic Province (WVP) and Cretaceous sedimentary basins. MB, maryborough Basin; NEO, New England Orogen. Hollow arrows represent generalised palaeocurrent directions for the surat Basin [9] and Otway/gippsland basins. Introduction Two aspects of eastem Gondwana in the Cretaceous have remained enigmatic. These are: 1) the nature of the eastem Gondwanan plate margin, whether it was undergoing rifling and plate breakup, or convergence; and 2) the source of large volumes of volcanic sediment now preserved in the Otway/Gippsland basins and Great Artesian Basin system of southeastem and northeastem Australia, respectively. Due to 4 e lack of early Mesozoic structural and geologic 124

information in southeastern Australia, and the widespread coverage of northeastern Australia by the poorly studied Great Artesian Basin system, the Cretaceous tectonic setting of eastem Australia has remained poorly constrained. A convergent margin tectonic setting has been argued on the grounds of relating Cretaceous volcanics (e.g. Whitsunday Volcanic Province and Grahams Creek Formation of the Maryborough Basin; Fig. 1) in eastem Queensland to a magmatic arc, roughly coincident to the modem Queensland coastline [1, 2, 3]. The abundant volcanogenic detritus in the form of lithic fragments and plagioclase grains of apparent "andesitic" composition in Early Cretaceous Great Artesian Basin sediments, and the calc-alkaline afiSnities of volcanics supported this argument [4, 5, 6, 7]. The long history of plate subduction, magmatism and tectonism during the Palaeozoic and early Mesozoic for eastem Australia [e.g. 8] added a bias to this interpretation. At the same time as volcanism along the central and southem Queensland coast and volcanogenic sedimentation in the interior, large volumes of feldspathic-volcanic lithic sediment were being shed into the Otway, Gippsland and Bass basins in southeastem Australia. A major problem has been to identify the source of this volcaniclastic sediment as there are no interbedded volcanic rocks within the successions. Two theories on the origin of volcanic sediment are: 1) derivation from a continental "andesitic" arc to the east [10]; or 2) an intrarift volcanic source [11,12,13]. Altematively, seismic reflection profiles across the margins of Austraha [e.g. 14,15,16] identified marginal plateaus and complex rifl-basin systems adjacent to the (eastem) Australian continental shelf These rifl-basins, based primarily on rift-fill interpreted to be Cretaceous in age, indicated that rifling and development of the eastem Australian passive margin began in the Cretaceous, prior to seafioor-spreading in the Tasman Basin (Fig. 1). Detachment models applied to the margins of the Australian continent [e.g. 16,17] interpret eastem Australia as an underplated upper plate passive margin. Importantly, these models imply no prior convergence in the Jurassic-Cretaceous. Any continental arc probably became inactive during the Early Cretaceous when the Pacific Plate began moving northwest with respect to Austraha along a dominantly transform plate boundary [18]. The aim of this paper is to provide a comparative study of these Early Cretaceous provinces along the eastem Australian margin. A substantial isotopic age, geochemical and petrological database exists for the Whitsunday Volcanic Province [7,19], while fission track dating of detrital minerals [11], and petrological studies [12,13,20] provide a complimentary database for the Otway/Gippsland basins. Previous studies on detrital mineral compositions for the Otway/Gippsland basins have been supplemented in this study by 500 new mdcroprobe analyses. The mineral chemistry data, which provides the only available control on volcanic source characteristics for the Otway/Gippsland basins, will be evaluated, and compared to the temporally equivalent Whitsunday Volcanic Province. It will be shown that a number of similarities exist between these provinces, and collectively, place important constraints on the nature of eastem Gondwana during the Early Cretaceous. Although spatially isolated, we believe that the Whitsunday Volcanic Province and the contemporaneous inundation of volcanogenic sediment into the Otway/Gippsland basins record the same, continent-wide event. That is, volcanism and volcanogenic sedimentation was associated with a roughly N-S trending rift system more than 2,500 km in length, that developed along the present eastem Austrahan margin, approximately 125-120 Ma. Large volumes (>1,500,000 km3) of volcanic and volcanically-derived material characterised the early stages of dispersal of Gondwana from the eastem margin of the Australian continent. Whitsunday Volcanic Province The Whitsunday Volcanic Province comprises volcanic rocks and related granites exposed in the Whitsunday, Cumberland and Northumberland Island groups, and onshore exposures (Proserpine Volcanics of Clarke et al., [21]), along the central Queensland coast (Fig. 2). The province is part of a volcanic belt over 900 km long, --100 km wide and locally more than 2 km thick, that extends southwards through the Shoalwater Bay area near Rockhampton, and Maryborough Basin (Grahams Creek Formation). Volume estimates for the province are >30,000 km^ but exceeds 100,000 km^ when the full extent of this volcanic belt is taken into consideration. Volcanic and intrusive rocks range in age from 132 to 95 Ma, just predating seafioor-spreading in the Tasman Basin beginning ~ 96 Ma [2]. The main period of activity occurred between 120-105 Ma. Intrusive activity was largely coeval with volcanism, and a significant period of e?q)losive volcanism occurred -120-115 Ma. Volcanic lithologies in the Whitsunday Volcanic Province are dominated by dacitic to rhyolitic lithic ignimbrite, with intercalated surge, fallout, lag breccias, and phreatomagmatic deposits. Rhyolitic and dacitic domes and andesite lavas are subordinate, while basalt lavas, uncommon on the islands, are volumetrically more abundant in mainland exposures. Ignimbrite depositional units are commonly 10-100 mthick, and ignimbrite sequences often exceed 1 km on some islands (e.g. Hook, Whitsunday islands). Associated with the volcanics are locally significant thicknesses of coarse volcanogenic sedimentary rocks (conglomerate and sandstone) exposed at Cape Conway (-550 m) and South Molle (75 m). The sedimentary rocks are texturally and compositionally immature, reflecting the local provenance, and sedimentation appears to have been in a poorly confined alluvial environment in which upper flow regime conditions dominated.

125

Figure 2. Extent and generalised volcanic geology of the Whitsunday Volcanic Province. Numbers are K-Ar and Rb-Sr dates from Ewart et al. [19]. Locations of islands referred to in text are shown. The volcanic sequences are interpreted as subaerially deposited, with localised water-magma interaction [cf. 21]. The Hook, Whitsunday, Hamilton and Lindeman island sequences (Fig. 2) are interpreted by Ewart et al. [19] as "intracaldera" or proximal facies, based on: 1) the >1 km thickness of ignimbrite-dominated sequences on these islands; 2) a pervasive low grade alteration; 3) the abundance of intermediate to silicic dyke swarms; and 4) the occurrence of very coarse lag deposits with clasts up to 6 m in diameter. The paucity of debris-flow deposits and the burial of near vent deposits such as basaltic agglomerate by ignimbrites are interpreted to indicate an overall low-rehef depositional environment. The interstratification of proximal/near-vent lithofacies such as rhyolite domes and lavas, basaltic agglomerate and agglutinate, and medial to distal lithofacies including ignimbrite and surge deposits suggests a multiple vent volcanic environment. Stratigraphic analysis from the sequence on South Molle shows that volcanism progressed from an initial explosive phase, producing a >1 km thick sequence of dacitic to rhyolitic ignimbrites, to a later, bimodal, mixed pyroclastic-effusive assemblage, >1 km thick, comprising rhyolite ignimbrites and basalt and rhyolite lavas. Volcanism is concluded to have occurred in a low-relief extensional environment, with no evidence from the facies architecture to support a high-standing 'Andean'-type magmatic arc. The phenocryst mineralogy comprises plagioclase, augite, Fe-Ti oxides, with lesser hypersthene (altered), homblende, quartz, biotite and sanidine. Importantly, plagioclase is the primary phenocryst phase in all volcanic rocks, regardless of whole rock composition. Although plagioclase is commonly altered to albite, remnant primary compositions are generally well preserved to determine compositional ranges (Fig. 3). Dolerites, basalts and andesites typically range from bytownite to labradorite, while phenocrysts from dacite to rhyolite lavas and dykes are labradorite to andesine in composition. Significantly, plagioclase compositions in the volumetrically dominant dacitic and rhyolitic ignimbrites are primarily andesine to oligoclase, and are distinctfromthe coherent lithologies. Calcic pyroxenes are relatively magnesian and show little variation in Fe-Mg-Ca ratios amongst the dolerite to low-silica rhyolite compositional range [19]. The calcic amphiboles (Fig. 4) are also relatively magnesian, ranging from edenite through to ferroan pargasite for the majority of volcanics. A small subpopulation of magnesian to ferro-homblendes occurs in the rhyolite ignimbrites, while amphibole compositions from granites are distinct, exhibiting a trend towards Fe-enrichment. The volcanic suite exhibits a broad spectrum of compositions from basalt to high-siUca rhyolite, with high-K calcalkaline affinities and geochemical signatures similar to modem convergent-margin volcanics. Trace element and isotopic studies indicate that the broad spectrum of compositions are generated by two-component magma mixing and superimposed fractional crystallisation in the high-silica rhyolites. The two magma sources are defined as: 1) a large volume, partial melt of relatively young crust with a non-radiogenic, calc-alkaUne character, and 2) a tholeiitic basalt (near E-MORB) with a geochemical character similar to the Tertiary within-plate basalts of eastem Australia [19, 22]. Three important implications of the geochemical study are: 1) that the calc-alkaline signature is inherited from the crustal source; 2) the basaltic end-member is distinct from basalts that floor back-arc basins which are transitional between NMORB and island arc or calc-alkaline basalts [cf. 23]; and 3) that a large thermal flux into the cmst is required to maintain volcanism.

126

Otway/Gippsland basins The Otway and Gippsland basins are two of three WNW-trending, transtensional rift basins which form the eastern extremity of a complex rift system that extended along the length of the southem margin of Australia during the Late Jurassic - Early Cretaceous. Rift basin formation was a precusor to the post-Middle Cretaceous break-up ofAustralia and Antartica. Despite being separated by a 25 km wide basement high (Fig. 5), palynological dating of sediments on the basement high, coupled with vitrinite reflectance and apatite fission track data, indicate the two basins were connected during the Late Jurassic-Early Cretaceous and the basement high is a Late Cretaceous - Miocene inversion structure. The Late Jurassic - Early Cretaceous sequences in the Otway and Gippsland basins are referred to as the Otway and Strzelecki groups respectively (Fig. 5). The sequence in each basin can be divided into three tectonolithostratigraphic units representing: pre ± syn-rift, syn-rift, and post-rift (sag) phases of deposition. Of specific interest are the post-rift (sag) phase, fluvial-lacustrine Eumeralla and Wonthaggi formations (Fig. 5). Both units are >2,500 m thick and dominated by sandstone and mudstone. The sandstones are lithic-rich with an average Q:F:Rratio of 15:10:75 (Eumeralla Formation) and 8:26:66 (Wonthaggi Formation), and reflect a major change in sediment provenance from the underlying Crayfish Subgroup (>4,500 m thick) and Tyers Subgroup (>425 m thick), that consist predominantly of basement-derived quartz sandstones with an average Q:F:R ratio of 84:10:6 (Crayfish Subgroup; Whittle, [25]) and 95:1:4 (Tyers Subgroup). Petrographic analysis of the Eumeralla and Wonthaggi formation sandstones indicate volcanic lithic grains are of andesitic, dacitic and rhyolitic composition, and average 78% (Eumerella Fomiation) and 83% (Wonthaggi Formation) of the total lithic component. Detrital minerals are predominantly plagioclase, with lesser homblende, pyroxene, apatite, sphene, and zircon. Fission track dating of apatite, sphene, and zircon in these units (Fig. 5) indicates a contemporaneous volcanic source for the sediment [11]. This is supported by palynological dating that also indicates the Eumeralla and Wonthaggi formations are Aptian - Albian in age. Unaltered, detrital plagioclase grains are primarily andesine to oligoclase in composition, while calcic amphiboles are relatively magnesian in composition, ranging from actinolites through homblendes to pargasite (Figs 4 & 5). Calcic pyroxenes are also relatively magnesian and exhibit a similarity to (high-K) calc-alkaline volcanic suites [12]. Palaeocurrent measurements from the exposed Aptian section of the Wonthaggi Formation in the Gippsland Basin, and the exposed Albian section of the Eumeralla Formation in the Otway Basin, indicate the source of the volcaniclastic sediment lay to the east of the Gippsland Basin (Fig. 5). This palaeocurrent data precludes an intra-rift volcanic source for the sediment [c.f. 11,12], as does the complete absence of interbedded volcanic rocks within these units. Furthermore, these units cannot have been sourced from volcanics associated with the onset of basin development, i.e. the early- to mid-Tithonian Casterton Formation and Duck Bay Volcanics. These units consist of olivine basalt, vitric tuff, volcanic breccia, conglomerate, dark red-brown shale and sandy siltstone [24] and are considerably older than the fission track ages obtained from detrital minerals in the Eumerella Formation [11].

127

251-

WHITSUNDAY VOLCANIC PROVINCE

WHITSUNDAY VOLCANIC PROVINCE 20--

>s o §

Dacite and rhyolite ignimbrites

Dolerite/gabbro dykes and sills

N = 180

N = 162 15--

D CT

1 0 -

IL.

j^Jkui <2

8-10

28-30

48-50 %Ab

68-70

88-90

>98

WHITSUNDAY VOLCANIC PROVINCE

Andesite dykes and lavas

N = 33

CT £

LJ.

8-10

28-30

48-50

68-70

88-90

>98

%Ab 25

251

SURAT BASIN

WHITSUNDAY VOLCANIC PROVINCE 20-

Dacite dykes and lavas N = 20

E£ 10

C 0) D

Wallumbilla Fm (Rolling Downs Group), Bungil Fm and Mooga Sst N = 13

JUUU <2

8-10

28-30

68-70

48-50

88-90

<2 8-10

>98

28-30

68-70

88-90

>98

WHITSUNDAY VOLCANIC PROVINCE 20

& C

48-50

%Ab

Rhyolite dykes and lavas

WHITSUNDAY VOLCANIC PROVINCE Granites N= 9

N = 57 15

<D 13 £

JUUSL <2 8-10

28-30

48-50

%Ab

58-70

U l

88-90

>98

JULfL <2 8-10

28-30

48-50

68-70

88-90

>98

%Ab

Rgure 3, Histograms showing microprobe analyses of plagioclase phenocrysts in coherent and pyroclastic compositional groups from the Whitsunday volcanic Province, and detrital grainsfromthe Otway/Gippsland basins and Surat Basin (Great Artesian Basin system). Data for the Surat Basin from Hawlader [9]. N is number of analyses, and note the change in scale for the Otway/gippsland basins dataset.

128

• (A2)

Alkali

. O Granite, Scawfell Is. . • Qtz porphyry, Derwent Is.

Na-Ca

Fe-Mg-Mn

Calcic ,

2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 BCa+BNa

Ca+Na+K ( 0 = 23) 8.0 /p- \ ^ '

Orogenic andesites, dacites rhyolitesi Bimodal itiyolites'

• (B2)

Edenite-E]

fTj fS

I ^

Alkali Fe-Mg-Mn

6.0cj-Tschermakite

lite •£]

• A Ignlmbrite, Penrith Is. . 0 Ignimbrite, Carlisle Is. • Doierite, Esk Is. • a Ignimbrite, Undeman Is. X Ignimbrite, Keswick Is. < Ignimbrite, Whitsunday Is. • • Ignimbrite, Esk Is. A Ignimbrite, Penrith Is. ' (B Rhyollte dyke, Shaw Is.

Na-Ca

Calcic

2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3.0 BCa+BNa

Ca+Na+K ( 0 = 23) 8.0 (CI)

• (C2)

Whitsunday Volcanic Province (volcanics) Edenite i ]

PQ ^

AlkaU

6.0CI-Tscheimakite

Pargasite -f ]

• • This study - O Jaski (1994)

Na-Ca

Fe-Mg-Mn Calcic

N = 25

.ill.

2.0 2.1 2.2 23 2.4 2.5 2.6 2.7 2.8 2.9 3.0 Ca+Na+K ( 0 = 23)

BCa+BNa

Figure 4. Compliation of calcic amphibole data from the Whitsunday Volcanic Province (phenocryst) and Otway/Gippsland basins (detrital grains): Al) & A2) granites; El & B2) volcanics; and CI) & C2) Wonthaggi Formation. Note the complete ovelap of homblende compositions between these two regions and with the field of "orogenic" intermediate and siHcic volcanics. 1) Data from Ewart [26]. All homblende analyses recalculated on the basis of all Fe and Fe^^ N is number of analyses. Great Artesian Basin system The Great Artesian Basin system (Fig. 1), comprising the Eromanga, Surat, Clarence-Moreton and Carpentaria basins, is a large intracratonic sag that developed initially in latest Triassic times, and was active during much of the Mesozoic era [28]. In the context of this paper, the Great Artesian Basin system is significant for containing Early Cretaceous (AptianAlbian), largely volcanogenic sedimentary rocks (Rolling Downs Group) which cover some 2,000,000 km' to an average thickness of-500 m [5]. The Rolling Downs Group is dominated by mudstone, siltstone and sandstone, with depositional environments ranging from fluvial/lacustrine to coastal plain and shallow marine [6,28]. Sandstones are feldspathic-lithic in composition, with more than 90% of the lithic grains volcanic, and unaltered, relatively fresh feldspar grains are 129

predominantly andesine in composition ([6]; Fig. 4). The Rolling Downs Group records a major change in sedimentary provenance from the underlying (Neocomian), basement-derived quartzose sandstones. Limited palaeocurrent data from the Surat Basin [8] indicate an easterly source for the sediment (Fig. 1).

Undifferentiated Palaeozoic metasedlmentary and granitic basement (Lachlan Fold Belt)

90-

Spore-pollen Zorws

100-

500

Otway Basin

Rpannosus Albian

110-

120-

C.paradoxa C.striatus

Aptian

Rnotensls

(Ma) 130-

Barremian Hauterlvian

140-

D.spedosus

Valanglnian Berralslan Tithonlan

i.spheneand m fission track ages with 2 Sigma errors < 1 2 Ma

C.australlensls R.watherooensis

KImmerldglan M.florlda Oxfordlan C.cooksonlae Callovlan

—I—I—I—I—r(Ma)

Number of grains

Figure 5. A) Outcrop and subsurface extent of the Otway Group (Otway Basin) and Strzelecki Group (Gippsland Basin). B) Stratigraphic subdivision of the Otway and Strzelecki groups. Outcrop age ranges are illustrated by vertical black bars beside stratigraphic columns. Palaeocurrent measurements are for the R notensis Zone (Wonthaggi Formation) and C striatus-R pannosus zones (Eumeralla Formation). Fisson track data from Gleadow & Duddly [11] and time scale after Harland et al. [27]. Discussion and Comparison The outline above points to a major volcanic episode along the present eastem Australian margin during the Early Cretaceous. The volcanic architecture of the Whitsunday Volcanic Province argues against a hi^-standing 'Andean'-type magmatic arc. Deltaic/coastal plain to shallow marine facies associations from the adjacent Styx Basin (Fig. 1) and Maryborugh Basin [28] further indicate a low rehef depositional environment for eastem Australia in llie Early Creatceous. Several points also argue against volcanism being related to continental back-arc extension. Geochemical data from basaltic rocks provide the most compelling evidence, which are clearly distinct from basalts that are associated with back-arc basins. If volcanism was related to back-arc extension, then this would imply that the Tasman Basin system (including the Cato Trough and Coral Sea Basin, Fig. 1) is a marginal/back-arc basin. This contradicts a substantial body of work that considers the Tasman Basin system to be formed by normal seafloor-spreading, and bordered by a passive 130

continental margin [e.g. 14,15,16,17]. A characteristic feature of back-arc extensional environments is the migration of the locus of extension and volcanism as extension proceeds (e.g. mid-Tertiary "ignimbriteflare-up"of westem USA, Best & Christiansen [29]; Axen et al. [30]; Early Pliocene of Lau Basin, Clift et al. [31]). The Whitsunday Volcanic Province is notable for its prolonged magmatic history, spanning 37 my, and elongate, narrow geometry. We want to emphasize that the Otway/Gippsland basins and Whitsunday Volcanic Province show a number of similarities. These include: 1) a similarity in mineral compositions, which also extends to the Great Artesian Basin system; 2) both regions showing similar age ranges, with a main period of activity between -120 and 105 Ma; and 3) the production of large volumes of primary volcanic and volcanically-derived material. The fresh nature of the detrital grains supports the age data that the sources of sediment were active volcanoes. As noted by Ewart et al. [19], the phenocryst composition and mineralogy of volcanics from the Whitsunday Volcanic Province are similar to modem "orogenic" destructive plate margin volcanics (e.g. Fig. 5). Previous studies on detrital mineral compositionsfromthe volcanogenic sediments of the Great Artesian Basin [6] and Otway/Gippsland basins [12, 13,20] illustrated the same similarity to convergent margin volcanics. The main feature of the mineral chemistry plots of Figures 4 & 5, is the close similarity of feldspar and homblende compositions between the dacitic to rhyolitic pyroclastic rocks of the Whitsunday Volcanic Province and detrital grainsfromthe volcaniclastic sediments of the Otway/Gippsland basins. Limited feldspar datafromthe Surat Basin [9] also correspond to compositions from dacitic-rhyolitic ignimbrites of the Whitsunday Volcanic Province. Pyroxene datafromthe Whitsunday Volcanic Province [19] and Otway/Gippsland basins [12] illustrate the same relationships. Duddy [12] noted that plagioclase compositions from the Otway Basin are distinctly more sodic than a range of low-K to high-K calc-alkaline basalts and andesites, and only the more sihcic compositions approach the Otway/Gippsland basins detrital mineral chenistry. The feldspathic nature of the sedimentary rocks may be a reflection of plagioclase being the dominant phenocryst phase in the volcanics. The overlaps between the Whitsunday Volcanic Province, Otway/Gippsland basins (and Great Artesian Basin) impHes a remarkable consistency in mineral composition, and consequently, whole rock chemistry, for Early Cretaceous volcanism. Based on these relationships, a (dacitic-) rhyolitic pyroclastic source, analogous to the ignimbrites from the Whitsunday Volcanic Province, is tentatively proposed for the volcaniclastic sediment of the Otway/Gippsland basins. Note that the slightly more sodic plagioclase compositions from the Otway/Gippsland basins may imply a somewhat more evolved volcanic source than that of the Whitsunday Volcanic Province. Age data from both regions are consistent in showing a broad range of activity: 126 -103 Ma for the Otway/Gippsland basins [11], and 132 -95 Ma for the Whitsunday Volcanic Province [19]. Early Cretaceous volcanics exposed at Cape Portland in NE Tasmania (Fig. 1), dated at 95 - 99 Ma [32, 11], suggest that the duration of volcanism in southeastem Australia may be closer to that of the Whitsunday Volcanic Province. Importantly, this age distribution correlates with the main period of activity identified for the Whitsunday Volcanic Province (120 - 105 Ma), and thus, a major volcanic episode appears to have been initiated 125-120 Ma along the present eastem Australian margin. The base of lithicfeldspathic sediments of the Rolling Downs Group in the Great Artesian Basin is near the Barremian-Aptian boundary (124.5 Ma; Fielding [28]), and provides another record of a major volcanic episode beginning 125-120 Ma. The third important factor regarding these Early Cretaceous provinces, is the large volumes of primarily volcanic-derived sediment accumulated in the Otway/Gippsland basins (>400,000 km^ and Great Artesian Basin system (>1 miUion km^; Smart & Senior, [5]), as well as >30,000 km^ of volcanics in the Whitsunday Volcanic Province. Of particular note is the rapid inflix of such large volumes of volcanogenic sediment into the Otway/Gippsland basins, where volcanogenic sediment was ahnost to the exclusion of basement-derived material [11]. Conclusions From the above discussion and comparison, we draw several conclusions. 1) A major volcanic episode was initiated approximately 125-120 Ma, and continued for another 25-30 my. 2) Volcanism occurred along the length of the present eastem Australian margin, a distance of >2,500 km, with the Whitsunday Volcanic Province interpeted to represent the northem extension of this volcanic belt. 3) Although the Otway/Gippsland basins formed in response to AntarcticAustralia rifting, volcanogenic sediment of the Eumerella/Wonthaggi formations was derived from a contemporaneous volcanic source further to the east, outside the basins. 4) Phenocryst compositionsfromvolcanics (Whitsunday Volcanic Province) and detrital grains in volcanogenic sediments (Otway/Gippsland basins and Surat Basin) consistently overlap, suggesting an apparent uniform chemical character to volcanism. Plagioclase compositions are distinct from basalts and andesites from both the Whitsunday Volcanic Province, and modem orogenic volcanic suites, and closely approximate the (dacitic-) rhyolitic ignimbrite compositional grouping from the Whitsunday Volcanic Province. Dacitic to rhyolitic ignimbrites volumetrically dominate the Whitsunday Volcanic Province, and the overlap with this compositional group suggests that silicic pyroclastic volcanism characterised the Early Cretaceous volcanic event. 5) Despite the volcanic architecture showing many similarities to volcanic arc complexes in extensional environments (e.g. Taupo Volcanic Zone), the within-plate geochemical signature to mafic volcanics of the Whitsunday Volcanic Province precludes volcanism from being related to back arc extension. These conclusions agree with the interpretation of the eastem Australian margin as a passive margin. Our observations add weight to the notion that break-up of eastem Gondwana began in the Early Cretaceous. 131

The critical point is that volcanism occurred along the entire length of the present margin of eastern Australia, and was sudden in its initiation, and voluminous in its products. We suggest an active volcanic rift system, trending roughly N-S, and extending for over 2,500 km, was initiated approximately 125-120 Ma, along eastem Gondwana, This rifting event led to the fragmentation of eastem Gondwana, now dispersed as segments of continental cmst in the SW Pacific. The intersection of the Antarctic-Australian rift and this -^N-S rift system to the east of the Gippsland Basin, allowed the shedding of volcanic material towards the west into the Otway/Gippsland basins. Volcanic rocks, absent along the NSW section of the continental margin, may be present on the Lord Howe Rise, and the -96 Ma rhyolites recovered from deep sea drilling [33], may be the youngest expression of this volcanism. References 1 Jones JG, Veevers JJ, Mesozoic origins and antecedents of Australia's Eastem Highlands. J Geol Soc Aust (1983) 30: 305-322 2 Veevers JJ, Powell C McA, Roots SR, Review of seafloor spreading around Australia. 1. Synthesis of the patterns of spreading. Aust J Earth Sci (1991) 38: 373-389 3 EUiot LG, Post-Carboniferous tectonic evolution of eastem Australia. APEA J (1993) 215-236 4 Exon NF, Senior BR, The Cretaceous of the Eromanga and Surat Basins. BMR J Aust Geol Geophys (1976) 1: 35-50 5 Smart J, Senior BR, The Jurassic-Cretaceous basins of northeastern Australia. In: Henderson RA, Stephenson PJ (eds) The Geology and Geophysics of Northeastern Austraha. Geol Soc Aust, QLD Div, Brisbane (1980) 1-26 6 Halwader HM, Diagenesis and reservoir potential of volcanogenic sandstones — Cretaceous of the Surat Basin, Australia. Sed Geol (1990) 66: 181-195 7 Webb AW, McDougall I, The geochronology of the igneous rocks of Eastem Queensland. J Geol Soc Aust (1968) 15:313-346 8 Stephens CJ, Tectonic evolution of the Gondwanaland margin during the Late Palaeozoic and Early Mesozoic. Abstr Mesozoic 96 (1996) this volume 9 Hawlader HM, Petrology, diagenesis, and reservoir potential of the Surat Basin sandstones with special reference to Hydrocarbon e:?q)loration. Unpubl. PhD thesis, Macquarie University, Aust (1989) 348 p 10 Veevers JJ, Eittreim SL, Reconstruction of Antarctica and Australia at breakup (95±5 Ma) and before rifting (160 Ma). Aust J Earth Sci (1988) 35: 355-363 11 Gleadow AJW, Duddy IR, Early Cretaceous volcanism and the early breakup history of southeastern Australia: Evidence from fission track dating of volcaniclastic sediments. Fifth Intemational Gondwana Symposium, Wellington (1980) pp 11-16 12 Duddy IR, The geology, petrology and geochemistry of the Otway Fomiation volcanogenic sediments. Unpubl PhD thesis, Melboume University, Aust (1983) 426 p 13 Felton A, Sedimentation and facies analysis of the Otway Basin. Unpubl PhD thesis. University of Wollongong, Aust (1992) 307 p 14 Falvey DA, Mutter JC, Regional plate tectonics and the evolution ofAustralia's passive continental margins. BMR J Geol Geophys (1981) 6:1-29 15 Symonds PA, Fritsch J, Schluter HU, Continental margin around the western Coral Sea Basin: Stmctural elements, seismic sequences and petroleum geological aspects. In: Watson ST (ed) Trans Third Circum-Pacific Energy and Mineral Resources Conf, Hawaii. AAPG (1984) 243-252 16 Lister GS, Etheridge MA, Symonds PA, Detachment models for the formation of passive continental margins. Tectonics (1991) 10: 1038-1064 17 Lister GS, Etheridge MA, Detachment models for uphft and volcanism in the Eastem Hi^lands, and their implication to the origin of passive margin mountains. In: Johnson JW, Taylor SR (eds) Intraplate volcanism in Eastem Australia and New Zealand. Cambridge University Press, Cambridge (1989) pp 297-313 18 Taylor LWH, Falvey DA, Queensland Plateau and Coral Sea Basin - stratigraphy, sturcture and tectonics. Aust Petrol Expl Geophys (1977) 6: 33-35 19 Ewart A, Schon RW, Chappell BW, The Cretaceous volcanic-plutonic province of the central Queensland (Australia) coast - a rift related "calc-alkaline" province. Trans Roy Soc Edin: Earth Sci (1992) 83: 327-345 20 Jaski C, The sedimentology, provenance and stmcture of the Otway Group in the vicinity of the Merino High, west Victoria. Unpubl Hons thesis. La Trobe University, Aust (1994) 86p 21 Clarke DE, Paine AGL, Jensen AR, Geology of the Proserpine 1:250 000 Sheet area, Queensland. BMR Geol Geophys Rep 144 (1971) 22 Stephens CJ, Ewart A, Bryan S, Schon RW, Rift-related, large-volume siUcic volcanism associated with Lower Cretaceous continental breakup, eastem Australia. Abstr 1995 lUGG XXI General Assembly, Boulder, (1995) A443 23 Saunders AD, Tamey J, Geochemical characteristics of basaltic volcanism within back-arc basins. In: Kokelaar BP, Howells MF (eds) Marginal basin geology: volcanic and associated sedimentary and tectonic processes in modem and ancient marginal basins. Geol Soc Spec Pub No. 16 (1984) 59-76 24 BMR, A preliminary review of the Otway Basin. Unpubl Rec Bur Miner Resour Geol Geophys Aust (1966) 170 132

25 26

Whittle AP, E}q)loration in the Otway Basin. APEA J (1968) 8: 78-87 Ewart A, A review of the mineralogy and chemistry of Tertiary-recent dacitic, latitic, rhyolitic, and related salic volcanic rocks. In: Barker F (ed.) Trondhjemites, Dacites, and Related Rocks. Elsevier, Amsterdam (1979) pp 13-

Harland WB, Armstrong RL, Cox AV, Craig LE, Smith AG, Smith DG, A geologic time scale 1989. Cambridge Univeristy Press, Cambridge (1989) pp 172-174 Fielding CR, A review of Cretaceous coal-bearing sequences in Austraha. Geol Soc Am Spec Pap (1992) 267: 303-324 Best MG, Christiansen EH, Limited extension during peak Tertiaryvolcanism, Great Basin of Nevada and Utah. J Geophys Res (1991) 96: 13509-13528 Axen GJ, Taylor WJ, Bartley JM, Space-time pattems and tectonic controls of Tertiary extension and magmatism in the Great Basin of the western United States. Geol Soc Am Bull (1993) 105: 56-76 Clifl PD, ODP Leg 135 Scientific Party, Volcanism and sedimentation in a rifling island-arc terrain: an example from Tonga, SW Pacific. In: Smellie JL (ed.) Volcanism associated with extension at consuming plate margins. Geol Soc Spec Pub No. 81 (1995) 29-51 Sutherland FL, Corbett EB, The extent of Upper Mesozoic igneous activity in relation to lamprophyric intrusions in Tasmania. Pap Proc R Soc Tasmania (1974) 107: 175-190 McDougall I, van der Lingen GJ, Age of the rhyolites of the Lord Howe Rise and the evolution of the southwest Pacific Ocean. Earth Planet Sci Lett (1974) 21:117-126

28 29 30 31 32 33

121

133

BENTONITES IN QUEENSLAND David Carmichael, Geoscientist Queensland Department of Mines and Energy, Brisbane Summary The importance of bentonite as an industrial mineral in Queensland is seen in the 1994-95 production figures of 226,981 tonnes valued at $20 million. Bentonite produced in Queensland is used in a wide range of applications in both the primary and manufacturing industries. The many characteristics of bentonite include its ability to exchange cations, its swelling and hydration capacity, its ability to act as a binder and its impermeability. Other specific applications such as in drilling muds use its properties of viscosity and thixotropy. There are six producers of bentonite in Queensland, all located in the southeast of the State. Of these six producers, five mine bentonite from either the Late Jurassic Orallo Formation or the Middle Jurassic Walloon Coal Measures. Introduction Bentonite is a clay consisting domdnantly of smectite minerals, most commonly montmorillonite. This clay is soft and highly plastic, with a waxy appearance and soapy texture. Its colour may be white, grey, green, or brown. The important properties of bentonite include; its ability to exchange cations; its swelling and hydration capacity; its ability to act as a binder, its impermeability; and its viscosity and thixotropy. Most bentonites have high cation exchange capacities ranging between 60-170meq/100g. The many uses for bentonite are based on the chemical activity provided by the cation exchange characteristics. The most common and commercially significant varieties produced in Queensland are sodium, and calcium-magnesium bentonites. Sodium bentonites have very high swelling capacities, swelling 15 to 20 times their original volume, forming stable gel-like masses. As well, sodium bentonites are excellent absorbents, absorbing up to 10 times their own weight of water. Sodium bentonites also exhibit good dry strength, plasticity, lubricity, impermeability, and low compressibility. Calcium-magnesium bentonites have lower swelling and absorbency capacities than sodium bentonite. Magnesium bentonites have varying swelling capacities. In Queensland, the main uses of bentonite are in stock feed, in pet litter, and as a sealant. More specialised uses for higher quality products are in foundry sands, drilling muds, civil engineering applications, horticulture, wine clearing, and ceramics. Sodium bentonite is used as a pellet binder in stock feed. It also increases retention times in the digestive tract to aid food digestion and helps to prevent acidosis. Bentonite is added to cattle, sheep, pig, and poultry feed. Sodium or calcium-magnesium bentonite is used in pet litter as an absorbent and to control odour. Sodium bentonite is used in clumping or scoopable pet litter, whereas calcium-magnesium bentonite is used in non-clumping pet litter. In civil engineering projects, sodium bentonites are excellent water sealants and are used for sealing dams, irrigation ditches, waste disposal ponds, and tunnel walls. Sodium bentonites are also used in the excavation of unconsoUdated rock or soil especially in landfill areas, using the 'slurry trench' or 'diaphragm wall' method. Sodium or calcium bentonite is used as a binding agent in foundry sand moulds. In water-based drilling muds, sodium or calcium bentonite is used to lubricate the drill bit, seal the walls of the hole, and keep the cuttings in suspension. Occurrences Bentonite deposits in Queensland occur mainly in sediments deposited in marine, stream, lake, or swamp environments. They are a product oftihein situ alteration of volcanic ash or tuff after contact with water. The timing of this alteration of the minute volcanic glass particles in the ash or tuff is not known. Alteration occurs either as soon as the ash or tuff comes in contact with the water, or when it reaches the bottom of the water body, or after burial. Commercial bentonite deposits in Queensland range in age from Late Permian to Tertiary. These deposits occur in the Late Permian Black Alley Shale; the Middle Jurassic Walloon Coal Measures; the Late Jurassic Orallo Formation; the Tertiary Petrie Formation; and in altered Tertiary volcanics in the Yarraman region. The locations of the bentonite deposits referred to are given in Table 1.

134

Table 1: Bentonite occurrences in Queensland. NAME LOCATION LONGITUDE Australian Bentonite Mine 30kmNWofMnes 150°03'55"E Bungeworgorai Creek prospect 31 km NW of Roma 148°35'16"E Ebenezer Mine 6 km SE of Rosewood 152°38'45"E Jeebropilly Mine 7 km ESE of Rosewood 152°39'27"E Malabar Mine prospect 6 km N of Rosewood 152°36'09"E Miles Bentonite Mine 5 km SW of Miles 150°08'57"E Oakleigh Colliery 3 km N of Rosewood 152°35'17"E Queensland Bentonite Mine 36 km NW of Miles 149°59'44"E Sandy Creek prospect 30kmSEofMUes 150°19'47"E Yuleba Creek prospect 70 km NE of Roma 149°26'33"E

FORMATION

LATITIIDE

Orallo

26°27'01"S

Orallo

26°2r41"S

Walloon Coal Measures

27°40'03"S

Walloon Coal Measures 27°39'14"S Walloon Coal Measures 27°35'40"S Orallo

26°4r44"S

Walloon Coal Measures

27°37'39"S

Orallo

26°24'37"S

Orallo

26°53'24"S

Orallo

26°2r30"S

Most of the production of bentonite in Queensland comesfromthe Upper Orallo Formation. Commercial grade sodium bentonite occurs near the surface in the Miles district, about 350km west of Brisbane. The Late Jurassic Orallo Formation of the Surat Basin comprises thickly bedded cross-stratified lithic sandstone, thin bedded siltstone and mudstone, and minor conglomerate, bentonite and coal (Exon, 1). This Formation was deposited in streams, lakes, and swamps. The presence of tree trunks in the bentonite and other rocks within the Orallo Formation, indicates periods of rapid deposition. Lateral extensions of the bentonite layers are largely controlled by the surface area of the water body in which the tuff was deposited. Three mining operations produce bentonite from the Orallo Formation. The largest and longest operating mine in the Miles district is AustraUan Bentonite at Gurulmundi, about 30km north of Miles. This company, a division of Cudgen RZ Ltd, has a combined production capacity of 120,000t/year. Out of the total reserves of 12Mt, there is over 3Mt of premium grade sodium bentonite. An average of 9m of sandstone overburden is ripped and removed by bulldozers. Three bentonite seams, with a combined average thickness of 6m, are selectively mined and stoclq)iled by elevating scrapers. The bentonite is spread out in drying areas and a rotary hoe is used to break down the larger pieces to aerate the material. Drying in the sun reduces the moisture contentfroman original 30-35%, to 15%. This process takes about three days in summer and a week in winter. From the drying areas, three grades are stored undercover. This bentonite is further dried to a moisture content of 12% in a gasfiredrotary kiln, then crushed in a roller mill and the different sizes cycloned off. The processing is accredited to International Standard ISO 9001 and to Q1 through the American Petroleum Institute. Products are used in a number of applications such as: an additive in stock feed; in foundry moulds; in the manufacture of refractory linings; in the manufacture of ceramics, in civil engineering; for clearing wines; as an additive in drilling muds; and in horticulture. Soda ash and a polyanionic cellulose is added to the premium quality bentonite for the drilling mud and engineering grade product. The bentonite product is sold in paper sacks, in bulkabags, and in bulk. Another significant bentonite operation is Queensland Bentonite, located about 6km northwest of Gurulmundi. Production of a high swelling sodium bentonite, a medium swelling sodium bentonite, and a magnesium bentonite began in 1993. Proven reserves are 2Mt, with a further lOMt of probable reserves indicated by drilling. The processing plant has a production capacity of 60,000t/year. The bentonite is sold in bags or in bulk, and is used in stock feed, in cat litter and as an oil absorbent. The third and most recently developed operation in the district is the Miles Bentonite mine, located 5km southwest of Miles. Production of a sodium bentonite from an open cut started in 1994. Proven reserves are 0.4 to 0.5Mt. About 5m of conglomerate and sandstone overburden is ripped and removed, and the bentonite extracted by scraper and stockpiled. This product is sold in bulk or bags and used in stock feed, in medicines, in cat litter and for sealing dams. Many outcrops of bentonite occur in the Orallo Formation in the Miles-Gurulmundi-Wooleebee districts, and other isolated outcrops of bentonite also occur 31km northwest of Roma at Bungeworgorai Creek, 29km north ofYuleba at 135

Yuleba Creek, and about 30km southeast of Miles at Sandy Creek. These outcrops form part of the 'bentonite belt' which extends from north of Roma, east to Miles and south to Moonie. Seams of bentonite occur in the Middle Jurassic Walloon Coal Measures around Rosewood, 50km southwest of Brisbane, at the Oakleigh, Jeebropilly and Ebenezer coal mines and in a small disused underground mine at Malabar north of Rosewood. The Middle Jurassic Walloon Coal Measures of the Moreton Basin and Surat Basin, comprises volcanolithic sandstone, carbonaceous siltstone, shale, mudstone, coal, bentonitic siltstone and minor oil shale (Wells & O'Brien, 2). The depositional environment has been interpreted as channels, overbank and backswamps of meandering streamis and flood plain deposits. In the Rosewood-Walloon area, Fielding (3), interpreted the depositional environment as an extensive alluvial plain crossed by meandering streams. In the peat-forming wetlands of the plains, ash from periodic volcanic eruptions was preserved. In many cases the ash has been altered to bentonite. At New Hope Corporation Ltd's, Jeebropilly open cut mine, about 7km southeast of Rosewood, a calcium-magnesium bentonite is mined from a layer approximately 0.5-1.Om thicL The bentonite occurs in the 7 Tops seam at the bottom of the Amberley Series in the Walloon Coal Measures. Proven reserves of 0.5Mt of bentonite have been calculated at the currently worked open cut. Production of a sodium bentonite began in 1994 at Idemitsu South Queensland Coal Pty Ltd's Ebenezer mine, 3km southwest of the Jeebropilly mine. There are proven reserves of 1.9Mt of bentonite. The bentonite is extracted from the E6-E7 Ebenezer Sequence of the Walloon Coal Measures. The bentonite is minedfroma layer averaging 0.4m thick. The processing plant has a production capacity of 30,000t/year. This sodium bentonite is traded using a subsidiary company. Clay Resources Company, and it is sold as a stock feed additive, and is used for clarifying wine. At Oakleigh Colliery 3km north of Rosewood, a calcium-magnesium bentonite has been mined from thin seams in the upper parts of the Walloon Coal Measures. No resource estimates have been made for the bentonites at Oakleigh Colliery. At Malabar, 3km north of Oakleigh Colliery, bentonite was mined underground from a layer up to 0.6m thick overlying the New Malabar Seam in the upper part of the Walloon Coal Measures.. The deposit, no longer worked, has been mined intermittently since 1960. The bentonite is predominantly a calcium-magnesium montmorillonite. Conclusions Although there has been extensive e?q)loration for bentonites in the Orallo Formation since 1969, there is still potential for future discoveries. Exploration for commercial bentonite deposits in the Walloon Coal Measures has been minor but the potential for discovering new deposits in this formation is high. The world market for bentonite is considered to be steady. An ejq)anding bentonite market is dependent on economic growth. This growth will result in an increase in the manufacture of iron and steel foundry products, an increase in drilling activity and an expansion of the feed lot industry which are all major users of bentonite. Advances in technology will also provide new uses for bentonite in recent patented applications for use in air filters for septic systems, and as an odour absorbent. These new applications are expected to contribute to an increase in demand. References 1. EXON, N.F., 1976: Geology of the Surat Basin in Queensland. Bureau of Mineral Resources, Geology and Geophysics, Bulletin, 166. 2. WELLS, A.T., & O'BRIEN, RE., 1994: Lithostratigraphicframeworkof the Clarence-Moreton Basin. In, Wells, A.T., & O'Brien, RE. (Compilers & Editors): Geology and Petroleum Potential of the Clarence-Moreton Basin, New South Wales and Queensland. Australian Geological Survey Organisation Bulletin, 241,4-47. 3. FIELDING, C.R., 1993: The Middle Jurassic Walloon Coal Measures in the type area, the Rosewood-Walloon coal field, southeast Queensland. Australian Coal Geology, 9,4-16.

136

INTERPRETATION OF INSITU STRESS MEASUREMENTS FOR BASIN ANALYSIS

LH. ClarkS GX. BoydS J.R Enever^, M.Bockiiig^ and C. Weber^ 1 The Minserve Group, Qld 2 CSIRO-Petroleum Resources, Vic 5 Pacific Power, NSW SUMMARY: Available insitu stress measurement data has been compiled to present a stress map of the Sydney Basin. Palaeodepth and stress analysis of the data has provided insist to the evolutionary structure of the basin and the regional distribution of stress. Both of these aspects are of fundamental interest to developing exploration strategies and for designing mining layouts. INTRODUCTION As part of an overall strategy to develop improved methods for coalbed methane exploration this paper presents a method to better understand the manner in which stress and deformation history have influenced the evolution of the regional geology and hence the extent and distribution of minerals and hydrocarbons. Increased permeability in rock materials (including coal seams) at reservoir scale commonly results from mechanical deformation associated with local tectonic history and uplift [1]. The local tectonic history represents an overprint of the broader geological (burial/upUfl) history and the modem state of stress in the rockmass has its origins with the total stress history of the area. The methodology of basin analysis for exploration or mining involves reconstruction of the major stages of basin evolution with emphasis on defining maximum depth of cover, structural development, deformation history, and determination of peak tectonic stresses during basin evolution. In the course of this study the following objectives were considered: • Development of a fundamental understanding of basin evolution parameters. • Subdivision of the basin into morphotectonic units representing particular stress histories. • Separating the significance of near surface and deep basin structure on local stress fields. • Determination of boundary stress conditions pertinent to any selected area of the basin for subsequent stress modelling. The association between basin evolution and measured stress data was analysed by collating the available insitu measured stress data and analysing for palaeodepth and stress relationships. On meeting the above objectives, the essential information will be available to carry out a simulation of basin scale deformation using numerical modelling techniques. By initialising and applying the correct boundary stresses the model is capable of yielding significant insight into current stress distribution across the basin. In strata representing prospective gas reservoirs, the state of stress can be interrogated to indicate areas of highest rock dilatancy, hence potential gas reservoirs can be targeted [2,3]. For mining applications advance knowledge of the local stress conditions will provide significant input for the design of mining layouts. SYDNEY BASIN STRESS MAP Presentation of Map A stress map has been compiled [4] which summarises a substantial body of data describing the horizontal stress field in Sydney Basin sediments excluding coal. The Sydney Basin Stress Map formed part of a study to develop an understanding of the relationship between Basin geological environments and areas of altered coal permeability conducive to the formation of gas reservoirs. (A copy of this map can be purchased from the principal authors). The measurements were derived from direct stress measurements conducted over the pastfifteenyears using either overcoring techniquesfi-omunderground excavations, or the hydraulicfiracturingtechnique, mainlyfiromsurface holes [5]. The stress data are superimposed on a structural and geological map compiled by E. Scheibner [6], see Figure 1. This map presents a comprehensive synthesis of the major basin structures and lineaments. Included on this map are superimposed earthquake epicentre data and directions of maximum stress directions derived firom earthquake focal mechanisms, [7]. For each location on the Sydney Basin stress map, the resultsfiroma number of point measurements have been combined to give average magnitudes and orientations for the horizontal principal stress components. The stress data are presented as a vector indicating the principal stress direction and a symbol indicating the ratio of horizontal stresses (sg /s^). The magnitudes of stress ratios are indicated on the map Legend. An information box is attached to each vector listing the magnitude of the major horizontal stress (MPa) and the depth (metres) of the stress measurement. The colour of the box refers to the stratigraphy. On the side of the box afilledgrid provides an indicator of the major horizontal to vertical stress ratio. 137

ORIGIN OF STRESSES The insitu stress field can be described in terms of the following components [8]: Gravitational Stress: geostatic loading associated with burial and consolidation of sediments. Tectonic Stress: essentially represents regionally uniform, broad scale tectonically induced stress fields which are due to forces acting on lithospheric plate boundaries. Residual Stress: consists of the stress state which remains in the rock mass even after original causes for the stress have terminated or diminished. Superimposed local effects can modify the regional tectonic and gravitational stress fields; erosion, for instance, can play a significant role in modifying the natural stress field near the earth's surface, probably up to regional scale [9]. If during gravitational loading (say on deep burial) complete lateral restraint (i.e. perfect confinement) can be assumed, then the normal horizontal stress components for an isotropic elastic rock mass can be given by: a^ = CTy =

V (1-v)

[1]

where a^ ~ horizontal stress a^ = vertical stress = yz V = Poisson's ratio y = unit weight of rock column z = depth below surface. It has been shown [8] that for typical values of Poisson's ratio of 0.20 - 0.25, Equation 1 gives a horizontal to vertical stress ratio (K^) of 0.25 - 0.33. However, these values are lower than those found at shallow crustal depths. It is well established [1] that on erosion (uplift), a commensurate drop in horizontal stress does not follow the near linear drop in vertical stress. In assessing the nature of near surface stress fields, local effects such as erosion, topography, rock mass anisotropy's and structural discontinuities, may exert an overriding influence on measured stress magnitudes and orientation. Of these, burial history and uplift have been shown to dominate. The final stress state will be considerably affected by the order in which the various geological processes occurred, i.e. the stress path to which the material has been subjected. To understand the current stress state, and its implications for basin evolution, it is necessary to deconvolute the current total stress in terms of the components (gravity, tectonism and uplift) and stress path. Interpolation of stresses between field measurement locations must consider, in addition to the geological history components, the structural features present, for example, faults and thrusts. ANALYSIS OF STRESS MEASUREMENTS Of importance to the determination of the complete stress history of the basin is the question of maximum depth of burial. The sedimentary record stops at the late Triassic Waianamatta Group, which occurs in three out of the four stratigraphic sub-divisions in the basin. Some authors have advanced various estimates of thicknesses of Jurassic cover over the Sydney Basin. Schmidt et al [10] refer to evidence based on palaeo-magnetism, fission track dating, vitrinite reflectance, and fluid inclusion plus K/Ar datingfiromthe Narrabeen Group that a major thermal event occurred at about 90Ma involving uplift and supra-cmstal cooling from elevated temperatures. Vitrinite reflectance values indicate palaeo-temperatures of 130°C to 180°C which relates to a burial depth of 1,500m to 2,100m in the mid Jurassic. Also based on evidence of vitrinite reflectance of coaly bands in Waianamatta shales, Mallett and Russell [11] suggest up to 1,000m of Jurassic succession over the current Sydney Basin. Timeline Analysis Timeline analysis attempts to resolve the spatial significance of measured stress data in terms of: • stress history related to burial depth; • stress history relative to periods of known tectonism; • stress history related to uplift. These phases of stress change are shown schematically in Figure 2. The vertical stress history is dominated by the cycle of burial then uplift, with stress magnitudes dependent upon the depth of cover. At the local scale vertical stresses may be amplified and may exceed equivalent cover depth when located about certain structures, e.g. blind ended faults. Horizontal stresses increase equally and in constant proportion to the vertical stress during burial. However, during the stage of active tectonism, anisotropy of the horizontal stresses will develop with the major horizontal stress aligned in the direction of tectonism.

138

Timeline analyses are valid when the data can be related to a reference surface (e.g. top or base of geological formation). The full burial history of the basin therefore needs to be resolved. In our analysis, a Jurassic cover depth of 1,000m was added to the present day stratigraphic succession. The Jurassic surface shown in Figure 3 is assumed to be a flat topographic surface. Any major step change in elevation (due to faulting, non-deposition, etc) will rapidly alter the continuity of the separation between reference surfaces. The manner in which consistency of timeline surfaces can be analysed is to consider vertical stress difference between conditions at maximum burial (palaeodepth) and current day depths of cover. Before presenting analysis of the stress data it is useful to present a palaeo-depth analysis. The palaeo-depth function adopted for the analysis was: f(D) = d / D [2] where d = current day depth of each insitu stress measurement D = depth of burial of each location test beneath selected palaeo-surface This formula tests whether the strata in which stress measurements were taken represent laterally continuous and regular surfaces beneath a uniform palaeo-surface. Assuming that all measurement depths represent a consistent stratigraphic position beneath the palaeo-surface, then a consistent trend should be found on a plot of f(D) versus current day measurement depth. Chart 1 on Figure 3. To better discriminate the data, f(D) was modified as follows: f(D) = [D-d] [3] D Chart 2 on Figure 3 shows that for this relationship, three distinct trends in stress measurement palaeo-depths are present. The consistency in trend along each curve suggests that the maximum burial depth corresponding to the top of the Jurassic cover was relatively consistent but varied in at least three areas about the basin. The formula was applied to data from each of the basin regions and plotted. The data corresponds with the basin regions as follows: Curve la: Central region Curve lb: Newcastle, Westem and Southem regions Curve 2: Hunter region Curve 3: Greta Coal Measures region. The separation of data is also associated with definite stratigraphic boundaries, viz: • Central region represents shallow measurements in the Mesozoic succession; • Newcastle, Westem and Southem are all contemporaneous, upper coal measures; • The Hunter region represents the comparative deeper Wittington Coal Measures; • Gretas: stratigraphically deeper coal measures. These results suggest a Jurassic cover of 1,000m thickness over the Waianamatta Group is stratigraphically plausible and consistent. Curve 1 data represents widespread geographic location of data, and discriminates a regular variation in maximum burial depth beneath the Jurassic palaeo-surface. Since Curves 2 and 3 (Chart 2, Figure 3) are similarly uniform they must also describe a different palaeo-depth surface. Different palaeo-depth cover can be accounted for by (i) variation in topography / thickness of the Jurassic cover, or (ii) difference in stratigraphic separation between coal measures. Should the difference in palaeo-depth equal the known separation depth of coal measures, then it can be concluded that the timeline surface (i.e. top of Jurassic) is a flat, near planar surface. The difference in the palaeo-depths represented by the curves can be computed by analysing the change in the value of f(D) to cause curves to merge. It can be shown that about 600 metres of stratigraphic separation is suggested for stress measurements in the Newcastle - Illawara coal measures and the deeper Wittingham coal measures of the Hunter Valley. Reference to the present day stratigraphic column confirms that 600 metres of separation is consistent within the indicated formation thickness of these coal measures. The significance of the above result is that the stress history, as shown in Figure 2, can be quantified, (i) A uniform palaeo-surface can be assumed below which palaeo-stresses due to deep burial alone can be computed. (ii) Maximum horizontal stresses due to deep burial can be computed and compared to present day measured stresses for evidence of tectonic over-print, or stress relief due to uplift. (iii) The discrimination of horizontal stresses into tectonic versus upUft controlled stresses about the basin allows greater insight into the geologic / tectonic evolution of the basin.

139

Horizontal Stress Measurements Palaeodepth Stresses Associated with Deep Burial It is well established in the soil mechanics literature that stresses develop during burial according to defined relations. The vertical stress increases with depth according to the following relationship: where p is the bulk density of the material, g is gravity and h is depth of burial. Horizontal stresses develop in the sediments in constant relation to the vertical stresses as follows: ajj = KQ . ay where the value of K^ is typically 0.3. Analysis of Post Deep Burial Stress History To differentiate between current day horizontal stresses that reflect maximum depth of burial, uplift, or residual tectonic impacts, analyses were again completed using maximum depth of burial beneath a palaeo-surface as the starting point (i.e. scheme of Figure 2). The current day measured stress represents the sum of the residual of past burial stress and past tectonic influences, i.e. ^H = ^ b where As^^ is the residual stress due to uplift, and As^ is the residual stress due to (any) tectonic effect To discriminate between burial or tectonic stress effects on present day measurements, a similar stress difference fiinction was computed as f(S) = (S-SH)/S,

[4]

where S is the maximum horizontal stress due to deep burial alone assuming KQ=0.3, and sjj represents the current day measured stress. Thefimctionf(S) is plotted against current day test depth in Figure 4 for all measure units. Values of f(S) have the following significance: f(S) = 0 Current day stress measurements reflect uplift and tectonic influence on horizontal stress as per the equation sg = Asj^ + As^ f(S) = +ve Occurs for values of sjj > 0; or for sjj to become very small compared to S; this represents real reductions in horizontal stress due to uplift and/or erosion. f(S) = -ve Occurs when current day stresses exceed the equivalent deep burial stress i.e. the stress includes a tectonic component, which increases with increasing negativity. The critical depths below which excess stresses occur for each region are: f(S) Range d" Region d-1 to +1.0 90.00 250.00 Hunter -0.6 to + 0.8 475.00 425.00 Greta -1.8 to+0.8 60.00 200.00 Newcastle -2.4 to+1.0 200.00 Central -3.1 to+0.9 300.00 Southem -1.7 to+0.7 200.00 Westem where d' represents the more strongly indicated depth of partition, and d" represents a less well supported depth of partition. Sites where excess residual stress is likely to be preserved at shallowest depths occur against the Hunter Thrust system; and in Permian Newcastle Coal Measures. It is also noted that Newcastle, Central, Southem and Westem regions all report current day horizontal stresses that are more than twice that which would be expected from deep burial alone. Observations on Stress Orientations Palaeodepth and stress analyses have provided insight to the major evolutionary processes in basin development. However, the following observations relating to orientation of measured stresses also contribute to this understanding: o • The dominant orientation of the major horizontal stress component in Permian horizons is NNE-SSW, occasionally swinging to NE-SW as a local change. • Overlying Triassic sequences carry stress orientations that mimic underlying Permian stress orientations. • Proximity to the northem basin ttost margin is believed to control orientations in the Hunter region. Orientations are generally normal and parallel to the thrust firont. 140

• Stress orientations in the lUawarra Coal Measures in the Southern region are the possible product of the confluence of a number of morphotectonic boundaries. • Stress orientations are radially disposed along the southwestern and western basin margins, in deference to the regional trend, and suggest a sub-regional tectonic influence. INTERPRETATION OF STRESS CONDITIONS IN BASIN UNITS Following Scheibner's morphotectonic boundaries, there are interesting but no conclusive associations to be drawn. Observations drawn from the critical depth data of relevance to specifying stress conditions in each of the morphotectonic units are summarised as: Hunter Valley Thrust and Fold Belt: Residual tectonic stresses expected at all depths in Permian coal measures; significant impact of local structure associated with Hunter Thrust system. Results fall into two sub-areas, namely: Muswellbrook Area: measurements radially orientate to Hunter Thmst trend and thus reflect remnant tectonic stresses aligned parallel to direction of tectonic transport. Singleton Area: Major horizontal stress parallels Hunter Thrust trend suggesting stress reUef has occurred in direction of tectonic transport probably due to local stmctures. It is concluded that for this unit, stress magnitudes and orientations will be controlled by proximity to the thrust system with a high chance of modification due to relief along local structure. Lochinvar-Kulnura Ridge: Excess residual stress occurs in Permian horizons at depths >400m in the north and at depths greater than 200m in south. It is concluded that Triassic horizons are stress relieved due to uplift and/or shallow depths. Lake Macquarie Trough: All Permian horizons deeper than 200m carry excess horizontal stresses consistently aligned NNESSW. All Triassic measurements mimic Permian stress orientations and carry no excess residual stresses. It is concluded that horizontal stresses have adjusted due to uplift. Woronora Ridge (Southern Lake Macquarie Trough): Limited and contrasting information was available for Permian horizons. It is possible that proximity of sites to the niawarra scarp may account for indicated stress relief. Western MacDonald Trough: Excess residual stress is carried in Permian coal measure beyond 250m of depth; relieved stress is carried in Triassic horizons. The most notable pattern of stress measurements in the basin lies along the southwestem margin. The general alignment of major horizontal stress direction is at right angles to the basin boundary. This orientation suggests possible ongoing expansion of the basin to the southwest along a buried detachment surface. This notion is anecdotally supported by: • the intensity of earthquake epicentres about the southwestem and mid-westem basin margins; • the projection of an interpreted transcurrent fault system trending WSW into the area; • co-orientation of focal solutions to earthquake events. A suspected buried feature representing the southem extension of the Meandarra Rifl Zone, Tadros [10] in the Gunnedah Basin, is thought to be associated with stress direction rotations, and epicentre distribution along the western flank of the basin. As plotted in Figure 1 the lineament swings to align with a WSW transcurrent fault system to define a "nose" that represents the confluence of the: (i) (ii) (iii)

southem extension of the Lochinvar-Kulnura rise; convergent west basin margin; convergent Lake Macquarie Trough.

The trend of the buried lineament along the western margin and around the "nose" also bears a great similarity to the change in trend of the Hunter Thrust system about Muswellbrook. From all the above associations, it would appear this west basin lineament (un-named) may in fact represent activity along a developing detachment surface associated with ongoing thrusting to the SSW of the New England Fold Belt. The only continental e^qjression for ongoing tectonism is the suspected buried feature recognised by Scheibner located along the west margin of the basin. It is therefore concluded that the stress regime along the western and southem westem flanks of the basin may vary from usual basin trends due to ongoing deep cmstal processes that find current day surface expression in the Hunter MookiThrust system. COMPUTATIONAL STRESS MODELLING In order to demonstrate the dependence of regional stress on the geological history (stress path) and the role of major stmctures to alter the stress field locally, a simple numerical model was set up. The analysis simulated cmstal shortening on a rigid cmstal block broken up by a series of linear, pervasive strike-slip features generally mimicking the style of those postulated by Scheibner. As a starting point, a gravity load was generated to simulate the burial process. The 141

western boundary of the model block was rigidly clamped to simulate the Lachlan Fold Belt. Active loading was then applied in a N-S, followed by an E-W orientation to reflect the commonly understood tectonic history of the basin, allowing sliding along intemal sub-block boundaries. The final stress state is shown in Figure 5. There is clearly a general correspondence in the stress orientations with those measured insitu throu^out the basin. The following observations are made: • a general swing from E-W toward N-E in the southem collieries district consistent with available focal plane solutions. • a pronounced N-S orientation along the eastem boundary consistent with measurements parallel to the coast through the central region. • north of the Hunter Thrust, the projected low stress magnitude is consistent with the stress field measured in the Gloucester Basin. • a pervasive ENE stress orientation through the central portion reflects both the insitu stress measurements and focal plane solutions for the Newcastle and Cessnock earthquakes. The model also suggests complex block movements in the south-west comer consistent with the concentration of seismic activity in this region. Substantial shear deformation is shown along the Hunter Thrust. CONCLUSIONS As part of an overall strategy to develop improved methods for coalbed methane coloration this paper presents a method to better understand the manner in which stress and deformation history have influenced the evolution of the regional geology and hence the extent and distribution of potential mineral or hydrocarbon accumulations. A study of limited scope into the significance of measured basin stresses has provided insight to the evolutionary structure of the basin and the regional distribution of stress. Both of these aspects are of fundamental interest to developing exploration strategies and for designing mining layouts. It is inferrred that regional tectonic elements of major influence on basin evolution relate to the thrusted New England Fold Belt to the north, NE trending transcurrent faults, and a possible buried detachment surface active along the southwestern basin margin. Structural influences were not assessed specifically, but there is evidence that near surface structures (<200m) along northem basin margins result in significant variation in stress magnitude and orientation but in a manner probably reconcilable to the local structure detail. We are confident that stress conditions in any one morphotectonic region, or across structural zones in any one region, can be determined for accurate, area-specific simulation of coal or rockmass deformation. In addition the analyses have i) highlighted regional differences in the setting of basin stresses; ii) indicated depths in each region at which uphft, rather than tectonic effects dominate in the measured stress values, and iii) provided a basis for defining robust boundary conditions stress modelling applications. Knowledge of basin stress conditions is fundamental to exploration and reserve targeting particularly in the case of gas reservoirs where the stress history is considered a controlling factor in developing reservoir properties. In strata representing prospective gas reservoirs, the state of stress can be interrogated to indicate areas of highest rock dilatancy, hence potential gas reservoirs can be targeted. For mining applications advance knowledge of the local stress conditions will provide significant input for the design of mining layouts. REFERENCES 1. Clark, I.H. and Boyd, G.L., 1995. Geological controls on coalbed methane accumulations. Symposium of Management and Control of High Gas Emissions and Outbursts in Underground Coal Mines, Wollongong. pp 369-374. 2. Enever, JR, CJ Pattison, RH McWatter and IH Clark, 1994. The relationship between in-situ stress and reservoir permeabiUty as a component in developing an exploration strategy for coalbed methane in Austraha. SPE/ISRM Eurock '94: pp 163-171, Balkema, Rotterdam. 3. Enever, JR Bocking, MA and IH Clark, 1994. The appUcation of insitu stress measurement and numerical stress analysis to coalbed methane exploration in Australia. SPE Asia Pacific Oil and Gas Conference, Melboume, Paper SPE 28780. 4. SYDNEY BASIN STRESS MAPiCompiled by CSIRO-Petroleum Resources, Pacific Power, SCT and The Minserve Group. Availablefi-omDr IH Clark, The Minserve Group, Suite 3 Level 1,1 Swann Road, Taringa Qld 4068. Tel: (07) 3870-1946 Fax: (07) 3371 4962 5. Enever, J.R. and Walton, R.J. 1995. Assessment of Ground Stresses. Geomechanical Criteria for Underground Coal Mines Design, (Ed. D. Krzyston) Intl. Bureau of Strata Mechanics, Central Mining Inst. Poland, pp 7-34. 6. Scheibner, E. 1993. Structural Framework of NSW. Qtly Notes, Geological Survey ofNSW, Oct. 142

7. 8. 9. 10. 11. 12.

Shepherd, J. and Huntington, J., 1981. Geological fracture mapping in coalfields and the stress fields of the Sydney Basin. 1 Geol Soc. Aust, 28, 299-310 Brown, E.T. and C.R.Windsor, 1990. Near Surface Insitu Stresses in AustraUa and their Influence on Underground Construction, Inst Eng Australia, Tunnelling Conference, Sydney pp.18-48 Hoek, E. and E.T.Brown, 1980. Underground Excavations in Rock, Inst Min Met, London. Schmidt, PW MA Lackie and JC Anderson, 1995. Palaeomagnetic evidence for the age of the Lapstone Monocline, NSW. Aust Coal Geology, VIO, pp 13-22 Mallett, C.W. and N. Russell, 1990. Thermal maturation modelling in the Rangal coal Measures - Dawson River area. CSIRO Div Geomechanics Internal Report No 55. Tadros, V. 1993. Gunnedah Basin, Memoir Geology 12 Geological Survey, NSW.

Figure 1. Geological structure map of the Sydney Basin, NSW (after Scheibner, 1993) 143

Vertical Stress

Burial

Shortening (Tectonlsffi) ^ .

Uplift

Extension

Major Horizontal Stress

Q.cc (0 o

Minor Horizontal Stress

Cf)

If - JJ

-f

Trace 1

. — J

Time History Figure 2.

Stress changes associated with Basin Evolution

D.f(er,nc» (D-(J/D)

FOR

144

GEOLOGICAL MODEL PALAEODEPTH ANALYSIS

0 A

©c

• A.

-200

•

•J

-400 Q. 0) Q

-600

• •

iiil >

•

o >Oo

)

•

r \

^oO

• lii iiiilO: liiliii iiiiiiii

•

iL: •

' A

-800

-1000 -2

-1

Horizontal Stress Difference (S-S^/S) o SOUTHERN

• WESTERN

0 HUNTER D' • HUNTER D"

® CENTRAL

•

A NEWCASTLE D' A NEWCASTLE D"

GRETAS

Figure 4. Palaeodepth Analysis of stress data from the Sydney Basin.

145

Figure 5. Simulated stress distribution associated with major structures and geological history in Sydney Basin

146

INDUSTRIAL MINERAL OPPORTUNITIES IN QUEENSLAND. Wilson Cooper^ and Grahame L. Baker 1 Director, Queensland Department of Tourism, Small Business and Industry Wilson Cooper, Queensland Department of Tourism, Small Business and Industry Summary Some of the enduring developments in industrial minerals in Queensland are from materials from the Mesozoic formations, and emerging opportunities have material sourced from this same era. The clay industry in its many facets is largely based on kaolin , bentonite and structural clay from the Triassic Ipswich Coal Measures to the Jurassic to Cretaceous Orallo Formation. Magnetite from the Mount Biggenden mine was emplaced by the intrusion of the Triassic Degilbo Granite into the Permian Biggenden Beds. And opal is hosted by the Cretaceous Winton Formation from the NSW border to north of Winton. Emerging opportunities are in areas such as mineral exploration and developments for clays, gypsum, ilmenite, magnetite and lump silica. However, major opportunities arise in processing for the manufacturing sector. The tme value of industrial minerals to the State's economy is in their essential contribution to current and potential manufacturing applications, and their worth is now being accorded recognition with the establishment and implementation of the Industrial Minerals Strategy for Queensland by the Department of Tourism, Small Business, and Industry. Introduction Queensland has extensive energy and metallic mineral resources of coal, oil shale, petroleum and gas, and base and precious metals. Mining and mineral processing of these resources generates a major component of the State's economic wealth. However, well developed and established operations of bauxite, heavy mineral sands, siUca sands, limestone and dolomite, the traditional clay-based ceramics, dimension stone and gemstones, newer developments of magnesite and kaolin, and proposals for phosphates are ensuring that industrial minerals are increasingly contributing to the State's well being. The State's abundant supply and quality of mineral resources are complemented by an infrastructure network which includes a transport system increasingly integrated with fully developed deep water ports, a modem, efficient communications system, as well as energy supplies and distribution networks for power and gas currently in place or planned for theftiture.When these factors are linked with a skilled workforce, advanced research and technology facilities, and political and economic stability, Queensland presents as an encouraging environment for investment and development. The mining industry in Queensland generates $5.43 billion in mineral and energy production in 1994-95, accounting for some 54% of the State's e^qjort income (Queensland Department of Minerals and Energy, 1995). The value of industrial minerals production was only about 5% of all minerals. However, this reflects only the value of ore at mine site. Mineral preparation and end-use applications are well recognised as adding value to the mineral, but this is not recorded in production figures. The true value of industrial minerals to the State's economy is the wealth created for the community through current and potential manufacturing appHcations, with consequent industrial development. This is now being accorded recognition with the establishment of the Industrial Minerals Strategy for Queensland by the Department of Tourism, Small Business and Industry. That strategy recognises and identifies the opportunities in Queensland in the industrial mdnerals field . These opportunities can be realised in mineral exploration and development, minerals processing, manufacturing using intermediates or processed minerals, and the maximising of opportunities for both investment in Queensland and for export. However, the success in capitalising on those opportunities depends on a knowledge and utilisation of the resource base of industrial minerals in the State. Raw materials Queensland has an inventory of industrial minerals, shown in the accompanying table, which is composed of over 50 commodities, ranging from abrasives to zircon. Some of Queensland's major developments of industrial minerals have been from Mesozoic formations. The most important of these are clays (including the structural and industrial clays), magnetite, dimension stone, opal and gypsum, and potential developments in ilmenite, lump and microspherical silica, and feldspar. Clays Clays are the basis of a range of applications, which includes ceramics, absorption, coating and filling. The clays include structural clays for brick, paver, pipe, tile and pottery manufacture, industrial clays which include kaolin, bentonite, raw materials suitable for whiteware, refractory flint clay and fireclay, and the plastic, kaolin-rich ball clay. 147

Structural Clays Structural clay deposits were described by Cooper and Carmichael (1992). In Queensland, most ceramic industries are established in the southeast in the Brisbane-Ipswich-Toowoomba area. There is additionally a substantial brick, paver and tile industry in the Bundaberg-Maryborough area, and small works service the major centres at Yeppoon, Mackay and Townsville along the coast of Queensland. Sources The most important clay resources in southeast Queensland occur as Mesozoic aged, pale buming shales and carbonaceous shales from coal measures and associated formations around Ipswich, Brisbane and the Toowoomba-Warwick area. Clay is also present in these rock types as deeply weathered residual clays, such as the kaoUn-rich lower paUid zones of laterite profiles. Production in this area is dominated by four main manufacturers, PGH, Boral, Austral Bricks and Nubrik, with plants in Brisbane (PGH at Strathpine, Oxley; Boral at Darra; Austral Bricks at Rochedale; Nubrik at Oxley), Cooroy north of Brisbane (PGH), Ipswich (Nubrik at Riverview), and Toowoomba (PGH). Smaller producers are at Ipswich (Claypave at Dinmore, Ebbwvale), Toowoomba (Qayware at Kleinton) and Warwick (Warwick Brickworks). Shales from the Triassic Ipswich Coal Measures and overlying Woogaroo Sub-Group are the main pale buming sources for PGH, Boral, and Nubrik from the Dinmore-Swanbank area and Rochedale area for Austral Bricks. Landsborough Sandstone shales are used by PGH from Narangba, and Jurassic Marburg Formation shales are used by the Toowoomba and Warwick works. In the Bundaberg-Maryborough area, major clay sources servicing the industry are derived from carbonaceous shales and siltstones of the Cretaceous Burrum Coal Measures and Maryborough Formation. Production occurs mainly at Bundaberg (Wide Bay Brickworks Ltd) for bricks and pavers, and Maryborough (Marcotta Tiles) for floor tiles. White buming material for the Bundaberg works is derived from the kaolinitic pallid zone of the lateritised binary Broomfield Granite of Middle Triassic age. Production in the smaller works at Yeppoon east of Rockhampton (PGH), Mackay and Townsville (Nubrik) is not dependent on Mesozoic aged clay materials. The small but strong pottery manufacturing sector is located mainly in southeast Queensland, with the largest production of mostly nurseryware from JT Sandison and Co, Wulkuraka, west of Ipswich. Mesozoic white buming shales of the Ipswich Coal Measures and Marburg Formation are important components in nurseryware and studio clay production. Industrial Clays These clays include ball clays, refractory fire and flint clays, kaolins and bentonites from Mesozoic formations and have a range of appUcations. Kaolin, refractory clays High kaolin low plasticity clays of the Ipswich Coal Measures shales were used for the original sanitaryware, porcelain insulator and pottery industries in the Ipswich area. Refractory flint clay andfireclay deposits were used by the refractory industry in the Dinmore area, where fireware is still produced by Claypave. Kaolin clay from Ravensboume, Kingaroy and Crows Nest in southern Queensland is also used in blends for studio clay and pottery. These deposits are derived fromkaolinised granites and kaolin-rich sedimentsfromthe Late Permian to Early Triassic Boondoomba Igneous Complex and Triassic Tarong Beds. The Ravensboume workings for kaolin are sourced from tailings from a silica sand working in the Late Triassic to Early Jurassic Helidon Sandstone. At Pierces Creek near Crows Nest, a primary deposit of kaohnized Late Permian to Early Triassic Crows Nest Granite has been used for kaolin and sihca for porcelain manufacture (Sawers & Cooper, 1985). Ball clay Ball and semi-ball clay have been recorded in Queensland in the Ipswich area associated with the Ipswich Coal Measures shales, but are uncommon in the Mesozoic formations. Bentonite Bentonite is mined in Queensland for a range of uses, mainly for absorbent properties and ceramics. Bentonite occurrences are formed by in situ alteration of minute glass particles in volcanic ash or tuff deposited in sedimentary sequences. Significant deposits in Queensland occur as thick, near flat lying beds within the Late Jurassic Orallo Formation in the Miles district, and in persistent seams in the Middle Jurassic Walloon Coal Measures around Rosewood west of Ipswich from the Oakleigii, Ebenezer and Jeebropilly coal mines (Carmichael, 1995). Queensland's production of bentonite is mostlyfromthe Miles and Rosewood areas. Recent developments in the Ipswich area could see significant amounts incorporated in ceramic bodies. Silica Silica as microspheres and in lump form is derived from Mesozoic formations. 148

Lump silica At Bajool south of Rockhampton, high purity quartz deposits exist in pipe form as a late stage of an unnamed granitic intrusion of late Permian to early Triassic age. These deposits offer excellent potential for processing to fused silica, and in servicing the specialised electronics markets in Asia. Microspheres Microcell Australia Pty Ltd processes white hollow silica microspheres as functional fillers in a broad range of applications. The material is sourced from fly ash produced from the Late Triassic Tarong Beds by the Tarong Power Station and processed in Nanango. Some apphcations include surface coatings, sealants, caulking, resin extender in plastics, and in refractories (Cooper and others, 1996). Feldspars and felspathic raw materials Currently, there is no Queensland supply of feldspar and felspathic raw materials, important components in glass manufacture and ceramics. Various potential feldspar resources in Queensland are currently being assessed for development, including a deposit near Duaringa west of Kingaroy from the Permian to Middle Triassic Boondoomba Igneous Complex, which in earlier times was a source of feldspar for ceramic flux (Sawers & Cooper, 1985). Opportunities could be available for a ceramic feldspar source. Gypsum Deposits have recently been worked in Cretaceous and overlying Tertiary sediments south of Winton (Cork, Mayneside) and at Hughenden. Almost 25,000 t were produced in 1994-95. Deposits of gypsum are also known at Richmond from the Early Cretaceous Toolebuc Formation, and from Lower Cretaceous marine clayey sediments of the Eromanga and Surat Basins, in the Boulia, Eulo and Roma areas. Umenite New developments onshore include the Goondicum ihnenite project west of Bundaberg, where Monto Minerals NL has identified a large ilmenite resource within the alluvium of the upper Burnett River. The resource, which is derived from the weathering of titaniferous magnetite-rich Cretaceous (Cec Murray, personal communication) Goondicum Gabbro is considered to be potentially significant in size, grade and quality. Monto Minerals is also undertaking studies of the resource for its processing potential at Gladstone for synthetic mtile and pigment production. Dimension Stone The main dimension stones obtained in Queensland of Mesozoic age are sandstone and granite. Sandstone Sandstone from the Helidon district west of Brisbane is the most important dimension stone in Queensland. Helidon Sandstone has been used in the building industry since the late 1800s. The sandstone is of fluvial origin and of Late Triassic to Early Jurassic age. The major producers include Australian Sandstone Industries, LH. Wagner & Sons and Comerford Sandstone. Over 27,0001 of sandstone was produced in 1994-95. Granite In the Crows Nest area, batholiths of Permian to Triassic age provide the coarse grained 'Austin Red'. Historically, the Greymare Granodiorite of Early to Middle Triassic agefromthe Warwick area and Enoggera Granite in Brisbane were used as dimension stone in early Brisbane architecture. Opal Australia produces over 90% of the world's precious opal. Three major types of precious opal are available - black opal from Lightning Ridge in New South Wales, white opalfromSouth Australia, and boulder opalfromQueensland. Boulder opal is found within ironstone boulders, filling concentric, radial, or random cracks, or as a kernel in the chemically altered Cretaceous sediments of the Winton Formation of the Eromanga Basin in westem Queensland. Opal is produced from numerous small mining operations. Queensland production in 1994-95 was estimated to be worth $1.5 million. Extensions to formations containing the Lightning Ridge type of black opal occur in southem Queensland. Redfire Resources successfully conducted innovative exploration in the Cretaceous Griman Creek Formation in the Hebel Dirranbandi area. Magnetite Magnetite in Queensland is used as a heavy medium for stone separation in coal washing plants. The Mount Biggenden mine of Commercial Minerals Limited accounted for all production for 1994-95 of 24,500 t. High grade magnetite ores occur as skam mineralisation with other economic minerals such as gold and bismuth at the Mount Biggenden mine. The magnetite was emplaced by the intrusion of the Triassic Degilbo Granite into the Permian 149

Biggenden Beds. Similar skam deposits are known throughout the state. Magmatically derived magnetite and titaniferous magnetite segregation deposits are known from the Late Permian to Early Triassic Hawkwood Gabbro and in the layered Eulogie Park Gabbro of central Queensland. Exploration activities are currently seeking new resources of high quality magnetite to supply local and overseas markets. Future prospects include the production of magnetite as a by-product of poly-metallic mines in the Mount Isa region. Opportunities The Queensland Government has instigated a policy of encouraging companies to upgrade raw materials to higher value added precursors and products. This is based on the State's own sources of good quality raw materials and intemationally competitively priced energy sources of electrical power and gas, essential for successful processing. It is highly supportive of investments in mining, processing and refining of industrial minerals, especially those mineral products that have an export component. There is a significant industrial base in Queensland which includes world scale mining, processing and smelting operations in coal, base and precious metals, a comparatively modest but strategic industrial minerals mining, processing and manufacturing sector. Many business opportunities exist in the industrial minerals field. These include:• mineral exploration and development, involving search for new resources, or extensions of currently known deposits, the development of those deposits, and the supply of minerals with some preliminary processing; • processing of raw minerals, forming and processing intermediates; • manufacturing products using intermediates or processed minerals; • research, development and education; • maximising opportunities for investment; and • maximising export opportunities of raw and processed minerals, and manufactured goods. These opportunities are based on the mineral prospectivity, availability of quality raw materials, competitive economics of location of processing plants, the continued growth of manufacturing, world class technical infrastructure to support sophisticated process industries, and the closeness to the fastest growing region in the world. Some of those opportunities are detailed below. Mineral exploration, development, processing Geological prospectivity of Queensland for industrial minerals has been recognised by mineral exploration companies, and significant exploration is continuing for a number of industrial minerals. Expenditure in Queensland exploration is spread across a range of commodities, including kaolin, ilmenite, magnetite, silica sand, lump silica, soda ash, wollastonite and zeolite. Extensions of currently worked deposits are expected in bauxite. Some examples of exploration and development projects are as follows. Those based on materials from Mesozoic formations are:— • Clay sources for ceramics in the Ipswich-Rosewood area • Iknenite at Goondicum; • Kaolin resource extensions around Ravensboume and Kingaroy; • Lump silica investigations at Bajool; • Soda ash (trona) and nahcolite from the Denison Trough; • Magnetite for coal washing in central Queensland. As well, some other investigations not in Mesozoic formations are:• Skardon River kaolin deposit. A pilot plant for this project is based in Caims, which is also available for toll processing and testing; • Magnesite deposits at Yaamba adjacent to the Kunwarara deposits; • Inland sources for zircon placers; • Wollastonite deposits identified west and southwest of Caims; • Zeolite deposits west of Emerald. Major developments in mining and processing occur in a number of projects. • Comalco's proposed second alumina refinery of 5.0 Mtpa capacity, doubling production of Weipa bauxite; • Queensland Metals Corporation (QMC) mines and beneficiates magnesite from Kunwarara, and processes further to caustic calcined, deadbumed and electrofused magnesia in Rockhampton; • Western Mining Corporation Ltd (WMC) proposed high analysis fertiliser project based on northwest Queensland rock phosphate resources and sulphuric acid utilising smelter gases at Mount Isa. Mining and processing of silica sand on North Stradbroke Island and Cape Flattery are major operations. 150

• Cape Flattery Silica Mines Pty Ltd mines and processes silica sand for export for the glass, foundry, and chemical industries (Cooper & Sawers, 1990). • ACI Resources works silica sand on North Stradbroke Island for glass and foundry use (Cooper, 1993). Processing raw minerals, intermediates There is a range of processing opportunities, some current, as well as some potential ventures. Those which depend on materials from Mesozoic aged sources are described. Consumable Supply: The ceramics industry in Australia is supported by suppHers of consumables such as frit, colours and glazes, and clay bodies. Investment opportunities exist for the manufacture of>• specialty ceramic slip and ceramic powders, using high quality silica and kaohn resources for the manufacture of frit and glazes; • Clay bodies for domestic use and export, prepared specifically for manufacturers. Currently, clay bodies for use in studio throwing are prepared in Queensland. Similarly, clay bodies for pavers, tiles and pottery, and specialist industrial clays based on kaolin, bentonite, and refractory clays could be prepared here. The pigment industry is based on titanium white from ihnenite supplies. • Synthetic rutile from Goondicum ilmenite could be a source material for a pigment plant Refractory raw materials, with some potential ventures in:• chamotte from calcining prepared kaolin, for both the local and export markets; • fused silica, from lump and sand silica sources using traditional plasma technology. For non-Mesozoic materials, other important opportunities are:Refractory raw materials:— • mullite from topaz, corundum: a proposal to process a topaz resource from Torrington in New South Wales at the Waraluck plant at Kilkivan is under investigation; • spinel from magnesia and alumina, for use in the proposed magnesium metal smelter; • silicon carbide, produced by electroflising silica with high quality carbon; • high alumina cements, prepared using high quality limestone, kaolin, calcined bauxite and alumina; • glass cullet manufacture from silica, limestone, soda ash. Magnesite — caustic caldned, deadbumed and electrofused magnesia:QMC processes magnesia for refractory intermediates, using caustic calcined, deadbumed and electrofused magnesia products for ceramic intermediates at the QMAG plant in Rockhampton. Zircon- refractories and glazes. Zircon and zirconia products for ceramics and refractories, with an increasing percentage being consumed in partially stabilised zirconia. Zircon is micronised to zircon flour in Consolidated Rutile Limited's Brisbane plant for ceramic uses. Most is consumed by the joint venture partner in ceramics, SEPR Australia, who cast monolithic zircon-alumina ceramics (ZAC) shapes for the refractory and abrasive amelioration industries. Manufacturing products Opportunities for manufacturing are concentrated in the ceramics market for traditional clays, for refractories, glass and in advanced materials. As well, a major development will be in phosphate manufacture and chemical derivatives from natural gas and smelter-derived sulphuric acid in northwest Queensland. Traditional structural clay products such as ceramic tiles, bricks and pavers, sanitaryware, tableware, pottery and artware are the most common ceramic production. • Ceramic Tiles: Three large manufacturers supply about 20% of the local market. The rest is imported, indicating there are still opportunities for investment in Queensland. There is proposed a terracotta tile plant based on material from the Walloon Coal Measure derived from the Jeebropilly coal mine; • Sanitary ware: Caroma Industries Ltd and Fowlerware in Sydney and Melbourne satisfy local demand, but there is a capabihty for export based on lowering production costs through automation and upgrading of casting, drying and firing technology; • Tableware: There is room in the market for specialist high quality products. Australian Fine China is the only major manufacturer of ceramic dinnerware in Austraha, mainly for the institutional market;

151

• Pottery and Artware: There are numerous medium to small scale manufacturers producing functional and decorative pottery, and industrial ceramics for the local market. Most generally fill niche markets although some are developing export markets for their products. • Export brick and paver products: Wide Bay Brickworks Pty Ltd from Bundaberg and Claypave in Brisbane have developed export markets. Wide Bay Brickworks utilise a state-of-the-art roller kiln to produce newly designed, h i ^ quality cyclone standard products targeted for the Asia-Pacific region. Smelting industry refractories are manufacturedfromtraditional clay-based and advanced materials. The Australian refractory industry is geared to serve mainly the domestic miarket in the mineral processing, ferrous and non-ferrous metal, building products md power generation industries. Bricks account for 90% and castables 10% of the market. The Queensland industry consists of production of refractory bricks and shapesfromfireclayshales by Claypave in Dinmore, and monolithic castables by SEPR Australia in Brisbane. The smelting industry is open to new refractory products. The local aluminium smelting industry is largely supplied from Europe and North America, with high performance bricks for the aluminium smelters and the more speciahsed refractory shapes and artefacts. New recent production facilities for specialist product manufacture should reverse this trend, such as SEPR. Wear and abrasive resistant ceramics for linings in mining, manufacturing industries. • SEPR Australia has developed ZAC wear resistant ceramics for the fast growing mining industry market in Australia and South East Asia, and in processing plant, shiploaders and power stations. Technical ceramics for components in manufactured goods, with possibilities in equipment manufacture. • Furnaces, kilns, materials handling equipment, andfiringtechnology. Australia's brick industry exports equipment and knowledge. Brick plants in Malaysia, China, Sri Lanka, Indonesia, New Zealand and Russia have been built or modemised with Australian technical assistance and equipment. Glass industry developments infloatand laminated glass specialties include :• coated reflective float glass for the commercial building sector, mirrors, toughened and laminated glass, drinking glasses, table and kitchenware, and glass fibres. Fertiliser industry. A high analysis fertiliser project based on the manufacture of mono-ammonium phosphate (MAP) and di-ammonium phosphate (DAP) fertilisersfromnorthwest Queensland resources is undergoing feasibility studies by Western Mining Corporation Ltd. The project involves reopening the Phosphate Hill mine, constructing a sulphuric acid plant based on utilising smelter gases at Mount Isa and a phosphoric acid plant, ammonia plant and fertiliser plants near Mount Isa. Research, Development and Education Major research and development initiatives in Queensland are invested in the Queensland Centre for Advanced Technologies and the Queensland Manufacturing Institute. As well, materials research is carried out at the QUT with clay chemistry, sol-gel, corrosion and polymer science, and at the University of Queensland by Advanced Ceramics Development - UniQuest Ltd, investigating advanced ceramics technologies through processing, catalysts, materials characterisation, aimed at materials for environmental, agricultural, mining and manufacturing industries (Cooper and others, 1996 b). However, further input is envisaged with establishing a CSIRO Division of Materials Science and Technology Networking OflSce in Brisbane; Tertiary education in industrial minerals and processing is invested in QUT, and the U of Q as various materials related courses. Part of the Queensland Government Industrial Minerals Strategy is to promote and encourage more ceramic education, including TAFE level courses allocated to focus on hobby and studio ceramics and how ceramic operator and technician training could be implemented in Queensland. Investment and export Opportunities The Queensland Government is currently encouraging investment in industrial minerals processing industries by actively seeking potential linkages with overseas companies, with strategies of bilateral joint ventures in countries where there is synergy of creative inputs in raw material, technologies and mmufactured products. Such opportunities have already been discussed at intemational forums on investment in the region. Export opportunities can be the result of the investment scenarios just presented, including equity participants bringing a market to a plant. There are substantial export opportunities for minerals commodities in raw, processed, and intermediate form for the growing Asian industries, ranging from bulk loaded material down to small lot packaged and processed minerals. 152

Queensland Advantages There are considerable competitive advantages in the industrial minerals area for Queensland. Most industrial minerals are either already available or prospective in the State. The state is geographically located very close to the rapidly growing Asian markets, and allfliesecountries have increasing demand for most industrial minerals to support their rapid industrial development. Queensland is a major exporter of kaolin, sihca, magnesite products, bentonite, heavy mineral sands and limestone, and industrial minerals product development in. such industries as mining and smelting can provide substantial opportunities for export and investment. The integrated transport system of road, rail and air is linked to deep water ports. Competitively priced energy for processing is available as electricity, coal and natural gas. Communication networks are well established, extensive, sophisticated but rehable. Land has been allocated for major industrial processing with appropriate zoning in most centres around the State. The Department of Tourism, Small Business and Industry can assist companies looking for industrial land, and facilitate their entry into markets. Conclusions Queensland has a substantial base in industrial minerals of Mesozoic age. There are primarily the clays for structural purposes and industrial uses, including kaolin, bentonite, and kaolin-rich refractory clays, with sandstone and granite dimension stone, gypsum, ilmenite, magnetite, silica and gemstone opal. However, there are opportunities for investment and export in mineral coloration and development, processing and manufacturing, as well as in research, development, and education. Acknowledgments The authors gratefully acknowledge industry's contribution for information on their operations, and wish to thank the Queensland Department of Mines and Energy for providing commodity and cartographic input with the map of Queensland's industrial minerals. References CARMICHAEL, D.C., 1995: Queensland Mineral Commodity Report - Bentonite. Queensland Government Mining Journal 1128, 14-24. COOPER, W., 1993: Queensland Mineral Commodity Report - Silica sand. Queensland Government Mining Journal, 1102, 7-15. COOPER, W. & CARMICHAEL, D.C., 1992: Ceramic resources in Queensland. Queensland Government Mining Journal 1090, 29-34. COOPER, W. & SAWERS, J.D., 1990: Silica sand. Cape Flattery and Shelbume Bay sihca sand deposits. In Hu^es, EE. (Editor): Geology of the mineral deposits of Australia and Papua New Guinea, Australasian Institute of Mining and Metallurgy, Melbourne, Monograph 14, 1665-1667. COOPER, W., CULLEN, J.E & BAKER, G.L., 1996 a: Queensland's industrial minerals - on top down under. Industrial Minerals, April, 343, 85-107. COOPER, W., CULLEN, J.E & BAKER, G.L., 1996 b: Queensland - Australia's gateway to the Asian ceramic industry. In Press (presented at PacRim2, the 2nd International Meeting of Pacific Rim Ceramic Societies, 15-17 July 1996 Cairns, Australia). QUEENSLAND DEPARTMENT OF MINERALS AND ENERGY, 1995: Queensland Minerals and Energy Review, Queensland Department of Minerals And Energy. SAWERS, J.D. & COOPER, W., 1985: Some Queensland industrial materials. Queensland Government Mining Journal 86, 188-195.

153

INDUSTMiO. MMERAVLS ^

Vrilya Point

^^ ^^^

Major Mining Projects and Resources

Skardon River Sheibume Bay Si ANDOOM Bx K& WEIPA Bx Ka

COMMODITIES

GEMSTONES

MINE

Resource

Potential for Opal

Potential for Bauxite/Kaolin Bentonite Bauxite a Brick clay Dt Diatomite Do Dolomite Fl Fiuorite Gr Granite Gp Gypsum Ka Kaolin Ls Limestone Mg Magnesite Mt Magnetite Ma Marble Pe Perllte Ph Phosphate Na Salt Ss Sandstone SI Silica St Slate Wo Wollastonite Ze Zeolite

Bn Bx

CAPE FLATTERY Si I Melody Roclcs River of Goid St NYCHUM Pem Mungana Worn CHILLAGOE Ma Ls FT^y

Cairns

Karumba

les Fl

MOUNT SURPRISE Do«

Ch Op Sa

MINERAL SANDS : n MS

« LAVA PLAINS Sa ^ Conjuboy Dt ts> ^ ^ Christrria^Creek Ls Broken River Lsm HILLGROVE Dom ^ I CALCIUMS Fanning River Ls^ BOWEN Na -

: Undllla Ls D Tme PA^ e u d y Annie Ph Lily CtwkJ^f^ y SPRING CREEK Si « m \ Cloncuny — UNDINA Ls MOUNT ISA m Gr GIDYA Qrm

MINE Resource nmenite Mineral sands

Export port Population centre Railway Mesozoic Basins

# Richmond Gp ^

Chrysoprase Opal Sapphire

Tasman Orogenic Zone

*HUGHENDEN Gp

Georgina Basin

Ardmore Ph^ / ) P l W h a t e HIII P/J^

Proterozoic Shield .Winton .Dysart

i^Boulia Gp

Herbert Creek Mg^ MARLBOROUGH Ohm % Yaamba Mg-' m

•CORK Gp

OPALT9N^ MAYNESIDE Op

Willows Z©^

KUNWARARA Mg Shoalwater Bay MS Byfield MS Si INT ETNA Ls

SOUTH ULAM Ma Ls„ ladstoneWitaHib' BAJOOL Ls« / Idle Island MS EAST END Ls^ Agnes Water MS TARAGOOLA Ls •TANNUM SANDS Si Goondlcum//# j, BUNDABERG COONARR S i k y ?

Mantuan Downs 6n# # Springsure Gp Bn /tm.

KYABRA-::; EROMANG^M^

TtSHE Op

BIGGENDEN

Charleville

ir^WKILLA^

^t^uijpie^

Cunnamulla

^N(teryborougli

HIVESVILLE Do, GURULMUNDIS/,^ M U R G O N MILES B/?« MURGON Ls KINGAROY YARRAMAN Bn- _ STRATHPJNE CI Crows Nest Gr''IPSWICH C I \ Moonie. HELIDON JI MOUNT SYLVIA NUMINBAH Pe

Hebel«

ROSEWOOD Bn Base map and information supplied by the Queensland Department of Mines and Energy

154

NORTH STRADBROKE ISLAND MS Si

INDUSTRIAL MINERALS P + ABRASIVES

CORUNDUM,GARNET

DOLOMITE

•

MAGNESITE |

ASBESTOS

FELDSPAR

MAGNETITE

MICA

MINERAL PIGMENT

BAUXITE

FLUORITE

GEMSTONES

OPAL, SAPPHIRE, CHRYSOPRASE, GARNET AGATE

BROMINE CHROMITE CLAYS • BALL CLAY; CLAY & SHALE FOR BRICKS, POHERY; • CEMENT; REFRACTORY CLAY • BENTONITE • KAOLIN • PALYGORSKITE

DIATOMITE

NEPHELINE SYENITE

BORON MINERALS

LITHIUM MINERALS

ALGINITE

BARITE

P *

DIMENSION STONE

GRAPHITE

•

GYPSUM

•

HEAVY MINERALS ILMENITE, RUTILE ZIRCON. MONAZITE

^ - Produced in Queensland

LIMESTONE

NITROGEN (NITRATES) OLIVINE

* *

PEAT PERLITE

P + PHOSPHATES

]

P - Past production

POTASH

REFRACTORY MINERALS ANOALUSITE, KYANITE, SILLIMANITE, MULLITE, CHAMOHE

P+ *

RARE EARTHS SALT SILICA . SAND; • LUMP

SODA ASH ++ SULPHUR TALC VERMICULITE

++ WOLLASTONITE ZEOLITE

PYROPHYLLITE

I - Major production

+ - Potential production

155

GEOLOGY AND MINERALISATION OF THE GYMPIE PROVINCE L.C.CRANFIELD Department of Mines & Energy, Brisbane Geological evolution and mineralisation of the Gympie Province have been a conjectural subject for the past 20 years. Present ideas on the tectonostratigraphy of southern (^eensland limit the areal extent of the Gympie Province. Amamoor and Good Night beds formerly considered parts of the Gyii5)ie Province are now considered as part of the New England Orogen (NEO). These units are deformed and metamorphosed remnants of a Late Devonian to Middle Carboniferous accretionary complex formed during Andean-style subduction at the eastem margin of the Australian continent. The Gympie Province is currently considered as a package of arc-related mafic to felsic volcanics, volcaniclastics, and marine and non-marine sediments of Early Permian to Early Triassic age. Units of the Gympie Province are the Gympie Group (Permian) and the Keefton Formation, Brooweena Formation and Kin Kin beds (Triassic). The probable depositional environments and tectonic history of units of the Gympie Province and its extent in space and time are outlined. The mode of emplacement of the Gympie Province to its present position and its structural complexity are reviewed in terms of present models for evolution of the Bowen Basin Mineralisation in the Gympie Province is mainly epigenetic of Early to Middle Triassic and Late Triassic age. The earliest mineralisation is coincident or immediately post-dates foreland loading deformation associated with the Hunter-Bowen orogeny. This event is represented by thrust-related veining and porphyry-style copper-silver-lead-gold and gold deposits. Skam mineralisation and vein deposits are associated with this event, but there was little wall rock alteration. Precious metal-bearing quartz veins (some associated with granitic intrusive rocks) are hosted by Permian volcaniclastic and sedimentary rocks and Early Triassic sedimentary rocks. Base metal-bearing quartz veins are hosted by Triassic intrusive rocks. Antimony mineralisation occurs in Permian limestone and Early Triassic sedimentary rocks. Gold mineraUsation in the Gympie Province occurs in five separate environments • Quartz veins in volcaniclastic, sedimentary and metasedimentary host rocks with no direct relation to intrusive rocks (eg Gympie Goldfield, Stanton Harcourt) • Quartz vein mineralisation hosted by the Brooweena Formation at the contact to small diorite intrusions (Glenbar, Dawn, Moonlight and Mount Scougall) • Skam deposits in Permian Limestone at the contact to the Early to Middle Triassic diorite (Mount Allen, Mount Suthers and Munna). • Holocene alluvial and eluvial deposits. Late Triassic mineralisation is associated with metasomatic skams and development of more extensive wall rock alteration (argillic, propylitic, and sericitic) in vein and intrusive contact-brecciated deposits. Mineral commodities include precious metals, copper, lead, zinc, antimony, manganese, cobalt, molybdenum, mercury, barite, and gemstones.

156

THE GEOLOGY AND MINERALISATION OF THE MOUNT CANNINDAH PORPHYRY COPPER AND GOLD SYSTEM

Patrick Creenaune, District Geologist, Qld & SW Pacific Newcrest Mining Limited, Brisbane, Qld and Ken Harvey, Manager Gold Projects, Eastern Australia, M M Exploration Pty Ltd, Brisbane, Qld Summary The Mt Cannindah porphyry copper mineralisation is contemporaneous with the intrusion of a PermoTriassic biotite granodiorite stock into a sequence of Carboniferous shallow water marine sediments. The mineralisation is characterised by a central copper molybdenum stockwork with extensive brecciation and skams. The copper molybdenum stockwork occurs in the intrusive and adjacent country rock and is comprised of chalcopyrite, pyrite and molybdenite in thin quartz veinlets within potassic alteration. The intmsion caused hydraulicfracturingof the host rock resulting in a large area of brecciation which consists of angular clasts with voids which have been infilled with quartz-carbonate-chlorite-pyrite-chalcopyrite. Proximal calcareous units have been replaced to form gamet-epidote-carbonate-magnetite skams. Gold mineralisation post-dates the porphyry copper mineralisation and occurs in north-east trending phyllic altered shears and quartz veins. Introduction The Mount Cannindah porphyry copper-gold system is located 90 km south of Gladstone in south-east Queensland at latitude 24°40'S, longitude 151°16'E. Alluvial gold was first discovered in 1889 and some rich patches were found and worked over a six month period after the discovery. Hard rock mining of high grade copper at the Mt Cannindah mine followed. Intermittent small scale copper mining continued in 1896, 1904, 1913 and 1930. In the 1950s a copper precipitation plant was erected which produced between 10 and 20 tonnes of copper per year. The precipitation plant ceased operation in 1973. Total recorded production from the Mt Cannindah district is approximately 10001 of copper and 30,000 oz of gold mined primarily from the Mt Cannindah mine. Recent exploration has defined three prospects where inferred resource estimates have been calculated; at the Mt Cannindah mine 4.4 m.t. @ 0.93% Cu, 20 g/t Ag and 0.4 g/t Au; at the Cannindah East prospect 250,0001 @ 2.82 g/t Au; and the Appletree prospect 30,000 t @ 1.7 g/t Au, 20g/tAgand2.1%Cu. The Mt Cannindah Mineralfieldwasfirstdocumented by Rands in 1896 and more recently by MacAHster (1) and Fletcher (2).

Regional Geology Mt Cannindah is located within the Yarrol Trough which contains a sequence of middle Paleozoic shallow marine sediments, intermediate volcanics and limestones. The sequence has been folded along north northwest trending axes. Folding was accompanied by high angle north northwest trending reverse faults and low angle east over west thrusting, indicative of major compression from the north east. In south east Queensland, a series of late Permian to early Triassic batholiths and stocks with associated copper and molybdenum mineralisation were intruded in a north-west corridor extending over a distance of 400 km from Rockhampton in the north to Gympie in the south. Horton (3). Mt Cannindah lies within this corridor. The emplacement of the Permo-Triassic intrusions was followed by a period of rapid erosion which exposed some of the stocks prior to deposition of the Muncon Volcanics in the middle Triassic. The Muncon Volcanics are comprised of a sequence of tufifaceous sandstone, pyroclastics and basaltflows.Partial erosion of the Muncon Volcanics has exposed a well preserved porphyry copper system. Ore Deposit Features Geological Setting The oldest rocks within the Mt Cannindah area are the Carboniferous Caswell Creek Group which consists of a sequence of siltstones, greywackes, calcarenites and limestones. This sequence dips shallowly to the south-east and forms the westem limb of a north trending syncline (Figure 1). At Mt Cannindah the sediments were intruded by the Cannindah Intrusive, a body comprising diorite, granodiorite and tonalite which occurs in the north west comer of the porphyry copper system. This intrusion forms an elongate, north trending body which dips 65° to the west and forms the structural hanging wall to the copper mineralised breccia at the Mt Cannindah mine. The Cannindah Intrusive is propylitically altered with pyrite occurring in veinlets and as disseminations but does not contain copper-molybdenum mineralisation. Emplacement of the Cannindah Intmsive was closely followed by intrusion of the Monument Intrusive. The Monument Intrusive is an equigranular biotite granodiorite stock which is central and contemporaneous with the stockwork copper157

molybdenum mineralisation and associated skams and Infill Breccia. The Monument Intrusive is a northwest oriented elliptical stock with steep contacts on the south-westem and southem margins and with a shallow dipping contact on its north easterly margin. A series of east-west trending felsic dykes post-date the Monument Intrusive and associated copper mineralisation but are cut by later epithermal quartz veining. The Muncon Volcanics unconformably overlie the Carboniferous sediments. A block of Muncon Volcanics in the northeast portion of the Mt Cannindah area is separated from the outcropping Carboniferous sediments by a major northwest trending fauh, the Kalpower Fault, whose position is interpreted from ground magnetics. The Kalpower Fault tenninates both the outcropping copper-molybdenum minerahsation and surface geochemical anomaly. The north-eastem portion of the porphyry copper system probably lies beneath this block of Muncon Volcanics. There are three main structural trends north, north-west and north-east. All of these structures probably date back to the early Carboniferous to Pennian and have been reactivated through to the Triassic. These structures manifest as shears, mineralised vein breccias and veins, some of which are gold bearing. Brecdadon A large area of brecciation occurs between the Monument Intrusive and the Cannindah Intmsive (Figure 1). The brecciation has surface dimensions of 1200m x 500m and a minimum of 350m depth extent. The breccia comprises angular clasts and an infill assemblage of quartz-chlorite-pyrite-carbonate with accessory chalcopyrite-sphalerite-galenasiderite. Rock flour matrix is absent. The clasts are phyUic altered and exhibit little evidence of rotation. This breccia has been termed the Infill Breccia and is thought to have formed during emplacement of the Monument Intrusive. At the time of intrusion the sediment host rock was hydraulically fractured. Close to the contact, biotite granodiorite is observed permeating fractures and forming irregular dykes. Further away from the contact the fracturing developed voids which were infilled with a hydrothermal assemblage of quartz-chlorite-pyrite-carbonate. Along the westem contact of the breccia the infilled void constitutes up to 50% of the rock and chalcopyrite with minor galena and sphalerite mineralisation occur as additional infill sulphides that post-date the earlier quartz-chlorite-pyrite-carbonate assemblage. A small elongate dyke-like milled breccia occurs on the westem contact of the Monument Intrusive. This breccia consists of subrounded polymict clasts in a fine grained rockflour matrix and post-dates the Infill Breccia and the copper molybdenum mineralisation. This breccia has similarities with late stage pebble dykes which are common in porphyry copper systems elsewhere. Gold mineralised phyllic altered shear breccias occur in the Cannindah East area. The shear breccias post-date Infill Breccia and trend north-east. On the northem margin of the Monument Intrusive a series of northeast trending quartz vein breccias post-date the stockwork copper-molybdenum veins. These breccias comprise chalcedonic quartz, pyrite arsenopyrite and small polymict clasts and have open cavities.

158

V ,V

-I^-ICANNINDAH MINE! COCKADE j i^'^isi^^^Z''Z-^ 2-1-1- J^ JTHEOSORE

-r-T

- -/+ + _-_-f +

:-:: T^v V V

'^X +

+ +

+ +>

• - L I T T L E - - - - - -WONDEFI -JMONUMENT

DUNNO

^LIFESAVER I - " -

^APPLETREE: • 500m Tritsslc Muncon voletnics Uonwntnt Intjys/v9 SkMtn zones Qusrtz V9ins tnd qvtrtz v»/rt brtedMS & shesrs

CtiuUndMh Intrvsiv9 Ctr6onlf0rous S«dlm«nts

Cannindah Mine Infni Breccia

Cannindah East

United Aines

Apple Tree

Cross Section Through the Mt. Cannindah Systenn

Figure 1. Geology of Mount Cannindah The shear breccias, milled breccias and quartz vein breccias form a minor part of the overall system. Mineralisation and Alteration Surface soil geochemistry defines a typical porphyry copper signature centred on the Monument Intrusive (Figure 2). There is an inner zone comprising a low order molybdenum geochemical anomaly surrounded by a concentric zone of copper-molybdenum which straddles the Monument intrusive contact. Both arsenic and gold soil anomalies trend northeast and are considered to be related to later north-east trending structural controls. Johansen (4)

159

SURFACE SOIL GEOCHEMISTRY 0 ^ iliming Low order Mo anomaly ygM Cu'Mo anomaly l\\\\\1 AU'As anomaly Figure 2. Sxirface Soil Geochemical Anomalies Geological mapping and drillcore observations have identified six distinct styles of mineralisation which are described as follows: 1. Disseminated pyrite phase: a pervasive disseminated pyritic halo some ten square kilometres surrounding the hydrothermal system. This disseminated pyrite phase is considered to be a product of homfelsing caused by intrusion of the earlier Cannindah and nearby Kalpower intrusive (Fletcher, 1975). 2. Stockwork copper-molybdenum quartz veinlets: the stockwork forms a wide zone spatially related to the Monument Intrusive contact, with values of greater than 500 ppm copper extending for up to 1 km from intrusive contact and averaging 2000 ppm copper near the contact. A potassic alteration zone with secondary biotite is associated with the stockwork which is overprinted by sericite-quartz-Kfeldspar vein selvage alteration. The stockwork veins are considered to immediately predate the Infill Breccia as the copper-molybdenum quartz stockwork has not been observed overprinting the breccia. 3. Infill Breccia: A large volume of Infill Breccia occurs to the north-west of the Monument Intrusive. The breccia contains angular clasts and voids which have beenfilledwith a hydrothermal assemblage of sulphides and gangue. Significant copper mineralisation occurs in areas of the breccia where the percentage of infill void is high (i.e. exceeding 30%). This occurs along the hanging wall contact between the breccia and the Cannindah Intrusive at the Mt Cannindah mine. In this area, copper has been mined historically and a resource of 6 m.t. @ .94% Cu has been identified. The breccia clasts contain disseminated pyrite, while sulphides within the infill material include 160

pyrite, chalcopyrite with minor sphalerite and galena. The chalcopyrite, galena and sphalerite occur later and enclose the pyrite and other gangue minerals. The observation of chalcopyrite being introduced into the areas of the breccia where the portion of void is high, late in the paragenesis, is crucial. It is interpreted that the brecciation along the contact margin was more 'open' and therefore had more void left when the chalcopyrite mineralisation was introduced. It also indicates that diere was an earlier copper phase which was closely associated with molybdenum in the stockwork style of mineralisation and then a later copper mineralisation which had no associated molybdenum in the Infill Breccia. 4.

Skam mineralisation: Skam mineralisation occurs selectively replacing limestone and calcareous units to the immediate south and north-east of the Monument Intrusive. In the southem area 50m wide skam zones dip 15-20° to the south and crop out over a 1.5 km strike length between the Monument and Appletree prospects. Proximal to the Monument Intrusive the skams consist of gamet-chlorite-epidote-magnetite-pyrite-chalcopyrite. Where intersected in drill core, the skam zones have returned average grades of between 0.35% to 0.45% Cu with minor intervals of greater than 1% Cu.

Gold bearing quartz veins and shears: Gold mineralisation occurs in narrow quartz-pyrite-arsenopyrite veins and phyllic altered shears at Little Wonder, Cannindah East, Blockade, Appletree and Dunno areas. The veins and shears trend north-easterly, dip at steep angles to the south-east and post-date the copper-molybdenum stockwork and Infill Breccia mineralisation.

Arsenopyrite vein breccias: At the northem contact of the Monument Intrusive a series of north-east trending chalcedonic vein breccia structures are observed post-dating the copper-molybdenum stockwork veins. The vein breccias are low in gold are rich in arsenopyrite. These vein breccias have textural and geochemical similarities to north-east trending veins which transect the Muncon Volcanics implying that some of the north-east trending quartz zones cutting the copper-molybdenum mineralisation may be mid Triassic in age or younger.

Conclusion Surface geochemistry and examination of ore textures have indicated two distinct mineralisation events. A classic porphyry copper-molybdenum mineralisation which is spatially and temporally associated with the Monument Intrusive and an overprinting structurally controlled gold-arsenopyrite event. Emplacement of the biotite granodiorite resulted in development of a high temperature copper molybdenum quartz stockwork and proximal gamet-epidote magnetite skam which replaced calcarenite and limestone units. Extensive hydraulic fracturing accompanied emplacement and created angular voids in the surrounding country rock. The voids were filled with a quartz-chlorite-pyrite-carbonate assemblage. A second phase of chalcopyrite mineralisation occurs infilling voids in areas where the void space was highest, along the contact of the breccia with the Cannindah Intrusive. A separate structurally controlled gold mineralisation event occurs in north-east trending shears and veins which transect the earlier copper-molybdenum veins and the Infill Breccia. It is interpreted that at the time of the gold mineralisation the Infill Breccia had sealed and therefore restricted fluid access into the breccia. Acknowledgement The authors would like to thank both Newcrest Mining Limited and MIM Pty Ltd for their permission to publish this paper. Dave Munt and Jeff Rayner are thanked for reviewing the paper. The work by other geologists associated with the history of the project is also acknowledged. References 1. 2. 3. 4.

MacAlister, T, 1964; Technical Report No. 43 "Mt Cannindah Copper Deposit" for CEC Pty Ltd, pp 23. Fletcher, R.J., 1975; Geology and Mineralisation of the Mt Cannindah Mineral Field. Unpublished MSc thesis, JCU, pp 70. Horton, D.J., 1978; Porphyry type copper-molybdenum minerahsation belts in eastem Queensland, Australia, Econ. GeoL, 73, pp 904-921. Johansen, G.J., 1995; Report on Exploration carried out on the Mt Cannindah mining leases during the period 1 January 1993 and 30 June 1995, pp 17.

161

RECENT DEVELOPMENTS IN THE GEOLOGY OF THE GYMPIE GOLDFIELD

Ron Cunneen, Chief Geologist Gympie Eldorado Gold Mines Pty Ltd Summary The Gympie Goldfield is historically the sixth largest goldfield in Australia, with a hard rock production of 116 tonnes of gold bullion between 1867 and 1923. Mineralisation occurs as gold bearing quartz calcite veins and is of two styles, Gympie Veins and Inglewood Veins. Both vein styles developed contemporaneously, with the Inglewood Veins being the feeder structures for the Gympie Veins. Source of the hydrothermal fluids forming the Gympie Goldfield is believed to be a buried intrusion located immediately south-west of the Goldfield. This intrusive is une^osed, but has been well defined by airbome magnetics. The age of mineralisation is mid Triassic. 1. Introduction The Gympie Region forms part of a significant extensive metallogenic province developed in the eastem Australian New England Fold Belt between the Permian and the Triassic. Throughout the region there is a strong spatial and genetic link between base and precious metal mineralisation and the Later Permian to Late Triassic plutons. MineraUsation is strongly associated with structural trends of north-west and north. The historical Gympie Goldfield lies within the Gympie Group of rocks, a conformable stratigraphic succession characterised by waning volcanic activity with increasing sedimentary characteristics towards the top of the succession. Cycling upward from sub-aerial volcanics and volcaniclastics through to shallow water marine limestones and deep water turbidites, the sequence reflects the progressive development and drowning of an island arc system (Figure 2). The Goldfield extends over an area of ten kilometres by four kilometres elongate north-south and is situated on the eastem side of a domal feature dismembered by thrusting and faulting. Within the confines of the Goldfield the stratigraphy dips shallowly to the east and is essentially conformable. Faulting occurs on all scales and has functioned to break the Goldfield up into a number of discrete blocks (Figure 3). The Goldfield, discovered in 1867, was worked continuously until 1923 and is historically the sixth largest goldfield in Australia (behind Kalgoorlie, Bendigo, Mt Morgan, Charters Towers and Norseman). Over 1500 shafts are known (the deepest extending to beyond 900 metres) servicing an estimated 120-150 kilometres of underground workings. Total hard rock production was 116 tonnes of gold bullion from an estimated 4.5 million tonnes of ore. 2. Mineralisation All primary mineralisation within the Goldfield consists of quartz calcite veins with free gold. These veins are of two distinct types: a) Gympie Veins Gympie Veins consist of an array of parallel fissurefilledreefs which strike parallel to the stratigraphy and dip 30° to 80° to the west, normal to bedding. Economic concentrations of gold are restricted to the Rammutt Formation, a unit containing interbedded fine-grained sediments and volcanics deposited during the waning stages of volcanism. A striking example of wallrock composition on ore deposition occurs where the Gympie Veins cut sequences of carbonaceous shale in the upper portion of the Rammutt Formation (the Productive Horizons). Spectacular gold grades were obtained in quartz veins intersecting beds of carbonaceous shale. With the strike of the Gympie Veins being sub-parallel to the stratigraphy the intersection of the veins with the "Productive Horizon" formed sub-parallel pencil-like ore shoots (Figure 3). The width of these ore shoots varied considerably depending on the geometry of the Gympie Vein but was generally narrow (0.30m-3.0m). Vertical extent was that of the thickness of the "Productive Horizon" (20-60 metres). Strike extent was considerable, with the larger Gympie Veins being mined over a length of 1.0-2.0 kilometres. Average grade recovered from the Gympie Veins was 33.7 g/t Au (head grade in the order of 37.4 g/t Au). b. Inglewood Veins In contrast to the Gympie Veins the Inglewood Veins strike north-west and are sub-vertical. Mineralisation occurs in tabular quartz reefs of considerable horizontal and vertical extent in large strike/slip fault structures and in strong association with diorite and dolerite dykes. Several of these structures occur within the Goldfield. The best known of 162

these, the Inglewood Structure (Figure 3) has been traced by a combination of historical workings, diamond drilling and underground development over a strike extent of 3500 metres and to a vertical depth of 1000 metres. Unlike the Gympie Veins economic gold grades are not restricted to the intersection with the "Productive Horizons" but occur throughout the Fault structure. The Inglewood Structure is a large complex strike/slip fault system that has been intermittently active over a considerable time span. The fault channel is sub-vertical and varies in width from 10-25 metres. Multiple dolerite and diorite dykes have intruded the fault during its activity. Gold-bearing quartz veins are hosted within the dolerite dykes, with the dolerite dykes spanning the time of emplacement of the quartz veins. The diorite dyking is younger than the mineralising event, and in places cross cuts the dolerite dyking and quartz veining. Typically two quartz veins are developed with individual width ranging between 20 and 140 centimetres. The quartz is commonly laminated reflecting the crack seal formation with successive fault movement. On the eastem side of the Inglewood a number of subsidiary mineralised splays roll out (peel off) to the east of the main fault channel. Historical mining on the Inglewood Fault Structure produced 1.8 million tonnes of ore at a recovered grade of 9.6 g/t Au (head grade in the order of 12.1 g/t Au)froma combination of quartz veins within the Fault and a stratabound stockwork formed by the quartz vein rolling out (peeling off) into the plane of the stratigraphy on the eastem side of the Fault. With the mining techniques and cost structures of the time this was a sub-economic grade. Most mines operating on the Inglewood also mined the Gympie Veins, using the Inglewood to provide the tonnage and the Gympie Veins to provide the grade. With the cost increases brought on by the First World War, mining of the Inglewood was abandoned leaving large bodies of stone behind. 3. Formation of the Gympie Goldifield a. Gympie Vein/Inglewood Relationships. Formation of the Gympie Goldfield has traditionally been interpreted as a two-phase event with early Gympie Vein formation post dated and offset by the Inglewood Fault Vein system. Recent work by Gympie Eldorado Gold Mines Pty Ltd has led to a major re-interpretation of the mineralisation of the Gympie Goldfield. The Inglewood Fault developed contemporaneously with the Gympie Veins, this fault being the feeder and controlling structure for the southem portion of the Goldfield. Evidence for this interpretation is: LA large degree of overlap in the trends of Gympie and Inglewood veins, with a smooth transition between them. 2. Detailed underground mapping which shows Gympie Veins, on approaching the Inglewood Structure, to alter orientation and merge with the quartz veins in the Inglewood Structure. All Gympie Veins mapped which were truncated by the Inglewood Stmcture were found to truncate either against post mineralisation faults or post mineralisation dykes on the outer contact of the Inglewood Structure. 3. The two main quartz textural types (massive and laminated) occur in both systems. 4. Both show identical associated mineralisation (pyrite, galena, chalcopyrite). 5. Both show identical associated sericite/clay alteration. 6. Both have the same stable isotope signature and the same fluid inclusion homogenisation temperature. b. Mineralisation Source Source of the gold mineralisation at Gympie has historically been thought to be related to the Permo-Triassic intrusive episode. Supporting evidence for intrusive related minerahsation is: 1. Dyking - Throughout the Goldfield there is a strong spatial and timing relationship between the episodes of dyking and quartz vein formation. Timing relationships show the dyke ages to span the time of formation of both the Gympie and Inglewood systems. The magmatic source of the dykes would seem to be the most likely source of the auriferous quartz veins. 2. The presence of granitic xenoliths in the diorite dykes of the Inglewood Fault. 3. Oxygen isotope data for quartz is consistent with a magmatic of metamorphic source. 4. Gold bullion fineness for Gympie is 820 which is close to the average of plutonic types (fineness 825). The most commonly proposed intrusive source for the Gympie mineralisation has been the outcropping Woondum Granite located some 16 kilometres south-east of the Goldfield. Recent image enhancement of airbome magnetics, however, shows a buried intrusive mass immediately south-west of the Goldfield which is a more likely source of the hydrothermal fluids forming the Gympie Goldfield than the Woondum Granite. Associated with this buried intrusive are a number of well developed north-west trending magnetic linears with the Goldfield itself occupying the northem side of one of these linears. c) Age of Mineralisation Age dating by Gympie Eldorado Gold Mines Pty Ltd gave 245Ma ± 14Ma for hydrothermal sericite in andesite adjacent to the Inglewood Fault and 229Ma ± 6Ma for hydrothermal sericite in dolerite within the Inglewood Fault. With regard 163

to the dolerite age the sample was taken close to the contact with a diorite dyke which is younger than the mineraUsing event. Emplacement of this diorite dyke would have reset the sericite, so die age reading is younger than that of the hydrothermal event. Given the error parameters it is likely that the two ages obtained reflect the time frame enclosing the main episode of hydrothermal activity forming the Gympie Goldfield, ie: Mid Triassic. 14-fr

2**10

2610-

2410'

2d*20'

0 1

Scale 1:250 0 0 0 5 10 I I

ISkm

LOCATION MAP Figure 1

164

nOIMEl

TURBIDITE SANDSTONE,SHALE 1 — r

ii X u 5 2

LIMESTONE o O

o D o

o O

COARSE ROUNDED PEBBLE CONGLOMERATE

O VOLCANIC CLASTIC SEDIMENTS, CARBONACEOUS SHALEJNTERCALATED A N D E S f t l C LAVASjPYROCLASTICS

ANDESITIC LAVAS,PYROCLASTICS INTERCALATED VOLCANISTICS, CARBONACEOUS SHALE 3 2 Z < (T

ANDESITE L A V A S , PYROCLASTICS V

V V

A A

Scale N T.S.

BASALT TO ANDESITE AGGLOMERATE, PILLOW LAVA

Project No.

GENERALIZED STRATIGRAPHIC COLUMN FIG. 2

Dat* MCH 9S

Drg.Nc.

Figure 2

165

Drawn by Robin Voss (014 048887)

AN OUTLINE OF PALAEOFORMATION WATER COMPOSITIONS AND FLOW PATTERNS IN THE EROMANGA AND COOPER BASINS THROUGH MESOZOIC TIME. Peter EadingtonS Mark Person^ Denah Toupin', Joe HamUton' L CSIRO Division of Petroleum, North Ryde, NSW, 2113 2. University of Minnesota, Minneapolis, USA 3. GEI Consultants, Winchester, MA, USA Summary Measurement of the salinities of fluid inclusions, the isotope composition of minerals, and two dimensional computer simulations of basin scale flow of formation water have been used to investigate the composition and flow pattems of palaeoformation waters in the Eromanga and Cooper Basins through Mesozoic time. Three main phases were identified. 1. Compaction driven flow during rapid sedimentation in Cretaceous time. Flow was cross formational from the underlying Cooper Basin and confinement in Jurassic aquifers under the transition bed aquitards produced radial flow to the basin margins. Fluid inclusions in Jurassic aquifers trapped palaeoformation waters at this time with salinities to 18,000 ppm which is higher than salinities of current formation waters (<4,000 ppm). Unlike current formation waters which retain a meteoric character, isotope compositions of palaeoformation waters are enriched in consistent with a long residence time in the basin. Fluid inclusions in the Cooper Basin trapped palaeoformation waters with salinities commonly about 60,000 ppm which is consistent with alteration of formation water during burial diagenesis. Current formation waters have salinities of 10,000 to 19,000 ppm. 2. Hydrodynamic flow at about 50 Ma due to a mid basin uplift forming a recharge zone near Mt. Howitt in south west Queensland and bifurcating flow to discharge zones located over the Eromanga Basin north east of Mt. Howitt and to the south west over the Cooper Basin. Flow paths penetrated into the Cooper Basin and contributed to flushing of high salinity formation waters. 3. Hydrodynamic flow from Pliocene time having the current configuration from a recharge zone at the uplifted eastern margin of the basin to the south west. Initially this formation water flow was more intense as reflected in large fossil mound spring deposits that indicate greater discharge rates. Recharge of meteoric water during the Pliocene or mdd Tertiary hydrodynamic regimes flushed high salinity formation waters from aquifers at stratigraphic levels from lower Cretaceous to basal Permian. Introduction Palaeoformation water flows are of interest in hydrocarbon exploration for modification of geothermal gradients in relation to maturity and oil generation^ hydrodynamic effects on oil migration directions, tilting of oil water contacts and flushing", stripping of gas from hydrocarbon accumulations by dissolution in formation and their influence on diagenetic reactions that affect reservoir quality. Previous investigations of the composition and flow of palaeofomiation waters in the Eromanga Basin have been carried out using fluid inclusion and isotope measurements at CSIRO"'''' and simulations of basin scale fluid flow at the Universities of New Hampshire^ and Minnesota®. The current formation water flow system in the Eromanga Basin has been reviewed by HabermehP while variations in the intensity of the flow of formation waters and the implications for hydrocarbon migration, tilting and flushing of oil accumulations has been examined by Bowering". Methods Fluid inclusions occur in diagenetic minerals and trap palaeoformation fluids at the time of closure of the fluid inclusions. Fluid inclusions are time specific in terms of the fluid they trap and the temperature they yield. Thermometric measurements are made by visually observing phase changes at controlled temperatures by microscopy. The isotope composition of palaeoformation waters is determined by referring the isotope composition of diagenetic minerals and the temperature at which the mineral crystallized to e^qjerimentally determined fractionation equations between the mineral and water ^Eadington et al, 1989). Computer simulations of two dimensional basin scale flow of formation water were made along NE-SW and NW-SE transects crossing the Patchawarra Trough of the Cooper Basin (Fig. 1). Flow of formation water occurs in response to changing thickness of sediments in basins, surface topography (gravitational potential) and water density (thermal or haline) effects. Burial histories from the age and thickness of sediments and depth porosity relationships provide the basic empirical data. This is supplemented by shale depth-velocity-compaction relationships and sedimentary environments to constrain uplifts during periods of nil or superficial deposition. This information was collated, for the two transects along which formation water flow was investigated, as data for a computer program to simulate formation water flow 167

incorporating these influences^ Temperatures were computed using conductive and convective heat flow and used to compute maturity and oil generation. The algorithms and underlying theory of the simulations are outlined by Toupin et The effects of hydrodynamics on migration of oil and high salinity fomiation waters were investigated by computing oil heads and salt water heads'^

SOUTMERN PAOFIC OCEAN

SOUTHERN OCEAN

Figure 1 Constraints on simulations are provided by the ability of the simulation to reproduce current formation water flows and temperatures, to predict vitrinite reflectances as a maximum palaeotemperature indicator, and to account for fluid inclusion and isotope data which constrain the origin of palaeoformation waters and changes in the composition of palaeoformation waters through time. Sediment thicknesses were from formation tops in databases of the Queensland Department of Mines and Energy, South AustraUan Department of Mines and Energy and Santos Ltd. The composition and flow of current formation water, porosity depth relationships and thermal maturities were from published data. Uplifts were derived from shale velocity depth relationships'^ structure maps and sedimentary environments of Tertiary sediments. Results Salinities of Palaeoformation Waters from Fluid Inclusions Measurements of aqueous fluid inclusions in quartz overgrowths (Table 1) in Hutton Sandstone made at Jackson 1 and Bogala 1 yield salinities from 0 ppm to 6,975 ppm and inclusions in magnesian siderite yield sahnities from 1,800 ppm to 18,900 ppm4. In Hutton Sandstone at Tintaburra 1 salinity measurements for aqueous inclusions in quartz overgrowths range from 3,500 ppm to 10,413 ppm. Current day formation waters in the Eromanga Basin have salinities from 600 ppm to 4,000 ppm^ Fluid inclusion measurements®'^® have been made in Upper Permian (Toolachee Formation) and Permian (Patchawarra Formation and Tirrawarra Sandstone) aquifers at Tindilpie 1, Kanowana 1 and Tirrwarrra 5 in the Patchawarra Trough of the Cooper Basin. The formations and fluid inclusion salinities with current salinities in brackets are Toolachee Formation 9,000 to 68,000 ppm, median 12,000 ppmi, (2,500 -10,000 ppm). Top Patchawarra Formation 15,000 ppm to 167,000 ppm, median 48,000 ppm,(10,000 ppm) and Tirrawarra Sandstone 7,000 ppm to 203,000 ppm, median 60,000 ppm (19,000 ppm).

168

Formation Current

Salinity of

Oxygen Isotope

Hydrogen

Salinity of

Palaeoformation water (ppm)

Composition of Palaeoformation

Formation Water (ppm)

Hutton Sandstone Toolachee Fm.

0 to 18,900 -4.5 -90 600 to 4,000 9,000 to 68,000, 2,500 to 10,000 median 12,000 Top Patchawarra Fm. 15,000 to 10,000 167,000 ppm. median 48,000 Tirrawarra Ss. 7,000 to 19,000 203,000 median 60,000 Table 1. Salinities of fluid inclusions and isotope compositions of palaeofoimatim waters in Jurassic and Permian aquifers of the Eromanga and Cooper Basins. Isotope Composition of Palaeoformation Waters Hydrogen isotope measurements on samples in the Hutton Sandstone^ indicate formation waters during kaolinite crystallization had 5D of-90%o. Current formation waters have 5D of-50%o. Oxygen isotope measurements on samples of Hutton Sandstone indicate formation waters during crystallization of quartz overgrowths and siderite had of - 4 . 5 % o . Current formation waters have of-6.7%o. Simulations of Basin Scale Formation Water Flow Simulations of formation water flow (Fig. 2) show that during mid Cretaceous time (90 Ma) high sedimentation rates during deposition of marine sediments caused cross formational compaction driven flow from the Cooper Basin into the Eromanga aquifers with confinement beneath the transition bed aquitards producing radial flow from depocentres of Cretaceous sedimentation. In early Tertiary time (50 Ma) a mid basin uplift centred at Mt. Howitt (south west Queensland) resulted in a regional recharge area in which temperatures were suppressed by up to 40°C by down flowing formation w'""^ This resulted in a bifurcating flow pattem radiating from Mt. Howitt to discharge zones in mid basin topographic lows which were areas of mid Tertiary sedimentation. Temperatures in discharge zones were increased by up flowing formation water. This was an energetic formation water system with flow through Cretaceous formations and deeply penetrating flow paths that flushed high salinity formation waters from the Cooper Basin. This intensity of formation water flow decreased as topography was levelled by erosion. In Pliocene time (5 Ma) uplift on the eastern margin of the Eromanga Basin produced a configuration of formation water flow similar to the present. Initially this was a high energy system during which the flow of formation water reached maximum values of 127 m/y which further flushed formation waters in Eromanga and Cooper Basin aquifers. Flow rates decreased as topography was eroded to current levels. Computed salt water heads indicate brine flushing deep within the Cooper Basin during the 50 Ma hydrodynamic system and to a lesser extent during the Pliocene to recent hydrodynamic system. Computed oil heads indicate hydrodynamically promoted reservoir charging above the southern end of the Cooper Basin during the 50 Ma hydrodynamic system only.

169

90 Ma

50 Ma

5 Ma

Figure 2 Discussion In the Jurassic Hutton Sandstone at Jackson 1, Bogala 1, and Tintaburra 1 the maximum salinities of fluid inclusions are 10,400 ppm in quartz overgrowths and 18,000 ppm in magnesian siderite. The flxiid inclusion data indicate salinities for palaeoformation waters were up to 18,000 ppm. SaUnities have since decreased to current values of 600 to 4,000 ppm. The enrichment in ^^O and depletion in Deuterium in the Hutton Sandstone compared with current formation water indicates modification of water by burial diagenesis. This is in accord with simulations indicating compaction driven flow in mid Cretaceous time when palaeoformation water had a long residence time in the basin and increased in salinity and accumulated from mineral reactions. Current formation waters in the Eromanga Basin have a short residence time and have low salinities and retain an isotope composition characteristic of modem meteoric water. The salinities of palaeoformation waters in fluid inclusion in the Permian aquifers of the Cooper Basin are higher than salinities previously reported in the Cooper and Eromanga Basins. Salinities of 50,000 to 60,000 ppm are common and are attributed to modification by burial diagenesis. Ultrahigh salinities of 167,000 to 200,000 ppm occur infrequently in the Tirrawarra Sandstone aquifer and are attributed to cross formation flow of hypersaline brines from the calcareous Warburton Basin underlying the Cooper Basin. No ultrahigh salinities have been measured influidinclusions in samples above the Permian Murteree Shale regional seal. The dispersion of high and ultrahigii salinity palaeoformation waters from Cooper Basin aquifers (current formation waters have salinities of 10,000 ppm to 19,000 ppm) requires flushing by flow of low salinity waters and is consistent with simulation results indicating energetic hydrodynamic flow of formation waters in mid Tertiary and Pliocene time when meteoric water recharging the basin flowed along deeply penetrating flow hnes including through aquifers in the Cooper Basin. The simulation result of energetic flow of formation waters in mid Tertiary time may correlate with evidence in fission track lengths'^ for mid Tertiary heating by fluid flow along faults at the margin of the Frome Embayment of the Eromanga Basin. The evidence for an energetic fluid flow system in Phocene time is consistent with large fossil mound spring deposits in South Australia' that indicate higher discharge rates than at present. The simulation results are in accord with the observation that sedimentary basins with sub-aerial exposure frequently have formation water flows due to topographic effects. The current active formation water system is illustrative of the 170

Eromanga Basin at different times through the Tertiary Period. However, differences in the distribution of uplifts have resulted in different configurations and intensity of flow of palaeoformation waters. Conclusions Palaeoformation water flow pattems in the Eromanga Basin were controlled by compaction during high rates of sedimentation in the Cretaceous Period and by changing topography due to two main periods of uplift during the Tertiary Period. The salinity of palaeoformation waters in the Cooper Basin of about 50,000 ppm and the enriched isotope composition of palaeoformation waters in the Eromanga Basin are consistent with alteration by burial diagenesis, a long residence time in the basin and slow flow rates associated with compaction driven flow. An early Tertiary uplift near Mt. Howitt south west Queensland caused a recharge zone and radial flow towards the margins of the basin. Meteoric water recharging the basin flowed along deeply penetrating flow paths and flushed high salinity formation waters in aquifersfromlower Cretaceous to basal Permian rocks. The current configuration of formation water flowfromthe eastem uplift to the south west occurred from about 5 Ma. Initially flow rates were faster than current which is manifest in large fossil mound spring deposits consistent with greater discharge rates in the past. Acknowledgements The authors thank all organisations that have contributed data and discussed the design and progress of the investigations, particularly Santos Ltd., Queensland Department of Mines and Energy, Mines and Energy, South Australia, and Australian Geological Survey Organisation. Financial support has been provided by QDME through a NERDDC research grant and by the Petroleum Research Fund of the American Chemical Society. References 1. Person, M., Toupin, D., and Eadington, P., 1995. Effects of convective heat transfer on the thermal history of sediments and petroleum generation within continental rift basins. Basin Research, v. 7., p. 81-96. 2. Bowering, O.J.W., 1982, Hydrodynamics and hydrocarbon migration - a model for the Eromanga Basin, APEA Joumal, V. 23, p. 227-236. 3. Dunlop, E.C., Browne, M.V., and E.F. Tadiar, 1992. Depletion of gas by molecular diffusion-A case study. APEA Joumal, V.32, p. 369-390. 4. Eadington, P.J., Hamilton, P.J., & Green, P., 1989. Hydrocarbonfluidhistory in relation to diagenesis in the Hutton Sandstone, south - west Queensland, in O'Neil, (ed.). The Cooper and Eromanga Basins Australia. Proc. of Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Australian Society of E^q)loration Geophysicists Conference, 601 - 618. 5. Eadington, P.J., Hamilton, RJ., and Bai, G.P., 1991. Fluid History Analysis - A new concept for prospect evaluation. APEA Joumal, v.31, p. 282-294. 6. Eadington, P.J., Hamilton, RJ., Lisk, M., Toupin, D., Person, M., Wamer, D., 1993. Hydrologic controls on petroleum generation within the Cooper and Eromanga Basins, Australia. 11 Fluid inclusion, isotopic, and geothemial data, in, Pamell, J., Ruffell, A.H., and Moles, N.R., Geofluids '93, Conference Proceedings, p. 338 341. 7. Toupin, D.K., 1993. The effect of formation waterflowpattems in evolving intracratonic sedimentary basins on heat flow and petroleum generation. M. Sc. Thesis, University of New Hampshire. 8. Person M., Toupin, D, Wieck, J, Eadington, R Wamer, D., 1993. Hydrological constraints on petroleum generation within the Cooper and Eromanga Basins, Australia: I Mathematical modelling, in, Pamell, J., Ruffell, A.H., and Moles, N.R., Geofluids '93, Conference Proceedings, p.264- 266 9. Habermehl, M.A., 1986. Regional formation water movement, hydrochemistry and hydrocarbon migration in the Eromanga Basin, in Gravestock, D.L, Moore, RS., and Pitt, G.M., Contributions to the Geology and Hydrocarbon Potential of the Eromanga Basin Geological Society of Australia Special Publication No. 12, p. 353-376. 10. Toupin, D., Eadington, RJ., Person, M., Morin, R, Wieck, J., and Wamer, D., in press. Petroleum hydrogeology of the Cooper and Eromanga Basins, Australia: some insights from mathematical modelling and fluid inclusion data. AAPG Bull. 11. Rodgers, J., Wher, F.L., and Hunt, J.W., 1991. Tertiary uplift estimationfromvelocity data in the Eromanga Basin. Exploration Geophysics, v. 22, 321-324. 12. Foster, D.A., Murphy, J.M., and Gleadow, A.J.W., 1994. Middle Tertiary hydrothermal activity and uplift of the northern Flinders Ranges, South Australia,: Insights from apatite-track thermochronology. Australian Joumal of Earth Sciences, V. 41, p. 11-18.

171

FLUORIDE ANOMALIES IN AQUIFERS OF QUEENSLAND SECTION OF THE GREAT ARTESIAN BASIN AND THEIR SIGNIFICANCE

Peter A. Evans Principal Environmental Officer, Waste Management Branch, Queensland Department ofEnvironment. Summary: In some areas of the Great Artesian Basin the fluoride content of the groundwater is significantly elevated. The elevated fluoride is a significant issue for water resource development because both humans and stock can be adversely effected. An investigation to delineate the areas of fluoride enrichment using both water bore samples and petroleum well drill stem samples revealed some areas of anomalously elevated groundwater fluoride across the Great Artesian Basin, and in particular a significant fluoride anomaly in the eastem areas of the Surat Basin (eastem Great Artesian Basin). The recharge areas for the aquifer units of the Great Artesian Basin are generally low in fluoride, and a possible influence on this a geochemical evolution downflowpath in the basin with a relative reduction in calcium and increase in sodium. The concentration offluoridein solution partly depends on the solubility of fluorite. Where there is very low calcium concentrations there is scope for elevated fluoride concentrations in solution. It appears likely that those fluoride anomahes which cannot be attributed to groundwater circulation into weathered granitic basement are due to the upward migration of "geochemically mature" groimdwater from the deeper aquifers. The linear fluoride anomaly in the eastem Surat Basin which coincides with the Leichhardt - Burunga and Goondiwindi - Moonie Fault systems effects all of the major aquifers and this is indicative of vertical migration of groundwater. It has been previously postulated that the hydrocarbons which accumulated in the Moonie structure (30 million barrels) migrated up the Goondiwindi - Moonie Fault. It may be possible to use subtle fluoride anomalies in younger aquifers above prospective reservoir sandstones as an exploration tool to deUneate areas of upward migration of groundwater. 1.0 Introduction The Great Artesian Basin is one of the largest artesian groundwater systems in the world covering 1,711,000 km^ (660,617 sq. miles). The basin occupies approximately one fifth of Australia and extends across parts of Queensland, New South Wales, South Australia and the Northem Territory (Habermehl, 1) (see figure 1). The groundwater resources provide the only rehable water supply for the generally semi-arid inland area underlain by the basin. The basin consists of Triassic to Cretaceous age sedimentary rocks. Three major sub-basins, the Surat Basin, Eromanga Basin, and Carpentaria Basin, are recognised within the Great Artesian Basin (seefigure1). These sub-basins are partially separated by basement highs. The Great Artesian Basin consists of a multi-layered system of sandstone aquifers and siltstone, mudstone and shale aquicludes. Figure 2, which details the generalised stratigraphy of the basin illustrates the complexity of the groundwater system. Generally groundwaterflowsfromthe uplifted eastem areas, where the aquifer units outcrop, towards the southwestem, westem and southem margins, (Habermehl, 1) where springs and diffuse upward discharge occurs (see figure 1). It has been long recognised that within the aquifers of the Great Artesian Basin there are some areas where the fluoride content of the groundwater is significantly elevated. The elevated fluoride is a significant issue for water resource development. Elevated fluoride in water can cause fluorosis in mammals leading to problems with teeth and skeletal development.

172

NORTHERN TERRITORY

- —

Direction Of Flow

- - - Structural Ridges

Figure 1. Great Artesian Basin indicating generalised flow direction Both humans and stock can be effected, although the susceptibility to fluorosis is greatest during early development. The drinking water limit for fluoride is relatively low at 1 mg/L. Although adult stock can usually tolerate fluoride up to 10 mg/L, fluorine in excess of 2 mg/L can be harmful to stock during the first 3 years of life. A plan of Queensland (see figure 3) was compiled indicating all areas of elevated groundwater fluoride by plotting all of the groundwater chemical analyses stored in the Department of Primary Industries (DPI) groundwater data base irrespective of the individual aquifers used. The plan revealed some areas of anomalous groundwater fluoride across the Great Artesian Basin, and in particular a significant anomaly in the eastem areas of the Surat Basin. This work was followed up with searches of analytical data held by the Department of Minerals & Energy (DME) for the Surat Basin area as well data held for petroleum exploration well drill stem test samples (Seefigure4). 2.0 The Geochemistry of Fluorine Fluorine is the lightest member of the halogen elements, and is the most electronegative of all of the elements (Hem, 2). In solutions fluorine is usually present in the form of the free fluoride ion F-, although it may form solute complexes with certain cations. Although fluorine and chlorine are halogen elements, fluorine behaves very differently in that fluoride it is far less soluble than chloride. The solubility of calcium fluoride limits the concentration of fluoride in waters with more than 10 mg/L calcium. According to Hem (2) fluorine concentrations in most natural waters are low and are generally less than 1.0 mg/L (where TDS < 1000 mg/L) although fluoride concentrations in groundwater of up to 67 mg/L have been reported. The concentration of fluoride in seawater is relatively low at approximately 1.3 mg/L (Mason, 3). In waters with pH values less than neutral fluoride ions would normally complex with aluminium (Hem, 2). Hem (2) noted that fluoride is often a convenient constituent of water to use as a tracer in establishing paths of movement of water. Fortunately, although fluorine has a tendency to form complexes in water with changes in pH, the Queensland Government Chemical laboratory routinely adds a "decomplexing" agent to samples before fluoride determinations. This agent liberates most of the complexed fluorine although this process becomes less efficient where dissolved iron exceeds 4.0 mg/L. In most cases therefore, the fluoride analyses reported for groundwater samples collected by most Queensland Government agencies are likely to be reasonably reliable (except in cases where dissolved iron levels exceed approximately 4.0 mg/L).

173

AGE

FARSOUTHWEST MOBTHWEST

CAONAO -WE FORMAT O IN

WNITON FORMATKM <saS!SS> MAOQNJDA FORMATOIN csHS> ALAIRUMUDSOrCI TOQLESUC FORMATOIN \ - z ; ^OONCASTER\ IBBSt <3SniS> CAONAO -ME FORMAT O IN 4 4

RONLOW LONCS G IHT BEDS UB fSTONE UWSTCNE (Baa^Mlifsin) (Bail M«ii facta) S (BsiiMigia faein) s

AGE

BASEMENT

UNPBAFLEABEEONZTOAICITED PABRUUYN GnAM ITSSt PRPNfElCOAM MOTEORNDM MSBETNAM CS

1 WYANDRA SANDSTONE MBBER

HOORAY SAWSTOK

ii 11 sSrnjS^m^m FORMATOIN XOHAUi

OFORMATIOli

GUBBERAMUNOA SANDSTONE

4 WESTBOURK FORMAT O IN AOORS IANDSTOC I 4 URO ICAO FORMAT O IN HUTTON SANDSTONE

A • BASAL UJKASSEKV*DGiREa FORMATOIN omiEBAsai ci m f BASMoBfASaENT MFOOROHAUTmKM BER lfAMNDTES BIUTO GROUP CEJISM NOAQUFBSt KTAGAM SANDSTOGNREOSUP ANADTSSTIONGESROUPaouns NDSTONES MPMCS \S^ABEDS EWAATKNM FORRM EWAATKNM 4 FORRM FORMEWUAHNON BASIN. E

GRHAN CREK FORMATOIN SURATSLTSTONE

COOWI«GAULH BASNvBASOENT

EVER6RSN4 FORMATKMaCSSMM

PREQPICE SANDSTONE

BOWENBASM or BASEMENT

LEGEND I

I Significant Aquifer Confined Aquifer as fiydraufically unconnected tenses 4

THE GREAT ARTESIAN BASIN

QUEENSLAND STRATIGRAPHY

Significant economic tiydrocart)on occurrences

Figure 2. Stratigraphy of the Queensland sections of the Great Artesian Basin 3.0 Natural Sources of Fluorine in Groundwater The most common fluorine bearing minerals are fluorite [CaF2] and apatite [Ca5 (PO4, 003)3 (F, OH, CI)]. There are alsofluorinebearing micas such as lepidolite and phlogopite and minor minerals such as topaz, amblygonite and cryolite (rare). Fluorine can also be present in amphiboles as a replacement for an OH- in the mineral structure. Fluorine is often associated with volcanic gasses (Hem, 2). According to Rankama & Sahama (4) rocks rich in alkali metals, and obsidian are significantly higher in fluorine than other igneous rocks. Sedimentary rocks can be sources of fluorine. Phosphatic rock can contain appreciable fluorine in the order of 3% and Hem (2) noted that fluoride bearing mineral particles are widespread constituents of resistate sediments. Traditionally it has been regarded that most of the fluoride in Queensland aquifers is sourced from fluorine bearing accessory minerals in granitoids. 4.0 Fluoride Anomalies In The Great Artesian Basin Some areas of significant fluoride anomahes can be discemed in Figure 3.

174

KILOMETRES

500

625

Fluoride Ranges 2.01 - 3.00 mg/L 3.01 - 4.00 mg/L > 4.00 mg/L Great Artesian Basin

Central-eastern Eromanga Basin Anomaly

LongFeach^yt

Manaroo^. .Platform^

Southwestern Eromanga Basin Anomaly

Roma Naccowlah West 1 A Oilfield Jackson • Oilfield .

• •

^ o

o ~

Eulo Ridge Anomaly

Eulo

See Figure 4

Figure 3. Distribution of elevated fluoride values in the Queensland sections of the Great Artesian Basin 4.1 The Carpentaria Basin Anomalies The Carpentaria Basin has generally been regarded as the basin most effected by fluoride problems. Fluoride concentrations up to 37.5 mg/L have been recorded from the basin. It has been postulated that the circulation of groundwater into granitic and metamoiphic basement beneath the basal Gilbert River Formation causes fluoride enrichment. Altematively a process of dissolution offluorinebearing minerals in the sediments which have a granitic or metamorphic provenance has been postulated. Mineral exploration carried out on a fluoride anomaly in the southem Carpentaria Basin (Esson & Burban, 5) revealed that both the underlying basement granite (0.11% F) and host aquifer sediments (generally less < 0.02% and max. 0.21% F) had lowfluorineconcentrations. This of course does not rule out a basement or host aquifer source, because of possible dispersion of the fluoride due to groundwater flow. 4.2 The Eulo Ridge Anomaly — Southem Eromanga Basin Fluoride levels in the vicinity of the Eulo Ridge (a basement high which outcrops as a granite inlier in the vicinity of Eulo) reach up to 8.1 mg/L. The Eulo Ridge fluoride anomaly can be rationalised in a similar manner to the anomalies of the Carpentaria Basin, as being either due to the circulation of groundwater from basal units into underlying granite 175

basement or the dissolution of fluorine rich accessary minerals in the aquifers. Muller (6) has suggested that micas in the granites of the Eulo Ridge may be the source of the fluoride anomaly in the vicinity of Eulo. 4.3 The central-eastern Eromanga Basin Anomaly The aquifers overlying the granitic basement of the Manaroo Platform have been noted as having relatively high fluoride, in particular the Longreach town water bores have fluoride ranging between 5-7 mg/L. The occurrence of elevated fluoride over the Manaroo Platform itself can be rationalised with the model of deep groundwater circulation into the underlying granitic rocks. However not all of the elevated fluoride values occur over the platform area, or down flow gradient of the platform. A broadly Unear belt of elevated fluoride values from water bores tapping the Hooray Sandstone and Hutton Sandstones occurs to the east and northeast of the basement platform over the Galilee Basin. The reasons for this distribution of values is not yet understood, although the it is possible that the Tara Structure may exert some control on groundwater flow. 4.4 Southwestern Eromanga Basin Anomaly Muller (7) noted the existence of a NE - SW trending anomaly of elevated fluoride concentration in the Hooray Sandstone in the southwest Eromanga Basin. Muller (7) also pointed out that in the Hooray Sandstone of the southwest Eromanga Basin the elevated fluoride values tend to occur in zones of relatively lower salinity. The groundwater produced from the Jackson (mainly Hutton Sst.) and Naccowlah fields is enriched in fluoride and this is of concern in respect to the practice of disposal of by-product water to natural water courses. 4.5 The Surat Basin Anomaly Fluoride values close to the outcrop of the major sandstone aquifers of the Surat Basin are low (< 1 mg/L) and even in the central areas of the basin fluoride rarely exceeds 2 mg/L. However water bores in the vicinity of the Undulla Nose feature adjacent to Leichhardt - Burunga Fault system are significantly enriched in fluoride up to 5 mg/L. This trend of elevated fluoride appears to continue to the south, in a manner approximiately coincident with the Goondiwindi - Moonie Fault although the fluoride levels are lower to the south. This trend is exhibited in all of the five major sandstone aquifers adjacent to the fault system. Figure 4 illustrates this trend. The data from both water bores and petroleum well drill stem tests (DSTs) suggests that the source of the elevated fluoride in the groundwater is within the Surat Basin sequence, although probably in the lower units. Quarantotto (8) has noted that apatite inclusions within quartz grains in the Gubberamunda Sandstone are common. Apatite inclusions in quartz grains could be the source of much of the fluoride in the other aquifers.

176

8S - + ^ + +

•Umkilla Nose*

zroc

27*00:

a oo o ^

+-H+

OlfFIELD,'

\ of

^ ^ — X QUEENSLAND —^ NEW SOUTH WALES

Fluoride Ranges <2.01 mg/L 2.01 - 3.00 mg/L o 3 . 0 1 - 4 . 0 0 mg/L • > 4.00 mg/L

Fault after Exon(1976) ( D , U indicate relative movement down.up)

Fault from Seismic data after Exon(1976) Oil Field after Exon(1976) Great Artesian Basin

Distribution of fluoride in the Surat Basin, Queensland One possible implication of these fluoride anomalies may be that groundwater from the lower aquifers has migrated (or continues to migrate) up along areas of the Goondiwindi - Moonie / Leichhardt - Burunga Fauh system into the younger overlying aquifers. Initial work on groundwater temperatures in the Mooga Sandstone has not revealed a significant positive temperature anomaly coincident with the fluorine anomaly along the fault system, however more data is required to be conclusive. Trilinear plots of groundwater chemistry of the high fluoride water bores adjacent to the Goondiwindi Moonie/Leichhardt Burunga Fault; the Miles Town Bore which taps the Precipice Sandstone (san5)les for F > 10 mg/L); and the by-product water from the Moonie Oilfield (F up to 4.7 mg/L) indicate a similar sodium bicarbonate dominant water type although the salinity of the groundwater in the shallow aquifers is considerably less than for the deeper aquifers. This in itself does not confirm the hypothesis of intermixing of shallow and deep groundwater, however it does not exclude it as a possibility. 5.0 Discussion Chebotarev (9) studied the groundwater chemistry of more than 10,000 samplesfromAustralian water bores and concluded that groundwaters tend to geochemically evolve along flow pathsfromcalcium bicarbonate types towards sodium chloride types. Quarantotto (8 & 10) demonstrated that the generalised concept of geochemical evolution of groundwater from anionic bicarbonate type to chloride type was not valid for the major aquifer units of the Surat Basin and southem Eromanga Basin, however cationic evolution from calcium and magnesium types to sodium types was valid for the Surat Basin and southem Eromanga Basin. F i g u r e 4.

177

The recharge areas for the Great Artesian Basin are generally low in fluoride, and a possible influence on this is the geochemical evolution down flow path in the basin with a relative reduction in calcium and increase in sodium. Where there is very low calcium concentrations there is scope for h i ^ fluoride concentrations in solution. It is also possible that the groundwater in the deeper Surat Basin aquifers are enriched in fluoride relative to the groundwater in the shallow aquifers because of increased temperature and pressure which could enhance fluoride solubility as well as the increased residence time in the aquifers. The migration of groundwater from deeper Great Artesian Basin aquifers into younger formations via fault pathways has been suggested by Muller (7) who postulated the movement of water from the Hooray Sandstone and Cadna-Owie Formation up into the Wallumbilla Formation, Mackunda Formation, and Winton Formations in locations in the southwestern Eromanga Basin on the basis of bulk groundwater chemistry similarities as well as trace elements (ie. Boron). 6.0 Conclusions There is a relatively complex distribution of fluoride values within the aquifers of the Great Artesian Basin, however these may be explained in terms of both geochemistry and geology. The fluoride carrying capacity of groundwaters is largely determined by the solubility of fluorite (CaF2), and where geochemical changes result in low calcixun concentrations there is scope for dissolution of fluorine bearing minerals. Such a geochemical evolution has been documented in the aquifers of the Great Artesian Basin. Where sandstone aquifers directly overhe granitoid basement there is a strong association between high groundwater fluoride values attributable to the dissolution of fluorine bearing minerals from the basement. Other anomalies which are not associated with granitic basement probably reflect the upward migration of "geochemically mature" low calcium groundwater into younger aquifers. The linear fluoride anomaly in the eastem Surat Basin which coincides with the Leichhardt - Burunga and Goondiwindi - Moonie Fault systems effects all of the major aquifers and this is indicative of vertical migration of groundwater. It has been previously postulated that the hydrocarbons which accumulated in the Moonie stmcture (30 million barrels) migrated up along the Goondiwindi - Moonie Fauh. It may be possible to use subtle fluoride anomalies in younger aquifers above prospective reservoir sandstones as an exploration tool to delineate areas of upward migration of groundwater. 7.0 References 1. Habermehl, M.A., The Great Artesian Basin, Australia. Bureau of Mineral Resources Journal of Geology & Geophysics. 5,1980, 9-38. 2. Hem, J.D., Study and Interpretation of the Chemical Characteristics of Natural Water. 3rd. edn. US Geological Survey Water-Supply Paper 2254. US Govt. Printing Office. 1989. 3. Mason, B., Principles of Geochemistry. 3rd. ed. John Willey & Sons, New York, 1966. 4. Rankama, K & siiama T.G., Geochemistry. Chicago Univ. Press. Chicago, Illinois, 1950. 5. Esson, B. & Burban, B., First Annual & Final Report on A-P 3591M, "Toolebuc" and A-P 3592M, Warburton Creek", Qld Metals Corporation N.L., Qld Dept. of Minerals & Energy Open File Company Report No. 3384., 1984 6. Muller, P. J., Hydrogeology of the Southwestem Eromanga Basin, Queensland. Queensland Department of Mines Record 1989/16,1989. 7. Muller, P.J., Aspects of the hydrogeology of the southem Eromanga Basin, Queensland. In The Cooper and Eromanga Basins, ed. O'Neil, B.J. Proceedings of the Cooper & Eromanga Basins Conference, Adelaide, 1989, pp 493-505 8. C^arantotto, P., Hydrogeology of the southeastern Eromanga Basin, Queensland. Geological Survey of Queensland, Record 1986/38. unpublished, 1986. 9. Chebotarev, I.I., Metamorphism of natural water in the crust of weathering. Geochimica and Cosmochimica Acta, 8, 1955,22-48, 137-170,198-212. 10. Quarantotto, P., Hydrogeology of the Surat Basin, Queensland. Queensland Department of Mines Record 1989/26. unpublished, 1989.

178

REGIONAL CLAY AND CARBONATE MINERALISATION OF LATE PERMIAN COAL MEASURES CLEATS, BOWEN BASIN, AUSTRALIA

Basim S.M, Faraj'S Chris R. Fielding' and Ian D.R. Mackinnon' 1 Centre for Microscopy and Microanalysis, The University of Queensland, St Lucia, Q 4072, Brisbane, Australia 2 BHP Australia Coal, Riverside Centre, 123 Eagle St, Q 4001, Brisbane, Australia 3 Earth Sciences Dept, The University of Queensland, St Lucia, Q 4072, Brisbane, Australia Cleats are natural fractures found in coal seams which are perpendicular to bedding, and are similar to joints in other sedimentary rocks. These are the main conduits for gas and water movement in coal seams. Mineralisation of cleats and fractures may occur during the diagenetic history of the coal seams. This minerahsation impedes the movement of gases and water in coal seams and has been found to negatively influence methane producibility from coal. This study provides details of the composition and spatial distribution of cleat minerals in the Late Permian Rangal/Baralaba and German Creek Coal Measures in the Bowen Basin. The nature and origin of cleat mineralisation have been investigated using a variety of complementary analytical techniques. A total of nineteen authigenic mineral phases have been found in cleats and joints of the Rangal/Baralaba and the German Creek Coal Measures in the study area. These are illite, calcite, kaolinite, ankerite, quartz, chlorite, siderite, barite, pyrite, illite/smectite mixed-layer, strontianite, wollastonite, whewellite, halite, barium phosphate, bjarebyite, analcime, albite, fluorite. However, the overall cleat mineralogy in the studied area is rather simple. The most frequently occurring minerals belong to two groups, (i) clay wzTierafe-dominated by illite found mostly in face cleats and (ii) carbonateS'dommdXQ& by calcite and found mostly in butt cleats. The regional pervasive clay mineralisation occurred in thefirstcycle of subsidence and uplift of the Bowen Basin during the Triassic in response to the Hunter-Bowen Orogeny. K/Ar ages and stable isotope geochemistry of face cleat-fill illites indicate three phases of illite formation during the Triassic from deeply circulating meteoric waters as follows: (1) an early phase with ages clustered around 245 Ma B.R, (2) a second phase about 232 Ma B.P. and (3) the latest phase about 219 Ma B. P. Face cleat coals from Banana and Moura areas are dominated by illite, while coprecipitation of illite, chlorite and kaolinite assemblages is noted in face cleat coals from Bluff, Blackdown and Dawson River areas. This change in mineralogy reflects a change in the chemistry of the mineralising fluids during the Triassic. This change was most probably caused by depletion and reduction in activity of K, Ca, Mg and Fe cations within the mineralising fluids as they moved away from their source along the orogenic eastern basin margin towards shallower regions of the basin. The Bowen Basin area experienced a second cycle of subsidence that commenced in Early Jurassic with formation of the Surat Basin. During this second cycle of burial widespread carbonates, mainly calcite mineralisation in butt cleats and joints took place. Calcite fluid inclusions and stable isotopes indicate that calcite was precipitated from meteoric water at temperatures between 70 and <100°C. The widespread carbonate mineralisation in butt cleats and the near absence of carbonates in the face cleats is here attributed to permeability anisotropy, caused by a change in the direction of lateral conq)ressive stress during the Jurassic-Cretaceous time relative to that during the Triassic Hunter-Bowen Orogeny.

179

MESOZOIC SEDIMENTARY BASINS AND RESOURCES IN EASTERN AUSTRALIA - A REVIEW OF CURRENT UNDERSTANDING Christopher R, Fielding Department of Earth Sciences, The University of Queensland, Qld 4072, Australia Summary The Mesozoic sedimentary history of eastern AustraUa can be rationalised in terms of five discrete periods of tectonic activity: 1) Late Permian to Middle Triassic crustal contraction (Hunter-Bowen event), 2) Late Triassic extension, 3) earliest Jurassic to mid Cretaceous, mainly passive, thermal subsidence with localised extension, 4) mid Cretaceous contraction, and 5) Late Cretaceous denudation and extension. Despite many gaps in understanding, the stratigraphic record and contained mineral resources (particularly fossil fuels) of each phase reflect their tectonic context, such that stratigraphic models predicting the distribution and characteristics of mineral resources can be constructed. Introduction This paper is intended as a review of our current knowledge of Mesozoic sedimentary basins and their contained mineral wealth. I will attempt to highlight recent advances in understanding, and areas where data are still lacking. I will attempt to place the Mesozoic sedimentary mineral resources of Queensland and New South Wales in a geological context. Inevitably, this is a personal review, based on my own experience: it does not pretend to be exhaustive, nor to provide definitive answers to all issues. My intention, rather, is to focus attention on recent achievements in pure and applied research, and on those areas where future effort might be profitably devoted. The research of my colleagues, students and I over the past ten years has led us to the conclusion that the Mesozoic record of eastem Australia may be resolved in terms of five periods of tectonic activity: 1) Late Permian to Middle Triassic crustal contraction (Hunter-Bowen event), 2) Late Triassic extension, 3) earliest Jurassic to mid Cretaceous, mainly passive thermal subsidence with locaUsed extension, 4) mid Cretaceous contraction, and 5) Late Cretaceous denudation and extension. The basis for this subdivision, and the geological record of each phase are discussed below in order of decreasing age. The various events and their stratigraphic record are framed in terms of the absolute time scale of Jones

(1). Prelude - The Permo-Triassic Extinction Event The base of the Mesozoic era is marked by a severe, worldwide biotic extinction event that in terms of generic diversity is regarded as the most profound in Phanerozoic Earth history (Erwin, 2). The cause or causes of this crisis are under dispute, but many researchers have drawn attention to coincidence between the age of the boundary as established by studies of the Chinese type section (251 Ma: Ym et al, 3) and that of voluminous continental flood basalts erupted in Siberia. Across much of Gondwana a virtually synchronous change occurs at or near the Permo-Triassic boundary from grey, in many cases coal-rich sedimentary facies into primary-reddened alluvial strata (eg.. Day et al, 4; Barrett et al, 5; Casshyap & Tewari, 6; Webb & Fielding, 7; Veevers et al, 8). This change coincides roughly with a major change in floral composition from the Glossopteris flora of the Permian to the Triassic assemblage. While individual basins could be explained in terms of eg., tectonic processes, the great extent and synchroneity of this change argue for a climatic cause. The precise placement of the Permo-Triassic boundary within the preserved successions of eastem Australia remains controversial. Many workers have nominated the facies change at the top of the latest Pennian coal measures as the boundary, but this change has been shown to be diachronous, and Permian flora have been found in strata ascribed on lithological grounds to the Triassic (eg., Foster, 9; Dehghani, 10). Attempts to place the Permo-Triassic boundary from variations in oxygen isotope values within strata of the Sydney Basin appear ambiguous, and recent SHRIMP dates of 250 and 251 Ma have been derived (Roberts et al, 11) from samples of the Black Alley Shale in the Bowen Basin, a unit that underlies the latest Permian coal measures. Furthermore, the main faunal extinction in the Bowen and Sydney Basins occurred some time before the end of the Permian period, and appears to have been driven by occlusion of marine environments. Clearly, the stratigraphic position of the Permo-Triassic boundary in eastem Australia, and its implications, require considerable further research. Critical to resolution of boundary problems is the development of more robust biostratigraphic zonations for the Permian and Triassic successions. 1) Late Permian to Middle Triassic contraction (Hunter-Bowen event) From Kazanian times until the late Middle Triassic, the meridional Bowen-Gunnedah-Sydney Basin system of Queensland and New South Wales acted as a (?retroarc) foreland basin. To the west of this system, epicratonic basins (notably the Cooper and Galilee Basins) were also subsiding and, to the east, Triassic volcanic and sedimentary rocks are preserved in parts of the New England fold Belt (eg., Esk Trough in SE Queensland: see Campbell, this volume). The

180

foreland basins were controlled by a long-lived, probably pulsed, crustal contraction which saw progressive encroachment of thrust sheets westward across the basins and reversal of earlier extensional faults to form hi^-angle reverse faults. The structural style of the eastem basin margin varies north to south from that of a broad belt of thinskinned thrust deformation to a single, high-angle reverse fault (see Korsch & Totterdell, 12). It is evident from fault offsets, fold pattems and stratigraphic relationships that much of this deformation occurred during the latter stages of the Hunter-Bowen event, in the latest Middle to early Late Triassic. The Triassic stratigraphic record is one of continental, mainly alluvial and lacustrine environments in the Bowen and Gunnedah Basins, with some possible marine influences preserved in the uppermost units of the Sydney Basin (Wianamatta Group), and in strongly deformed rocks of the northem New England Fold Belt in Queensland (Brooweena Formation, Kin Kin Beds). Previous interpretations of marine lithofacies from the Triassic of the Bowen Basin (eg. Butcher, 13; Schroder, 14) have proved to be unrehable, and recent mineralogical investigations by Baker et al (15) have confirmed a non-marine source for early diagenetic siderite in Rewan Group strata which contain spinose acritarchs. Sediment supply to the Bowen Basin was principallyfromthe rising, volcanically active orogenic highland to the east (Kassan, 16; Fielding et al, 17). At various times, the palaeo-Austrahan craton to the west also acted as a sediment source, giving rise to quartzose alluvial facies preserved in the Rewan Group of the Roma Shelf (a prospective target for hydrocarbon exploration), and the basin-wide sheet sandstones of the Clematis Group/Showgrounds Sandstone (the latter another prolific hydrocarbon reservoir). The basin was at different times through the Early and Middle Triassic either underfilled or overfilled, in response to variations in the rate of sediment supply and/or subsidence. During periods of overfilling, sediment spilled overfromthe Bowen Basin into the Galilee and Cooper Basins to the west (see Kassan & Fielding, this volume). The Triassic record of the Gunnedah and Sydney Basins of NSW is similar to that of the Bowen Basin, and it is my contention that major stratigraphic units and petrographic/facies changes can be correlated over the entire length of the basin system. Sediment contriWonsfromthe Australian craton are perhaps better represented in the NSW basins (eg., Cowan, 18), although the occurrence and distribution of orogen-derived sediments have also been well-documented from these areas (eg., Jian & Ward, 19; Tadros, 20). Regional sediment dispersal pattems inferredfrompalaeocurrentdata (Cowan, 18; Tadros, 20; Fielding et al, 17) suggest the existence of a major, north to south (axial) drainage system during periods when sediment supply outstripped or matched subsidence. At certain times, this system degenerated into intemally draining alluvial systems or in the case of the Middle Triassic Snake Creek Mudstone (basal Moolayember Formation) of the Bowen Basin was drowned by a regional lacustrine transgression. Such disruptions may be tentatively attributed to increases in the rate of subsidence relative to that of sediment supply, and suggest a strong tectonic control on stratigraphic architecture in the Triassic of the Sydney-Gunnedah-Bowen Basin system. The age range of contractional deformation in Queensland is constrained by stratigraphic pattems within the Bowen Basin. Discrete pulses of thmst propagation are suggested by the occurrence of coarse clastic wedges shed into the eastem part of the basin. Successive units extend further westward into the basin, reflecting advancement of the thrustfront.Thefinalphase of deformation that closed the sedimentary record of the Bowen Basin occurred at about 232 Ma, and is interpreted by Holcombe et al (21) to be responsible for the development of the Connors Arch and possibly the Aubum Arch. Holcombe et al (21) further suggest that fluidflowwithin the thmst stmctures and sediments of the Bowen Basin at this time may have been responsible for gold mineralisation in basement rocks of the northem New England Fold Belt, and for widespread clay mineralisation in Bowen Basin sedimentary rocks (notably coals: see Faraj et al, 22, and this volume). During the final stages of the Hunter-Bowen event, up to 4 km of section were erodedfromthe Bowen Basin (eg., Beeston, 23; Korsch & Totterdell, 12). 2. Late Triassic extension A suite of mainly small, north-south elongate, partly fault-bounded sedimentary basins of Late Triassic (Camian to Rhaetian) age occur scattered across eastem AustraUa. The predominantly alluvial strata preserved within these basins are interbedded locally with lavas and volcaniclastic rocks of bimodal mafic and felsic composition. These basins formed apparently coevally with extensive siUcic granite intrusions and explosive volcanic complexes of rhyolite with minor mafic lava and ignimbrite (see Stephens, this volume). Many of these basins (eg., Tarong, Ipswich, Callide, Nymboida/Red Chff, Leigh Creek) host thick, localised coal bodies that are economically important. The Tarong Basin has a half-graben geometry (Williams, 24), as do a number of less well-described basins in the subsurface beneath the Great Artesian Basin (eg., Wiltshire, 25). Although the cross-sectional geometry of the other basins is not wellconstrained, recent research suggests that these basins are probably similar in geometry. On the basis of the crosssectional basin geometry, and the regional association between coarse alluvial deposits formed in tectonically active environments with extensive, coeval, bimodal volcanic sequences, Holcombe et al (21) have suggested an extensional tectonic setting for this suite of rocks. Facies analysis studies of Late Triassic coal measures in Queensland suggest that transverse (ie., east and west-directed) sediment disperal into formative basins was largely internal to individual basins, and often involved mass flows and even 181

rockfall processes on tectonically active, unstable slopes (eg., O'Brien et al, 26; Williams, 24; Jorgensen & Fielding, this volume). Basins in SE Queensland also preserve a record of more long-term, large-scale, axial fluvial systems that drained predominantly southward (eg., Falkner, 27; WilUams, 24; Jorgensen & Fielding, 28). Coal bodies are bestpreserved in the uppermost parts of grosslyfining-upwardmegasequences within these formations, and range up to 20m+ in thickness. Coals show evidence of accumulation in areas and at times of rapid subsidence. An active tectonic control on coal distribution and thickness is also suggested by the occurrence of diamictites, conglomerates and breccias of debris flow origin interbedded with coals in the Callide (Jorgensen & Fielding, this volume), Tarong, Ipswich and Red CUff Coal Measures. Potential may exist for the discovery of further such thick, localised coal bodies in eastem Australia, perhaps concealed by later Mesozoic and Cainozoic deposits. 3. Early Jurassic to mid Cretaceous passive thermal subsidence with localised extension The Jurassic and Cretaceous geological history of eastem Australia is marked by the development of extensive sedimentary basins, notably the Great Artesian Basin system. This basin system is of great economic importance as a host for coal, oil and gas, and water reserves. Preserved remnants of this intracratonic basin system cover much of eastem Australia, and the stratigraphy of other, now isolated areas of Jurassic/Cretaceous outcrop suggest that the Great Artesian Basin (GAB) system may have originally covered much of Queensland and northem New South Wales. The various component basins of the GAB (Eromanga, Surat, Carpentaria, Clarence-Moreton) preserve a thick Jurassic succession of mainly alluvial and lacustrine sedimentary rocks, which gave way in the Early Cretaceous to mainly marine strata. Individual formations are continuous across the component basins and in most cases between the basins, giving the impression of a true "layer-cake" stratigraphy. A number of depocentres have been recognised, some of which coincide with the axes of older sedimentary basins such as the Bowen Basin (Exon, 29). The stratigraphy within the GAB shows an altemation between 1) formations of quartzo-feldspathic composition (eg.. Precipice Sandstone: Martin, 30) which were derivedfi-omelevated basement terrains in central Australia, and 2) units of volcanic lithic and feldspathic composition (eg., Walloon Coal Measures: Fielding, 31; Yago & Fielding, this volume), which were derived from the east. The boundaries between some formations show a degree of petrographic mixing, indicating interference between drainage systems of opposite polarity. Such contacts have provided complex hydrocarbon sealing relationships in some producing oil/gas fields (eg.. Watts, 32). In at least one area, the same phenomenon has formed a sandstone with properties that render it a valuable dimension stone resource (the Helidon Sandstone). The volcanic lithic formations (notably the Jurassic Walloon Coal Measures, but also the Cretaceous Winton Formation) are host to significant coal and bentonite resources. The complex cross-sectional geometry and modest thickness of coal seams reflects the active nature of the alluvial plain depositional environment. The tectonic context of the GAB has been controversial. All stratigraphic evidence points towards a regime of passive, relatively even subsidence over an immense area. Jones & Veevers (33) interpreted the Clarence-Moreton and Surat Basins as part of a foreland basin system, developed to the west of a continental margin volcanic arc situated off the present coast of Queensland. In this model, floods of volcanic lithic detritus were shed periodically westward from the arc across eastem Australia. This interpretation has been accepted without question by several subsequent workers (eg.. Watts, 32; Hawlader, 34; Elliott, 35; Russell & Gumis, 36). Fielding (37), however, pointed out that there is no preserved record of a Jurassic arc within the AustraUan continent, and no unambiguous evidence as to its location or even existence. Recent research on Middle to Late Jurassic strata in the Clarence-Moreton Basin (Matheson, 38; Minifie, 39; Fielding, 31; Wells & O'Brien, 40; Yago & Fielding, this volume) has clarified a number of issues relevant to the problem. While it is evident that volcanic sediment was dispersed in an overall westerly direction (as suggested by Jones & Veevers, 33), recent studies show that quartzose, basement-derived sediments were also shed from sources to the south and east of the Clarence-Moreton Basin (eg.. Kangaroo Creek Sandstone). Peripheral basins along the present Queensland coast (eg., Namboxir, Maryborough) also preserve thick successions of quartzose sandstones. Furthermore, five lithostratigraphic members of the Walloon Coal Measures can be traced as continuous sheets across the Clarence-Moreton, Surat and Eromanga Basins (see Yago & Fielding, this volume). Such characteristics are inconsistent with formation in a retroarc foreland basin. The distribution of volcanic ash beds within the Walloon Coal Measures (now preserved as bentonite) suggests that volcanic sources may have been intra-basinal, and concentrated along a north-south elongate belt of possible coeval eruption centres, lavas and intrusives, of bimodal mafic and felsic composition, that can be recognised stretching from the Garrawilla Volcanics in NSW at least as far north as Tiaro in SE (^eensland. This belt coincides with a series of north-south elongate depocentres in NE NSW and SE Queensland. It is also noteworthy that accumulation of the Walloon Coal Measures coincided with emplacement of voluminous and extensive mafic intmsions in NSW, Tasmania and West Antarctica which are regarded as recording the early stages of rifling of the SE Australian continental margin. Research by Ewart, Holcombe, Stephens, Fielding and others suggests that the eastem margin of the Australian Plate has been broadly convergent since at least the mid-Palaeozoic, that the locus of subduction has been stepping progressively east over that time, and that several cycles of (broadly back-arc) cmstal extension have affected the present eastem 182

Australian continent. In this sense, all Mesozoic sedimentary basins in eastern Australia may have formed in a broadly convergent plate margin setting. The studies cited above and others, however, argue against a foreland origin for either the Clarence-Moreton and Surat Basins or for the various Jurassic and Cretaceous basins along the Queensland coast (eg., Nambour, Maryborough, Styx, etc.: see Fielding, 41), and suggest that the locus of any subduction must have been a considerable distance east of the present coastline. The sheet-like external and intemal geometry of formations within the GAB suggests formation during a period of predominantly passive, thermal subsidence. I suggest that this subsidence regime succeeded the widespread but modest. Late Triassic extension noted above, and furthermore suggest that the north-south elongate Jurassic and Early Cretaceous depocentres preserved within the Clarence-Moreton Basin and other Queensland coastal basins may record incipient or failed rifting episodes that preceded opening of the Coral Sea in the Late Cretaceous and early Tertiary (see also Fielding, 41). Early Cretaceous volcanic successions preserved within some of these basins have previously been attributed to crustal extension (eg., Ewart et aL, 42; Parianos et al, 43; Bryan, this volume). The extent to which the various coastal basins were connected to the GAB is as yet largely unresolved. 4. mid Cretaceous contraction The youngest preserved strata in the GAB and in the other basins mentioned above are early Late Cretaceous (top Cenomanian: c. 90 Ma). All Jurassic and Early Cretaceous basinfillsshow evidence of a mild contractional deformation that caused widespread reactivation and reversal of older faults, mainly broad, low amplitude folding, and local thrusting. This event is believed responsible for the formation of structures that have trapped oil and gas in the GAB. A midCretaceous age is favoured for the deformation, based on fault offsets and stratigraphic/unconformity relationships (eg., Hill, 44; Elliott, 35; Korsch & Totterdell, 12). While most authors beUeve this event to have been mild relative to the earlier, Hunter-Bowen event, some (eg., Elliott, 35) beUeve that the mid-Cretaceous event was responsible for thrust deformation of the eastem Bowen Basin. The latter interpretation is not bome out by stratigraphic relationships: for example, largely undeformed, flat-lying Late Triassic to Early Cretaceous strata (Callide Coal Measures, Precipice Sandstone, Razorback Beds, Stanwell Coal Measures) unconformably overlie intensely deformed Devonian to Middle Triassic rocks of the New England Fold Belt in the Rockhampton-Biloela region of central Queensland. Coals within the younger association display vitrinite reflectance values that indicate considerably lower rank (thermal maturity) than underlying strata. 5. Late Cretaceous denudation and extension There is little stratigraphic record of the Late Cretaceous in onshore eastem Australia. Hill (44), however, suggested that extensional grabens and half-grabens imaged by recent seismic surveys offshorefromthe Queensland coast may be filled by strata of Late Cretaceous and Tertiary age. Hill attributed these basins to rifting associated with opening of the Tasman and Coral Seas. It is possible that certain extensional basins preserved within coastal central Queensland may also preserve an as yet unrecognised Late Cretaceous record. Across much of eastem Australia, the Late Cretaceous was a period of denudation. Apatite Fission Track Analysis of the Bowen and Surat Basins by Raza et al (45) suggests regional uplift in the mid-Cretaceous as discussed above, and subsequent erosion of 1-1.5 km of strata. The extent to which tectonic events associated with formation of the Tasman and Coral Seas are recorded in the landscape has not yet been fully addressed, but may be considerable. Conclusion The Mesozoic history of sedimentary basin formation andfillingin eastem Australia can be rationalised in terms of an Early to early Late Triassic contractional orogeny (the Hunter-Bowen event), followed by a succession of mild crustal extension events and consequent periods of passive thermal subsidence. The initial, Late Triassic, Jurassic and Early Cretaceous failed rifts were succeeded by the extensional events that led to the formation of the Coral and Tasman Seas. The nature and distribution of Mesozoic fossil fuel resources to some extent reflect their geological (and particularly, tectonic) context. Studies of coal bodies preserved within the different sequences have now reached the stage that geological models with some predictive value can be generated. It is anticipated that such models may be of use in guiding future exploration, and in assisting mine planning and operations. Genetic stratigraphic studies will also assist in the delineation of possible stratigraphic traps for future oil and gas exploration, particularly as untested structural traps become more scarce. Indeed, the future development of all sedimentary mineral resources must benefitfroman improved understanding of their geological context. References 1. Jones, P.J., AGSO Phanerozoic Timescale 1995 and E?q)lanatory Notes, AGSO, 1996, 32pp. 2. Erwin, D.H., The Permian-Triassic extinction. Nature, 367,1994, pp.231-236. 3. Ym, H.F., Wu, S.B., Ding, M.H., Zhang, K.X., Tong, J.N. & Yang, F.Q., Suggestion of Meishan section as global stratotype section and point (GSSP) of Permo-Triassic boundary, Joumal of the China University of Geosciences 6,1995,pp.l-15. 183

4. 5.

6. 7.

9. 10. 11. 12.

13. 14. 15. 16. 17.

18. 19. 20. 21.

22.

23. 24. 25.

26. 27. 28. 29. 30.

Day, R.W., Whitaker, W.G., Murray, C.G., Wilson, LH. & Grimes, K.G., Queensland Geology, Geological Survey of Queensland Publication 383,1983,194pp. Barrett, PJ., Elliot, D.H. & Lindsay, J.H., The Beacon Supergroup (Devonian - Triassic) and Ferrar Group (Jurassic) in the Beardmore Glacier area, Antarctica, In: Turner, M.D. & Splettstoesser, J. (Eds.) Geology of the central Transantarctic Mountains, American Geophysical Union Antarctic Research Series 36,1986, pp.339-428. Casshyap, S.M. & Tewari, R.C., Depositional model and tectonic evolution of Gondwana basins. The Palaeobotanist 36,1987, pp.59-66. Webb, J.A. & Fielding, C.R., Permo-Triassic sedimentation within the Lambert Graben, northem Prince Charles Mountains, East Antarctica, In: Findlay, R.H., Unrug, R., Banks, M.R. & Veevers, J.J. (Eds.) Gondwana 8 Assembly, Evolution and Dispersal, Balkema, 1993, pp.357-369. Veevers, J.J., Conaghan, P.J. & Powell, C.McA., Eastem Australia, In: Veevers, J.J. &, Powell, C.McA (Eds.) Permian - Triassic Pangean Basins and Foldbelts along the Panthalassan Margin of Gondwanaland, Geological Society of America Memoir 184,1994, pp.11-171. Foster, C.B., Permian plant microfossils of the Blair Athol Coal Measures, Baralaba Coal Measures, and basal Rewan Formation of Queensland, Geological Survey of Queensland Publication 372, 244pp. Dehghani, M.H., Sedimentology, genetic stratigraphy and depositional envoronment of the Permo-Triassic succession in the southem Sydney Basin, Australia, PhD thesis. University of Wollongong, 1994. Roberts, J., Claoue-Long, J. & Foster, C.B., SHRIMP zircon dating of the Permian system of eastem Australia, Australian Joumal of Earth Sciences 43, 1996, in press. Korsch, R.J. & Totterdell, J.M., Structural events and deformational styles in the Bowen Basin, In: Follington, LL., Beeston, J.W. & Hamilton, L.H., (Eds.) Bowen Basin Symposium 1995 Proceedings, Geological Society of Australia Queensland Division, 1995, pp.27-35. Butcher, P.M., The Showgrounds Formation, its setting and seal in ATP145P, Austrahan Petroleum Exploration Association Joumal 24,1984, pp.336-357. Schroder, R., Exploration results and future activities in ATP 377P, Fairymount Oilfield, In: Queensland 1988 Exploration and Development, Petroleum Exploration Society ofAustralia Queensland Branch, 1988, pp.113-125. Baker, J.C., Kassan, J. & Hamilton, P.J., Early diagenetic siderite as an indicator of depositional environment in the Triassic Rewan Group, southem Bowen Basin, eastem Australia, Sedimentology 43,1996,77-88. Kassan, J., Basin analysis of the Triassic succession, Bowen Basin, Queensland, PhD thesis, University of Queensland, 1993. Fielding, C.R., Stephens, C.J., Kassan, J. & Holcombe, R.J., Revised palaeogeographic maps for the Bowen Basin, central Queensland, In: Follington, LL., Beeston, J.W. & Hamilton, L.H. (Eds.) Bowen Basin Symposium 1995 Proceedings, Geological Society of Australia Queensland Division, 1995, pp.7-15. Cowan, E.J., Longitudinal fluvial drainage pattems within a foreland basin-fill: Permo-Triassic Sydney Basin, AustraUa, Sedimentary Geology 85,1993, pp.557-577. Jian, F.X. & Ward, C.R., Triassic depositional episode. In: Tadros, NZ. (Ed.) The Gunnedah Basin, New South Wales, Geological Survey of New South Wales Memoir Geology 12,1993, pp.297-326. Tadros, N.Z., Lithostratigraphy, In: Tadros, N.Z. (Ed.) The Gunnedah Basin, New South Wales, Geological Survey of New South Wales Memoir Geology 12,1993, pp.95-133. Holcombe, R.J., Stephens, C.J., Fielding, C.R., Gust, D.A., Little, T.A., Sliwa, R., Kassan, J., McPhie, J.C. & Ewart, A., Tectonic evolution of the northem New England Fold belt: the Permo-Triassic Hunter-Bowen event, Australian Joumal of Earth Sciences, submitted. Faraj, B.S.M., Fielding, C.R. & Mackinnon, I.D.R., Cleat mineralisation of Upper Permian Baralaba/Rangal Coal Measures, Bowen Basin, Australia, In: Gayer, R. & Harris, I. (Eds.) Coalbed methane and coal geology. Geological Society of London Special Publication 109,1996, pp.151-164. Beeston, J.W., Coal rank variation in the Bowen Basin, Queensland, International Joumal of Coal Geology 6, 1986,pp.l63-180. Williams, B.M., The geology of the Meandu coal deposit and the central Tarong Basin, southeast Queensland, Bsc Honours thesis. University of Queensland, 1993. Wiltshire, M.J., Late Triassic and Early Jurassic sedimentation in the Great Artesian Basin, In: Moore, P.S. & Mount, T.J. (Eds.) Eromanga Basin Symposium, Summary Papers, Geological Society of Australia and Petroleum Exploration Society of Australia, 1982, pp.59-67. O'Brien, P.E., Wells, A.T., Smyth, M. & Russell, N.J., Alluvial fan and flood-basin sedimentation in the Triassic Red Cliff Coal Measures, New South Wales, Australian Joumal of Earth Sciences 35,1988,491-503. Falkner, A.J., Sedimentology of the Blackstone Fomiation, Ipswich Coal Measures, southeast Queensland, PhD thesis. University of Queensland, 1986. Jorgensen, P.J. & Fielding, C.R., Facies architecture of alluvial floodbasin deposits: three-dimensional data from the Upper Triassic Callide Coal Measures of east-central Queensland, Australia, Sedimentology 43,1996, in press. Exon, N.F., Geology of the Surat Basin in Queensland, Bureau of Mineral Resources Australia, Geology and Geophysics Bulletin 166,1976. Martin, K.R., Early Jurassic sedimentation in the Surat Basin, Australian Coal Geology 1,1980, pp.71-81. 184

31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

42.

43.

44. 45.

Fielding, C.R., The Middle Jurassic Walloon Coal Measures in the type area, the Rosewood-Walloon coalfield, SE Queensland, Australian Coal Geology 9,1993,4-16. Watts, K.J., The Hutton Sandstone-Birkhead Formation transition, ATP 269P(1), Eromanga Basin, Austrahan Petroleum Exploration Association Journal 27, 1987, pp.215-228. Jones, J.G. & Veevers, J.J., Mesozoic origins and antecedents of Australia's Eastern Highlands, Journal of the Geological Society of Australia 30, 1983, pp.305-322. Hawlader, H.M., Reservoir properties of some Surat Basin sandstones as a function of diagenetic clay-mineral assemblage, Queensland Government Mining Journal 91,1990, pp.180-189. Elhott, L.G., Post-Carboniferous tectonic evolution of eastem Austraha, Australian Petroleum Exploration Association Joumal 33, 1993, pp.215-236. Russell, M. & Gumis, M., The planform of epeirogeny: vertical motions of Australia during the Mesozoic, Basin Research 6,1994, pp.63-76. Fielding, C.R., The geological setting of Queensland coal. In: Queensland Coal Symposium, Australasian Institute of Mining and Metallurgy, 1991, pp.97-103. Matheson, S.G., Exploration and coal resources of the Moreton Basin, Queensland, In: Queensland Coal Symposium, Australasian Institute of mining and Metallurgy, 1991, pp. 105-116. Minifie, S., Sedimentology of the Kangaroo Creek Sandstone, Pillar Valley, south Clarence-Moreton Basin, BSc Honours thesis, Queensland University of Technology, 1992. Wells, A.T. & O'Brien RE. (Eds.), Geology and petroleum potential of the Clarence-Moreton Basin, New South Wales and Queensland, Austrahan Geological Survey Organisation Bulletin 241,1994. Fielding, C.R., A review of Cretaceous coal-bearing sequences in Austraha, In: McCabe, P.J. & Parrish, J.T. (Eds.) Controls on the distribution and quality of Cretaceous coals. Geological Society of America Special Paper 267, 1992,pp.303-324. Ewart, A., Schon, R.W. & Chappell, B.W., The Cretaceous volcanic-plutonic province of the central Queensland (Australia) coast - a rift-related "calc-alkaline" province. Transactions of the Royal Society of Edinburgh 83, 1992, 327-346. Parianos, J., Bryan, S., Ewart, A. & Schon, R.W., Early Cretaceous rift volcanics of the central Queensland coast, In: Flood, P.G. & Aitchison, J.C. (Eds.) New England Orogen, Eastem Australia, University of New England, 1993,pp.655-663. Hill, P.J., Maryborough and Capricom Basins - new geophysical data. In: Draper, J. (Ed.) Queensland 1991 Exploration and Development, Petroleum Exploration Society of Australia Queensland Branch, 1991, pp. 70-82. Raza, A., Hill, K.C. & Korsch, R.J., Mid-Cretaceous regional uplift and denudation of the Bowen-Surat Basins, Queensland and its relation to Tasman Sea rifting. In: Follington, I.L., Beeston, J.W. & Hamilton, L.H. (Eds.) Bowen Basin Symposium 1995 Proceedings, Geological Society of Australia Queensland Division, 1995, supplementary paper (8pp).

185

AN APPLICATION OF GROUNDWATER GEOCHEMISTRY TO THE DETECTION OF PROSPECTIVE BASEMENT BENEATH MESOZOIC COVER IN NORTH QUEENSLAND

Angela Giblin, Principal Research Scientist CSIRO Division ofExploration and Mining. NORTH RYDE NSW Summary Crystalline basement, covered by Mesozoic sediments adjacent to the Eastem Mt. Isa and Georgetown Blocks, are important targets for mineral exploration. Groundwaters are effective indicators of the composition of concealed basement, due to their mobility and chemical reactivity. Comparative concentrations of major constituents contrasted groundwaters from the deep Mesozoic sediments (NaHCO3) with those in the crystalline basement (Ca-Mg-S04). Between these extremes, basin-groundwater compositions reflected mixing of the two. The degree of mixing, depicted as a Principal Component score calculated from the major constituent concentrations, reflects the degree of influence of the basement at each sample location. Where, within the basin, this influence is significant, the depth of basement was interpreted as sufficiently shallow to warrant exploration for ore deposits. Depletion of Ca in deep basin waters allowed unusually high concentrations of fluorine, molybdenum and tungsten. Other trace element indicators of concealed lithologies, including possible ores such as gold, were identified. Introduction Although ore deposits may not be common in the Mesozoic of the Australian mainland, exploration companies increasingly have to operate in terrain where targeted ore deposits are concealed beneath in situ or transported Mesozoic cover. Exploration techniques that can detect geochemical signals through this cover have thus become important. Among these, geochemical exploration techniques using groundwaters as the sample media, have particular potential. Groundwaters interact chemically with aquifer lithologies. They also move with varying degrees of freedom through subsurface zones. Groundwater chemistry therefore provides a geochemical reflection of potentially large volumes of subsurface rocks. A single groundwater sample can reflect the geochemistry of substantially larger rock volumes than would a single conventional solid rock or mineral sample. Modem analytical techniques provide low detection limits for a wide range of chemical elements. Groundwater elemental abundances, speciation and distribution pattems provide a series of geochemical indicators useful for identifying aquifer lithologies and proximity to ore bodies. Groundwater sampling and analyses provide a rapid and cost effective method of assessing the exploration potential for large areas, that hitherto have received little systematic exploration. Study Area This paper presents results from ongoing investigations at the CSIRO Division of E?q)loration and Mining aimed at determining whether groundwater geochemistry can be used to identify "blind" ore deposits, overlain by Mesozoic cover in the vicinity of the Eastem Mt. Isa Block The investigation has used groundwater data from several economically interesting regions of the Eastem Mt. Isa Block, the western side of the Georgetown Inlier and the intervening Mesozoicsediment covered region of the Carpentaria and Eromanga Basins (Fig. 1). The sources of groundwater samples included exploration drill holes and stock bores, some of which were artesian. Within the Eastem Mt. Isa Block, groundwater sample locations included zones around mineralized areas at Dugald River, Lorena, Monakoff, Eloise and Osbome. Most samples were acquired during the last five to ten years through collaborative research projects with exploration companies. Other data has come from Queensland Government compilations of bore water chemistries. Important questions addressed by the research include whether concealed deposits sensed by groundwaters, would be within an economically feasible depth for mining. The basement rocks of the Mt. Isa and Georgetown Blocks have been studied extensively and provide a model for potential extensions beneath the intervening basin sediments. The part of the study area covered by Mesozoic sediments, is a large region of relatively flat, featureless terrain, constituting an ideal test site for groundwater geochemical exploration techniques. From an exploration viewpoint, basement beneath this part of the basin is an attractive target because: • The depth of cover sequences is generally less than 300m. • Stock water bores provide a reasonable distribution of groundwater sampling sites, a large percentage of which intersect the basement. • The region includes geophysically interpreted basement highs such as the Fort Bowen Ridge and the Euroka Arch, considered to be the boundary between the Carpentaria Basin in the north and the Eromanga Basin in the south (Fig 1). • The region includes areas of outcropping basement of the Mt. Isa and Georgetown Blocks, outcrop on the Fort Bowen Ridge and two areas of Precambrian and/or Palaeozoic basement outcrop in the south east. 186

• Airborne magnetic data for the area reveals major structural features which could have acted as channel ways for mineralizing fluids. • Gravity data and water bore logs indicate that portions of the area may not be as deeply buried as previously believed. 8000000-r

7950000+ + +

C A R P E N W l A BASIN K

7900000-

++•

7850000^-

7750000-

(UT+ttv A" ^

^^^

* ^ ++ +* *+ J-

* V **

7700000-

EROMANGA BASIN 7650000-

7600000-

7550000400000

450000

500000

550000

600000

650000

700000

AMG Easting

Figure 1. Location map showing groundwater sample sites ('+'), eastem margin of the Mt. Isa Block, westem margin of the Georgetown Block and the Euroka Arch separating the Carpentaria and Eromanga Basins. Sample Collection Most groundwaters were sampled from stock bores, usually via pump outflow pipes. Groundwater samples from exploration drill holes and from bores without pumps were bailed from 5m below the water table, a depth at which changes to the water chemistry due to contact with air, would be minimal. Three samples were collected at each site, a 1 litre sample for gold, a 500 mLsample for other analyses and a small sub-sample forfieldmeasurements. The sample for gold analysis was field treated with lime, cyanide and charcoal to preconcentrate the gold to bring it into the range suitable for neutron activation analysis. Measurements of pH, Eh, conductivity and reduced iron (Fell) were made on the sub-sample immediately after each sample was collected. Field measurements provide some immediately usefiil exploration indicators. For example, acid pH and detectable Fell are immediate field indications that sulfides are being oxidized in the groundwater flow path. This process constitutes the principal mechanism by which traces of sulfideassociated base and precious metals enter the groundwater. Conductivity also provides an immediate field indication of groundwater salinity that indicates whether a series of samples is derivedall derive from possibly the same, or clearly different aquifers. Abundances of Al, Si, Fe, Mn, Cu, Li, P, Ti, B, Ba, Be, Sc, V, Sr, Cu, Pb, U, La, Y, Yb, Co, Cr, Cd, Ni, Zn, Mo, Tl, Th, Rb, Cs, W, Ga, Ge, Zr, Sb, Bi, CI, SO4, total carbonate (TCO3), F with an ion selective electrode andAs . , and F were determined in the laboratory using standard modem analytical procedures, e.g. spectrometry and ion chromatography. Major Elements Variations in groundwater concentrations of major element species Mg^"", Na% CI", 804^' TCO3) ^^^^^t variations in aquifer lithologies. In most studies this has been adequately illustrated by Schoeller plots that depict absolute and relative concentrations of major groundwater constituents. Figure 2 shows examples which illustrate the contrast between the major element composition of groundwatersfromtheEastem Mt. Isa Block and those from locations within deep Mesozoic sediments. The major ions in waters in deep parts of the basin were predominantly Na-HC03, whereas groundwatersfromthe adjacent Mt. Isa and Georgetown Blocks also contained signifcant concentrations of Ca-Mg-S04. Crystalline basement rocks are apparently better sources of Ca-Mg-S04 dominant groundwaters than are deep basin 187

sediments. The examples in Fig. 2 represent end members of a series of groundwater compositions that would result from mixing varying proportions of watersfromEastem Mt. Isa basement rocks, with watersfromMesozoic sediments. Eromanga Basin

Eastem Mt Isa Block

1000 100 10 mE/L

mBL

0.1

1000 100 10 1

0.1

0.01 0.001 Ca

Na+K

S04

TC03

0.01 0.001

Na+K

S04

HC03

Figure 2 Schoeller plots depicting the major element composition of representative groundwatersfromtheEastem Mt. Isa Block and thosefromlocations within deep Mesozoic sediments. Schoeller Plots are a means of depicting the major solute composition of a groundwater by plotting logarithmic concentrations of major solutes (in miUiequivalents/litre (me/L)) in a water, in the order (Na^+K"^), CI", SO4'- and TCO3. The slope of each line joining the points represents the concentration ratios Ca/Mg, Mg/(Na+K), (Na+K)/Cl, CI/SO4, SO4/TCO3 respectively, and the resultant shape that derivesfromthese ratios constitutes a signature for the aquifer in terms of the major solute content of the water. A more sensitive means for demonstrating this series results from the use of a multivariate statistical procedure. The procedure involved compressing the informationfromall major element variables into single variables for each sample, using principal component (PC) calculations of PC scores. The spread in scoresfroma particular PC directly reflected the observed major element variation. Calculated loadings for this PC indicated that low PC scores were strongly influenced by Na and HCO3 and high scores by Ca, Mg and SO4. Relating the PC scores to sample locations indicated that groundwaters from the deep parts of the basins had the lowest PC scores, whereas groundwaters from mineralized locations on the Eastem Mt. Isa Block had the highest. Between these extremes, values of the PC score could be interpreted in terms of the relative contributions to the composition of each groundwaterfromthe Mesozoic cover and the crystalline basement. Assuming that the contribution from the cover was from sediments overlying crystalline basement, the PC score could be regarded as an indicator of the depth of the Mesozoic cover. The differences between basement and sediment hosted groundwaters directly reflects the fimdamental geochemical differences in the two rock types. Due to chemical weathering associated with erosion, transport and deposition, Mesozoic sediments are likely to be considerably depleted in Ca and Mg compared to crystalline basement. A consequence of this depletion is that carbonate concentrations in associated groundwaters would increase, as Ca controls groundwater carbonate levels via the formation of calcium carbonate minerals. Present day groundwaters that contact basement beneath very deep Mesozoic cover would dissolve Ca and Mg. However such deeply sourced Ca and Mg would be lost as carbonates within the longflow-pathsbetween the source and surface sample points. Conversely, Ca and Mg derived from shallow crystalline rock sources would deplete the available groundwater carbonate and accumulate to the concentrations observed in Eastem Mt. Isa Block groundwaters.

188

SOOOOOOi-

7550000 400000

450000

500000

550000

600000

650000

700000

AMG Easting

Figure 3. Contoured image of Principal Component scores that expresses the variation in concentrations of Ca, Mg, Na, SO4 and TCO3 for all groundwater samples. Increasing depth of shading depicts the transition from groundwaters strongly influenced by Ca, Mg and SO4 to those dominated by Na and HCO3. To visualize the lithological variation implied by the variation in PC scores, an image was created using a gray series to represent the varying values of the PC score across the study area (Fig 3). The image was derived by calculating a contour grid (using Kriging) from values of the PC score derivedfrommeasured groundwater values. Because sample locations are unevenly distributed across the area covered by the image, its veracity is variable. Calculated contour values are represented in the image by white and pale grays, where the PC score is more strongly influenced by Ca, Mg and SO4 than by Na and HCO3, to dark grays and black where the opposite applies. The image illustrates a strong contrast between groundwater located within the two basement blocks, and those of intervening basin sediments. Basement groundwater signatures, apparent within the basin, may reflect the presence of shallow basement concealed by basin sediments. In some cases, such as the Fort Bowen Ridge these accord with sub-surface structures or basement variation previously identified, in others they suggest concealed features not otherwise recorded.

189

8000000-

7950000-

7900000-

7850000-

7800000-

o 775000077000007650000-

7600000-

7550000-

450000

500000

550000

600000

650000

AMG Easting

Figure 4. An illustration of the main features of the image variation together with locations of some Eastem Mt. Isa Block ore deposits and some major basement features including the Euroka Arch, the Fort Bowen Ridge and the Canobie Depression. In Fig. 4 the PC imiage agrees with some of the implications that each of the depicted basement feature have with regard to basement depth. The groundwater image indicates a linear transition zone between Na-HC03 waters and those with more Ca, Mg and SO4 along what is probably the northem edge of the Euroka Arch; shallow ridges and platforms of older sedimentary, metacnorphic or igneous rocks. The same transition zone changes direction, following the south easterly direction of the Fort Bowen Ridge, after its intersection with the Euroka Arch. Hence, north and west of this transition zone the image reflects deeper basin sediments. However, south and east the image implies widespread shallow sediments, with significant influence of basement rocks on groundwater compositions. Altematively, it could imply a marked change in the chemical composition of cover sequences for this region compared with other Mesozoic sediments in the study area. An additional small locality of shallow basement is suggested by the image in the north west, indicated in Fig. 4 by a question mark. Overall, the groundwater image, which maps relative influences of Na-HC03 waters compared with Ca-Mg-S04 waters, illustrates a broad reflection of basement topography, the structure of which accords with some major basement features. Of special note is that this image is created using only the most commonly recorded groundwater data. Trace Elements Some trace solute concentrations reflect lithological variations, whilst others can indicate the presence of potentially economic mineralization. Examples of each type are shown in Figures 5 - 7 which show the variation across the study area of groundwater concentrations of gold, arsenic(mineralization) andfluorine(lithology),plotted as contoured images. Similar images could be drawn for all the trace elements that were determined. Enhanced groundwater concentrations of trace elements that are indicative of mineralization mi^t indicate actual, or down-flow locations of precious and/or base metal ores. The value of groundwater trace element enrichments to exploration is increased if they extend beyond the immediate locations of the mineralization, thus enlarging the exploration target. These enrichments include those directly derived from groundwater interaction with a point source of mineralization, whether the water is subsequently sampled close to, or away from the deposit, and those derived from groundwater interaction with a weathering or alteration envelope that may surround and geochemically reflect the mineralization. The value of gold as an exploration indicator has been establishedfrommany groundwater studies across Australia. Gold is not detectable (< 1 ng/L) in the large majority of collected samples, and any detectable gold, is a very good 190

groundwater indicator to the presence of gold in aquifers, whether as an. This may be as gold in its own right, or as a component of other mineralization. In this study (Fig. 5), gold was encountered in groundwaters associated with mineralization on both the Mt. Isa and Georgetown Blocks. A modest gold halo is implied also in Fig. 5, in sediments close to each block. However, from a greenfields e}q)loration viewpoint, more interesting are the possibilities of "blind targets" associated with zones with detectable gold in both east and central regions of the basin.

GokingO.

400000

450000

500000 550000 AMGEasbng

600000

650000

Figure 5. Variation across the study area of groundwater concentrations of gold in ng/L. In geochemical coloration, an association between anomalous arsenic and anomalous gold concentrations can be indicative of primary or sulfide associated gold deposits. This concept has some appUcation on the Mt. Isa Block. However the large zone of elevated arsenic in groundwaters in the western part of the basins (Fig 6), probably corresponds with the presence of the Toolebuc Shale, which includes large reserves of low grade oil shale that outcrop near Julia Creek. Pyritic sulfides associated with the oil shale (Ozimic, 1986) could result in a high background of As in groundwaters in this area. In the northem part of the large zone, the hi^er As values may be related to mineralization, as could the zone near the Georgetown Block.

191

Arsenic ug^

7550000400000

450000

500000

550000

600000

650000

AMG Easting

Figure 6. Variation across the study area of groundwater concentrations of arsenic in ug/L. Fluorine concentrations in many of the samples in this study were exceptionally high, ranging up to 36 ppm. It is rare to encounter fluorine in natural groundwaters in excess of 3 ppm. Figure 7 demonstrates that the higher fluorine concentrations are predominantly in the northem part of the Carpentaria Basin. Low fluorine concentrations in groundwaters correlate with groundwaters containing significant concentrations of Ca and/or Mg. An inverse relationship between fluorine and calcium concentrations in groundwaters results from the suppression of fluorine concentrations by the formation of insoluble calcium fluoride or fluorite. Similarly, groundwaters that are low in Ca contain elevated concentrations of carbonate. Not all regions of very low Ca and/or Mg had groundwaters with the high fluorine concentrations; although most exceeded the 3ppm level. The higher fluorine in the northem region may be related to leaching of felsic igneous rocks of the Georgetown Block. Such rocks are enriched in fluorine, which would have accumulated in the basin, and hence in the low calcium groundwaters. The distribution of tungsten and molybdenum concentrations in northem groundwaters also fit the same pattem for similar reasons, with a few exceptions in the case of molybdenum. In addition, Georgetown felsic igneous rocks could explain a zone of groundwaters enriched in lithium that occurs adjacent to the Georgetown block. The distribution of sample sites where groundwaters have higher mbidium (and cesium) concentrations are interesting in that they cluster into two areas. This suggests that the potassium content of aquifer lithologies in these areas is higher than in other parts of the study area. The area where rubidium values are highest coincides with the westem edge of the Georgetown Block. A more southern area of higher rubidium may relate to the Euroka Arch. The significance of this is unknown except that both Mt. Isa and Hilton groundwaters are regionally enhanced with respect to rubidium, as are groundwaters in other mineralized parts of the Mt. Isa Block. The few high concentrations of manganese (another trace element enhanced in mineralized Mt. Isa groundwaters) coincide with samples from the north-east group of higher rubidium

192

7950000-

R u o r i n e rngfL

400000

450000

500000 550000 AMG Easting

600000

650000

Figure 7. Variation across the study area of groundwater concentrations offluorinein mg/L. Higher values of boron cluster in the north-west of the study area. To some extent this reflects higher salinity of groundwaters and correlates with higher concentrations of chloride. However, a region of samples in the east with higher concentrations of chloride shows no comparable highs in boron. Since boron is an indicator of marine sourced sediments, elevated boron in the north west is consistent with Cretaceous shallow-marine sediments on most, but not the eastem side, of the study area (Ozimic, 1986) Two water-mineral equilibrium programs THERMODATA (Tumbull and Wadsley, 1984) and MINTEQA2 (Allison et al, 1991), have been used to test the compositions of all waters in the study for saturation with respect to minerals, such as carbonates, sulfates, oxides and chlorides of major cations and trace elements. Locations of groundwater samples that were found to have common suites of saturated minerals, might indicate concealed lithological boundaries that may have exploration significance. All samples were saturated with one or more iron oxides, and most with an aluminium oxide and a fluorine-carbonate apatite. Groundwaters at many locations throughout the study area were also saturated with calcite which confirms the role of calcium carbonate in depleting Ca concentrations in groundwatersfiromdeep basin sequences. In addition to calcite, groundwaters in the north and the west were also saturated with dolomite, and at some sites, magnesite. The source of the additional Mg indicated by these latter two minerals may relate to local basement variation, or to relative enrichments of Mg in the local sediments. Sulfate minerals such as jarosite, a weathering product of base metal sulfides, may also be indicators of specific types of lithology including the presence of sulfide minerals. In two zones a few groundwaters were saturated with jarosite; one was a disused mine shaft confirming the association of sulfides and jarosite saturated groundwaters. Conclusions Groundwater sampling and analyses provided a rapid and cost effective method of assessing the exploration potential for a large area, that hitherto has received little systematic exploration. Groundwater elemental abundances, speciation and distribution pattems provided a series of geochemical indicators usefiil for identifying aquifer lithologies and proximity to ore bodies. Variation in concentrations of major dissolved components, expressed as Principal Component scores, formed an image that clearly distinguished sample locations within crystalline basement of the Eastem Mt. Isa and Georgetown Blocks, firom those in intervening Mesozoic Basin sediments. Between these extremes, basin-groundwater compositions reflected mixing of the two. Within the region covered by Mesozoic sediments, image variations that could be interpreted as depths of cover sequences, were also evident. Shallow regions so interpreted included previously documented basement highs, Euroka Arch and the Fort Bowen Ridge. Similarly, the Canobie Depression was interpreted as deep sediments by the image. This 193

image, derived from the most commonly recorded groundwater data, could provide valuable preliminary information to exploration companies who would wish to avoid exploration expense at locations where any basement hosted ore deposit would be too deep to mine. Groundwaters in deep Mesozoic sediments had high concentrations of element species, F, Mo and W, that normally are controlled by groundwater Ca concentrations. Unusually high levels of these elements can then also indicate deeply covered basement. In most instances, sites within the Mesozoic-covered region where groundwaters contained detectable concentrations of trace elements with potential as economic mineral indicators such as gold, were located in zones, interpreted as shallow. Acknowledgments This study was possible because of the encouragement, financial and practical support of several exploration companies, in particular North Ltd. References Allison, J.D., Brown D.S. and Novo-Gradac, K.J. MinteqA2/ProdefA2, A geochemical assessment model for environmental systems: Version 3.0 User's Manual. U.S. EPA Contract 600/3-01/021 1991. Ozimic, S. The geology and petrophysics of the Toolebuc Formation and its time equivalents, Eromanga and Carpentaria Basins. Geological Society of Australia Special Publication No 12,1986, ppl 10-137. Tumbull A. G. & Wadsley M. W.. Thermodynamic modelling of metallurgical processes by the CSIRO-SGTE THERMODATA system. In Proceeding of Institute of Mining and Metallurgy Symposium, 1984, pp 79-85. Australasian Institute of Mining and Metallurgy, Melboume.

194

IMAGING THE THERMOTECTONIC EVOLUTION OF EASTERN AUSTRALIA DURING THE MESOZOIC FROM FISSION TRACK DATING OF APATITES Andrew Gleadow', Barry Kohn', Kerry Gallagher^ and Simon Cox^ Australian Geodynamics Cooperative Research Centre, 1 School of Earth Sciences, La Trobe University, Bmdoora, Victoria 3083, Australia, 2 Department of Geology, Imperial College, London SW7 2BP, England. 3 CSIRO Exploration and Mining Nedlands, WA 6009, Australia Summary Apatite fission track thermochronology is an unrivalled tool for reconstructing the low temperature thermal and tectonic evolution of the continental crust margins. Analysis of samples representing a range of depths within the crust, such as a suite from a deep drill hole or from outcrops over a range of surface elevations, can be used to document the variation of fission track ages and distribution of fission track lengths within the upper crust. These parameters are primarily controlled by cooling which may be initiated by broad earth movements and consequent denudation at the Earth's surface and/or by changes in the thermal regime. Using numerical forward modelling procedures these parameters can be matched with time-temperature paths, which enable the thermal and tectonic processes to be mapped out in considerable detail. An approach is described whereby large regional fission track data sets are modelled simultaneously and the resulting time-temperature solutions visualised as a series of time-slice images which depict the cooling history of present day surface rocks during their passage through the upper crust. In addition, the data can be extended and combined with other data sets to image the denudation history and the evolution of palaeotopography. The methodology is demonstrated for a large data set from SE Australia which shows, in particular, how important Mesozoic tectonic events affecting the region can be visuahsed in time and space. Introduction An understanding of the thermal history of rocks in the upper part of the Earth's crust provides one of the best methods of constraining the thermal and tectonic processes that have controlled the evolution of such environments, and the resulting pattems of denudation that have occurred at the surface in response to these. The most important method available today for reconstructing such histories in the region of the upper 3-5 kilometres of the crust (below about 120°C) is apatite fission track analysis. Like other thermochronological methods, fission track analysis relies on a geological dating technique where the retention of the products, or effects, of radioactive decay is sensitive to elevated temperatures. Monitoring the degree to which a particular dating system has remained closed enables the history of exposure to elevated temperatures in the geological environment to be quantified. In most cases such thermochronometers give rise to apparent ages which reflect regional pattems of cooling, rather than original formation ages of the rocks involved. Mostly the apparent ages obtained are mixed ages which reflect more than one component of the cooling history and only in relatively unusual circumstances do they directly date a particular cooling episode. Apatite fission track studies have been carried out over the past 20 years over progressively larger areas of eastem Australia from Tasmania to southem Queensland, from southeastem South AustraUa to the east coast and from sea-level to the summit of Mt Kosciusko (1-10). Although the apparent fission track ages in this broad regional pattem range from mid-Palaeozoic to Tertiary, they only rarely reflect the formation ages of the host rocks sampled, which were mostly of Palaeozoic age. The most obvious features of the pattem of apparent fission track ages have resultedfi-omthe tectonic and thermal effects of continental rifting and basin formation in eastem Australia during the Mesozoic. The pattem of fission track ages in southeastem Australia shows broad regional variations which can be contoured as shown in Figure 1, which is updated from Dumitru et al (4). The existence of such regional pattems, and the development of methods for quantitatively reconstmcting the thermal histories underlying them, suggests that much more powerful approaches to visualising the thermotectonic evolution of this area are now possible. In this paper we will explore the development of imaging techniques which not only provide new ways of examining and interpreting the fission track data but for the first time also allow them to be combined with other regional data sets.

195

Sample Locality Fission Track -300- isochrons (Ma)

Adelaide

100

200

Murray Basin

350-

•••I

iliii«iliiiiiiiii

if? Hobart •

Figure 1. The distribution of apatite fission track ages in southeastern Australia (modified after Dumitru et al., (4)). Sample point are shown as black dots and contours indicate the apparent ages in millions of years. Virtually all of the samples were obtainedfromrocks of Palaeozoic age. Fission track analysis in apatite Fission track dating relies on the accumulation of radiation damage in minerals from the spontaneous nuclear fission of ^U over geological time. The damage occurs as linear zones, calledfissiontracks, of highly damaged material in uraniumbearing crystals, such as apatite and zircon. These tracks can be enlarged by a simple chemical etching procedure on a polished surface of the mineral so that they can be observed and measured by optical microscopy. The number of tracks that have accumulated in a mineral gives a measure of the time over which they have been quantitatively retained, ie. a measure of geological age . In addition thefissiontracks have some important properties which form the basis of their use in thermochronology. When exposed to elevated temperatures the radiation damage making up the tracks is progressively repaired, or annealed, over a temperature interval which is characteristic for each particular mineral. In the case of apatite, this temperature interval occurs below ~120°C for geological heating times of the order of 10^ years or more. Some annealing in apatite occurs even down to ambient surface temperatures but below about 60°C this is relatively insignificant. For convenience the temperature interval between about 60° and 120°C is often referred to as the apatite fission track annealing zone. Another important property of thefissiontracks is their length. For the most part,fissiontrack annealing takes place by a progressive shortening of the tracksfromtheir ends (11), and all the tracks in a particular mineral have very nearly the same length when they arefirstproduced. During annealing each track will shorten to a length which is characteristic of the highest temperatures to which it has been exposed so thatfissiontrack lengths can be used to give a measure of the amount of annealing that has occurred. Each individual track is added by radioactive decay at a different time, and thus experiences a different fraction of the thermal history. The distribution of track lengths in a sample therefore gives a record of the history of cooling througfi thefissiontrack anneaUng zone (12). Fission track annealing is a kinetic process that can be studied at higher temperatures for much shorter times (hours to years) in the laboratory using controlled heating experiments on fresh, neutron-induced 235U fission tracks. These are essentially identical to the natural 23 8U tracks used in dating. Such laboratory annealing studies have given rise to quantitative models offissiontrack annealing in apatite, the most widely used of which is that of Laslett et al. (13). These 196

annealing models can then be used to calculate the fission track age and track length distribution that would result from any given thermal history on a geological time scale. In this way plausible thermal histories can be tested against actual fission track measurements. More recently, Gallagher (14) has automated this procedure to give a quasi-reversed modelling approach which combines a Monte-Carlo simulation of numerous possible thermal histories with statistical testing of the outcome against the observed fission track measurements. A genetic algorithm is also used to provide rapid convergence to an acceptable fit. Such approaches tofissiontrack modelling have mostly been applied to simulating the thermal history for one sample at a time, but increasingly, there is a demand for simultaneously modelling the thermal history of arrays of samples. A complication to the study of apatite annealing is that small variations in composition (particularly chlorine) between apatites cause shght differences in their annealing properties (15). This means that where there is a range of compositions between individual apatite grains there will be a corresponding spread in their annealing behaviour and the apparent ages that they will exhibit, particularly after a more complex thermal history (16). However, in most cases this is a secondary effect which does not unduly compromise the ability to model apatite annealing behaviour satisfactorily. It is likely, however, that future modelling approaches will incorporate this factor e^licitly. Modelling fission track sample arrays An important consequence of fission track annealing is that fission track ages gradually decrease from some observed value at the Earth's surface to an apparent value of zero at the depth where nofissiontracks are retained. This depth to the base of the fission track annealing zone will depend on the thermal gradient and the actual annealing properties of the particular apatites being studied. For typical fluorapatites, as found in most granitic rocks, this will usually occur at a depth where temperatures are --100-110°C. The shape of the profile of apatitefissiontrack age with sample depth below the surface will reflect the thermal history of the rocks as they have cooled throu^ the annealing zone. Gleadow (17) has discussed how these profiles will vary for different thermal history styles. Figure 2 shows the profile of apparent fission track age and mean track length that would result from a two-stage thermal history including a distinct episode of rapid cooHng. The apparent age profile in such cases shows an obvious break in slope at an age which approximates the time of onset of the rapid cooling episode, 80 Ma in the example in Fig 2. Samples below this break in slope were deeper than the annealing zone prior to the cooling event and, as a result, retain nofissiontracks from before this time. Samples at higher levels record a mixture of tracksfrombefore and after the cooling event and give intermediate apparent ages. Such profiles are relatively common and may be examined in deep drill holes or, at the surface, in areas of high topographic relief (eg 7,18,19). The importance of such vertical arrays of samples is that they contain more information than that which can be obtained from any individual sample alone. Because of thefixedgeometric relationship that the samples have to each other, their thermal histories are constrained to have followed essentially parallel temperature-time paths. In areas where block uplift has dismpted an earlier cooUng pattern, the fission track profiles will be offset relative to each other (eg 19, 20). The apparent ages form palaeodepth markers which can be used as an invisible 'stratigraphy' within the rocks to determine relative uplift between different blocks and the amount of throw on bounding faults. In this situation a common thermal history is assumed to be responsible for the variation infissiontrack parameters between all of the samples in the area. This approach has been successfiilly apphed in block mountains in Antarctica, East Africa and the Snowy Mountains in southeastem Australia and gives important insights into the underlying structural evolution of the mountain ranges (1822).

197

200

160 1 1

9 20

£ 40

3 ?

S. 60

100

: -

Time (Ma) 120 80

2.5 km 25°C/km

40 1 1 1

Track Length (|im)

- too 0

80 120 160 200

Apparent Age (Ma)

Figure 2. Modelling of the vertical profile of apparent fission track age (open squares) and mean track length (filled squares) in apatite which would result from the two stage thermal history shown on the left. Temperature increases downwards and can also be regarded as a proxy for depth in the crust. To visualise such afissiontrack data set as an image would ideally require an extremely high sampling density, which in most cases is quite impractical to collect. However, in such areas, the measured relationdiip between apparent fission track age and sample elevation means that the sample elevation itself can be used to interpolate the fission track age between sample localities, at least within a particular fault block. Combining the regional thermal history and a digital elevation model enables a series offissiontrack images of the area to be produced. The principal assumption made is that the variation in apparent fission track ages at the land surface is caused by the depth of denudation and the pattem of tectonic offsets. The Snowy Mountains We have applied this approach to a study of the granitic rocks of the Snowy Mountains region, collecting samples at regular 100m intervals in elevation along a series of vertical profiles (21,22). These cover --ITOOm of topographic rehef in a number of different blocks in the range. A positive correlation between apatite fission track age and elevation describes segments of a profile for each measured section. These are offsetfiromone another due to block faulting but can be restored to form a common regional reference profile which preserves a pre-uplift partial fission track annealing zone. Apparent fission track ages for apatites fromfiveof these profiles are shown in Figure 3a. Li each one the apatite ages form highly correlated arrays, with apparent ages increasing towards the higher elevations, but all of these are offset vertically from one another by greater or lesser amounts. If the individual profiles are allowed to move vertically then they converge on a single profile (Fig 3b) which has a similar fomi to that shown in Figure 2. The vertical offsets between the different profiles required to produce this convergence varies by amounts up to 83 Om. On this basis it is thought that a single thermal history is applicable for all of the rocks of this area and that the vertical offsets required to restore the individual profiles to the master reference profile reflect actual tectonic offsets of the fault blocks. The relatively simple relationship between elevation and age of the reference profile may be used as the basis of interpolation between those points where samples have been analysed. Starting with a gridded digital elevation model, the Kosciusko massif and adjacent Monaro Tableland can be divided into a set of blocks on the basis of mapped and inferred faults. The age-elevation profile determined for each block is then applied to estimate the relative offsets and predict the local apatite fission track age. Even for an unknown fault block within the area, the apatite ages can be predicted on the basis of even a single sample on the reasonably secure assumption that the same underlying thermal history has applied to this block as well. The result is a complete image of the fission track age at the surface, tied to a set of sampled control points. This predicted apatite fission track age can also be converted into an image showing regional denudation as illustrated in Figure 4. An additional image of the area can be constructed showing the relative tectonic uplift of the various fault blocks. The age mismatch across each fault may in tum be used to infer vertical offsets, based on the reference profile, which range up to several hundred metres in most cases. The resulting set of images may be visualised in a variety of ways, including shaded relief maps and pseudo-perspective views with various overlays of faults and two dimensional profiles showing relative offset of blocks (21).

198

2600 2400

' • Brindle Bull Hill , O Porcupine Rocks Kosciusko-Thredbo • 0 Geehi-Hannels ^ A Alpine Way

•f

A. i

2200

C 2000 O "5

S 1800

1600

1400

- O O

1200

-0-

—

1000 50

(a)

100

150

200

Apatite Age(Ma)

250

300

Apatite Age(Ma)

Figure 3. The variation of apparent fission track age with sample elevation in apatitesfromfivevertical profiles within the Snowy Mountains. The left hand diagram (a) shows the raw data plotted against actual sample elevation for the profiles while in the right hand diagram (b) each individual sampling profile has been translated vertically to show how they all represent a single underlying profile. The vertical offsets required are indicated relative to the Brindle Bull Hill profile as a reference on the ri^t side of (b). The master apatite age profile in Fig 3b profile shows a prominent break in slope at around 90 Ma, below which there is very little variation in the apparent age for further changes in elevation. The common thermal history responsible for this master profile reveals that an episode of rapid cooling began in this area at around 90 Ma. Similar periods of rapid cooling at about this time (Mid-Cretaceous) are observed elsewhere in southeastem Australia such as on the east coast south of the Sydney Basin (3), on the eastem flanks of the Lachlan Fold Belt inland from the Sydney Basin (8) and around the margins of the Bass Strait Basins (4). On this basis it is likely that cooling at this time was widespread and may have been related to the first upUfl of the present day highlands. However the movements responsible for creating the present topographic expression of the Snowy Mountains must be much younger than this as it has disrupted the cooling pattem established at this time. Subsequent denudation has now exposed different levels within the master profile on the various fault blocks. The specific time of these younger movements cannot be ascertained clearlyfromthefissiondata alone. The pattem of fission track ages in the granitic rocks of the Snowy Mountains can thus be explained as the result of a common Mesozoic thermal history dominated by a significant mid-Cretaceous cooUng followed by faulting and differential uplift to produce the block mountain structure. Finally, the actual pattem of apatite ages is revealed through differential erosion into the uplifted blocks. This differential denudation can also be calculated across the area from the fission track and digital elevation data to give a third kind of image showing the depth of erosion across the area. This is essentially the inverse of the predicted age image and is based on the amount of material which must be removed within the master apatite age profile to expose the observed pattem offissiontrack ages.

199

Figure 4. A greyscale image showing the estimated depth of denudation in the Snowy Mountains region based on the relative depth of exposure indicated within a single reference profile of apparent fission track ages. The image shows the total amount of denudation in metres that is interpreted to have occurred over approximately the last 90 million years. Modelling regional sampling arrays in southeastern Australia The approach above does not lend itself so readily to large areas where there is relatively little relief, and for which the assumption of a common cooling history is clearly not appropriate, a situation which is typical of much of the Austrahan continent. An altemative approach, suitable for the interpretation of large regional arrays of surface samples, is to simultaneously model the thermal histories for all the samples in the array. In this case the thermal histories are not constrained as they would be in a vertical profile but are free to vary independently of each other. The result of this modeUing is a set of thermal histories over a consistent set of time steps. These time-temperature solutions may be visualised as a sequence of regional time-slice images which depict the regional cooling history of present day surface rocks during their passage through the upper crust. Combining the fission track modelling with other digital data sets can reconstruct not only palaeotemperatures, but also regional denudation pattems, on the assumption that cooling in the near-surface environment is dominated by denudation at the land surface. Byfiirthercombining this information with digital elevation data, it is possible to model the evolution of palaeotopography. The palaeotopography is modelled by "backstackiag" the amount of section removed by denudation in a given time period and allowing the cmst to regain isostatic equilibrium, as described by Brown (23). At present this is done simply using local isostasy but an allowance could also be added for some degree offlexuralrigidity in the crust. In this way a set of images can be constructed for any particular time slice for which the modelled temperature remains within the apatite fission track annealing window. Time sequences showing how temperature, denudation and topography have varied through time can be visualised as digital movies.

200

Some of the additional information required to make these images is not known explicitly, such as past variations in thermal gradient, and various assumptions must be made. Some of these assumptions may be quite reasonable, but at the least may have significant uncertainties associated with them. Thus it is clear that the farther removed the calculated images are from the primary data, the greater will be the uncertainties associated with them. Two greyscale examples of these images are shown in Figures 5 and 6 for two different times in the Cretaceous. The images show the distribution of palaeotemperature experienced by rocks now at the surface (Figure 5) and an estimated reconstruction of palaeotopography (Figure 6). No allowance has been made for flexural rigidity in the palaeotopographic images, which has the potential to significantly alter the outcome, but the images are nonetheless illustrative of the kind of results that can be generated through combining the regional fission track data with various other data sets.

Figure 5. Palaeotemperature images for rocks now at the surface in southeastem Australia in the Early Cretaceous (125 Ma) and at the end of the Cretaceous (65 Ma). Crosses show sample localities. Despite the uncertainties still inherent in the procedure, this modelling and imaging approach has the potential to greatly enhance the interpretation of the thermal and tectonic evolution of broad regions of the crust. Imaging also makes objective use of all the available data and may reveal aspects of the overall pattem which are not obvious from traditional methods of interpretation. Visualisation of denudation and palaeotopography may also have an important role to play in understanding landscape evolution on a broad scale, and can potentially be combined with further data sets to constrain some of those inputs which remain poorly understood. The addition of information arising from studies of the flexural rigidity of the lithosphere, past variations in heatflow, the tectonic subsidence of sedimentary basins and constraints from palaeogeography, for example, could greatly enhance the usefulness of this approach. Studies along the eastem Australian margin using these methods reveal a striking imprint of tectonic activity related dominantly to late Mesozoic continental breakup, but the effects of early Mesozoic compression may also be recorded (24). An obvious feature of the overall pattem is a decrease in apparent fission track age towards the continental margin as a result of generally greater palaeotemperatures in this direction. This is interpreted to be due to generally greater depths of denudation in this direction, probably combined with elevated heatflow near the coast during breakup (3, 9). Significantly elevated palaeotemperatures are also observed within and adjacent to the various Mesozoic sedimentary basins in the region, including the Bass Strait basins, the Sydney and Clarence-Moreton Basins (9, 10). Within some basins the possibility of elevated mid-late Mesozoic geothermal gradients suggests the apatite fission track data pattem miay result from a complex interplay between erosion and palaeotemperature. In general, the lowest palaeotemperatures

201

are found in the most elevated areas of the current highlands, and the highest palaeotemperatures near the east coast. Significant denudation is also revealed on the inland slopes of the highlands.

65 Ma

125 Ma

Figure 6. Reconstructed palaeotopography of southeastern Australia for the early Cretaceous (125 Ma) and end Cretaceous (65 Ma), based on the estimated pattems of denudation, digital elevation data and assuming local isostasy. Crosses show sample localities. Distinct differences in thermal history are visible between different major terranes in the region. In westem Victoria the rocks have seen relatively little thermal disturbance during the Mesozoic and there is no evidence for higher palaeotemperatures towards the south coast and the Otway Basin. (5, 6) Coming eastwards, this pattern gives way fairly abruptly to the style of coastwards-increasing palaeotemperatures which appears characteristic of central and eastem Victoria, and most of the eastem margin in New South Wales. There is also a general decrease in apparent fission track age northwards along the eastem margin towards southem Queensland possibly reflecting the diachronous opening of the Tasman Sea in that direction (9). To the south, Tasmania shows generally deeper erosional levels which may reflect its position within the evolving rift systems during the separation of Australia and Antarctica, and also the ceiling effect created by widespread Jurassic igneous activity there (4). Conclusions The modelling of large regional arrays of apatite fission track data is an important new tool for the visualisation of palaeotemperature, pattems of denudation and the evolution of topography over periods of time ranging back at least to the Palaeozoic. In southeastern Australia these images reveal key events in the Mesozoic thermotectonic evolution of the Eastem Australia Plate. The pattem of apatite fission track ages in the Snowy Mountains reveals a common Mesozoic thermal history dominated by a significant mid-Cretaceous cooling. This pattem imprinted a distinctive profile of apparentfissiontrack age with depth in the cmst across the region which has later been disrupted by block upUft to form the Snowy Mountains. Differential uplift of the fault blocks and displacements across bounding faults may be visualised in a series offissiontrack derived images showing the depth of denudation and relative uplift. Simultaneous modelling of the thermal history of several hundred fission track samples in a broad regional array across southeastern Australia gives rise to a series of fission track derived images. These include sequences of images of palaeotemperature, denudation depth and palaeotopography throu^ time. These images reveal a pattem of relatively high palaeotemperatures during the Cretaceous for rocks now at the surface. This pattem is related primarily to tectonic activity associated with continental breakup and the development of Mesozoic sedimentary basins. Suchfissiontrack derived images 202

have the potential to greatly enhance the usefulness of large regionalfissiontrack data sets and to enable their combination with other quantitative information in novel ways which will provide new insists into the thermal and tectonic evolution of the upper crust. Acknowledgments: Much of the earlier phase of this work was funded by the Australian Research Council and more recently by the Australian Geodynamics Cooperative Research Centre. Neutron irradiation costs were covered by the AustraUan Institute of Nuclear Science and Engineering. This paper is pubhshed with the permission of the Director of the Australian Geodynamics Cooperative Research Centre. We &ank various members of the La Trobe Fission Track Research Group for their contributions to this on-going study, especially to Paul O'Sullivan and Roderick Brown. References 1. 2. 3.

5. 6.

7. 8.

9. 10. 11. 12. 13. 14. 15. 16.

17. 18. 19.

20. 21.

22.

Gleadow A.J.W. and Lovering J.F., 1978a. Thermal history of granitic rocks from westem Victoria: afissiontrack dating study. Joumal of the Geological Society of AustraUa 25, 323-340. Gleadow A.J.W. and Lovering J.F., 1978b. Fission track geochronology of King Island, Bass Strait, Australia: relationship to continental rifling. Earth and Planetary Science Letters 37,429-437. Moore M.E., Gleadow A.J.W. and Lovering J.F., 1986. Thermal evolution of rifled continental margins: new evidence fromfissiontrack dating of apatites from southeastem Australia. Earth and Planetary Science Letters 78, 255-270. Dumitru T.A., Hill K.C., Coyle D.A., Duddy I.R., Foster DA., Gleadow A.J.W., Green P.F., Laslett G.M., Kohn B.P. and O'Sullivan A.B., 1991. Fission track thermochronology: application to continental rifling of southeastem Australia. Australian Petroleum Exploration Association Joumal, 31,131-142. Foster DA. and Gleadow A.J.W., 1992. Reactivated tectonic boundaries and imphcations for the reconstruction of southeastem Australia and northem Victoria Land, Antarctica. Geology 20,267-270. Foster D.A. and Gleadow A.J.W., 1993. The architecture of Gondwana rifling in southeastem Australia: evidence from apatite fission track thermochronology. in R. Findlay, Unmg, R, Banks, R.H., and Veevers, J.J., (eds) Gondwana 8 - Assembly, Evolution and Dispersal, Balkema, Rotterdam., 597-603. Gleadow A.J.W. and O'Brien P.E., 1994. Apatite fission track thermochronology and tectonics in the ClarenceMoreton Basin of eastern Australia. AGSO Bulletin 261, 189-194. O'Sullivan RB., Kohn, B.R, Foster D.A. and Gleadow A.J.W., 1995. Fission track data from the Bathurst Batholith: evidence for rapid middle Cretaceous uplifl and erosion within the eastem higjilands of Australia: Australian Joumal of Earth Sciences 42, 597-607 O'Sullivan P.B., Coyle DA., Gleadow A.J.W. and Kohn B.R, 1996. Late Mesozoic to early Cenozoic thermotectonic history of the Sydney Basin and the eastem Lachlan Fold Belt, Australia, - this volume Raza A., Hill K.C., Korsch R.J. and Brown RW., 1996. Mesozoic denudation and tectonics of the central eastem margin of Australia: fission track thermochronology - this volume Green RE, Duddy LR., Gleadow A. J. W., Tingate P.R. and Laslett G.M., 1986. Thermal annealing of fission tracks in apatite: 1 - A qualitative description. Isotope Geoscience 59,237-253. Gleadow A.J.W., Duddy LR, Green P.F. and Lovering J.F., 1986. Confined fission track lengths in apatite - a diagnostic tool for thermal history analysis. Contributions to Mineralogy and Petrology 94,405-415. Laslett G.M., Green P.F., Duddy LR, and Gleadow A.J.W., 1987. Thermal annealing of fission tracks in apatite: 2 — A quantitative analysis. Isotope Geoscience 65,1-13. Gallagher, K., 1995. Evolving temperature histories from apatitefission-trackdata. Earth and Planetary Science Letters, 136,421-435. Green P.F., Duddy LR, Gleadow A.J.W. and Tingate P.R., 1985. Fission track annealing in apatite: track length measurements and the form of the Arrhenius plot. Nuclear Tracks 10, 323-328. O'Sullivan P.B. and Parrish RR., 1995. The importance of apatite composition and single-grain ages when interpreting fission track data from plutonic rocks: a case study from the Coast Ranges, British Columbia. Earth and Planetary Science Letters 132,213-224. Gleadow A.J.W., 1990. Fission track thermochronology-reconstructing the thermal and tectonic evolution of the crust. Proceedings of the Pacific Rim Congress 1990, Gold Coast, Queensland May 1990, Vol III, 15-21. Foster D.A. and Gleadow A.J.W., 1996. Structural framework and denudation history of the flanks of the Kenya and Anza Rifts, East Africa. Tectonics 15, 258-271. Gleadow A.J.W. and Fitzgerald P.F., 1986. Uplift history and stmcture of the Transantarctic Mountains: new evidence from fission track dating of basement apatites in the Dry Valleys area, southern Victoria Land. Earth and Planetary Science Letters 82,1-14. Fitzgerald P.F. and Gleadow A.J.W., 1988. Fission track geochronology, tectonics and structure of the Transantarctic Mountains in northem Victoria Land, Antarctica. Isotope Geoscience 73, 169-198. Cox, S.J.D, Kohn, B.R and Gleadow A.J.W., 1994. Towards a fission track image of Australia: model based interpretation in the Snowy Mountains using a GIS. Geological Society of Australia Abstracts 37, 72-73. (Also at http:wwwjied.dem.csiro.au:80/AGCRC/projects/2005LO/snowys/) Kohn B.R, Gleadow A.J.W. and Cox S. J., 1994. The rise of the Snowy Mountains - when and how much: evidence from apatite fission track thermochronology. Geological Society of Australia Abstracts 37,226. 203

23. 24.

Brown, R.W., 1991. Backstacking apatite fission track "stratigraphy": a method for resolving the erosional and isostatic rebound components of tectonic uplift histories. Geology 19, 74-77. 0'Sullivan P.B., Foster D.A., Kohn, B.P. and Gleadow A.J.W., 1996. Tectonic implications of Early Triassic and middle Cretaceous denudation in the eastem Lachlan Fold Belt, NSW, Australia. Geology (in press).

204

NATURE AND SOURCE OF CARBONATE MINERALISATION IN BOWEN BASIN COALS: IMPLICATIONS FOR THE ORIGIN OF COALSEAM GASES S.D. Golding, M. GUkson, KD. Collerson, J. X. Zhao, K Baublys and J. M. Crossley Department of Earth Sciences, The University of Queensland, QLD 4072 Summary Pervasive carbonate-rich veins in Permian coal measures in eastern Australia are the product of magmatismrelated hydrothermal activity. This interpretation is supported by the unradiogenic isotopic compositions (0,70352 to 0.70506) and C and 0 isotopic ratios of carbonates that indicate mixing between magmatic and meteoric fluids. Vitrinite reflectances, coking and the presence of excessive bitumen confirm rapid heating of the coal to temperatures locally approaching 300°C. Advection of heat associated with Cretaceous (and possibly Tertiary) magmatism by C02-rich hydrothermal fluids was responsible for the generation of bitumen and methane. Igneous activity during break-up of Gondwana rather than burial induced thermal maturation of the coals thus explains gas generation in Australian Permian coals. Results have major consequences regarding the source and timing of coalseam gas generation and the formation of inertinite or char-rich coals in eastern Australia and other Gondwana basins. Introduction The Permo-Triassic Bowen Basin is the northernmost structural element of the ca. 2000 km long Bowen-Gunnedah-Sydney Basin System in eastern Australia (Fig. 1). The Bowen Basin fomied through back-arc extension and foreland subsidence prior to the break-up of Gondwana, resulting in a series of distinct domains reflecting basement morphology and the effects of tectonism (1-3). It contains up to 10000m of marine and non-miarine siliciclastic sediments with coal seams and widespread accumulations of coalseam and natural gas in the basin. Coalseam gas is a potentially valuable hydrocarbon resource, yet it is a major hazard for coal miners. The formation and liberation of coalseam gas is related to the thermal and tectonic history of the host coal, and an improved understanding of its origin has consequences for resource development and mine safety. Coalseam gas is widely believed to occur largely as a result of burial induced thermal maturation of organic material (4), although recent studies have demonstrated that biogenic gases may be generated in permeable coals subject to flushing by groundwater (5). According to Mallett et al (6), thermal maturation of Permian coals in the Bowen Basin occurred during the Triassic when seams were buried to a maximum depth of 3000m. Methane was interpreted to be generated at a maximum burial-induced temperature of around 120°C for ca. 10 Myr. This model is inconsistent with new petrographic and isotopic data from the Bowen Basin that require a radical reinterpretation of the maturation paths of Gondwana coals. Coal Petrology Coals from the Bowen Basin show the following features: (I) erratic variations in rank within and between individual coalseams typical of hydrothermal regimes (Fig. 2); (II) the presence of a variety of coking structures and high contents of inertinite (char) in certain coalseams; and (III) excessive bitumen as infill of inertinite cavities and vitrinite cleats. There is a direct correlation between these features and the distribution of mafic dykes and sills in the basin. In fact, thermal aureoles with VRo >3.0% values extend for over 0.5 km adjacent to major intrusions of Cretaceous age (e.g.. Fig. 3; (7)). Changes in the character and rank of the coal are therefore interpreted to be the result of magmatic heat and associated hydrothermal activity, rather than due to the effect of burial alone. Hydrothermal inprint on the coal has resulted in irregular reflectance profiles similar to those recorded from the Midland Valley, Scotland where coals were influenced by igneous bodies and hydrothermal fluids (8). Similarly erratic reflectance profiles are the norm in active hydrothermal systems such as the Red Sea (9), and Proterozoic basins with hydrothermal ore-grade mineralization (10). Hydrothermally unaffected Permian and Triassic coals have reached sub-bituminous rank only, indicating low geothermal gradients during Triassic burial. Mean reflectance gradients of individual boreholes and individual seams also support a low geothermal gradient in the Triassic (11). The irregular reflectance profiles and widespread charring indicate uneven heat distribution with temperatures locally approaching 300°C. Furthermore, excessive bitumen is typically associated with hydrothermal systems where light and heavy oils and gases are seen to generate simultaneously (12); the heavy oil is polymerised as bitumen in vitrinite cleats and after diffusion in inertinite cavities. Coals in the Bowen Basin are extensively mineralised, with Ca-Mg-Fe carbonates, pyrite, clays, haematite and barite occurring as cleat infillings, veins and stratiform bands. Carbonate infilling cleats and in veins is intimately associated with bitumen, indicating that carbonate emplacement occurred synchronous with hydrocarbon generation. The significant presence of cleat mineralization has previously been documented in Permian coals from the Bowen-Gunnedah-Sydney Basin and its effect on coal utilization has been recognised (e.g., 13-16). It is generally accepted that nodular siderite formed during early diagenesis, whereas Ca-Mg-Fe carbonates were deposited after maturation of the coals was largely

205

complete. Carbonate mineralization has generally been interpreted to result from diagenetic processes associated with basin development; however, the source of the mineralising fluid was equivocal. Textural relationships at a variety of scales indicate that the formation of calcite and other carbonates postdates cleat-fill clay mineralisation. Cleat-filling illites in the Baralaba/Rangal Coal Measures are of Triassic age with K-Ar dates from 244 to 219 Ma (16). On the basis of these dates and preliminary fluid inclusion studies of cleat calcite, Faraj et al (16) has proposed that carbonate mineralisation precipitated from meteoric water during Jurassic-Cretaceous times. K-Ar dating of intrusions We have identified three periods of Cretaceous magmatism, using K-Ar geochronology on plagioclase separates from intrusions in the northem and central Bowen Basin. Isotopic results are given in Table 1. The earhest magmatic event involved the emplacement of plagioclase-phyric microdiorite and rhyolite dykes and sills with ages about 130 Myr. A second group of plagioclase-phyric dioritic dykes and sills yield crystallization ages from 110 to 103 Myr. Emplacement of a suite of alkali basaltic dykes occurred at about 78 Myr. The earlier periods of dioritic magmatism in the Bowen Basin are broadly equivalent in age and petrological character to the Whitsunday-Cumberland Islands calc-alkaline volcanic-plutonic province at 105-132 Myr (17). According to these workers, this magmatism occurred in the early stages of continental breakup during rifling associated with opening of the Tasman Basin. Isotope geochemistry C, O and Sr isotopic data have been obtained for vein and cleat carbonates from different parts of the basin (Table 1). Calcite 5'^C values exhibit a wide range from -10.7 to 17.9 per mil, suggesting that the positive carbonate C isotopic compositions reflect deposition from a mixed CO2- and CH4-bearing fluid. In an aqueous fluid near the CO2/CH4 boundary, the least positive values are representative of the bulk C isotope composition of the fluid (18). These values indicate a mixed source including organic carbon and either sedimentary carbonate or magmatic carbon. A dominantly magmatic carbon source with some input of organic carbon is considered likely because the coal seams are remote from fossiUferous miarine units and the area has been the site of major Cretaceous (and Tertiary) igneous activity. Calcite values exhibit a narrower range from 9.9 to 20.9 per mil and are broadly positively correlated with the values (Fig. 3). Such covariation of C and 0 isotope composition is to be e^q)ected in hydrothermal carbonates deposited over a range of temperatures; however, the slope of the broad correlation trend is too steep to be explained by temperature variation alone. The two processes which can e:q)lain the observed correlation trend are carbonate deposition from a mixed CO2- and CH4-bearing fluid of relatively constant C and 0 isotope composition and/or fluid mixing. The occurrence of distinct subpopulation trends within carbonatesfromdifferent sites indicates that both processes are locally significant. The steep slope of the C and 0 isotope correlation trend also suggests that the carbonates were deposited over a very narrow temperature interval basin-wide, or at relatively h i ^ temperatures

206

Table 1. Stable and radiogenic isotopic data for Cretaceous intrusions, coalseam carbonate mineralisation and reflectance data for host coals from the Bowen Basin. No

Sample

1 2 3 4 5 6 7 8

Intrusion 300B E720C Ramp 6 P996 77 Aquila2 Rib Rlc

9 10 11 12 13 14 15 16 17 18

Carbonate QOOl Q005 Q007 R437C/10 OTO-l OCK-la OCK-lb QOlOa QOlOb QOll

Rank'

K-Ar Age 2

(% VRo)

(Ma)

residue

(ppm)

133±2 132±3 103±4 110±6 78±I 129±2 103±1 no±2

0.8 1.4 1.3 0.9 1.1 0.9 0.9 0.5 0.5 0.5

1.1 30.4 8.5 0.8 16.2 0.1 0.1 27.5 1.0 26.1

2240 1768 2650 4564 1071 1580 1441 1761 1229 1307

•Tlb/^r

«Sr/«Sr '

Initial

nuid6'«0«

Fluid

SMOW

@150'C

@250'C

-5.6

19.8

7.1

12.5

-8.6

22.2

9.5

14.9

15.3 1.1 .9.9 12.9 17.9 3.1

14.6 14.1 9.9 17.3 16.6 20.9

1.9 1.4 -2.8 4.6 3.9 8.2

7.3 6.8 2.6 10.0 6.3 13.6

-7.4 -0.7 -10.7

10.8 12.4 11.6

-1.9 -0.3 -I.l

3.5 5.1 4.3

"Sry^Sr^

PDB

0.0827

0.703954±09

0.703798

0.0095

0.703634±16

0.703620

0.0735

0.703439±11

0.703358

0.0000 0.0011 0.0003 0.0001 0.0003 0.0001 0.0000 0.0000 0.0000 0.0000

0.704814± 10 0.705061±10 0.704662±11 0.704710± 11 0.704711±11 0.703516±10 0.703532±ir 0.703809±11 0.703807±08' 0.703786±10

0.704814 0.705059 0.704662 0.704710 0.704711 0.703516 0.703532 0.703809 0.703807 0.703786

•

1. Petrological studies were carried out on polished coal blocks in reflected white light, and fluorescence mode, using an MPV-2 photomicroscope at wave length of546 nm. The same polished blocks were used for maceral analysis and reflectance determinations. A Jeol-6400F field emission SEM equipped with Energy Dispersive Spectrometer (EDS) was used for elemental analysis of carbonate mineralisation. 2. Ar isotopic ratios were measured on a VG 8-80 single inlet gas source mass spectrometer using a static routine. The K2O content of samples was determined in duplicate by atomic absorption spectrophotometry. Ages were calculated from mass spectrometer data corrected for mass discrimination and system blanks using decay constants of Steiger and Jager (19). Replicate analyses of separate loads of the ANU standard biotite GA1550 yielded a mean K-Ar age of 97.8±2.1(la) Ma. 3. Carbonate was dissolved using distilled 7N nitric acid. The sample was then centrifuged and the leachate was split into two aliquots, a smaller one for ICP-MS trace element analysis, and a larger one for Sr isotopic analysis. The residues were dried and weighed with the residue % presented in the table. The Sr aliquots were converted to chloride using sub-boiling HCl. Sr and a rare earth element enriched fraction were separated using cation exchange resin. The total procedural blank for Sr was less than 50 pg. Sr isotopic ratios were measured on a Fisons 54-30 Sector thermal ionisation mass spectrometer using a dynamic routine. Sr isotopic ratios were corrected for mass discrimination using ^®Sr/®®Sr = 0.1194. Replicate analyses of separate loads of NBS-987 yielded a mean ''Srf^Sr = 0.710248±16(2a). 4. Initial ®'Sr/®®Sr was calculated at t = 110 Ma. ® W S r ratios were measured by ICP-MS. The decay constant for is 1.42*10' m.yr\ 5. Carbonates were reacted with orthophosphoric acid @ 25° for 1 day to extract carbon dioxide for carbon and oxygen isotope analysis (20). An acid fractionation factor of 1.01025 was used for the calculation of values (21). Analyses are reported in per mil relative to PDB for carbon and V-SMOW for oxygen with analytical uncertainities of better than ±0.2 per mil (2a). 6. Fluid

values were calculated using the calcite-water fractionation equation of O'Neil et al (22).

7. Replicate analysis with separate dissolution, separation chemistry and mass spectrometry. (i.e., greater than 150°C) where mineral-fluid oxygen isotopefractionationsare small. The high temperature hypothesis is consistent with the common observation of coking structures adjacent to carbonate minerahsation and carbon isotopic evidence for '^C exchange between CO2 and CH4. Calculated fluid oxygen isotope compositions in the temperature range 150 to 250°C are ^®0-enriched and overlap the ranges of magmatic and meteoric fluids. Calcite is a strontium-rich phase that provides a good indication of the source of Sr in the mineralisingfluid.Vein and cleatfilling calcites have strongly unradiogenic Sr isotopic compositions which show significant regional variations (Table 1). Samples from the northem Bowen Basin have "Sr/®^Sr ratios between 0.70466 and 0.70506, whereas samples from the central and southem part of the basin are even less radiogenic, with ^'^Sr/^Si ratios between 0.70352 and 0.70471. The ^^Sr/^^Sr isotopic composition of the carbonates overlaps the initial ratios reported for the 105-132 Myr. Whitsunday-Cumberland Islands volcanic-plutonic province to the east (0.7031 to 0.7044; (17)) and for Tertiary volcanic rocks in central (Queensland (0.70300 to 0.70428; (23)). By contrast, argillites and greywackes of the Palaeozoic basement 207

Neranleigh-Femvale Group have distinctly more radiogenic ®'Sr/^Sr ratios of 0.71100 and 0.71750 calculated at 100 Myr (23). The unradiogenic ®^Sr/®^Sr isotopic compositions indicate that Sr in the carbonate veins in Bowen Basin coals was not derived from Palaeozoic or Mesozoic continental sediments during diagenesis as these sources would have higher (more radiogenic) ''Sr/^Sr ratios typical of felsic upper continental crust. The isotopic data therefore suggest that the Sr in the fluid depositing carbonates was derived largelyfroma magmatic source or sources, rather than a crustal source. To evaluate whether this magmatic source of carbonate was the mafic intrusions in the basin, Sr isotopic data was obtained for plagioclase separates from the 78 to 133 Myr intrusions in the northern part of the basin (Table 1). Initial ^'Si/^Sr ratios range from 0.703358 to 0.703798 and are similar to the least radiogenic initial isotopic compositions of rocks from the Whitsunday-Cumberland Islands volcanics-plutonic province (17). These initial ratios are significantly less radiogenic than the ®'^Sr/®®Sr isotopic composition of carbonate veins from the same part of the basin, suggesting that the Sr isotopic composition of the carbonate may have been displaced towards more radiogenic compositions by contamination with radiogenic cmstal Sr. An altemative source or sources of juvenile Sr is also possible as the Sr isotopic compositions are similar to carbonatites (Fig 3), suggesting a possible genetic relationship between carbonatite magmatism and the carbonate veins. Conclusions C, O and Sr isotopic data indicate that the carbonate mineralisation is the product of magmatism-related hydrothermal activity. Hydrothermal inprint on the coal has resulted in irregular reflectance profiles rather than gradually increasing reflectance with depth as would be expected with simple, burial-induced thermal maturation. Thennal aureoles around intrusions as well as significant rank variations within and between seams, together with textural relationships between carbonate mineralisation and bitumen indicate that the irregular thermal effect of hydrothermal fluids was the determining factor in hydrocarbon generation from the coals. Extensive studies of active hydrothermal systems have demonstrated that during hydrothermal activity light oil, heavy oil and methane generation occurs simultaneously, the heavy end is deposited as bitumen (12,24). Furthermore, high yields of bitumen are considered typical in organic matter subjected to rapid heating at relatively high temperatures (25). High bitumen content has been noted for Indian coals (26), and other Gondwana coals. The strong correlation observed between native gas desorption and bitumen (or pyrobitumen) content suggest that the cracking of this bitumen during hydrothermal maturation was the major source of coalseam gases in Bowen Basin coals (7). This study demonstrates for thefirsttime that thermal maturation of coal in the Bowen Basin to form oil and gas is caused predominantly by magmatically induced thermal events. The heavy fraction of this hydrocarbon generation process occurs as bitumen infill of cleats and cavities in all Bowen Basin coals. Changes in the character and rank of the coal are therefore a result of magmatic heat and associated hydrothemaal fluid activity in the Cretaceous (and possibly Tertiary) rather than burial alone. The results provide a method for assessing the potential of gas accumulation in coal seams and hazards due to gas outburst based on petrographic observation and isotopic geochemistry. Acknowledgments Financial support for this research from the Queensland Transmission and Supply Corporation and the Energy Research and Development Corporation is gratefiilly acknowledged. Our sincere thanks to The Shell Company of Australia, Capcoal, Newlands Coal and MGC Resources for assistance with access to sampling sites and drill-core material. References 1. Fielding C.R., Falkner A.J., Kassan J. & Draper, J. 1990. Permian and Triassic depositional systems in the Bowen Basin. In: Beeston J.W. (ed) Proceedings Bowen Basin Symposium, Geological Society of Australia, Queensland Division, Brisbane, pp. 21-25. 2. Murray C. G. 1993. Tectonic evolution and metallogenesis of the Bowen Basin. In: Proceedings of the Symposium on Permian Geology of Queensland, Geological Society of Australia, Queensland Division, Brisbane, pp. 1-32. 3. DeCaritat P. and Braun J. 1992. CycHc development of sedimentary basins at convergent plate margins: 1. Structural and tectono-thermal evolution of some Gondwana basins of eastem Austraha. Journal of Geodynamics, 16, 241-282. 4. Tissot B.P. and Welte D.H. 1984. Petroleum Formation and Occurrence, Springer Verlag, Beriin. 5. Rice D.D. 1993. Controls of coalbed gas composition. Proceedings Intemational Coalbed Methane Symposium, Birmin^am, Alabama, USA, The University of Alabama, pp. 207-221. 6. Mallett C.W., Russell N. & McLennan T. 1990. Thermal history of the Bowen Basin. In: Beeston J.W. (ed) Proceedings Bowen Basin Symposium, Geological Society of Austraha, Queensland Division, Brisbane, pp. 157.

20.

Glikson M., Golding S.D., Lawrie G., Szabo L.S., Fong C., Baublys K., Saxby J.D. and Chatsfield P. Hydrocarbon generation in Peraiian coals of Queensland, Australia: source of coalseam gases. In: Follington I.L., Beeston J.W. and Hamilton L.H. (eds) Bowen Basin Symposium 1995, Geological Society Australia Inc Coal Geology Group, Brisbane, pp. 205-216.

208

8. 9. 10. 11. 12.

13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

28.

Murchison D.G. and Raymond A.C. 1989. Igneous activity and organic maturation in the Midland Valley of Scotland. Intemational Joumal of Coal Geology, 14,47-82. Robert R 1988. Organic geochemistry and geothermal history (Elf-Aquitane and D. Reidel PubHshing Company, Boston McConachie BA. & Dunster J.N. 1996. Sequence stratigraphy of the Bowthom block in the northem Mount Isa Basin, Australia: Implications for the base-metal mineralisation process. Geology, 24(2), 155-158. Barker M.R. 1995. An interdisciplinary study of coal measures from the Reids Dome Beds, Lexington Dome, Minerva Hills, central Queensland. Unpubhshed Honours Thesis, the University of Queensland. Simoneit, B.R.T. 1994. Organic matter alteration and fluid migration in hydrothermal systems. In: Pamell, J. (ed) Geofluids: Origin, migration and evolution of fluids in sedimentary basins. Geological Society Publication, 78, 261-274. Ward C.R. and Christie P.J. 1994. Clays and other minerals in coal seams of the Moura-Baralaba area, Bowen Basin, Australia. Intemational Joumal of Coal Geology, 25,287-309. ^ Patterson J.H., Corcoran J.F. & Kinealy K. 1994. Chemistry and mineralogy of carbonates in Australian Bituminous and subbituminous coals. Fuel, 73, 1735-1745. Tarabbia P.J. 1994. Authigenic Sr-Ba-Ca carbonate minerals in coals of the Hunter Valley, New South Wales. Australian Joumal of Earth Sciences, 41, 617-620. Faraj B. S., Fielding C. R. and Mackinnon 1. D. R. 1996. Cleat mineralisation of Upper Permian Baralaba/Rangal Coal Measures, Bowen Basin, Australia. Geological Society Special Publication, 109,151-164. Ewart A., Schon R.W. and Chappell B.W. 1992. The Cretaceous volcanic-plutonic province of central Queensland (Australia) coast - a rifl related calc-alkaline province. Transactions of Royal Society of Edinburgh: Earth Sciences, 83, 327-345 Ohmoto H. and Rye R.O. 1979. Isotopes of sulphur and carbon. In: Bames J.L. (ed) Geochemistry of Hydrothermal Ore Deposits, 2nd edition, Wiley Interscience, New York, pp. 509-567. Steiger R.H. and Jager E. 1977. Subcommission on Geochronology: convention on the use of decay constants in geochronology and cosmochronology. Earth and Planetary Science Letters, 36,359-362. McCrea J.M. 1950. On the isotopic composition of carbonates and a paleotemperature scale. Joumal of Chemical Physics, 18, 849-857. Sharma T. and Clayton R.N. 1965. Measurement of 180/160 ratios of total oxygen carbonates. Geochimica et Cosmochimica Acta, 29,1347-1353. O'Neil J.R., Clayton R.N. and Mayeda T.K. 1969. Oxygen isotope fractionation in divalent metal carbonates. Joumal of Chemical Physics, 51, 5547-5548. Ewart A., Chappell B.W. and Menzies MA. 1988. An overview of the geochemical and isotopic characteristics of the eastern Australian Cainozoic volcanic provinces. Joumal of Petrology Special Lithosphere Issue, pp. 225-273. Simoneit B.R.T. 1985. Hydrothermal petroleum: Genesis, migration and deposition in Guaymas Basin, Gulf of California. Canadian Joumal of Earth Sciences, 22,1919-1929. Barth T., Borgund A.E. and Hopland A.L. 1989. Generation of organic compounds by hydrous pyrolysis of Kimmeridge oil shale - bulk results and activation energy calculations. Organic Geochemistry, 4(1), 69-76. Saxena R., Navale G.KLB., Chandra D. and Prasad Y.V.S. 1990. Spontaneous combustion of some Permian coal seams of India: an explanation based on microscopic and physico-chemical properties. Palaeobotanist, 38,58-82. Nelson D.R., Chivas A.R., Chappell B.W. and McCulloch M.T. 1988. Geochemical and isotopic systematics in carbonatites and implications for the evolution of ocean-island sources. Geochimica et Cosmochimica Acta, 52, 1-17. Bell K. And Blenkinsop J. 1989. In: Bell K. (ed) Carbonatites Genesis and Evolution, Unwin Hyman, London, pp. 278-300.

209

Figure 1. The distribution of coal measures and sample localities in the Bowen Basin (open squares with numbers corresponding to samples in Table 1). The insert map shows the extent of the Bowen-Gunnedah-Sydney Basin System.

210

100 125 150 175 200 225 250 275

Depth (m) Figure 2. Examples of typical vitrinite reflectance profiles for: a) coalseams in coal measure sequence, central Bowen Basin; and b) a single coalseam, northem Bowen Basin. These changes in the rank of the coals reflect magmatic heat and associated hydrothermal activity rather than the effect of burial alone.

211

0.702

0.703

-©

0.704 0.705 Initial 87Sr/86Sr T

0.706

CQQ 8 C-. 4 o

T—(

Sedimentary carbonates 0 -4 -^^X^v^v^v^v^v^'^^'^Worldwide carbonatite field -8

-12

0.702

0.704

0.708 0.706 Initial 87Sr/86Sr

0.710

Figure 3. C-O-Sr isotope systematics of carbonate veins in Bowen Basin coals. Shown for comparison are the fields for sedimentary carbonates, carbonatites and the depleted mantle (27-28), the Cretaceous Whitsunday-Cumberland Islands volcanic-plutonic province (17) and central Queensland Cainozoic alkali basalts (23).

212

COBALT AT MT MANGANESE CENTRAL QUEENSLAND Rod Gould Summary An Early Jurassic sequence of predominantly quartzose sandstones in the easternmost margin of the Eromanga Basin has been overlain in places by a thin veneer of Tertiary sediments and intruded by Tertiary basaltic and gabbroic rocks. In the area to the immediate south-west of the Carnarvon National Park, central Queensland, outcrops referred to as manganese "wads" and manganiferous quartzose sandstones and grits contain cobalt in concentrations of 0.05% to over 2.8%. At Mt Manganese, drilling has shown that the manganese-cobalt mineralisation extends downwards beneath prominent minerahsed mounds, apparently confined within joint-bounded diatreme-Uke structures containing brecciated coarse sandstone blocks. Interaction of the ascending magmas with groundwater in the underlying strata probably produced the diatreme-like structures and the hot waters could have mobilised and concentrated anomalous manganese and cobalt from certain horizons within the Evergreen Formation. A later phase of Tertiary weathering may have been responsible for further enhancement of the mineralisation. Introduction Outcrops referred to as manganese wad and manganiferous sandstone are widespread in the Mt Tabor area to the immediate south-west of the Camarvon National Park in central Queensland (Figure 1). The manganese oxides commonly contain cobalt in concentrations greater than 0.05%. Investigations of these occurrences have been carried out since the mid-1970s, with drilling programs undertaken in 1980-81 and 1993 at a locality known as Mt Manganese. Location Mt Manganese is about 125km north-west of Augathella in the Bally Lethbridge paddock of "Babbiloora" Station. The Universal Grid Reference for the summit of Mt Manganese on the 1:100,000 Sheet 8447 "Warrong" is 55J EN (5)580 (72)290. Previous Investigations Samples of manganese wad and manganiferous sandstone from the area were submitted to the Geological Survey of Queensland in 1952 and the wad material assayed 38.9% manganese and 2.4% cobalt. Mr Denys Faulkner, who was a station-hand employed in the area at the time, tells me he submitted the samples from the Mt Manganese locality. Subsequent reference was made to the occurrence in reports on the Bureau of Mineral Resources/Geological Survey of Queensland mapping of the 1:250,000 Eddystone geological sheet (1-3).

Figure 1. Location map, Mt Manganese A definitive article on cobalt resources in Queensland published by the Geological Survey of Queensland early in 1979 also noted the occurrence (4), as did the more recent publication on nickel and cobalt from the same organisation (5). Mineral Deposits Ltd explored the area 1978-1982 (6-10) and Southem Ventures NL held exploration title around Mt Manganese 1983-1985 (11-14). Cobalt Resources NL has worked on this project since 1992 (15-17). 213

Mineral Deposits drilled 62 percussion holes totalling 1,091m at Mt Manganese and estimated the resource to a depth of about 20m as 73,000 tonnes @ 0.3% cobalt (cut-off 0.05% Co). Mineral Deposits obtained a 200kg bulk sample and the CSIRO Division of Mineral Chemistry was commissioned to undertake metallurgical testing (18,19). Cobalt recoveries of close to 100% were obtained from the material in relatively short time periods when treated with agitated sulphurous acid. Cobalt Resources drilled 139 percussion holes totalling 4,974.5m at Mt Manganese and estimated the resource to 50m depth as 175,000 tonnes @ 0.15% cobalt (cut-off 0.05% Co). In order to test the applicability to locating additional minerahsation. Cobalt Resources commissioned an airborne geophysical orientation survey in 1994 over a 120 square kilometre area around Mt Manganese, recording multichannel radiometric observations and magnetometer readings (16). The map of residual magnetic intensity clearly identified intrusive/eruptive centres, some of which were already known to have associated manganese-cobalt mineralisation. General Geology The geology of the area is taken from the Eddystone 1:250,000 geological sheet (2). A sequence of shallow dipping but undulating Early Jurassic sediments of the eastem margin of the Eromanga Basin is followed by Tertiary basalt flows and intrusives. The general trend of the undulating Early Jurassic sediments indicates a regional shallow dip to the south-west. The Jurassic sediments comprise a conformable sequence starting with the Precipice Sandstone, followed by the Boxvale Sandstone and Westgrove Ironstone overlain by the Hutton Sandstone. The Precipice is a white, cross-bedded pebbly quartzose sandstone with some conglomerates. The Boxvale and Westgrove are generally considered to be Members of the Evergreen Formation which is known to contain ferruginous manganiferous horizons that exhibit anomalous cobalt; the Boxvale contains quartzose and micaceous sandstone and is characterised by upper and lower scarp-forming sandstones, often somewhat iron-rich, while the Westgrove is characterised by concretionary ironstone and some limonitic and haematitic quartzose sandstones. Fine quartzose sandstones and thinly-bedded siltstones occur between the upper and lower scarp-forming sandstones of the Boxvale Sandstone. The Hutton Sandstone is a buff coloured mediumgrained quartzose and felspathic sandstone. Remnants of Tertiary olivine basalt flows, sills, dykes and gabbroic plugs are widespread in the area, usually forming distinctive peaks and small tablelands. Tertiary metaquartzite or silcrete and quartzose gravels are also evident in some places immediately beneath the base of the basalt flows. Mt Manganese Cobalt Prospect The Tertiary basaltic plug and associated flows form a small peak at Mt Manganese. The flows are underlain in places by silcrete and some thin quartzose pebble conglomerates. Not far below the base of the flows, sporadic outcrops of manganiferous sandstones and grits occur as small scarps and bluffs around the circumference of the peak and some of this material also has a high iron content. However in an area extending 300-450m to the north-west from the summit (Figure 2) there are three distinctive mound-like outcrops exhibiting massive manganese wad as well as manganiferous quartzose sandstones and grits, and showing signs of brecciation. These three obvious mineralised outcrops, identified as outcrops 2,3 and 4, more or less aUgn with the summit of Mt Manganese. At three other locations in the general Mt Tabor area, somewhat similar manganese-cobalt mineralisation is displayed on the north-westem side of the respective basaltic intrusions. The mianganese minerals form the matrix between the quartz grains in the sandstones and grits, and some quartz grains are present even in the more massive "wad". Mammilated surfaces of manganese oxide are developed in the open fractures and joints in the brecciated material.

214

(D Outcrop 4 Outcrop 3

0 Outcrop 2 Outcrop 5 Q

t) Outcrop 1

0 I Outcrop 6 10.200mN

Outcrip 0

MT MANGANESE

lO.OOOmN

LEGEND Q

Suiyoyed Grid line Outcrop - >0.05% Cobalt

100m

FIGURE 2

Figure 2. Outcrop location plan, Mt Manganese A typical elemental analysis of a composite grab sample of wad from the surface of the mounds is 2.88% Co, 31.30% Mn, 0.25% Ni, 0.22% Cu, 0.10% Zn, 0.77% Fe, 4.47% Al, 6.48% Ba, 11.20% Si, while a typical analysis of a composite grab sample of the manganiferous sandstone is 0.65%) Co, 9.68% Mn, 0.05% Ni, 0.10% Cu, 0.04% Zn, 2.19% Fe, 1.59% Al, 1.88% Ba, 35.20% Si. Mineralogical analysis (19) indicates two phases of manganese oxide are present and, although intimately associated, only one contains a significant amount of cobalt. There is a barium-rich phase considered to be romanechite (psilomelane) and the cobalt-bearing phase of pyrolusite (asbolan). The Mt Manganese prospect has been extensively drilled, particularly in the vicinity of the prominent manganiferous mounds. The manganese-cobalt minerahsation extends downwards beneath the three obvious surface mounds to depths of at least 50m, with the higher cobalt grades occurring near or at the surface (Figure 3). The mineralisation is apparently confined by joints and the material within the joint-bounded regions is brecciated and mostly coarse-grained quartzose sandstone or grit. This contrasts with the adjacent rocks which are generallyfiner-grained,bleached or pallid and contain sericite 215

in some horizons. It seems that vertical or steeply dipping structures at Mt Manganese have acted as conduits for mineralising fluids. The mineralisation associated with the sporadic outcrops, small scarps and bluffs around the circumference of the peak, appears to have a sub-horizontal component with the mineralising fluids permeating along joints, fractures and partings. Some of this mineralisation has a relatively high iron content. W

- 700mR.L

. 680m R.L

T.D. 30m

- 660m R.L.

T.D. 52m

LEGEND >0.05% Cobalt assay values

_j f"

20m

0.01-0.05% Cobattassay values

r *. .

Cobalt anomaly

0.26/30

% of Cobalt over distance in metres (0.26%Co/30metres)

MT MANGANESE OUTCROP 2 CROSS SECTION 10,325m N FIGURE 3

Figure 3. Cross section of Outcrop 2 along grid line 10,325m N, Mt Manganese Interpretation and Conclusions The origin of the manganese-cobalt mineralisation at Mt Manganese is uncertain, but throughout the general area such mineralisation is obviously related in some way to the Tertiary igneous activity. However the basalt itself lacks any significant manganese or cobalt. The mineralogy of the resource indicates a complex genetic process (18). The deposit does not fit comfortably into any of the accepted designations for manganese oxide metallogenesis (20) or with known associations recorded for other AustraUan manganese oxide occurrences (21). Drilling of the prominent mounds displaying manganese-cobalt mineralisation at Mt Manganese has shown the mineralisation extends downward apparently confined within joint blocks containing coarser sandstones and grits. These occurrences are reminiscent of the breccia filled diatremes recently described from the Sydney Basin (22); it was noted that a series of such structures may occur along the line of a dyke or lineation associated with intrusions. The diatremelike structures possibly resulted from the interaction of the ascending basaltic magma with groundwater contained in the underlying Precipice Sandstone aquifer. Anomalous concentrations of manganese (e.g. 1.5%) and cobalt (e.g. lOOppm) are known to occur in certain horizons within the Evergreen Formation and it is possible that this material was mobihsed by the hot waters. A later phase of Tertiary weathering may have been responsible for some further concentration of the mineralisation. 216

References 1. Mollan, R. G., Exon, N.F., & Forbes, V.R. Notes on the geology of the Eddystone 1:250,000 sheet area. Bureau of Mineral Resources, Australia, Record 1965/98.1965. 2. Exon, N.F. Eddystone, Queensland. Sheet SG/55-7. 1:250,000 geological series. Bureau of Mineral Resources, Austraha, E;q)lanatory Notes. 1968. 3. Mollan, R.G., Forbes, V.R., Jensen, A.R., Exon, N.F., & Gregory, C.M. Geology of the Eddystone, Taroom, and western part of the Mundubbera sheet areas, Queensland. Bureau of Mineral Resources, Australia, Report 142. 1972. 4. Brooks, J.H. Cobalt resources of Queensland. Queensland Government Mining Journal 80:17-25.1979. 5. Walhs, D.S. Nickel and cobalt. Queensland mineral commodity report. Queensland Government Mining Journal 95:28-36. 1994. 6. Wynn, R.G. Report on A to P 2040M - Oiarleville for the six months May 15, 1980 - November 15, 1980. Mineral Deposits Limited. 1981. (Queensland Department of Mines & Energy CR8953.) 7. Grenning, RJ. Report on A to P 2040M - Charleville for the six months November 15, 1980 - May 14, 1981. Mineral Deposits Limited. 1981. (Queensland Department of Mines & Energy CR9138.) 8. Report on A to P 2040M - Charleville. Drill logs, holes MT5-61. Mineral Deposits Limited. 1981. (Queensland Department of Mines & Energy CR9242.) 9. Grenning, RJ. Report on A to P 2040M - Charleville for the six months May 15, 1981 - November 14, 1981. Mineral Deposits Limited. 1981. (Queensland Department of Mines & Energy CR9582.) 10. Cook, F.W., & Grenning, RJ. Final Report on A to P 2040M - Charleville. Mineral Deposits Limited. 1982. (Queensland Department of Mines & Energy CR11075A.) 11. Fisher, D. A to P 3405M. Mt Tabor cobalt project, Queensland. Report for six months ended 4th July 1983. Southem Ventures NL. 1983. (Queensland Department of Mines & Energy CR12568.) 12. Flesher, C.J. A to P 3405M. Mt Tabor cobalt project. Report for six months ended 4th January 1984. Southem Ventures NL. 1984. (Queensland Department of Mines & Energy CR13167.) 13. Flesher, C.J. A to P 3405M. Mt Tabor cobalt project. Six monthly report for the period ended 3rd July 1984. Southem Ventures NL. 1984. (Queensland Department of Mines & Energy CR14084.) 14. Flesher, C.J. A to P 3405M. Mt Tabor cobalt project. Final report, January 1985. Southem Ventures NL. 1985. (Queensland Department of Mines & Energy CR14085.) 15. Cobalt Resources NL. Prospectus. 1993. 16. Cobalt Resources NL. Annual Report. 1994. 17. Cobalt Resources NL. Annual Report. 1995. 18. Canterford, J.H., & Joye, M.-C. Leaching of Mt Tabor non-ferrous metal-bearing manganese wad. Report No.l: January - March 1980. CSIRO Division of Mineral Chemistry. IR 1115 R. 1980. (Queensland Department of Mines & Energy CR8301 & CR11075B.) 19. Canterford, J.H. Processing of Mt Tabor manganese wad. CSIRO Division of Mineral Chemistry. MCC 313.1981. (Queensland Department of Mines & Energy CR9138 - Appendix.) 20. Nicholson, KL Contrasting mineralogical-geochemical signatures of manganese oxides: guides to metallogenesis. Economic Geology 87: 1253-1264.1992. 21. Ostwald, J. Genesis and paragenesis of the tetravalent manganese oxides of the Australian continent. Economic Geology 87: 1237-1252. 1992. 22. Dove, A., & Lee, G. Breccia filled diatreme in Permian Illawarra Coal Measures and Triassic strata, Kandos, New South Wales. Joumal & Proceedings, Royal Society of New South Wales 127: 39^5. 1994.

217

STRATIGRAPHIC RELATIONSfflPS BETWEEN LATEST TRIASSICEARLY CRETACEOUS BASINS OF SOUTHERN QUEENSLAND Peter M. Green & John L, McKellar Department of Mines and Energy, Queensland Summary The latest Triassic-Early Cretaceous succession in the Clarence-Moreton, Surat and Eromanga Basins can be subdivided into six stratigraphic sequences. These sequences are generally complete in the east, but are separated by hiatuses of varying duration in the west. The origin of these sequences is considered in relation to tectonic activity, eustasy, sediment supply, and depositional setting. Introduction In north-eastern Australia, during the latest Triassic-middle Cretaceous, fluvial,fluvio-lacustrineand marine conditions developed in a series of interconnected intra-cratonic basins, which have generally similar depositional histories. An understanding of the lithostratigraphic and biostratigraphic relationships of the strata in these basins has highlighted regional trends in sediment distribution, depositional environments, and local tectonic activity. Geological setting The latest Triassic-middle Cretaceous basins of north-eastem Australia developed westward of the Eastem Australia plate boundary [Shaw (1); Veevers (2)]. An active volcanic arc existed along the margin of this boundary, offshore from the present coastline. Volcanic material was shed from the arc and transported westward. Contingent upon periods of increased tectonism and volcanicity in the arc, supply and deposition of this material altemated with the supply of quartzose sediment from cratonic areas in the south and west. The influence of altemating sediment sources is apparent in six sedimentary basins: the Maryborough Basin, Nambour Basin, Clarence-Moreton Basin, Surat Basin, Eromanga Basin and the Carpentaria Basin. Lithostratigraphic interconnection of these basins reflects the development of a largely continuous depositional system. Their sedimentation history records an initial period of fluvial and fluvio-lacustrine deposition in the latest Triassic through to the Early Cretaceous. This was followed by marine, deltaic and fluvial sedimentation in the early-middle Cretaceous (not considered here). The presence of sedimentation cycles in the Surat Basin was initially recognised by Whitehouse (3). Veevers (2) highlighted the presence of tectono-eustatic cycles, which he considered to be related to changes in tectonic activity and the development of a volcanic arc along the east coast of Australia. Periods of increased tectonic activity produced a greater quantity of volcanic detritus that fomied lithic sandstones and mudrocks in these basins. In the Surat and Eromanga Basins, Exon & Burger (4) and Burger (5) also identified a number of sedimentary cycles. Their cycles consist of a lower quartzose sandstone-dominated interval and an upper mudstone-dominated interval. They suggested that these cycles were related to global sea-level changes. The stratigraphic sequences recognised by Totterdell & others (6) coincide with the cycles recognised by Exon & Burger (4) and Burger (5). Braided streams dominated the initial quartzose sedimentation, whereas meandering streams and lacustrine deposition is apparent in the upper mudstone interval in these sequences. The Surat Basin sequences recognised by Totterdell & others (6) are related to groups of formations; and the sequence boundaries coincide generally with either a top or bottom of a formation. The continuation of lithostratigraphic units in the Surat Basin with units in the Clarence-Moreton and Eromanga Basins indicates that the same sequences are present in these basins. However, biostratigraphic data point to variations in the timing, and thus local base level control, of initial sedimentation of the same sequence in different basins. Inter-basin relationships and the influence of source, tectonism, eustasy and depositional setting on the formation of the sequences are discussed below. Inter-basin sequences Only the Eromanga, Surat and Clarence-Moreton Basins will be considered here, as they dominate the Mesozoic geology of southem Queensland. The relationships between these basins are indicated in Table 1. Their sedimentation history is not complete for the whole of the latest Triassic-Early Cretaceous interval under consideration. The latest Triassic-Early Jurassic succession is complete in the east, whereas the Middle Jurassic-Early Cretaceous succession is better developed in the west. The Middle Jurassic-earliest Cretaceous stratigraphic succession in the Eromanga Basin is, however, not as fully developed as that in the Surat Basin.

218

Table 1: Stratigraphic sequences in the Eromanga, Surat and Clarence-Moreton Basins in Queensland. Sequence Age Eromanga Basin Surat Basin Clarence-Moreton Basin EARLY Wallumbilla Formation Wallumbilla Fomiation CRETACEOUS Cadna-Owie Formation Bungil Formation Mooga Sandstone Hooray Sandstone (including uppermost Orallo Formation Westboume Formation) non-deposition Gubberamunda Sandstone LATE Westboume Formation Westboume Formation D JURASSIC Adori Sandstone No sedimentary record non-deposition Springbok Sandstone Birkhead Formation Walloon Coal Measures Walloon Coal Measures MIDDLE JURASSIC Hutton Sandstone Hutton Sandstone *Heifer Creek Sandstone Mbr non-deposition 'basal Jurassic unit' *Ma Ma Creek Mbr B EARLY JURASSIC Evergreen Formation Gatton Sandstone non-deposition Precipice Sandstone mid-upper Ripley Road Sandstone lower Ripley Road Sandstone LATE TRIASSIC Cuddapan Formation Undifif. latest Triassic unit (upper part) Raceview Formation non-deposition? Aberdare Conglomerate * Members of the Koukandowie Formation Sequence A The Clarence-Moreton Basin, together with the Nambour and Maryborough Basins, record the only continuous succession of latest Triassic-Early Jurassic sedimentation on mainland Australia. A complete section of Sequence A is present in the Clarence-Moreton Basin and is represented by the Aberdare Conglomerate, Raceview Formation and the lower part of the Ripley Road Sandstone. The latest Triassic strata in the Surat and Eromanga Basins are separated from overlying Early Jurassic units by an unconformity. The correlative of this unconformity in the eastem part of the Clarence-Moreton Basin is located somewhere in the lower part of the Ripley Road Sandstone. It may coincide with the change in palaeocurrent direction in the Ripley Road Sandstone noted by O'Brien & Wells (7), from south to north-north-east. Sequence A in different areas of the Surat Basin contains strata of slightly variable palynological age within the latest Triassic [McKellar (8); Price (9)]. Nonetheless, these ages are within the limits determined for the sequence in the Clarence-Moreton Basin and reflect preservation of different parts of the sequence in the undulations of the unconformity surface of the underlying Bowen Basin. Lithologically, the sequence is diflScult to differentiate from the Precipice Sandstone, the basal unit of the succeeding sequence. Detailed palynological studies are needed to identify Sequence A in the Surat Basin and additional investigations are likely to show that it is more widespread than is presently known. In the Eromanga Basin, equivalents of Sequence A are found in the upper part of the Cuddapan Formation (including the 'Beanbush beds'). The Cuddapan Formation is easily recognisable, occurring over a large area in westem Queensland [Powis (10)]. The formation is dominantly sand-prone, but the presence of siltstone interbeds and minor coal reflect deposition by anastomosing or high sinuosity streams. However, the sUghtly older (early Late Triassic/APT4) palynological age for the lower Cuddapan Formation [Powis (10)] suggests that the formation as a whole may be more related to sedimentation in the west than in the east. Sequence B Sequence B records the westerly extension of continuous sediment accumulation from the Clarence-Moreton Basin into the Surat and Eromanga Basins. In the lower part of the sequence, the easterly flowing palaeocurrents in the Precipice Sandstone in the Surat Basin and the north-easterly palaeocurrents in the Ripley Road Sandstone in the Clarence-Moreton 219

Basin suggest deposition by the same stream system [O'Brien & Wells (7)]. Other units in the Clarence-Moreton Basin included in this sequence are the Gatton Sandstone and the Ma Ma Creek Member of the Koukandowie Formation. In the Surat Basin, Sequence B unconformably overlies the Bowen Basin and Sequence A. Sequence B encompasses the Precipice Sandstone and the Evergreen Formation. Sequence B in the Eromanga Basin is represented by the 'basal Jurassic unit'. This unit commonly consists of lower sandstone and upper mudstone intervals, which have been correlated respectively with the Precipice Sandstone and the Evergreen Formation. This correlation was based on the continuity of outcrop from the Surat Basin into the Eromanga Basin and the recognition of this subdivision in the subsurface in both the Eromanga and Surat Basins. UnpubUshed palynological data from the upper mudstone interval generally support correlation with the upper Evergreen Formation in the Surat Basin. The underlying sandstone interval of the 'basal Jurassic unit' is difficult to differentiate palynologically from the Precipice Sandstone in the Surat Basin. However, deposition of the 'basal Jurassic unit' was continuous. Thus, the lower sandstone interval is largely, if not entirely, equivalent in age to the lower Evergreen Formation in the central Surat Basin. The mid-upper Ripley Road Sandstone (Clarence-Moreton Basin), the Precipice Sandstone (Surat Basin), and the lower sandstone interval of the 'basal Jurassic unit' (Eromanga Basin) represent a highly diachronous sandstone unit. The widespread deposition of the Ma Ma Creek Member (Clarence-Moreton Basin), the upper Evergreen Formation (including the Boxvale Sandstone Member and the Westgrove Ironstone Member in the Surat Basin), and the mudstone interval of the 'basal Jurassic unit' (Eromanga Basin) was controlled by the significant base-level increase to an Early Jurassic peak in the Toarcian [McKellar (11)]. Major changes in sediment source and depositional setting occur in the upper part of the sequence. The Gatton Sandstone in the Clarence-Moreton Basin reflects tectonic rejuvenation of the hinterland to increase the supply of labile material [O'Brien & Wells (7)]. The type of stream system also changed with a more varied style being apparent, in comparison with that which deposited the Ripley Road Sandstone [O'Brien & Wells (7)]. Low-gradient streams and the development of large lake systems dominated the final depositional setting. The Ma Ma Creek Member was deposited in an environment characterised by low-gradient streams crossing a plain with extensive lakes and swanps [O'Brien & Wells (7)].The largest lake is reflected in the deposition of the Evergreen Formation in the Surat Basin. Lacustrine deposition is also apparent for the upper mudstone interval of the 'basal Jurassic unit' in the Eromanga Basin. Sequence C Sequence C in the Clarence-Moreton Basin consists of the Heifer Creek Sandstone Member of the Koukandowie Formation and the Walloon Coal Measures. The Hutton Sandstone-Walloon Coal Measures interval and the Hutton Sandstone-Birkhead Formation interval are the corresponding strata in the Surat Basin and Eromanga Basin respectively. The Heifer Creek Sandstone Member and the Hutton Sandstone are the quartz-rich units in the lower part of Sequence C, whereas the Walloon Coal Measures and Birkhead Formation are the volcano-lithic units in the upper part. Major regional changes are apparent in some units. Gray (12) and O'Brien & Wells (7) noted that the Heifer Creek Sandstone Member in the Clarence-Moreton Basin merges westwards into the Hutton Sandstone of the Surat Basin. Elsewhere in the Clarence-Moreton Basin, the Heifer Creek Sandstone Member is thinner and lower in the stratigraphic succession. O'Brien & Wells suggested, based on consistent palaeocurrent directions, that the change in sandstone composition associated with the Heifer Creek Sandstone Member reflects a change in weathering conditions in the hinterland, rather than a change in provenance. The thinner development of the Heifer Creek Sandstone Member and the presence of lithic sandstones in the eastem part of the Clarence-Moreton Basin reflect the early arrival of volcanic debris (characteristic of the overlying Walloon Coal Measures) from the east [O'Brien & Wells (7)] and closer proximity to the arc. A generally older palynological age for the Walloon Coal Measures in the Logan Sub-basin (Clarence-Moreton Basin) near Beaudesert, compared with the age of the formation in the Surat Basin, is also consistent with the incoming of material from the east. There is poor regional age control on, and imprecise lithostratigraphic definition of, the base of the Hutton Sandstone in the Eromanga Basin. Unpublished palynological data broadly suggest that, in some areas of the basin, equivalents of the lower Hutton Sandstone of the Surat Basin may not be represented. The indications are that the base of the Hutton Sandstone is younger in the west than in the east; and that a time break may occur between this formation and the underlying 'basal Jurassic unit'. The Walloon Coal Measures are present in the Clarence-Moreton and Surat Basins. The absence of coal in the Birkhead Formation in the Eromanga Basin is difiScult to explain, as coal development would be expected to occur at the distal ends of a stream system where floodplains and lakes are common. McKerron (13) has shown that the predominant current direction in the Birkhead Formation is towards the south-west, with channel orientation along a north-east to south-west trend. This is consistent with the supply of volcanic material from the east. In the Birkhead Formation, coal seams are 220

less than 10cm thick, which is in contrast to its correlative, the Walloon Coal Measures, in the Surat and ClarenceMoreton Basins. McKerron (13) suggested that 'What must have been absent was the poor drainage that allows peat to accumulate or, more likely, thefrequentinundation and reworking by floodwaters'. Sequence D Sequence D and later sequences are not represented in the Clarence-Moreton Basin. Their absence in this region is probably due to erosion during the Cainozoic, rather than non-deposition. The sequence in the Surat and Eromanga Basins is represented by the Springbok Sandstone-Westboume Formation interval and the Adori Sandstone-Westboume Formation interval respectively. In the Eromanga Basin, the uppermost part of the Westboume Formation, embracing spore-pollen unit APKl.l [Price & others (14)], is not included in this sequence. The units assigned to Sequence D are laterally continuous in outcrop between the Surat and Eromanga Basins and are readily recognisable in the subsurface. The palynological data suggests that there is a hiatus in the Eromanga Basin before the Adori Sandstone was deposited. The Adori Sandstone-Westboume Formation interval in the Eromanga Basin (excluding strata of APKl.l age) is a time equivalent of the mid-upper Westboume Formation in the Surat Basin. The supply of quartz-rich material from the west is reflected in the deposition of the Springbok Sandstone and the subsequent westward migration of the facies to form the Adori Sandstone. This supply continued during the development of sandy facies in the Westboume Formation in the west and south. An additional quartz-rich sediment source may have existed in the north as indicated by palaeocurrent directions in the Adori Sandstone in outcrop [Williams (15)]. Petrological investigations indicate that the source of sand in the Adori Sandstone and the Westboume Formation in the Eromanga Basin was similar [Shield (16)]. The quantity of quartz-rich material being supplied from all sources was insufiBcient to infill the lakes that developed. The lacustrine conditions in the Eromanga Basin region were relatively unifomi, whereas greater variation existed in the Surat Basin. Interestingly, conditions in both basins during deposition of the Westboume Formation were not conducive to accumulation of large quantities of organic matter. Sequence £ Sequence E records a major change in tiie depositional style between tiie Eromanga and Surat Basins. This change is reflected in the lack of similarity of depositional history, with the common grouping of lithostratigraphic units between basins being no longer apparent. Sedimentation of Sequence E in the Surat Basin began with the deposition of the Gubberamunda Sandstone in a braided stream system. The overlying Orallo Formation represents a change tofluvio-lacustrinedeposition, with the final phase of deposition being the fluvial Mooga Sandstone. The latter unit is provisionally assigned to Sequence E. The palynological data indicate that, in places, there is a significant hiatus at the base of Sequence E in tiie Eromanga Basin. These data suggest that strata (of unit APKl.l age) assigned to the uppermost Westboume Formation are equivalent to the upper Orallo Formation in the Surat Basin (McKellar, unpublished data). The top of the Hooray Sandstone is equivalent to tiie top of the Mooga Sandstone. The Gubberamunda Sandstone and the lower part of the Orallo Formation in the Surat Basin appear to have no time correlative in some areas of tiie Eromanga Basin in Queensland. The uppermost part oftiieWestboume Formation consists mainly of interbedded lithic sandstones and shales. This part was assigned to the Westboume Formation, based on lithic sandstones being overlain by quartzose sandstones (of the Hooray Sandstone). The presence of cross-beds and ripple laminations associated witii the interval is indicative of deposition by rivers [Shield (16)]. The uppermost part of the Westboume Formation has a relatively restricted distribution, being present mainly in the central part of the Eromanga Basin in Queensland. The hiatus in deposition, which occurred in tiie Eromanga Basin while the Gubberamunda Sandstone and part of the Orallo Formation were being deposited in the Surat Basin, corresponds to an erosion event that removed previously deposited Westboume Fomiation. This event has been recorded in the northern Eromanga Basin [Burger (5)] and is also present in the central Eromanga Basin in Queensland. The erosion resulted in the development of an incised valley (in the top of the Westboume Formation), which was subsequentiy infilled with volcano-lithic labile sandstone to form the upper part of the Westboume Formation. The distribution oftiiesesandstones reflects the extent of the incised valley. The facies associations in Sequence E in the Eromanga Basin are variable. In the south-westem part of the basin in Queensland, the sequence is represented bytiieNamur Sandstone and overlying Murta Formation. These formations pass laterally into the Hooray Sandstone. The Namur Sandstone is the quartzose sandstone unit in the lower part of Sequence E. Intiienorth, in the lower part of Sequence E, quartzose sedimentation is poorly developed. This part of the sequence contains a thin interval of pebbly quartzose sandstone, which is succeeded largely by volcano-lithic sandstones. In the upper part of Sequence E in this area, quartzose sandstones are common. These regional variations suggest that the supply of sediment into this depositional system changed and domination of the original quartzose source in the west waned. 221

Sequence F Sequence F, which represents a major change in the depositional style in both basins, encompasses the final stage of nonmarine deposition and is represented by the Bungil Fomiation in the Surat Basin and the Cadna-Owie Formation in the Eromanga Basin. These units were probably deposited in low-lying coastal environments, which enabled the rapid transgression of the sea across the land. This transgression is reflected in deposition of the Wallumbilla Formation in the Surat and Eromanga Basins. Biostratigraphic evidence suggests that the base of Sequence F is approximately the same age in both basins. Discussion The relationships shown in Table 1 reflect the maximum extent of the hiatuses present between the sequences in the Clarence-Moreton, Surat and Eromanga Basins in southem Queensland. Each hiatus is of variable extent and may not be present throughout the respective basins. The westward migration of braided stream deposition is a recurring theme in the deposition of all the sequences. The hiatuses, which are more abundant and of greater duration in the west than in the east, may be related, in part, to rates of subsidence. Subsidence in the Clarence-Moreton Basin was relatively uniform, but only about 50% of that of the Surat Basin [Gallagher (17)]. Likewise, subsidence in the Eromanga Basin, although greater than in the ClarenceMoreton Basin, was still only 75% of the Surat Basin rate. Subsidence also may have influenced the development of depositional environments in the individual basins. The greater rate in the Surat Basin may have produced the large lake system associated with the Evergreen Formation. Also, higher subsidence rates would have permitted the accumulation and preservation of vertical accretion deposits associated with fluvial systems. Thus, the proportion of mudstone in the fluvial units may be hi^er in areas of greater subsidence. The presence of a hiatus at the base of quartzose units provides information on the way in which sediments accumulated. The easterly flowing rivers transported the quartzose sediments to the east, with the sediments initially accumulating at the downstream end, the lowest point below base level. Accumulation of sediments began at the lowest point and gradually backfilled the accommodation space. This fill mechanism resulted in quartzose sediments initially accumulating in the east and backfilling to the west. This accounts for the development of a hiatus at the base of quartzose xinits and the apparent diachronous nature of these units between the basins. Although marine conditions are not directly associated with the sequences considered here, sea-level change appears to have influenced the depositional systems. This apparent influence may be related to the landward shift of meandering and fluvio-lacustrine depositional systems associated with coastal plains during times of hi^er global sea-level. Many of the palaeocurrent directions are inconsistent with the provenances envisaged for the rock types present in the basins. These directions may be a reflection of the intemal drainage pattem of the basins, rather than an indication of the direction from which the material was suppUed. The ideas presented here are based mainly on the regional geographic distribution and palynological ages of stratigraphic units in the Clarence-Moreton, Surat and Eromanga Basins in southem Queensland. Additional information is required, especially on provenance of the sediments, depositional settings, palaeocurrent directions, and palynostratigraphic relationships, in order to confirm the validity of these views. Conclusions The six stratigraphic sequences recognised here in the Mesozoic basins in southem Queensland have the following characteristics: • Bases of the sequences are variable, being continuous in the east and generally containing an unconformity in the west. • Increasing-decreasing activity in the arc controlled the eastward migration of quartzose material and the westward spread of volcanic detritus. • The local tectonism was a reflection of an upsurge in global tectonic activity that was accoii5)anied by eustatic sea-level changes. Thus, there appears in general to be a direct relationship between the formation of the sequences and relative sea-level increases when they are the result of the same tectonic activity. • A rise in sea-level may have assisted the landward migration of meandering andfluvio-lacustrinedepositional systems, resulting in the formation of the upperfine-grainedunits at the top of each sequence. • Subsidence may have influenced the intemal drainage systems and channel stacking patterns, resulting in the development of regional facies trends. • A source of quartz-rich sediment from existing cratonic areas was present miainly in the south and west, whereas volcano-lithic material was supplied from the east. • Accumulation of sediments began at the lowest point below base level and gradually filled the accommodation space. This fill mechanism, accompanied by the supply of quartzose material from cratonic areas, resulted in the westerly migration of quartz-dominant units with time. 222

References 1. SHAW, R.D.: The seismo-tectonic of southeastern Queensland. In Finlayson, D.M., (Compiler & Editor): The Eromanga - Brisbane transect: a guide to basin development across Phanerozoic Australia in southem Queensland. Bureau ofMineral Resources, Geology and Geophysics, Australia, Bulletin 232 (1990). 2. VEEVERS, J.J. (Editor): Phanerozoic Earth history ofAustralia. Oxford University Press, New York, 418 pages (1984). 3. WHITEHOUSE, EW.: The geology of the Queensland portion of the Great Australian Artesian Basin. Appendix G. In Artesian Water Supplies in Queensland. Department of the Co-ordinator General of Public Works, Queensland (1954). 4. EXON, N.F., & BURGER, D.: Sedimentary cycles in the Surat Basin, and global changes of sea level. BMR Journal ofAustralian Geology and Geophysics, 6, 153-159 (1981). 5. BURGE]^ D.: Palynology, cyclic sedimentation, and palaeoenvironments in the Late Mesozoic of the Eromanga Basin. In Gravestock, D.I., Moore, P.S., & Pitt, G.M., (Editors): Contributions to the geology and hydrocarbon potential of the Eromanga Basin. Geological Society ofAustralia, Special Publication, 12, 53-70 (1986). 6. TOTTERDELL, J.M, BRAKEL, A.T., WELLS, A.T. & HOFFMANN, K.L.: Basin phases and sequence stratigraphy of the Bowen Basin. In Follington, LL., Beeston, J.W. & Hamilton, L.H. (Editors): Proceedings of the Bowen Basin Symposium 1995,1-3 October, Mackay, Qld. Geological Society of Australia Coal Geology Group, Brisbane, 247-256 (1995). 7. O'BRIEN, P.E., & WELLS, A.T.: Sedimentology of the Bundamba Group. In Wells, A.T., & O'Brien, P.E., (Compilers & Editors): Geology and petroleum potential of the Clarence Moreton Basin, New South Wales and QnQQUslaiii, Australian Geological Survey Organisation, Bulletin 241, 72-137 (1994). 8. McKELLAR, J.L.: Palynostratigraphy of samples from GSQ Eddystone 1. Queensland Government Mining Journal, 79,424-434 (1978). 9. PRICE, P.L.: Triassic-Early Jurassic palynostratigraphic reference sections, Bowen-Surat Basins (including Taroom #12 and Eddystone #1). APG Consultants Report, 633/3 (unpubUshed) (1995). 10. POWIS, G. D.: Revision of the Triassic stratigraphy at the Cooper Basin to Eromanga Basin transition. In O'Neil B.J., (Editor): The Cooper & Eromanga Basins Australia. Proceedings of Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA Branches), Adelaide, 265-277 (1989). 11. McKELLAR, J.L.: Palynofloral and megafloral indications of palaeoclimate in the Late Triassic, Jurassic, and Early Cretaceous of southeastem Queensland. In Proceedings of MESOZOIC 96. Mesozoic geology of the East Australia Plate Conference, 23-26 September 1996, Brisbane (this volume). 12. GRAY, A.R.G.: Bundamba Group - stratigraphic relationships and petroleum prospects. Queensland Government Mining Journal, 76, 310-324 (1975). 13. McKERRON, A.J.: An investigation of the depositional environment, mineralogy and provenance of the Hutton Sandstone and Birkhead Formation sequence, central Eromanga Basin, south-west Queensland. B.Sc. (Honours) Thesis, University of Sydney (1991). 14. PRICE, P.L., FILATOFF, J., WILLIAMS, A.K., PICKERING, S.A. & WOOD, G.R.: Late Palaeozoic and Mesozoic Palynostratigraphical Units. CSR Oil & Gas Division, Palynology Facility, Report 274/25. Held by the Department of Mines and Energy, Queensland as CR14012 (1985). 15. WILLIAMS, L.J.: Sedimentology of the Adori Sandstone in the Tambo region, central Queensland. Queensland Department ofMines, Record, 1989/25 (1989). 16. SHIELD, C.J.: Sedimentology and palynology of the Westboume Formation, Augathella area, south-central Department ofR^^^ Queensland. Record, 1991/29 (1991). 17. GALLAGHER, K.: Permian to Cretaceous subsidence history along the Eromanga-Brisbane geoscience transect. In Finlayson, D.M., (Compiler & Editor): The Eromanga-Brisbane transect: a guide to basin development across Phanerozoic Australia in southem Queensland. Bureau ofMineral Resources, Geology and Geophysics, Australia, Bulletin 232, 133-175 (1990).

223

EARLY MESOZOIC MAGMATISM OF THE NEW ENGLAND OROGEN: SPACE, TIME, AND COMPOSITIONAL RELATIONSHIPS David Gusts PhO BlevinS Bruce ChappelF and Chris Stephens' 1 School of Geology, Queensland University of Technology, Brisbane,Q 2 Key Centre for Geochemical Evolution and Metallogeny of Continents, Geology Department, Australian National University, Canberra ACT 3 Norseman Gold, Norseman, WA Summary Late Permian to early Triassic magmatism of the New England Orogen is voluminous throughout the entire fold belt. Collectively termed the New England Granitoids (NEG), it is dominantly composed of I-type granitiods. Supersuites are defined for the southem NEG (sNEG), based primarily upon geochemistry. Plutonic suites of the northem NEG (nNEG) have similar geochemical characteristics and may be easily accomodated into the existing supersuite nomenclature. Supersuites in the sNEG define diffuse NNEtrending belts while those of the nNEG are constrained to packages with a NNW-trend. Magmatism of the sNEG and nNEG is broadly contemporaneous, fine scale temporal variations require the acquisition of a more extensive dating database. The NEG relationship to subduction is difficult to establish, with a subduction model more approriate for the nNEG in comparison to the sNEG. Introduction The study of granitoids as records of past magmatic events is critical in interpreting the tectono-magmatic history of continental evolution. Studies of granite geochemistry on a regional scale such as in the Lachlan Gold Belt of southeastem Australia (Chappell and White (1)) and earlier references therein) and the Peninsular Range Granitoids in westem North America (Taylor (2)) provide examples of the role of granitoids in interpreting the tectono-magmatic history of continental regions. Good preservation of Permian and Triassic plutonic rocks ^lerein collectively termed the New England Granitoids (NEG)) within the New England Fold, a region regarded as an active continental margin firom the Middle Palaeozoic through to the Mesozoic (Day et al. (3)), provides an opportunity to apply such magmatic studies to the long term tectonic evolution of this region. Shaw and Hood (4), McCulloch and Chappell (5), Hensel et al. (6) and Shaw et al. (7) provided geochemical and isotopic interpretations of the temporal, spatial and geochemical characteristics of intrusive rocks in the sNEG. Continuing study by Blevin and Chappell (8) has further refined these interpretations and provided afirameworkfor understanding the metallogenic evolution of the region. In contrast, there has been no comprehensive review of the temporal, spatial and geochemical characteristics of igneous rocks in the nNEG since Webb and McDougall (9) and the publication of the Queensland Geology map. This paper extends the work of Gust et al. (10) to provide a preliminary synthesis of the entire NEG in eastem Australia. Tectonic framework The nature and timing of the major volcanic, sedimentary and structural cycles that have affected rocks of the NEO are also the subject of continued debate (e.g. Day et al. (3); Harrington and Korsch (11), Murray et al. (12), Fergusson and Leitch (13)). The only significant developments in terms of tectonic models are those for the Carboniferous (Murray et al.(12)) and elements of the Henderson et al. (14). model for the Devonian-Triassic. Although the interpretation as to the nature of individual tectonic elements of the NEO has differed in the recent past, all previous workers followed quite closely the scheme of Day et al. (3) to include a) the Early Permian Camboon Volcanic Arc (CVA) associated with a deep, forearc or strike-slip basin (the Grantlei^ Trough. At the same time, or just after the CVA, the Bowen Basin developed as a retroarc foreland basin b) the Hunter Bowen Orogeny, a contractional (fold-thrust) event of mid-Late Permian to Triassic age (Fergusson (15)), c) Late Permian to Late Triassic I-type magmatism and d) the Gympie Block of potentially exotic origin. A well-developed model for the tectonic evolution since the mid-Carboniferous involves repeated cycles of convergent and extensional tectonics related to active subduction and subsequent roll-back of the slab, with subduction renewing to the east (Holcombe et al. (16), Stephens et al. (17)). This model suggests in part that the Hunter-Bowen Orogeny was a -25 my period of crustal contraction with at least two distinct pulses of initiating at about 260 Ma, and that the Early Triassic (250-230 Ma) I-Type magmatism records the migration of arc magmatism westward onto the continent during contraction of the continental margin. A transition to extension-related sihcic magmatism occurred in the Late Triassic, marking the onset of the prolonged break-up of Gondwana. The New England Granitoids In this paper, we use the term New England Granitoids (NEG) to include but not restricted to, all plutonic rocks of Permian through Triassic age that lie within the New England Orogeny. As such, the NEG can be divided into two parts, 224

separated by the Clarence-Moreton Basin in southeastern Queensland. The southern part of the NEG (sNEG) is described as consisting of two late Carboniferous S-type suites (Bundara, now called Copeton, and Hillgrove), three Late Permian to Triassic I-type suites (Uralla, Moonbi and Qarence River) and an unclassified group of leucogranites (Shaw and Flood (4)). We utilise the term OsupersuiteO for the description of these suites. The northem part of the NEG (nNEG) is considerably less well studied and has been the focus of ongoing research. Gust et al. (10) reviewed the temporal and spatial arrangements of nNEG magmatism, noting a progressive change in the loci of activity with time. To date, the nNEG has not been divided into supersuites. Time and space relations The Permian and upper Triassic I-type plutonism of the sNEG began about 255 m.y. ago. The emplacement of the major I-type hornblende biotite plutons, and continued until approximately 225 Ma. Shaw and Flood (4) suggests that I-type plutonism of the Upper Permian consists mainly of early mafic adamellites and granodiorites (255 - 240 Ma) and a later leucoadamellites (230 - 220 Ma). At the current time there is not a recognised distinctive age zonation for the three Itype supersuites. However, examination of the spatial distribution of these intrusions indicate that they He more or less in distinct linear belts that trend to the north/northeast. The Clarence River supersuite is situated on the most eastem portion of the sNEG, with the Moonbi and Uralla supersuites broadly overlapping. All of these supersuites He to the east of the Copeton supersuite a S-type late Carboniferous intrusive complex. Within the nNEG, Late Palaeozoic and Early Mesozoic plutonic rocks in the Queensland segment of the New England Orogen define two regions that show restricted overlap in the temporal evolution and productivity of plutonism. The most northem part of the nNEG is dominated by the Urannah Igneous Complex (305 Ma- 275 Ma). Intmsion of granites (s.s.) occurred closer to the present coastHne and offshore during the Triassic between 240-228 Ma. The southem area shows greatest activity during a broad magmatic cycle between 265 - 220 Ma. This cycle began with a bimodal plutonic event between 265 - 250 Ma, peaked at approximately 240 Ma associated with a convergent margin, and waned through a silicic magmatic event corresponding to extension between 230 - 220 Ma. This magmatic activity spatially defines a broad zone with a pronounced northwest trend in contrast to the north/northeast trend of the sNEG. Plutons of 230200Ma is displaced slightly to the east of the 230-250Ma plutons. 210 - 220 Ma anorogenic-style plutonism occurred throughout the central nNEG. Compositional relations The Moonbi supersuite covers a broad compositional range from diorite through granite, but consists dominantly of granites. It has high concentrations of large iron lithophile elements, including K20, Rb, Sr, and as well as high concentrations of U. Th, and Pb. The Moonbi supersuite has very high P205 which exhibits a strong negative correlation with silica. Its mineralogy is consistent with its high K nature there is a dominance of monzogranite, quartz monzonite, and leucogranite. The suite is also highly oxidised with high MgO/FeO*, high MgO contents, and lower FeO* contents. It is low in CaO, high in Cr and Ni; Nb and are similar to the Uralla supersuite. The Moonbi Supersuite has steep rare earth pattems, with low HREE (and Y).Isotopically, the Moonbi suite is moderate, having ®'Sr/®'Sr of 0.704 - 0.705. The Uralla supersuite is compositionally diverse, ranging from gabbro to diorite to monzogranite, but is dominantly intermediate in composition. Its P205 content is lower than the Moonbi, but higher than the Clarence River. It has a very low redox state. LREE are similar to the Moonbi, however, the HREE are higher, thus REE pattems are less steep. Initial ratios are higher than that of the Moonbi and vary between 0.7046 and 0.707. The Clarence River Supersuite (Bryant et al. (18)) consists of both the Nundall suite, the Clarence River suite and the Barrington Tops granodiorite. The Clarence River Supersuite has a great diversity, with gabbros through monzo-granites; the average mode is quartz diorite to tonalite. This supersuite is low K20, Rb, Ba, U, Pb, and Th. It has low P205, Ti02, Zr, and Nb and essentially flat to sHghtly LRRE- enriched pattems. It is isotopically primitive, with ®'Sr/®'Sr ratios varying between 0.7036 and 0.7037. Overall, the Clarence River supersuite is the most mafic of the NEG supersuites, having a large gabbroic component in individual intrusions. Many of the unclassified leucogranites (Shaw and Flood (4)) of the northem portion of can be assigned to the Moonbi supersuite. However, other leucogranites remain unassigned, their relatively young ages (226 Ma. to 195 Ma.) suggest that they are younger other I-type granites of the sNEG. Other plutonic groups identified in the sNEG include a series of small tholeiitic and calc-akalic gabbroids, spatially associated with the HiUgrove S-type supersuite and a minor high-K granite group. Examination of the plutonic rocks from the nNEG suggest that there are broad similarity of these I-type suites of late Permian to mid Triassic age, with those of the sNEG.. Rocks which are older (240 - 260Ma) are most similar to those of the Uralla, however, some plutons have high P205 contents with a strong negative correlation with Si02 and may have Moonbi affinities. Plutons of the 220 - 240 Ma in the nNEG are very similar to those of the Clarence River Supersuite. This nNEG is compositionally diverse, with a significant gabbroic/dioritic component and dominant granodiorite/tonalite signature. Granites are not common. P205 variations with Si02 overlap Clarence River supersuite trends. A group of plutons less then 220Ma is bi-modal, consisting of gabbros and mildly A-type granites. These are enriched in P205 in the gabbroic bodies but depleted in the granitic bodies. Strontium isotope data show that the Late Palaeozoic granitoids 225

and some of the earliest Mesozoic granitoids include an evolved, probably crustal, isotopic component in their source Q'Sr/^Sr >.704). Late Triassic granitoids, in contrast, have low initial strontium ratios C'Sr/'^Sr<.704). Tectono / magmatic comments and conclusions In a general way, Shaw and Flood (4) attributed the generation of the I-type suites of the sNEG to partial melting of Itype source rocks in the lower crust, underlying the continental margin. This partial melting resulted from the combined effects of an increase in temperature, due to the recovery of normal continental geotherms post-subduction, and the crustal thickening associated with over-thrusting of the trench complex. One of the major issues that appears to confront the interpretation of the granitoid magnetism of the NEG is simply, is it related to subduction? Shaw and Flood (4) correctly point out that if there was a trench, it would have been considerably offshore to our presently exposed location. That offshore location however, is unspecified, and with regard to Cretaceous extension, that subduction zone may have been considerably displaced or disrupted. In the nNEG, the possibility that the granitoid magmas, with their linear-type distributions were related to a subduction zone seems strong. Recent structural stories have documented (as suggested previously) a contractional or deformation event in the mid Triassic. The interpretation of the NEG granitoids as strictly post orogenic (Day et al.(3)) cannot be correct. The NEG I-type granites are similar to the Cordilleran granites of westem America which were associated with active subduction. The occurrence of significant amounts of gabbro and diorite indicates the importance of mantle melting. Within the framework of modem arc evolution, the process of fractional crystallisation relative to cmstal anatexis becomes a real issue. Pending the development of a more comprehensive geochemical database for the NEG, the relative contributions of mantle and cmst to the generation of the NEG remain unresolvable. References 1 2 3 4 5 6

8 9 10

11 12 13

Oiappell, B. W., and White, A. J. R., 1992,1- and S-type granites in the Lachlan Fold Belt: Trans. Roy. Soc. Edinburgh: Earth Sci., v. 83, p.1-26. Taylor, H.P Jr. (1988) Oxygen, hydrogen and strontium isotope constraints on the origon of granites. Transactions of the Royal Society of Edinbur^, 79, 317-338. Day, R.W., Murray, C.G. and Whitaker, W.G. (1978) The eastern part of the Tasman Orogenic Zone. Tectonophysics 48, 327-364. Shaw, S.E. and Hood, R.H. (1981) The New England Batholith,Eastem Australia; Geochemical variations in time and space. Joumal of Geophysical Research 86,10530-10544. McCulloch, M. T, and Chappell, B. W., 1982, Nd isotopic characteristics of S- and I-type granites: Earth Planet. Sci. Lett., V. 58, p. 51-64. Hensel H.D., McCullough M.T.. and Chappell, B.W. (1985) The New England Batholith: constraints on its derivation from Nd and Sr isotopic studies of granitoids and country rocks. Geochimica et Cosmochimica Acta 49, 369-384. Shaw, S.E., Conaghan, R.H. and Flood, R.H. (1991) Late Permian and Triassic igneous activity in the New England Batholith and contemporaneous tephra in the Sydney and Gunnedah Basins. In Diessal, C.F.K. ed. 25th symposium on advances in the study of the Sydney Basin. University of Newcastle, Department of Geology Publication, 413,44-51. Blevin, P.L. and Chappell, B.W. (1996) Permo-Triassic granite metallogeny of the New England Orogen. this volume. Webb, A.W. and McDougall, 1. (1968) The geochronology of the igneous rocks of eastem Queensland. Joumal of the Geological Society of Australia 15, 313-346. Gust, DA., Stephens, C. J. and Grenfell, A.T. (1993) Granitoids of the northem NEO: their distribution in time and space and their tectonic implications. In Aitchinson, J.C. and Flood, P.G. eds. New England Orogen, eastem AustraUa, NEO ^93 Conference Proceedings, University of New England, 565-572. Harrington and Korsch (1985) Late Permian to Cainozoic tectonics of the New England Orogen. Austrahan Joumal of Earth Sciences 32, 181-283. Murray, C.G., Fergusson, C.L., Flood, RG., Whitaker, W.G., and Korsch, RJ. (1987) Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Austrahan Joumal of Earth Sciences 34, 213-236. Fergusson, C.L. and Leitch, E.G. (1993) Late Carboniferous to Early Triassic evolution of the New England Fold Belt, eastem Australia, in Flood, P.G. and Aitchison, J.C. eds New England Orogen, eastem Austraha NEO'93 Conference Proceedings, UNE Australia, 53-59 Henderson, R.A., Fergusson, C.L., Leitch, E.C., Morand, V.L, Rheinhardt, J.J. and Carr, P.F. (1993) Tectonics of the Northem New England Fold Belt. In Flood P.G. and Aitchinson J.C., eds New England Orogen, eastem Austraha. Department of Geology and Geophysics, University of New England, Armidale, 505-515. Fergusson, C.L. (1991) Thin-skinned thmsting in the northem New England Orogen, central Queensland Austraha. Tectonics 10,797-806. 226

Holcombe, R.J., and Little, T.A. (1994) Blueschists of the North D^Aguilar Block: Structual development of the Rocksberg Greenstone and associated units near Mt Mee, southeast Queensland. Australian Journal of Earth Sciences, 41 115-130. Stephens, C.J., Ewart, A., Fielding, C.R., Holcombe, R.J. and Schon, R.W. (1994) Location and timing of convergent margin volcanism during the late Palaeozoic-Mesozoic evolution of eastem Australia. Geological Society of Australia Abstracts 37,417. Bryant, C. L, Arculus, R. J. & Oiappell, B. W. (1996) The Clarence River Supersuite: a 250 Ma Cordilleran tonalitic I-type analogue in Eastem Australia. Submitted Joumal of Petrology.

227

GROUNDWATER MOVEMENT AND HYDROCHEMISTRY OF THE GREAT ARTESIAN BASIN, AUSTRALIA

MA. Habermehl Australian Geological Survey Organisation, Canberra, ACT. Summary Groundwater in the Lower Cretaceous-Jurassic confined aquifers of the Great Artesian Basin, Australia flows from the recharge areas in the eastern margins to the westem, southwestem, southem and northem margins where artesian groundwater dischargesfi-omsprings. The chemistry of the artesian groundwater in the Basin is dominated by Na-HC03-Cl and these ions contribute more than 90 percent of the total ionic strength of solutes in the eastem, central and northem Basin areas. In the westem parts of the Basin groundwater is chemically characterised by Na-S04-Cl type water, and the groundwater flow direction is fi-om the recharge areas in the westem margin towards the southwestem discharge margins. Introduction The Great Artesian Basin is a confined groundwater basin comprising aquifers in quartzose sandstones of continental origin and Triassic, Jurassic and Cretaceous ages. The main confining unit is a thick argillaceous sequence of sediments of marine origin and Early Cretaceous age, which is underlain by the Lower Cretaceous-Jurassic aquifers, and overlain by the confined aquifers of Early to Late Cretaceous age. The Lower Cretaceous-Jurassic and Triassic aquifers altemate with confining beds of siltstone and mudstone, which are of continental and marine origin and Triassic and Jurassic in age. The Great Artesian Basin occupies an area about one-fiflh of the AustraUan continent, and underlies parts of Queensland, New South Wales, South Australia and the Northem Territory. The Basin is up to 3000m thick, and forms a large syncUnal stmcture, uphfted and exposed along its eastem margin, and tilted southwest. Recharge occurs mainly in the eastem marginal zone, an area of relatively high rainfall, and large-scale regional groundwater movement is generally towards the southwestem, southem and westem margins. Recharge also occurs in the westem margin of the Basin, and groundwater flow directions are towards the soutiiwestem discharge margin. Natural discharge occurs in those areasfiromflowingartesian springs, most of which have built up mound-shaped deposits of sediments or carbonates. Dischargefiromthe artesian aquifers near the discharge margins also occurs by diffuse discharge where the overlying confining beds are thin. Many springs are associated with stmctural features, abutment of aquifers against bedrock or thin confining beds near the discharge margins. Abundant artesian groundwater supplies of good quaUty are obtainedfiromflowingartesian waterbores, andfi-ompumped artesian waterbores in the Basin. Groundwater in the most exploited aquifers in the Lower Cretaceous-Jurassic sequence generally contains about 500 -1000 mg/L total dissolved sohds, and is therefore of good quality, making it suitable for domestic and town water supply, stock use in the pastoral industry and water supplies for the mining and petroleum industries. The Great Artesian Basin underlies arid and semi-arid regions, where surface water is sparse and unreliable. The discovery of the Basin's artesian groundwater resources around 1880, made settlement possible, and led to the establishment of an important pastoral industry. Pastoral activity and town water supplies are to a very large extent dependent on artesian groundwater in the Basin area. In recent years artesian groundwater has been used increasingly in the mining and petroleum industries located both inside and outside of the Basin area, and most of these industries are largely or totally dependent on the Basin's artesian groundwater resources, eg. Olympic Dam and the town of Roxby Downs in South Australia, several mines in the NW margin of the Basin, oil and gas production in NE South Australia, SW Queensland and E Queensland. Use in some areas for irrigation is also on the increase, though generally most of the artesian groundwater is unsuitable for irrigation because in much of the Basin area it is chemically incompatible with the soils. Hydrocarbon source and reservoir rocks are abundant in the sedimentary sequence of the Basin, and commercial and subcommercial oil and gas discoveries have been made in several Jurassic and Cretaceous sandstones, contradicting earher behefs that the Basin-wide groundwaterflowhadflushedhydrocarbons out of the system. Dissolved hydrocarbons in the artesian groundwater are generally dry gases and are usefiil petroleum e^qjloration indicators. Geology The hydrogeological Great Artesian Basin comprises the sedimentary Eromanga, Surat and Carpentaria Basins and parts of the Bowen and Galilee Basins (Habermehl, 1980). The geology of the Basins has been reviewed in Habermehl (1980, 1986). The constituent sedimentary basins are continuous across shallow ridges and platforms of older sedimentary, metamorphic and igneous rocks (Figure 1). The Basin consists of several broad synclinal stmctures trending north and northeast, overlying sedimentary, metamorphic and igneous rocks of pre-Jurassic or pre-Triassic ages. 228

The Mesozoic sedimentary sequence in the central part of the Basin reaches a maximum total thickness of about 3000m Parts of the marginal areas of the Basin have been eroded, in particular along the eastern border, which was uplifted during Cainozoic times. Sheet-like, conformable rock bodies extend relatively unchanged for hundreds of kilometres and are aknost horizontal. The Great Artesian Basin is an asymmetrical basin elongated northeast-southwest, and tilted towards the southwest. Four centres of basin subsidence are present, two coinciding with the Surat and Carpentaria Basins, and two within the Eromanga Basin, separated by the Birdsville Track Ridge, and overlying the Cooper Basin and the Pedirka Basin (Habermehl, 1980). Cainozoic and earher uplift along the eastem margin, and subsidence in several parts of the Basin, particularly in the central and southwestem parts, led to the Basin's asymmetry. Many of the near surface folds, particularly monoclinal features, grade downwards into faults and are the product of draping and differential compaction of the sediments over fault bounded basement blocks. Several major fault and fold systems occur in the Basin, in places forming en echelon structures. Throws of up to 300m affect some major faults, thou^ the displacement of Jurassic-Cretaceous sediments along normal faults is usually much less. The stratigraphic succession in the Great Artesian Basin is given in Habermehl (1980,1986), and shows the distribution and correlation of the rock units of Middle Triassic to Late Cretaceous ages in the constituent sedimentary basins. The Jurassic sequence comprises continental deposited quartzose sandstones, with lesser siltstone and mudstone. Siltstone, mudstone and lithic sandstone were deposited in shallow marine environments during Early Cretaceous times. During the Late Cretaceous more sandy sediments were laid down in lacustrine and fluviatile environments. The Eromanga Basin is deepest where it overhes Palaeozoic and older Mesozoic sedimentary basins. Thinner sequences are present across the shallow ridges and platforms connecting the Eromanga Basin with the Surat and Carpentaria Basins. The southeastem parts of the Great Artesian Basin includes the sedimentary Surat Basin, and the Coonamble Embayment, and consist of an altemation of Jurassic continental sandstone, siltstone, mudstone and some coal. The Cretaceous sediments are partly continental, but mainly shallow marine lithic sandstone and mudstone. The Carpentaria Basin contains continental rocks of Jurassic age, and marine sedimentary rocks of Cretaceous age. The deeply weathered erosional surface of the Cretaceous sediments in these basins are overlain by Tertiary sediments, which are also partly weathered and sihcified, and by mostly unconsolidated Quatemary sediments. The latter overUe parts of the Great Artesian Basin, and are usually up to several tens of metres in thickness, but form shallow basins as much as 150 m deep in some regions. Tertiary basalts cover some areas of Mesozoic rocks in the northeastem, eastem and southeastem parts of the Basin. Hydrogeology The confined aquifers of the Great Artesian Basin are present within a rock sequence, which, where complete, is bounded by the Rewan Group at the bottom, and the Winton Formation at the top (Habermehl, 1980). Aquifers are present in the Clematis, Precipice, Boxvale, Hutton, Adori and Hooray Sandstones, and the Cadna-owie Formation and their equivalents, and in the Mackunda and Winton Formations. Most of the individual aquifers are relatively uniform in their hydrogeological characteristics, and they are continuous and hydraulically connected across the constituent geological basins. The major confining beds consist of the Rewan Group, Moolayember, Evergreen, Birkhead, Westboume, Wallumbilla and Toolebuc Formations, and their equivalents, and the Allaru Mudstone, and parts of the Mackunda and Winton Formations. The hydrogeological basement comprises impervious sedimentary, metamoiphic or igneous rocks, and this basement forms in part an aquiclude or aquifuge. The hydraulic characteristics of the confined aquifers have been determined since the early development of the Basin from a large number of tests carried out on the flowing artesian waterbores, and during the last 30 years from rocksamples and wire-line logs obtained from petroleum exploration wells in many parts of the Basin. Hydraulic conductivity values range from 0.1 to 10 m/day, the majority being in the lower part of that range, and mainly related to the Cadna-owie Formation, Hooray, Algebuckina and Pilliga Sandstones and Iheir equivalents (Habermehl, 1980). Transmissivity values ascertained from periodic systematic tests by the State Water Authorities, range from 1 to 2000 mVday. Storage coefficient values, as calculated from petroleum log data, range from 10^ to 9 x 10"^ Intrinsic permeability rangesfi-omseveral tens to several thousands of millidarcys. Porosity values range fi-om 10 to 30 percent

229

(Habermehl, 1980). Average vertical hydraulic conductivities of the leaky, very low permeable, confining beds range from 10-^ to 10^ m/day. Recharge Recharge of the aquifers by infiltration of rainfall through overlying unconsolidated sediments occurs mainly in the outcrop areas of the aquifers along the eastem, elevated, margins of the Basin, which mainly are located on the westem slope of the Great Dividing Range. Recharge along the westem margins of the Basin, in the arid centre of the continent, takes place where aquifers are exposed or overlain by sandy sediments. The potentiometric surface of the confined groundwater in the aquifers of the Lower Cretaceous-Jurassic rock sequence indicate the directions of flow of the groundwater. Environmental isotope and other hydrochemical studies of groundwater from drill-holes in the recharge areas, and from waterbores located downgradient of the recharge areas, and towards the centre of the Basin, have confirmed the increase in residence times and the flowrates and flowpattems of the artesian groundwater predicted from the hydrogeological analyses and the potentiometric surfaces (Airey et al., 1979, 1983, Calf & Habermehl, 1984, Bentley et al., 1986, Torgersen et al., 1991, Herczeg et al., 1991, Habemiehl et al., 1993, Cresswell et al., 1996). These studies support an assumption of continuing recharge from geological to modem times, and also show that the artesian groundwater is of meteoric origin. Discharge Discharge from the Great Artesian Basin takes place as natural discharge in the form of concentrated outflow from springs, vertical leakage towards the regional watertable, subsurface outflow into neighbouring basins, and as artificial discharge by means of free or controlled artesian flow and pumped abstraction from waterbores drilled into the aquifers. Difiuse discharge from the artesian aquifers through the confining beds towards the surface occurs in the marginal areas were the confining beds are relatively thin, and potentials are high, and watertables shallow (Woods et al., 1990). Springs and areas of seepage are abundant in the marginal areas of the Basin, particularly in the southem, southwestem, northwestern and northem areas. Most springs are concentrated in groups, covering relative small areas. Eleven groups have been identified (Habermehl, 1982). Rates of discharge from the springs are generally low, and range from less than 1 L/s to about 150 L/s (the latter from a spring at Dalhousie Springs, northem South Australia); temperatures of the springwater range from about 20° to 45°C. Flowing artesian springs are generally associated with faults along which the groundwater flows upwards, the abutment of aquifers against impervious bedrock, and pressure water breaking through thin confining beds near the discharge margins of the Basin. Many springs have built up conical mounds several metres to several tens of metres in diameter, and up to several metres high. The mounds are formed by deposition of particles brought up from the aquifers and the confining beds, and by the chemical and biological precipitation of solids dissolved in the artesian groundwater. The artesian springs are terminal evaporitic systems, usually comprising a central carbonate mound, with outer zones dominated by sulphate and chloride salts. Mound morphology is controlled by several factors, including groundwater discharge rates, hydrochemistry, evaporation, influence of inorganic versus organic carbonate precipitation, local subsidence of the mound and micro-tectonics. In some areas the mounds consist mainly of particles brought up from the confined aquifers and the confining beds, and form mud mounds and mud volcanoes. Many mounds, particularly those built by springs in the westem and southwestem margin of the Great Artesian Basin, consist of carbonate. The latter are dominated by calcite and dolomite, and occur as tufa, travertine and veryfine-grainedor crystalline carbonate, which was deposited as a chemical precipitate out of the artesian groundwater, and precipitated by a combination of chemical, algal and bacterial action. Terraced mounds and waterfall or cascade deposits produced by algae are common, though many accumulations consist of steeply sloping mounds. Artesian springs and their deposits in the Lake Eyre region range from topographically high springs to younger, topographically low springs as a result of the lowering of the land surface and spring outlet levels in Quatemary times (deposits of extinct pre-Quatemary springs occur more than 40 m above the present springs). This also indicates that the potentiometric surface in the Lake Eyre region has declined considerably during recent geological time (Habermehl, 1982). Morphological diversity and lithofacies patterns indicate that the spring complexes have developed over several climatic cycles. The age of several basal spring deposits suggests that some springs might have been activated or reactivated as a result of major climatic changes. Fossil carbonate spring deposits of Pleistocene age, consisting of veryfine-grainedcarbonates with abundant reed casts, gastropod shells and algal structures, rise several tens of metres above the present land surface where the present active springs (and Recent spring deposits) occur. The higher, older spring carbonates cap circular mesas and hills and overlie pedestals of Cretaceous mudstones. Spring discharges have declined as a result of waterbore development in many parts of the Basin during the last 100 years. Springs are quite common in the recharge areas along the eastem margins, but most of these springs are the result of "overflow" or the "rejection" of recharge into the aquifers, or result from the intersection of the local topography and aquifers. 230

Vertical leakage from the aquifers upwards through the semi-pervious confining beds occurs throughout the Basin, and despite the low percolation rates, involve a considerable volume of water, which constitutes a major part of the groundwater throughflow of the Basin. A deep phreatic surface, usually at several tens of metres (60 to 80m) below the groundsurface, generally conceals the vertical leakage. Groundwater in the Great Artesian Basin has been exploited from flowing artesian waterbores since artesian water was discovered in 1878, allowing an important pastoral industry to be established. Waterbores are up to 2000m deep, but average about 500m. Flowsfromindividual bores exceed 10,000 mVday (more than 100 L/s), but the majority have much smallerflows.About 3100 of the 4700flowingartesian waterbores drilled in the Basin, remainflowing.The accumulated discharge of these waterbores is about 1.5 x lOW/day, compared to the maximumflowrate of about 2x10' mVday from about 1500flowingartesian waterbores around 1918. Flowing artesian bores obtain their groundwaterfromaquifers in the Lower Cretaceous and the Jurassic sequence (mainly the aquifers in the Cadna-owie Formation, Hooray, Algebuckina and Pilliga Sandstones and their equivalents). The original non-flowing artesian waterbores generally tap the aquifers in the Winton and Mackunda Formations. These non-flowing bores, which number about 20,000, are generally shallow i.e. several tens to hundreds of metres deep. It is estimated that these generally windmill-operated pumped waterbores supply on average 10 mVday. High initialflowrates and pressures of artesian waterbores have diminished as a result of the release of waterfromelastic storage in the groundwater reservoir, and approach a steady-state condition in many areas. Exploitation of the aquifers has caused significant changes in the rate of various discharges in time (Habermehl & Seidel, 1979; Habermehl, 1980; Seidel, 1980). Prior to development, the Basin was in a natural steady-state condition, with an equilibrium between recharge and natural discharge from springs and vertical leakage. Following development, natural discharge diminished. A visible effect has been the diminution inflowfromsprings in the south-central, southwestem and northem parts of the Basin. Abstraction by waterbores caused a steepening of the hydraulic gradient and allowed more recharge water to enter the system. At present a new approximate steady-state condition has been reached in which total recharge and discharge are approaching equilibrium again. Further decrease of the diminution of the artesian pressures and the reduction of the uncontrolled discharges are achieved during the last 7 years since the introduction of the Great Artesian Basin Bore Rehabilitation Program (Hillier et al., 1995). Completion of the Bore Rehabilitation Program could result in the better control and management of theflowingartesian waterbores, particularly if a Basin-wide management program is implemented. With fully controlled flowing artesian waterbores and reduced outflows, artesian pressures in most areas could increase significantly. The past and present distribution of artesian groundwater from the flowing artesian waterbores by open earth drains, which have lengths of many tens of kilometres, is extremely wasteful, owing to seepage, transpiration and evaporation of the water. Introduction of (polythene) piping to replace the earth drains will significantly reduce the demand on flowing artesian waterbores for groundwater, and could ahnost eliminate the wastage of water, if piping is combined with float valve controlled tanks and trou^ systems. Piping will also reduce the environmental effects caused by the introduction of large amounts of water and watering points in the semi-arid and arid landscape, including land degradation, the spread of introduced weeds, shrubs and trees, and greatly increased numbers of feral and native animals attracted by the water. Groundwater Movement The potentiometric surfaces of the confined aquifers in the Lower Cretaceous-Jurassic sequence were above the groundsurface over the whole of the Basin before exploitation began around 1880. Since then the regional potentiometric surface of the exploited aquifers in the sequence has dropped by several tens of metres in many heavily developed areas. It is still above groundlevel in most of the Basin, though in some areasflowsfromartesian waterbores ceased and water has to be pumped. The potentiometric surface of the confined aquifers in the upper part of the Cretaceous sequence has always been below the groundsurface, consequently waterbores tapping these aquifers are non-flowing artesian and have to be pumped. Potentiometric maps showing the conditions during the early years of development and the 1970s for the main aquifers in the Lower Cretaceous-Jurassic sequence, which produceflowingartesian wells, are given in Habermehl (1980), and were produced from the large number of measurements carried out on the flowing artesian waterbores since the early development of the Basin, and computer simulation modelling of the hydrodynamics of the Basin (Seidel, 1980). Development of the artesian groundwater resources has led to considerable changes in the pattems of the potentiometric contours, and the changes in waterlevels as a result of the regional drawdowns are shown in Habermehl (1980, Fig. 7). Predicted drawdowns, changes in discharges and predicted potentiometric surface maps as a result of possible future developments are shown in Habermehl & Seidel (1979) and Seidel (1980). A new computer simulation modelling study of the whole of the Basin is being carried out by the Austrahan Geological Survey Organisation at present (1996), and will refine and enhance the earlier work. 231

Hydraulic gradients of aquifers in the Lower Cretaceous-Jurassic sequence range from 1:2000 to 1: 4000 in the central-southwestern part of the Basin. Hydraulic gradients of the aquifers in the upper part of the Cretaceous sequence are about 1:1800. The confined aquifers in the Lower Cretaceous-Jurassic sedimentary sequence are sheet like deposits, which are relatively uniform, and extend for hundreds of kilometres. The aquifers are continuous across shallow ridges and platforms of older rocks. In some areas faults locally displace or disconnect aquifers, and obstruct part or all of the groundwater flow in the main Lower Cretaceous-Jurassic aquifers, which is normally directed to these structures. These faults could act as permeable or impermeable barriers, either to groundwater or to hydrocarbons migrating in the sandstones (Senior & Habermehl, 1980). Other impermeable barriers could occur in the aquifers, and be barriers of stratigraphic or diagenetic origin. Part of the Canaway Fault appears to be a preferential permeable zone along which groundwater movesfromJurassic aquifers upwards into Cretaceous aquifers (Habermehl, 1986). Regional groundwater movement in the Basin has been interpretedfromthe potentiometric surface maps of the aquifers in the Jurassic and Lower Cretaceous sequences (Figure 2). Flow directions are generally towards the south, southwest, west and north. In the westem part of the Basin regional groundwater movement is towards the southeast and south. Groundwater movement is slow, and based on hydraulic data probably around 1 m/year, as hydraulic conductivities and gradients are low and porosities high. The flow rates and groundwater residence times in the Lower Cretaceous-Jurassic Hooray Sandstone aquifer (and its equivalents) as calculatedfromhydraulic data are consistent with the residence times derivedfromenvironmental isotope studies on artesian groundwaterfromflowingartesian waterbores tapping this aquifer, carried out throu^out the Basin (Airey et al., 1979, 1983, Calf & Habermehl, 1984, Bentley et al., 1986, Torgersen et al., 1991, Herczeg et al., 1991, Habermehl et al., 1993, Cresswell et al., 1996). Residence times determined from carbon-14 and chlorine-36 studies range from several thousands of years near the recharge areas to more than one million years near the centre of the Basin. Hydrochemistry Groundwater in the most widely exploited confined aquifers in the Lower Cretaceous-Jurassic sequence generally contains about 500 to 1000 mg/L total dissolved solids. pH values of the artesian groundwater are ahnost uniformly between 7.5 and 8.5. The artesian groundwater is chemically of the Na-HC03-Cl type, and these ions contribute more than 90% of the total ionic strength of solutes in the main Basin area. Evolution of the groundwater chemistry along the flowlines is characterised by the removal of Na and K by reconstitution reactions involving kaolinite, a Na-smectite and illite (Herczeg et al., 1991). In the southwestem part of the Basin the groundwater is characterised by Na-Cl-S04 type water, and the two regional groundwater flow directions show different hydrochemical characteristics, with westward flowing water being of the Na-HC03-Cl type and eastwardflowingwater being of the Na-Cl-S04 type. Theseflows,within the same aquifer meet and mix, and are directed towards the main discharge area near the Basin's southwestem margin (Habermehl, 1986). Near the recharge areas Ca, Mg and SO4 concentrations are proportionally higher, but these decrease basinwards. Variations of the major ion concentrations and ratios occur along theflowlines.Na and HCO3 concentrations generally increase along theflowlinesin most parts of the Basin. CI and SO4 also increase in most areas, though an initial decrease occurs basinwards of the northeastern marginal area. Along the southerly directed flowpaths from the northeastem recharge area to the New South Wales border, concentrations of all four ions decrease, and then increase or remain approximately constant. In the centre of the Basin some deep waterbores with high CI concentrations occur, and very low to semi-stagnant flows along the deep and long flowpaths might account for these high CI concentrations and high salinity values. In the northwestem part Na, HCO3 and CI increase along the flowlines, but SO4 concentrations are constant (Habermehl, 1983, 1986). Total dissolved solids values generally show an increase downgradient in the Basin, and this is probably the result of mixing of the dilute recharge water with more saline groundwater in the deeper parts of the Basin, ion-filtration through mudstone membranes, and dissolution of evaporites, carbonate minerals or incongment dissolution of feldspars, micas or clay minerals (Herczeg et al., 1987,1991). The increase of HCO3 and decrease in SO4 concentrations in the Basin might result from biochemical reduction of carbon dioxide to produce methane rather than the dissolution of carbonate minerals. The aquifer system is open to CO2 and the addition of CO2 is accompUshed by fermentation processes occurring in situ. Some of the added CO2 is a byproduct of methanogenesis. The addition of CO2 drives the carbonate dissolution reaction, and so exerts an important control on the evolution of the Na-HCO^ groundwater within the Basin (Herczeg et al., 1991). Fluoride values in many parts of the Basin are high, with values up to 10 mg/L and more, which are a problem for domestic and stock water supplies. High fluoride concentrations in the artesian groundwater have been attributed to 232

groundwater being in contact with underlying basement rocks, in particular igneous rocks (Evans, 1995, Habermehl & Lau, 1993, Habermehl et al., 1996). Contact of the artesian groundwater with igneous rocks has been interpreted to be the reason for the occurrence of surprisingly unradiogenic ''Sif^Sr ratios (Collerson et al., 1988). Cretaceous aquifers have hi^er salinities and CI values than the Lower Cretaceous-Jurassic aquifers, and the high NaC1 values in the Cretaceous aquifers probably reflect the non-flushing characteristics of these mainly lenticular shaped aquifers. The marine origin of the adjoining mudstones in the Cretaceous sequence might also contribute. The separation of the Lx)wer Cretaceous-Jurassic aquifers from the Cretaceous aquifers based on their hydrochemical characteristics is quite possible, but the distinction between individual aquifers in the Lower Cretaceous-Jurassic sequence is less obvious (Muller, 1989, Quarantotto, 1986, 1989). Most artesian waterbores produce varying amounts of gases. The main constituents of the gases include N2, CO2 , Ar and small amounts of H2 and He. In addition many waterbores produce artesian groundwater containing small amounts of hydrocarbons. The hydrocarbons are mainly CH4 and lesser amounts of C2H6 to CyH^g but hquid hydrocarbon fractions have also been detected (Habermehl, 1986,1989). Hydrocarbon source and reservoir rocks are abundant in the sedimentary sequence of the basin, and commercial and subcommercial oil and gas discoveries have been made in several Jurassic and Cretaceous sandstones, contradicting earlier beliefs that the basin-wide groundwater flow hadflushedhydrocarbons out of the system. Water quality improves with the lower stratigraphic location of the aquifers in the Lower Cretaceous-Jurassic sequence, with groundwater obtained from aquifers in the older part of the Lower Cretaceous-Jurassic sequence having better quality water than the upper aquifer, which underlies the main confining bed of Early Cretaceous age and marine origin. Groundwater from all of the aquifers in the Lower Cretaceous-Jurassic sequence are of good quality and suitable for domestic, town water supply and stock use, though it is generally unsuitable for irrigation because in much of the Basin area it is chemically incompatible with the soils. Water from the upper. Late Cretaceous, aquifers has a higher sahnity, thou^ it is still acceptable as stockwater. Groundwater temperatures Groundwater surface temperatures of waterbores tapping aquifers in the Lower Cretaceous-Jurassic sequence generally range from about 30° to 100°C, and springs have temperatures from about 20° to 45°C. Geothermal gradients show a wide range and give a mean of about 3 9°C/km, and a range of about 15°C/kfn to 100°C/km (Polak & Horsfall, 1979), as obtained from temperature wireline logs in waterbores. References to earlier authors are shown in Habermehl (1980). Waterbore values for the geothermal gradients, though internally consistent, are too high, based on data from deeper petroleum exploration wells in the central part of the Basin (Cull & Conley, 1983). Geothermal gradients derived from petroleum exploration wells are given in Pitt (1986). The heatflowin the Basin is attributed to heat produced in the earth crust by uranium and thorium, and by recent volcanic activity (Torgersen et al., 1992). References AIREY, P.L., CALF, G.E., CAMPBELL, B.L., HABERMEHL, MA., HARTLEY, P.E., & ROMAN, D., 1979 -Aspects of the isotope hydrology of the Great Artesian Basin, Australia. In: Isotope Hydrology 1978, 1, p. 205-219. Proceedings International Symposium on Isotope Hydrology - Intemational Atomic Energy Agency and United Nations Educational, Scientific and Cultural Organisation, Neuherberg, Fed. Rep. Germany, 19-23 June 1978. International Atomic Energy Agency^ Vienna, 1979. AIREY, P.L., BENTLEY, H., CALF, G.E., DAVIS, S.N., ELMORE, D., GOVE, H., HABERMEHL, M.A., PHILLIPS, F., SMITH, J., & TORGERSEN, T., 1983 - Isotope hydrology of the Great Artesian Basin, Australia. In: Papers of the Intemational Conference on Groundwater and Man, Sydney, 5-9 December 1983. Australian Water Resources Council Conference Series, No. 8, vol. 1, p. 1-11. Australian Government Publishing Service, Canberra, 1983. BENTLEY, H.W., PHILLIPS, F.M., DAVIS, S.N., HABERMEHL, M.A., AIREY, P.L., CALF, G.E., ELMORE, D., GOVE, H.E., & TORGERSEN, T., 1986 - Chlorine 36 dating of very old groundwater. The Great Artesian Basin, Australia. Water Resources Research, 22 (13), p. 1991-2001. CALF, G.E., & HABERMEHL, M.A., 1984 - Isotope hydrology and hydrochemistry of the Great Artesian Basin, Australia. In: Isotope Hydrology 1983, p. 397-413. Proceedings Intemational Symposium on Isotope Hydrology in Water Resources Development, Intemational Atomic Energy Agency and United Nations Educational, Scientific and Cultural Organisation, Vienna, Austria, 12-16 September 1983. Intemational Atomic Energy Agency, Vienna, 1984. COLLERSON, K.D., ULLMAN, W.J., & TORGERSEN, T., 1988 - Ground waters with unradiogenic 87Sr/86Sr ratios in the Great Artesian Basin, Australia. Geology, 16, p. 59-63. 233

CRESSWELL, R.G., FIHELD, L.K., KEYWOOD, M.D., HABERMEHL, M.A., KELLETT, J.R., JACOBSON, G., & WISCHUSEN, J., 1996 - Theory, Rainfall and Recharge: A Chlorine-36 Story. In: American Geophysical Union, 1996 Western Pacific Geophysics Meeting, Brisbane, 23-27 July 1996 CULL, J.R, & CONLEY, D., 1983 - Geothermal gradients and heat flow in Australian sedimentary basins. BMR Journal ofAustralian Geology and Geophysics, 8, p. 329-337. EVANS, PA., 1996 - Fluoride anomalies in aquifers of the Queensland section of the Great Artesian Basin and their significance. Mesozoic Geology of the Eastern Australia Plate Conference, Brisbane HABERMEHL, M.A., 1980 - The Great Artesian Basin, Australia. BMR Journal of Australian Geology & Geophysics, 5, p. 9-38. HABERMEHL, M.A., 1982 - Springs in the Great Artesian Basin, Australia - their origin and nature. Bureau ofMineral Resources, Australia, Report 235, 50 pp. HABERMEHL, M.A., 1983 - Hydrogeology and hydrochemistry of the Great Artesian Basin, Australia. Invited Keynote Paper. In: Papers of the International Conference on Groundwater and Man, Sydney, 5-9 December 1983. Australian Water Resources Council Conference Series, No. 8, vol. 3, p. 83-98, Australian Government Publishing Service, Canberra, 1983. HABERMEHL, MA., 1986 - Regional groundwater movement, hydrochemistry and hydrocarbon migration in the Eromanga Basin. In: GRAVESTOCK, D.I., MOORE, RS., & PITT, G.M., (Editors), 1986 - Contributions to the geology and hydrocarbon potential of the Eromanga Basin. Geological Society of Australia Inc. Special Publication No. 12, p. 353-376. HABERMEHL, M.A., 1989 - Hydrogeology of the Great Artesian Basin. In: BMR89 - Yearbook of the Bureau of Mineral Resources, Geology and Geophysics Australian Government Publishing Service, Canberra, p. 115 -117. HABERMEHL, MA., ALLAN, G.L., FIFIELD, L.K., & DAVIE, R.F., 1993 - C1-36/C1 Ratios in the Great Artesian Basin, Australia. In: 6th International Conference on Accelerator Mass Spectrometry, Canberra - Sydney, 27 September - 1 October 1993, Abstract Volume, p. 64. HABERMEHL, MA. & LAU, J.E., 1993 - Fluoride in groundwater of Cape York Peninsula. In: Papers of the Conference Aquifers at Risk: Towards a National Groundwater Quality Perspective, Canberra, 13-15 February 1993, Abstract AGSO Journal ofAustralian Geology and Geophysics, 14 (2 & 3), p. 316 - 317. HABERMEHL, M.A., LAU, J.E., MACKENZIE, D.E., & WELLMAN, R, 1996- Sources of fluoride in groundwater in North Queensland, Australia. In: 13th Australian Geological Convention, Canberra, 19-23 February 1996 Geological Society ofAustralia Abstracts No. 41, p. 176. HABERMEHL, M.A., & SEIDEL, G.E., 1979 - Groundwater resources of the Great Artesian Basin. In: HALLSWORTH, E.G., & WOODCOCK, J.T., (Editors), 1979 -Proceedings of the Second Invitation Symposium Land and Water Resources of Australia-Dynamics of utilisation, Australian Academy of Technological Sciences, Sydney, 30 October-1 November 1978. Australian Academy of Technological Sciences, Melbourne, p. 71-93. HERCZEG, A.L., TORGERSEN, T., CHIVAS, A.R., & HABERMEHL, MA., 1987 - Geochemical evolution of groundwaters from the Great Artesian Basin, Australia. American Geophysical Union 1987 EOS Transactions American Geophysical Union 68 (44), p. 1275-1276. HERCZEG, A.L., TORGERSEN, T., CHIVAS, A.R., & HABERMEHL, MA., 1991 - Geochemistry of ground waters from the Great Artesian Basin, Australia. Journal of Hydrology, 126, p. 225-245. HILLIER, J.R., HAZEL, C.R, WILLIAMS, M.R., HARRIS, B., JOLLY, R, & HABERMEHL, MA., (Interstate Working Group on the Great Artesian Basin), 1995 - The Great Artesian Basin - Technological Advances. In: Proceedings of the 16th Federal Convention of the Australian Water and Wastewater Association Inc. -Delivering the Vision for the Next Century, Sydney, 2-6 April 1995, Volume 2, p. 245-251. MULLER, P.J., 1989 - Aspects of the hydrogeology of the southem Eromanga Basin, Queensland. In: O'Neil, B.J. (editor), 1989 - The Cooper and Eromanga Basins, Australia. Proceedings of the Cooper and Eromanga Basins Conference of Petroleum Exploration Society ofAustralia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA Branches), Adelaide, 26-27 June 1989, p. 493-505. PITT, G.M., 1986 - Geothermal gradients, geothermal histories and the timing of thermal maturation in the EromangaCooper Basins. In: GRAVESTOCK, D.I., MOORE, RS., & PITT, G.M., (Editors), 1986 - Contributions to the geology and hydrocarbon potential of the Eromanga Basin. Geological Society of Australia Inc. Special Publication No. 12, p. 323-351. POLAK, E.J., & HORSFALL, C.L., 1979 - Geothemaal gradients in the Great Artesian Basin, Australia. Bulletin of the Australian Society of Exploration Geophysicists 10, p. 144-148. QUARANTOTTO, P., 1986 - Hydrogeology of the southeastem Eromanga Basin, Queensland. Geological Survey of Queensland Record 1986/38, 58 p. QUARANTOTTO, P., 1989 - Hydrogeology of the Surat Basin, Queensland. Geological Survey of Queensland Record 1989/26, 34 p. SEIDEL, G.E., 1980 - Application of the GABHYD groundwater model of the Great Artesian Basin, Australia. BMR Journal ofAustralian Geology & Geophysics, 5, p. 39-45. SENIOR, B.R., & HABERMEHL, MA., 1980 - Structure, hydrodynamics and hydrocarbon potential of the Central Eromanga Basin, Queensland, Australia. BMR Journal ofAustralian Geology & Geophysics, 5, p. 47-55. 234

TORGERSEN, T., HABERMEHL, M.A., PHILLIPS, EM., ELMORE, D., KUBIK,R, JONES, B.C., HEMMICK, T., & GOVE, H.E., 1991 - Chlorine-36 dating of very old groundwater 3. Further studies in the Great Artesian Basin, Australia. Water Resources Research, 27 (12), p. 3201-3213. TORGERSEN, T., HABERMEHL, M.A., & CLARKE, W.B., 1992 - Crustal helium fluxes and heat flow in the Great Artesian Basin, AustraUa. Chemical Geology (Isotope Geoscience Section), 102, p. 139-152. WOODS, RH., WALKER, G.R., & ALLISON, G.B., 1990 - Estimating groundwater discharge at the southem margin of the Great Artesian Basin near Lake Eyre, South Australia. Proceedings of the Intemational Conference on Groundwater in Large Sedimentary Basins, Perth, 9-13 July 1990 Australian Water Resources Council Conference Series No. 20, p. 298-309.

Figure 1. Location and extent of the Great Artesian Basin and its constituent sedimentary basins, intermediate ridges and underlying basins.

235

22°-

f 30°-

Confining Cretaceous

200km

bed aquifer Boundary between

Confining

bed

Jurassic

aquifer

Basement

confined

Potentiometric

surface Jurassic

aquifer,

1880

Potentiometric

surface Jurassic

aquifer,

1970

Petroleum

exploration

Watertable

Direction

of groundwater

Fault

Spring

aquifers and confining

— 200— Topographic

beds

contour (m) above

MSL

Concentration

well

of springs

flow

19/A/207

Figure 2. The Great Artesian Basin, extent and cross-section of the simplified hydrogeological units, including the Late - Early Cretaceous confined aquifer ("Cretaceous" aquifer) and the Lower Cretaceous-Jurassic confined aquifer ("Jurassic" aquifer) sequences. The directions of the regional groundwater flow in the "Jurassic" aquifer, as determined from potentiometric contour maps, and the locations of flowing artesian springs, are shown.

236

APPLICATION FOR MAGNETOTELLURICS IN PETROLEUM EXPLORATION Lynn Hastie Physics Department, The University Of Queensland Summary Natural field magnetotelluric methods have long been used in the regional study of sedimentary basins, however the data have often not been capable of providing the sort of accuracy that is necessary to trace structures of interest in petroleum exploration. This paper points out some of the problems that can occur due to noisy data and the depth-dependent resolution of the method, and highlights typical methods of minimising such problems. Introduction Natural field miagnetotelluric (MT) methods have traditionally been used in sedimentary basin investigations, the seminal paper in this regard being that by Vozofif (1). Natural field MT rehes on eddy currents induced in the ground by fluctuations in the surface electromagnetic fields. Impedance properties of the Earth may be deduced firom the field strengths due to these induced electric currents relative to their inducing magnetic fields (Cagniard, (2)). Both reflection and refi-action occurs at the Earth-air interface with the refiracted component propagating normally downwards due to the higher refiractive index of the ground con5)ared to the atmosphere, so that interpretation in the magnetotelluric method is based on vertically transmitted signals. If such a method is to be valuable in petroleum exploration it must be capable of resolving and tracing deep marker horizons. It is potentially capable of this, however care must be taken in the data recording and basic processing methods in order to overcome the typically high noise levels, which are particularly severe in tropical regions. The Principle of the MT Method In MT, the noise-Hke magnetic fields are used as the input signal and the electric field is measured as the output signal that has been modified by the underlying conductivity structure. If we consider recording with sensors ahgned with a horizontal orthogonal coordinate system represented by x- and y-coordinates.

ZxxCf) z^if) 'HM) (1)

where / is a firequency, the are electric field components, the H^^ are magnetic field con^nents and the Zjj are electromagnetic (EM) input impedances components which are the basic ground response parameters. These are usually expressed in temis of the apparent resistivity and phase, as the stmctural modelling is in terms of a distribution of resistive bodies.

For a one-dimensional layered model,

where =

(3)

and paff) is the apparent resistivity at thefirequency/ The apparent resistivity (the resistivity of an equivalent half-space atfirequencyJ), and the impedance phase (pij) are the quantities that are being estimated for use in modelling the resistivity structure. A number of very effective one-dimensional (Whittall and Oldenburg (3)) and two-dimensional modelling programs (Smith and Booker (4)) exist and are readily available. Three-dimensional modelling can be carried out, however it is currently not very effective in practice. Effects of Noise MT data, particularly the electric field recordings, are very noisy. This is a particular problem in the tropics and all measures have to be taken to enable satisfactory data to be recorded in a reasonable length of time. Remote reference recording, where a magnetic recording is made at a site that is sufficiently distant that the local noise is independent of the main recording site is essential, however the noise on the electric field is still a problem. In order to investigate the effect of noise on the estimates obtainedfiromMT recording it is usefiil to examine an artificial example. Consider the two-dimensional structure shown in Figure 1(a). The apparent resistivity and phase this generates, corresponding to a recording of the electric field in the cross-strike direction illustrated in Figure 1(a), are shown iti Figure 1(b), and are also reproduced as reference curves in Figures 2, 3 and 4. The latter three curves were generated by 237

passing an artificial MT signal consisting of random pulses of noise-like signal through the 'filter' corresponding to the model, and adding unifomi white noise to the result. This has then been processed using two standard data processing methods, to reduce the data to apparent resistivity and phase curves using increasing data lengths. The two methods will be referred to here as the conventional (CONVMT) and robust remote reference (RRRMT) methods. The CONVMT method is essentially that proposed by Madden and Nelson (5) and further developed by Vozoff (1), but with a robust estimator used instead of the mean. The RRRMT method is a more statistically sophisticated method based on robust M-estimators, and was developed by Chave et al (6). In progressing from Figure 2 to Figure 4 the data length used changesfirom4096 to 20480 to 81920. The basic 4096 length is chosen to provide estimates spanning two decades and is used in each case as the transform length. The additional data was used to provide redundancy so the results in Figure 2 are based on one 4096 point section of data, in Figure 3 on averaging the results from 5 sections and in Figure 4 on averaging the results from 20 sections. The signal to noise ratio on the electric channel is 0 dB (1:1) while that on the magnetic field recording is 20 dB (10:1). This is considered to correspond to typical expectations for MT. The results provided by both methods can be seen to improve dramatically with the increase in data length, however it can be seen that CONVMT has resulted in resistivity estimates that are biasedfi-omthe expected result while the RRRMT estimates are unbiased. Thus this example illustrates two points, the data sequence to provide reasonable estimates over two decades should be at least 80,000 long, and the data reduction method chosen does affect the bias present in the estimates. It should be noted that the actual frequencies shown and the time length represented by the data sequence depend only on the sampling rate. In the example the sampling rate is 10,000 samples per second which means that the 81920 sequence represents 8.2 seconds, however if the objective was to examine deeper structures the sampling rate may be 10 samples per second and then the sequence would represent 2 hours 17 minutes, but the uncertainties and biases would be the same. Depth Sounding and Resolution MT is often perceived as an EM equivalent of seismic surveying. This point of view has some validity, however it also leads to a number of misconceptions as to the interpretation of EM results. The EM signal propagates very rapidly, and in MT is considered to be established immediately, so there is no time-of-fiight measurement. Thefieldis very strongly attenuated as it propagates down into the Earth, so high frequency fluctuations are progressively removed from the signal. Thus a particularfirequencyonly penetrates to a certain depth, typically represented by the skin depth (see Figure 5), and a reconstruction algorithm can build up a model of the conductivity structure versus depth because lower frequencies penetrate to progressively greater depths. However the lowfirequencycomponents will average over an increasing larger volume of the sub-structure causing the resolution to necessarily worsen. This is very different from the situation in seismic surveying. As the EM field is strongly attenuated, estimates at low frequencies have to be very accurate to maintain sensitivity, however the power present in the natural field increases with period, and partially offsets the more stringent accuracy requirement. Now, because the penetration depth increases as the square root of the period (see Figure 5) it is necessary to increase recording times dramatically in order to estimate the power at the low frequencies to obtain extra depth information. This increase is amplified by the necessity to record greater amounts of data to enable sufficient averaging to be carried out to overcome the noise present in the signal. Roughly speaking, to estimate the power at afrequencyin a low noise signal to one percent, ten periods are needed, and up to one hundred are required to achieve comparable accuracy in noisy situations such as typical MT signals. Because the lower frequencies penetrate the shallow structure, the modelling accuracy of deeper stmctures is strongly affected by the accuracy of the shallower structures, especially if these are highly conducting. Thus it is important for the shallow structures to be well determined in order to avoid bias in the deeper model. This consideration, together with the occurrence of surface channelling of local electric currents, means that the use of shallow exploration methods such as dc-resistivity, TEM or CSAMT to resolve this surface structure is very desirable, if high quality MT results are needed. In order to demonstrate the depth resolution problem, the model shown in Figure 6(a), consisting of a 200 metre, 10 Dm layer, overlying a 100 Qm half-space in which there is a 100 metre layer at 2 kilometres depth, was processed using the RRRMT algorithm and the results are shown in Figure 7. The data were generated with the same noise levels as for the example shown in Figures 2 to 4, and 20 sections of 4096 long data sequences were averaged, as for Figure 4. The apparent resistivity and phase for the model are shown in Figure 6(b) for two cases. The soUd line corresponds to the expectation when the deep layer is absent, while the dashed hne is the model with the deep layer present. These curves shown in Figure 6(b) are reproduced as reference curves in Figure 7. The presence of the layer can be seen to have caused a reasonably subtle change in the expected results. It can be seen that the difference in the phase response is considerably greater than that in the apparent resistivity curves, and it is quite obvious that the apparent resistivities calculated from the data resolve the layer quite well. However it can also be seen that the extra accuracy and lack of bias in the RRRMT method was necessary, and that the amount of averaging could not have been significantly reduced if the layer were to be resolved. 238

Conclusions MT is a relatively cheap method for cnistal sounding that has a resolution which reduces with depth. The main problem is the lack of predictability of the data quality that can cause unexpected increases in the time that is necessary for carrying out a survey. However with proper care taken in the data recording and in choosing the most effective available method to reduce the field data to impedance estimates, the results can be very useful in for gaining the type of information required in petroleum exploration, where environments approximate such one and two-dimensional situations. While the current situation is satisfactory, processing and modelling methods are continually improving. References 1. 2. 3. 4. 5. 6.

Vozoff K. The magnetotelluric method in the exploration of sedimentary basins. Geophysics, 37, 1972, 98-141 Cagniard L. Basic theory of the magnetotelluric method. Geophysics, 18,1953, 605-635. Whittall P. and Oldenburg D.W. Inversion of magnetotelluric data for a one-dimensional conductivity. S.E.G. Geophysical Monograph Series, 5, 1992. Smith J.T. and Booker J.R. Rapid inversion of two and three-dimensional magnetotelluric data. 1 Geophys. Res., 96, 1991, 3905-3922. Madden T. and Nelson P. A defence of Cagniard's magnetotelluric method. Project NR-371'401, Office of Naval Research, Geophysics Laboratory, MT, Boston, Mass., 1964. Qiave A.D., Tliompson D.J. and Ander M.E. On the robust estimation of power spectra, coherences, and transfer functions. J. Geophys. Res., 92,1987, 633-648.

Figure Legends Figure 1. (a) A single two-dimensional model of a 10 Hm layer overlying a vertical contrast between 100 Qm and 10 Qm material is used to illustrate the advantage of averaging, when reducing data, to overcome the effects of noise. (b) The apparent resistivity and impedance phase corresponding to the electric field measurement in the cross-strike direction of the model shown in (a). Figure 2. The apparent resistivity and phase for a single 4096 long section of data generated from the model in Figure 1(a), and processed using CONVMT and RRRMX The solid smooth line in each plot is the curve shown in Figure 1(b) and is the ideal result. The error bars represent two standard deviation confidence Umits and the estimates are joined by strai^t line segments. Figure 3. The apparent resistivity and phase forfive4096 long sections of data generatedfiromthe model in Figure 1(a), and processed using CONVMT and RRRMT. The solid smooth line in each plot is the curve shown in Figure 1(b) and is the ideal result. The error bars represent two standard deviation confidence limits and the estimates are joined by straight line segments. Figure 4. The apparent resistivity and phase for twenty 4096 long sections of data generatedfiromthe model in Figure 1(a), and processed using CONVMT and RRRMT. The solid smooth line in each plot is the curve shown in Figure 1(b) and is the ideal result. The error bars represent two standard deviation confidence limits and the estimates are joined by straight line segments. Figure 5. Electromagnetic skin depth (d) plotted against signalfi-equency(f), and the resistivity (r) of the half-space being penetrated by the signal. The relationship is given by. 5= 'i 4 ; r " x l O - 7 Figure 6. (a) A one-dimensional model with a conductive layer at depth to illustrate the problems with resolving details at depth. (b) The apparent resistivity and impedance phase curves corresponding to the model shown in (a). The solid curves are the results when the layer is not present and the dashed curves are the results when the layer is present. Figure 7. The apparent resistivity and phase for twenty 4096 long sections of data generatedfiromthe model in Figure 6(a), and processed using RRRMT. The solid and dashed smooth curves are the curves from Figure 6(b) and represent the ideal result for the model respectively, without the deep layer and with the deep layer. The error bars represent two standard deviation confidence limits and the estimates are joined by straight line segments. The processed results can be seen to have clearly resolved the layer.

239

AIR Om 10 n m 100 m

loonm

10 a m

(a)

Resistivity

-0.5

1.5

Log Frequency (Hz) (b) Figure 1.

240

Resistivity - CONVMT Method

Resistivity - RRRMT Method

2.5

3.5

2 2.5 3 Log Frequency (Hz) Phase - RRRMT Method

3.5

Log Frequency (Hz) Phase - CONVMT Method 80 i60 bO

^40 ^20

1.5

2.5

Log Frequency (Hz)

Resistivity - CONVMT Method

2.5

3.5

2.5

3.5

Log Frequency (Hz) Phase - CONVMT Method

Figure 2.

2.5

Log Frequency (Hz)

3.5

Resistivity - RRRMT Method

2.5

Log Frequency (Hz) Phase - RRRMT Method

3.5

^60 E u00 Q :4o C3 JZ cu 20

Log Frequency (Hz)

Figure 3.

Log Frequency (Hz)

241

Resistivity - CONVMT Method

Resistivity - RRRMT Method

2 2.5 3 3.5 Log Frequency (Hz) Phase-CONVMT Method

2 2.5 3 Log Frequency (Hz) Phase-RRRMTMethod

2 2.5 3 Log Frequency (Hz)

Figure 4.

Figure 5.

242

AIR Om 200 m

10 Q m

lOOam

10 Clm

2000 m 2100m (a)

Resistivity

-0.5

0.5

Log Frequency (Hz) Phase

-0.5

0.5

Log Frequency (Hz) (b)

Figure 6.

243

Resistivity

-0.4

-0.2

0.4

0.6

0.8

Log Frequency (Hz) Phase

-0.4

-0.2

0.4

0.6

0.8

Log Frequency (Hz)

Figure 7.

244

1.6

SEQUENCE STRATIGRAPHIC AND PALAEONENVIRONMENTAL INTERPRETATION OF THE BATHURST ISLAND GROUP, MONEY SHOALS BASIN Bob Henderson James Cook University of North Queensland The Cretaceous System of the Money Shoals Basin consists of the Bathurst Island Group which ranges in thickness to some 700m. In the southern part of its distribution it is nonconformable on Proterozoic basement but to the north it is disconformable on older Mesozoic systems. Four component formations are recognised. The oldest of these comprise the Darwin and the MarUgar Formations which are lateral facies equivalents and interpreted as a transgressive systems tract. A thin quartzose basal conglomerate overlying the nonconformity is typically developed and represents shoreface sedimentation on a pediment surface of very low relief. An interval of glauconitic sandstone commonly follows. X-ray diffi-action and micro-probe analyses indicate that the glauconite is structurally and compositionally mature, implying that this stratum represents a condensed sequence. The bulk of the Darwin Formation consists of radiolarite, mudstone and bioturbated quartz sandstone of the shallow shelfal origin. It contains little benthic fauna suggesting that turbid bottom conditions prevailed. The radiolarite is interpreted as the product of nutrient upwelling driven by wind-generated currents from the shelf edge, 350km to the north. Top of the of Darwin Formation is marked by a widely developed nodular phosphorite horizon interpreted as an omission surface marking the maximum flooding. A thick succession of Wangarlu Mudstone follows, interpreted as a high-stand systems tract and it is overlain by glauconitic and quartzose sandstone of the Moonkinu Formation which is considered to represent a prograding offlap systems tract developed during regression. In broad terms, the Bathurst Island Group is consistent with a global second-order eustatic cycle recognised by Vail and his co-workers. Transgression commenced in the late Albian and regression was complete by the mid- Turonian. This eustatic cycle contrasts strikingly with the sequence apparent for the Great Artesian Basin which abuts the Money Shoals Basin. Here the late Albian Toolebuc Limestone records the same transgressive episode but regression and emergence was complete in the Cenomanian. It is considered that sediment supply exceeded the capacity for accommodation for the Great Artesian Basin, forcing regression that was out of cycle wi& stratigraphic systems controlled by eustacy.

245

MESOZOIC REGOLITH AND PALAEO-LANDSCAPE FEATURES IN SOUTHEASTERN AUSTRALIA S.M. Hfll Centre for Australian Regolith Studies, do The Department of Geology, The Australian National University, Canberra, ACT, 0200 Summary The long term survival of regolith and landscape features of Mesozoic origins in southeastern Australia is supported by the sedimentary record, pre-volcanic remnants, the oxygen isotope signature of secondary minerals and associated geomorphological evidence. Their preservation places significant constraints on interpretations of the late Mesozoic to early Cainozoic geology and landscape history of the region, suggesting that denudation was restricted. Apparent discrepencies between landscape based geological interpretations and the interpretations of widespread kilometre-scale denudation in the late Mesozoic from Apatite Fission Track results need to be resolved with an appreciation for the local variations in landscape denudation and inherent differences in the basis of these techniques. Bedrock lithology and structure are major controls on restricting denudation, allowing for the preseservation of these features by influencing local erosional baselevels and, in the case of volcanic and sedimentary burial, providing protective cappings. INTRODUCTION Australian landscapes are well known for their antiquity. Geomorphological and geological accounts of Australian landscape evolution mostly extend back in time to beyond the Quatemary, frequently extending into the mid and early Cainozoic and in some cases into and beyond the Mesozoic. The suggested implications of the preservation of these palaeolandscape features have been of long-term landscape stability and low denudation rates. Their recognition in southeast Australia is therefore important for models of the geological and landscape evolution since these times. The models must account for the preservation of these features within the context of the established regional geological and landscape history. There is a need for a synthesis of the evidence for the preservation of Mesozoic regolith and landscape features, particularly in light of new regional data and models for the geological and landscape evolution. The region considered here is limited to southeastem mainland Australia, particularly focusing on Victoria but also extending into the southern parts of New South Wales. MESOZOIC REGOLITH AND LANDSCAPE FEATURES Early Observations Early debate focussed on the age of the upland palaeosurfaces and their subsequent correlation and stratigraphic significance. Most early interpretations suggested relatively young ages for landscape features, such as Pliocene/Pleistocene or Late Tertiary. This was derived from the assumption that the highland surface represented an uplifted Miocene ''peneplain" of continental extent ("The Great Peneplain"), and also the "youthful" appearance of coastal gorges by analogy with relatively young northem hemisphere landscapes (e.g. Andrews, [1], p.457). In contrast, some early reports proposed much greater landscape antiquity. Hart (2) recognised that the oldest sediments overlying the "peneplain" were early Tertiary in age and therefore the hi^er, older, upland surfaces were related to Mesozoic denudation. Skeats (3) identified two "peneplains" in Victoria, the highest of which defined the simimits of Mt Macedon and the Dandenong Ranges predating the Middle Tertiary "Older Volcanics". Working in the Aberfeldy - Mt Baw Baw area of Victoria, Baragwanath (4) identified two pre-Older Basalt surfaces, one directly underlying and immediately pre-dating the basalt residuals and the other ("Older Peneplain") 1,000 - 2,000 feet higher defining the Baw Baw Plateau. F.C. Craft emphasised the great age of the landscape showing that some of the river systems of the central and southem tablelands of NSW originated in the Middle Triassic (5; 6) and parts of the landscape of the Monaro region pre-dated the Cainozoic (7). Greater antiquity was gradually more widely accepted. Professor E.S. Hills (8;9) used mainly Triassic and Early Tertiary sub-volcanic and sub-Cretaceous sedimentary evidence (see following discussion) to promote and estabUsh the recognition of a Trias-Jura "palaeoplain" forming an upland surface across much of the Victorian hi^lands and into southem NSW. One of the main instigators of later interpretations favouring landscape antiquity was the development of better chronological controls, such as the radiometric dating of basalts and the refinements in the regional stratigraphic framework. Observations of Mesozoic palaeosurfaces have since been made in areas including: i) Cretaceous, pre-basaltic drainage in the Kiandra area (10), ii) the Healsville-Warburton region and the Dandenong Ranges (11;12), iii) the Cobberas and the Nuniong Plateau, Mt Gibbo, the Bogong High Plains, Mt Skene, the Baw Baw, Wellington and Buffalo Plateaux (13; 14; 15), iv) highland surfaces and accordant summits in the West Victorian Highlands, such as the Mt Cole-Mt Buangor plateau, Mt Macedon and Langi Ghiran (15; 16), v) The Gippsland Basin margins, such as at Wilsons Promontory (17; 18; 19).

246

Sedimentary Evidence Early Cretaceous Sediments The Early Cretaceous sediments in the Gippsland and Otway Basins (the Strezlecki and Otway Groups respectively) are essentially non-marine, feldspathic, volcaniclastic, sandstone, mudstone, shale, and basal conglomerates largely derived from a volcanic complex at the eastern margin of the Gippsland Basin (20). This is a period characterised by minimf)] sedimentation of materials derived from the present basin margins suggesting that these areas were characterised by relatively low erosion rates. Within these basins and in parts of the basin margins, where the rates of erosion were less than the rates of weathering, deep mantles of weathered bedrock developed. Examples include basement highs often composed of resistant granitic lithologies such as at Wilsons Promontory and Cape Woolami on Phillip Island (19). Many of these ancient landscape features would have undergone some post-Cretaceous modification by weathering and erosion, although their basic form must have developed during the Mesozoic. This is supported by the often fresh nature of the Cretaceous sediments overlying and flanking weathered bedrock and also by the contribution of weathered lithologies to the midCretaceous and early Tertiary sediments. Late Cretaceous Sediments During the mid-Cretaceous there was a major change in sedimentation pattem and tectonics in parts of southeastem Australia that has been related to the major evolution of the Eastern Highlands (20; 21; 22). Late Cretaceous sediments limited to the Otway (Sherbrooke Group) and Gippsland (Latrobe Group) Basins and also within the Ceduna Depocentre, consist of quartzose sandstones, siltstones and shale derived from denudation of weathered highlands along the basin margins (20; 17). The dramatic change from essentially Early Cretaeous volcaniclastic sediments to Late Cretaceous terrigenous sediments signifies the onset of erosion of the Mesozoic regolith and landscape. Although these sediments have been linked with widespread denudation (e.g. 20) the limited extent of these sediments is a problem for applying this interpretation across all of southeastem Australia (e.g. as discussed by Branagan [23]). Early Tertiary Sediments Early Tertiary sedimentation within or adjacent to the highlands in southeastem Australia reflects the progressive erosion of a deeply weathered landscape. Early Tertiary terrestrial sediments frequently overlie eroded (and therefore preexisting) weathering profiles. For example, Palaeocene and pre-Palaeocene sediments and Eocene basalts in the Monaro overlie truncated ferruginous, h i ^ y weathered palaeosurfaces (24; 25). The lithology of many of these sediments, containing an abundance of resistant quartzose lithologies, reflects a derivation from a highly weathered landscape (25). The lithologies of other Cretaceous to early Tertiary, buried (such as in sub-basaltic and sub-sedimentary deep leads) and high-level gravels, in the southern Eastem Highlands support the presence of a highly weathered landscape pre-dating the Cainozoic (18; 26; 27). In the Colac Trough, north of the Otway Ranges, the Aptian - Albian sequence is truncated and overlain by Palaeocene sediments representing mid-Cretaceous uplift and denudation (29). In the Broken Hill region, and along the westem part of the Canobolas Divide (27), early Tertiary, Murray Basin sediments (e.g. The Warina Sand), that overlie truncated deep weathering profiles, have quartzose and kaohnitic lithologies consistent with the stripping of pre-existing weathering profiles. North of Broken Hill in the Callabonna Sub-basin, the early Tertiary Eyre Formation has similar features (29; 30). Pre- and Sub-Volcanic Regolith and Palaeosurfaces Perhaps the oldest volcanic extrusions with afifiities with the present landscape in southeastem AustraUa are the Triassic volcanics near Benambra, north-eastem Victoria. Trachytic lavas and tuffs have been stratigraphically and radiometrically dated as Triassic (31; 32; 33). The surface on which they rest must therefore pre-date the Triassic, and adjoining areas are possibly modified relicts of a palaeosurface originating at least by this time (the Trias-Jura palaeoplain of Hills [9]). Although some of these volcanics and associated sediments have been shghtly disrupted by later intrusion of related magmas (Meredith Orr, personal communication, 1996), the presence of these volcanics at the present landsurface supports the notion of a Triassic landsurface approximating the present landsurface and of at least local landscape stability since the Triassic. Early Cretaceous lavas occur on the coastal plain near Mt Dromedary between sea level and approximately 70m a.s.L and also at approximately 780m a.s.l. on the flank of Mt Dromedary (34). If these lavas do infact conformably rest on the present coastal plain surface then it appears that the basic framework of the Eastem Highlands terrain had developed by the Early Cretaceous in at least the Mt Dromedary region (34). Further work is needed to determine the regional extent and significance of these interpretations. "Older Volcanics" lavas have been the most widely used benchmark in landscape studies in the Eastem Highlands. The lavas have been shown to range from Late Cretaceous to Late Miocene in age (35; 36). Regolith and related palaeosurfaces beneath at least the oldest of these basalts would in many cases date back to Mesozoic origins. Present valleys as well as relief inverted surfaces containing Early Tertiary basalts have also been used to recognise higher-level, older landscape features. For example Skeats (3), Baragwanath (4), Singleton (14) and Hills (9) identified apparent erosion surfaces at higher levels than surfaces capped by "Older Volcanics" in Victoria. Extrusion of Eocene basalts in the Endrick River valley, west of Ulladulla, show that dissection of the highland surface was underway at least by the beginning of the Early Tertiary (37). Young (37) extended this observation by extrapolating post-basaltic rates of valley247

side retreat to determine a Mesozoic (possibly Early Cretaceous or Jurassic) age for the onset of valley incision and consequently at least the equivalent antiquity of the plateaux surfaces. Oxygen-Isotope Evidence Bird & Chivas (38; 39; 40) interpreted several kaolinite samples from southeastem Australia as having a pre-Mid Cretaceous age. They obtained low 5180 values, indicative of pre-Late Mesozoic weathering, from weathering profiles in the Brown Mountain - Bombala region and also on the coastal plain at Jervis Bay. Bird & Chivas (40) and Nott & Purvis (34) used the preservation of ancient weathering products on the coastal plain to support the development of the coastal plain before Cretaceous, followed by minimal erosion. Pre-Late Mesozoic 5180 signatures have also been recognised in kaolinite samples from Tertiary sediments in southeastem Australia, indicative of erosion and redeposition of clays originally derived from widespread pre-mid Cretaceous weathering profiles (40). Apatite Fission Track (AFT) Evidence Studies by Moore et al (41), Duddy & Green (42), Dumitru et al (43) and Foster & Gleadow (44) reported a regular pattem of apparent apatite ages across the southeastem margin of Australia dominated by a trend of rapidly decreasing ages towards both the eastem and southem continental margins. Kohn & Gleadow (45) and O'Sullivan et al (21) have used AFT interpretatios to suggest kilometre-scale (at least l-2km) denudation of the Eastem Highlands at 90±10Ma (Mid-Cretaceous). These authors highlighted the corresponding change in sedimentation pattem in southeastem Austraha and also the timing of rifling prior to breakup and opening of the Tasman Sea. AFT studies on basement highs along the northem margins of some of the Bass Strait Basins, such as Wilsons Promontory and the Momington Peninsula in southem Victoria, suggest that they were buried to depths of 1.5-2km during the Early Cretaceous followed by denudational cooling (42). These interpretations appear to be at odds with many of the previously discussed evidence for landscape antiquity and stability. If this kilometre scale denudation was as great and also as widespread as AFT interpretations suggest then the preservation potential of these ancient landscape remnants would be low. The implications of this apparent discrepency needs to be considered further. DISCUSSION: Preservation of ancient landscape features and regional synthesis R^ional Synthesis and Implications There is a general consensus among geological and palaeolandscape interpretations in southeastem Australia supporting limited denudation and the preservation of a hi^ly weathered landscape in the Mesozoic up until the mid and late Cretaceous. After this time the landscape-based interpretations suggesting continued landscape stability and minimal denudation (e.g. 34; 28), and AFT interpretations of widespread, kilometre-scale denudation (e.g. 21; 45) appear to be in conflict. Rather than having to accept one interpretation in preference to the other, there is an argument for considering the vaUdity of both. As discussed by Branagan (23) for the Sydney Basin area, and is also apphcable to much of this region, it is difficult to account for great depths of denudation by invoking removal of thick, late Mesozoic sedimentary cover across aU of southeastem Australia. The preservation of these Mesozoic palaeo-landscape remnants after kilometrescale exhumation is therefore possibly only applicable to a few localities (at most only on the margins of the Bass Strait basins). Some of the Mesozoic regolith is also etched remnants from formerly more extensive weathering profiles. Although weathering profiles of several hundred metres depth have been recorded from the region (17; 18; 46), it is unlikely that stripping of weathering profiles at least one kilometre deep would have been widespread across the entire region. Both the landscape-based and AFT interpretations are based on data sets that have been extrapolated across much of the region. By extending these interpretations, their ability to account for local variations in the amounts of denudation and therefore the existence of landscape facets that have been relatively stable over long time frames adjacent to parts of the landscape that have experience large-scale denudation is limited. This sort of variation can be seen in areas such as Wilsons Promontory (17; 18) and the Southem Tablelands of NSW (37) where hundreds of metres of denudation have been restricted to deeply incised valleys, alongside ancient palaeosurface remnants. Inherent differences in the basis of these two interpretations may also contribute to the apparent conflict. Palaeolandscape studies inherently emphasise areas of greatest landscape stability (e.g. palaeosurface remnants and depositional sequences), for they feature the greatest amount of preservation of field evidence, over the longest time frame. In contrast AFT studies tend to favour areas of greater denudation, where regolith materials are not an obstruction to obtaining relatively fresh apatite samples. Rather than being a weakness of either discipline, these features should be exploited to provide more complete and complimentary information on the evolution of these landscapes. Preservation of Mesozoic Regolith and Landscape Features The preservation of ancient palaeosurface and regolith materials may appear as unejq)ected within a landscape attributed to a history of uplift and denudation. Denudation in many areas, however, has been been restricted, accounting for areas of long term landscape preservation alongside areas of major denudation. Several mechanisms for restricting denudation can be invoked to allow for the preservation of these regolith and landform features. Lithology has a major influence on the differential erosion of parts of the landscape. Many of the Mesozoic palaeosurfaces are associated with relatively resistant lithologies, such as Devonian volcanics in the Healsville, Warbourton and 248

Mt Dandenong regions (3; 8; 9; 11; 12), Palaeozoic granitic intrusions (4; 8; 9; 13; 19) and Devonian sedimentary rocks (13). In these cases major denudation is accomodated within adjacent, less resistant lithologies or along structural weaknesses just as joints. These resistant lithologies also are a major influence on local erosional baselevels, where the persistence of ancient palaeosurfaces may be due to erosional grading to resistant bedrock lithologies downstream. This is certainly the case in many granitic areas, where variations in the reUef of the weathering front has a major influence on local baselevels (47; 19). For example at Wilsons Promontory, and the Baw Baw and Buffalo Plateaux, "topographic step rises" and plateaux margins often consist of fresh granite enclosing highly weathered material. Sedimentary and volcanic burial protecting palaeosurfaces from extensive denudation may also explain the preservation of these features. Many of these may be presently expressed as exhumed features, although others still remain buried. These depositional areas would also tend to be characterised by considerably less erosion than adjacent highlands and also provide a protective veneer to later erosion. Examples include some of the sub-volcanic regolith described earUer. CONCLUSION Preservation of Mesozoic regolith and landscape features suggests long-term landscape stability and restricted denudation. This is supported by a wide range of landscape-based interpretations, but is in apparent contrast to AFT interpretations of widespread, kilometre-scale denudation in the mid to late Cretaceous. Detailed field studies in areas such as Wilsons Promontory suggest that apparent discrepencies in regional extrapolations by both landscape and AFT based research may not account for local variations in denudation and over-en^hasise particular landscape settings favoured by these techniques. Collaborative, detailed studies by landscape-based and AFT scientists with an appreciation for the local variations in denudation within a landscape as well as the inherent differences in basis of these disciplines has the potential to greatly increase our understanding of the landscape history of this region. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Andrews, E.C., 1910. Geographical unity of eastem Australia in the late and post Tertiary time. Journal and Proceedings of the Royal Society of New South Wales, 67, 251-350. Hart, T.S., 1908. Highlands and Main Divide of Western Victoria. Proceedings of the Royal Society of Victoria, 250-273. Skeats, E.W., 1909. Volcanic rocks of Victoria. Report of the Australian Association of the Advancement of Science, 12,173-235. Baragwanath, W., 1925. The Aberfeldy district. Memoir of the Geological Survey of Victoria, 15. Craft, F.A., 1932a. The physiography of the Shoalhaven River Valley, v. The upper valley and stream system. Proceedings of the Linnean Society of New South Wales, 57,197-212. Craft, F.A., 1932b. The physiography of the Shoalhaven River Valley, vi. Conclusion, Proceedings of the Linnean Society of New South Wales, 57, 245-260. Craft, F.A., 1933. The surface history of the Monaro, N.S.W.. Proceedings of the Linnean Society of New South Wales, 58, 229-244. Hills, E.S., 1934. Some ftindamental concepts in Victorian physiography. Proceedings of the Royal Society of Victoria, Al, 158-174. Hills, E.S., 1975. Physiography of Victoria. Whitcombe and Tombs, Melboume, 373pp. Gill, E.D., & Sharp, K.R., 1957. The Tertiary rocks of the Snowy Mountains, Eastem AustraHa. Journal of the Geological Society ofAustralia, 4,21-40. Garrat, M.J., 1973. Faulting and the physiography of the Croydon Sunkland, Victoria. Proceedings of the Royal Society of Victoria, 86, 15-18. Vandenberg, A.H.M., 1973. Geology of the Melboume district. In: McAndrew, J., & Marsden, M.A.H. (eds). Regional Guide to Victorian Geology, 14-30. Neilson, J.L., 1962. Notes on the geology of the high plains of Victoria. Proceedings of the Royal Society of Victoria, 75,277-284. Singleton, 0.R, 1968. Outline of the Geology and Physiography of Victoria. In: McAndrew, J. & Marsden, MA.H. (eds). Regional Guide to Victorian Geology, University of Melboume, Geology Department, pp.1-13. Jenkin, J.J., 1988. Geomorphology. In: Douglas, J.G. & Ferguson, J.A. (editors). Geology of Victoria, Geological Society of Australia, Victorian Division, 403-419. Joyce, E.B., 1992. The west Victorian Uplands of Southeastern Australia: Origin and history. Earth Surface Processes and Landforms, 17,407-418. Hill, S.M., Oilier, C.D., & Joyce, E.B., 1995. Mesozoic deep weathering and erosion: an example from Wilsons Promontory, Victoria. Zeitschriftfur Geomorphologie, 39, 331-339. Hill, S.M. & Joyce, E.B., 1995. Granitic regolith and landscape evolution of Wilsons Promontory, Victoria. Proceedings of the Royal Society of Victoria, 107, 1-10. Hill, S.M., 1996. The differential weathering of granites, with particular reference to Victoria, AustraHa. AGSO Journal of Geology and Geophysics, 16, 271-276. Jones J.G. & Veevers J. J. 1983. Mesozoic origins and antecedents of Austraha's Eastem Highlands. Journal of the Geological Society ofAustralia 30, 305-322. 249

21.

22.

23. 24. 25.

26. 29. 27. 28.

30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

42. 43.

44. 45.

46. 47.

O'SuUivan, P.B., Kohn, B.R, Foster, D.A., & Gleadow, AJ.W., 1995. Fission track data from the Bathurst Batholith: evidence for rapid mid-Cretaceous uplift and erosion within the eastern highlands of Australia. Australian Journal of Earth Sciences, 42, 597-607. O'Sullivan, P.B., Foster, D.A., Kohn, B.P., Gleadow, A.J.W., & Raza, A., 1995. Constraints on the dynamics of rifting and denudation on the eastem margin of Australia: Fission track evidence for two discrete auses of rock cooling, PACRIM '95 Abstracts, 441-446. Branagan, D.F., 1983. The Sydney Basin and its vanished sequence. Journal of the Geological Society of Australia, 19, 345-349. Browne, W.R., 1964. Grey Billy and the age of tor topography in Monaro, NSW. Proceedings of the Linnean Society of NSW, 89, 322-325. Taylor, G., 1994. Palaeoweathering and Bauxites. In: McQueen (ed.). The Tertiary Geology and Geomorphology of the Monaro: The perspective in 1994, Centre for Australian Regolith Studies Occassional Publication No. 2, 51-56. Williams, G.W., 1983. The Tertiary auriferous alluvial deposits of north central Victoria. Bureau of Mineral Resources Geology and Geophysics Record, 1983/27,137-143. Wopfher, H., Callen, R., & Harris, W.K., 1974. The Lower Tertiary Eyre Formation of the southwestem Great Artesian Basin. Journal of the Geological Society of Australia, 21,17-51. Oilier, C.D. & Pain, C.F., 1994. Landscape Evolution and tectonics in southeastem AustraUa. AGSO Journal of Geology & Geophysics, 15, 335-345. Hill, K.C., Hill, K.A., Cooper G.T., O'Sullivan, A.J., O'Sullivan, RB., & Richardson, M.J., 1995. Inversion around the Bass Basin, SE Australia. In: Buchanan, J.G. & Buchanan, P.G. (eds). Basin Inversion, Geological Society Special Publication No. 88, 525-547. Hill, S.M., Taylor, Graham, & Eggleton, R.A., 1994. Field guide and notes on the regolith and landscape features of the Broken Hill region, western NSW. AGSO Record 1994/57, 39pp. Singleton, O.R, 1970. Geology and mineralization of Victoria. In: McAndrew, J. (ed.). Geology of Australian Ore Deposits, pp.440-449. Bowen, K.G., 1974. Potassium-argon dates - determinations carried out for the Geological Survey of Victoria. Geological Survey of Victoria Report, 1974/79. UnpubUshed. McDougall, I., & Wellman, P., 1976. Potassium-argon ages for some Australian Mesozoic igneous rocks. Journal of the Geological Society ofAustralia, 23,1-9. Nott, J.F. & Purvis, A.C., 1995. Geomorphic and tectonic significance of Early Cretaceous lavas on the coastal plain, southern New South Wales. Australian Journal of Earth Sciences, 42,145-149. Wellman, P. & McDougall, L, 1974. Cainozoic igneous activity in eastem Australia. Tectonophysics, 23,49-65. Price, R.C., Gray, C.M., Nicholls, lA., & Day, A., 1988. Cainozoic Volcanic Rocks. In: Douglas, J.G. & Ferguson, J.A. (editors). Geology of Victoria, Geological Society of Australia, Victorian Division, 439-452. Young, R.W., 1983. The tempo of geomorphological change: evidence firom southeastem Australia. Journal of Geo/ogy, 91,221-230. Bird, M.J. & Chivas, A.R., 1988. Oxygen isotope dating of the Australian regolith. Nature, 331, 513-516. Bird, M.J. & Chivas, A.R., 1989. Stable-isotope geochronology of the Australian regolith. Geochimica et Cosmochimica Acta, 53, 3239-3256. Bird, M.J. & Chivas, A.R., 1993. Application of oxygen-isotope geochronology to the Australian regolith: geomorphic and palaeoclimatic implications. Australian Journal of Earth Sciences, 40, 345-358. Moore, M.E., Gleadow, A.J.W., & Lovering, J.F., 1986. Thermal evolution of rifled continental margins: new evidence fromfissiontracks in basement apatitesfiromsoutheastem Australia. Earth and Planetary Science Letters, 78,255-270. Duddy, I.R. & Green, P.F., 1992. Tectonic Development of the Gippsland Basin and Environs: Identification of Key Episodes Using Apatite Fission Track Analysis (AFTA). Gippsland Basin Symposium, pp. 111-120. Dumitru, T.A., Hill, K.C., Coyle, D.A., Duddy, I.R., Foster, DA., Gleadow, A.J.W., Green, RF., Kohn, B.R, Laslett, G.M. & O'Sullivan, A.J., 1991. Fission track thermochronology: Apphcation to continental rifting of South-Eastem Australia, 1991,131-142. Foster, D.A. & Gleadow, A.J.W., 1992. Reactivated tectonic boundaries and implications for the reconstruction of south eastem Australia and northem Victoria Land, Antarctica. Geology, 20,267-270. Kohn, B.P. & Gleadow A.J.W. 1994. Thermo-tectonic evolution of the Snowy Mountains: An apatite fission track study (abstract). Specialist Group in Tectonics and Structural Geology Field Conference, p.88. Geological Society of Australia. Oilier, C.D., 1984. Weathering (2nd Edition), Longman, London, 270pp. Oilier, C.D., 1965. Some features of granite weathering in Australia. Zeitschriftfiir Geomorphologie, 9,285-304.

250

THE GREAT ARTESIAN BASIN MANAGEMENT OF WATER RESOURCES AFTER 100 YEARS OF DEVELOPMENT John HilUer Resource Sciences Centre, Department of Natural Resources, 80 Meiers Road Indooroopilly QLD Summary The Great Artesian Basin underlies 22% of the Australian continent, and is the major source of water over most of this area. It was first tapped in 1878 and over the next 30 years numerous bores were drilled to tap the water which flowed plentifully to the surface. Little thought was given to the longevity of the resource and many bores were poorly constructed and were permitted to run freely into creeks and constructed drain systems. This early lack of control has caused problems for water managers up to the present day. Control of drilling and water use in the Great Artesian Basin is the responsibility of the individual States. Only in recent times have the States collaborated and some uniform management practices been introduced. Management is now moving into a new phase with the formation of a Community Consultative Council to advise on management options. However at present about 80% of the water which flows from the Basin is wasted, mainly from uncontrollable bores and ineflQcient distribution systems. There are many options for management, and these include complete replacement of distribution drains with pipelines, increased pumping in non-flowing areas, or increased use for irrigation, mining and industrial purposes. Other options include management to maintain the artesian pressures or to provide maximum benefit from water use now. With the hydrology of the Basin still not fully understood but with knowledge of the enormous reserves of water which the Basin holds, management decisions are difficult. Conservative management can stifle development and while large-scale waste continues, can hardly be justified. However, decisions today will determine the value of the Basin tomorrow. Management of the Great Artesian Basin is one of the major challenges facing water resource managers in the next 20 years and will determine the future of this generally semi-arid area. Introduction The Great Artesian Basin is the most extensive of the Mesozoic Basins in Australia. It consists of layers of porous sandstone with interbedded shale and mudstone layers. The sandstone contains a resource of great value over the vast area of basin - the largest source of freshwater on the Australian continent This Basin occupies about 1 711 000 km^ on about 22% of the land surface of Australia (Figure 1). It extends from Cape York Peninsula to the Simpson Desert, to northem South Australia and New South Wales, and covers about 60% of Queensland -. mainly the semi-arid westem area. The sediments of the basin extend down to over 2000 m below the current land surface. Water in the sandstone is confined by the shale and mudstone aquicludes. Over much of the basin the piezometric surface of the water is above ground level, ie the water will flow naturally from the bore - it is artesian. This has been and remains the greatest management challenge - the freeflowingbores impart the impression that the supply is virtually inexhaustible. Wastage of water has been enormous as the very large flows encountered in individual bores are far in excess of the actual waters requirements.

251

i Great Artesian Basin

Figure 1. Location of the Great Artesian Basin Although bores were first drilled into the artesian aquifers in 1878, management of this resources is still evolving. The basic, overall hydrogeological system was understood many year ago, but recent advances in technology means that many of the assumptions can now be quantified. Management of this huge resource has many challenges and, because of its size, many options. Early Development Following the drilling of the first bore to tap the artesian aquifers near Bourke, NSW, in 1878, graziers were quick to chase this reliable water source. Within 20 years, the extent of the basin was well defined - a remarkable feat in the poorly explored and sparsely settled areas of central Australia. By 1900, some 670 artesian bores had been drilled in Queensland, with a total flow of about 365 000 megalitres/year (Hillier, (1)). Just a few years after bores were drilled, it was evident that the high initial flow rates, some in excess of 10 megalitres/day, were declining rapidly. There was much debate as to the cause of the decline and much conjecture that the basin was running dry. An interstate conference was held in 1912 to consider the long-temifiitureof the Great Artesian Basin and other artesian basins in Australia, and to attempt to gain an understanding of their hydrogeology. Further conferences were held in 1914, 1921, 1924 and 1928. However, as water resources are a State responsibility, no basin wide management plan was conceived despite the involvement of all relevant States in this most detailed conference series. However, the series ended with general agreement and clarification of generalised basin geology, the effects of elastic storage on diminution of flows and probable long term steady state basin discharge. The Queensland Government continued investigation into the Great Artesian Basin, culminating in the 1954 publication of a report. Artesian Water Supplies in Queensland (Queensland Government, (2)). This most comprehensive report considered most aspects of the basin and concluded with recommendations that:• reinforced existing policies on licensing of artesian bores. • considered preference should be given to domestic, stock and irrigation use in that order, with irrigation use to be restricted. • preference should be given to pipeline distributionfi-omnew bores. • a strict program of water conservation should not be undertaken, butflowsshould be regulated to actual requirements. • rehabilitation of bores be optional and be assisted by the provision of technical assistance. • consideration be given to the construction of a working model of the Great Artesian Basin to assist in understanding of the hydrology. Most of these recommendations were either in place or were quickly implemented and it was on this basis that the basin was managed for the next 30 years. No computer model was developed as at that time as only physical models were possible. The quest for more knowledge of the geology and hydrology of the basin has expanded as modem technology has greatly improved the tools available. Gamma ray logging of bores, continued remeasurements and testing of bores to determine 252

static heads and aquifer characteristics, isotopic analyses and age dating of water samples, and computer simulation modelling of the aquifer hydrology have, over the last 20 years, led to a better understanding of the Basin. This greater understanding, together with a greater community reaUsation that water resources in Australia are limited has changed general opinion on Great Artesian Basin Management strategies. Current State of the Basin Currently outflow from the Great Artesian Basin is about 420 megalitres/year. Of this, over 80% is wasted because of uncontrollable bores and inefiScient earth channel distribution systems. There has been a continued decrease in the artesian pressure in the Basin - a decrease caused mostly by this enormous waste of water. Head losses of up to 80 metres in southem Queensland and northem New South Wales have occurred in the 100 years since resource development occurred (Habermehl, (3)). The rate of decline of the artesian pressure is reducing as the basin hydrology approaches steady state and the flow from bores decreases as the pressure falls. In some areas, notably the Flinders Sub-basin, equilibrium has almost occurred, despite the large outflow from numerous uncontrolled bores. Recent Developments In 1987, the Australian Water Resources Council established an interstate working group in an attempt to have the Great Artesian Basin assessed, monitored and managed on a basin wide basis without the interruptions of State boundaries. This group, with membership of the relevant States and the Northem Territory has brought about a more national approach to many aspects relating to the Great Artesian Basin. In 1989, a program aimed at the rehabilitation of uncontrollable bores was commenced, involving the Commonwealth Government and the States of Queensland and South Australia, with New South Wales joining shortly after. The program provides generous subsidies from the two levels of Government (up to 80%) to repair or plug and replace bores which are flowing uncontrollably. The effects of this program are now being seen as the reduced outflow from bores is resulting in an increase in artesian pressures. (Hiller, (4)). Integral to the success of this program is the replacement of the inefficient bore drain distribution with a fully controlled pipeline system. However, there are still in excess of 25 000 kilometres of drains in Queensland, and subsidies for pipelines are limited. Many individual landholders are replacing drains with pipeUnes as their bores are rehabilitated and finances are available. Management Alternatives The Great Artesian Basin contains a huge reservoir of water. Use at the current rate would result in the water lasting about 20 000 years if no recharge occurred, but obviously most of this would need to be pumped out. As recharge of the aquifers is occurring, there is little danger of running out of water, but more water will eventually have to be pumped. Management over the past 100 years has variedfrombeing wasteful to being conservative. Wasteful management has allowed the use of bore drains and has accepted the uncontrollable bores and swanps that have resulted from these. Conservative management has been driven by the importance of the grazing industry. The cost of water conservation is too large to be bome by individual landholders, but to protect their livelihoods, priority of use has been allocated to their industry. Other users have not been totally excluded. Some, such as the oil industry, are equally as wasteful with oil production wells producing more than 95% water which is discharged to waste. However, the general trend in the past has been to discourage other users where possible. The management altematives are obvious. They could be summarised as below:• stop all waste from uncontrollable bores and drain distribution systems. • allow use of water for high value uses such as mining and industrial as required. • allow irrigation use in areas of water and soil suitability. In the long-term, water will need to be pumped, ie will not flow. There would then be little need to stop wastefrombores, as waste would be minimal wastage will be reduced, as water will have a greater value as pumping costs are incorporated. • economics will be the over-riding management control, as the cost of pumping will govern the use of water. High value uses will predominate. • Thus there are really two management options:1. Management to maintain the artesian pressure in the Basin. This option is somewhat Utopian. The pressure had to fall once the Basin was tapped. Besides the elastic storage factor, any use was going to reduce the pressure, and over most of the Basin the transmissivity of the sandstone, ie the ability of 253

the aquifer to move water, is the limiting factor in the yield. Water cannot travel through the sandstone quickly enough to supply the outflow, so the head falls to cause an increase in the gradient to increase the flow. Such head falls can be arrested by decreasing use, but over much of the Basin the artesian conditions would, in the long-term, become subartesian. Current consumptive use may be accommodated but wastage would need to cease and little other use would be possible. Management to utilize the water volume in the Basin. Most groundwater systems are managed on the basis that, in the long-term, use should not exceed average annual recharge. However, tiie rate of movement of water in the Great Artesian Basin is very slow and recharge volumes have not been quantified. Monitoring indicates that in some areas in the Basin a steady-state situation is being attained, but there is no evidence that this is being achieved by a balance of recharge to outflow. Such balances are difficult to quantify due to the large storage volume involved and it is probable that the perceived steady-state situation is a balance between outflows and transmission of water through the aquifer rather than recharge. Attempts are now underway to quantify recharge, but current monitoring indicates recharge is dependant on widely spaced, very much above average rainfall events. The utilization of such recharge water is very long-term when isotopic dating shows that much of the water stored in the Basin is in excess of 1 million years old (Torgersen et al, (5)). An extremely large volume of water is stored and available for use — a volume that cannot be used in conceivable time. Current Management Practices Management of the water resources in the Great Artesian Basin is, to a certain extent, still based on the recommendations which came from the 1954 report. However, several important changes have been made though some of these changes are being introduced by encouragement rather than regulation. The current management aims are:• To reduce and eventually eliminate wastagefirombore and distribution systems. • To preserve pressures so artesian conditions are maintained over the majority of the Basin. • To give priority of use towards domestic and stock purposes, while allowing limited irrigation where water and soil are compatible and use is unlikely to cause a reduction in pressure. Such areas are intake areas or areas where pressure depletion is small and water is being saved. • To encourage use for high value purposes such as mining or manufacturing, where such use does not have a significant detrimental effect on other users. • To consider the environmental value of the water resources in all management decisions. One of the most important aspects of the strategies for management of the Basin has been the introduction of bore construction standards. These standards are basin wide and now ensure the integrity of new and rehabilitated bores. Cement grouting of casing in bores has been mandatory for many years to ensure that bores can be controlled. However, the main cause of uncontrollable bores has been deterioration of steel casing caused by corrosive water. Areas have been delineated where water is known to be corrosive and bores in these areas must be cased with inert materials such as FRP, PVC or ABS. Stainless steel headworks are now utilized in all areas containing corrosive water. Although all bores drilled since 1954 have had to reticulate water using pipelines, the replacement of drain systems by pipes is still optional. The advantages of pipeline distribution are quite diverse, from control of feral animals and woody weeds, to better water quality at watering points, ability to water more of the property, decreased land degradation, watering point mustering, and better water supplies at homesteads. Most bores which are rehabilitated are being piped as landholder finance becomes available. The effects of the rehabilitation program and piping can be seen in Figure 2. There has been a saving in outflow from the Basin in Queensland since 1990 of over 30 000 megalitres/annum. Pressures are starting to rise in areas where rehabihtation of bores has been significant and some non-flowing bores are starting to flow. Management of the Basin is controlled by the Water Resources Act (1989) which is administered by the Department of Natural Resources. To bring some community involvement into the management of the Basin, a community advisory committee is being estabUshed, with representatives from all states and user groups. This group will advise the relevant State authorities on management options.

254

2000-n

1880

- 4000

I I I • I > I I I I I

1940

YEAR

1994

Figure 2. Artesian Bore Flow History Conclusions It is obvious that management of the Great Artesian Basin is still evolving. Knowledge of the Basin is increasing especially in terms of Basin recharge and discharge. The advent of a computer simulation model of the whole Basin will allow various use scenarios to be simulated and the long-term effects to be evaluated. This will have a major effect on management as the consequence of management policy will be able to be considered at the time of policy formulation. The Community Advisory Committee will ensure that the user groups have input into management policy. The formalisation of community involvement is considered an important process in the management of this extremely large resource. References HiUier, J.R., The Great Artesian Basin, Water Journal, Volume 19, No. 6, December 1992, Australian Water & Wastewater Association, p. 42-43. Queensland Government, Artesian Water Supplies in Queensland, Report following First Interim Report (1945) of Committee appointed by the Queensland Government to investigate certain aspects relating to the Great Artesian Basin (Queensland portion) with particular reference to the diminishing supply., 1954, Parliamentary Paper a561955. Habermehl, M.A., The Great Artesian Basin, BMR Journal ofAustralian Geology & Geophysics, 5,1980, p. 9-38. HiUier, J.R., The Australian Great Artesian Basin- Management for Resource Sustainability, Proceedings of the International Association ofHydrogeologists Congress, Edmonton, Canada, June, 1995. Torgersen, T., Habermehl, M.A., Phillips, EM., Elmore, D., Kubik, P., Jones, B.G., Hemmick, T, and Gove, H.E., Chlorine 36 Dating of Very Old Groundwater, 3, Further Studies in the Great Artesian Basin, Australia, Water Resources Research, Vol 27,1991, p. 3201-3213.

255

THE MESOZOIC AQUIFERS OF CAPE YORK PENINSULA Anthony HornS Elizabeth DerringtonS Graham Herbert", Rob Lait' and John Hillier' Department of Natural Resources, Brisbane I MareebaS, InnisfaiU, Consultant, Brisbane (formerly AGSO, Mareeba)2 Summary In Cape York Peninsula significant groundwater supplies are obtained from the Mesozoic Laura Basin for town and stock use. Groundwater, while available in significant quantities, is not commonly abstracted from the Mesozoic Carpentaria Basin because of the ready availability of surface water supplies in the northern part of the basin and because supplies are more easily obtained from the overlying Tertiary Karumba Basin in the south. Groundwater quantities available from the Mesozoic Basins are generally large and water quality is generally good except for some sections of the Carpentaria Basin where high fluoride levels, corrosive water and high Sodium Adsorption Ratio water can be present. Comparatively few bores have been drilled through the Mesozoic sediments in the Peninsula and the precise hydrogeological nature of these basins is not well understood. Structure contours are presented for the top of the Gilbert River Formation, the main aquifer unit in both of these basins and a brief assessment of groundwater quantity and quality is provided. Introduction The Cape York Peninsula study area covers some 143 000 kilometres in northem Queensland (Figure 1) and information presented here is based on field work undertaken throughout this area as part of the Cape York Peninsula Land Use Strategy (CYPLUS) program. The study area lies exclusively within the tropical region with the climate varying from humid along the eastem coastal strip to semi arid in the central westem Peninsula area. The area has a small sparsely settled population of around 10 000 (1986 census) with few major population centres. The principle resource uses are rangeland grazing, mining and tourism. The study area can be divided into several major hydrogeological provinces that represent both formal and informal tectonic units. These are the Proterozoic and Lower Palaeozoic Coen and Yambo Inliers, the Devonian to Permian Hodgkinson Province, the Carboniferous Cape York Pyroclastics, the Jurassic/Cretaceous Carpentaria and Laura Basins, and the Tertiary Karumba Basin. Tectonic unit boundaries are illustrated in Figure 1. and age relationships of the tectonic units in Figure 2. Figure 3 provides a stratigraphic table of the major Cainozoic and Mesozoic units. The Jurassic/Cretaceous Carpentaria and Laura Basins were deposited contemporaneously as epicratonic basins following basement downwarping and local block faulting. They are composed of equivalent formations and are lateral and contemporaneous equivalents to the Eromanga Basin (the largest sub-basin of the Great Artesian Basin). The Carpentaria Basin is also part of the Great Artesian Basin. The Carpentaria and Laura Basins are composed principally of two predominantly quartz sandstone formations overlain by an extensive and thick argillaceous unit. These units generally thicken towards the centre of their respective basins and usually overlie an uneven basement topography. The major porous aquifers which occur in the Carpentaria Basin are contained in the Gilbert River Formation and the Garraway and Helby Beds and the major porous aquifers in the Laura Basin are contained in the Gilbert River Formation and the Dalrymple Sandstone. Minor aquifers occur within the lower Rolling Downs Group but they are generally not hydraulically connected to each other or the underlying formations. Overall, reserves of groundwater from the Mesozoic Basins appear to be large with groundwater discharges in these basins of up to 43 L/s being recorded. Generally, flow rates of up to 15 L/s could be expected from artesian and subartesian bores abstracting from the major Mesozoic aquifers. QuaUty of groundwater throughout the study area is usually very good with commonly low salinity levels. The most common water types are sodium bicarbonate and sodium chloride type or minor variations of these. The groundwater is predominantly acidic with some alkaline groundwater occurring in the Karumba (Tertiary) Basin and in the Gilbert River Formation/Dalrymple Sandstone aquifer in the Laura Basin. In areas of the Karumba and Carpentaria Basins, high fluoride levels have been recognised in the groundwater. In some areas this water is used for human consumption where no altemative supplies currently exist.

256

l h a n d a y Island THUmOAYI. IMQPA 1

•

COOKTOWN

fioimmm N lOKHiwnLU ^^ MTOA

BRISBANE

CYPLUS STUDY AREA

LEGEND

Northern Limit of Karumba Basin Tectonic Unit Boundary Bores

CARPENTARIA

Pormpuraaw

( Wujul Wulul

Figure 1. Location Map and Tectonic Units Laura Basin The onshore Laura Basin covers approximately 23,000 kilometres^ and is connected to the Carpentaria Basin across the Kimba Arch. The Dalrymple Sandstone and the Gilbert River Formation, in combination, contain the most significant aquifer system in the Laura Basin. This aquifer system is continuous across the basin and provides reliable supplies of fair to excellent quality water. The thickness of the combined formations is over 800 metres in the central part of the basin (Williams, (1)). The Basin increases in depth and thickness principally to the north but also to the east (Smart and Rasidi, (2)). Basin sediments dip gently to the north towards the depocentre off Princess Charlotte Bay (Passmore, (3)) with basement highs occurring in the southem part of the basin. The Laura Basin Mesozoic sequence began with the predominantlyfluvial/deltaicDalrymple Sandstone, overlain by the fluvial and marginal marine Gilbert River Formation. The sequence temiinates with the entirely marine Rolling Downs Group. The Palmerville Fault is the major structural feature in the basin and was initiated in the Jurassic (Hawkins and Williamis, (4)) and remained active until the Late Cretaceous (Smart and Rasidi, (2)). The general nature of the hydrogeological units in the Laura Basin is outlined:

257

South

North

Recent Cretaceous

Carpentaria Basin

Laura Basin

Jurassic Triassic Permian

Karumba Basin

Tertiary

Carbonitferous Devonian

KUt^HI

Coen/Yambo Inliers

Hodgkinson Province

Silurian Ordivician Cambrian Precambrian

Figure 2. Tectonic Unit Age Relationships Dalrymple Sandstone/Gilbert River Formation The Dalrymple Sandstone is in general, conformably overlain by the Gilbert River formation which shares a similar lithology and fluviatile origin in its lower sections thus making the boundary between these units difficult to distinguish. For this reason the two formations are treated together as one hydrogeological unit. Within the study area the Dalrymple Sandstone occurs only in the Laura Basin, however, the unit is lithostratigrapically similar to the Garraway Beds in the Carpentaria Basin. The Dalrymple Sandstone is composed of fluvial, paralic and shallow marine sands. Generally, the Dalrymple Sandstone fines upwards from conglomerate into sandstones and siltstones. It consists of fine to coarse grained pebbly in part) quartzose to sub-labile sandstone with minor mudstone, siltstone and conglomerate. It is thick bedded, unstratified and cross bedded and generally well sorted. The Dalrymple Sandstone appears to maintain a relatively uniform lithology throughout the basin. The Formation thickness variesfromaround 60 metres near the outcrop in the eastem part of the basin to over 500 metres near the coast (WilHams, (1)). The depositional environment of the Gilbert River Formation varies from fluviatile at the base to marginal marine at the top (Hawkins and Williams, (4)). Drilling has established formation thicknesses from 60 metres in the east to over 280 metres in the central part of the basin (WiUiams, (1)). The Gilbert River Formation in the Laura Basin has a lower interbedded siltstone and sandstone interval and an upper sandstone unit, with more minor siltstone and conglomerate, especially where it grades into the Rolling Downs Group. Structure contours for the Gilbert River Formation are presented in Figure 4. Artesian flows occur from aquifers in the Gilbert River Formation and Dalrymple Sandstone within 20 kilometres of outcrop in the southem and south westem parts of the Laura Basin. Artesian conditions persist across the basin up to the eastem onshore margin. These formations also provide artesian water in the north westem onshore area at similar distances from the margins. The total available water quantity from these formations has not been determined as few bores penetrate the entire Mesozoic sequence. The Utah Development Company recorded artesianflowrates rangingfrom2.5 L/s to 12.6 L/s from bore holes partially penetrating the Mesozoic Sandstones in the Bathurst Range and Kalpowar areas.

258

GENERALIZED STRATIGRAPHY OF CAINOZOIC / MESOZOIC ARTESIAN BASINS OF CAPE YORK PENINSULA (Hev.ed s.an et ai isso KARUMBA

WYAABA BEDS .0 — 7 7

CARPENTARIA

YAM CREEK ? —BEDS ?—?

CAINOZOIC

ULYVALE BEDS 7 —

7 —

—

7 —

LAURA

TERTIARY 7

UNDIFFERENTIATED

-? — ?— ?— ?— ?— ?— ?— ?— ?— ?— ?— ? — ?— ?— ?— ?-

BULIMBA FORMATION

-? — ? —? —? — ? — ? — 7 —7 —7 — ? — ? — ? — ? — ?—•? — ? Do M CO

SOUTHERN AREA NORMANTON FORMATION

NORTHERN AREA

SOUTHERN AREA

ALLARU MUDSTONE ROUJNG DOWNS GROUP

WELGUNYA SUB-GROUP

TOOLEBUC FORMATION

I NORTH-EASTERN AREA

ROLUNG DOWNS GROUP

IfALLUMBILLA FORMATION GILBERT RIVER FORMATION

COFFIN HILL MEMBER EQUIVALENT YAPPAR MEMBER EQUIVALENT

GILBERT RIVER FORMATION

LOTH FORMATION

EULO QUEEN GROUP

o j j CO

—

UNNAMED SEQUENCE

HELBY BEDS

GARRAWAY BEDS

HAMPSTEAD SANDSTONE

? —

—

DALRYMPLE SANDSTONE

L 7 _ 7 _

Figure 3. Stratigraphic Table (revised Smart et al. 1980) The quality of water from these units is usually acceptable for most purposes. The conductivity varies betvêen 50 |iScm-' and 3200 j^Scm"' vîth an average figure of only 900 |iScm-^ The majority of vâter samples from the Gilbert River Formation/Dalrymple Sandstone aquifer are of sodium bicarbonate type. Isolated bores in outcrop areas near the sea (i.e. at Bathurst Head) supply water of the Na-Cl type probably as a result of higher NaQ content of rainwaterfromsea spray. Water quality records indicate that there is a slow increase in total dissolved solids as water migrates away from intake areas towards the basin axis i.e. T.D.S. 40 mg/L to T.D.S> 350 mg/L over a distance of approximately 50 kilometres. Bores that supply water with a T.D.S. of greater than 1000 mg/L are, it is proposed, probably subject to influences extemal to the aquifer. These may include structural features that allow interformational water mixing, the proximity of the aquifer to basement rocks or lithostratigraphic interrelationships.

259

Figure 4. Laura Basin - Gilbert River Fonnatioii Structure Contours Groundwater sourced from the Dalrymple Sandstone/Gilbert River Formation has an average pH of 6.4, and further water quality parameters are indicated in Table 1. Table 1. Average Groundwater Analyses - Dalrymple Sandstone/Gilbert River Formation, Laura Basin LAURA BASIN-AVERAGE GROUNDWATER ANALYSES * indicates data forms a skewed distribution and significant variationsfrommean values occur Cond Hard SAR Alk Na Ca Mg CI HCO3 SO4 NO3 F Sample Size Dalrymple 80 Sandstone/Gilbert 900* 137* 7 288* 170* 32* 14* 142* 348* 5* 0.7* 0.6* River Formation Rolling Downs Group The Rolling Downs Group is generally flat lying, and exhibits little deformation. However a ?Late Cretaceous movement of the Palmerville Fault has locally displaced the group down to the east by about 100 metres (Smart and Rasidi, (2)). The Group was deposited in shallow marine and paralic environments (Smart and Senior, (5)). It's thickness ranges from around 40 metres near Kalpowar to over 200 metres in the central part of the basin (Williams, (1)). The general sequence of mudstone-siltstone sequences are usually of low permeability, but contain isolated sandy layers that may act as aquifers. However, the Rolling Downs Group varies lithologically across the basin. The Rolling Downs Group, where drilled, generally provides small quantities of water at depth but is not considered to be a regional groundwater target because of its limited permeability. Most bores either stop at the top of this Group or continue through it to better prospects in lower formations. Little data is available regarding the potential supply of water from the Rolling Downs Group. The isolated lenses of sandier lithology encountered produce only limited subartesian supplies, generally less than 1 L/s. However, reasonable supplies of fair quality water have been obtained from fractures within the Rolling Downs Formationfromone bore on Olive Vale Station. While, significantfracturinghas been reported in the Rolling Downs Formation in the southem part of the basin in mining company reports and private driller's logs. An artesian flow, sourced from afracturezone on Olive Vale, was measured at 1.2 L/s witii a positive head of 3.5 metres. A subsequent water analysis revealed that this water is probably sourced from contact with a granitic basement. The quality of water sourced from Rolling Downs Group aquifers is generally poor, though it is usually suitable for stockwatering. Conductivity ranges from 1400 jiScm-' to over 7000 [iScm"' and the water is predominantly of the Na-Cl 260

type. Samples from the fracture zone drilled at Olive Vale Station revealed concentrations of fluoride and boron normally associated with granitic rocks rather than the mudstones and siltstones of the Group. Carpentaria Basin The Carpentaria Basin is, along with the Surat and Eromanga Basins, one of three inter-connected basins that con:5)rise the Great Artesian Basin. The Carpentaria Basin covers some 85 800 kilometres2 of the study area and underlies all of the westem side of Cape York Peninsula. The Tertiary Karumba Basin overUes most of the Carpentaria Basin in the south westem and central westem study area. The basin reaches a thickness of 1700 metres at its depocentre offshorefromWeipa and has a maximum thickness of 900 metres onshore (McConachie et al., (6)). The depositional environment was dominantly continental (riverine) except for a major Lower Cretaceous marine transgression (GSA/IWSC, (7)). On a basin-wide scale, the Carpentaria Basin has reasonable fonnational and broad lithological continuity. However, there is significant regional geological variation, resulting from basement highs and valleys and from transgressive and regressive depositional environments. The main aquifer unit of the Carpentaria Basin is the Gilbert River Formation. In the northem part of Cape York Peninsula the Garraway Beds and Helby Beds also contain significant aquifers. Minor aquifers also occur within the lower Rolling Downs Group but these are generally not hydraulically connected to each other or to the underlying formations. There is significant distances between boreholes penetrating the deeper Mesozoic sequences in the Carpentaria basin. Most data from these deeper holes is clustered around Weipa and Kowanyama and little data is available over other parts of the basin. Because of this paucity of direct geological data the local subsurface geology of the basin is still not well defined. As an example, two bores, drilled as part of the CYPLUS project, terminated in the middle of the Allaru Mudstone (an upper unit of the Rolling Downs Group) at depths of 327 metres and 331 metres respectively. It has been assumed that approximately 300 metres of Rolling Downs Group sediments would exist below these depths. It was originally expected that the top of the Gilbert River Formation (which directly underlies the Rolling Downs Group) would have been intersected at depths of less than 300 metres in these bores. Insufficient bores have been drilled in the Gairaway Beds and Helby Beds to allow proper definition of these formations as hydrogeological units though these units are considered to be, in a regional sense, lithologically similar throughout their extent. For the purposes of this work the Gilbert River Formation, the Garraway Beds and the Helby Beds are considered as one hydrogeological unit as it is difficult to distinguish these conformable units in the subsurface. The general nature of the hydrogeological units of the Carpentaria Basin is outlined: Garraway Beds, Helby Beds and Gilbert River Formation The Garraway Beds are equivalent to the Eulo Queen Group in the southem Carpentaria Basin, are of Late Jurassic age and were deposited in a fluvial environment. The formation rests unconformably on the Pre-Mesozoic basement and could be up to 130 metres thick in the Sir William Thompson Range area (Smart and Senior, (5)). The general lithology is described as "clayey, micaceous, quartzose sandstone, granule and pebble conglomerate, locally carbonaceous, increasingly conglomeratic towards base" (Smart and Senior, (5)). Very little information is available on the aquifer potential of the Garraway Beds. North of 12°S the Gilbert River Formation merges with the Helby Beds. Smart and Senior, (5) consider the Helby Beds to be equivalent to parts of the Gairaway Beds and the Gilbert River Formation and described the general lithology as quartzose sandstone, conglomerate, and siltstone. Tuff and tufifaceous sandstone occur in the lower part. Powell et al., (6) consider the thickness of the Helby Beds to be about 300 metres. The Helby Beds generally contain subartesian aquifers in their outcrop areas that have the potential to supply usefiil quantities of good quality water. However few bores exploit this aquifer owing to the ready availability of surface water. A 40m deep bore, drilled as part of the CYPLUS program, obtained a supply of 3 L/s from this formation. The Gilbert River Formation is the main Carpentaria Basin aquifer unit in Cape York Peninsula. It correlates with the Hooray Sandstone, a major aquifer in the Eromanga Basin. The formation was deposited principally in fluviatile conditions in the Late Jurassic to Early Cretaceous and conformably overlies the Garraway Beds in the study area. Smart et al., (8) sub-divided the Gilbert River Formation into the lower Yappar Member, which accumulated in freshwater, terrestrial conditions, and the upper Coffin Hill Member which accumulated in marine and marginal marine conditions. The Yappar Member is composed of clayey coarse to medium-grained, quartzose sandstone grading upwards to a medium and fine-grained quartzose sandstone. In its type area it is pebbly at its base and tends towards conglomerate. The Coffin Hill Member conformably overlies the Yappar Member, It consists of coarse to fine-grained, clayey, quartzose sandstone with interbedded siltstone. Glauconite is a common accessory mineral.. These two members however, were indistinguishable during drilling undertaken as part of the CYPLUS program. 261

Figure 5 is a structure contour map of the top of the combined Gilbert River Formation/Helby Beds hydrogeological unit. These contours indicate that there is a gentle dip towards the west from the outcrop area. However, it should be noted that few data points give accurate stratigraphic control between Weipa and GSQ Rutland Plains 1. Very few bores extract groundwater from the Gilbert River Formation where it is confined by the Tertiary Karumba Basin. Flowing supplies of between 25 L/s and 43 L/s were recorded from the Gilbert River Formation and Garraway Beds in the deep Weipa Peninsula bores. Non-flowing supphes of up to 5 L/s are generally obtainable from the lower Carpentaria Basin units. The quality of groundwater from the major Carpentaria Basin aquifers, the Gilbert River Formation and the Garraway Beds, when they are unconfined, is generdly good and suitable for most purposes. When these aquifers become confined, the water quality (conductivity) deteriorates. This deterioration in conductivity is probably due to mixing with saline groundwater from the overlying Rolling Downs Group. Generally water from the Rolling Downs Group is quite saline and, therefore, unsuitable for purposes other than stockwatering. Deep bores drilled at Weipa and Aurukun were intended for bauxite mining purposes but are no longer used for that purpose owing to the corrosive nature of their water, partially because of high CO2 levels, and the ready availability of good quality groundwater from younger Karumba Basin aquifers. The groundwater from the lower units of the Caipentaria Basin (the Gilbert River Formation, the Garraway Beds and the Helby Beds) is generally suitable for stockwatering purposes. However the elevated levels of sodium and bicarbonate in the groundwater generally make it unsuitable for irrigation of the clayey soils derived from the Rolling Downs Group (the water commonly has a h i ^ Sodium Adsorption Ratio). Fluoride levels in waterfromthese lower units can often exceed the recommended limits for drinking water for humans and young stock. Generally the conductivity content variesfromaround 400 laScm"' in outcrop areas to 1400 liScm-' in deep artesian bores. Water from the Gilbert River Fomiation and Helby Beds are of sodium bicarbonate and sodium chloride type while the Garraway Beds aquifers contain a more sodium-chloride type water.

Figure 5. Carpentaria Basin - Gilbert River Formation/Helby Beds Structure Contours Rolling Downs Group The Rolling Downs Group is mainly of lower Cretaceous age and conformably overhes the Gilbert River Formation. The Group is in general, hydrogeologically similar in the Carpentaria and Laura Basins. Because of its finer grainsize compared with the underlying Gilbert River Formation and Garraway Beds, the Rolling Downs Group serves as a major confining unit for the regional aquifer systems throughout the Great Artesian Basin (GSA/IWSC, (7)). The Rolling Downs Group contains sandstone interbeds formed in lacustrine conditions which support local aquifers that produce low discharges. Minor aquifers are present in the lower part of the Wallumbilla Formation (Smart et al. (9)). These aquifers are usually hydraulically unconnected and are of relatively minor importance compared with the underlying Gilbert River Formation and overlying Karumba Basin aquifers. 262

Few bores produce water from the Rolling Downs Group in the Carpentaria Basin. Those that do are subartesian and usually have quite saline water except in its very lowest part of the formation where it grades into the underlying Gilbert River Formation. The known conductivity of water from the Rolling Downs Group ranges from approximately 200 liScm-^ to 11 200 iiScm"^ in the study area. Groundwater sourced from the Gilbert River Formation has an average pH of 6.4. Further, average water quality parameters for the Carpentaria Basin are presented in Table 2. Table 2. Average Groundwater Analyses, Carpentaria Basin CARPENTARIA BASIN-AVERAGE GROUNDWATER ANALYSES * indicates data forms a skewed distribution and significant variations from mean values occur

Garraway Beds Gilbert River Formation Helby Beds Rolling Downs Group

Cond Hard SAR 1170

Alk

CI HC03 S04 N03

Sample Size

204

247

2.1*

227

245

61*

1.58

1340 87* 94 27 3800* 566*

17 6 20

223 81 277

258 23* 7.6* 276* 42.2 82* 80* 7 667* 119* 65* 975*

267 99 335

36* 1.0* 2.6* 2 0.1* 0.7* 336* 0.07* 2.6*

26 7 8

0.3

Conclusions The Mesozoic Carpentaria and Laura Basins provide significant and generally good quality groundwater over much of the Peninsula. The Laura Basin is a major source of groundwater in the southeast of the study area while the groundwater resources of the Carpentaria Basin are not widely used. This is principally because of the reliability of surface water supplies in the northem section of the study area and the occurrence of more easily accessed supplies from the overlying Tertiary Karumba Basin in the central and southem sections. Water quality from the principle Mesozoic aquifers is generally of low salinity though problems exist with high fluoride levels, corrosive water and high Sodium Adsorption Ratio (SAR) levels that can leave the water unsuitable for irrigation on clayey soils. Principle Mesozoic aquifers in Cape York Peninsula include the Dalrymple Sandstone, the Gilbert River Formation and the Garraway and Helby Beds with some minor supplies occurring in the Rolling Downs Group. However, few deep bores intersect these aquifers and in comparison to other areas of Queensland, the groundwater processes operating in the Peninsula are not well understood. Acknowledgments This paper is based on work undertaken throu^ the Cape York Peninsula Land Use Strategy (CYPLUS) program. The support and permission of the Queensland and Commonwealth Governments to present this work is gratefully acknowledged. References 1. 2. 3. 4. 5.

6. 7. 8. 9. 10.

Williams, L. J. GSQ Ebagoola 1: Preliminary Lithologic and Composite Log, Queensland Deapartment of Mines. Record 1988/14 Smart, J. and Rasidi, J.S., Geology and Petroleum Potential of the Laura Basin, Torres Shelf and Papuan Basin, Queensland. Queensland Government Mining Journal. June 1979. Passmore, V.L., Carpentaria and Karumba Basins Explanatory Notes and Stratigraphic Columns. Bureau of Mineral Resources, Record 1979/22. Hawkins, P. J. and WilUams, L.J., Review of the Geology and Economic Potential of the Laura Basin. Geological Survey of Qld Record 1990/2. Smart, J.R and Senior, B.R., Jurassic-Cretaceous Basins of Northeastern Australia In: Henderson, R.A., Stephenson, P.J. (Editors), The Geology and Geophysics of Northeastem Australia, Geological Society of Austraha Incorporated, Queensland Division, July 1980. McConachie, B.A., Filatoff, J. and Senapati, N., Stratigraphy and Petroleum Potential of the Onshore Carpentaria Basin, Queensland. Austrahan Petroleum Exploration Association Joumal, 1990. GSA/IWSC, Groundwater Resources of Queensland Explanatory Notes 1:2500000 Map. Geological Survey of Queensland, Irrigation and Water Supply Commission, Brisbane, 1973. Powell, B.S., Gibson, D.L., Smart, J., Grimes, K.G. and Doutch, H.F., New and Revised Stratigraphic Nomenclature, Cape York Peninsula. Queensland Government Mineral Joumal, April 1976. Smart, J., Ingram, J.A., Doutch, H.F. and Grimes, K.G., Recent Geological Mapping in the Carpentaria Basin New Stratigraphic Names. Qld Government Mineral Joumal, June 1971. Smart, J., Grimes, K.G., Doutch, H.F., Pinchin, J., The Mesozoic Carpentaria Basin and the Cainozoic Karumba Basin, North Queensland. Bureau of Mineral Resources Australia, Bulletin 202,1980

263

THE CONTRIBUTION OF AIRBORNE GEOPHYSICS TO EXPLORATION FOR MAGNETITE IN THE NORTHERN NEW ENGLAND FOLD BELT. Richie D Hubert John Siemon^ and Fred Bruvel^ 1 Department of Mines and Energy, Queensland. 2 Siemon JE Pty Ltd, Brisbane, Queensland 3 Department of Mines and Energy, Queensland Summary A simple demonstration of GIS technology focused at optimising areas suitable for the exploration of magnetite in the AIRDATA 94/95 program area is presented. This exercise used three selection criteria to distinguish suitable areas: 1) the presence of a strong magnetic anomaly coincident with an intrusive unit, 2) this intrusive unit is adjacent to a volcanic unit, and 3) the area is not the focus of ciurent exploration activity. DME datasets used by this exercise are the AIRDAIA 94/95 aerial geophysical data, the digital 1:2.5 million scale Queensland Geology data, the QSET tenure data and the histroical tenures data These datasets provided complete, continuous data over the area under consideration. The optimisation process identified four sites. Introduction Magnetite is a strategic mineral to the Queensland coal export industry. The mineral is economically significant to Queensland with production statistics for 1994-95 of 24 551 tonnes valued at $2.9 million. The Mount Biggenden mine is Queensland's only producer of magnetite. Exploration activities are currently seeking new resources of high quality magnetite to supply our local and overseas markets. AIRDATA is a $3.5 million Queensland Government program, to acquire high resolution airbome magnetic and radiometric data over 102 000 square kilometres of central eastem Queensland. These data are critical to re-mapping of the New England Fold Belt (NEFB) in Queensland as part of the Department of Mines and Energy (DME) GEOMAP 2005 program. Magnetic and radiometric results have been particularly valuable in re-mapping intrusive rocks of the NEFB, which range from granites to layered gabbros. The recognition of distinct plutons in previously undifferentiated batholiths, such as the Urannah Complex, the Rawbelle Batholith, and the Miriam Vale Granodiorite, and changes to the mapped geology of poorly known intmsive con^lexes, will significantly increase their prospectivity. The stated role of the GEOMAP 2005 program is "to generate new information needed to underpin new exploration programs...and so ensure that the recent discoveries ...are repeated."(Day, (1)) Our aim is to stimulate interest in the new datasets being generated by GEOMAP 2005 by a simple demonstration of GIS technology focused at optimising areas suitable for the exploration of magnetite in the AIRDATA 94/95 program area. Our desire is that this paper generates some form of action, either: • finitfiil discussions, or • raised level of awareness about available infomiation sources, or • experimentation with current and new datasets, or • the generate exploration tenure activity. The principles demonstrated here may be applied througji out the NEFB subject to the availability of suitable datasets. Optimisation Criteria & Data Sets Optimisation involves selecting the best areas based on a set of evaluation criteria. The contribution from airbome geophysics to exploration for magnetite in the NEFB is base on the expected strong magnetic field associated with the ore deposit. Geological models associated with this ore environment indicate the importance of adjacent intrusive and volcanic rock units coupled with limestone occurrences. The final criteria is dictated firom DME's perspective to maximise exploration activity This exercise uses three selection criteria: 1. 2. 3.

The presence of a strong magnetic anomaly coincident with an intrusive unit, This intrusive unit is adjacent to a volcanic unit, and The area is not the focus of current exploration activity.

The primary difficulty in relating magnetic results to geological entities is the ambiguous nature of this inversion. Clark & others(2) indicates there are many geological factors that influence magnetic properties including: lithology, depositional environment, tectonic setting, geochemical aflSnities, hydrothermal alteration, metamorphic grade, structure and rock age. Mclntyre(3) identified the major magnetic minerals as magnetite, pyrrhotite, ma^emite, native iron, 264

magnetic hematitie and some ilmeno-hematites. Magnetite is the most abundant of these minerals, and most magnetic anomalies can be associated with this mineral. In Queensland, magnetite occurrences have been found in a number of different geological settings but most have proved to be uneconomic (Bruvel & others(4)). The Mount Biggenden mine is the only currently operating magnetite mine. The magnetite body is a typical example of a contact metasomatic skam. This skam developed where the Degilbo Granodiorite intruded the older Biggenden Beds. Massive magnetite was formed through the replacement of limestone by reaction with hydrothermal solutions originatingfromthe intrusive unit (Weekes & others(5)). Bruvel & others(4) pubhshed a table of magnetite occurrences in Queensland which is reproduced below. TABLE 1. Magnetite occurrences in Queensland Name Latitude/Longitude Alma Creek: 23®54'S; 150^21'E Mount Biggenden 25°32'S; 151°59'E Calcifer 17°12'S; 144°34'E Cloncurry River 20°22'S; 140°40'E Cut Creek 21°30'S; 148°46'E Emest Henry 20°26'S, 140°43'E Gillian 17°43'S; 145°02^E Glassford Creek 24®35'S; ISriS'E Hawkwood 25^48'S; 150^51T Iron Island 22°00'S; 150°08'E Iron Range 12°40'S; 143°I5'E Mount Cardwell 17°38'S; 144°57'E Mount Leviathan 20°44'S; 140°28'E Mount Lucy 17°2rS; 144°39'E Mount Moss 19°07'S; 145°50'E Mount Philp 21°00'S; 139°57'E Mount Podge 19°18'S; 145°57'E Mount Wyatt 20°53'S; 147°17'E Osbome 22°06'S; 140°35'E Police Creek 21°25'S; 147°20'E Selwyn 21°3rS; 140°30'E Wild River 17°34'S; 145n9'E Woodstock 19°37S; 146°42'E The bold type indicates deposits that occur in the AIRDATA 94/95 survey area. Bmvel & others(4) briefly discusses most of these deposits and we refer you to that discussion for additional information. DME datasets used by this exercise are the AIRDATA 94/95 aerial geophysical data, the digital 1:2.5 million scale Queensland Geology data, and the QSET tenure dataset. These datasets provided con5)lete, continuous data over the area under consideration. Optimised Areas The optimisation process identified four sites. These geologic environments are summarised in Table 2. A check of DME's historical exploration tenure dataset indicates that all of the areas had previous base metal exploration activity. Only Area 3 was evaluated as a potential magnetite deposit by the Coal Division of CSR. The grade was found to be unsatisfactory (Biggs(6)). There are a number of other potential targets identified by this process. Ignoring the third criteria, all of the identified deposits listed in Table 1 (metasomatic and magmatic types) in die survey area were selected by this process.

265

TABLE 2. Magnetite optimised areas. Descriptive Easting/northing Area 000 Location Zone 56 Sheet East of 1 271037/7406610 Gladstone Rockhampton 2 Mount

215028/7388691

Morgan 3

251531/7331826

Biloela 4 Scoria

269707/7254468

Volcanic

Intrusive

Unit

Berserker

Unnamed basic to

Beds

NWof

Unnamed volcanic

ultramafic mass. Mainly serpentinite. Bouldercombe

Mt Morgan

unit, Kv.

Complex

WSWof

Owl Gully

Mount Gerard

Gladstone

Volcanics

Complex

SE of

Tertiary

Wingfield

basalt/Rannes Beds

Adamellite

1:100

Banana

Biloela Conclusion We set out to demonstrate GIS technology focused on optimising areas suitable for the e}q)loration of magnetite in the AIRDATA 95/96 program area. Based on the defined set of criteria, four areas of interest were identified. A cursory review of previous base metal exploration efforts confirmed the presents of magnetite in three of the four areas. One of the areas was evaluated as an unsuitable source of miagnetite. The technique is suitable first pass exploration filter. All of the known magnetite deposits (Bold print, Table 1) in the program area were also identified and are the subject of current exploration programs. It was also our intention to focus interest on the new datasets being generated by GEOMAP 2005 mapping programs. The AIRDATA geophysical data set is thefirstof these datasets. This exercise has demonstrated that GIS technology will require the DME to redesign there Geoscience Information Products. The traditional map product is simply a picture. New products will need to have a relational database associated with this picture (a multi-dimensional map). Optimisation processes will become more sophisticated as our geoscience data sets mature. The type of queries applied to a datasets is determined by the information content available in digital product, that is the amount of information per spatial unit and the aerial extent of consistent information. References 1. Day, R W, GEOMAP 2005 Program: The Key to unlocking Queensland's Mineral Wealth, Queensland Government Mining Joumal, September, 1995, p 21. 2. Clark, D A, French, D H, Lackie, M A, Schmidt, P W, Magnetic petrology: Application of integrated rock magnetic and petrological techniques to geological interpretation of magnetic surveys. Exploration Geophysics, 23,1992, pp 65-68. 3. Mclntyre, J I, Geological Significance of Magnetic Pattems related to Magnetite in Sediments and Metasediments - A Review, Bulletin Australian Society of Exploration Geophysicists 11 no 1/2 June 1980, ppl9-34. 4. Bruvel J, Ewington, D J, Jones M R, Magnetite in Queensland, Queensland Government Mining Joumal, July, 1995, pp 23-31. 5. Weekes G J, Robertson, A C, Magnetite production by Commercial Minerals Limited , Australasian Mining and 6.

Metallurgy, The Sir Maurice Mawby Memorial Volume 2, Monograph 19, 1993, pp 1393-1395.

Biggs, M S, A-P 3210 Collards Creek, Final Report including Progress Exploration Report for Six Months Ending 4.02.83, CSRLtd., Unpublished DME company report number 12093, April 1983.

266

GEOLOGY OF THE GOLD RIDGE EPITHERMAL GOLD DEPOSIT, GUADALCANAL ISLAND, SOLOMON ISLANDS R. D. James, L. Johnston and P. A. Ruxton Ross Mining (Solomon Islands) Ltd, PO Box 1556, Honiara, Solomon Islands Abstract - Gold mineralisation at Gold Ridge, central Guadalcanal, Solomon Islands is hosted within a distinctive volcaniclastic facies of a widespread epiclastic unit of the Lower Pliocene Epoch. H i ^ temperature, low sulphidadion epithermal mineralisation is characterised by a gold-pyrite association in quartz-calcite veins within illite-carbonate-quartz-pyrite alteration. The primary control on mineralisation is a set of extensional faults and associated low angle shears and tensional microfractures localised along a NE trending arc-normal fault zone. Exploration to March 1996 outlined proved and probable ore reserves of 19.54Mt at 1.65g/t for 1.035 million ounces of gold. A bankable feasibility study on Gold Ridge was recently submitted to the Government of the Solomon Islands. Pending agreements scheduled for completion in the short temi, the project will become the first mining operation in the Solomon Islands. INTRODUCTION SOLOMON ISLANDS

NEWZE/^^

Figure 1: Location Figure 1. Location The Gold Ridge gold deposits are in the Malango mountains of central Guadalcanal Island in the Solomon Islands, at latitude 9°40'S, longitude 160°08'E about 30km SE of the national capital, Honiara (Figure 1). Access from Honiara is by 35km of sealed and 10km of unsealed road. The main project area covers about 5km^ of steep jungle clad hills between 300m and 600m above sea level on the northem slopes of the central mountain range. The climate is warm and humid with rainfall ranging between 4,000-5,000mm per year in the project area. Gold was traced to its bedrock source in 1936 by panning in the streams and tributaries north of the Gold Ridge project area. Substantial modem exploration by CRA, AMOCO, Cyprus and ARIMCO culminated in two feasibility studies in 1990 and 1992. Ross Mining purchased the project in March 1995 and completed a bankable feasibility in July 1996 following field and test work. Drilling in the project to date totalling 66,000m identified three ore bodies, Valehaichichi, Kupers and Dawsons within a diamond shaped hydrothermally altered zone 2.5km long by 1.4km wide. Pending approvals from the Solomon Islands Government and the Gold Ridge Landowners Association, which represents the interests of the Gold Ridge people, construction is expected to commence early in 1997 and mining in mid 1998. A 2Mt per annum open cut gold mine is proposed with processing of ore at the same rate as mining in a treatment plant close to the mine site. The first gold pour is scheduled for the last quarter of 1997. Expected mine life on current ore reserves is 10 years at a production rate of about 100,00 ounces of gold per year.

267

REGIONAL SETTING PAPUA NEW GUINEA

Subduction Zone (Plate Boundary)

Major Gold Deposit

Figure 2: Tectonic Setting & Major Deposits The Solomon Islands and Papua New Guinea are part of a continental island arc system formed by the collision of the N moving Australia-India Plate with the SW moving Pacific Plate. Subduction of the Pacific Plate beneath the AustraUaIndia Plate resulted in partial melting of the Pacific Plate and diapiric rise of magmas into the domed Australia-India Plate. Fracturing of core rocks due to coUision and subduction generated a set of lineaments trending ESE and NNE. These primary structural elements were reactivated through time by ongoing tectonism and are referred to as arc- parallel and arc-normal structures. Arc-normal faults are transfer structures that are being increasingly recognised as important first order controls on the localisation of magmatic activity and mineraUsation (Etheridge & Henley, 1995). Several of the well known PNG mineral deposits (Ok Tedi, Porgera, Mt Kare) are located along transfer structures where they intersect with arc-parallel structures (Figure 2). Gold Ridge comprises three closely spaced deposits within hydrothermally altered volcaniclastic rocks over an area of about 5km^ astride the NNE to NE trending Melango fault (Figure 3). The Melango fault system corresponds to the short axis of the sigmoid-shaped island of Guadalcanal, presumably offsetting the uphfled Pre-Miocene oceanic basement to the SE from the younger rocks to the NW. In addition to locaUsing the Gold Ridge volcaniclastics, the Melango fault

268

Figure 3. Map of Guadalcanal (after Hackman, 1980) showing hot springs and major faults. system corresponds to a series of ultramafic and other basement fault blocks, minor volcanic accumulations and high level intrusives that include a porphyry prospect to the south of Gold Ridge. Etheridge & Henley (1995) recognised the influence of such deep seated arc-normal faults on the distribution of gold and gold-copper mineralisation in island arc settings in Indonesia, New Guinea and New Zealand, and consider these structures to be a key control on the distribution of large porphyry copper deposits in regions such as Chile. The most likely structural setting for the Gold Ridge volcaniclastic sub-basin was a pull-apart basin along a step in the Melango fault system, which appears to have originated as a transform fauh offsetting the Mio-Pliocene arc of Guadalcanal (Etheridge & Henley, 1995). PROJECT GEOLOGY Gold mineralisation at Gold Ridge is associated with argillic alteration (illite-carbonate-quartz-pyrite) and pervasive low order silica-pyrite alteration (quartz-illite-pyrite) usually as matrix infill. The three deposits are hosted by the Gold Ridge Volcanics (GRV) of the Lower Pliocene Epoch - a distinctive, shallow dipping volcaniclastic facies at least 500m thick (probably 900m) at the base of a more widely distributed epiclastic formation. An alteration map of the known mineralised zone with soil auger geochemical values plotted shows the strong association of gold with the argillic alteration (Figure 4).

269

Proposed Mine and Construction Camp

VALEH'AICHICHI

KUPERS

Argillic alteration boundary DnI! hole collar

Figure 4. Alteration & Gold Geochemistry The GRV sequence underwent pervasive argillic alteration characterised by illite and illite-kaolinite assemblages. Argillic alteration overprinted an earlier and more widespread propylitic alteration event (chlorite-smectite-carbonate-quartz). Valehaichichi hosts the most intense and concentrated argillic alteration. Propylitic alteration survived at Kupers and Dawsons where argillic alteration is less intense. Primary porosity of shallow dipping, clast supported lithologies was an important factor in the lateral dispersion of gold mineralising fluids away from subvertical, narrow fissures that provided the conduits. Steep dipping displacement zones trending NNW to NNE were recognised throughout the project area, although overall the GRV sequences are not significantly disrupted by faulting. Mapping and drill core orientations indicate broad, open folding caused by compressional tectonics. Anticlinal folding along a ENE to ESE trending axis was interpreted at Kupers, and a broad synclinal fold along a NNE axis at Dawsons. A combination of compressional and strike-slip deformation produced most of the moderate to shallow dipping secondary fractures and veins. Shallow dipping lithological controls as well as moderate to shallow dipping fractures and veins combine to impart a strong subhorizontal distribution to gold mineralisation (Figure 5).

270

ALTERATION I Tgc I

Clast supported Volcaniclastics

|-Tgm|

l^al"*supported Volcaniclastics

Intense silica pyrite - alteration ^ ^ H

Undifferentiated Volcaniclastics

Moderate silica pynte overprint alteration

]

Argillic alteration (lllite and lllite - kaolinite assemblages)

Propylitic alteration

East Kupers, Central Dawsons Unaltered ( D

N.W. Dawsons Supergene mineralisation zone

(§)

S E Dawsons

West Kupers

Valehaichichi

Auriferous veining quartz sulphide, cartx>nate sulphide, quartz cart)onate sulphide Penetrative faulting / fracturing Plane of broad open folds

Figure 5. Schematic Mineralisation Model Gold mineralisation at Gold Ridge is associated with argiUic alteration and pervasive low order silica-pyrite alteration. High gold grades are most frequently observed in the following: • Zones of high order silica-pyrite alteration. • Quartz, quartz-carbonate and carbonate-sulphide veinlets (pyrite > sphalerite, galena, chalcopyrite). • Zones of intense illite-kaolinite-pyrite alteration that are oftenfractured,locally crushed, or brecciated. ACNOWLEDGEMENTS The management of Ross Mining NL are thanked for their permission to pubUsh this paper. REFERENCES Etheridge & Henley Geoscience Consultants, 1995. Audit of Geological Model and Practices at Gold Ridge, Solomon Islands. Hackman, B.D., 1980. The geology of Guadalcanal, Solomon Islands. Overseas Memoirs, Inst. Geol. Sci. (Lx)nd.), No. 6.

271

TRIASSIC SEDIMENTATION IN THE GUNNEDAH AND SYDNEY BASINS, NEW SOUTH WALES Feng Xu Jian' and Colin R. Ward' 1 Western Australian Petroleum, GPO Box SI580, Perth WA 6001 2 Department ofApplied Geology, University of New South Wales, Sydney NSW 2052 Summary The Early to Middle Triassic strata of the Guimedah Basin have been subdivided on the basis of extensive drill-core studies into three formations, the Digby, Napperby and Deriah Formations. The Digby Formation, further subdivided into a Conglomerate Interval and a Sandy Interval, represents the product of a large-scale wet alluvial fan or braidplain system, derived from the uplifted orogen of the New England Fold Belt on the eastem side of the basin. This was succeeded by a south-flowing braided to meandering fluvial system with sediment derived from both the New England Fold Beh in the east and the cratonic Lachlan Fold Belt in the west. The Napperby Formation overlies an extensive palaeosol horizon at the top of this sequence. It represents a series of deltas derived from the New England Fold Belt building into an extensive lacustrine system. The Deriah Fomiation was deposited on top of these sediments by a lowsinuosity fluvial system, again derived mainly from the New England side. Comparison to the Triassic sequence in the well-known Sydney Basin suggests that the Digby Fomiation is equivalent to the alluvial fan and fluvial deposits of the lower and middle Narrabeen Group, whcih were also deposited mainly by southerly-flowing river systems. Additional sediment was input to the Sydney Basin succession from an uplifted area of intermediate to mafic volcanics a few kilometres east of the present coastline. The Napperby Formation is equivalent to the upper Narrabeen Group, the Hawkesbury Sandstone and the Wianamatta Group, which were deposited mainly by easterly and north-easterly flowing streams discharging into what appears from the Gunnedah Basin study to have been an extensive lacustrine system. The transition from Digby to Napperby type sedimentation (and the respective equivqalents in the Sydney Basin) is thought to be due to a major tectonic readjustments, which essentially reversed the pdaeoslope and re-arranged the relation of the water table to basin topography in the Sydney-Gunnedah block. Introduction The Sydney and Gunnedah Basins occupy the southem part of the Sydney-Bowen Basin. They are separated from each other by the Coricudgy Anticline, and from the Bowen Basin in the north by the Moree High (Tadros, 1993). Although they had a more complex earlier history (Tadros, 1993), in the Early to Middle Triassic they were part of an extensive north-south trending foreland basin, with the basin fill faulted by the Hunter and Mooki Thrust Systems against the newly emergent orogen of the New England Fold Belt in the east and lapping unconformably on to the older cratonic Lachlan Fold Belt in the west. Stratigraphy The Triassic succession in the Gunnedah Basin overiies with local unconformity the Permian coal-bearing strata of the Black Jack Group (Tadros, 1993; 1995). It is overlain, also with local unconformity, by the Jurassic beds of the Garrawilla Volcanics and the Purlawaugh Formation, which are part of the overlying Surat Basin sequence. Although other names have been used by different workers for different parts of the sequence, Jian and Ward (1993) used basin-wide borehole studies to divide the Gunnedah Basin Triassic sequence into three units: the basal Digby Formation, the overlying Napperby Formation and the uppermost Deriah Formation (Table 1). The Digby Formation is equivalent in age and general lithofacies to the lower and middle Narrabeen Group (Ward, 1972), the Napperby Formation to the upper Narrabeen Group (McDonnell, 1974), the Hawkesbury Sandstone (Standard, 1969; Cona^an and Jones, 1975), and the lower Wianamatta Group (Herbert, 1980), and the Deriah Formation to the upper Wianamatta Group of the better-known Sydney Basin succession.

272

Table 1: Triassic Stratigraphy in the Sydney and Gunnedah Basins (after Jian, 1991) GUNNEDAH BASIN

SYDNEY BASIN

DERIAH FORMATION

Interval E Interval D

Bringelly Shale

WIANAMATTA

NAPPERBY

Interval C

Ashfield Shale

GROUP

Interval B

HAWKESBURY SANDSTONE

FORMATION

Interval A

Newport Formation

DIGBY

Sandy Interval

Bald Hill Claystone to Munmorah Conglomerate (and equivalents)

FORMATION

Conglomerate Interval

NARRABEEN GROUP

Digby Formation The lower part of the Digby Formation is made up of persistent beds of lithic conglomerate, forming a section referred to by Jian and Ward (1993) as the Conglomerate Interval This is overlain by a more sandy succession, referred to as the Sandy Interval. Quartz-lithic sandstone dominates the basal portion of the Sandy Interval, especially in the eastem part of the basin, and quartzose sandstone dominates the remainder of the Sandy Interval succession. Conglomerate Interval The overall distribution of the Conglomerate Interval and the aggregate thickness of conglomerate within the unit both show a lobate geometry, with two separate areas of thick accumulation against the Hunter-Mooki Thrust System on the eastem side of the basin and a progressive thinning of both lobes towards the west. The lower part of the sequence in the east consists mainly of massive, clast-supported pebble conglomerate (resembling the Gm fades of Miall, 1978 and other workers) with pebbles up to 60 mm in diameter, minor horizontally-bedded to cross-bedded sandstone (sometimes pebbly or granule-bearing), and a small proportion of shale or mudstone. On the basis of its lithofacies features and its relationship to other facies in the sequence (Jian and Ward, 1993), this section is interpreted as a proximal fan and braidplain deposit (Figure la), formed by braided streams mainly of the Scott type (Miall, 1978) derived from the elevated uplands of the New England Fold Belt. The upper part of the Conglomerate Interval, especially in the west, contains a significantly higher proportion (up to 70%) of massive, parallel-bedded and cross-bedded sandstone, with minor grey to grey-brown shales showing flat bedding, lenticular bedding and in some places burrowing. These sediments are interpreted as more distal fan deposits, formed in a braided stream environment such as that represented by the Donjek model of Miall (1978) or in a highly sinuous river system like that on the humid alluvial fan of the Po plain in northem Italy (Ori, 1982). Sandy Interval The quartz-lithic sediment of the Sandy Interval has a similar distribution to the beds of the underlying Conglomerate Interval. Two lobate depocentres are developed at the base of the interval in the north-east and south-east of the basin, thinning westwards away from the Hunter-Mooki Thrust. The quartzose sandstone and associated shales that make up most of the Sandy Interval, however, are much more widely distributed, with an axis of thickening along the centre of the basin and maximum deposition in the south-central region. Several different types of lithofacies are present, including conglomeratic and sandy deposits identified by Jian and Ward (1993) as channel-fill deposits, thin sandstones or interlaminated sandstone-shale sequences interpreted as crevasse-splay and levee deposits, and shaly sequences of different types referred to as lake-fill, well-drained and poorly-drained swamp deposits. Distinctive light-colored kaolinite clayrocks (flint clays) are also present in parts of the Sandy Interval sequence. The Sandy Interval of the Digby Formation is interpreted as the product of a southerly-flowing braided to meandering river system (Figure lb), with the quartz-lithic material derived from the New England Fold Belt on the eastem side of the basin and the more quartzose sediment derived from the Lachlan Fold Belt on the western side. At the top of the sequence is a persistent off-white mudstone, interpreted by Jian and Ward (1993) as a basin-wide palaeosol horizon. This unit apparently marks a hiatus between the alluvial fan and fluvial deposition of the Digby Formation and the more quiescent lacustrine to lacustrine delta deposits of the overlying and generally finer-grained Napperby Formation.

273

Napperby Formation The Napperby Formation is an overall coarsening-upwards sequence, with dark grey shale and interlaminated to thickly interbedded sandstone and shale in its lower part grading to a dominantly sandstone succession in its upper part. Althou^ detailed subdivision is difficult, the sequence has been broken up for genetic discussion into three separate intervals, referred to by Jian and Ward (1993) as Interval A at the base. Interval B in the middle and Interval C at the top. The top of the unit is taken at the base of a distinctive off-white sandstone with abundant perthitic fragments or in its absence the base of a distinctive green lithic sandstone, either of which mark the base of the overlying Deriah Formation. Lithofacies recognised in the Napperby Formation include: • Dark grey shales with sideritic laminae and normally-graded sandstone/shale laminites with load casts, flame structures and horizontal burrows; these are identified by Jian and Ward (1993) as offshore lake and prodelta deposits; • Wavy-bedded sandstone/shale laminites, usually strongly bioturbated, which are interpreted as near-shore lake deposits; • Coarsening-upwards laminite sequences, with sandstone increasing upwards in abundance in each case, interpreted mainly as delta-front deposits. These are moderately to strongly bioturbated and commonly contain ripple crosslamination; hummocky cross-stratification has also been identified in some intervals; • Similar sequences, usually heavily bioturbated and often with shale clasts near the top and root structures on the upper surfaces are identified as interdistributary bay, levee and crevasse-splay accumulations; • Fine to medium-grained sandstone with parallel-bedding, ripple cross-lamination or climbing ripples structures, interpreted as stream-mouth bar deposits; • Fining-upwards sandstone sequences with medium-scale cross-bedding, intemal erosion surfaces, shale clasts, claydraped ripples and other features; these are thought to represent fluvial distributary channel deposits; • Dark grey rooted mudstones with siderite nodules and abundant organic matter, and hghter grey-brown, often mottled shaly sediments, thought respectively to represent poorly-drained and well-drained swamp accumulations; The relative abundance of sandstone within the various Intervals is highest along the eastem side of the basin, against the Hunter-Mooki Fault, and decreases steadily towards the Lachlan Fold Belt in the west. On the basis of both this regional lithofacies pattem and the features of the individual strata as seen in drill core, the sediments are interpreted as having been formed by deltas flowing from the New England Fold Belt into a basin-wide lacustrine system. Detailed study suggests that the deltas forming Interval A were mainly of a broadly lobate type (Figure Ic), possibly reflecting supply from a series of coalescing fans and sheet flows, while those forming Interval B were of a more digitate form (Figure Id^ fed by a relatively small number of master channel systems. Interval C is thought to represent the final stage of river progradation, with the lake being almost totally filled and the area converted to a fluvial depositional system. Deriah Formation Due to removal from other areas by post-Triassic erosion, the Deriah Formation is mainly confined to the northem part of the basin. The lower part is dominated by a distinctive green lithic sandstone, and the upper part by off-white lithic sandstone with interbedded mudstone and minor coal bands. The lithofacies are similar to those represented by channel deposits in the underlying Napperby Formation, and the sequence as a whole is interpreted essentially as a low-sinuosity fluvial deposit. The geometry and lithofacies distribution suggest sediment influx mainly from the New England Fold Belt. Relation to Sydney Basin Deposition The Digby Formation is very similar in lithology and inferred depositional environment to the lower beds (the Clifton Sub-group) of the essentially coeval Narrabeen Group in the Sydney Basin to the south (Ward, 1972). Both sequences are interpreted as representing alluvial fan and braided stream deposits, fiUing the basin initially from its faulted margin with the New England Fold Belt. They also pass laterally and upwards into more sandy fluvial successions, with streams sourced from both the orogen and the craton sides flowing southwards along the axis of the foreland basin system. In the Sydney Basin, however, there is also a secondary source of sediment, mainly consisting of reworked volcanic material, that was input from an elevated area to the south-east, offshore of the present coastline (Ward, 1972; Bradley, 1993). The palaeosol at the top of the Digby Formation is probably equivalent to the Bald Hill Claystone, a Sydney Basin unit with an unusual mineralogy that apparently represents widespread input into the otherwise quiescent south of the basin from the southeastem volcanic source area. Taken together, the two markers are thou^t to represent weathering during a major depositional hiatus, possibly developed after the basin had reached equilibrium and before the next phase of deposition commenced. The Napperby Formation in the Gunnedah Basin represents a succession of relatively localised deltas input to an extensive lake system from the New England Fold Belt. In the Sydney Basin this phase of sedimentation is represented by the lithologically similar Gosford and Newport Formations, which probably represent part of a deltaic succession built into a similar lake increasingly from the west (McDonnell, 1974), and the overlying Hawkesbury Sandstone, a quartzose deposit formed by a large-scale, north-east flowing braided stream pattem, comparable to the modem Brahmaputra system (Conaghan and Jones, 1975). 274

Although a large distance separates the present-day remnants of the respective units, the upper Napperby Formation is equivalent to the Ashfield Shale of the Wianamatta Group, a dark grey shaley sequence interpreted as representing lacustrine to possibly shallow-marine deposits (Herbert, 1980). The Deriah Formation is essentially coeval with the overlying but also widely separated Bringelly Shale (Herbert, 1980), the upper part of which contains an increasing proportion of fluvio-deltaic lithic sandstone beds. Conclusions Integration of studies in the Sydney and Gunnedah Basins suggest two separate phases of Triassic sedimentation. During the Early Triassic, large-scale alluvial fans consisting mainly of lithic debris were built into the basin from the orogen side (Figure 2), followed as sediment input declined by the more basin-wide development of braided to meandering river systems. Input of quartzose sediment from the Lachlan Fold Belt became progressively more significant as this phase proceeded, possibly due to post-overthrust uplift of a forebulge to the west of the New England Fold Belt. The second phase of sedimentation is represented by an extensive lake system across the Sydney-Gunnedah Basin, into which the deltas of the Napperby Formation were deposited in the north and the braided streams forming the Hawkesbury Sandstone flowed from the south-west (Figure 3). The transition between the two phases is ascribed to a regional change from a mainly southerly to a mainly northerly palaeoslope in the Sydney Basin, with the resulting adjustment of the ground and surface water levels and a more impeded drainage system giving rise to Napperby lake development. Acknowledgements The Gunnedah Basin study described in this paper was completed as a PhD project at the University of New South Wales, sponsored in part by the National Education Commission of the People's Republic of China. Access to drill cores and other records, as well as considerable other advice and assistance, was provided by the New South Wales Department of Mineral Resources. Additional financial support was provided by the UNSW/CSIRO Collaborative Research Fund. Particular thanks are expressed to Vic Tadros, Doug Hamilton, Lionel Etheridge, Andrew McMinn, Michelle Smyth, John Roberts and Dick Evans for assistance with different aspects of the project. References Bradley, G.M., 1993. Evolution and hydrocarbon prospectivity of the offshore Sydney Basin - NSW/PIO. In: Proceedings of New South Wales Petroleum Symposium, 2 June, 1993, (ed. C.J. Swarbrick and D.J. Morton), Petroleum Exploration Society of Australia (NSW Branch), 33pp. Conaghan, P.J. and Jones, J.G., 1975. The Hawkesbury Sandstone and the Brahmaputra: a depositional model for continental sandstones. Journal of the Geological Society of Australia 22, 275-283. Herbert, C. 1980. Wianamatta Group and Mittagong Formation. In: A Guide to the Sydney Basin (ed. C. Herbert and R.J. Helby), Geological Survey of New South Wales Bulletin 26, 254-272. Jian, EX. and Ward, C.R., 1993. Triassic Depositional Episode. In: The Gunnedah Basin, New South Wales (ed. NZ. Tadros), Geological Survey of New South Wales, Memoir Geology 12, 297-326. McDonnell, KLL., 1974. Depositional environments of the Triassic Gosford Formation, Sydney Basin. Journal of the Geological Society ofAustralia 21, 107-132. Miall, A.D., 1978. Lithofacies types and vertical profile models in braided river deposits: a summary. In: Fluvial Sedimentology (ed. A.D. Miall), Canadian Society of Petroleum Geologists Memoir 5, 597-604. Ori, G.G., 1982. Braided to meandering channel pattems in humid region alluvial fan deposits. River Reno, Po Plain (Northem Italy). Sedimentary Geology 31, 231-248. Standard, J.C., 1969. The Hawkesbury Sandstone. In: Geology of New South Wales (ed. G.H. Packham), Journal of the Geological Society ofAustralia 16,407-417. Tadros, NZ., 1993. The Gunnedah Basin, New South Wales. Geological Survey of New South Wales, Memoir Geology 12, 649 pp. Tadros, NZ., 1995. Gunnedah Basin. In: Geology of Australian Coal Basins (ed. C.R. Ward, H.J. Harrington, C.W. Mallett and J.W. Beeston), Geological Society of Australia Coal Geology Group, Special Publication 1,247-298. Ward, C.R., 1972. Sedimentation in the Narrabeen Group, southern Sydney Basin, New South Wales. Journal of the Geological Society ofAustralia 19, 393-409.

275

Figure 1: Palaeogeographic reconstructions for key intervals of the Gunnedah Basin Triassic sequence (Jian and Ward, 1993). a) Conglomerate Interval, Digby Formation; b) Lower Sancfy Interval, Digby Formation; c) Interval A, Napperby Formation; d) Interval B, Napperby Formation

276

Figure 2: Palae<^eographic reconstruction of Sydney-Gunnedah Basin deposition in the Early Triassic (Digby Formation and Lower Narrabeen Group). Arrows represent palaeodrainge directions. After Ward (1972), Jian (1991), Jian and Ward (1993) and Bradley (1993).

277

Figure 3: Palaeogeographic reconstruction of Sydney-Gunnedah Basin deposition in the Middle Triassic (Napperby Formation and Hawkesbury Sandstone). Arrows represent palaeodrainage directions. After Standard (1969), Ward (1972), Jian (1991), Jian and Ward (1993) and Bradley (1993). 278

GOONDICUMILMENITE DEPOSITS Laurie G. Johnson^ and Graham Lee^ 1 Managing Director Monto Minerals NL 2 Mining & Geological Consultant, Peter Stitt & Associates Pty Ltd Summary Traditionally ilmenite sands have been mined from coastal environments. The beginning of the titaniferous feedstock industry in Eastern Australia was during the 1930's when rich deposits of heavy minerals were mined and processed from coastal areas. This has now all but ceased and exploration has been directed in many instances towards less sensitive areas inland. The Goondicum flmenite Deposits occur approximately 90km west of Bundaberg in a highly mineralised area in the Monto District. Once in production these non-marine deposits will be one of few such deposits operating in the world. Introduction The Monto Minerals Group holds five Exploration Permits for Minerals covering two ilmenite and titano-magnetite bearing gabbroic intrusions and alluvial deposits derived from them, in central Queensland, near the town of Monto. Size, grade and quality considerations are such that these tenements cover what is potentially a world-class titanium resource. The ilmenite exhibits high FeO and low Fe203 which makes it suitable for sulphate-route Ti02 pigment production, since high Fe0:Fe203 ratios contribute to high reactivity. Further, chemical consistency, low Cr203 content and low content of the radioactive elements uranium and thorium compare favourably with major deposits currently in production elsewhere in the world. The results of analyses have indicated that ilmenite from Goondicum is suitable for titanium dioxide (Ti02) pigment production - preliminary acid-leach synthetic rutile process tests have given very encouraging results with products derived from the deposit containing greater than 96% Ti02. The Goondicum Complex is a plutonic intrusion situated within the northem New England Geosyncline. The Goondicum Complex and Bumett Valley host five types of ilmenite-bearing resource. Two of these, the alluvials in the Burnett channel and the flood plain deposits beside this channel, will be the most important resources for mining in the early years of Monto Minerals Group operations. Geology Located 90km west of Bundaberg and 30km east of Monto the Goondicum Complex outcrops as a circular body covering 30km2. The Complex is a layered mafic intrusion, mostly comprised of gabbros, leucogabbros, with some oxide gabbros. The layers appear to represent segregations of the intrusion and are fiirther intruded by late-stage oxide (magnetite and ilmenite) gabbros. According to Groen (1993) the Goondicum Complex is one of a number of intrusions, which probably represent the final stage of subduction. He obtained a K-Ar age of Cretaceous (96 Ma) for homblende separated from the gabbro. Groen says that based on similar gabbro intrusions elsewhere in south-east Queensland he would expect an age of 200250 Ma (Triassic). A marginal contact homfels zone forms erosion-resistant relief around the Complex and is cut by both the East and West Bumett Rivers draining the intrusion. About 1km south of the intrusion the East and West branches combine to form the Bumett River. The drainage direction is to the south-west and about 25km downstream the channel breaks out of rugged terrain into more gently undulating terrain, where large quantities of ilmenite-bearing alluvial sediments have been deposited to form an extensive flood plain on each side of the modem channel. The five resource types associated with the Goondicum Complex are: i.

Oxide Gabbros which occur as sill-like bodies appear to be the major source of primary ilmenite and contain the highest Ti02 content of any rock type con:5)rising the complex. In addition to ilmenite (order of 20%) they also contain titaniferous magnetite. Two main zones of oxide gabbro and numerous smaller discontinuous outcrops have been observed. One of the oxide gabbro bodies, occurring as part of the innermost circular sill, has a higher opaque mineral content than the average, with 19.9% Ti02 and 69.3% Fe304. Testing has not yet been completed to determine the titanium mineralogy of this rock, but the chemical analysis suggests it could be suitable as a lump smelter feed material.

279

In addition, some biotite gabbro and laminated gabbro rocks may be of lesser economic interest as a source of much of the detrital ilmenite. Typically the opaque oxide contents are in the order of 10% in these rocks and they outcrop over larger areas than the oxide gabbros. Chemical analysis of ilmenite from the Complex is shown in Table 1. They have low Cr203, U308 and Th02 which are desirable for Ti02 pigment production. Table 1. Goondicum Ilmenite Chemistry Ti02

FeO

FejOj

52.8 Gabbro (Rock) Eluvial (Goondicum Soil) 50.8 Tertiary (Tellebang Plateau) 53.7 50.6 Alluvial (Bumett River) Flood Plain (Bumett River) 50.9

42.8

19.2 35.8

21.7 9.2

SAMPLE

AI2O3

Cr203

CaO

0.02 0.26 0.28

0.03 <0.02

MgO

P2O5

<0.02 0.18

2.3 2.90

MnO 0.70

3.81 0.04 0.08

2.4 0.84

* SEM analysis reported total Fe as FeO. The gabbros outcrop at the surface and, where of economic grade, should be amenable to conventional open pit mining. In terms of the overall project, the gabbro has significant potential, but requires more extensive and costly geological investigation than any of the other resource types. Ilmenite resources contained in the hard rock of the Goondicum Complex are potentially large and may ultimately be shown to be in the range of some tens to hundreds of million tonnes. ii.

Eluvial Deposits containing ilmenite occur on and just below the surface over large areas of the Complex. Sampling and testing covering an area exceeding 10km2 has been undertaken. The eighteen samples tested indicated mean ihnenite grades of 7.8% (range 3.5 to 19%) and mean thickness 2.6m (range from 1 to 7m); however this does not represent the full thickness of the eluvial deposits.

iii.

Tertiary Alluvials of the Tellebang Plateau are shown on Figure 1. Chemical alteration of this deposit is reflected in the different ilmenite composition, with higher Ti02 and oxidised iron compared to iknenite from the Goondicum Complex. During Tertiary times the ancient Bumett followed a course to the west of the modem river. It deposited sand and gravel which now fomi the Tellebang Plateau shown on Figure 1. Subsequently these sediments were lateritised, producing a cap which has protected underlying ilmenite-bearing alluvials from later erosion. When the Bumett River later changed course, it eroded areas to the east. The plateau now provides a secondary source of ilmenite draining into the river principally via Tellebang Creek.

iv.

Alluvial Channel Deposits occurring in the modem Bumett River contain ilmenite. Monto Minerals has directed most of its exploration effort into defining this resource. The deposits become finer and better sorted downstream from the Goondicum Complex, while ilmenite grades decrease slowly as the Bumett River alluvium is diluted with smaller quantities of barren sediment from tributaries Ilmenite distribution within the sediments is influenced by local geomorphology of the stream bed and banks, and appears to be preferentially concentrated into the sandier fractions of the total sediment load. This gives the alluvial deposits a heterogeneity which requires a more rigorous evaluation than was initially thought necessary.

Flood Plain Deposits beside the modem channel of the Bumett River have formed where the drainage, due to lower gradient, is unable to carry the full sediment load. The deposits cover an area exceeding 50km2, into which the modem drainage channel is incised. Recent drilling and testing has shown that the flood plain material generally comprises a gravel rich basal deposit, overlain by fine silts which in tum are covered by overbank sand and silt. Ancient sand levee banks which occur as surface ridges parallel to the river have been observed at a number of locations within these flood plains.

Comparing the Goondicum deposit with other commercial hard rock ilmenite deposits it is important to note that nearly all are associated with anorthosites or gabbros. In the Western World there are two producers from hard rock deposits, Titania A/S at Telhies, Norway, and QIT at Allard Lake, Quebec, Canada.

280

Gladstone 128 km

— 25" 2 3 ' 0 0 " S

Monto Minerals tenements, Goondkum area. Figure 1. Tenement map, goondicum area

281

Table 2 nmenite Composition Comparison BEACH DEPOSITS HARD ROCK DEPOSITS Eneabba Goondicum Bunbury Tellnes AUard Lake WA(3) WA(3) Canada(2) Qld Norway (1) (Weight Percentages) Element 60.8 54.6 45.0 37.7 50.6 TiO 23.2 3.7 35.8 34.0 28.8 FeO 29.5 9.2 16.8 38.9 12.5 Fe^O 0.84 1.07 0.2 1.5 0.3* MnO 0.80 0.70 0.6 0.53 2.8 Si<y 1.00 0.65 1.0 0.28 AlHy 0.6 0.22 0.14 0.16 0.36 0.11 0.27* 0.15 5.0 2.9 2.9 MgO 0.08 <0.04 0.035 0.01* 0.10 P^05 na na 0.1 0.18 0.25 CaO <0.02 0.035 0.20 <0.076 0.1 Cr03 150ppm* <lppm 20ppm* <10ppm* <10ppm* U 300ppm* <lppm 35ppni* <10ppm* <10ppm* Th (1) Lynd, (2) Lee (3) Towner na: No analysis available for this element. The Table above compares the Goondicum ilmenite composition with other products. The Ti02 content of Goondicum ilmenite is significantly higher than either Tellnes or Allard Lake ihnenites; MgO is less than Tellnes and comparable with Allard Lake. Compared to Bunbury and Eneabba, Westem Australia, Goondicum has lower Ti02 content, very h i ^ FeO and a low Fe203 content. The high FeO content is desirable for many consumers, particularly sulphate-route pigment producers, due to the higher reactivity of these ilmenites. Most importantly, Goondicum ilmenite is significantly lower in the radioactive elements uranium and thorium. In addition, Goondicum ilmenite is chemically more consistent than many other deposits mined currently. Discussion Goondicum is a large hard-rock deposit with extensive associated eluvials and alluvials. Quantities of contained ilmenite are very large and are considered as potentially world-class titanium resources. The work undertaken by the Monto Minerals Group to assess the Goondicum Indicated Resources of the alluvial channel reUed principally upon pitting and bulk sampling totalhng 633 tonnes, followed by pilot plant processing. This effectively is trial mining. Conventional small diameter drilling does not produce sufficient sample volume to reliably measure the grade of deposits containing heavy mineral concentrations associated with gravel. The coarse gravel causes unreliable sampling and leads to erroneous grade information; for instance the Goondicum alluvial channel contains a considerable quantity of+50mm gravel and to get a representative sample with a maximum gravel size of 50mm requires at least a 250kg sample, with coarser gravel requiring even larger samples. The techniques employed by Monto Minerals NL therefore, give a higher level of confidence in the results than drilling does, where sample sizes are smaller than the minimum required for representative sampling. The geological complexity of these deposits is becoming increasingly apparent with each step taken in exploration and development. Environmental considerations are critically important to the success of this project as is landowner relations. The fact that this is an inland deposit does not reduce the requirement for research into any matters associated with the Goondicum Ilmenite Deposits. References Groen, S G, 1993. Petrogenesis, Petrography, Petrology and Field Relations of the Goondicum Gabbro. Queensland University of Technology B App Sc Hons Thesis. (2) Lee, H Y, 1986. The Future Role of Titania Slags and Upgraded Ilmenite on Ti02 Feedstock Supply and Demand. Keynote Address to "Australia; A world source of Ilmenite, Rutile, Monazite, and Zircon." AusIMM Conference, Perth 1986. (1) Lynd, L E and Lefond, S J, 1975. Titanium Minerals, in Industrial Minerals and Rocks, pages 1149-1208. Edited by Lefond, S J Society of Mining Engineers of American Institute of Mining, Metallurgical, and Petroleum Engineers Inc. (3) Towner, R, 1990. Australian Mineral Sands Industry Symposium. Product Specification and Market Outlook. Austrahan Bureau of Agricultural and Resource Economics. 282

DEBRIS FLOW DEPOSITS IN THE LATE TRIASSIC CALLIDE COAL MEASURES AND THEIR IMPACT UPON COAL MINING CONSIDERATIONS. Peter J. Jorgensen and Christopher R. Fielding Department of Earth Sciences, The University of Queensland, Brisbane, 4072, Summary This paper presents some results of research undertaken by the authors as part of a PhD project (PJJ) at The University of Queensland. The aim of the project has been to investigate the sedimentology of the coalbearing sequence contained within the Late Triassic Callide Basin in east-central Queensland. One of the major findings of this work has been the recognition and documentation of debris flow deposits in the Callide Coal Measures. The study has shown that the flows are a significant element within the coalbearing sequence, and in some instances have strongly affected the lateral continuity of the main coal horizon in the basin. The debris flows have been recognised in both outcrop and drill core. Examination of Company drilling records identified other potential sites but evaluation of this information was difficult due to the wide variety of nomenclature used to describe these lithologies. However, the discovery that the flows have a characteristic response on wireline geophysical traces has allowed positive identification of the units, particularly in rotary chipped boreholes. Much of the investigation into the deposits has been carried out by utilising workstation-based mine planning computer software (MINEX GMC 3D on Silicon Graphics workstations). The flows have been modelled in three dimensions so as to assess their lateral extent, orientation and volume. These parameters have been used to determine the degree of coal loss from the resource models in the various areas affected by the debris flow activity. Significant reassessment of some of the coal deposits has resulted firom this work. Introduction The Late Triassic (Camian-Rhaetian) Callide Coal Measures are preserved in a small (22.5 by 8km), partly fault-bounded basin remnant (the Callide Basin) near the town of Biloela in east-central Queensland (Fig. 1). The largely undeformed basin fill unconfonnably overlies a variety of Palaeozoic and Early Triassic rocks which were strongly folded and thrusted during the Hunter-Bowen Event. The Callide Basin is interpreted to have formed after the orogenic event, during a period of modest crustal extension that opened elongate, partly fault-bounded troughs parallel to the predominant regional stmctural trend. The preserved coal measure sequence comprises up to 150m of clastic sedimentary rocks interbedded with four coal seam horizons, and includes one of the thickest black coal seams in Australia (the Callide Seam Member, up to 23m thick; Fig. 2). At present, coal is mined by opencut methods in several discrete areas within the preserved Callide Basin (Fig. 1) and is utilised for power generation in central Queensland. The Callide Coal Measures form a grossly fining-upward sequence, with a lower, conglomerate-dominated association which contains virtually no coal, passing upward abruptly into a finer grained, coal-bearing association. The basal conglomeratic portion was deposited within high energy, coarse grained alluvial channel environments. Abrupt waning of the supply of coarse detritus then allowed finer grained, low sinuosity fluvial systems to become established. The sediments of the upper association were deposited within these channel systems and their associated floodbasin environments. The coal seams were the product of extensive, long-lived peat mires that expanded and contracted across much of the basin. The facies content of the sequence and the architecture of the sediment bodies are considered characteristic of alluvial systems, and together with a lack of evidence of marine influence indicates that the entire sequence was of continental origin. Debris flows in the CaUide Coal Measures One of the major findings of the present study has been the recognition of debris flow deposits within the Callide Coal Measures. While the rocktypes now assigned to this facies association have previously been logged in both drill core and outcrop by a number of geologists, no positive interpretation of these units has ever been made. In the past the deposits have been variously described as "washouts" (Shepherd, 1951a; Grimstone, 1977), coarse-grained flood deposits (Lumley, 1985), channel bank collapse deposits (Behets, 1987), tufifs (M. Biggs, pers. comm.), fault breccias and even igneous intrusions. Following a thorough re-examination of drilling records and detailed field mapping, these deposits have now been positively identified and a significant database has been established. The debris flow deposits are now known to be an important component of the Callide Coal Measures.

283

^Gladstone i Brisbane rSydney

Kilometres

Figure 1. Maps showing the location of the Callide Basin in east-central Queensland, and the distribution of coal deposits and mining areas within the preserved remnant of the basin.

284

150^

MARKER SEAM HORIZONS 2 < HU CcJ W ^ o

125

100

CALLIDE SEAM MEMBER

0 'o • o' o 'o 'o SAWMILL SEAM LOWER SAWMILL -WRIGHT SEAM

UPPER ASSOCIATION

u 00 00 < 2 wH <J

.0* 0 . 0 .

SAWMILL SEAM HORIZONS

•o- '0*

BOTTOM SEAM HORIZONS

0 6 0

50-

<U m a.

25 LOWER ASSOCIATION

BASAL CONGLOMERATE MEMBER

O® « OO -o ^ o O o < — - OL AAAAAAAA Scale in metres A A A A A A A A A A A A A A A A A A A A A A

Figure 2. Composite stratigraphic column for the Callide Coal Measures showing the subdivision into lower, conglomeratic and upper, coal-bearing associations (after Biggs et al., 1995). The debrisflowshave no preferential pattem of occurrence within the stratigraphic column, having been recognised in both the basal conglomerate member and the upper, coal-bearing portion of the coal measures sequence. Two debrisflowfacies have been distinguished (Table 1) which, while being different in composition, are interpreted to have formed by the same process. 285

Debris flow fades association Table 1: Essential features of the two debris flow lithofacies recognised within the Callide Coal Measures. Facies B Fades A Organic debris flows Interpretation Clastic debris flows Brecciated coal, often with abundant Diamictite, granule to pebble breccia Lithology tuffaceous claystone fragments, coal & conglomerate, and sandy siltstone, has high clastic content, sharply all with grey-black silt-grade matrix, based units <1.5m thick. some have pumiceous and other felsic volcaniclastic particles, sharply based units <15m thick, many units composite. Massive to strongly disoriented Dominantly massive, minor clast Sedimentary fabric, tuff clasts may show plastic imbrication, possible loaded layers, structures deformation features. reverse and normal grading. Laterally restricted, up to at least Elongate to lobate, possibly bi-lobate Plan 200m long and 60m wide. or multi-lobed, up to at least 1500m Geometry long and 600m wide. Fades A - Clastic debris flow deposits Facies A comprises diamictites, granule to pebble conglomerates and breccias, and sandy siltstones of varying texture, fabric and composition. However, each lithology is characterised by the presence of the same distinctive, argillaceous, sandy siltstone matrix. The matrix consists of an unsorted mixture of silt, carbonaceous material, fine sand grade quartz and fragments of siltstone and tuff. It is either dark brown or dark grey-black in colour. Diamictite is the most common lithology and consists of small to moderately sized (less than 8cm) pebbles and granules of intraformational siltstone and sideritised siltstone, kaolinised tuff (tonstein) and coal, scattered throughout a poorly sorted siltstone matrix that may also contain some sand sized material and variable amounts of carbonaceous debris. The conglomerates are of a similar coii5)osition to the diamictites and exhibit a matrix-supported fabric in most instances. The breccias are the least common variant and contain abundant ragged felsic tuff clasts plus minor trachyte and intraformational siltstone clasts in a finegrained, dark grey siltstone matrix. The breccia is predominantly clast supported (70-80% clasts) and displays local clast imbrication. Facies A deposits are preserved as elongate to lobate bodies in plan form, some possibly bi-lobate or multi-lobed, many of which thicken distally and appear to have an abrupt, high profile termination. Individual bodies are up to 15m thick, with the majority in the range 0.5 to 2.0m Composite sequences up to 35m thick have been encountered in drillholes. The bases of units are sharp and tend to be planar but may locally be erosional. The upper surfaces of the deposits are also sharp and planar. Intemally the units are apparently unstratified in most instances, but some show a variety of vaguely formed sedimentary structures including indications of clast imbrication, normal grading, flat to low angle bedding and loading. In some instances the clasts within the basal few centimetres of the deposits are strongly reverse graded, but this fabric dissipates upward into the more typical unstratified fabric. Interbedded within some of the Facies A deposits are well stratified units 0.05 to 1.5m thick, composed of a variety of lithologies from siltstone to conglomerate and breccia. Their compositions are similar to those of the deposits they overlie. However, the beds are strongly horizontally laminated and some also exhibit well developed normal grading. Their contacts with the underlying bodies are commonly erosive. The tuff clasts contained within the various lithologies have a wide variety of shapes and sizes including many with elaborate, plastically-deformed edges which were probably deposited while still soft. The preservation of these fragile clast shapes probably reflects the largely non-turbulent nature of the depositional system. This, in conjunction with the unstratified character of the deposits and the ubiquitous presence of silt-grade matrix suggests that the units were formed through the action of debris flows. This interpretation is reinforced by the lobate shape of the deposits and the presence of abrupt, high profile terminations similar to those in modem day debrisflows(Johnson, 1970). Many of the bodies are thought to be composite, formed where several individual flows have followed the same, or a similar, path and were superimposed upon one another. In some instances appreciable time must have elapsed between successive pulses because significant thicknesses of other facies have accumulated between the debrisflowbodies. The features of the well-stratified units manthng some of the debris flows are indicative of tractional current activity. Thus these units are interpreted to represent portions of the original debris flows which underwent reworking by postdepositional stream flow activity across the top of the flows.

286

Fades B - Organic debris flow deposits This is a poorly exposed fades observed to date only in two opencut highwall faces; one within The Hut mining area and the other at Trap Gully (see Fig. 3). Further examples have been tentatively identified from the lithological logs of cored boreholes in the same general areas as these two highwall exposures. However, the primitive nature of the borehole logs precludes confident recognition since the chief characteristic of the facies (a strongly brecciated fabric) is a feature that can be artificially produced by the core drilling process. Facies B comprises thin, lensoidal bodies of coal which exhibit a prominent brecciated fabric and lack any bedding or lamination. Up to 1.5m thick, the bodies are symmetrical lenses in cross-section. The two examples identified in highwall closure were laterally restricted, extending over lateral distances of between 40 and 60m. Examination of the lithology logs from boreholes adjacent to the highwalls has allowed recognition of the bodies over longitudinal distances of approximately 200m. Although the coal within these bodies appears to be predominantly dull, no traces of the depositional cycles so prominent in the normal coal facies are preserved. Instead, Facies B consists of a chaotic mixture of mainly dull coalfragments(up to 25cm long), with lesser amounts of interbanded and bright coal fragments contained within a matrix of fine-grained coal debris and some clastic detritus. In places, abundant angular and elongate tuffaceous claystone (tonstein) fragments up to 12cm long are also contained within the brecciated coal. These clasts have no preferred orientation and some of them have delicate and intricate shapes which seems to indicate they were still plastic when incorporated into the coal debris. Some of the clast shapes are similar to those seen in the tuffaceous clast-rich breccias of Facies A. The bodies are sharply bounded at both top and base, with the basal surfaces either planar or slightly irregular. In places, the lower boundaries are marked by a thin (<5cm) clastic-rich parting which contains small sand to granule sized intraformational siltstone and claystone clasts as well as tonstein fragments in a silty coal matrix. This parting typically grades up into the main portion of the coal breccia. Although Facies B is composed dominantly of coal, the chaotic intemal fabric of the deposits is unlikely to have been formed during normal peat accumulation processes. We suggest that sections of in situ peat bodies underwent postdepositional transportation which induced mixing and brecciation of the peat, and any contained tonstein bands, before the mass was re-deposited in its current position. The preservation of the deUcate tufffragmentsindicates that the flow was probably non-turbulent. For these reasons a debris flow origin is invoked to explain the formation of this facies. Distribution of debris flow deposits in the CaUide Basin Utilising a combination of both new exploration and a re-assessment of archived Company and Government drilling records, eight areas of debris flow activity have been identified within the Callide Basin to date (Fig. 3). While minor outcrops of the debris flow facies have been discovered across the basin, the majority of the deposits have been recognised solely from subsurface information. Evaluation of this borehole information was initially hampered by the wide variety of nomenclature used to describe the debris flow lithologies. However, the discovery that the deposits have a characteristic response on wireline geophysical traces has enabled positive identification of the units, even in rotary chipped boreholes. The best results have been obtained with a SIROLOG neutron-gamma "ash rat" sonde which records an abnormally high response of between 0.30 and 0.45Mev for the debris flow sediments (normal clastic sediments register 0.20-0.275Mev). The elevated neutron-gamma reading probably reflects the high content of volcaniclastic debris within the deposits. Once the characteristic geophysical anomaly was recognised for the deposits, any existing boreholes suspected to have intersected debris flow deposits were re-logged with the neutron-gamma sonde and the intervals of interest were re-examined. This method of checking for the presence of debrisflowsusing the SIROLOG sonde is now used as a routine part of exploration activities, particularly in areas where it is thought that debrisflowactivity may have affected coal reserves.

287

;srGladstone To

To Gladstone

Kilometres

Figure 3. Map showing the location of the debris flow deposits identified in the Callide Basin. Areas with hatched pattern indicate clastic debris flows (Facies A). Solid fill indicates organic debris flows (Facies B). Central Reserve Debris Flow Deposits The Central Reserve debris flow deposits are the best known examples in the Callide Basin and have the most available data (all of it subsurface - Fig. 4). The Central Reserve coal deposit is a small resource of approximately 40 milHon tonnes located immediately to the north of the Dunn Creek Mine, the site of the most intensive mining activity in the basin. The current research has revealed that this is the area where debris flow activity has had the most significant affect on coal reserves. This finding, and the close proximity of the deposits to current operations, has made the detailed understanding of the debris flows a high priority for the mining company as it is anticipated that mining will commence in the area within the foreseeable future. At Central Reserve there are two clastic debris flow intervals in the sequence between the Marker Seam Horizons and the Callide Seam Member (Fig. 5). These deposits have characteristic elongate, multi-lobed shapes in plan view (Fig. 6) and are between 1.1 and 2.6km' in area. They appear to have formed as sheets of sediment, probably through the coalescence of two or more separate debris flows. At times there is a significant thickness of non-debris flow deposit between the two intervals, but in other places the two intervals form one continuous sequence. In cross-section the lowermost debris flows can be seen to have affected the lateral continuity of the coal seams in the Callide Seam Member (Fig. 5). Erosion by the flows of this DF2 interval has removed significant portions of the upper seams, and in one borehole to the east of the line of section (R6008), the complete coal-bearing portion of the CalUde Seam Member is missing and has been replaced by a 22.7m thick accumulation of debris flow sediment. In total, some 3 million tonnes of coal is thought to have been removed by debris flow activity from the Central Reserve area. This represents ahnost 7% of the original, pre-debris flow discovery, coal reserves calculated for the deposit. With further exploration and modelling, particularly in the eastern section of the deposit, this coal loss figure may yet be increased.

288

Central Reserve Debris Flow Deposits o Limit of the Central Reserve coal resource model

Cored hole Rotary chip

N Kilometres Figure 4. Location plan of the Central Reserve area indicating the outline of the Central Reserve coal resource and the distribution of boreholes in the area. Cross-section AA' is illustrated in Figure 5.

Central Reserve Debris Flow Deposits:Diagrammatic cross-section MARKER SEAM HORIZONS CALLIDE SEAM MEMBER

Debris flow DFl Debris flow DF2 Figure 5. Diagrammatic cross-section through the upper part of the Callide Coal Measures in the Central Reserve area illustrating the shape of the two debris flow intervals. See Figure 4 for location of cross-section.

289

Central Reserve Debris Flow Deposits

Debris flow deposit DFl ^^m Interpretedflowdirections ^ ^ (based on thickness isopachs) Debris flow deposit DF2 ^

Interpreted structural lineament

^ ^ Measured fault

i N 1

Kilometres Figure 6. Distribution maps of the debris flows in the Central Reserve area showing the close proximity of the deposits to prominent structural features. Interval DFl shown at top and interval DF2 at bottom. Debris Flow Formation Initial investigations into the mode of formation of the Calhde Basin debris flows deposits have highhghted a close association between the location of the deposits and the position of structural features (faults, lineaments) within the basin (Fig. 6). It is thought that the flows may have been initially generated by fault movements, perhaps "shocking" the heavil)' 290

water saturated sediments, causing them to move down slight slopes on the edges of peat deposits. Because of the sizeable component of primary volcaniclastic material within the deposits, it is apparent &at some volcanic activity was also occurring contemporaneous to the debris flow propagation. The fault movements may also have had some relationship to the volcanic activity. Unlike the deposits of true, laminar debris flows which are non-erosive (Selby, (7)), the Callide Basin examples were often associated with areas of significant erosion, particularly where the substrate was peat. There are two possible explanations for this; (1) the erosion occurred prior to the propagation of the debris flows and the flows utilised preexisting topography, or (2) the Callide Basin debrisflowswere partly turbulent in nature and eroded the substrate during their movement down slope. The high coal content in some of the deposits tends to suggest that the latter scenario was the dominant one, althou^ the presence of pre-existing flow paths cannot yet be fully discounted. Conclusions Recent investigations in the Late Triassic Calhde Basin of east-central Queensland have revealed the presence of prominent debris flow deposits within the main basinfillunit, the Callide Coal Measures. Two types of debris flow have been identified; a typical clastic-dominated form and a unique, coal-rich variety. Both are interpreted to have formed by similar processes which probably involved a close association with movements along faults within the basin. The debris flows appear to have been quite erosive in nature and in places have strongly eroded portions of the main coal seam interval in the coal measures. In one small coal deposit in the basin, some 3 million tonnes of the originally deposited coal have been redistributed by the action of the debris flows, thereby significantly reducing the in situ coal reserves of the deposit. Because of this effect on the coal deposits, the presence of debrisflowdeposits is to be routinely assessed in all future exploration activities in the basin. Acknowledgments Callide Coalfields Pty Ltd are acknowledged for granting permission to carry out this study and to pubhsh the results. Their provision of logistical and computing support has also been invaluable. Senior Geologist, Mark Biggs, is especially thanked for his technical input and personal interest in the project. References 1. Shepherd, S.R.L., The Calhde Coalfield. Drilling in the upper Dunn's Creek section. Queensland Government Mining Journal, 52, 1951, pp 133-141. 2. Grimstone, L.R., The geology of APCl 88 and the evaluation of the Callide Coalfields. Theiss Bros Pty Limited Mining Division report (unpubhshed), 1977, 57pp. 3. Lumley, D., A palaeoenvironmental analysis of the Callide Coal Measures in the north-eastem comer of the Callide Basin. Honours thesis. University of Queensland, 1985. 4. Behets, R.A., Geology of the southernmost Callide Basin, east-central Queensland. Honours thesis. University of Queensland, 1987. 5. Biggs, M.S., Burgess, A.W., & Patrick, R.B., Mesozoic and Tertiary Coal Deposits: Callide Basin. In, Ward, C.R., Harrington, H.J., Mallett, C.W., & Beeston, J.W. (editors), Geology ofAustralian Coal Basins, Geological Society of Australia Coal Geology Group Special Publication, 1,1995, pp 471-488. 6. Johnson, A.M., Physical Processes in Geology. Freeman, Cooper: San Francisco, 1970, 577pp. 7. Selby, M.J., Hillslope sediment transport and deposition. In, K. (ed.), Sediment Transport and Depositional Processes, Blackwell, 1994, pp 61-87.

291

ASPECTS OF TRIASSIC SEDIMENTATION IN THE BOWEN BASIN, QUEENSLAND

Jochen Kassan^^ and Christopher R Fielding^ 1 Consultant, 34-56 Whistler Court, Spring Mountain, Qld. 4124 2 Department ofEarth Sciences, University of Queensland, Qld, 4075 Summary Basin analysis of the continental Triassic section in the Bowen Basin of eastern Queensland, Australia, has allowed an improved understanding of controls on the basins evolution. Major depositional episodes are recognised, each of which gave rise to a distinctive suite of alluvial and lacustrine lithofacies. Each package is bounded by surfaces of erosion, non-deposition, flooding or their conformable correlatives. Comparison with the gradients of major rivers in modem orogenic belts and other tectonic settings, suggests that the depositional surface of the Bowen Basin may have been elevated several hundred to over 1000 meters above sea-level and consequently absolute variations in sea-level cannot be invoked to e^lain the observed facies patterns. Furthermore, significant evidence is here recorded in support of a tectonic control on basin infilling, including the presence along the eastern basin margin of thick, coarse alluvial fan deposits at several stratigraphic levels, and the apparent westward migration of these units through geological time. The alluvial fan deposits, which overlie major changes in facies that define depositional episode boundaries, are interpreted to have formed following discrete compressional tectonic events. While these events can be shown to have affected the entire basinfiromeast to west, it is evident that the cratonic westem basin margin preserves a different record to that of the orogenic eastem margin. Introduction The Bowen Basin of eastem Queensland is a Permo-Triassic sedimentary basin, which was positioned between an active orogenic belt to the east (New England Fold Belt) and a stabilised craton to the west (Fig.l). The Bowen Basin is contiguous southward with the coeval Gunnedah Basin and the Sydney Basin. The entire Bowen-Gunnedah-Sydney basin complex spans some 2000 km in length and attains a maximum preserved onshore width of ca. 200 km. The basin history is complex andfiromearly Late Permian times on the basin developed a strong east-west cross-sectional asymmetry with a depocentre located in proximity to the tectonically active eastem basin margin (Harrington et al (1)). During this time the basin acted as the foreland basin to the New England Fold Belt, as did the Gunnedah and Sydney basins to the south. Sediment accumulation terminated during a climax in compression associated with westsouthwestward directed thrusting in the late Middle Triassic (Ladinian). The stratigraphic nomenclature used in this paper is summarised in Fig.2 together with summary interpretations of lithostratigraphic units covered. Litho-stratigraphic units are defined by significant changes in sediment composition and depositional facies, and can be correlated to biostratigraphic zonation schemes.

292

Stages NariHiiafA im

237.5-

239.r 241.2" 241.9243.4_ 245.0,

Lad

upper Moolayember Fm

TraT3 Spa u12 5£

lower Moolayember Fm Snake Creek Mudst Showgrounds Fm

Expedition Sst Glenidal Fm

Arcadia Formation

3.4 3.3 3.2 M

2.2

2.1

Sagittarius Sandstone

Gri

Tat

Figure 1. Schematic location map of study area.

1= ®

oo .o

Ans

Nam

Price etal 1985

Lithostratigraphy

So m

Bandanna Formation

Figure 2. Summaty of top Permian and Triassic stratigraphy of the Bowen Basin

Stratigraphy Data from total of452 boreholes was utilised. Wireline logs from 140 stratigraphic and petroleum exploration boreholes were examined and correlated in detail. In excess of 2500 m of drill core was logged in the Triassic section with a significant proportion of this in fully cored stratigraphic boreholes. For the remaining wells formation tops picked by staff of the Queensland Department of Minerals and Energy were utilised in the construction of isopach and pinch-out maps. Our approach has been to identify genetically related stratigraphic intervals, that are bounded by regionally extensive surfaces. These may comprise unconformities and their correlative conformable surfaces, flooding surfaces, and surfaces of non-deposition, as well as major changes in sediment provenance or dispersal, and fundamental changes in environments of deposition. The stratigraphic units (depositional episodes) recognised are considered to represent the highest resolution time sHces that can currently be reconstructed for the entire basin. The resultant Triassic depositional episodes are numbered I to V in Fig.2. Higher resolution subdivision is possible in areas on the westem basin margin, with favourable well spacings and significant amount of exploration drill core. In particular the interval of Showgrounds/Snake Creek offers the potential for high-resolution stratigraphy based on maximum and subordinate flooding surfaces. The array of depositional environments for the Triassic of the Bowen Basin includes a variety of fluvial and standing water systems. Depositional Environments Drill core and outcrop sections were subdivided into lithofacies (e.g. laminated mudrock, ripple-cross-laminated sandstone, etc.) and subsequently grouped into genetic facies assemblages (e.g. interchannel lake, lateral accretion deposit, etc.). The spatial distribution of depositional environments and the development of drainage patterns during each depositional episode is summarised in a series of maps in Fig 3. A total of 10 depositional environments is differentiated on this basin-wide scale. Each is briefly discussed below. Conglomeratic alluvial deposits - Clast and matrix supported well-rounded pebble and cobble conglomerates of basement lithology clasts. Imbrication and crude stratification is occasionally discemible. Occasional interbeds up to Im thick of massive to flat laminated medium to coarse grained sandstone and mudrocks, exhibiting incipient palaeosol formation. Rounding of clasts and the observed stratification pattems suggest deposition in bedload-dominated streams. Thick successions (tens to hundreds of meters) of conglomerate have only been encountered in drill core on the eastern (orogenic) basin margin. We interpret them to form part of large alluvial wedges, which derived material from the orogenic hinterland and prograded into the Bowen Basin. No evidence was found for massflow deposits, such as would be expected to dominate piedmont-style alluvial fans. Meandering rivers - Channel fades: trough cross-bedded and ripple-cross-laminated sandstones, often with scoured bases and erosional lags of pebbles and intra-fomiational rip-up clasts. Overbank facies: laminated and ripple-crosslaminated mudrocks and fme to veryfinegrained sandstones. Palaeosols are indicative of periodic drainage of the alluvial 293

surface. Geometry: moderate sized, ribbon-like channels dissecting extensive vegetated floodplains. Inclined master bedding planes (epsilon cross-beds) can readily be recognised in many exposures Fluvial deposits -low sandstone/mudrock ratio (from core) -Facies characteristics closely resemble those of meandering rivers, but the lower level of interpretation reflects the limitation of core data Sandy, mobile rivers - Channel fades: trough and planar cross-bedded, individual preserved set sizes range from few dm to reported sizes of 7m. Sets of 1-3 m are common in many parts of the basin. Downstream and lateral accretion occurred. Overbank facies: Preservation potential of overbank deposits is extremely low and often the only manifested as intraformational clasts. Geometry: Sheet-like with minor laterally discontinuous lenses of overbank facies. Fluvial deposits - high sandstone/mudrock ratio (from core) - Facies characteristics closely resemble those of sandy mobile rivers, but the lower level of interpretation reflects the limitation of core data Lacustrine deltaic sediments - Deltaic deposits include components of other facies differentiated here, such as finegrained fluvial associations. Thin coals form a minor component of lacustrine delta deposits in the study area. The main criterion for the identification of deltaic deposits was the presence of prograding facies patterns overlying lacustrine deposits. Indeterminate wetlands - Where exposure is absent or not of suitable quality to differentiate between lacustrine, overbank or deltaic deposits in intervals of interbedded coarse and fine-grained facies these were assigned to be indeterminate wetlands Open circulation lacustrine - Comprises nearshore and offshore laminated facies. Nearshore deposits contain a significant proportion of winnowed and well sorted sandstone interbedded with bioturbated and interlaminated mudrocks. Offshore facies is dominated by fine-grained low energy suspension fall-out deposits of organic-rich siltstones and claystones, locally bioturbated. Occasional influxes of fine sand occur, most likely a consequence of storm activity, and are preserved as isolated laminae. Geometry: laterally extensive and sheet-Uke thin deposits. Taroom-12/12a

Taroom-14

Tiggriggie Creek-1

Baraiaba-1

Taroom-16

Saline Lake - The only indication for the presence of potential high salinity environments comes from palynological constraints from one well (Tiggriggie Creek 1). An interval occurs in this well with a very restricted algal kerogen facies while overlying and underlying successions yielded a rich and well preserved microflora (Heckel 1983). Alluvial peat mires - Limited to Permian sediments only. Compositionally banded coals and intervening sandstones and mudrocks. Geometry: Coals are continuous over tens of kilometres. Major channel deposits display an overall sheet-like geometry and are up to 2.2 km wide 294

Eustacy and base-level fluctuations No direct evidence exists for a connection between the Bowen Basin and the palaeo-Pacific ocean through the actively rising New England Fold Belt. Southward directed drainage systems in the Bowen and Sydney Basins, however, have been interpreted in the past to suggest a connection with the sea in that direction (Jensen (2)). Palaeocurrent and facies patterns in coeval strata of the Sydney Basin suggest marginal marine conditions were developed, at least episodically around the latitude of Sydney (Hamilton & Galloway (3), Herbert (4)). To constrain the elevation of the basin surface and the geometry of the depositional slope at least at an order-of-magnitude level, an estimate of the elevation of the basin floor was conducted, based on an "average river slope estimate". To determine this the courses of 16 rivers world-wide were determined from the 1:2.500 000 world map series (Karta Mira,, Budapest 1969-), plotting distance along the river course (approximate thalweg) against altitude. Most of the rivers included currently deposit sediment over most of the length of their thalweg, or at least do not appear to exhibit net erosion. The resultant average river slope for world rivers is 0.00033. For rivers draining tectonically active areas the slope is 0.00088. With a drainage distance of 1300 km for an axial Bowen to Sydney Basin drainage and these estimated slope ranges, basin floor elevation in the study area is estimated to have been between 434 and 1075 metres. Both figures serve as an order-of-magnitude only, and show that basin elevation was 10^ to 10^ meters above sea-level, assuming southward axial drainage to the sea, and a position of the palaeo-shoreline not far from the current position of Sydney. The determined basin elevations impose significant constraints on a possible connection to the sea through the New England Fold Belt. If such a connection is invoked to have existed contemporaneously with the axial southward drainage it would have had a slope of between 0.002 and 0.004, which is an order of magnitude steeper than average slopes on modem river systems. Any inundation by the sea and or influx of even brackish waters into the basin would require the sea level to have risen several hundred meters. This would result in widespread development of fully marine conditions, at least in the lower part of the system (i.e. Sydney and Gunnedah Basins). Based on the above arguments we do not regard baselevel fluctuations in the Triassic successions of the Bowen Basin to result directly from eustatic variations, but predominantly reflect the interplay of basinal subsidence and sediment supply. This can be expressed as variations in the basins status ranging from underfilled (subsidence > sediment supply) and overfilled (subsidence< sediment supply). The indirect influence of eustacy on sedimentation pattems by variation of climatic controls cannot be excluded and may have played a significant role in parts of the succession. Discussion of depositional episodes Episode I — Lower Rewan

Depositional Environments: On the eastem basin margin widespread accumulation occurred of lithic pebble conglomerates and pebbly sandstones (Cabawin Formation, Mack (5)). On the westem basin margin the lower Rewan Group is characterised by thick successions of lithic and volcanic-lithic medium to very coarse grained sandstones with generally high sandstone to mudrock ratios. In the southwestern subsurface areas the basal part of the lower Rewan Group consists of quartzose to quartz-lithic medium to very coarse grained sandstones. These sediments were most likely derived from nearby exposed cratonic areas to the northwest and west. Mixing with the more typical volcanic lithic sandstones of the lower Rewan Group is evident where that transition is recorded in drill core. Nature of the basal bounding surface: Low-relief erosional disconformity with no apparent angular difference in bed attitudes between the Bandanna Formation and the Rewan Group, both at outcrop and in core. On regional seismic lines a basal Triassic unconformity has been illustrated (Skilbeck & Lennox (6)). Evidence exists in drill core for an unconformity surface underlying basal Rewan Group sediments. Many cores intersect the unconformity, which often is often characterised by basal Rewan Group sediments lying directly on Permian Bandanna Formation coal seams. In areas closer to the depocentre the contact between the Bandanna Formation and the Rewan Group is gradational. Recent advances in the biostratigraphic subdivision of the Triassic of the Bowen Basin (Price pers. comm. 1993), suggests that most commonly the biostratigraphic zones APP6 (uppermost Permian) and APTl (lowermost Triassic) are missing, where the Permo-Triassic contact is unconformable (Fig. 2). Relative base-level variation: In the many areas of the basin sediments of the Rewan Group can be differentiated from the Bandanna Formation only by the absence of coals and the increasing primary reddening of fine grained intervals, and no unconformity is apparent. Jensen (2) interpreted sedimentary environments during early Rewan times to have been a continuation of those of the latest Permian coal measures, but with overall better drained conditions. This implies a significant lowering of base-level on a basinwide scale.

295

15r

^llllll

27*

28*

Basemap

and line of section

Episode 1 - Lower Rewan underfilled basin base-level low sediment influx is transverse (cratonic and orogenic)

in Fig.3

[-5

1: Permian, undiff. 2: Rewan Group 3: Clematis Group 4: Moolayember Formation 5: Basement, undiff 6: Bowen Basin covered by Surat Basin succession Borehole locations indicated by black dots 150-

Episode 2 - Upper Rewan underfilled to overfilled basin base-level low, initially internal drainage sediment influx is axial & transverse (orogenic)

151*

150*

isr

Episode

3(a) - Clematis Glenidal underfilled basin base-level low sediment influx is transverse (cratonic)

Figure 4. Palaeogeographic maps for depositional episodes during the Triassic of the Bowen Basin. Three maps are provided for Clematis Episode, reflecting the higher stratigraphic resolution in that interval. Legend for depositional environments provided overleaf. Basemap shows the distribution of outcrop and subsurface areas and the location of wells utilised dring this study. Also illustrated is the approximate position of line of section for time-space plot (Figure 3).

296

148*

150»

150*

149*

Episode 3(b) - Clematis Expedition Sst

Episode 3(c) - Clematis/Snake Creek

overfilled basin base-level low sediment influx is axial & transverse (cratonic)

underfilled basin base-level highstand, internal drainage most of basin covered by large lake

Episode 4 - Lower Moolayember underfilled to overfilled basin initially internal drainage, rel. high base level major clastic progradation fi-om cratonic and orogenic sides

Episode 5 - Upper Moolayember ?overfiiled basin base-level high major axial drainage, transverse input from both sides

Figure 4 (contd.) Legend for palaeogeographic maps. Fluvial, low sand/mud ratio Fluvial, high sand/mud ratio

sandy, mobile rivers

deltaic sediments fcrgriz:^ ^ ^ ^

indeterminate wetlands

III! —

nil —

= 1111 = nil i = Mil = nil =

lacustrine

Direct indications of structural influence: Onlap of basal Rewan Group sediments may be interpreted from some correlation lines on the westem basin margin. Sediment composition of the basal Rewan Group varies significantly along the westem basin margin along strike (N-S) ranging from quartz conglomerates to volcanic lithic, mudrock dominated 297

successions. This is interpreted to reflect the influence of local source areas and structuring on sedimentation style and sediment composition. On the eastem basin margin the Cabawin Formation comprises volcanic-lithic conglomerates and sandstones, derived from the adjacent New England Foldbelt to the east, indicating the presence of significant relief in that area at the time, most likely resulting from structural upheaval. The stratigraphic position of the Cabawin Formation is not clear from the original definitions as virtually all of the biostratigraphic framework in the Triassic of the Bowen Basin was established after definition of the unit (Mack (5)). By comparison with stratigraphic units elsewhere in the basin and examination of core from Cabawin 1 and Cabawin East 1 during this study, we are confident that the Cabawin Formation corresponds to the lower part of the Rewan Group. Direct evidence of climatic influence: Similar changes from coal measures to reddened fine grained fluvial successions are widespread along the eastem margin of Gondwana and are also reported from the Gunnedah and Sydney basins (Herbert & Helby, (7); Tadros 1993) and the Prince Charles Mountains, Antarctica (Webb & Fielding (8)), among other areas. Interpreted origin of surface: The extent of similar facies variations in continental sediments across the Permo-Triassic boundary along virtually the entire eastem margin of Gondwana indicates the inter-regional nature of the change from coal measures to continental red beds at this time. The basal Rewan unconformity surface and its conformable correlatives are interpreted to result from widespread changes in climatic conditions associated with the Permo-Triassic boundary, which were fiirther accentuated by structural upheaval on the margins of the Bowen Basin. Episode 2 — Upper Rewan Depositional Environments: Upper Rewan Group sediments are volcanic-lithic and exhibit significantly lower sandstone to mudrock ratios than in the Rewan. Fluvial sediments of the upper Rewan Group (Arcadia Foraiation) exhibit well defined channel sandstones encased in thick successions of preserved overbank deposits. Laterally extensive (sheet) sandstones do not occur. Overbank deposits commonly exhibit palaeosols with well developed horizonations. Primary reddening occurs in both the sandstones and mudrocks of the upper Rewan and indicates oxidising conditions, which resulted in very poor preservation potential of organic detritus. No bioturbation or wave-ripple dominated successions were identified, which could serve as evidence for widespread standing water conditions. Palynological analysis in Tiggriggie Creek 1 suggests the presence of hypersaline standing bodies of water near the basin's depocentre (Heckel (9)). Nature of the basal bounding surface: A mid-Rewan unconformity is indicated on the western basin margin by a mapped angular unconformity (up to 15°) in the crestal area of an anticline in Arcadia Valley (Woolley (10)). On the eastem basin margin thickening of Rewan Group strata over reverse faults can be identified from regional seismic (Kassan (11), Elliott (12)), possibly indicating an episode of structural deformation during approximately mid-Rewan times. Relative base-level variation: Unknown. The higher proportion of fine grained intervals preserved may indicate a higher rate of aggradation than in the underlying Episode 1, and possibly reflects elevated base-levels. Direct indications of structural influence: Thickening across faults in seismic from the eastem basin margin. It is suspected, but cannot be confirmed here, that part of the conglomerate succession of the Cabawin Formation (Mack (5)) on the eastem basin margin may be associated with the Mid-Rewan structuring. Direct evidence of climatic influence: Data on this surface are insufficient to determine the role, if any, of climatic variations. Interpreted origin of surface: Based on what little evidence is available on this bounding surface, it most likely reflects an episode of structural deformation, which affected both the eastem and western side of the basin. Episode 5 — Clematis Depositional Environments: Depositional environments during the Qematis depositional episode show a broad subdivision into fluvial systems with low sandstone/mudrock ratio (Glenidal Formation) grading upward into sandy mobile rivers (Expedition Sandstone/Showgrounds Formation), which in turn are replaced over much of the basin by laterally extensive lacustrine flooding (Snake Creek Mudstone). Facies variations within these environments are limited during the Clematis episode, and units exhibit sheet-like character across the basin. Outcrop analyses of the Glenidal Formation indicate a predominance of fluvial environments, with some evidence for minor bodies of standing water present at the time. Fluvial channels were of moderate size and are interbedded with commonly preserved overbank facies. Some floodplain sediments show primary reddening and palaeosol development. No conglomeratic alluvium or significant lacustrine deposits occur within the Glenidal Formation. The Expedition Sandstone is dominated by deposits of sandy mobile rivers, which most likely had a braided planform. Individual channel belts were in excess of 1.5 km wide and tens of meters thick. Regional analysis of presented cross-set 298

thickness indicates a concentration of large cross-sets (>1.5 meters) in the central and eastern parts of the outcrop belt, where set thicknesses of up to 4m were commonly observed. This suggests the presence of large trunk rivers near the basin's depocentre. The upper 10-15 m of the Expedition Sandstone and Showgrounds Formation show a marked change to lower energy conditions and are comprised of interpreted marginal lacustrine facies, grading into the overlying lacustrine Snake Creek Mudstone. The succession is dominated by well sorted medium to fine-grained sandstones interbedded and interlaminated with organic-rich siltstones. Sedimentary structures in this interval include small scale cross bedding, ripple cross-lamination (including wave and combined flow ripples) and bioturbation. The overlying lower part of the Snake Creek Mudstone reflects the predominance of progressively more distal lacustrine environments. Based on the distribution of shaly, low energy deposits of interpreted open lacustrine setting the maximum extent of the Snake Creek Lake is estimated as 54,000 kml This is approximately the size of the modem Lake Victoria (East Africa). On the eastem basin margin the correlative of the Snake Creek Mudstone is formed by a thinner (<10m) succession of mudrocks and interbedded sandstone, which is assigned to a undetermined wetlands setting. The interval may reflect accumulation in distal floodplain settings with occasionally developed small interchannel lakes. Nature of the basal bounding surface: In outcrop areas in the west of the basin the transition from the uppermost Arcadia Formation (Rewan Group) to the Glenidal Formation (Clematis Group) is apparently conformable and transitional over several meters of section. It is defined on the basis of a change to more quartzose sediment in the Glenidal Formation (Jensen (2)). The Glenidal Formation as a whole is transitional from the lithic and volcanic lithic sandstones and fine-member-dominated sediments of the Arcadia Formation, to the quartzose and quartz-lithic, increasingly coarse-member dominated deposits of the Glenidal Formation. In the subsurface southwest of the basin, the Glenidal Formation is absent and Showgrounds Formation rests disconformably on lower Rewan sediments. Totterdell et al. (13) identified a sequence boundary (unconformity with truncation of underlying reflectors) at the base of the Clematis Group (their sequence boundary B48). In core and outcrop we find evidence for a break in sedimentation between the two units limited to basin marginal areas onlapping structural highs, and more typically observe a gradual change to quartzose sandstone lithology over tens of meters of section. Relative base-level variation: The change to progressively higjier connectedness of channel sandstones in the Clematis Group (Glenidal Formation) possibly indicates a lowering of base-level relative to upper Rewan (Arcadia Formation) times. Direct indications of structural influence: The amount of missing section between the lower Rewan Group and the Showgrounds Formation (Expedition Sandstone equivalent) on the Roma Shelf and adjacent areas indicates uplift and erosion in that area (Fig.2). Structuring occurred in the west of the basin at the base Qematis level, mostly through reactivation of pre-existing faults {see Totterdell et al. (13), Elliott (12)), and locally leads to Clematis resting directly on Upper Permian Bandanna Formation. Where upper Clematis Group (Showgrounds or Expedition) rests unconformably on Rewan Group the stratigraphic gap between the two units is substantial (1-2 biozones, equivalent to approximately 3.5 Ma), and basal Clematis relief can be significant. The development of incised valley fills on the northern Roma Shelf is interpreted to be controlled by the distribution of basement lithologies and is persisted from Permian times. The stratigraphic record on the eastem basin margin of the Bowen Basin lacks evidence of increased topographic relief during Clematis times, when no significant conglomerate sections were accumulated in the basin. Direct evidence of climatic influence: Qimatic differences between the Rewan Group and Glenidal Formation have been demonstrated by Jensen (2), based on palaeosol characteristics. Interpreted origin of surface: Widespread erosional thinning of underlying Rewan sediment and pinchout of lower Clematis Group strata can be demonstrated on the (cratonic) western basin margin (Elliott (12)), but such evidence is sparsefi-omthe (orogenic) eastem margin. Similarly sediment inputfromthe New England Foldbelt was greatly reduced during Clematis times and no thick successions of conglomerate are associated with the Rewan/Clematis boundary. The development of a forebulge has been shown in other foreland basins to result in increased influx of detritus from the cratonic side of such basins (e.g. Sinclair et al. (14)). Forebulges develop as a crustal response toflexureby thmst loading, and their position and amplitude depend on theflexuralrigidity of the cmst. The higher the load on the orogenic side (foldthrust belt) is, the deeper the basin-forming inflection, and the more pronounced the positive relief of the forebulge. Maximum development of a forebulge on the cratonic western side of the basin during upper Rewan times would coincide with interpreted high subsidence rates in the depocentre at that time (Kassan (10)). The apparently transitional nature of the Rewan/Clematis boundary in outcrop areas in the west of the basin may result from reworking of uplifted Rewan Group sediments. Rather than documenting stmctural upheaval within the Bowen Basin or on its margins, sediment accumulation during the Clematis Episode is considered here to reflect a period of tectonic quiescence in the area. In particular the absence of coarse clastic input on the (orogenic) eastem side of the basin suggests by comparison with the lower Rewan and lower Moolayember intervals much reduced rejuvenation of topographic relief in the source areas. Reduced uplift in the New 299

England Fold Belt would have resulted in a diminished influx of volcanic lithic detritus into the basinfromthe east, which dominated sedimentation during Rewan times. Episode 4 — Lower Moolayember Depositional Environments: Lacustrine deposits and indeterminate wetlands initially dominated accumulation during this episode and were gradually infilled by prograding clastic wedges in the form of interpreted deltaic and coarse clastic alluvial settings. Lacustrine sediments of the upper Snake Creek Mudstone comprise progressively sand-dominated shoaling successions of interbedded siltstone and fine sandstone. Widespread lacustrine deposits are restricted to the basal part of the lower Moolayember Episode. Deltaic successions overlying the lower lacustrine interval are characterised by interbedded lithic sandstones and organicrich mudrocks. Sandstones are commonly coarsening-up and are laterally extensive at outcrop over several hundred meters. Fluvial processes with abundant evidence of incision and reworking become progressively more important towards the upper part of the unit. Small meandering channels and associated organic rich overbank deposits are exposed in a series of roadcuts on the southwestem outcrop area and form the later stages of Lower Moolayember accumulation in that area. Cobble and pebble conglomerates occur on the eastem basin margin and in thefiillycored stratigraphic borehole Taroom16 some 300 meters of predominantly conglomeratic section occur. Both matrix and clast supported fabrics were identified during this study, with occasional imbrication and generally well rounded clasts. Intervals of mudrock exhibit mottling associated with incipient palaeosol formation. Nature of the basal bounding surface: Maximumfloodingsurface. A regional marker, the Snake Creek Mudstone, was first recognised in the Roma Shelf area (Hogetoom (15)). It comprises organic rich laminated claystones and siltstones and is readily recognised on wireline logs. In drill core and at rare outcrops it can be demonstrated that the top Clematis — base Moolayember surface represents the maximum transgression of a large lake (Snake Creek Lake). Totterdell et al. (1992) identified the basal Moolayember surface as a sequence boundary. We do not find any evidence for incision or truncation at this level, and can document a transitionfromfluvialExpedition Sandstone and Showgrounds Formation to fiilly lacustrine conditions in the Snake Creek Mudstone. Relative baselevel variation: A rise in relative baselevel is indicated by the transgression of the Snake Creek Lake. Direct indications of structural influence: No direct indications of structural upheaval could be identified at this stratigraphic level. The documented rise in baselevel itself, however, is likely to be structurally controlled. Interpreted origin of surface: Lacustrine maximum flooding surface. We interpret the Snake Creek flooding to result from renewed onset of deformation in the New England Fold Belt with associated load-induced subsidence in the basin. This interpretation is consistent with the change of petrography across the flooding event, from quartzose to more lithic and immature compositions. Furthermore the virtual cut-off of quartzose sediment supply from the west and a rearrangement of dispersal directions to an axial dominated system in lower Moolayember times coincide with the SnakeCreek bounding surface. Episode 5 — Upper Moolayember Distribution: The upper Moolayember Formation is present in deeper syncUnal areas of the basin and is assumed to have originally covered the entire basin. Most of the information presented here is based on examination of drill cores. Depositional Environments: Where accessible deposits of the Upper Moolayember Episode are dominated by fluvial accumulations with a low sandstone to mudrock ratio. Sandstones are lithic and dominated by medium scale cross-bedding. Nature of the basal bounding surface: On the eastem margin (GSQ Taroom 16) this bounding surface corresponds to a surface of non-deposition, with potentially some minor erosion. Relative base-level variation: A rise in base-level is consistent with the observations of the apparent dramatic lowering of stream gradients on the eastem margin. Nearer the basin centre the change from prograding facies patterns in the lower Moolayember Formation to the aggradation of the upper Moolayember Formation is equally consistent with a rise in base-level. Direct indications of structural influence: The sudden cut-off of coarse clastic detritus may indicate contemporaneous structuring in areas more proximal to the sediment source area in the New England Fold Belt. Direct evidence of climatic influence: Data on this surface are insufficient to determine the role, if any, of climatic variations. Interpreted origin of surface: An interpretation of controlling mechanisms is hampered by the sparsity of data. Structural controls are considered likely in view of the thick clastic deposits on the eastem (orogenic) margin of the basin. Top-Moolayember Deformation A major deformation affected the entire Bowen Basin and resulted in a series of N-S trending broad open folds in the southem part and imbricated thrust penetrating deep westward into the basin in the northern part. 300

The timing of deformation is constraint by the youngest deformed rocks within the basin (Ladinian upper Moolayember) and the oldest non-deformed rocks in the New England Fold Belt (Camian to Rhaetian Callide Coal Measures). The oldest rocks overlying the Moolayember Formation belong to the Sinemurian Precipice Sandstone (biozone PJl, Price et al. 1985), which form the basal succession of the overlying Surat Basin. Sharp lowering of local base-level was associated with the uplift and resulted in large sections of the Moolayember Formation, and locally also Clematis and Rewan Groups, being eroded. Conclusions • The Bowen Basin acted as a foreland basin during Late Permian to Triassic times underwent the broadly westward encroachment of an orogenic belt (New England Foldbelt). • The Triassic fill of the Bowen Basin is here divided into five basinwide recognisable genetic stratigraphic "packages", each bounded by surfaces of erosion, non-deposition, flooding or their conformable correlatives. • Sediment accumulation in the Triassic of the Bowen Basin was fundamentally controlled by stmctural deformation of the basin margins. • On the eastem basin margin a systematic association of facies and stmctural deformation is interpreted. Major stratigraphic changes are typically overlain by coarse clastic wedges of interpreted alluvial fan origin. A progressive westward migration of the edge of coarse clastic accumulation may follow the trend of the migration of depocentres. • On the westem basin margin the stratigraphic evolution was somewhat different, but also controlled by structural upheaval. Uplift and erosion occurred at least episodically throughout Triassic times and may be linked to the establishment of a forebulge in that area. • Assessment of palaeo-slopes based on modem-day examples has provided additional constraints on the interpretation of some stratigraphic surfaces. References 1.

2. 3. 4. 5. 6. 7. 8.

10. 11. 12. 13.

14.

15.

Harrington, H.J., Brakel, A.T., Hunt, J.W., Wells, A.T., Middleton, M.F., O'Brien, P.E., Hamilton, D.S., Beckett, J., Weber, C.R., Radke, S., Totterdell, J.M., Swaine, D.J. & Schmidt, P.W. 1989, Permian Coals of Eastem Australia, Bureau of Mineral Resources, Australia, Bulletin 231, 412pp. Jensen, A.R.1975, Permo-Triassic Stratigraphy and Sedimentation in the Bowen Basin, Queensland, Bureau of Mineral Resources, Geology and Geophysics, Australia, Bulletin 154 Hamilton, D.S. & Galloway, W.E. 1989, New Exploration Techniques in the Analysis of Diagenetically Complex Sandstones, Sydney Basin, NSW. APEA Journal 29, p.235-257 Herbert, C. 1993, Early to Middle Triassic Fluvial to Marine Transition in the Sydney Basin, 5th International Conference on Fluvial Sedimentology, Brisbane, Australia, 5-9 July 1993, Abstracts, p.49 Mack, J.E. Jr 1963, Reconnaissance Geology of the Surat Basin, Queensland and New South Wales, Bureau of Mineral resources, Australia, Petroleum Search Subsidies Act Publication No. 40 Skilbeck, C.G.&Lennox, M.J.1984, The Seismic Atlas of Australian and New Zealand Sedimentary Basins, University of Sydney, Earth Resource Foundation, Sydney, 301pp. Herbert, C. & Helby, R. (eds) 1980 A Guide to the Sydney Basin, Geological Survey of New South Wales, Bulletin 26 Webb, J A. & Fielding, C.R. 1993 Permo-Triassic Sedimentation in the Lambert Graben, northem Prince Charles Mountains, Antarctica. In: Findley, B.H., Unrug, R., Banks, M.R. & Veevers, J.J. (eds) Gondwana Eight Assembly , Evolution and Dispersal Proceedings of the eighth Gondwana Symposium, Hobart, Tasmania, Australia, p.357-369 Heckel H. 1983, Palynology and Biostratigraphy of Core and Cuttings Samples from COE Tiggriggie Creek 1, Company Report, Open File, held by Queensland Department of Minerals and Energy Price, P.L., Filatoff, J., Williams, A.J., Pickering, S.A. & Wood, G.R.1985, Late Palaeozoic and Mesozoic PalynoStratigraphical Units CSR Oil and Gas Division, Report No. 274/25 Open File, held at Queensland Department of Minerals and Energy Woolley, J.B.1944, Geological Report on Arcadia, Shell (Qld) Development Pty, Geological Report 12 (unpublished) Kassan, J. 1993, Basin Analysis of the Triassic Succession, Bowen Basin, Queensland, PhD Thesis, University of Queensland (unpublished) Elliott, L.G. 1993 Post-Carboniferous Tectonic Evolution of Eastem Australia, APEA Journal, p.215-236 Totterdell, J.M., Wells, A.T., Brakel, A.T. Korsch, R.J. & Nicoll, M.G. 1992, Sequence Stratigraphic Interpretation of Seismic Data in the Taroom Region, Bowen and Surat Basins, Queensland. Bureau of Mineral Resources, Australia, Record 1991/102, 61pp. Sinclaire, H.D., Coakley, B.J., Allen, PA. & Watts, A.B.1993, Simulation of Foreland Basin Stratigraphy using a Diffusion Model of Mountain Belt Uplift and Erosion: An Example from the Central Alps, Switzerland, Tectonics 10/3,p.599-620 Hogetoom, D. 1970, Pine Ridge and Raslie Gas Fields, Geological Survey of Queensland, Record 55

301

ARC EVOLUTION AND ORE DEPOSIT MODELS Stephen E. Kesler Department of Geological Sciences, University of Michigan, Ann Arbor, MI, USA 48109 Summary Models for individual ore deposits as well as larger-scale patterns of metallogenic evolution are in a constant state of change, reflecting new information on both the deposits and their geologic environments. Factors that control patterns of metallogenic evolution include crustal thickness, crustal composition and petrologic evolution. Among existing models for individual deposits, acid-sulfate, alkalic porphyry coppergold, sedex, and El Laco-Olympic Dam-type deposits appear to have changed the most. In addition, evidence is accumulating to suggest that the timing and distribution of many deposits reflect global forcing mechanisms such as ocean anoxia and plate reorganizations. Introduction Most mineral deposit models comprise two parts, one dealing with the geology and geochemistry of the deposit itself and another dealing with the relation between the deposit and its geologic environment (Thompson, 1993). As our understanding of geologic features and environments for individual deposits advances, deposit models must change. An added complication for those of us who work on Mesozoic and older systems resultsfromthe fact that we must try to see through the veil of deformation, metamorphism and other processes that obscures original rock relations. Deposit types for which recent changes have been greatest include acid-sulfate precious metal, alkalic porphyry copper, sedex, and massive iron oxide, as discussed below. Deposit Models Acid-Sulfate Models Although models for acid-sulfate mineralization in subaerial volcanic environments have been discussed widely in recent years (Hayba et al., 1985; Arribas, 1995), the Pueblo Viejo deposit remains difi&cult to classify. The significance of this problem is underscored by the fact that Pueblo Viejo has produced more gold than all other acid-sulfate deposits combined. Despite the fact that Pueblo Viejo is the largest of the acid-sulfate deposits, it appears to stand alone in its class. Whereas most other acid-sulfate deposits are associated with subaerial calc-alkaUne rocks, Pueblo Viejo is found in submarine volcanic rocks with considerably lower K20 contents. Pueblo Viejo is hosted by a maar-diatreme complex in the upper part of the Los Ranchos Formation, a volcanic pile consisting of roughly equal amounts of basalt and dacite with almost no andesite (Figure 1). The Los Ranchos and related formations comprise the oldest unit in the Greater Antilles and they have undergone extensive alteration by seawater to produce spilites and keratophyes, for which the original rock composition must be inferred from altered phenocrysts and silica contents (Donnelly, 1966; Kesler et al., 1981). Donnelly and Rogers (1980) suggested that Los Ranchos and related rocks differ from calc-alkaline and tholeiitic rock series, and named them the "primitive island arc (PL\ ) series". Lebron and ges, especially ones that have undergone extensive seawater alteration. Viewed in its petrologic-tectonic context, the association of Pueblo Viejo with PIA-type bimodal volcanism appears to be more significant than its association with seawater alteration. This is particularly true in view of the demonstration from stable isotope relations that seawater was kept outside the Pueblo Viejo hydrothermal system by a wall of gypsumsealed fractures (Vennemann et al., 1993). Accordingly, Pueblo Viejo-type acid-sulfate deposits will probably be found in bimodal volcanic sequences ranging in age from early to late-arc. Rocks with strong similarities to the PIA series volcanic rocks have been described from the Himalayas (Dietrich et al., 1983), but are rare in the Phanerozoic record. In contrast, the ?IA series is similar to many Archean greenstone terranes, where volcanism is thought to have begun on thickened plateau-like crust. In younger terranes, however, it might be related to deposits such as Chinkuashih, which is associated with very young dacite that lacks related mafic rocks (Alkaline (Shoshonitic) Porphyry Copper Deposits Although porphyry copper deposits are most commonly associated with calc-alkalic rocks, important deposits are also associated with alkalic, and even shoshonitic, rocks (Muller and Groves, 1995). These deposits are of exploration interest because of their relatively high gold grades. The curious question is why they are not more common. This is particularly true in view of the apparent tendency toward copper enrichment in alkalic and shoshonitic igneous systems. Copper concentrations in Nicola-group volcanic rocks, which host many of the Canadian alkalic porphyry copper deposits, are considerably above average for basalt, and late-stage magmas from alkaline intrusions associated with porphyry copper mineralization in the Nicola group have even higher copper values (Figure 2). Alkalic and shoshonitic volcanic rocks are found throughout westem North America (Mortimer, 1986), as well as in Puerto Rico, Fiji, New Guinea, Italy, and Mexico, and some of these show evidence for enrichment of copper. Alkaline miagmas might even play a role in the generation of porphyry copper deposits in calc-alkaline environments. Keith et al. (1995) have reported that shoshonite and related alkaline volcanic rocks associated with the intrusive rocks 302

that formed Bingham are unusually enriched in copper and have suggested that this might be the source of copper found in the ore deposits. The fact that Bingham also occupies a far back-arc tectonic setting that would place it over the deeper part of any genetically related subduction zone lends further support to the possible role of alkaline magma in formation of these deposits. Finally, and perhaps most importantly, although intrusive rocks directly associated with the Bingham porphyry copper deposits are not subsilicic, they are unusually feldspar-rich and quartz-poor (Moore, 1973). A final aspect of alkaUne porphyry copper deposits that merits consideration is their relation to epithermal gold mineralization associated with alkahc igneous rocks, including such important deposits are Cripple Creek, Porgera, and Emperor (Mutschler et al., 1991; Muller and Groves, 1995). Most models for alkaline epithermal gold deposits involve an underlying (hypothetical) alkaline porphyry system (Richards, 1995). Simple metal abundances suggest that separation of copper and gold must take place during formation of Ikaline porphyry copper deposits. The average Cu:Au (weight) ratio in fresh intrusive rocks associated with ore at Aflon-Ajax and Copper Mountain is about 1:5,000, whereas the ore bodies have a ratio of about 1:30,000 or more (Stanley et al., 1995; Ross et al., 1995). Thus, if gold and copper separate from the magma in approximately equal proportions, large amounts of gold must escape from porphyry copper deposits into lower temperature hydrothermal systems. Basin-Related Sedex and MVT Deposit Models In most Mesozoic settings, basin-related mineralization is closely related to rifling, and the evolution of this rifling determined the nature and distribution of mineralization. Critical factors in the history of rifling include the rate and periodicity of clastic input, the extent of carbonate development along stable shelves, and possibly the degree and timing of basinal compression. The Cayetano basin of western Cuba, which hosts numerous sedex deposits, underwent continuous subsidence until complete continental break-up took place (Feoktistov et al., 1983). During formation of the Caribbean arcs this basin was broken up and part of it was carried along as the crustal fragment that currently underhes westem Cuba. MVT mineralization did not form during break up of the basin, probably because of the absence of suitable carbonate host rocks, a lack of basinal brines due to the absence of shelf carbonate environments for evaporative concentration of seawater, and the absence of a compressive or uplifl phase to dewater the basin in its later stages. In contrast, the BasqueCantabrian basin of northem Spain underwent episodic rifling and subsidence followed by compression that eventually formed the Pyrenees (Srivastava et al, 1990). Sedimentation was controlled largely by fault blocks formed by rifling and flow of underlying Triassic evaporites, and includes several sequences of clastic similar sulfur isotope compositions suggest that all of these deposits formed as part of a basinwide[AUl] hydrothemaal system that was much larger than that seen in the Cayetano basin. The massive expulsion of brines from this basin was probably caused either loading of thick flysch deposits late in the life of the basin or by compression related to collision of Africa and Europe. £1 Laco-Olympic Dam(?)-type Iron Oxide Deposits El Laco and other massive iron oxide ores, along with possibly related copper, uranium, fluorine, gold and rare-earth element deposits such as Olympic Dam (Oreskes and Einaudi, 1988) comprise a poorly understood class of deposits with considerable exploration potential (Figure 3). These deposits are associated with silicic magmatism ranging in age from Proterozoic to Pleistocene and debate centers on whether they formed by crystallization of immiscible iron-oxide magmas or precipitation from iron-bearing hydrothemial fluids. Recognition of magnetite mineral inclusions or aqueous fluid inclusions in ore minerals would help resolve this controversy, but inclusion work on these deposits has been impeded by a lack of transparent minerals. Most progress on inclusion work has been made at Vergenoeg, a cylindrical body of magnetite, fayalite and fluorite in the Rooiberg Felsites, the siUcic volcanic pile that fomaed immediately prior to intrusion of the Bushveld Igneous Complex (Crocker, 1985). Fluorite at Vergenoeg contains abundant aqueous fluid inclusions and few, if any, possible magmatic inclusions, thus supporting an important role for hydrothermal activity, at least in the latter stage of the life of the system (Borrok et al., 1996). The chemistry of these solutions, as indicated by the preliminary fluid inclusion data, suggests that they would be able to carry gold only as a chloride complex and thus, that they would be effective mineralizing agents only at relatively high temperatures and salinities. Deposits of this type are found largely in continental collisional zones, where they are closely associated with silicic ignimbrites (McDowell and Clabaugh, 1979). Most magmas related to these deposits are highly oxidized, a feature that has been shown to enhance formation of immiscible iron-rich magmas (Naslund, 1983). Formation of oxidized melts in collisional margins is poorly understood, but one possible mechanism would be incorporation of meteoric water, as shown experimentally by Baker and Rutherford (1996). Some of these deposits, such as La Perla and Cerro Mercado in Mexico, are in calderas that fed ignimbrite terranes (Swanson et al., 1978), a setting that would probably facilitate introduction of meteoric water. Global Forcing Mechanisms It is beginning to look like many ore deposits are related to larger-scale, global forcing mechanism such as ocean anoxia or plate reorganizations. The pivotal role of anoxic bottom waters in the formation of sedex deposits has been stressed 303

by Goodfellow (1987) and Turner (1992). Relations in the Cayateno basin in Cuba, a major Mesozoic-age sedex basin that is discussed below, provide support for this generalization, in that most mineralization is related to a widespread black shale thought to represent anoxic conditions (Maynard, 1995). These anoxic conditions may well have been related to global events. Rampino and Caldeira (1993) have shown that global anoxic events occurred at 208 and 193 Ma, which coincide approximately with ages of ore-hosting rocks in the Cayetano basin. Major plate reorganizations may also have controlled the formation of mineral deposits. Solomon (1990) has proposed the porphyry copper-gold deposits in the southwest Pacific formed during reversals in subduction direction. Cox (1973) suggested that the porphyry copper deposits in Puerto Rico formed when north-directed subduction beneath the Greater Antilles gave way to west-directed subduction beneath the Lesser Antilles. Magmatism related to the Mexican Pb-Zn-Ag deposits formed at the same time. Similarly, formation of the widespread Mexican ignimbrite terrane and probably many of the Au-Ag epithennal vein deposits coincided with a later global plate reorganization. References Cited Arribas, A., 1995, Characteristics of high-sulfidation epithermal deposits and their relation to magmatic fluid: Mineralogical Association of Canada Short Course Volume 23, p. 419-454. Baker, L.L. and Rutherford, M.J., 1996, The effect of dissolved water on the oxidation state of silicic melts: Geochimica Cosmochimica Acta, v. 60, p. 2179-2188. Borrok, D.M., Kesler, S.E., Boer, R.H., and Crocker, IT., Fluid Inclusion Chemistry of the Vergenoeg Massive Iron Oxide/Fluorite Deposit, Bushveld Layered Igneous Complex, South Africa: 1995 PACROFI Meeting, University of Wisconsin, Madison, WL, p. 21. Crocker, LT., 1985, Volcanogenic fluorite-hematite deposits and associated pyroclastic rock suite at Vergenoeg, Bushveld Complex: Economic Geology, v. 80, p. 1181-1200. Cox, D.P., 1973, Porphyry copper deposits in Puerto Rico and their relation to arc-trench tectonics: U.S. Geological Survey Open File Report 73-51, 9 p. Donnelly, T.W., 1966, Tectonic significance of the spilite-keratophyre association: Transactions of Third Caribbean Geological Conference, Kingston, Jamaica, p. 13-26. Donnelly, T.W. and Rogers, J.J.W., 1980, Igneous series in island arcs: The northeastem Caribbean compared with worldwide island-arc assemblages: Bulletin Volcanologique, v. 43, p. 347-382. Feoktistov, V.R, Aniyatov, LA. and Norman, A., 1983, Metallogeny of westem Cuba: International Geology Review, v. 25, p. 309-318. Goodfellow, W.D., 1987, Anoxic stratified oceans as a source of sulphur in sediment-hosted stratiform zinc-lead deposits (Selwyn Basin, Yukon, Canada): Chemical Geology, v. 65, p.359-382. Hayba, D.O., Bethke, P.M., Heald, R and Foley, N., 1985, Geologic, mineralogical and geochemical characteristics of volcanic-hosted epithermal precious metal deposits: Society of Economic Geologists Reviews in Economic Geology, v. 2, p. 129-168. Johnson, R.W., MacKenzie, D.E., and Smith, LE.M., 1978, Delayed partial melting of subduction-modified mantle in Papua-New Guinea: Tectonophysics, v.46, p. 197-216. Keith, J.D., Christiansen, E.H., Best, M.G., Barr, D.I., Moore, D.K., Waite, K., Tingey, D.G., Whitney, JA., Kim,. C-S., 1995, The role of mafic alkaline magmatism in formation of porphyry and vein-type mineraUzation: examples from the Bingham and Tintic mining districts, Utah: Geological Association of Canada Abstracts Volume, v. 20, p.A-52. Kesler, S.E., Russell, N., Seaward, M., Rivera, J.A., McCurdy, K., Cumming, G.L. and Sutter, J.F., 1981, Geology and geochemistry of sulfide mineralization underlying the Pueblo Viejo gold-silver oxide deposit, Dominican Republic: Economic Geology, v. 76, p. 1096-1117. Lebron, M.C. and Perfit, M.R., 1994, Petrochemistry and tectonic significance of Cretaceous island-arc rocks, Cordillera Oriental, Dominican Republic: Tectonophysics, v. 229, p. 69-100. Maynard, J.B. and Morton, J., 1995, Lead-zinc-barite deposits in Jurassic rocks of westem Cuba: Indicators of a cratonic rift tectonic setting: Geological Society of America Abstracts with Programs, v. 29, p. A-239. McDowell, F.W. and Clabaugh, S.E., 1979, Ignimbrites of the Sierra Madre Occidental and their relation to the tectonic history of westem Mexico: Geological Society of America Special Paper 180, p. 113-124. Moore, W.J., 1973, Igneous rocks in the Bingham mining district, Utah: U.S. Geological Survey Professional Paper 629B,42p. Mortimer, N., 1986, Late Triassic, arc-related, potassic igneous rocks in the North American Cordillera: Geology, v. 14, p. 1035-1038. Muller, D. and Groves, D.L, 1995, Potassic igneous rocks and associated gold-copper mineralization: Lecture notes in Earth Sciences, v. 56, Springer-Verlag, 210 p. Mutschler, EE., Mooney, T.C., and Johnson, D.C., 1991, Precious metal deposits related to alkaline igneous rocks - a space-time trip through the North American Cordillera: Mining Engineering, v. 43, p. 304-309. Naslund, H.R., 1983, The effect of oxygen fugacity on Uquid immiscibility in iron-bearing silicate melts: American Joumal of Science, v. 283, p. 1034-1059. 304

Oreskes, N. and Einaudi, M.T., 1988, Origin of rare earth element-enriched hematite breccias at the Olympic Dam CuU-Au-Ag deposit, Roxby Downs, South Australia: Economic Geology, v. 85, p. 1-28. Rampino, M.R. and Caldiera, K., 1993, Major episodes of geologic change: correlations, time stmcture and possible causes: Earth and Planetary Science Letters, v. 114, p. 215-228. Richards, J.R, 1995, Alkalic-type epithermal gold deposits - a review in Thompson, J.EH., Magmas, fluids and ore deposits: Mineral Association of Canada Short Course Volume 23, p. 366-400. Ross, K.V., Godwin, C.I., Bond, L. and Dawson, K.M., 1995, Geology, alteration, and mineralization of the Ajax East and Ajax West copper-gold alkalic porphyry deposits, southem Iron Maskbatholith, Kamloops, British Columbia in Schroeter, T.G., ed.. Porphyry deposits of the Northwestem Cordillera of North America, CIM Special Volume 46, p.565-580. Sloman, L.E., 1989, Triassic shoshonite from the Dolomites, northem Italy: alkaline arc rocks in a strike-slip setting: Journal of Geophysical Research, v. 94, p. 4655-4666. Smith, A.D., Brandon, A.D., and Lambert, R. StJ., 1995, Nd-Sr isotope systematics of Nicola Group volcanic rocks, Quesnel terrane: Canadian Journal of Earth Science, v. 32, p. 437-446. Solomon, M., 1990, Subduction, arc reversal, and the origin of porphyry copper-gold deposits in island arcs: Geology, v. 18, p. 630-633. Srivastava, S.P., Roest, W.R., Kovacs. L.C., Oakey, G., Levesque, S., Verhoef, J. and Macnab, R., 1990, Motion of Iberia since the Late Jurassic: results from detailed aeromagnetic measurements in the Newfoundland Basin: Tectonophysics, v. 184, p. 229-260. Stanley, C.R., Holbek, P.M., Huyck, H.L.O., Lang, J.R., Preto, V.A.G., Blower, S.J. and Bottaro, J.C., 1995, Geology of the Copper Mountain alkalic porphyry copper-gold deposits, Princeton, British Columbia in Schroeter, T.G., ed.. Porphyry deposits of the Northwestem Cordillera of North America, CIM Special Volume 46, p. 537-564. Swanson, E.R., Keiser, R.P., Lyons, J.I. and Clabaugh, S.E., 1978, Tertiary volcanism and caldera development near Durango City, Sierra Madre Occidental, Mexico: Geological Society of America Bulletin, v. 89, p. 1000-1012. Tan, L-R, 1991, The Chinkuashih gold-copper deposits, Taiwan: SEG Newsletter, no. 7, p. Iff. Tumer, R.J.W., 1992, Formation of Phanerozoic stratifomi sediment-hosted zinc-lead deposits: evidence for the critical role of ocean anoxia: Chemical Geology, v. 99, p. 165-188. Vennemann, T.W., Muntean, J.L., Kesler, S.E., O'Neil, J.R., Valley, J.W. and Russell, N., 1993, Stable isotope evidence for magmatic fluids in the Pueblo Viejo epithermal acid-sulfate Au-Ag deposit, Dominican Republic: Economic Geology, v. 88, p. 55-71.

305

Crustal Thickness vs. Volcanic Rock Composition (D

sz CC +

<D 60 B Q+ o 'co <D •o <c c <D ©

20 —

O Mid-Ocean Ridge 20

(Leeman, 1983)

Crustal Thickness (km) Figure 1. Schematic cross-section through a island arc showing the relation between Pueblo Viejo-type and calcalkaline-type acid sulfate deposits. Metals vs. Crustal Thickness-Middle America

Figure 2. Cu vs. Si02 plot for shoshonitic volcanic rocks showing that some rock suites are enriched in copper. Although not shown here, copper usually correlates with potassium in these rocks (data from Jolly, 1971; Bloomer et al., 1989; Sloman, 1989; Smith et al., 1995)

306

Lead Deposits Cenozoic Silicic Volcanic Rocks

vein-Type

—

u ' l f e Manto-Type Parral

Western Limit of Sierra Madre Terrane

Sierra Madre Oriental

Figure 3. Schematic models for El Laco type deposits, showing location of both massive iron oxide and possible related hydrothermal deposits.

307

MESOZOIC DEFORMATIONAL EVENTS IN EASTERN AUSTRALIA AND THEIR IMPACT ON ONSHORE SEDIMENTARY BASINS R J. Korsch and J.M. Totterdell Marine, Petroleum & Sedimentary Resources Division, Australian Geological Survey Organisation, GPO Box 378, Canberra, ACT 2601 Summary Eastern Australia in the Permian to Middle Triassic was part of a convergent plate margin system that was related to the coalescence of continental fragments to form Gondwana. In the Cretaceous it formed part of a major continental extensional system related to the fragmentation and dispersal of Gondwana. Within this jframework, several major interplate and intraplate tectonic events occurred, the most significant being extension in the Early Permian and Early Cretaceous, and contraction in the mid-Permian, Late Permian, Early Triassic, Middle-Late Triassic and early Late Cretaceous. These events had a profound impact on the Bowen, Gunnedah and Surat sedimentary basins that were forming adjacent to the New England Orogen. Introduction During the Mesozoic in eastem Australia, several major interplate and intraplate tectonic events converted the New England Orogen (Fig. 1) from an active subduction-related plate margin to a cratonic regime. These events had a profound impact on the sedimentary basins that were forming adjacent to the orogen, and some of the events have important implications for the generation and trapping of hydrocarbons in the basin system. The present day tectonic pattern of eastem Australia originated in the Devonian, although there are indications that eastem Austraha was an active plate margin as far back as the Cambrian. From the Early Devonian, a convergent plate margin related to a west-dipping subduction system produced (from west to east) a magmatic arc, a forearc basin and an accretionary wedge. This pattem continued until the Late Carboniferous when subduction ceased at its former position and jumped east, setting up the convergent system that dominated the Pennian to Cretaceous tectonics of the former greater eastem Australia. The Permian-Triassic Bowen and Gunnedah basins and the Jurassic-Cretaceous Surat Basin formed inboard of the New England Orogen and thus evolved in a back-arc tectonic setting (Figure 1). The Permian to Cretaceous record of the convergent plate margin is best preserved in former pieces of the eastem Australia plate (e.g. New Zealand, New Caledonia) that have since been rifled off Australia due to sea-floor spreading in the Late Cretaceous and Early Tertiary. Stmctural and sequence stratigraphic mapping of a regional grid of seismic data in the Bowen, Gunnedah and Surat basins was undertaken as part of the National Geoscience Mapping Accord (NGMA) project Sedimentary Basins of Eastem Australia. The seismic mapping has provided the geometry and timing of several deformational events that have occurred in the Mesozoic in eastem Australia (Figures 2, 3), as well as the geometry of several sequence boundaries. These events are mainly contractional in nature and occxirred in the Late Permian, in the Early Triassic, in the Middle Triassic, and in the early Late Cretaceous. The cessation of sedimentation in the Bowen Basin and in the Surat Basin was caused by two of these deformational events. As well, a major extensional event commenced in the Early Cretaceous, and though the main effects were focused on the present offshore region, this event had a considerable impact on the onshore part of the Eastem Australia plate. Here we provide a description of the nature of the Mesozoic deformational events and their impact on the sedimentary basins in the region, and speculate on the causes of the deformational events in terms of their links to the New England Orogen and to continental fragments that were once a part of continental Australia but have since been rifted off the eastem margin. Evolution of the Bowen, Gunnedah and Surat basins The Bowen and Surat basins have been subdivided into several basinal phases (Korsch & Totterdell, 1995), which are also applicable to the Gunnedah Basin (Figures 2, 3; see also Korsch et al., 1993a). In the western part of the Bowen Basin, in the Denison Trough (Figure 1), initial rapid subsidence in the Early Permian was due to mechanical extension and formation of a series of half graben. At the same time in the Taroom Trough, there was the emption of a thick volcanic pile, probably also in an extensional environment (Murray, 1990). Two-dimensional modelling of the Meandarra Gravity Ridge in the Taroom Trough by Krassay et al. (pers. comm.) suggests that the gravity anomaly can be explained by an asymmetric pile of mafic volcanics beneath the Permian sediments. This implies that a large magmatic-dominated extensional system was operating to the east of, and simultaneously with, the much smaller, non-magmatic extensional half graben in the Denison Trough. The extensional events were followed by a period with slower subsidence rates driven by thermal relaxation of the lithosphere. This phase was interrupted in the Late Permian by the onset of rapid subsidence in the Taroom Trough, interpreted as a foreland loading phase due to thrusting and cmstal thickening in the New England Orogen to the east This phase of rapid subsidence is obvious in the tectonic subsidence curve for seismic line T82-L-102 (shotpoint 13 T 308

from the eastern side of the Bowen Basin (Figure 4), in comparison with the very limited amount of subsidence in the petroleum e^q^loration well White 1 on the westem side of the basin (Figure 4). The onset of foreland loading was diachronous across the basin, being earlier in the east than in the west. Also, sediments deposited during the foreland loading phase thin dramatically from east to west (Figure 5). Eventually, in the Middle Triassic, the thrust front in the New England Orogen propagated westwards into the basin, effectively terminating sedimentation. In the Late Triassic, erosion produced a marked peneplain on which the Surat Basin was deposited, when subsidence was again driven by thermal relaxation of the lithosphere. Following the Late Triassic unconformity, three phases of subsidence can be seen in the well White 1 (Figure 4). The first two are 'concave up', which is a typical thermal subsidence pattem. It is uncertain, however, whether the third phase, commencing at or before about 112 Ma, represents a further phase of thermal subsidence or whether it is due to flexure and/or loading as a response to the continental extension phase occurring at this time farther to the east. Thus the complicated histories of the Bowen, Gunnedah and Surat basins involved volcanism, mechanical extension, thermal cooling and thrust-related flexuring of the lithosphere during foreland loading. Deformational events in the Bowen, Gunnedah and Surat basins Late Permian Contraction The last event in the Permian, prior to deposition of Early Triassic fluvial sediments, is best observed in the Gunnedah Basin. Here, a Late Permian contractional episode can be recognised where coarse-grained fluvial-alluvial fan sediments of the Early Triassic Digby Formation (APTl-2) unconformably overlie Permian sediments. On the east-west oriented seismic line 85-Bl-G, this is seen as a strongly angular unconformity that truncates dipping Permian reflectors (Figure 6). The Digby Formation overlies progressively older Permian sediments to the west and eventually onlaps basement. In the well DM Bellata 1, located on this line, the Digby Formation directly overlies Watermark Formation sediments of APP4.2 palynological zone (Figure 1). The youngest rocks below the base-Digby unconformity form the Black Jack Group (e.g. in the coal borehole DM Springfield 1) of APP5 age, thus the deformation took place during the Late Permian. Recent SHRIMP dating of a tuff in the Black Jack Group (J.C. Claoue-Long & S. Edgecombe, pers.comm.) suggests that a significant part of the Late Permian section in the Gunnedah Basin has either been eroded or was never deposited. Uplift and erosion caused by this event appear to be greatest in the northem part of the Gunnedah Basin, south of the Moree High. In contrast, seismic lines from the Queensland portion of the southem Bowen Basin do not show any angular unconformity between the Rewan Group (Digby equivalent) and the Permian succession. Hence, the pre-Digby event may be equivalent to a deformational event seen on some seismic lines on the westem edge of the Taroom Trough. On these lines, most of the westerly-thinning Permian section has been eroded from the hanging wall blocks of northstriking, west-dipping thrust faults. Movement on these faults must pre-date deposition of the Late Permian coal measures in the Bandanna Formation (i.e. late APP5) because these rocks have not been affected by the faulting and overlie the entire structure. It is possible, therefore, that this Late Permian (pre-Digby) event is responsible for the progressive erosion southwards of pre-Bandanna Permian sediments in the southem Bowen Basin in a similar fashion to the progressive erosion northwards seen in the Gunnedah Basin. The absence of Bandanna Formation and Rewan Group in the vicinity of the Queensland-New South Wales border could be due to either onlap and non-deposition or erosion following the Early Triassic contraction, and prior to the deposition of the Showgrounds Sandstone (see below). Early Triassic Contraction A contractional event in the Early Triassic deformed the sediment pile prior to the deposition of the Clematis Group, resulting in a significant regional unconformity. Reactivation of earlier faults resulted in inversion of the sedimentary succession, and new thrust faults also formed. Uplift due to the faulting, and subsequent erosion removed up to a few hundred metres of mainly Early Triassic Rewan Group (see seismic section C83-T-04, in Totterdell et al., 1995, figure 4). Elliott (1993) considered that this contractional event was dominated by thrusting and that it was the most important deformational event within the basin system. We consider, however, that the Middle-Late Triassic contractional event, mentioned below, was a magnitude greater in size and was more important in terms of the tectonic development of the region. Middle-Late Triassic Contraction The Middle-Late Triassic contractional event deformed the youngest unit of the Bowen Basin, the Moolayember Formation, and older sediments prior to deposition of the Surat Basin succession. This event was complex and produced significantly different geometries in different parts of the basin. It was not a simple folding event as proposed by Elliott (1993). In the Denison Trough, the Early Permian extensional faults were again reactivated as thrusts. These faults propagated upwards into the younger part of the Bowen Basin, with the shortening being accommodated by thrusting at depth, and by folding in the upper levels of the pile. This produced fault-propagation folds, with the amount of displacement on the faults decreasing upwards (Korsch & Totterdell, 1995, figure 2). New thrusts and backthrusts also formed. For further description of the structural style see Korsch & Totterdell (1995).

309

The eastern margin of the Bowen Basin has generally been regarded as a faulted margin, defined by the near meridional Hmter-Mooki-Goondiwindi-Moonie-Leichhardt-Burunga fault system (Fig. 1). The northem segment, the Burunga Fault, is a west-dipping backdirust (Korsch & Totterdell, 1995, figure 3) associated with a basement duplex to the east (Totterdell & Korsch, 1992; EUiott, 1993). The Burunga Anticline, above the thrust, is a fault propagation fold, and the Burunga Fault appears to be a new fault that formed during the contractional event and not an earlier extensional fault that was reactivated. At its southern end, a displacement transfer zone separates the Burunga Fault, by en echelon overlap, from an east-dipping thrust, the Leichhardt Fault. Farther south, the Moonie Fault consists predominantly of a low-angle thrust fault with a flat-ramp geometry (Korsch & Totterdell, 1995,figure4), showing a typical fault-bend fold style. The major movement on this thrust occurred after deposition ceased in the Bowen Basin, but prior to commencement of deposition in the Surat Basin, that is, in Middle-Late Triassic time. The succession in the anticline above the thrust fault has been eroded so that the Surat Basin succession sits directly on a basement of Tamworth BeU immediately southeast of the fault, but with a remnant of the Bowen Basin occurring farther to the southeast. Frontal or short cut thrusts with limited displacement commonly occur. To the south, a displacement transfer zone separates the Moonie Fault from the Goondiwindi Fault. Thus, the present eastem margin of the Bowen Basin in Queensland and in northernmost New South Wales is an erosional remnant located to the east of the thrust faults. It is only farther south in New South Wales that the Hunter and Mooki faults mark the present eastem extent of the Sydney and Gunnedah components of the basin system. There is some debate over the importance of the contractional events. Although Elliott (1993) stressed the importance of the Early Triassic event, we consider that uplift in the Middle-Late Triassic was a magnitude greater in size, particularly along the eastem margin of the basin where, in places, there has been uplift and erosion of at least 4 km of the stratigraphic succession. Harrington & Korsch (1985) speculated that the deformation was caused by the arrival, docking and suturing of the Gympie Province (Terrane) to the New England Orogen. At this time, deformation was widespread across AustraUa (e.g. Etheridge & O'Brien, 1994), implying that this event reflects a global change in plate configuration, which included accretion and collision of terranes in Asia, extension prior to sea-floor spreading in the Atlantic, and collision of Wrangellia with North America, as well as a sharp bend in the North American apparent polar wander path (the J1 cusp). Early Cretaceous Extension A continental extensional event that affected the eastem Australian plate commenced in the Early Cretaceous, and was part of the breakup and fi:agmentation of the Gondwana supercontinent. Continental extension continued through to seafloor spreading, which commenced at about 80 Ma with the opening of the Tasman Sea (Weissel & Hayes, 1977). In eastem AustraUa, the main effects of this extension are focused in the offshore component (see Symonds et al., this volume). In the Bowen and Surat basins, the extension is expressed as a series of Cretaceous plutons. Farther east, there is an extensive, extension-related suite of volcanic and plutonic rocks ranging in agefiromabout 130-100 Ma (Stephens et al., 1993). Tectonic subsidence curvesfiromthe Surat Basin (Figure 4) indicate that there is an increase in the rate of subsidence in the Early Cretaceous, and that this had commenced by about 112 Ma. There are several possible reasons for this increase, including renewed thermal subsidence following the extensional event, increased sediment loading in response to uplift associated with the continental extension phase occurring at this time farther to the east, or flexure of the lithosphere associated with the extension. Geohistory modelling of the Bowen and Surat basins in Queensland (Boreham et al., this volume) has indicated that the rapid subsidence rate continued beyond the presently preserved sedimentary record and that about 1000 m of Late Cretaceous (c. 99-95 Ma) sediments have been removed from above the Griman Creek Formation. This has been confirmed by apatitefissiontrack analysis (Raza et al, 1995; this volume). This extra sediment thickness is necessary to provide sufiBcient burial to model the current vitrinite reflectance trends without increasing the heatflow to unrealistic levels. The extra thickness of sediment is important because, for over much of the Bowen Basin, it allows the Permian source rocks to enter the oil window for thefirsttime, leading to the generation of the hydrocarbons now found in many of the traps (Boreham et al., this volume). Early Late Cretaceous Contraction Immediately following deposition of the Early Jurassic-Early Cretaceous Surat Basin, another contractional deformational event reactivated many of the earher stmctures in the Bowen Basin and in the basement. Renewed thrusting on these faults led to the propagation of some of the faults up into the Surat Basin succession (e.g. Korsch & Totterdell, 1995,figure4) but, more commonly, the deformation in the Surat Basin was principally by folding and uplift of the Surat succession above a reactivated thmst fault at depth (Figures 7, 8). This reactivation was important because many of the structures in the Triassic succession of the Bowen Basin and in the Surat succession, that now contain oil or gas, were formed by this event. The reUef on the structures on the western side of the basin is limited, in the order of tens of metres (Figure 7), in comparison to the relief generated above the fault system near the eastem margin, where it is several hundred metres (Figure 8). The lack of deformation in Eocene sediments of the Duaringa Basin (immediately west of Gogango Overfolded Zone on Figure 1) constrains the age of the post-Surat deformation to between post-Albian and preEocene. Apatite fission track dating by Raza et al. (1995) of samplesfiromseveral wells in the eastem Bowen and Surat basins indicates that cooling occurred in the early Late Cretaceous due to erosion of overlying sediments. The erosion 310

was induced by rapid uplift that was associated with this contractional event, which we consider occurred in the interval 95-90 Ma. In eastern Australia, this contractional event is best seen in coastal Queensland and in the offshore part of the Maryborough Basin (east of Gympie Province on Figure 1) where it is strongest in intensity (e.g. Hill, 1994). In New Zealand, the effects of the same event can be recognised by uplift and erosion, which by 90 Ma were widespread, resulting in the start of peneplanation (Korsch & Wellman, 1988). This contraction occurred within an overall continental extensional regime, where it is likely that the thermal bulge prior to sea-floor spreading led to the formation of passive margin mountains and the regional uphft. The deformation associated with this uplift was accommodated on pre-existing basement faults. Conclusions The conversion of the Bowen and Gunnedah basins from an extensional to a foreland basin system in the mid-Permian is related to the development of a major west-directed thrust system in the New England Orogen, which converted the westem part of the orogen into a major foreland thmst belt. This belt developed in a back-arc setting behind the convergent plate margin that had migrated to the east. The orogen is a classic example of a doubly-vergent orogen, with west-dipping subduction in the east and predominantly east-dipping thrusts in the west (Korsch et al., 1993b). The foreland thrust belt was initiated in the mid-Permian by a collisional event at the convergent plate margin farther to the east. The thrust front propagated westwards towards the Bowen and Gunnedah basins, with subsidence starting to be driven by foreland loading in the Late Permian. In the southem part of the Bowen Basin, and in the Gunndah Basin, thrusts propagated to the present basin margin by the Middle Triassic, but in the central part of the Bowen Basin thrusting had propagated well into the basin by this time. The propagation of the thrust front into the basin appears to have ended sedimentation. The foreland thrust belt existed for about 40 m.y., which is typical of foreland thmst belts described from elsewhere in the world. The start of the development of the foreland thrust belt in the mid-Permian was an in^ortant event in eastem Australia. In the New England Orogen it is represented by a major mid-Permian deformational event, which was termed the HunterBowen Orogeny by early workers. In New England, this event is marked by a major angular unconfomiity between intensely deformed early Permian and older rocks and weakly deformed to essentially subhorizontal Late Permian and younger rocks. In New Zealand, the change in tectonic regime is marked by the initiation of subduction, which represents an outboard jump for the system previously located closer to the Australian margin. The contractional deformational events in the Permian and Triassic in the Bowen and Gunnedah basins occurred during the foreland loading phase which itself occurred as a response to thrust loading in the orogen to the east. Deposition was punctuated by short, strong deformational events, the effects of which were felt across the basins system and often elsewhere in eastem Australia and beyond. Because of their widespread effects, the contractional events are hkely to be the result of events originating outside of the current New England Orogen, but located elsewhere along the convergent plate boundary. Eastem Austraha was affected by extension in the Cretaceous during breakup and fragmentation of the Gondwana supercontinent. Continental extension commenced in the Early Cretaceous and culminated in the Late Cretaceous with sea-floor spreading in the Tasman Sea and the rifting of continentalfragmentssuch as New Zealand and New Caledonia from Austraha. In New Zealand, this initially resulted in the termination of subduction and conversion to an extensional system (Bradshaw, 1989). In conclusion, deformational events in the Bowen, Gunnedah and Surat basins firstly record aspects of the coalescence of continentalfragmentsto form Gondwana in the Permian to Middle Triassic, and then thefragmentationand dispersal of Gondwana in the Cretaceous. Acknowledgements We wish to thank other members of the NGMA project from AGSO, GSQ and GSNSW for many useful discussions during the project, and B.R. Goleby and H.LM. Stmckmeyer for reviewing the manuscript. We thank the Cartographic Services Unit, AGSO, for drafting i e figures. Pubhshed with permission of the Executive Director, AGSO. References Bradshaw, J.D., 1989. Cretaceous geotectonic pattems in the New Zealand region. Tectonics, 8, 803-820. Elliott, L.G., 1993. Post-Carboniferous tectonic evolution of eastem Australia. APEA Joumal, 33, 215-236. Etheridge, M.A. & 0'Brien,G.W., 1994. Stmctural and tectonic evolution of the Westem Australian margin basin system. PESA Joumal, 22,45-63. Harrington, H.J. & Korsch, R.J., 1985. Deformation associated with the accretion of the Gympie terrane in eastem Australia. Geological Society of Australia, Abstracts, 14,104-108. Hill, P.J., 1994. Geology and geophysics of the offshore Maryborou^, Capricorn and northern Tasman Basins: results of AGSO Survey 91. Australian Geological Survey Organisation, Record, 1994/1, 71 pp.

311

Korsch, RJ. & Totterdell, J.M., 1995. Structural events and deformational styles in the Bowen Basin. In: Follington, I.W., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995.... 150 years on .... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 27-35. Korsch, R.J. & Wellman, H.W., 1988. The geological evolution of New Zealand and the New Zealand Region. In: Naim, A.E.M., SteWi, EG. & Uyeda, S. (Editors), The Ocean Basins and Margins, Vol. 7B, Plenum, New York, 411-482. Korsch, R.J., Wake-Dyster, K.D. & Johnstone, D.W., 1993a. Deep seismic reflection profiling in the Gunnedah Basin: regional tectonics and basin responses. In: Swarbrick, G.J. & Morton, D.J., (editors). Proceedings NSW Petroleum Symposium, PESA Exploration Society of Australia NSW Branch, 30 pp. Korsch, R.J., Wake-Dyster, K.D. & Johnstone, D.W., 1993b. The Gunnedah Basin-New England Qrogen deep seismic reflection profile: implications for New England tectonics. In: Flood, P.G. & Aitchison, J.A. (Editors), New England Orogen, eastem Austraha, University of New England, Armidale, 85-100. Murray, CG, 1990. Tectonic evolution and metallogenesis of the Bowen Basin. In: Beeston, J.W., (comp.), Bowen Basin Symposium 1990 Proceedings. Geological Society of Australia, Queensland Division, Brisbane, 201-212. Raza, A., Hill, K.C. & Korsch, R.J., 1995. Mid-Cretaceous regional upUft and denudation of the Bowen-Surat Basins, Queensland and its relation to Tasman Sea rifting. In: Supplement to Follington, I.W., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995.... 150 years on .... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 1-8. Stephens, C.J., Schon, R.W. & Ewart, A., 1993. Mesozoic crustal extension in the northem New England orogen: geochemical and isotopic evidence from large scale silicic magmatism. In: Flood, P.G. & Aitchison, J.A. (Editors), New England Orogen, eastem Austraha, University of New England, Armidale, 637-642. Totterdell, J.M. & Korsch, R.J., 1992. Structural configuration and tectonic development of the Bowen and Surat basins in the Taroom region. Australian Bureau of Mineral Resources, Record, 1991/102, 35-52. Totterdell, J.M., Hofi&nann, K.L., Brakel, AT. & Wells, A.T., 1995. Basin phases and sequence stratigraphy of the Bowen Basin, Queensland. In: Follington, I.W., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995.... 150 years on .... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 247-256. Weissel, J.K. & Hayes, D.E., 1977. Evolution of the Tasman Sea reappraised. Earth & Planetary Science Letters, 36, 7784.

312

Jurassic-Cretaceous sediments Permian-Triassic sediments Concealed margin of Penvo-Triassic sediments

Figure 1. Map of eastern Australia showing locations of the Bowen Basin, Surat Basin and New England Orogen. Also shown are the locations of the subsidence curves in Figure 4 (4a = seismic line T82-L-102; 4b=White 1) and the seismic sections shown in Figures 6, 7 and 8, correspondingly labelled here as 6, 7 or 8.

313

Bowen Basin Palynological zones Denison Trough Taroom Trough

Age

Tectonic Basin events phases

Gunnedah Basin

APT4 230 -

O CO tn < APT3

240 APT2 APT1 250 -

Showoroun T. Clematis Gp

APPe^

Bancianna Fml Black Alley Sh

'laralabaCM Burunga Fm

APP5

Peawaddy Fm

Scotia Mbr Banana Fm Flat Top Fm Barfield Fm Oxtrack Fm

Ingelara Fm Freitag Fm

260-

APP4 270 -

Moolayember Fm

S QC

upper Aldebaran Sst i i i i n i i i i i i i m

-V ; ^ Deriah Fm Napperby Fm

— B90—

-880

-870 — -865~ =C>4= -B60 — -855

Black Jack Gp Watermark Fm Porcupine Fm

-850 — -845 — 5 c E ®

lower Aldebaran Sst Cattle Creek Fm APP2 APP1

—840 —

Maules Creek Fm

APP3 280-

Reids Dome beds

Eo. CO.E

—

-C O (O

Buffel Fm Goonbri Fm Camboon Volcanics I I I I I I I I

-830— A -

9 -

-815 —

Boggabri Volcanics

Figure 2. Permian and Triassic correlation chart of the Bowen and Gunnedah basins showing relationship between lithostratigraphy, sequences and main tectonic events. Arrows indicate contractional () and extensional () tectonic events. For a description of the sequences see Totterdell et al. (1995); the sequence boundary names are used on the seismic sections (Figures 6 to 8).

314

Palynoiogical zones

Age

Surat Basin Queensland

I New South Wales

APK7

100-1

olU

Sequences

Sequence Tectonic boundaries events

=C><3=

Section removed

APK6 APK5

Basin phases

Griman Creek Formation Surat Siltstone

APK4

o <

fbS

Coreena Member

^ 5

Doncaster Member Not mapped seismically

APK3 cc

Bungil Formation APK2

Mooga Sandstone

APK1

Orallo Formation

-850 —

Gubberamunda Sst

150APJ6

a> CO <

Westboume

Pilliga

Formation

Sandstone

Springbok Sst

APJ5

Walloon CM

cc APJ4

-S40-

Purlawaugh Formation

-835-

Hutton Sandstone -830-

APJ3

I-

Evergreen Formation APJ2

-820-

Precipice Sandstone 200 -

-810-

APJ1 o

APT5

cn CO <

Eddystone beds

Garrawilla Volcanics

APT4

Figure 3. Triassic to Cretaceous correlation chart of the Surat Basin showing relationship between lithostratigraphy, sequences and main tectonic events. The base of this chart abuts the top of Figure 2. Arrows indicate contractional () tectonic event. The Early Cretaceous extensional event coincides with the volcano symbol.

315

Seismic Line T82-L-102 SP131

Whitel

Scotia Burunga

Buffel to Banana

Baralaba 2 Q.

t O

@ Doncaster to Griman Creek

e—e

1 290

1 190

Age (Ma)

\ Clematis

3 -

Y Moolayember Surat

Tectonic subsidence Decompacted Unconiormity 1

Rewan

® Evergreen to Westboume 3 (D Gubberamunda to Bungil

o—€> Tectonic subsidence Unconformity J

1 90

5 290

L 190

Age (Ma)

,1..90 16/A/319

Figure 4. Tectonic subsidence curves of a synthetic well calculated from seismic line T82-L-102 shotpoint 131 (RHS), showing the rapid subsidence of the Late Pemiian and Eaaly-Middle Triassic due to foreland loading, and of a petroleum exploration well. White 1 (LHS), showing an increase in the tectonic subsidence rate in the youngest preserved sediments of the Surat basin in the Early Cretaceous.

316

Figure 5. Isopach map (in two-way travel time) of the Rewan Group (B70-B80) between 23o30'S and 26oS reflecting subsidence due to foreland loading as a response to thrusting and mountain building in the New England Qrogen. Contours are in milliseconds. WEST SP1890 Onr

1 km

85-B1-G

EAST Bellata 1

SP2610

0.2H

I r ' ^

0.4-

0.6H

0.8

S30 B85 , B80 B70 845 830 815

1.0-

Figure 6. Seismic line 85-Bl-G showing the unconformity below Digby Conglomerate (B70) that has removed the sediments at the top of the Permian and truncated older units. This reflects a defomiational event just below the Permian-Triassic boundary. 317

83-Y19

Figure 7. Seismic line 83-Y19 showing basement faults that were reactivated during the post-Early Cretaceous deformational/contractional event. This event was important for the formation of many of the petroleum traps in Jurassic sediments of the Surat Basin. Ballymena 1

WEST

^mm

MOONIE FAULT

P81-112 GOONDIWINDI FAULT

1 km

EAST - 0

S40

S30 S10 B85 B70

B30

iiiii^

i-2

Figure 8. Seismic line P81-112 showing the geometry of the Surat Basin (SiO to top of section) across the relay zone between the Moonie and Goondiwindi faults to the south of Moonie Oil Field. This line shows significant reactivation of the thrusts, particularly the Goondiwindi Fault, in post-Early Cretaceous with about 455 ms (approximately 720 m) of uplift of the Surat succession east of the Goondiwindi Fault relative to the same succession to the west of the Moonie Fault. Although the faults were reactivated at depth, here they appear not to have propagated upwards into the Surat succession, that is, there appears to be no displacement of the Surat succession across the faults.

318

THERMAL EVENTS IN THE SYDNEY-BOWEN BASIN AS DEFINED BY PALAEOMAGNETISM MA. Lackie^ and P.W. Schmidt^ 1 School of Earth Sciences, Macquarie University, Sydney NSW 2109 2 CSIRO, Division of Exploration and Mining, P.O. Box 136, North Ryde NSW 2113 Summary Palaeomagnetic studies of sedimenta27 and igneous rocks in the Sydney Basin identify pronounced magnetic overprinting. Secondary magnetisations are observed in sediments of the Narrabeen Group and in Permian and Early-Mid Jurassic igneous rocks, such as the Gerringong Volcanics, the Milton Monzonite and the breccia diatremes. The magnetisations are entirely of normal polarity, suggesting they were acquired during the Cretaceous Normal Superchron (118 Ma - 83 Ma). The intensity of overprinting decreases away from the coast and is consistent with a thermal event, uplift, erosion and cooling of the basin, probably related to events leading up to initial rifting in the Tasman Sea. Rocks now at the surface have been heated to about 200°C. Fluid inclusion data and K/Ar dating of illites are also consistent with a thermal event having occurred along the sea-board at about 90 Ma. Deflection of the Cretaceous overprint directions by the Lapstone Monocline indicate that this structure post-dates the mid-Cretaceous. Palaeomagnetic studies of coal measures in the Bowen Basin identify a consistent NRM overprint direction throu^out the basin. Although the directions are similar throu^out the basin, the inclination is steeper in the south (Moura region) than in the north (Goonyella Mine) of the basin. The overprint directions in the south of the basin are similar to those observed in the Sydney Basin, and a similar age is inferred (--90 Ma), whereas the direction in the northem part of the basin indicates a younger age of overprinting. Remanence directions found in Cretaceous intrusions and associated homfels within the Bowen Basin show both reversed and normal polarities, in accordance with the ages of the intrusions (140 Ma -120 Ma), being sUghtly older than the Cretaceous Normal Superchron. These directions are similar to those observed in the south of the basin but differ from those observed in the north, although the intrusions outcrop in the north of the basin. SYDNEY BASIN During palaeomagnetic investigations of the Sydney Basin (Fig. 1) in the 80s, Embleton and McDonnell (1980), Schmidt and Embleton (1981) and Schmidt (1982) identified a pronounced overprint magnetisation present in various rock units throughout the basin. This led to the hypothesis of a thermal event, uplift, erosion and cooling of the basin 70-100 Ma ago, probably related to events leading up to initial rifting in the Tasman Sea. Rocks now at the surface have been heated to about 200°C. The cooling is required to explain the blocking, or locking-in, of the overprint magnetisations. Vitrinite reflectance data (Middleton and Schmidt, 1982) corroborate these events. Figure 2 is taken from Schmidt and Embleton (1981) and shows the NRM directions and the directions of the constituent magnetic components for the Homsby Breccia (25 km northwest of Sydney) and the Milton Monzonite (175 km south-southwest of Sydney). The low temperature (<450°C) components from both rock units correspond to an overprint acquired over a long time i.e. on a geological timescale (-10 Ma), albeit at a lower temperature (~200°C). The high temperature components correspond to the TRM or T-CRM acquired when these rock units formed. The distinctive north steep up overprint direction (Table 1) has been measured in Narrabeen Group sediments and breccia diatremes at many locations throughout the Sydney Basin (Fig. 1). The palaeomagnetic poles for the Milton Monzonite, Homsby Breccia and Hawkesbury sandstone overprints all fall on the Late Cretaceous section of the Australian apparent polar wander path (APWP) (Fig. 5).

319

5 10 15 20Km ~ SC^E

HAWKESBURY LOOKOUT

Figure 1. Localities in the Sydney Basin mentioned in the text and tables (after Schmidt et al, 1995). Lapstone Monocline Remanent magnetisations of tilted strata, such as the Lapstone Monocline, have been examined to determine the relative timing of the tilting and the magnetisation (Schmidt et al 1995). The amplitude of the Lapstone Monocline decreases to the south where it merges with the Nepean Fault (Fig. 1). The Nepean Fault generally strikes NNW. Schmidt et al (1995) studied samples from the Tahmoor CoUiery and from recent roadworks at both Lapstone and the Hawkesbury Lookout, where major roads cross the Lapstone Monocline (Fig. 1). The samples were all from grey shales near the top of the Hawkesbury Sandstone.

320

r-1

Table 1. Mean directions of Sydney Basin overprint magnetisations Locality N DecO Incn 0950 Norah Hd 16 17.0 -80.3 3.8 Toowoon 6.4 36 -74.8 1.8 Forester Nth 36 17.9 -77.5 2.6 Otford 16 0.8 -79.1 5.8 Coalcliff 26 355.2 -74.4 3.7 Tahmoor 131 -76.4 6.0 1.6 Homsby 70 5.5 -77.5 2.1 Dundas 13 349.7 -75.7 8.8 Marsden 36 43.5 -78.0 1.9 24 St Marys 358.1 -81.8 2.6 4* Milton 347.6 -79.0 5.9 Note: N, number of samples, Dec, declination, Inc, inclination, a95 , 95% confidence radius (Fisher, 1953), (Table after Schmidt etal 1995). *, Site mean direction taken from Schmidt and Embleton (1981). Presentfieldinchnation,-65°,^ -53°. HORNSBY BRECCIA

Figure 2. Directions of natural remanent magnetisation (NRM) and low temperature and high temperature components from the Homsby Breccia and Milton Monzonite (after Schmidt and Embleton, 1981). Closed (open) symbols represent lower (upper) hemisphere. Directionsfi-omthese samples are listed in Table 2. It is clear from these data that the directions of remanence from sediments at Lapstone and Hawkesbury show a deflection which is consistent with the sense of dip at each locality. Thus, after correction for bedding the Late Cretaceous north steep up direction is observed. The monocline has affected the directions of the overprint magnetisation by a considerable amount.

321

Table 2. Mean directions from monoclines 0950 N DipD Dip AzC) Dh(») Iho a95r) DbO I* wsw 286.3 -66.6 3.9 68.4 -74.1 4.1 41 40 352.4 -80.3 4.3 4.3 63.3 -56.5 90 32 2 37 5.5 5.5 334.9 -60.7 50.4 ^6.0 90 50 3 26 5.7 16.9 -73.8 7.8 37.1 -67.2 104 4 Note. 1,600 Panel Tahmoor Colliery (*variable dips, maximum dip given), 2 - Lapstone, 3 - Hawkesbury Lookout, 4 - All data. N, number of samples, D, declination, I, inclination, h, with respect to present horizontal, b, corrected for bedding, a95, 95% confidence radius (after Schmidt et al 1995). BOWEN BASIN Following the studies of the overprint magnetisation in the Sydney Basin a study was commenced in 1992 (Lackie & Schmidt 1992a; Lackie 1993; Lackie and Schmidt 1993b;) to analyse the remanence of the coal measures and Cretaceous intrusions of the Bowen Basin. Oriented samples were collectedfromfivecollieries in the Bowen Basin (Fig. 3). :)ollinsville

GOONY • SARAJp •GERMAhlvCREEK V Emerald

148

•BLACKWAT

_24

• Mt)URA

Figure 3. Sketch map of the Bowen Basin showing the collieries at which san^les were taken. Samples were taken from the Moranbah, Rangal and Baralaba Coal Measures, as well as the German Creek Formation. The NRM directions of all specimens, excluding coarse-grained sandstones,fromthefivecollieries are plotted in Fig. 4. The dominant NRM direction is northerly and up (Table 3) with the magnetic inclination steeper in the south of the basin (Moura region) than in the north of basin (Goonyella region). When AF demagnetised, samples displayed a magnetically hard (MDFy2>100 Oe) single component direction (Lackie and Schmidt 1992b). The overprint directions in the south of the basin are similar to those observed in the Sydney Basin, with the palaeomagnetic pole for the Moura samples falling on the Late Cretaceous section of the APWP with the Sydney Basin samples (Fig. 5); a similar age is inferred (-90 Ma). The direction in the northem part of the basin indicates a younger age of overprinting, with palaeomagnetic poles (GC, GY, SJ, BL) lying on the Tertiary section of the APWP ^ig. 5) suggesting a Tertiary age of magnetisation.

322

Moura

Blackwater

Figure 4. NRM directions from five collieries in the Bowen Basin, (a), Moura. (b), Blackwater. (c), German Creek, (d), Saraji. (e), Goonyella. Closed (open) symbols represent lower (upper) hemisphere.

323

Table 3. Mean NRM directions from Bowen Basin mine sites Mine

DecO

IncO

0950

Goonyella Saraji German Creek Blackwater Moura

73 92 28 84 78

2.2 8.6 2.0 9.4 1.1

-58.4 -59.2 -55.4 -62.3 -67.5

2.6 1.8 2.7 3.2 2.7

N, number of specimens; Dec, Inc, declination, inclination of the mean NRM direction; a95 - radius of 95% confidence cone (Fisher, 1953). Present field inclination -53°, dipole, -40°. (Table after Schmidt and Lackie 1995). As well as sampling the coal measures in the basin, the Cretaceous intrusions within the basin and the Hecate Pluton in the Urannah Complex were also sampled (Fig. 6). The purpose of sampling the Cretaceous intrusions and associated homfels was to compare their remanence directions with that of the overprint as observed in the coal measures.

Figure 5. Apparent Polar Wander Path (APWP) for Australia from the Triassic to the present. Data taken from Idnurm (1985) and Lackie and Schmidt (1993). KI, Cretaceous intrusions (this study). TH, Tahmoor and Coalclifif, MO, Moura, BL, Blackwater, SJ, Saraji, GC, German Creek (calculated from NRM directions, this study). MMO, HB, Milton Monzonite overprint, Homsby Breccia (Schmidt and Embleton 1981). Orthographic projection, centre of plot is 45°S, 150°E, grid 30°. Remanence directions found in Cretaceous intrusions and associated homfels within the Bowen Basin show both reversed and normal polarities (Table 4, Fig. 7), in accordance with the ages of the intrusions (140 Ma -120 Ma), being slightly older than the Cretaceous Normal Superchron. The Bundara Granodiorite and associated homfels shows both normal and reversed results (31,32), while all the other intrusions display only one polarity. These directions are broadly similar to those observed in the south of the basin (Moura, Table 3) but differ significantly from the directions observed in the north of the basin. The mean direction of the sites which give normal directions for the Cretaceous intrusions and homfels is Dec = 354.7°, Inc= -77.5° (N=10, a95 = 7.3°), while the mean direction of sites which give a reversed direction is Dec= 148.2°, Inc= 78.1° (N=7, a95 = 5.2°). Application of the reversal test of McFadden and McElhinny (1990) results in a "B" classification indicating that the directions are drawn from means that are antipolar and have an antipodal angle of 174°. The mean direction for the combined data set is Dec= 344.0°, Inc= -78.1° (N=17, ag^ = 4.6°) which gives a palaeomagnetic pole of Lat = 42.8°S, Long = 156.7°E (dp = 8.2, dm = 8.7). The resulting palaeomagnetic pole falls on the Australian APWP on the Early to mid Cretaceous section of the path (Fig. 5) indicating that Cretaceous intrusions in the Bowen Basin retain a primary remanence.

324

Table 4. Mean remanence directions after thermal demagnetisation of Cretaceous intrusions and associated homfels N Inc Site Dec a95 21.1 16 Harrow Crk Sill 9 -66 8.9 -80.4 4 8.4 32 Bundara Gd 235.7 341.1 -74 38 Daunia Granodiorite 3 10.1 62 Hfls Blenheim Ki 5 -76.8 299 3 64 Ki -73.4 5 319.9 4.3 4 72 Hecate 13.7 14.1 -78 -70.7 73 Hecate 5 7.1 8.5 74 Hecate 32 -64.2 13.2 5 -76.4 6 25 17.1 CI 6 Hecate C51Ki 5 307.9 -78.9 10.9 Mean 10 354.7 -77.5 73 17LKiIntr 5 n.i 168.0 74.6 18 Winchester Intr 5 182.9 78.9 7.7 122.4 31 BiandaraGd 3 69.9 6.6 54 Hfls RedcUff 78.2 3 168.6 8.0 78.2 55 Hfls RedcUff 5 153.3 8.9 149.4 56 Hfls RedcUfiF 5 80.1 5.0 63 Ki 4 105.1 77.8 5.1 Mean 148.2 7 5.2 78.1

344

-78.1

4.6

Note, N, number of samples. Dec, Inc, a95, Declination, Inclination and Alpha 95 of mean remanence direction. Resulting palaeomagnetic pole is Lat = 42.8°S, Long = 156.7°E (dp = 8.2, dm = 8.7). Geographic location taken as 21°S, 148.3°E. Discussion and Conclusions There is now overwhehning evidence from a variety of techniques that the Sydney Basin has been subjected to low temperature metamorphism during the Mid-Qetaceous, particularly in the southeast. Evidence has been derived from palaeomagnetism (Schmidt & Embleton, 1981, Schmidt, 1982), fission-track dating (Morley et al, 1980), vitrinite reflectance (Middleton & Schmidt, 1982),fluidinclusion data and K/Ar dating (Bai et al, 1993). These last two avenues of research support the inference based on palaeomagnetic data that a thermal event occurred at about 90Ma involving uplift and supracmstal cooling from elevated temperatures. Ruid inclusions in quartz overgrowths in Narrabeen Group samples from only a few hundred metres depth reveal minimum homogenisation temperatures of up to 100°C requiring some considerable overburden, may be coincident with an enhanced geothemial gradient. K/Ar dates of fine authigenic illite reveal a consistent eastward decrease towards the coast from 146Ma to 91 Ma, closely paralleUng iso-refiectance contours (P.J.Hamilton, personal communication, 1994). Modelling of vitrinite reflectance values indicates palaeotemperatures of 130°C to 180°C and requires burial depths of 1500m to 2100m in the mid-Jurassic (Bai et al., 1993). This modelling is independent of that used by Middleton & Schmidt (1982) which arrived at similar conclusions. Overprinting, or complete remagnetisation, related to the Mid-Cretaceous thermal event have been recognised in many different rock-types throughout the Sydney Basin, including red, grey and green fine sandstones, siltstones, claystones, Permian and Early-Mid Jurassic igneous rocks, such as the Gerringong Volcanics, the Milton Monzonite and the breccia diatremes. In sedimentary strata or intruded igneous rocks that have not been tilted, the Cretaceous overprint magnetisations are consistently directed northward, albeit steeply upwards. This property has afforded the coal industry a valuable orientation tool (Schmidt and Anderson 1992, Lackie and Schmidt 1993b).

325

Figure 6. Distribution of intrusions in the Bowen Basin (after Pattison, 1990). Sample localities cited in Table 4 are indicated by the relevant site numbers. The deflection of Cretaceous overprint directions by the Lapstone Monocline, and related structures, indicates that these tectonic structures post-date the Mid-Cretaceous, vis. --90Ma. This in turn implies that the major movement along the Lapstone Monocline post-dates '-90Ma and was probably related to precursor events leading to rifting in the Tasman Sea (Schmidt etal 1995). The overprint directions in the south of the Bowen Basin are similar to those observed in the Sydney Basin, and a similar age is inferred (-90 Ma), whereas the direction in the northem part of the basin indicates a younger age of overprinting (Fig. 5). The magnetisation of the Cretaceous intrusions is in accordance with the age of the intrusions (140Ma- 120Ma) and clearly differs from the overprint magnetisations measured in the north of the basin where the intrusions are found. As would be expected, the direction observed in the intrusions and associated homfels is similar to the Late Cretaceous overprint documented in the coal measures in the south of the basin. The magnetic overprint in the Bowen Basin is not as pervasive as that observed in the Sydney Basin and given the Bowen Basin's distance from the Australian margin it most likely has not been as strongly affected thermal events associated with rifting. Other studies document a Cretaceous 326

thermal event (Marshallsea et al 1985, Mallett et al 1990) and the remanence data agrees with this in the south of the basin, but younger magnetic overprints are indicated in the north of the basin. The consistent single polarity NRM direction observed in the coal measures does however suggest a Cretaceous age of remagnetisation and it is feasible that there are present field components in the NRM resulting in a shallower inclination. The palaeomagnetic results indicate that the Cretaceous plutonism did not directly remagnetise the Permian coal measures in the Bowen Basin except when those sediments were contact metamorphosed (Table 4). The overprint magnetisation in the south of the basin was clearly acquired during the Cretaceous Normal Superchron at -90 Ma (Fig. 5), indicating that thermal effects (uplift etc) other than those caused by Early Cretaceous intrusions were the cause of the remagnetisation.

Figure 7. Resultant site mean directions obtained from Cretaceous intrusions and homfels in the Bowen Basin. Closed (open) symbols represent lower (upper) hemisphere. REFERENCES Bai, G.R, Eadington, RJ. & Hamilton, RJ., 1993: Fluid history investigation of Permo-Triassic sandstones in the Sydney Basin. CSIRO Division of Exploration Geoscience, North Ryde, Restricted Report 417R. Embleton, B.J.J and McDonnell, K.L., 1980. Magnetostratigraphy in the Sydney Basin, Southeastern Australia, J. Geomagn. Geoelectr., 32, suppl. III., 1-10. Fisher, R.A., 1953. Dispersion on a sphere. Proceedings of the Royal Society A217, 295-305. Idnurm M., 1985. Late Mesozoic and Cenozoic palaeomagnetism of Australia - 1 . A redetermined apparent polar wander path. Geophys. J. R. astr. Soc. 83, 399-418. Lackie, M.A., 1993. Drillcore orientation using palaeomagnetism: preliminary results, Bowen Basin, Australian Coal Geology, 1993, 9,47-48. Lackie, M.A. and Schmidt, P.W., 1992a. Some examples of drill core orientation using palaeomagnetism in the south Sydney Basin. Restricted Report 300R, CSIRO Division of Exploration Geoscience, North Ryde, NSW. Lackie MA. & Schmidt P.W. 1992b. Preliminary study of drillcore orientation using palaeomagnetism, German Creek CoUiery. CSIRO Restricted Report 345R. pp. 7. Lackie M.A. & Schmidt P.W. 1993a. Remagnetisation of strata during the Hunter-Bowen Orogeny. Exploration Geophysics 24, 269-274. Lackie, M.A. and Schmidt, P.W., 1993b. Drill core orientation using palaeomagnetism, Exploration Geophysics, 1993, 24, 609-614. McFadden P.L. & McElhinny M.W., 1990. Classification of the reversal test in palaeomagnetism. Geophys. J. Int. 103, 725-729. Mallett C.W., Russell N. & McLennan T. 1990. Thermal history of the Bowen Basin. Proceedings of the Bowen Basin Symposium 1990, 15-20. Marshallsea S.J., Green P.F., Duddy LR. & Gleadow A.J.W. 1985. The thermal history of the southem Bowen Basin: an apatite fission track study. Proceedings of the Bowen Basin Symposium 1985, Geol. Soc. Aust. Abstracts 17,109113. Middleton, M.F. and Schmidt, P.W., 1982. Paleothermometry of the Sydney Basin, J. Geophys. Res., 87, 5351-5359.

327

Morley, M.E., Gleadow, AJ.W., & Levering, J.F., 1980: Evolution of the Tasman rift: Apatite fission track dating from the southeastern Australian continental margin. Proc. Vth Gondwana Symp., 289-293. Pattison, LP., 1990. Igneous intrusions in the Bowen Basin. Unpublished M.A.Sc. thesis, QUI. Schmidt, P.W. and Anderson, J.C., 1992. Orientation of drill core using palaeomagnetism in coal exploration. Aust. Coal GeoL, 8,18-20. Schmidt, RW. and Embleton, B.J.J., 1981. Magnetic overprinting in southeastem Australia and the thermal history of its rifted margin, J. Geophys. Res., 86, 3998-4008. Schmidt, P.W. and Lackie MA. 1995. ACARPC1615: Palaeomagnetic drill core orientation. CSIRO Division of Exploration and Mining report 144F, pp64. Schmidt P.W., Lackie M.A. & Anderson IC. 1995. Palaeomagnetic evidence for the age of the Lapstone Monocline, NSW. Aust. Coal Geol. 10,13-22. Schmidt, P.W., 1982. Linearity spectrum analysis of multicomponent magnetisation and its application to some igneous rocks from southeastem Australia, Geophys. J. R. astr. Soc., 70, 647-665.

328

^ ^ {} *

MID AND EARLY LATE CRETACEOUS BREAK-UP BASINS OF THE SOUTH ISLAND, NEW ZEALAND M.G. Laird Department of Geological Sciences, University of Canterbury, ChristchurcK NZ Summary In the South Island, the change from a convergent margin to an extensional tectonic regime in mid to late Albian times is represented by a major unconformity. The onset of rifting and continental break-up resulted in the widespread development of grabens and half-grabens, the majority infilled with non-marine deposits; in Marlborough, however, the infill consisted of deep water marine sediments. Commonly, non-marine sediments adjacent to a bounding fault consist of locally-derived breccia deposited as talus or debris flows on alluvial fans, passing via fluvial deposits into a distal lacustrine environment. Active faulting was the dominant control on subsidence and sediment supply, and continued in some areas until the initiation of sea floor spreading in early Campanian times. In Marlborough, marine half-graben fill consists dominantly of debris flow deposits and turbidites at the base, fining upwards to mudstone. From early Cenomanian times onwards sedimentation in half-grabens ceased and accommodation was controlled by regionwide slow subsidence and episodic relative sea level changes resulting in deposition of paralic to shallow marine sediments. Sporadic volcanism shows a geochemical signature consistent with intraplate extension. Introduction The pre-mid Cretaceous basement terranes of New Zealand are usually divided into two major groups termed the Western and Eastem Provinces. The Westem Province represents a fragment of Gondwana, comprising mainly early Paleozoic sedimentary sequences cut by Late Paleozoic and Late Mesozoic granitoids, and, in the South Island, lying mainly west of the Alpine Fault; while the Eastem Province is made up of incomplete remnants of magmatic arcs, forearc basins, trench-slope basins, and accretionary complexes. From the Permian to the Early Cretaceous most rocks in the New Zealand region formed under the influence of convergent margin tectonics inferred by Bradshaw (1) to be driven by the oblique subduction of the Phoenix Plate. The later stages of this regime in the Early Cretaceous were accompanied in the Eastem Province by widespread compressional deformation, low-pressure metamorphism, uplift and erosion, and, in the Westem Province, by high-pressure metamorphism and calc-alkaline plutonism. A change of tectonic pattem, around 105±5 Ma (middle Albian), affected the whole New Zealand region. The older, subduction-related rocks are everywhere separated by a major unconformityfi-omyounger, less-deformed strata (Laird, 2). Throughout the South Island, the oldest strata overlying the unconformity contain fossils of mid to late Albian age. This unconformity marks the change in tectonic regime from convergent margin to extension, attributed by Bradshaw (1) to collision of the spreading ridge between the Phoenix and Pacific Plates. On the West Coast, the youngest subductionrelated granite gives a U-Pb zircon SHRIMP date for intrusion of 109.6±1.7 Ma (Muir et al, 3), while the basal tuffs forming part of the oldest extension-related cover strata in the area yield SHRIMP ages on zircons of 10L2±2.0 Ma (S.D. Weaver, pers.comm. 1996). There is thus an approximately 8 m.y. gap in age between basement and cover strata here. K/Ar dates on tuffs overlying the unconformity at Kyebum in Otago (Fig. 1) range between 105±2 Ma and 108.3±2.8 Ma, suggesting an earUer date of deposition: ignimbrites from correlative deposits at Shag Point give dates of 101±2 Ma and 103±2 Ma, similar to those of the West Coast (Adams and Raine, 4). However, the K/Ar method used on the Otago rocks is likely to be less precise than dating by the SHRIMP method, and it is possible that the Otago volcanic rocks are coeval with those of the West Coast. The onset of rifting and continental breakup resulted in the widespread development of grabens and half-grabens. Although the majority of these basins are infilled with non-marine deposits, in the northeastem South Island halfgrabens were infilled by marine sediments, attributed to deposition on the depressed thin continental crust of the Pahau subterrane, which underlies the East Coast of the North Island, Marlborough, and north Canterbury (Fig. 1). The remaining terranes of the South Island had normal continental thickness, and were relatively buoyant compared with the Pahau subterrane, resulting in non-marine sedimentation.

329

Figure 1. Locality map of the South Island, New Zealand, showing main areas of mid to early Late Cretaceous outcrop. Mid and early Late Cretaceous marine successions These are limited to the Marlborough region of the northeastern South Island (Fig. 1). Basin configuration determined from paleocurrent patterns, isopachs and lithofacies pattems indicates a succession of dominantly marine basins, terminating to the south near Kaikoura, and opening towards the northeast to merge with a similar basin succession in the southeast of the North Island. The marine succession in Marlborough begins with infill of half grabens, developed on the mid Cretaceous unconformity surface. Locally they accumulated in excess of 2000m of mid to late Albian deposits of mainly sediment gravity flow origin (Laird, 5). The basal deposits are coarse-grained, in many instances consisting of clast or matrix-supported conglomerate containing scattered shelly fossils, locally infilling channels between 500m and 800m deep which may represent submarine canyons. Infill includes olistostromes up to 240m thick, and containing outsize blocks up to 20m in diameter (Laird, 5). The basal coarse sediments are inferred to have been deposited as a result of active faulting associated with the formation of the half grabens. The basal coarse-grained deposits fine upwards throughout the region first into turbidite successions, commonly hundreds of metres thick (Fig. 2), and then into laminated mudstone-dominated deposits, suggesting waning of tectonic activity along bounding basinal faults through time. Paleodepths are inferred to be mainly bathyal, with possibly shallower depths in the south (Laird, 5).

330

Figure 2. Vertically-dipping turbidites of late Albian age in a marine succession, Marlborough The half-graben phase of deposition was followed by uplift and erosion in the early Cenomanian. Subsequent basinal sedimentation in the early Late Cretaceous was controlled by regional slow subsidence and relative changes of sea level, resulting in a series of unconformity-bounded sedimentary sequences (Fig.3). The first depositional cycle, of early Cenomanian age, rests with local slight angular unconformity and a notably scoured surface on both immediatelypreceding half-graben deposits and on basement. The sequence involved deposition of shallow marine and non-marine sandstones and mudstones, in sharp contrast to the underlying deep marine half-graben deposits. The eruption of extensive alkaline basaltic lavas accompanied and terminated the sequence, indicating still-active tectonism. Age INTERNATIONAL NEW ZEALAND (Ma) STAGE STAGE

Haumurian SO-

BS9095

Campanlan

EVENT

Santonian Coniacian

Turonian

Teratan

Mangaotanean

2 uj S

m <0 a

Kyeburn

Puysegur

West Coast

Volcanism (Marlborough and Canterbury)

Arowhanan Cenomanian

100

RSLF

z Volcanism p (Marlborough I and N. Canterbury) RSLF ^

1101

Pitt Island

Sea floor spreading begins

PIrlpauan

Ngaterian

105-

GENERALISED STRATIGRAPHY Marlborough

Motuan

Volcanism (West Coast)

Urutawan

Volcanism (aago) Halt grabens develop

Albian

Granite Intrusion (West Coast)

Figure 3. Correlation of events and generalised stratigraphy of South Island mid to early Late Cretaceous strata. RSLF = Relative Sea Level Fall. Resting concordantly but erosively on these deposits is a sequence of late Cenomanian to Coniacian age, which shows a further basinward facies shift, indicating another drop in relative sea level. The succession is dominated by shallow marine highly-burrowed glauconitic sandstone and mudstone, representing a condensed sequence not exceeding 100m in thickness. It passes into non-marine deposits in the extreme southwest. Alkaline basaltic lavas are present locally. This sequence is in turn overlain concordantly but erosively by glauconitic sandstone of Santonian to Campanian age which passes rapidly upwards into thick carbonaceous mudstone. This represents a deepening succession attributed to thermal subsidence in response to sea floor spreading as New Zealand separated from Australo-Antarctica at about 82 Ma. 331

Mid and early Late Cretaceous non-marine successions These successions form the oldest post-unconformity deposits throughout most of the South Island, and locally reach several thousands of metres in thickness. That these sequences were commonly developed in grabens or half-grabens is clearly shown on offshore seismic profiles and by the infill geometry of some outcropping successions. In many instances seismic reflection profiles indicate that the normal faults forming the margin of the half-grabens decrease in dip downwards, and probably form hstric faults at depth (TumbuU, Uruski, et al. 7). Half-grabenfillis also often fan-shaped on seismic profiles, suggesting that active faulting continued throughout sedimentation. Offshore mid Cretaceous grabenfill successions have so far been drilled only in the Great South Basin, where the oldest sediments penetrated were of Cenomanian age (Beggs, 8). However, the deepest parts of this succession have not been drilled, and sedimentation could well have started earher. The basins of the Chatham Islands (Fig. 1), which are paralic, represent a transition between the deep marine basins of the northeastern South Island and the wholly non-marine basins of the remainder of the Island. The only exposures occur on Pitt Island, where between 240m-400m of only sli^tly indurated late Albian to Santonian mainly sandy sediments occurring in a fault-disrupted succession, with the base unseen, have been described by Campbell et al. (9). The outcrops are correlated with the upper part of the infill of E-W oriented half-grabens seen in seismic profile on the Chatham Rise. The oldest portion of the exposed sequence, of late Albian age, and deposited in a fluvial environment, was succeeded by sediments with bimodal cross-stratification, dense burrowing, marine bivalves, and common flaser and linsen structures suggesting a tidally-dominated shallow marine environment. At the top of the marine succession, channels up to 2m deep infilled with layered sand and mud withflaserand linsen structures, clay drapes and rounded clay clasts, and the occurrence of coaly horizons, rootlet layers, and rare coahfied trees in apparent position of growth, suggest an estuarine to onshore environment. The latest Albian estuarine sequence is succeeded by up to 260m of section characterised by depositional cycles 1.6m-6m thick and consisting of clay, wavy-laminated very fine sand interbedded with mud, cross-bedded or parallel-bedded fine sand, sandy mud or muddy sand with rootlets and pyrite nodules, and lignite layers up to Im thick, with tree trunks in growth position. However, the presence of layers containing flaser and linsen structures and dinoflagellates indicates periodic marine influence. This portion of the section is interpreted to have accumulated in the terrestrial/marginal marine part of a delta system, with terrestrial environments dominating. This part of the section shows no observable stratigraphic breaks, and is of late Albian to Coniacian age. The sequence passes up rapidly but gradationally into shallow marine tuffs. Fully non-marine mid Cretaceous successions occur on the east coast of the South Island south of the Chatham Rise in the Canterbury, Otago, and Great South Basins, to the south in the Westem Southland Basins, and on the West Coast of the South Island (Fig. 1). The sediments infilling all these basins share similar characteristics, and the varying nature of the basement between Eastem and Westem Provinces has no detectable influence on basin development or type of deposition. Essentially complete stratigraphic successions occur at Kyebum (central Otago), Puysegur Point in westem Southland, and in the Paparoa Range on the West Coast (Fig. 1). O&er areas of more limited outcrop, however, show similar characteristics. All infill fault-controlled depressions. At Kyebum, approximately 4000m of coarse-grained breccia, conglomerate and sandstone, some of it reddish-brown in colour, make up the Kyebum Formation (Bishop and Laird, 10). The basin was controlled by two orthogonally-developed normal faults, one trending NW-SE, and the other NE-SW. The basal sediments rest with strong angular unconformity on schist of the Eastem Province. Sihcic tuff horizons low in the succession have given early to middle Albian ages (Adams and Raine, 4), while pollen assemblagesfiromthe middle of the unit suggest a late Cenomanian or early Turonian age. No erosional break is evident within the succession, which appears to represent continuous depositionfiromAlbian to ?Turonian times. Paleocurrent directions and Uthofacies variations show a close relationship to the basin-bounding faults. The gradational contact from unsorted breccia to poorly-sorted breccia-conglomerate is marked by a gradual increase in the degree of rounding of the clasts, a reduction in the maximum clast size, and the appearance of well-defined bedding surfaces, inferred to represent the change from talus deposits to water transport on alluvial fans. Eventually angular clasts disappear entirely, and breccia-conglomerate passes into poorly-sorted, poorly-rounded conglomerate with sandstone matrix, commonly infilling channels in upward-fining sequences. This is inferred to represent a progression into first a braided stream system, and then into the flood plain environment of a meandering stream of low sinuosity. Locally intercalated with the flood-plain Uthofacies is a thin unit of calcareous sandstone and siltstone, inferred to represent a lacustrine environment. In westem Southland, mid Cretaceous sedimentary rocks are assigned to the Puysegur Group, which is exposed only along a 20km wide coastal strip (Tumbull, Umski, et al., 7), although the Group is inferred from seismic evidence to be more extensive offshore. Palynological dating gives ages of late Albian to early Cenomanian. Outcrops of Puysegur Group unconformably overlie a weathered basement of granite and metasedimentary rocks, and are themselves overlain by a regional unconformity which separates them from overlying Cenozoic sequences. The lower part of the Group consists of up to 200m of interbedded granule to pebble conglomerates, cross-stratified sandstone, and variably carbonaceous mudstone which commonly contains thin bituminous coal lenses and rootlet horizons. Overall sedimentary features, including fining upward cycles, indicate deposition in shallow, meandering, fluvial channel and floodplain environments. The overlying strata, at least 1250m thick, are inferred to have formed in a lacustrine pro-delta fan and 332

channel system. Three facies associations have been identified: (1) a lacustrine basin-margin association of boulderconglomerate, pebbly sandstone and mudstone; (2) an inferred pro-delta channel association consisting mostly of thick channelled sandstone bodies enclosed by mudstone and thin, graded sandstone; and (3) a pro-delta fan association of graded sandstone-mudstone cycles.

Figure 4. Outcrops of matrix-supported (rear) and clast-supported (middle) Hawks Crag Breccia, coast west of Paparoa Range. Note imbrication. Hammer for scale. On the west coast of the South Island, west of the Alpine Fauk, outcrops of breccia collectively known as the Hawks Crag Breccia, a formation of the Pororari Group, are scattered but widespread over a NE-SW extent of approximately 320km (Fig. 4 ). Palynological studies indicate that the lower portion of the Pororari Group is mid to late Albian in age (Nathan et al., 11), while tuffaceous sediment from near the base of the Group in the northem Paparoa Range gives a U-Pb zircon SHRIMP age of 101.2±2.0 Ma (S.D. Weaver, pers. comm. 1996). The original upward time range of the group in the region is unknown as all exposures are either truncated by unconformities or are bounded by faults. The youngest recorded age gives a range between late Albian and Cenomanian (Adams, 12), and a possible correlative of the Pororari Group, lying approximately 100km northeast of the Paparoa Range, has yielded a poorly-constrained microfloral age in the range late Cenomanian to early Santonian (Johnston, 13). indicating that Pororari Group deposition may have extended locally well into the late Cretaceous. Hydrothermally altered granite immediately below a detachment fault associated with half-graben formation in the southem Paparoa Range yields a K/Ar age of 85 Ma, which is considered to date the hydrothermal activity associated with rapid uplift on the fault (Tulloch & Palmer, 14), and probable continued or renewed deposition of Pororari Group. This age is similar to that of the oldest Tasman Sea floor, and directly links the early break-up phase of extension to continental rifting associated with formation of the Tasman Sea basin. In the Paparoa Range (Fig. 1), deposition of the Pororari Group was associated with a NNE-SSW extensional event related to the formation of a metamorphic core complex (Tulloch and ELimbrough, 15). The detachment faults on the northeast and southwest sides of the Paparoa Core Complex had opposite senses of shear, with cover rocks on both sides moving away from the metamorphic core and forming opposite-facing half-grabens, into which the Pororari Group sediments were deposited. The most diverse range of lithologies and sedimentary facies occurs in the southem Paparoa Range, where over 2000m of strata have been divided into three facies assemblages (Laird, 16). Assemblage 1, composed of matrix- and clast supported breccia (Hawks Crag Breccia) containing commonly imbricated subrounded to subangular clasts embedded in a coarse sandy miatrix, is inferred to represent the deposits of debris flows and sheet flows on alluvial fans. Assemblage 2, which consists of pebbly sandstone and massive or graded sandstone, with interbeds of carbonaceous mudstone, is interpreted to be the deposits of debris flows and turbidity currents into a lacustrine setting. Assemblage 3, which consists dominantly of highly carbonaceous laminated mudstone, locally including horizons of thin-bedded graded sandstone, is inferred to represent distal lacustrine deposition. The three facies assemblages intertongue, and are considered to represent elements of a fan-delta system debouching into a lake (Fig. 5). This fan-delta system is inferred to have infilled an actively subsiding half graben, the paleocurrent pattem and direction of thinning of lithostratigraphic units indicating sediment transport from the NNE, and downthrow of the bounding fault to the SSW. This is consistent with the postulated movement and half-graben formation on detachment faults overlying the Paparoa Core Complex, and associated half-graben formation (Tulloch and Kimbrough, 15).

333

Figure 5. Inferred relationship of the lithofacies assemblages making up the Pororari Group in the southern Paparoa Range In the northem Paparoa Range, on the northeast flank of the Paparoa Core Complex, the basal Pororari Group comprises 60m of vitric tuff, dark carbonaceous shale, and minor arkosic sandstone interbedded locally at its base with layers of glassy hyalodacite in a lacustrine environment (Nathan, 17). It is overlain concordantly but sharply by up to 600m of nonvolcanogenic arkosic sandstone, dark carbonaceous shale with local coal lenses, and conglomerate consisting of rounded clasts of basement rocks. The depositional environment is inferred to be floodplain with scattered shallow lakes and swamps. The sandstones and conglomerates of this unit interfinger with and are overlain with rapid gradation by clast and matrix supported breccia (Hawks Crag Breccia), which may exceed 4500m in thickness (Nathan, 17). As in the southem Paparoa Range, the Hawks Crag Breccia is inferred to have been deposited largely by debris flow processes on a series of alluvial fans infilling a half-graben or grabens. Mid and early Late Cretaceous magmatic activity As noted earlier, early deposition in half-grabens was accompanied by silicic volcanism of Albian age in the West Coast and in Otago. This took tiie form of tuffs in the northem Paparoa Range and in the Kyebum area, and ignimbrite flows at Shag Point (Fig. 3). The next major phase of volcanism was limited to the northeast of the South Island. In Marlborough, widespread dike intrusion was accompanied by major outpouring of alkaline basalts of mainly Cenomanian age (Fig. 3). The geochemistry of the basalts shows that they contain no subduction-related signature, and they are consistent with intraplate volcanism (Laird, 5). At Mandamus, in north Canterbury, intrusive igneous rocks consisting of syenite and gabbro, together with associated trachytic lavaflows,have been Rb-Sr dated at 97.0±0.5 Ma (Weaver and Pankhurst, 18). They are the same age as and have similar geochemistry to the Marlborough Cenomanian igneous rocks, and are consistent with asthenospheric uprise and extension of thin continental crust, represented by the Pahau subterrane. The Mt Somers Volcanics Group of central Canterbury consists of garnet-bearing andesites and rhyolites, Rb-Sr dated at 89.3±2.0 Ma (Barley, 19). The geochemistry suggests high mantle heat flow and resultant crustal anatexis, and is compatable with an environment of crustal extension. The contrast in composition between these and Cenomanian alkaline magmas further north may be largely a consequence of the degree of interaction between mantle-derived magmas and the thicker continental crust underlying Mt Somers (Weaver and Pankhurst, 18). Locally-developed basalt flows of probably similar age also occur in Marlborough (Fig. 3). Their geochemistry clearly indicates an intraplate, alkaline basalt afiSnity, and is closely similar to that of the Cenomanian igneous rocks of Marlborough (Crampton and Laird, 20). Fault Patterns On the West Coast of the South Island, the faults associated with mid Cretaceous movements, and orientation of late Santonian to early Campanian dike swarms, fall into a distinct pattem of WNW to NW-oriented stmctures, which are essentially parallel to the trend of Late Cretaceous spreading axes in the Tasman Sea, and consistent with the NE to NNE direction of shear on detachment faults associated with metamorphic core complexes (Laird, 2). The faults are also parallel to the rift system occupying the westem edge of the southem Lord Howe Rise, and may represent its extension. This WNW to NW trend is therefore almost certainly related to early rifling associated with the separation of New 334

Zealand from Australia, although preceding the formation of the oldest sea floor in the Tasman Basin by approximately 20 Ma. It may represent an attempt by the Lord Howe Rise Rift System to propagate through the New Zealand region, its extensions being the EW faults bounding Cretaceous half grabens on the Chatham Rise, which were likely to have had a more WNE orientation prior to mid Cenozoic rotation due to Alpine Fault movement (Laird, 2). The 85 Ma K/Ar age on hydrothermally altered granite beneath a detachment fault associated with a WNW-trending half-graben suggests that semi-continuous tectonic activity accompanying breakup under NNE to NE extension may have continued until sea floor spreading in the Tasman Sea was initiated. In Marlborough, in the NE of the South Island, dike swarms of Cenomanian age, which are widespread, appear to have been intruded along a conjugate set offracturesoriented E-W and NNE to NE (Laird, 5). The former direction is parallel to the dominant fault system initiated in the mid to late Albian on the Chatham Rise to the south, and therefore likely to be related to early rifting in the Tasman Sea. The latter is parallel to the dominant NE direction of faulting in the Great South Basin. The NE trend in the Great South Basin of both basin axis and major bounding faults is parallel to the rifted Campbell Plateau edge and to the fiiture spreading axis developed between New Zealand and Antarctica, and the basin is thus likely to have developed as a result of initial rifting along this trend. Thus in the eastem South Island, rifting developed as a result of early extension both between New Zealand and Antarctica, and between New Zealand and Australia. Conclusions Although geophysical interpretation of marine magnetic anomalies indicates that seafloorspreading in the New Zealand area did not occur until about 82 Ma, extension accompanied by rifting began in the region 20-25 m.y. earlier, at about 105 Ma, supplanting a long period of convergent margin tectonics. The change to extension is characterised by a major regional mid to late Albian angular unconformity, and the widespread development of half-grabens. In most cases halfgraben infill is non-marine: in the northeastem South Island, however, half-grabens were infilled with deep water marine sediments deposited on thin continental crust. Non-marine half-graben infill exhibits a similar motif in most areas of outcrop: close to the active bounding fault, locally derived breccia was deposited as talus or debrisflowson alluvial fans, passing directly as fan deltas or viafluvialdeposits into a lacustrine environment. Active faulting was the dominant control on subsidence and sediment supply, and continued, at least intermittently, in some areas until the initiation of seafloorspreading in Campanian times. No significant breaks in deposition have been recognised, although this may be aftmctionof limited preservation of the younger part of the successions. Half-grabenfillin Marlborough consisted dominantly of deep marine sediment gravityflowdeposits forming large-scale fining-upward units, the basal deposits consisting of olistostromes or debrisflows.In contrast to the non-marine basins, major faults bounding half-grabens became inactive by early Cenomanian times, and accommodation was controlled by slow subsidence and episodic relative sea level changes, resulting in deposition of mainly shallow marine sediments forming a series of unconfomiity-bounded sequences. Sporadic volcanism shows a geochemical signature consistent with intraplate extension of a thin continental crust. The Pitt Island succession is distinctive in containing both shallow marine and non-marine sediments deposited along a fluctuating shoreline. No significant breaks in deposition have been recognised, and the late Albian to Coniacian section is thin and relativelyfine-grainedcompared with those investigated elsewhere in the South Island. A stable, very slowly subsiding platform is inferred, with neghgible tectonic activity. Faults active during sedimentation form a consistent and distinctive pattem. In the westem South Island they are consistently WNW to NW, sub-parallel to the Lord Howe Rise rift system and to the future late Cretaceous Tasman spreading axis. The E-W faults of the Chatham Rise are also likely to be part of this system. By contrast, the trend of major faults in the Great South Basin, as well as of the basin axis, is NE, parallel to the rifted edge of the Campbell Plateau and to the fiiture spreading axis developed between New Zealand and Antarctica. In Marlborou^, dike swarms of Cenomanian age appear to have been intruded along a conjugate set of fractures oriented E-W and NNE to NE, i.e., parallel to both East and West coast trends. It is clear that the Cretaceous fault pattem developed as a result of early extension both between New Zealand and Antarctica, and New Zealand and Australia. References 1. Bradshaw, J.D. Cretaceous geotectonic pattems in the New Zealand region. 1989. Tectonics 8 (4), pp 803-820 2. 3. 4.

Laird, M.G. Geological aspects of the opening of the Tasman Sea. In "Evolution of the Tasman Sea Basin" (Eds. G.J. van der Lingen, K.M. Swanson, and R.J. Muir), Balkema, 1994. pp 1-17. Muir, R.J., Ireland, T.R., Weaver, S.D., and Bradshaw, J.D. 1994. Ion microprobe U-Pb zircon geochronology of granitic magmatism in the Westem Province of the South Island, New Zealand. Chemical Geology 113, pp 171189. Adams, C.J. and Raine, J.I. Age of Cretaceous siliceous volcanism at Kyebum, central Otago, and Palmerston, eastem Otago, South Island, New Zealand. 1988. New Zealand Journal of Geology and Geophysics 31, pp 471475. 335

Laird, M.G. Cretaceous stratigraphy and evolution of the Marlborough segment of the East Coast region. Proceedings, 1991 New Zealand Oil Exploration Conference, Christchurch. Ministry of Commerce, 1992. pp 89-

Crampton, J.S., et al. An interim New Zealand geological time scale. Institute of Geological and Nuclear Sciences Science Report 95/9,1995. 5 pp. Tumbull, LM., Uruski, C.I., et al. Cretaceous and Cenozoic sedimentary basins of Westem Southland, South Island, New Zealand. Institute of Geological and Nuclear Sciences monograph 1. Institute of Geological and Nuclear Sciences, 1993. 86 pp. Beggs, J.M. Depositional and tectonic history of the Great South Basin. In "Sedimentary Basins of the World: South Pacific" (Ed. P.F. BaUance), Elsevier, 1993. pp 365-373. Campbell, H.J., et al. Cretaceous-Cenozoic geology and biostratigraphy of the Chatham Islands, New Zealand. Institute of Geological and Nuclear Sciences Monograph 2. Institute of Geological and Nuclear Sciences Ltd., 1993. 269 pp. Bishop, D.G., and Laird, M.G. Stratigraphy and depositional environment of the Kyebum Formation (Cretaceous), a wedge of coarse terrestrial sediments in central Otago. 1976. Joumal of the Royal Society of New Zealand 6 (1), pp 55-71. Nathan, S. et al. Cretaceous and Cenozoic basins of the West Coast region. South Island, New Zealand. New Zealand Geological Survey Basin Studies 1, DSIR, 1986. 90pp. Adams, D.P.M. Cretaceous and Eocene geology of south Westland. Unpublished M.Sc. thesis. University of Canterbury. 339 pp. Johnston, M.R. Geology of the St Amaud district. Southeast Nelson (Sheet N29). New Zealand Geological Survey Bulletin 99. New Zealand Geological Survey. 119 pp. Tulloch, A.J., and Pahner, K. Tectonic imphcations of granite cobbles from the mid-Cretaceous Pororari Group, southwest Nelson, New Zealand. 1990. New Zealand Joumal of Geology and Geophysics 33(2), pp 205-217. Tulloch, A.J., and Kimbrough, D.L. The Paparoa Metamorphic Core Complex, New Zealand: Cretaceous extension associated with fragmentation of the Pacific margin of Gondwana. 1989. Tectonics 8 (6), pp 1217-1234. Laird, M.G. Coarse-grained lacustrine fan-delta deposits (Pororari Group) of the northwestem South Island, New Zealand: evidence for Mid Cretaceous rifting. In "Sedimentary Facies Analysis" (Ed. A.G. Plint). Special Publication no. 22 of the Intemational Association of Sedimentologists, 1995. Blackwell. pp. 197-217. Nathan, S. Sheet S31 and part S32 - Buller-Lyell. Geological Map of New Zealand 1:63 3 60. New Zealand Geological Survey, DSIR. Weaver, S.D., and Pankhurst, R.J. A precise Rb-Sr age for the Mandamus Igneous Complex, North Canterbury, and regional tectonic implications. 1991. New Zealand Joumal of Geology and Geophysics 34(3), pp 341-345. Barley, M.E. Origin and evolution of mid-Cretaceous, gamet-bearing, intermediate and silicic volcanics fi-om Canterbury, New Zealand. 1987. Joumal of Volcanology and Geothermal Research 32, pp 247-267. Crampton, J.S., and Laird, M.G. The Burnt Creek Formation and Late Cretaceous basin development in Marlborough, New Zealand. In press. New Zealand Joumal of Geology and Geophysics.

7. 8.. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

100.

336

THE HUTTON SANDSTONE TO BIRKHEAD FORMATION TRANSITION, GIDGEALPA SOUTHERN SOME: SHOWING AN ASSOCIATED UNCONFORMITY AND PROVENANCE CHANGE FROM CRATON-DERIVED TO VOLCANIC-ARC-DERIVED SEDIMENT. £. Lanzilli and PJ. Boult University of South Australia, School of Engineering, Applied Geology Summary It has been determined that a sequence boundary exists within the Button Sandstone near the top of the formation. This erosional surface has been identified on seismic and from core logs. The sequence boxmdary separates the Hutton Sandstone into a lowstand systems tract below and a late stage lowstand systems tract above. The transition between the Hutton Sandstone and the Birkhead Formation is the first transgressive surface and marks the start of a transgressive systems tract and also the start of the Birkhead Formation. The lower part of the Birkhead Formation consists of reworked Hutton Sandstone sediments with the mid Birkhead Formation sediments consisting of lithic arenites. This change is related to a provenance change which has been proved by petrophysical data, Sm/Nd dating and Dipmeter data. Introduction In the Gidgealpa southem dome, seismic interpretation, sequence stratigraphy, Sm/Nd dating, dipmeter data and petrophysical techniques have provided a new data point for the interpretation of the Hutton Sandstone to Birkhead Formation transition. Stratigraphy and Sedimentology The bulk of Hutton Sandstone is a coarse grained quartz arenite representing a braided fluvial or low sinuosity fluvial environment. It was deposited during a period when river base level (lake level) was low, in a lowstand systems tract. Since subsidence at this time was either very slow or nil, accommodation space was at a minimum causing finer grained sediments to bypass large areas near the edge of the basin and deposit further into the basin depocentre. Seismic sections show erosional truncation below a seismic horizon near the top of the Hutton Sandstone, which correlates to an intraformational conglomerate on an erosional surface at Gidgealpa 17 (Figure 1). The model of Gravestock et.al.(l) is in agreement with our interpretation of this erosional surface or sequence boundary. The surface is probably formed by either regional tectonic uplift or an increase in subsidence causing a further base level fall with a negative accommodation space. Because tectonic activity at this time was greater to the east we would expect to find this sequence boundary extending into these areas. A sequence of finer quartz arenite beds with common shaly interbeds occurs between this previously mentioned erosional surface and the top of the Hutton Sandstone, it is of variable thickness and is interpreted as a late stage lowstand systems tract which filled in the palaeotopographic lows. At the top Hutton Sandstone to Birkhead Formation transition, a base level rise caused a transgressive surface (first flooding surface) to cover the Gidgealpa southem dome and surrounding areas. This formation change was formed by either a decrease in basin sag at the depocentre or an increase in sedimentation rate. The latter scenario is favoured, as it is associated with a sediment provenance change, seen fi-om conventional petrological studies, Sm/Nd dating, and dipmeter data. Petrology Petrological studies reveal a change in sediment type related to a provenance change, which occurs a short distance above the transgressive surface separating the Hutton Sandstone and Birkhead Formation. This data shows (Figure 2) that arenites of the late stage lowstand systems tract at the top of the Hutton Sandstone comprise medium to coarse grained quartz arenites where rock fragments are absent. Above the transgressive surface the basal part of the Birkhead Formation (transgressive systems tract) consists of reworked Hutton Sandstone sediments which are represented by a further reduction in grainsize and the intergranular porosity being filled by detrital clays and micas. Post-provenance-change, true Birkhead 'type' lithic arenites were deposited as a continuation of the transgressive systems tract with quartz contents typically less than 40%. These lithic arenites are associated with locally abundant calcite cement which is related to their diagenesis. Sm/Nd Dating To further prove that the change in sediment types was due to a provenance change, Neodymium model ages were calculated. Sm/Nd studies of a sediment involves estimating when its crustal precursors were fractionated fi:om the mantle. Weathering, diagenesis and erosion have little or no effect on Sm/Nd values. However, calculated age may not 337

have direct time significance since crustal reworking and mixing can effect values and thus only gives an indication of changes in provenance. Results (Figure 3) show that the lowstand systems tract sediments of the upper Hutton Sandstone and part of the base of the transgressive systems tract (base Birkhead Formation) have a Neodymium model age of approximately 1500 Ma. Provenance for this sediment is interpreted to be derived from the cratonic terrain to the southwest, with rocks varying in age from 1400 - 2000 Ma. The upper part of the transgressive systems tract, comprising lithic arenite sediments (above the provenance change) have Neodymium model ages of approximately 800 Ma. Similar results were reported by Whitford et al(2) at Bodalla South. Thus we suggest that these sediments are derived from the east. Dipmeter Data Dipmeter data (Figure 4) at Gidgealpa shows a change in palaeocurrent direction, the source direction changed from south-west to east, which is associated with this provenance change. This is in agreement with other workers (Watts(3) and Veevers(4)) who interpret sediment within the Birkhead Formation in the eastem Eromanga Basin and the equivalent Surat Basin sediments to be shed off a rapidly upHfting volcanic arc complex off the coast of Queensland. Conclusions The Upper section of the Hutton Sandstone has been divided into a lowstand systems tract and a late stage lowstand systems tract separated by sequence boundary. This sequence boundary can be seen as truncating reflectors on seismic sections and exists as an intraformational conglomerate in core logs. The sequence boundary was probably caused by tectonic uplift causing a further drop in an already low river base level. The Hutton Sandstone - Birkhead Formation transition is a change in depositional environments caused by a rise in river base level, probably due to an increase in sedimentation rate. The basal part of the Birkhead Formation is composed of reworked Hutton Sandstone sediments with a finer grainsize, as shown by petrological data. Dipmeter data also shows a change in palaeocurrent direction sourcing from the southwest to east. After deposition of the reworked sediments, petrological data shows the sediment type changed to a lithic arenite, with a palaeocurrent direction also to the west. The lithic arenite marks a change in sediment provenance from craton-derived to volcanic arc-derived. Sm/Nd dating show that the quartz arenites have a model age of approximately 1500 Ma and the lithic arenites have a model age of approximately 800 Ma. References 1. 2. 3. 4.

Gravestock, D., GrifiBths, M. and Hill, A., The Hutton Sandstone - Two Separate Reservoirs in the Eromanga Basin, South Australia, 1983, The APEA Joumal Vol 23 Pt 1, p 109 - 1 1 9 Whitford, D.J., Hamilton, P.J. and Scott, J, Sedimentary Provenance Studies in Australian Basins Using Neodymium Model Ages, 1994, The APEA Joumal, Vol 34 Pt 1, p 320 - 329 Watts, K.J., The Hutton Sandstone-Birkhead Formation Transition, ATP 269P(1), 1987, The APEA Joumal, Volume 27 Number 1, p 215 - 228 Veevers, J.J., Phanerozoic Earth History of Australia, 1984, Oxford .University Press, New York

Acknowledgments The authors would like to thank Santos and joint venture partners for access to data for this abstract.

338

Gidgealpa 17, cores 1 2 & 3 Hutton Sst. depth I ft 6000 depth ft

6003. 6006

6009 6012

6015 6018

6045

6048

ooo ®

- • * O oo® o O eo o o o o

o O ^

6051 6054 6057

3ZI3

6060 CO T3 C

c o

6021

6063 6066

6024

I a> c LU

6069

6027

6072

6030 6033

6075 Intraformational conglomerate

6078

6036

6081

6039

oo® ao' « oo' O O8 5 5 S Z o ..»» .«ooo® .»» . • eoo .••

6084

6042

0) (0 73 C (0(0 c o

£S oo (0

6087

6045 6090 Figure 1. shows the Gidgealpa 17 core log separated into two parts, the coarse grained Hutton Sandstone and the lower energy Hutton Sandstone. These two sections are separated by a intraformational conglomerate which marks the interHutton sequence boundary

339

mineralogy percentages

5944.0

a • B •

I o

a <h

?i + D-

& ®

O u

Figure 2. is a graph of percentages of rock constituents of sandstonesfromGidgealpa 32 and Gidgealpa 26. Rf = rock Fragments, Felds = Feldspars, Opq = Opaques, Hv = Heavy minerals, Org = organics, S-por = secondary porosity, Kaol = Kaolin, Qtz = Quartz, Ig-por = intergranular prosity, Calc = Calcite

340

DEPTH

MODEL AGE (millions of years)

(GISIALPA32)

Figure 3. shaded triangles indicate Sm/Nd dates and shows the age difference between the Hutton Sandstone and base of the Birkhead Formation and the mid Birkhead Formation.

341

GIDGEALPA SOUTHERN DOME - DIPMETER DATA N ^ 17

MID BIRKHEAD FORMATION (POST-PROVENANCE CHANGE/ TRANSGRESSIVE SYSTEMS TRACT)

r33*

2km

N 17

LOWER BIRKHEAD FORMATION (REWORKED HUTTON/ TRANSGRESSIVE SYSTEMS TRACT)

It" TOP HUTTON SANDSTONE (LOWSTAND SYSTEMS TRACT)

2kin

Figure 4. shows dipmeter direction for Gidgealpa 17, Gidgealpa 23, Gidgealpa 33 and Gidgealpa 34. Large arrows show overall palaeocurrent direction.

342

CHEMICAL CHARACTER OF GROUNDWATER IN THE WALLOON COAL MEASURES OF SOUTHEAST QUEENSLAND Jiaorong Li and Malcolm £. Cox Research Concentration in Environmental Earth Science, School of Geology, Queensland University of Technology Summary

The Middle Jurassic age Walloon Coal Measures (WCM) as the uppermost Mesozoic sedimentary formation in the Moreton Basin, Southeast Queensland, could be attractive sources for groundwater extraction. A number of boreholes were drilled into the formation by private and mainly intersect coal seams acting as aquifers. Quantity and quality of groundwater varyfromplace to place and over time. The overall problem is the existing salinity throughout aquifers. This has limited the use of groundwater suitable to stocks in most cases rather than irrigation. ITie formation is a typical alluvial deposits in fresh water of continental environment under warm and moist climate condition. The source of salinity, especially the high content of chloride ion in WCM groundwater are of interest in the study. Twenty boreholes were sampled which intersect aquifers in the formation. Samples were then analysed for major and minor constituents (Na, K, Ca, Mg, CI, HCO3, SO4 and Sr, Li, Si02, F) in order to exam the groundwater quality and to study the chemical evolution along flow path. It is concluded that the water chemical types are Na-C1,HC03 and Na-HC03,Cl; and alongflowpath the fomier gradually have evolved to the later one. Introduction

After 7 years the serious drought condition in Southeast Queensland eventually broke in November 1995. Many bores were drilled during the drought, in particular through the alluvial aquifers and into the underlying sedimentary formations. This trend continues due to a combination of greater extraction and greatly reduced recharge to alluvial aquifers. An example area is presented here, the Teviot Brook catchment, Boonah Shire, Southeast Queensland, where the underlying formation is mainly the Middle Jurassic Walloon Coal Measures (WCM). Geological Characteristics

The Teviot Brook catchment is within the broad Moreton Basin which developed from the Late Triassic to the Late Jurassic. The geology of the basin is comprised of Jurassic sedimentary rocks, the Early Jurassic Marburg Sandstone Formation and the Middle Jurassic Walloon Coal Measures, intruded by various Tertiary igneous rocks as sills, dykes, and plugs. The Walloon Coal Measures (WCM) was deposited underfluvialand lacustrine conditions in a continental environment. It consists of interbedded lithic, and sublabile fine grained sandstone, siltstones, carbonaceous mudstone, claystones and shale, and up to 37 thin coal seams (often calcareous) in different locations, and minor limestone, with occasional beds of bentonite (altered volcanic ash) [(1), (2), (3)]. Petrographically, the WCM appears as dominantly volcaniclastic and most sandstones may be classified as volcanic litharenites or lithic arenites (3). Mineralogical examination shows montmorillonite is the major clay species followed by kaolinite and illite within a matrix of mudstone, shale, coal and fine sandstones (4). The Jurassic Walloon Coal Measures has the potential to be a considerable source of groundwater. Existing drill logs for water and coal exploration show several water-bearing layers exist within the formation. The poorly permeable mudstones act as partial barrier to the movement of groundwater and form semi-confined aquifers (4). The combination of the features outlined above complicated the hydrology and chemistry of groundwater found within the formation. Information on Boreholes

Only limited borehole information was availablefromgovernment sources and most available drill logs are provided by drillers. These logs can be correlated to a certain extent because coal-bearing layers are quite common and can be readily recognisable. As shown in Table 1. the depth of aquifers is quite variable. The top of an aquifer could be at less than 8m immediately beneath the Teviot Brook Alluvium to depths of 49m, and aquifer thicknesses rangesfrom0.7m to 6m with about 2.5m in average. Most of the main aquifers are result offractureporosity, especially within the coal seams and interbeded basalt. Thisfractureporosity is also of significance to discharge rates and water quality (eg. as Conductivity). Lithological data show the main aquifers in the WCM are in horizons of shale, coal, and sandstone, but the overlying hydrogeological units and their topographic location vary. Boreholes 1-2 (Table 1.) are at upstream of Teviot Brook valley, near Croftby, through the alluvium into the WCM formations. They are quite close to each other (about 300m apart) but intersect aquifers at different depth; the aquifers comprise interbedded coal, shale andfracturebasalt. Hydraulic head in borehole 1 is much higher than in borehole 2. Borehole 1 is an only artesian in the area. Boreholes 3-6 are located in the Teviot Brook valley from upstream tc 343

downstream. Boreholes 3 and 4 intersected aquifers immediately beneath Teviot Brook alluvium with some hydraulic connection between two units. Boreholes 5 and 6 are in the town of Boonah, which is often subject tofloodingafter heavy rain. The water bearing-beds are immediately under the weathered zone. Groundwater levels are usually high here and close to ground surface. Boreholes 7-9 are located on eastem side topographically higher areas parallelled the valley. Above the aquifers are thick weathered zones of WCM. Borehole 10 is on a hill of westem side boundary of the catchment. Borehole 11 is at down stream, north to Coulson, on higher area. Above differences in the boreholes locations can be related to groundwater quantity and quality. This is also reflected in chemical evolution of groundwater along its flow path and the type of aquifer material of the formation. The various chemical characteristics that develop can be categorised to certain use of suitable purposes. Sampling and Chemical Analysis of Water Over 100 sites in the Teviot Brook catchment were investigated and sampled. These samples include surface stream water, alluvial groundwater and groundwaterfromthe WCM. A selection of 20 samples are presented, which are believed to be representatives of groundwater from the WCM. In the field, bore samples are taken by bailing, or pumped if the bore is equipped. Depth of groundwater table (SWL) and water quality parameters, such as pH, Eh, temperature, DO (ppm), EC(|iS/cm) and salinity of groundwater were recorded at well head by use of T.P.S. Microprocessor Field Analyser (Model 90FL). Samples were analysed for major ions and minor elements by Inductively Coupled Plasma(ICP) emission spectrometer. The CI is determined by Mohr's titration method. HCO3 ^^ determined by the titration method using hydrochloride acid. Table 2 shows thefieldmeasured water quahty parameters and chemical analysis of major and minor constituents. Some analysed data are not listed due to space, such as DO (ppm), salinity, Fe, Mn, Zn and Al. Results (a) Salinity Field measurements reveal that salinity in the area is variable, and depends on location of bore, depth of aquifer and hydraulic connection to other hydrogeological units. In general, TDS of WCM groundwater ranges from 230mg/L to 7100mg/L and averages 2580mg/L. According to Hart (5), the lowest salinity of WCM groundwater is in class 2 (salinity: 175-500mg/L) as medium salinity water; the highest one in class 5 as extremely high salinity water. Medium salinity water only account for 20% in WCM groundwater. There are 20% WCM groundwater is high salinity water (class 3: 5001500mg/L); others are very high (class 4:1500-3500mg/L) to extremely high (class 5: >3500mg/L) salinity water which are not suitable for agriculture use in the area. It was found that boreholes located on topographic h i ^ have groundwaters that are often more sahne (eg. boreholes 7, 8,10 in Table 1) than those on alluvial flats (eg. boreholes 1, 2,3,4). It is also found that deeper water-bearing units have higher salinity (compare with samples no. JWOl and JW02; JW08 and JW17 in Table 2). (b) Quality of Groundwater in the Walloon Coal Measures In most cases, pH of the WCM groundwater ranges from 6.4 to 7.4 with several more alkaline up to 8.7. The average pH is 7.1. Redox potential (Eh) ranges widely from -247mV to +218mV and averages +13mV. Sandstone aquifer has Eh ranges from +5mV to +21mV, shale and coal aquifers have wider Eh ranges (from -26mV to +218mV). Dissolved Oxygen ranges from 0 to 8.8ppm. Sodium concentrations in sample water ranges from 46mg/L to 2241mg/L and averages 662mg/L. The Sodium Adsorption Ratio (SAR) values are used to evaluate the sodium hazard potential to soils irrigated with waters containing high quantities of sodixim. The SAR values of sample water rangefrom2 to 59 and averages 23. 3 0% sample water show low level sodium hazard (SAR: 0-10); 10% have medium level sodium hazard (10-18), and the rest have high (18-26) to very high sodium hazard (>26). So most of WCM groundwaters are not suitable for irrigation purpose in long term point of view. Magnesium in sample water ranges from 6mg/L to 398mg/L and averages 126mg/L. Animals require magnesium salts in their diet. Certain magnesium salts, particularly the sulphate can cause scouring or diarrhoea in livestock. The suggestion of the upper concentration of magnesium in the drinking water for various animals are: less than 250mg/L for pigs and poultry; horses, cows in milk, ewes with lambs; less than 400mg/L for beef cattle and less than 500mg/L for adult sheep on dry feed (5). The WCM groundwater is suitable for all livestock in most cases. A few farmers dilute the higher magnesium groundwater with rain water or dam water to suit their stock drinking from safety point of view. Calcium concentration in sample water ranges from 7mg/L to 461mg/L and averages 147mg/L. Calcium is essential to livestock. The working level for calcium in stock drinking waters is l,000mg/L, but a level of less than 700mg/L is probably desirable for beef cattle especially if magnesium is also present (5) Calcium together with magnesium contribute to hardness. Hardness of the WCM groundwater ranges from 50mg/L to 2654mg/L and averages 870mg/L. Only 25% sample water have hardness within 0-180mg/L (soft to hard water). 75% sample water have total hardness exceed 180mg/L (very hard level); and over 60% exceed 500mg/L which is not suitable for domestic use (6). 344

Fluoride concentration from sample water ranges from under detection limit (<0.1mg/L) to 2.42mg/L and averages 0.5mg/L. Certain fluoride concentration in livestock drinking water is essential (0.5-2mg/L) (5). Walloon groundwater has generally low fluoride content with only one exception is great than 2mg/L. In brief, due to high sodium, magnesium plus calcium content, WCM groundwaters are mostly not suitable for irrigation and domestic purpose, but suitable for stock especially beef cattle drinking. Otherwise, individual consideration of suitable purpose is needed. (c) Water Chemical Types Groundwaters from WCM in the Teviot Brook catchment are mainly Na-HC03,Cl and Na-C1,HC03 types. Ca, Mg and SO4 ions are secondary contributors. As shown in Table 2, a low (Na+K)/(Ca+Mg) (mg/L) ratio ranges from 0.3 to 1.3 indicate that relatively higher Ca+Mg content is likely associated with sandstone as water-bearing bed (Table 1 and 2). One bore shows that Ca+Mg content related to shale and coal beds (borehole no. 11 in Table 1 and sample no. JWl 1 in Table 2). CI/HCO3 ratios in Table 2 show that 65% WCM waters are Na-HC03,Cl type water (CI/HCO3 ratio is at least 0.6 in mg/L). This type water are mainly distributed on both side topographic highs of the catchment and down stream area; and they also associated with the higher salinity of water in contrast to the others. Sulphate is generally low in WCM groundwater. It has a h i ^ correlation to TDS and CI, however, C1/S04 ratios vary greatly as shown in Table 2. The lower values mean hi^er sulphate content which range from 1.9 to 6.5 (lower than 7.2 of seawater) are strongly associated with shale and coal as water-bearing beds (Table 1; Figure 1). Thus, water from sandstones and basalt have higher C1/S04 ratio and low in sulphate content. (d) Ratios, trends, groups in relation to WCM groundwater evolution Na/Cl equivalents ration of WCM groundwater tends to decreasing as Mg+Ca (meq/L) is increasing. Two groups of WCM groundwater has been recognised; group I has Na/Cl (meq/L) ratio higher than that of head water (1.76); group n has Na/Cl (meq/L) ratios changing from above that of seawater (0.85) to below it along with the increasing of Mg+Ca (meq/L). This occurs from upstream recharge area (samples group 1.) towards downstream higher salinity waters gradually (samples group 11.) as shown in Figure 2. Group I waters are Na-C1,HC03 type; and group II are Na-HC03,Cl type. Samples no. JW03, JW04 and JW09 show mixing of WCM groundwater with overlying alluvial aquifer water. Discussion Sodium could be present as an impurity in the cementing material in sedimentary rocks. It is also present in sodium feldspars of volcaniclastic and other minerals, which is more susceptible to attack by solution and the consequent release of the cation than potassium feldspars. Upward leakage of saline water from deep aquifers, where the salinity can be either connate (fossil seawater) or evaporitic may inject sodium into essentially freshwater aquifers. When sodium dissolves, it tends to remain in solution. There are no important precipitation reactions to reduce its concentration, althou^ cation exchange can alter the concentration of sodium (relative to other cations) (5). The concentrations of CI ion in most surface waters are generally lower than those of sulphate or bicarbonate. Since igneous rocks, at least those available for sampling and analysis, cannot yield very high concentrations of CI to normally circulating natural water (6), Zahawi(7) suggested that the source of CI (in the Walloon Coal Measures) could therefore be attributed to the adsorption of the ion to the rocks of the Bundamba Group by the transgression of the sea water. The adsorption of the CI ion to the rocks takes place when porous rocks after their formation are submerged by the sea. The rocks then become impregnated with soluble salts which are mainly of the CI type. The WCM groundwater are generally high in Mg/Ca (meq/L) ratio; 75% of saddles have Mg/Ca (meq/L) ratio exceeded unity, in the case of low Na/Cl (meq/L) ratio (>1) indicates that Mg could be released by weathering of the Mg-rich minerals in the volcanic litharenites and basalt. The presence of Ca and HCO3 ^ WCM groundwater are because of the calcite dissolutionfromcarbonaceous mudstone and often calcareous coal seams and minor limestones in the rock units. Strontium has similar chemistry to that of calcium and its carbonate (strontianite) and the sulfate (celesite) are common in sediments. But, strontianite is less soluble than calcite (6). High Sr may require substantial water-rock contact times. Sr and Na exchange is almost certainly important as evidenced by the unusually high Sr concentration in low-Na water and general similarity of the Sr and Ca vs. CI plots (Figure 3). Lithium typically increases rapidly relative to CI and Na during the early stages of infiltration, in response to the release from clay and rock forming minerals. The increase in Li/Cl is often proportional to residence time as demonstrated in flow through confined aquifers. After long residence times (e.g. 10^-10^ yr) a typical Li/Cl ratio is reached for a particular formation, which signifies a new equilibriiun between water and rock (8). In WCM groundwater, Li is positively correlated to TDS, CI and Na. It appears to be affected by ion exchange reactions being greater in high-Na water and lower in the low-Na waters. Concentrations are in general high in comparison with seawater as shown in figure 4. This 345

may be due to the ease with which Li is displaced from exchange sites by Na and K by adjustment of Mg-rich clays by rejecting Li (8). Sources in the formation could include kaolinite, montmorillonite and feldspars. Fluoride concentrations are not easy to explain, because it does not show any clear pattems of variation. Its source could be fluorite (CaF2) occurring in both igneous an sedimentary rock. It has the poorest correlation with all the elements. It is interesting that fluoride always has opposite correlation coefficient value with Si02 to the others, ie, it has positive value to the others while the Si02 has a negative value. Fluoride has positive correlation to pH range mostly between 67.5. It also has positive correlation to Na, K, Li, and HCO3, SO4; however, it has negative correlation to Ca, Mg, Sr and CI, even if they are very poor. Si02 concentration is associated with Main Range Volcanics. It is positively correlated with Ca, Mg, and Sr; and negatively correlated with Na, K, CI and pH. Silica is dissolved during recharge (4). Conclusions The WCM groundwater chemistry reveals the fact of high Na concentration (accounts for 49%-99% of total cations). This has resulted high sodium absorption ratio of WCM groundwater mostly not to be suitable for irrigation purpose. Overall cations contents increase along with the increasing salinity level, so accompanied high Ca and Mg concentration result high total hardness of most WCM groundwaters, which are not suitable for domestic use. Most WCM groundwater are suitable for livestocks drinking with no fluoride hazard, but several boreholes produce extreme high level salinity water (TDS>3500mg/L) may need careful considerations. Chloride concentration generally accounts for 26%-48% of total anions as seconded to HCO3 ion; and 65%-87% as major anions in salt. It tends to increase from upstream towards down stream along the Teviot Brook catchment and thus alters the water chemical types. Whitehouse (9) suggested seawater transgression into the Moreton Basin in the early Aptian (early Cretaceous), but direct evidence of the exact area is not known due to postAlbian uplift and erosion (10). It is likely that lithology of WCM has influenced groundwater chemistry. The higher Mg+Ca content in WCM groundwater is associated with sandstone horizons. While SO4 content appears higjier in shale and coal horizons. An increasing Mg+Ca (meq/L) down flow path results decUning in Na/Cl ratio to even much lower than that of seawater, plus Mg/Ca equivalent ratio exceeded unity in majority of WCM groundwater, this may indicate extemal source of Mg related to seawater as parallel to the source of CI. The concentration of trace elements Sr and Li in WCM groundwater are generally high which indicate long residence time of groundwater in WCM aquifers (transmissivity of aquifer of borehole no.l is L6mVd). Most species have reacted with the clay minerals in WCM, therefore, limited value in indicating the source of salinity at this stage. However, overall trend is likely towards assumption of the seawater origin. Further works are needed to be done. Acknowledgment We thank the following people and organisations from the bottom of our heart: all those enthusiastic property owners in the study area we have met during field investigations; the Water Resources Commission of Queensland Dept. of Resource Management; all staff of the Geochemical Laboratory of School of Geology, QUT; the R & K Harch Drilling; and the Landcare group of local community, Boonah. References: 1. Cameron, J.B.: The Rosewood-Walloon Coalfield. Geological Survey of Queensland. Publication 344, 1970. 34pp. 2. Cranfield, L.C.; L.J.Hutton & RM.Green.: Ipswich Sheet 9442, Queensland 1:100000 Geological Map Commentary. Queensland Dept ofMines. 1989. 56pp. 3. Fielding, C.R.: Sedimentary environments of the Middle Jurassic Walloon Coal Measures in the Rose-Walloon Coalfield, Southeast Queensland. Ninth Australian Geological Convention. Geological Society of Australia Abstracts Number 21. 1988. ppl35-136. 4. Huxley, W.J.: The Hydrogeology, hydrology and Hydrochemistry of the Condamine river valley alluvium. Vol. 1 (Text). M.S'c. thesis, Queensland Institute of Technology. 1982. 600pp. 5. Hart B.T,: A Compilation of Australian Water Quality Criteria. Australian Water Resources Council TechnicalPaper No 7. Australian Government Publishing Service, Canberra 1974. 310pp. 6. Hem, J.D.: Study and Interpretation of the Chemical Characteristics of Natural Water. 3rd ed. United States Geological Survey Water-Supply Paper 2254. 1985. 264pp. 7. Zahawi: Report on the Hydrogeology in Lockyer Valley. Geological Survey of Queensland. 1975. 120pp. 8. Edmunds, W.M.: Characterisation of Groxmdwaters in Semi-Arid and Arid Zones Using Minor Elements, in Groundwater Quality. Edited by H.Nash and G.J.H. McCall. Chapman & Hall. 1994. ppl9-30. 9. Whitehouse, F.W.: The Geology of the Queensland Portion of the Great Artesian Basin. Appendix G in Artesian Water Supplies in Queensland. Dept of the Co-ordinate General ofPublic Works, Government Printer, Brisbane. 1955. 20pp. 10. Day, R.W., W.G.Whitaker, C.G. Murray: Queensland Geology. A Companion Volume to the 1:2,500,000 Scale Geological Map (1975). Geological Survey of Queensland Publication 383. 1983. 194pp. 346

Discharge Rate (Us) (gph)

Aquifer Thickness (m)

Aquifer Lithology

Nature of Porosity

Standing Water Level (m)

53.0

4.0

coal, mudstone, basalt

Fractures

+3.52

1.06

840

24.0

28.2

4.2

black shale + coal + basalt

Fractures

-15.6

2.78

2200

B107

7.7

8.5

0.8

sandstone

Semi-Consolidated

-4.74

- -

B109

12.3

13.9

1.6

sandstone

Consolidated

-5.11

- -

1980

Aquifer Depth (m) Bottom Top

No.

Name of Bore

49.0

3 4

B15

12.0

14.0

2.0

high weathered sandstone

- -

-0.85

2.50

B16

16.0

18.0

2.0

shale + coal

-0.75

2.00

1584

15.0

17.5

2.5

coal

Fractures

-4.5

1.26

1000

NFwbl

20.4

21.9

1.5

black shale + coal

-19.5

0.25

200

wb2

28.5

1.5

black shale + coal

- -

0.51

400

10.8

11.1

0.3

hard shale

- -

-8.1

1.14

900

sew

10.5

12.6

2.1

coal

-0.6

2.53

2000

IG wM

15.0

15.7

0.7

black shale

Fractures

0.13

100

0.32

250

0.38

300

1.24

980

wb2 wb3 Average:

26.7 45.0

32.7 51.0

6.0

black shale + coal

Fractures

-4.8

6.0

black shale

Fractures

2.5

Table 1. Lithological and Hydrogeological Information of Groundwater Aquifers in the Jurassic Walloon Coal Measures.

Temp.

Bore Name

Well Depth (m)

EC (US/cm)

JW01

TS (1)

1764

JW02

MR (2)

631

Sample No.

so4

HCO3

156

7.5

14.0

°C

6.8

24.0

331

4.6

8.4

18.8

100

2.3

Ions Ratios Cl/HCOj (Na+K)/(Ca+Mg)

TDS

C1/S04

0.48

889

20.8

0.2

5.1

0.24

307

4.1

0.2

3.2

Si0 2

654

1.24

0.064

246

0.08

0.009

842

6.9

19.1

1.1

121

15.7

280

0.45

0.009

0.35

454

7.7

0.6

B107 (3)

0.4

JW03

947

6.8

19.3

0.8

106

4.3

335

0.49

0.003

0.25

488

24.5

0.3

B109 (4)

0.3

JW04

7400

6.8

21.3

355

252

804

2.5

2141

210.0

567

5.90

0.086

0.22

4032

10.2

1.3

B15(5)

3.8

J WO 5

255

200

920

2.3

1978

220.0

638

3.50

0.049

0.29

3872

9.0

3.1

2.0

JW06

B16 (6)

7200

6.9

20.6

6050

6.8

20.0

105

131

1080

8.0

1446

184.8

855

3.65

0.227

0.75

3323

7.8

4.6

JE(7)

1.7

JW07

12830

6.9

19.4

307

398

1470

20.6

3765

575.8

817

8.60

0.497

0.65

6891

6.5

2.1

NF (8)

4.6

J WO 8

3.9

22.5

123

0.27

0.006

<0.1

230

1.9

0.3

1.7

158

168

2241

23.9

3545

633.5

780

9.39

1.084

0.39

7111

5.6

4.5

6.9

JW09

KW (9)

502

6.7

20.7

JW10

SCW (10)

12390

6.7

21.0

3300

6.9

19.7

192

130

268

0.7

674

116.1

451

2.43

0.074

0.55

1613

5.8

0.8

IG(ll)

1.5

JW11

6.4

22.2

461

366

731

15.1

3332

304.4

789

18.20

0.221

<0.1

5582

10.9

4.2

0.9

JW12

McLn

10470

5500

6.6

20.0

146

132

832

9.8

1383

108.2

718

6.54

0.135

0.43

2950

12.8

3.0

PDLC

1.9

JW13

7840

6.4

20.1

294

197

1012

9.5

2113

169.2

709

12.10

0.081

0.16

4105

12.5

2.1

WnTrf

3.0

JW14

7.4

23.7

111

165

285

4.1

730

15.9

539

2.10

0.058

0.31

1570

46.1

1.4

1.0

3260

13.5

794

1.57

0.348

2.42

2104

63.8

1.1

6.4

RK1

JW15 JW16

RK2

4230

8.2

22.6

731

4.4

858

6310

7.4

23.3

125

133

1034

7.2

1404

185.1

803

2.31

0.289

2.20

3241

7.6

4.0

NFw

1.7

JW17

SRMN

7.2

20.9

361

2.2

135

4.2

658

0.67

0.122

0.81

824

32.1

0.2

JW18

1560

10.5

72 72

1623

8.7

20.2

378

4.0

203

6.6

564

0.36

0.082

0.50

860

30.7

23.7

RWLD

0.4

JW19

2135

8.2

19.0

482

4.1

343

8.0

617

0.53

0.080

0.70

1125

43.0

23.0

JWLD

0.6

JW20

4839

7.1

20.8

141

126

662

6.5

1227

141.0

597

4.02

0.176

0.59

2578

18.2

1.8

5.2

20.6

1.3

2.1

123

0.13

0.002

<0.1

17C1

10.0

0.2

0.6

413

1290

10800

399.0

19500

2700.0

142

7.60

0.170

1.30

3500C1

7.2

137.3

6.6

Mean Value Head Water

303

8.0

Sea Water a. All units expressed in mg/L, unless otherwise indicated.

c. Values in border italic font indicate the maximum value. d. Values in light italic font indicate the minimum value.

Table 2. Chemical Analysis of Major and Minor Constituents of Groundwater Samples from the Walloon Coal Measures.

Figure 1. Plot of correlation of SO4 vs. CI concentration (mg/L) with reference to seawater and headwater. The higher SO4 water are associated with shale and. coal beds

10000 seawater 1000

O) E

higher SO4 water

100

S (0

lower SO4 water

10 ^

headwater

100

1000

10000

100000

Log CI (mg/L)

Figure 2. Plot of correlation of Na:Cl ratio vs. Ca+Mg with reference to seawater ratio and headwater ratio. Two groups are recognised. Group I waters are distributed at upstream area of the catchment. Group II waters are at down stream area and/or on topographic highs. The trend Une indicates WCM groundwater evolution along flow path. The increasing Ca+Mg result the decreasing in Na:Cl ratio.

10.0 seawater Na:CI = 0.85

) S

1.0

headwater Na:CI = 1.76 0.1

0.1

1 10 Log (Ca+Mg) (meq/L)

100

Figure 3. Plot of correlation of Sr vs. CI (mg/L). The concentration of Sr is positively correlated to CI. Overall concentration of Sr is higher than that of seawater. Group I waters are at upstream area; and group 11 waters are at down stream area.

100.0

d 10.0^

Ik. (0 q

1.0

0.1

100

1000 10000 Log CI (mg/L)

100000

Figure 4. Plot of correlation of Li vs. CI (mg/L). The concentration of Li is positively correlated to CI; and generally high in comparison with seawater. Group I waters are at upstream area group II waters are at the topographic medium high area, and group III waters are at down stream area or on topographic highs.

10 1 -

O) E

head water

0.1 -

O) o 0.01 0.001

—Tf' 10

100

1000 10000 Log CI (mg/L)

100000

349

PETROLEUM SYSTEMS OF THE BOWEN, SURAT AND GUNNEDAH BASINS T.S. LoutitS RJ. KorschS JM. TotterdeiF, CJ. BorehamS A.T. BrakelS A.T. WeUsS M,G. NicolF & R.D. Shaw^ 1 Marine Petroleum and Sedimentary Resources Division, Australian Geological Survey Organisation, GPO Box 3 78, Canberra, ACT 2601 2 Vanibe PtyLtd, 5A Mulbring Street, Mosman, NSW2088 Summary The Sedimentary Basins of Eastern Australia (SBEA) project is now complete. The study, a joint venture between the Geological Survey of Queensland, the Geological Survey of New South Wales and the Australian Geological Survey Organisation, completes Phase I of the petroleum studies in the National Geoscience Mapping Accord. Unlike the previous studies in Phase I, the SBEA project has undertaken an evaluation of the petroleum potential of part of the region. The evaluation was carried out to illustrate the usefuhiess of the geological framework produced by the project and to ensure that the basin analysis methods employed produced the appropriate data for petroleum assessment. The results of the petroleum evaluation suggest that significant volumes of hydrocarbons have been produced in the Bowen and Surat Basins. The bulk of the hydrocarbons generated after 140Ma and most of the generation occurred in the early late Cretaceous. As expected, the estimated volume of the hydrocarbons generated far exceeds the volume of discovered hydrocarbons, suggesting that preservation of accummulations may be the main risk factor in the region. However, the locations of the areas of maximum generation within discrete drainage divides or migration cells provide interesting insights into the prospectivity of a large proportion of the Bowen and Surat Basins, particularly the eastem margin. Introduction The petroleum systems of the Bowen, Surat and Gunnedah Basins have been documented as a result of four phases of basin analysis consisting of Province and Terrane Analysis, Regional and Basin Analysis, Play Element Evaluation and Play and Petroleum Systems Definition. Much of this work has focused on the structural and sequence stratigraphic mapping of a regional grid of seismic data, firstly to define the interplate and intraplate tectonic events that have helped to create the accommodation space (e.g. Korsch and Totterdell, 1995; this volume), and secondly to define the stratal geometry of the sedimentary units (e.g. Totterdell et al., 1992; Wells et al., 1994; Totterdell et al., 1995). Thus, the Sedimentary Basins of East Austraha study has helped to define the shape of the containers within which organic-rich rocks (ORRs) have been deposited, the timing of deposition of the ORRs, the character of each ORR, the timing of generation, the products generated, identified potential migration paths, and has documented the timing and distribution of seal and reservoir rocks. The location and style of traps has not been documented thoroughly except in the central Bowen and Surat Basins. This paper documents briefly the approach and the steps taken to evaluate the petroleum potential of the region. East Australian Petroleum Super Systems Bradshaw (1993) and Bradshaw et al. (1994) defined seven petroleum supersystems that have developed in Australia during the Phanerozoic. Two of these supersystems are present in the basins of eastem Australia. The Gondwanan Supersystem is generated from siliciclastic sequences of Late Carboniferous to Middle Triassic age deposited in high southerly latitudes. The Murta Supersystem is generated from Late Triassic to Cretaceous siliciclastic, dominantly nonmarine, sequences deposited at middle to high southerly latitudes. The Gondwanan Super System is the only supersystem present in the Bowen and Surat Basins. In other words, no petroleum was generated from the Murta Supersystem. The concept of the Supersystem was coined by Bradshaw et al. (1994) to describe the widespread deposition of similar organic-rich rocks in a number of basins in Austraha during discrete time periods. Bradshaw recognised that each supersystem consisted of a number of systems that would be defined with more detailed work in each basin. Here, based on our recognition of several source-rock units (see below), we are able to define six petroleum systems for the Bowen, Gunnedah and Surat basins. The effectiveness of the organic-rich rocks in the central Bowen Basin has been documented within drainage divides or source kitchen areas to define the petroleum systems of the region. The equivalent work in the northern and southem Bowen Basin and the Gunnedah Basin has not been undertaken and will only be discussed briefly at the conference. Bowen-Surat Petroleum Systems Boreham et al. (this volume) have documented six effective source-rock units within the Gondwanan Supersystem. The effective units (and their equivalents) are the Moolayember Formation, the Baralaba Coal Measures, the Burunga Formation, the Banana Formation, the Flat Top to Buffel Formations and the Reids Dome beds. The Baralaba and the Burunga have generated over 90% of the oil in the central Bowen Basin and the same two units plus the Flat Top-Buffel 350

and Banana Formations and equivalents have generated about 90% of the gas in the region (Figures 1 and 2). Each of these units has generated hydrocarbons in discrete structurally-defined regions of the Bowen Basin (Figure 3). Each of these source-rock units forms the basis of a separate petroleum system (e.g. Figure 4) that has now been evaluated in the central part of the Bowen and Surat Basins (Shaw, 1996). The information on the timing and products of generation coupled with an estimate of the relative yields of the more effective ORRs will allow the SBEA project team to look at the potential for new plays in the region. We consider that the present day structure is a reasonable representation of the geometry of the central Bowen Basin at the time of generation during the early Late Cretaceous. The distribution and character of reservoir and seal units and traps will be overlain on potential migration pathways from the centre of the basin particularly on the relatively unexplored eastem margin of thebasin. The timing of the emplacement of trap, reservoir and seal play elements relative to the timing of generation will be documented on timing charts for the region. Project Methodology The SBEA project, similarly to other projects within the NOMA, was designed to improve our understanding of the effect of geologic processes on resource systems. The project defined a set of questions about specific processes. These basic questions reflect our current understanding or conception of how geologic processes form, deform and destroy basins and the impact of those processes on petroleum and mineral systems. These concepts can be expressed as a set of rules either graphically (Figure 5) or in written form (see below). The basic rules or concepts used by the SBEA project are: • Interplate events produce intraplate deformation. • Intraplate deformation creates accommodation space for sediments. • Intraplate deformation controls trap timing and style. • Intraplate deformation may control timing of fluid movement. • Accommodation space is modified by sea-level and climate processes. • Accommodation space controls stratal geometry within local palaeogeography. • Stratal geometry controls the gross distribution and timing of source, reservoir and seal. The emphasis in the project has been on certain aspects of Province and Terrane Analysis, Regional and Basin Analysis and Play Element Evaluation (Figure 6). More recently we have begun to look at Play Element Evaluation and Definition in the central Bowen Basin in more detail. Conclusions The SBEA project has produced a consistent tectonostratigraphic framework over a large region of eastem Austraha that will form the basis for regional evaluation of the petroleum systems in the Bowen, Surat and Gunnedah Basins. A more detailed study of the central Bowen and Surat Basins suggests that there are at least six petroleum systems operating but that only two or three are producing significant volumes of hydrocarbons. The results suggest that there are a number of unexplored areas within the central Bowen and Surat Basins that may have seen significant volumes of hydrocarbons. Preservation of any accummulations may be the major rsik factor in the region. Acknowledgements We thank Robert Dabrowski and David Lund for help with the map production and with some of the data compilation and figures. PubUshed with the permission of the Executive Director, Australian Geological Survey Organisation. References Boreham, C.J., Korsch, R.J., and Carmichael, D.C., this volume. The significance of mid-Cretaceous burial and uplift on the maturation and petroleum generation in the Bowen and Surat Basins, eastem Australia. Proceedings of the Mesozoic geology of the Eastem Australia Plate conference. Bradshaw, M.T., 1993 Australian petroleum systems. PESA Journal, 21,43-53 Bradshaw, M.T., Bradshaw, J., Murray, A.P, Needham, D.J., Spencer, L., Summons, R.E., Wilmot, J. and S. Winn, 1994. Petroleum systems in western Australian basins. In: Purcell, P.G. & R.R. (editors). The sedimentary basins of Westem Australia. Proceedings of Petroleum Exploration Society of Australia Symposium, Perth, 93-118. Korsch, R.J. and Totterdell, J.M., 1995. Structural events and deformational styles in the Bowen Basin. In: Follington, LW., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995.... 150 years on .... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 27-35. Korsch, R.J. and Totterdell, J.M., this volume. Mesozoic deformational events in eastem Australia and their impact on onshore sedimentary basins. Proceedings of the Mesozoic geology of the Eastem Australia Plate conference. Loutit, T.S. 1996. Australian petroleum research and development: a problem driven approach to geoscience research management. APPEA Journal, 36, 500-515. Magoon, L.B. and Dow, W.G., 1994. The petroleum system. American Association of Petroleum Geologists, Memoir, 60, 3-24. 351

Shaw, R.D., 1996. An estimation of undiscovered hydrocarbon resources using Sedimentary Basins of Eastem Australia project data. Confidential Report prepared for Australian Geological Survey Organisation by Vanibe Pty Ltd, unpublished. Totterdell, J.M., Wells, A.T., Brakel, A.T., Korsch, R.J. and NicoU, M.G., 1992. Sequence stratigraphic interpretation of seismic data in the Taroom region, Bowen and Surat basins, Queensland. Bureau of Mineral Resources, Australia, Record 1991/102, 61 pp. Totterdell, J.M., Brakel, A.T., Wells, A.T. and Hofifinann, K.L., 1995. Basin phases and sequence stratigraphy of the Bowen Basin. In: Follington, I.W., Beeston, J.W. & Hamilton, L.H., (editors), Bowen Basin Symposium 1995.... 150 years on .... Proceedings. Geological Society of Australia, Coal Geology Group, Brisbane, 247-256. Wells, A.T., Brakel, A.T., Totterdell, J.M., Korsch, R.J. and Nicoll, M.G., 1994. Sequence stratigraphic interpretation of seismic data north of 26°S, Bowen and Surat Basins, Queensland. Australian Geological Survey Organisation, Record, 1993/51, 25 pp.

CONTRIBUTIONS OF OIL BY SOURCE ROCK INTERVAL 94.6% of oil generated post 141 Ma

70% n 60%-

50%OIL

40%30%-

IllBli

• Reids Dome • Buffel - Banana H Burunga

20%-

10%-

• Moolayember

0% STRATIGRAPHIC UNIT 352 Figure 1. The relative contributions of oil by source rock interval in the central Bowen Basin

CONTRIBUTIONS OF GAS BY SOURCE ROCK INTERVAL 91 A% of gas generated post 141 Ma • Reids Dome • Buffel - Banana m Burunga H Baralaba • Moolayember STRATIGRAPHIC UNIT Figure 2. The relative contributions of gas by source rock interval in the central Bowen Basin. 148

Relative oil yield per drainage cell

• • High Medium % Fields and prospects

Low

Figure 3. Illustration of the relative oil yield per drainage cell in the central Bowen Basin, the fields in the region and a number of prospects (Shaw, 1996) 353

Baralaba Petroleum System 300 1 1

200 1

100 1

Mesozoic

Palaeozoic Permian

Triassic

Jurassic

Cainozoic Cretaceous

Tertiary

Geological time/"''^ scale ^ / P e t r o l e u m system elements Source

Figure 4. Diagrammatic representation of the Baralaba Petroleum System for the central Bowen and northem Surat Basins in Queensland (methodology after Magoon and Dow, 1994).

Figure 5. Illustration of a suite of concept-based rules that have been tested by the SBEA project.

354

SBEA

Play Element Evaluation Play DeTmltion

and Evaluation

Petroleum Systems Definition and Evaluation

Figure 6. Illustrates major Systematic Approach to Basin Resource Evaluation (S.A.B.R.E.) tasks (Loutit, 1996) applied to the SBEA project. Darker shading shows that the main emphasis of the SBEA project was on Province and Terrane Analysis, Regional and Basin Analysis and Source and Maturation Evaluation.

355

GEOLOGICAL CONTROLS ON HYDROCARBON ACCUMULATIONS IN THE EROMANGA BASIN, SOUTH WEST QUEENSLAND - A PETROLEUM SYSTEM PERSPECTIVE B.S. Lowe-Young, SJ. Mackie and RS. Heath Santos Ltd, Brisbane, Queensland, Australia Summary The Eromanga Basin, located in central Australia, has been established as the most prolific onshore oil province in Australia. Hydrocarbon accumulations occur in reservoirs ranging from Late Cretaceous to Early Jurassic in age within structurally controlled anticlinal traps. Approximately 190 mmstb of recoverable oil with minor amounts of gas have been discovered in the basin to date. The majority of accumulations are characterised by proximity to underlying Cooper Basin source rocks of Permian age, and the occurrence of multiple reservoir pools. However, the processes of migration and entrapment of hydrocarbons from source to reservoir are controlled by a complex interplay of geologic factors which varies across the basin. Investigation of geologic controls within a petroleum system context provides a new perspective on understanding the synergy required between key components of the system for significant accumulation of hydrocarbons. Pre-Tertiary structural trap development, effective lateral and vertical migration of hydrocarbons and post-Tertiary preservation of accumulations are identified as major geologic controls on hydrocarbon accumulations in the Cooper-Eromanga petroleum system in south west Queensland. Introduction The Eromanga Basin, which is Early Jurassic to Late Cretaceous in age, is located in Central Australia. The basin unconformably overlies numerous Palaeozoic basins and extends over 1 million km2 (Figure 1). The formation of the Eromanga Basin has been attributed to crustal downwarping and intracratonic sag with the major structural elements of the basin influenced by underlying palaeostructural features (Figure 2). Deposition of Eromanga Basin sediments occurred in response to gradual basin subsidence, with major depocentres located over older trough areas. Petroleum exploration in the Eromanga Basin has established the central portion of the basin, which overlies the Permian Cooper Basin, as the most prolific onshore oil province in Australia. Since initial discoveries of oil in Eromanga Basin sandstones at Strzelecki, South Australia in 1978 and at Jackson, Queensland in 1981, approximately 190 mmstb recoverable oil have been discovered in the central area of the basin, with minor amounts of sales gas. Sedimentation within the Eromanga Basin occurred uninterrupted from the Early Jurassic to Late Cretaceous period. The Jurassic to Early Cretaceous section of the basin comprises continental fluvial and lacustrine sandstones interbedded with siltstones, mudstones and minor coal (Figure 3). A major marine transgression, following deposition of the Cadna-owie Formation, resulted in the widespread deposition of Early Cretaceous mudstones, shales and shallow marine sediments. The Late Cretaceous period is represented by a regressive sequence of fluvio-deltaic and lacustrine sediments. Several Tertiary tectonic events, following deposition of the Late Cretaceous Winton Formation, resulted in uplift and erosion of some of the Eromanga Basin section. Tertiary compression resulted in the fomiation of northeasterly anticlinal trends throughout the basin. Detailed stratigraphy and structural history of the Eromanga Basin have been documented in previous studies by Green et al.(l), Hofiftnan (2), Moore (3) and Wecker (4). Previous investigations of hydrocarbon source, migration, depositional systems, and tectonic evolution have provided aframeworkfor understanding major controls on petroleum accumulations in maturely explored areas of the Eromanga Basin. However, application of existing geologic models to recent discoveries in less explored areas of the basin requires a more comprehensive understanding of the interplay between major controlling factors. This paper applies an integrative approach to reviewing and examining controlling factors with respect to historically significant and recent discoveries in the Eromanga Basin in south west Queensland. Although significant oil discoveries have been made in South Australia, this study is restricted to examination of accumulations in the south west Queensland sector of the basin which contains approximately 55% of the volume of oil discovered in the basin. Investigation of geologic controls within a petroleum system context provides a practical framework for analysis of synergistic relationships between key components of the system.

356

Northernlern^tojy^ _ South Justralia ^

ADAVALE BASIN

j. ^.^.'.L.-. PEDIRKA BASIN

WARBURTON BASIN

\ . ^

f x

COOPER BASIN

:l\

Ballera, ^ Ballera^

//)""]

/ /

y '

' /

-28

\ —J

V ARCKARINGA RASTN

Aloomba •

•\Jackson

}

. \

28M '^eensland South'Wal^

Figure 1. Location of Eromanga Basin showing underlying Palaeozoic basins. Hydrocarbon Habitat Hydrocarbon accumulations in the Eromanga Basin are characterised by the common occurrence of multiple, vertically stacked reservoir pools in structurally controlled anticlinal traps. The majority of discoveries are located along the Jackson-Naccowlah structural high and southem edge of the underlying Cooper Basin, marginal to the Nappamerri Trough. Oil fields range in size from the Jackson field in south west Queensland, which contains an estimated 40 mmstb recoverable oil, to small single well fields containing less than 0.100 mmstb. Oil remains the dominant hydrocarbon type with minor occurrences of gas accumulations which are generally small in size.

357

140®

1 144®

142« \

EROMANGA

Queensland

BASIN

APPROXIMATE PERMIAN ZERO EDGE

-:: Naccowtsth

26®

•

28®'

100km

Figure 2. Major structural elements of the Cooper and Eromanga basins. Productive reservoirs have been proven to occur in all stratigraphic units within the Early Jurassic to Late Cretaceous Cadna-owie section (Figure 3). Eromanga Basin reservoirs are predominantly fluvial in origin, with the Jurassic Hutton Sandstone regarded as the most productive reservoir unit. Minor reservoirs are present in all other Jurassic formations. Nearshore and fluvio-deltaic sediments belonging to the Early Cretaceous Mooga Formation Murta Member and the Cadna-owie Formation, however, have formed economic reservoir pools in approximately 50% of fields as indicated by Wecker (4). The geographic extent of commercial hydrocarbons in the Eromanga Basin has been extended by the discovery of oil in several recent ejq)loration wells located outside of the previously proven hydrocarbon area. The most significant new discovery, the Inland field is located along the northwestem flank of the Windorah Trough, 70 km north of the northem edge of the Cooper Basin (Figure 4). In addition, discoveries in the Keleary-Telopea area in South AustraUa, and in the Bargie area in south west Queensland, located close to the underlying Cooper Basin edge, suggest that significant undiscovered hydrocarbon potential exists in less explored areas of the Eromanga Basin. The widespread occurrence of Eromanga Basin hydrocarbon accumulations in south west Queensland demonstrates the inherent complexity of factors which govern the hydrocarbon habitat of reservoirs across the basin. Petroleum System Perspective Geologic models developed to explain hydrocarbon accumulations in Eromanga Basin reservoirs have favoured a Permian source for the hydrocarbons, with proximity and vertical access to Permian source rocks identified as a major controlling factor by Heath et al.(5). However, source rock and oil biomarker studies conducted by Powell et al. (6) suggest that oil reservoired in Murta pools has been primarily sourced locally from Murta source rocks, whilst Jurassic biomarkers have been identified in oil samples from several wells north of the Nappamerri Trough area in studies by Hawkins et al. (7) and Jenkins (8). Althou^ it is recognised that Eromanga sourced oil has likely contributed to some accumulations, the existence of oil accumulations sourced solely from Jurassic rocks has been difficult to establish due to the lack of a definitive Permian biomarker. The majority of oil reservoired

358

AGE (HARLAND ET A L , 1989]

STRATIGRAPHY

S.W. (COOPER BASIN AREA,N.E

90-

SOURCE RESER VOIR

SEAL

PROD. ZONES

DEPOSITIONAL ENVIRONMENT

Fluvial - Lacustrine

100 -

110 120-

CO <

130^

140-

<ffl O

cc Lil

Shallow Marine

2? uia. o

Marine I CADNA-OWIE FM.

'-UcusMne

Fluvial - Paludal Fluvial BIRKHEAD FM.

Fluvial - Lacustrine

HUnON SST.

Fluvial

160-

180-

190200-

Fluvial

150-

170-

Stioreface

Fluvial - Shallow Marine

oCO

Q Q

s CE

O <C

POOLOWANNA FM.

BASAL JURASSIC

Fluvial - Paludal

210-

220-

230CO < m DC LU Q.

o o o

240250260-

Permian Source Rocks:

270-

Lacustrine Environments

Ruvial - Paludal-

280Minor Oil

Major Oil IwELAIFM v w MwERRM w. WZ Figure 3. Composite stratigraphic column showing major elements of the Cooper-Eromanga petroleum system. in Eromanga reservoirs is considered to have a Permian origin with multiple charging of reservoirs from both Permian and Jurassic source rocks postulated by Gilby et al. (9) for oil accumulations in which a Jurassic source affinity has been identified. Since it has been estimated by Jenkins (8) that over 80% of the volume of oil in Eromanga reservoirs has been

290-

359

sourced from Permian source rocks, investigation of the relationship between Cooper Basin source rocks and Eromanga reservoirs in the context of a petroleum system (Cooper-Eromanga system) provides a practical framework for investigation of geologic controls on Eromanga hydrocarbon accumulations across the basin in south west Queensland. Examination of an Eromanga sourced petroleum system is not considered within this study. Spatial Characteristics of Cooper-Eromanga Petroleum System Eromanga Basin hydrocarbon accumulations in south west Queensland extend from the Inland field southward to the Munro field, and northeast from Munro to Bargie (Figure 4). Accumulations occur throughout the Cretaceous Cadnaowie to Basal Jurassic section, which varies between 1330 feet and 3000 feet in thickness across the Cooper-Eromanga system area (Figure 5). The common occurrence of vertically stacked Eromanga reservoirs throughout the system indicate that although vertical migration has been eflFective through significant thicknesses of section, local seals are competent enough to support multiple reservoir zones within a single field. The functional description of lithologies within the Eromanga stratigraphic column is summarised in Figure 3. Due to the widespread deposition of Early Jurassic braided stream sediments, the Hutton Sandstone is regarded as the major reservoir of the system in all areas. Over 50% of the volume of oil discovered in the Eromanga Basin is reservoired in the Hutton Sandstone. The development of reservoir sandstones in the minor fomiations is locally controlled by depositional trends and lateral facies variations. Reservoir quality generally decreases with depth, with more significant loss of primary porosity due to compactional and diagenetic effects in the northem part of the study area. The spatial distribution and geometry of reservoir sandstones within individualfieldsis primarily controlled by thefluvialarchitecture of each reservoir unit. 144® 142® 140® • Oil Field d Oil Recovery / Show

EROMANGA

APPROXIMATE PERMIAN ZERO EDGE

BASIN

Morney q Inland • Cuddapan Q 26®-

COOPER BASIN

Queensland South Australia

26®

O Marengo • Cook Keleary^ Telopeaj9

BodallaSth^

Bargie / • Kenmore TintaburnT'"* Tarbat • • Talgeberry Toobunyah

Naccowlah Sth « ) Chookoo • # JacJso/7 Jarrarm ^^ JkThungo •'V •Maxwell Patroclus ( Tickalara Munro

28® 100km

144® 142® -L Figure 4. Geographic distribution of major oil fields and significant oil occurrences in the Cooper-Erommiga petroleum system, south west Queensland. 140®

360

MUNR01

Feet -2000'

TICKAURA2

JACKSON 1

KARMONM

C00K1

MARENG01

-4000' -

-6000' SOUTHERN BASIN PLAY Jurassic section overlies Permian source and carrier beds. No Triassic seal. -8000'

d ]

Eromanga Basin Sedimenjts

Cooper Basin

-10000'

COOK PLATFORM

Triassic

Permian Sediments Oil accumulation

Migration pathway

Lower Triassic seal absent Upper Triassic seal leaky or breached by faults.

NORTHERN BASIN PLAY Migration via Triassic or Basal Jurassic carrier beds.

Figure 5. Present day structural cross section illustrating spatial relationships between reservoir, source and seal rocks across the Cooper-Eromanga petroleum system in south west Queensland. ' I' 200

PERMIAN

TRIASSIC

Geological Time Scale

JURASSIC M

I I I Mil Eariy structural trap development

Compaction and drape over early stnjctures

Petroleum System Events

Source Rock Reservoir Rock Seal Rock Overburden Rock Trap Formation Generation, Migration and Entrapment Preservation (Tectonic events - T) Critical Moment

Figure 6. Events charts for the Cooper-Eromanga petroleum system showing the timing of essential elements and processes. Petroleum System Events - Temporal Relationships Maturation modelHng studies conducted by Hunt et aL(lO) have indicated that Permian source rocks commenced oil generation (Rv=0.7%) during the Late Cretaceous in the Nappamerri and Windorah trough areas with primary migration of hydrocarbons inferred to have occurred throughout the Late Cretaceous-Early Tertiary period. This timing of oil generation is coincident with attainment of maximum depth of burial of Eromanga Basin sediments during deposition of the Winton Formation. Pre-Tertiary structures with a history of early structural growth would have been most favourably positioned for entrapment of hydrocarbons generated at this time. This phase of primary oil generation and migration is defined by Magoon et al.(ll) as the critical moment of the petroleum system which emphasises the importance of early trap development and effective palaeo-migration pathways as key controls upon accumulations (Figure 6). 361

Three major episodes of tectonism have been recognised by Hoffman (3) in the Tertiary, which post-dates the critical moment of primary oil generation and migration. Two phases of major uplifl, erosion and deformation occurred between Late Cretaceous and Oligicene time, with a third mild compressive event occurring during Miocene time. The effects of Tertiary tectonism vary across the basin, with significant amounts of uplift estimated by Rodgers et al.(12) to be in excess of 2000 feet along the Mt. Howitt, Curalle-Momey, and Innamincka structural highs. The intensity of these Tertiary events have affected the preservation of primary (pre-Tertiary) hydrocarbon accumulations through breaching of reservoir seals from faulting and fracturing of seal rocks, alteration of migration pathways, and creation of late structural traps for receipt of hydrocarbon charge from secondary migration processes. Local disruption of petroleum generative processes has also resulted from Tertiary structuring with source rocks uplifted out of the oil window in several local source areas. However, the magnitude of upUft and deformation associated with individual Tertiary structural events is poorly understood. Local generation of hydrocarbons may have continued throughout the Early Tertiary up to the second phase of tectonism in Oligocene time, particularly in source areas where the effects of initial post-Winton tectonism were minimal. Direct evidence for secondary migration of hydrocarbons during the Tertiary is difBcult to clearly establish. However, seal breaching with leakage of hydrocarbons from primary accumulations has been indicated by the presence of strong residual hydrocarbon shows and indications of deeper palaeo-oil/water contacts in several areas, particularly where extensive faulting of the Cretaceous-Triassic section has occurred. Critical Moment Palaeostructure A schematic palaeostructural section across the Cooper-Eromanga system has been constructed to illustrate the relative structural positions of major accumulations at the critical moment of the system, when Eromanga sediments reached maximum burial depths (Figure 7). Estimates of Late Cretaceous burial depths have been determined utilising section loss estimates determined from Cretaceous shale velocity studies by Rodgers et aL(12) which aimed to quantify the magnitude of Tertiary uphfl in the Eromanga Basin. Comparison of present day and critical moment sections indicates that in the southem area of the system, the Jackson anticline has been a prominent high at the time of primary and secondary hydrocarbon generation and migration. The Munro and Tickalara structures appear to have retained similar relative structural positions pre- and post-Tertiary uplifl. By contrast, the critical moment section indicates that the Cook Platform was positioned higher relative to surrounding structures (vis a vis the present day section), such that the Cook area would have been a major focus for primary hydrocarbon migration and entrapment prior to Tertiary uplift. Preservation of hydrocarbon accumulations in areas which have undergone intense Tertiary uplifl would have been more susceptible to leakage from seal breach and trap deformation. Investigation of the palaeostructure of the Cooper-Eromanga petroleum system at the critical moment of the system provides a more comprehensive understanding of the structural controls on migration and entrapment at the time of primary hydrocarbon generation and migration, and provides a basis for examining the potential for preservation of accumulations post-Tertiary uplifl.

362

N MUNRO 1

TICKALARA2

JACKSON 1

KARMONAI

C00K1

MARENGO 1

Feet SS -4000' -

-6000'

-8000'

SOUTHERN BASIN PLAY Jurassic section overlies Permian source and carrier beds. No Triassic seal.

-10000'

CZ]

Eroraanga Basin Sediment

Cooper Basin iH

Triassic Permian Sediments Oil accumulation

Migration pathway

COOK PLATFORM Lower Triassic seal absent Upper Triassic seal leaky or breached by faults.

NORTHERN BASIN PLAY Migration via Triassic or Basal Jurassic carrier beds.

Figure 7. Palaeostructural cross section illustrating spatial relationships between reservoir, source and seal rocks at the critical moment of the Cooper-Eromanga petroleum system. Migration and Entrapment Processes Migration pathways linking Permian source rocks with Jurassic/ Cretaceous reservoirs necessitate a combination of lateral and vertical migration mechanisms which vary in complexity across the basin. The maturity of exploration in the southem area of the Cooper-Eromanga system has facilitated the development of models which adequately explain vertical migration of hydrocarbonsfromPermian source rocks into Jurassic/Cretaceous reservoirs where the Triassic section is thin or absent. In the southem area, Jurassic reservoirs commonly overlie Permian source and carrier beds; the absence of Triassic seals provides direct access between source and reservoir. Lateral migration of hydrocarbons from source to reservoir beyond the edge of Permian sediments occurs over relatively short distances, with most accumulations located within 30 km of the Permian edge. By contrast, in the mid-basin and northem sectors of the Cooper-Eromanga system, the Triassic section forms a barrier to vertical migration of hydrocarbons from the Permian into the Jurassic section (Figure 5). Triassic seals in these areas must be incompetent or breached to allow vertical migration to occur from underlying source beds into Jurassic reservoirs. Stratigraphic subdivision of the Triassic Nappamerri Group by Powis (13) has resulted in the identification of a basal Triassic seal unit (Paning Member) and an upper Triassic seal unit (Gilpeppee Member). In the mid-basin sector of the Cooper-Eromanga system, the Paning Member seal is thin or absent, such that migration of hydrocarbons from Permian beds into Triassic strata can be readily achieved. The recent discovery of oil in Triassic reservoirs at Keleary and Telopea in South Australia support this migration model. In addition, faulting with breaching of the upper Triassic Gilpeppee seal facilitates migration of hydrocarbons into overlying Jurassic reservoirs. This fault model, for vertical migration, is exemplified by the Cook field in south west Queensland where oil is reservoired in the Button Sandstone and Namur Sandstone. Strong residual hydrocarbon shows in the upper Triassic Doonmulla Sandstone Member in Cook1 suggest that Triassic sandstones may have acted as carrier beds for lateral movement of hydrocarbons. Long distance migration of hydrocarbons from Permian source areas has previously received little consideration due to what Jenkins (8) describes as low expulsion efficiencies of source rocks in the Cooper-Eromanga system. The commercial discovery of the Inland oilfield,approximately 70 km northfromthe Permian subcrop edge, has challenged previous views on migration of hydrocarbons into areas beyond Permian source areas. Effective lateral and vertical migration processes must be invoked to explain the Inland accumulation if a Windorah Trough Permian source is assumed.

363

Although models can be constructed in which combined lateral and vertical migration processes are be shown to be effective, prediction of migration pathways and focussed hydrocarbon charging of structures remains a major challenge in exploration of the mid-basin and northern areas. Future Considerations A more comprehensive understanding of geologic controls on accumulations in the Cooper-Eromanga petroleum system has emerged from the synthesis of spatial and temporal elements of the system. The synergy required between key elements and processes within the system for significant accumulation of hydrocarbons in the Eromanga Basin has indicated the critical moment of the system to be 90 my, at the end of Late Cretaceous sedimentation and immediately prior to Tertiary uplift. As a result, pre-Tertiary structural trap development, effective lateral and vertical migration of hydrocarbons and post-Tertiary preservation of accumulations have been identified as major geological controls on Eromanga Basin accumulations. The critical moment of the system highlights the importance of understanding the pre-Tertiary structural configuration of the system to identify early formed structural traps (for primary migration and charge) and to delineate migration pathways operative at the critical moment. Further detailed study is required to improve understanding of the effects of Tertiary tectonism upon preservation of primary accumulations, and to map out the palaeostructure of the system on a regional basis. Stratigraphic positioning of Permian source beds relative to Jurassic/Cretaceous reservoirs places strong emphasis on understanding processes for effective vertical and lateral migration of hydrocarbons throughout the system. These processes vary in complexity geographically across the system area. On the basis of migration entrapment mechanisms, three main play types can be identified in the Cooper-Eromanga Basin petroleum system; i) southem edge of the system where Jurassic reservoirs have direct access to Permian source beds; ii) mid-basin area of the system where Triassic seal units are thin to absent and/or breached to allow vertical migration of hydrocarbons to occur; and iii) northem area of the system where long distance lateral migration must be effectively combined with vertical migration for miajor accumulations to occur. Since e?q)loration in the mid-basin and northem basin areas is at a lower level of maturity, significant undiscovered hydrocarbon potential may still exist in these areas. Although investigation of the major elements of the Cooper-Eromanga petroleum system has improved current understanding of key factors which govem the habitat of the majority of Eromanga oil accumulations discovered to date, the possibility of primarily Eromanga sourced oil accumulations in south west Queensland should not be dismissed. Examination of a Jurassic/Cretaceous sourced Eromanga oil system would provide an additional perspective on hydrocarbon generation, migration and entrapment in the basin, and would create new opportunities for future petroleum e}q)loration in the Eromanga Basin. References 1.

4. 5.

GREEN, P.M., BRAIN, T.J. and JOHN, B.H., Possible stratigraphic controls on hydrocarbon distribution within the Jurassic-Early Cretaceous rocks, Eromanga Basin, southem Queensland. In O'Neil, B., (Editor): The Cooper and Eromanga Basins, Australia. Proceedings of the Petroleum E^loration Society of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (S.A. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 251-264. HOFFMAN, KLL., The influence of pre-Jurassic tectonic regimes on the stmctural development of the southem Eromanga Basin, Queensland. In O'Neil, B.J. (Editor): The Cooper and Eromanga Basins, Australia. Proceedings of the Petroleum Exploration Society of Australia, Society of Petroleum Engineers, AustraUan Society of Exploration Geophysicists (SA. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 315328. MOORE, P.S., An exploration overview of the Eromanga Basin. In Gravestock, D.I., Moore, P.S. and Pitt, G.M. (Editors): Contributions to the geology and hydrocarbon potential of the Eromanga Basin. Special Pubhcation No. 12, Geological Society of Australia Incorporated, 1986, pp 1-8. WECKER, H.R.B., Prospectivity in perspective: Eromanga Basin. The APEA Joumal, v.29,1989, pp 379-397. HEATH, R., McINTYRE, S. and GIBBINS, N., A Permian origin for Jurassic reservoired oil in the Eromanga Basin. In O'Neil, B.J. (Editor): The Cooper and Eromanga Basins, Australia. Proceedings of the Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (S.A. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 405-415. POWELL, T.G., BOREHAM, C.J., MCKIRDY, D.M., MICHAELSEN, B.H. and SUMMONS, R.E., Petroleum geochemistry of the Murta Member, Mooga Formation, and associated oils, Eromanga Basin. The APEA Joumal, v.29,1989, pp 114-129. HAWKINS, PJ., ALMOND, C.S., CARMICHAEL, D.C., SMITH, R.L and WILLIAMS, L.J., Kerogen characterisation, and organic and minerals diagenesis of potential source rocks in Jurassic units, southem Eromanga Basin, Queensland. In O'Neil, B.J. (Editor): The Cooper and Eromanga Basins, Australia. Proceedings of the Petroleum Exploration Society of Australia, Society of Petroleum Engineers, AustraUan Society of 364

10.

11. 12. 13.

Exploration Geophysicists (S.A. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 584599. JENKINS, C.C., Geochemical correlation of source rocks and crude oils from the Cooper and Eromanga Basins. In O'Neil, B.J. (Editor): The Cooper and Eromanga Basins, Austraha. Proceedings of the Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Austrahan Society of Exploration Geophysicists (S.A. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 525-540. GILBY, A.R. and MORTIMORE, I.R., The prospects for Eromanga oil accumulations in the northem Cooper Basin region, Austraha. In O'Neil, B.J. (Editor): The Cooper and Eromanga Basins, Australia. Proceedings of the Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 391-403. HUNT, J.W., GUTHRIE, D.A. and DODMAN, A.R, Jackson Field-Australia, Cooper-Eromanga Basins, Central Austraha. In Beaumont, E.A. and Foster, N.H. (Editors): Structural Traps IV. Atlas of Oil and Gas Fields, AAPG Treatise of Petroleum Geology, 1990, pp 217-253. MAGOON, L.B. and DOW, W.G., The Petroleum System. In Magoon, L.B. and Dow, W.G. (Editors): The Petroleum System - from Source to Trap. AAPG Memoir 60,1994, pp 3-24. RODGERS, J., WEHR, F.L. and HUNT, J.W., Tertiary uphft estimation from velocity data in the Eromanga Basin, Exploration Geophysics, v. 22,1991, pp 321-324. POWIS, G.D., Revision of Triassic stratigraphy at the Cooper Basin to Eromanga Basin transition. In O'Neil, B.J. (Editor): The Cooper and Eromanga Basins, Australia. Proceedings of the Petroleum Exploration Society of Austraha, Society of Petroleum Engineers, Austrahan Society of Exploration Geophysicists (SA. Branches), Cooper and Eromanga Basins Conference, Adelaide, 1989, pp 265-276.

365

PALYNOFLORAL AND MEGAFLORAL INDICATIONS OF PALAEOCLIMATE IN THE LATE TRIASSIC, JURASSIC, AND EARLY CRETACEOUS OF SOUTHEASTERN QUEENSLAND John L. McKeUar Department of Mines & Energy, Queensland Summary Several distinct climatic changes are recognised in the Late Triassic-Early Cretaceous megafloralpalynofloral succession of southeastern Queensland. Humid, warm temperate conditions are associated with the ginkgoalean-bearing floras of the Late Triassic and Neocomian. In contrast, climates during most of the intervening Jurassic were somewhat warmer and variably humid. In the earliest Jurassic (HettangianSinemurian), warm dry conditions with seasonal rainfall appear to have developed, enabling Corollinaproducing, xeromorphic cheirolepidiacean conifers to establish their dominance of the vegetation. As relative sea level rose quickly during the epoch to a peak in the mid Toarcian, warm humid conditions, favouring proliferation of the araucariacean conifers, are evident on a regional scale. Generally similar climatic conditions persisted into succeeding late Toarcian-late Oxfordian times in southeastern Queensland, althou^ there are some megafloral indications that are suggestive of a more tenq)erate climate. Overall, a transitional warm to warm temperate, humid climate is indicated for this part of the succession, which includes the rich megaflora of the Walloon Coal Measures, as general conditions throu^out this interval were clearly not cool enough to have permitted warm temperate plant associations with ginkgoaleans to exist in lowland depositional environments. The reappearance of these plants in strata of Neocomian age indicates that the climate had changed by that time. However, details of the Jurassic megafloral succession above the Walloon Coal Measures are generally lacking, and the retum to a warm temperate climate is assumed to have occurred somewhat earlier in the late Oxfordian, based on the significantly increased representation of trisaccate podocarpacean pollen in palynofloral assemblages. Introduction Preliminary results of an investigation into the palaeoclimatic imphcations of major changes in the Late Triassic-Early Cretaceous palynofloral and megafloral successions of southeastem Queensland are provided here. The study [McKellar (1)] firom which these data ensue was initiated in respect of the Jurassic palynofloras and megafloras of southeastem Queensland in order to acquire a better perception of associated cUmates, and to enhance comprehension of their placement in time and space. The large thermal and geoid anomaly, which had developed under Pangaea by the end of the Palaeozoic and is still represented today by the African-eastern Atlantic geoid h i ^ , led to significant uplift of the supercontinent at the close of the era; and this is reflected in the first-order lowstand of eustatic sea level at the Permian-Triassic boundary [Anderson (2, 3); Erwin (4)]. The slow development of this and other anomalies to a certain critical size in regions that were sited well away from the equator apparently effected substantial changes in the moments of inertia of the mantle, leading to a rapid change in the location of the planet's principal axis of inertia, and thus in the location (relative to Earth) of the rotational axis. This motion of the Earth relative to its essentiallyfixed-in-spacerotational axis is known as tme polar wander (TPW); and it has occurred at particularly fast rates at certain times in the geological past, as it is occurring at the present time, in order to distribute mass anomalies about the equator and thus minimise the kinetic energy of rotation [e.g.. Chase (5); Gordon (6); Anderson (2, 3); Besse & Courtillot (7)]. The TPW indicated at the close of the Permian and during the subsequent Triassic appears to have been responsible for the shift of Pangaea's centre of mass towards the equator during this interval. This resulted in movement of the eastem Australian region of the supercontinent away from the high latitudes that it was associated with during the Early Permian to mid latitudes during the Triassic and perhaps also the earliest Jurassic. Together with changes in relative sea level and thus base level of erosion (and their effects on regional climate and depositional environments), TPW accounts for the demise in eastem Australia of the Perniian cool temperate Glossopteris Flora and the succeeding development of the warm temperate climate associated with the Triassic Dicroidium Flora [McKellar (1)]. Phytogeography and palaeoclimatology Climates associated with the Late Triassic-Early Cretaceous megafloras of southeastem Queensland have been broadly assessed in terms of Krassilov's (8) palaeoclimatic zonation of Mesozoic megafloras, althou^ the published works of Vakhrameev (9-13) have also particularly influenced interpretation. Krassilov recognised two major vegetational types, deciduous Phoenicopsis forests and evergreen brachyphyllous forests, which respectively delimited his temperate and warm zones, the latter straddling the equator and extending substantially into both hemispheres. A third zone, restricted to a narrow belt between the other two in each hemisphere, was referred to as the warm temperate ecotone (and also as 366

the ecotonal warm temperate zone or simply the warm temperate zone). This zone, the vegetation of which was compositionally intermediate between that of the warm and temperate zones, is particularly relevant to palaeoclimatic interpretation of the eastem Australian Mesozoic. It denotes the restricted climatic conditions under whichfloralelements broadly characteristic of the two climatic extremes of the equator-to-pole thermal gradient were able to coexist. Although Krassilov referred to the transitionalflorabetween his warm and temperate zones as the warm temperate zone, he did not specifically embrace the term "cool temperate" for thefloraof his temperate zone, which was so delimited by Ziegler et al (14), who assigned the Early-Middle Jurassic Siberian flora to the cool temperate deciduous biome. Camian to early Norian-Rhaetian Strata of early Late Triassic (Camian to early Norian-Rhaetian) age in eastem AustraHa have restricted distribution. In southeastern Queensland, they largely encompass the Ipswich Coal Measures, as well as equivalents in the Tarong beds and partial equivalents in the lower part of Callide Coal Measures [de Jersey (15-18); Day et al. (19); Stevens (20); ages based on de Jersey & Raine (21)]. The associated palynofloras represent the Craterisporites rotundus Oppel Zone [de Jersey (22); Helby et al (23)] and also, for the upper section of the Callide Coal Measures, part of the succeeding Polycingulatisporites crenulatus Oppel Zone [de Jersey (22); Helby et al (23)]. These biozones constitute the upper part of the Late Permian to latest Triassic [or early Hettangian(?)] Falcisporites Superzone [Helby et al (23)], which is generally characterised by an abundance of non-striate bisaccate pollen (Alisporites/Falcisporites) of corystosperm affinities; and their megafloral equivalent is represented by the Dicroidium Flora [Gould (24)]. Provincialism developed in the Falcisporites Superzone in the Middle and Late Triassic [Dolby & Balme (25); de Jersey & McKellar (26); Helby et al (23)], as the Australian region moved into lower palaeolatitudes; and this allowed distinctive Tethyan elements to penetrate palynofloral associations in northern and westem Australia. These palynofloras were termed the Onslow Microflora by Dolby & Balme (25); and coeval, higher-palaeolatitude palynofloras from southeastem Queensland (including those from the Late Triassic Ipswich Coal Measures, Tarong beds and Callide Coal Measures), as well as palynofloras from southem Australia, which all lack the distinctive pollen of the Onslow Microflora, were referred to as the Ipswich Microflora. These Microfloras represent two broad megafloral realms that have been designated as the Onslow and Ipswich Provinces [Foster et al (27)]. For the Ipswich Microflora, based apparently on its latitudinal relationship with the Onslow Microflora, Jersey & GrantMackie (28) suggested a cool temperate climate. However, it is evident that the climate associated with the Ipswich Province (including the Late Triassic megafloras of the Ipswich Coal Measures in southeastem Queensland and the Red Cliff Coal Measures in northeastem New South Wales) would have been of humid, warm temperate aspect, considering records of numerous ginkgoalean fironds, cycadophytes such as Pterophyllum multilineatum, and the corystosperm, Pachypteris [Walkom (29); Jones & de Jersey (30); Hill et al (31); Houston (32); Hint & Gould (33); Gould (24)]. This accords with Krassilov's (8, pp. 213-214, Figure 1) assignment of the megaflora of the Ipswich Coal Measures to his warm temperate zone. Early Norian-Rhaetian to early Hettangian Palynologically, the latest Triassic-Early Jurassic transition has been detailed in the Clarence-Moreton Basin [de Jersey (15,34-39); McKellar (40)] and in the adjacent Nambour Basin [McKellar (41-45)]. In the basal Clarence-Moreton Basin (basal Bundamba Group) succession, palynofloras in the Aberdare Conglomerate-Raceview Formation interval [Assemblage A of de Jersey (39)] and in the succeeding lower Ripley Road Sandstone [Assemblage B of de Jersey (39)] can be considered to represent the final stages of the Ipswich Microflora and the Falcisporites Superzone (and thus the Ipswich Province and the Dicroidium Flora). Overall, these palynofloras are embraced by the Polycingulatisporites crenulatus Oppel Zone. Marginally higher in the Clarence-Moreton Basin succession, in the lower Ripley Road Sandstone (within Assemblage B and the upper part of the Polycingulatisporites crenulatus Oppel Zone), de Jersey & Raine (21, p. 66) have equated the first appearance of Retitriletes austroclavatidites and R. semimuris [de Jersey (35, p. 40)] with the level of their appearance at the Otapirian-Aratauran (Triassic-Jurassic) boundary in New Zealand. However, this is somewhat complicated by de Jersey & Raine's (21, pp. 13, 62) indication that stratigraphic continuity at the Triassic-Jurassic boundary has not been demonstrated in New Zealand, as it is presently uncertain if equivalents of the lower Hettangian are represented. Moreover, the first appearances in the Aberdare Conglomerate-Raceview Formation-lower Ripley Road Sandstone interval of these and other species (namely Corollina meyeriana, Foveosporites moretonensis and Zebrasporites interscriptus), which have been recordedfi-omthe Late Triassic of Europe, suggests that the climate of southeastem Queensland in the latest Triassic-earliest Jurassic was becoming somewhat warmer [de Jersey & Raine (21, p. 76)] and that there was a slight change awayfiromthe warm temperate climate associated with the Ipswich Coal Measures. Nonetheless, from the aspect of the megaflora, southeastem Queensland at this time remained in Krassilov's (8) ecotonal warm temperate zone. This is based on records of numerous ginkgoalean fronds in the Brisbane area from the latest Triassic (Norian-Rhaetian) basal Landsborough Sandstone (basal Nambour Basin), which unconformably overlies the Ipswich Coal Measures and is equivalent to the Aberdare Conglomerate-Raceview Formation succession at the base of the Bundamba Group in the Clarence-Moreton Basin [Gould (46); Houston (32); McKellar (45)]. Moreover, 367

in the basal Bundamba Group, Jones & de Jersey (30) have recorded several small ginkgoalean-bearing collections, which they considered to be essentially Ipswich in character. Late Hettangian to mid Toarcian Approximately in the mid Ripley Road Sandstone, there is a marked change in the palynofloral succession, represented by significantly increased frequencies of Corollina (Classopollis) and signifying a major change in climatic conditions. The level at which this occurs represents the base of the Applanopsis (Callialasporites) dampieri Superzone and the Corollina torosa Oppel Zone within it [Helby et al (23)]. According to the latter authors, the boundary between these units and the underlying corystosperm-dominated Falcisporites Superzone represents a substantial extinction horizon and commonly coincides with a major regression in the Hettangian. In the westem part of the Clarence-Moreton Basin in southeastern Queensland, the superzone boundary delimits the variable hiatus recognised by de Jersey (39). However, in the eastem area of the basin (east of the West Ipswich Fault), deposition of the Ripley Road Sandstone appears to have been continuous, but this is uncertain. In the Clarence-Moreton Basin, the C. torosa Oppel Zone encompasses the mid to upper Ripley Road Sandstone and the overlying lower part of the Marburg Subgroup. Moreover, the upper Ripley Road Sandstone and the C torosa Zone are continuous with the Precipice Sandstone at the base of the Surat Basin to the west. There, the C. torosa ZOUQ extends into the succeeding Evergreen Formation (approximately up to the level of the Boxvale Sandstone and Westgrove Ironstone Members). In these intervals in both basins, Corollina comprises up to 85% of the palynoflora [de Jersey (35-37, 39); Reiser & Williams (47); McKellar (48)]. However, its abundance varies considerably from sample to sample and averages 45% (based on 81 samples from statistical data given in the above references). A warm climate and assignment to Krassilov's (8) warm zone is thus evident for these Corollina-iich strata in the Clarence-Moreton and Surat Basins. Generally dry conditions, however, are suggested for the depositional environment of the Precipice Sandstone and the mid to upper Ripley Road Sandstone, considering the xeromorphic character of the cheirolepidiacean conifers from which these pollen were derived. As the Precipice and Ripley Road Sandstones were deposited by a fast-flowing, braided stream system on a land surface of generally low relief [Cranfield et al (49); Martin (50)], a monsoonal climate (facilitated by the lowstand of sea level at the beginning of the Jurassic) therefore appears most likely, with long periods of generally dry conditions being broken by presumably short periods of higji-energy wet conditions [McKellar (1)]]. It is suggested that these were the conditions under which the hydrological requirements of the sedimentary facies and the xerophilous nature of the vegetation were muttially satisfied. Moreover, the abundance of Corollina indicates that the Cheirolepidiaceae were able to become well established on the extensive, well-drained, early Jurassic (late Hettangian-Sinemurian) land surface with little competition from other plant groups. As eustatic sea level and base level increased to a peak in the mid Toarcian [indicated in the Surat Basin by the widespread development of largely fluvio-lacustrine and deltaic conditions in the upper Evergreen Fomiation at the level of the Boxvale Sandstone and Westgrove Ironstone Members {Bradshaw & Yeung (51, 52); McKellar (1)}], environmental conditions were no longer favourable for domination of the flora by the Cheirolepidiaceae. With the progressive development of a more humid climate during the time of deposition of the Evergreen Formation (Pliensbachian-Toarcian^ other plants, particularly the Araucariaceae (represented in the palynoflora by Araucariacites fissus and Applanopsis spp.), were competing with them in the area they had previously overshadowed under the climatically adverse conditions associated with the Precipice Sandstone-Ripley Road Sandstone regime. Thus, during the Pliensbachian-Toarcian, warm moist conditions replaced the warm climate and seasonal rainfall associated with the earlier Jurassic. Also coincident with the Toarcian high sea-level phase was the cessation of red-bed formation on the westem margin of the Australian continent, indicating high water tables throughout the year and termination of the more strongly developed monsoonal influence that characterised the early Jurassic of that region [Bradshaw & Yeung (51,52)]. The breakdown of the monsoonal circulation in the Australian region was presumably related, at least in part, to the significant net increase in global sea level during the Early Jurassic and the consequent reduction of the landmass area exposed in southem Pangaea (Gondwanaland), as a result of increased coverage by fresh, brackish and marine bodies of water. Late Toarcian to late Oxfordian A number of seemingly diverse palaeoclimatic indicators are evident in the palynofloral-megafloral data derived from the upper Evergreen Formation-mid Westboume Formation succession in the Surat Basin. However, the general composition of the palynoflora essentially remained static through this interval, indicating that climatic conditions did not change significantly during the associated late Toarcian-late Oxfordian period of time [McKellar (1)]. Thermophihc filicalean spores of the Marattiaceae, Dipteridaceae and Matoniaceae, suggestive of a moist warm climate [Vakhrameev (13); Lele (53)], occur throughout this part of the palynofloral succession, although they are generally not common components of assemblages. Suppression of the cheirolepidiacean conifers is indicated by the generally low abundance of Corollina in palynofloral assemblages from upper Evergreen Formation-Westboume Formation strata [very rarely up to 23% of the palynoflora, occasionally up to 10-14%, and generally embracing values of less than 10%, with some samples containing insufficient 368

cheirolepidiacean pollen to be registered in counts of 250 specimens {McKellar (1)}]. However, in the adjacent Eromanga Basin and contrasting with the palynological data recorded from the Surat Basin, sporadically distributed assemblages with abundant Corollina (reflecting isolated remnant occurrences of vegetation dominated by the Cheirolepidiaceae and areas probably with limited water availability) have been encountered in the Middle Jurassic Hutton Sandstone succession [McKellar (67); Balfe (68); P. Price, verbal communication]. Although low quantities of Corollina (<10%) are suggestive of temperate climates [Vakhrameev (11)], the sporadic occurrence of assemblages with increased frequencies approaching 20% or more supports the view that conditions were still relatively warm, and that it was the development of consistently humid conditions and generally high water tables, rather than reduced climatic temperatures, that significantly impacted on the Cheirolepidiaceae and contributed to their diminished representation in palynofloras of the upper Evergreen Formation-Westboume Formation interval in the Surat Basin. The rich megaflora of the Walloon Coal Measures is encompassed by the concept of the Otozamites-Ptilophyllum Flora of Gould & Shibaoka (54). Bennettitaleans attributable to Otozamites and Ptilophyllum are generally abundant in this flora [Gould (55)], suggesting that climatic conditions were warm to warm temperate [Krassilov (8)]. Of particular significance in this group of plants in the Walloon flora and implying the existence of a warm rather than warm temperate climate is the fact that they not only possessed larger fronds than the somewhat cooler-climate (warm temperate). Early Cretaceous, Victorian bennettitaleans (J.G. Douglas, verbal communication), but also embraced the occurrence of O. incurvatus Douglas (56), a species with incurved pinnae margins. This xeromorphic adaptation, according to Douglas, is not unusual in bennettitaleans of the Otozamites and Ptilophyllum type. However, it is more spectacularly developed in 0. incurvatus; and the degree of protection afforded the undersurface of the pinnae by the incurved margin averages about 50% over the entire lamina. The absence of the Ginkgoales from the megaflora of the Walloon Coal Measures [highligiited by Gould (24, 55, 57); McLoughHn & Drinnan (58)] also suggests that a warm rather than a warm temperate climate existed at the time of their deposition. Moreover, apart firom a rare occurrence (recorded as: Ginkgoites sp., possibly G, antarctica) reported by Jones & de Jersey (59)firomthe late Early Jurassic (Toarcian) Brighton beds in the Nambour Basin, this group of plants is not represented in collections (from lowland depositional environments) through most of the Jurassic in southeastem Queensland, appearing to have been generally incompatible with the climatic conditions associated therewith. Contrastingly, growth-ringed, fossil-wood occurrences in the Walloon Coal Measures of small to fairly large pentoxylalean stems/trunks, and, in the Walloon Coal Measures and overlying Springbok Sandstone, of massive coniferlike trunks have been used to imply the existence of a temperate climate, as favoured by Gould (55) for the Walloon Coal Measures, based on comparison of its floral elements (where possible) with living analogues. Seasonal leaf fall also may have occurred, considering the common occurrence at some localities in the Walloon Coal Measures of Elatocladus and Taeniopteris [Gould (55); McLoughlin & Drinnan (58)]. In the Victorian Early Cretaceous, leaf base morphology and massed foliage of the Elatocladus leaf type suggest derivation from plants with seasonal leaf fall [Douglas & Williams (60)]; and, as indicated by McLoughlin & Drinnan (58), Taeniopteris leaves from the Walloon Coal Measures near Mutdapilly (Clarence-Moreton Basin) frequently occur as complete specimens in matted layers, consistent with the interpretation [Drinnan & Chambers (61, 62)] that they represent the foliage of deciduous pentoxylaleans. Winters through a considerable part of the late Early to Late Jurassic of the Surat Basin and adjacent basins in southeastem Queensland would have been moderated by the geographically extensive development of swamps, lakes, rivers and deltas, as has been indicated for such a scenario by Yemane (63), Ziegler (64) and Kutzbach & Ziegler (65). However, at times of lowered base level [corresponding to the lower, largely fluviatile parts of the sedimentary cycles recognised in the Surat Basin {Exon & Burger (66)}], areas were not as extensively covered by water and this may have affected the climate of the region. Ginkgoaleans may then have existed in lowland environments under more temperate conditions, but this is only speculative, as the associated sandy lithofacies (Hutton Sandstone, Springbok Sandstone) generally lack well preserved plant megafossil assemblages. Considering the phytogeographic-palaeoclimatic megafloral zonation of Krassilov (8) in relation to the assorted megafloral and palynological evidence cited above for the upper Evergreen Formation-mid Westboume Formation interval in the Surat Basin, a transitional warm to warm temperate, moist climate is suggested. Conditions were clearly not cold enough to either effect the bennettitaleans (which persist into succeeding Early Cretaceous strata, although largely, if not entirely, in the absence of the more thermophihc Otozamites), or induce cooler-weather ginkgoaleans to infiltrate the lowland vegetation associated with the known collections. Late Oyfordian to Neocomian Higher in the Surat Basin succession, a slight change to a miarginally cooler (warm temperate) climate is apparent in the mid Westboume Formation (late Oxfordian) at the boundary between the Applanopsis dampieri Superzone and the SMZCQtiingMicrocachryidites Superzone [of Helby et al (23)]. At and above this level, trisaccate podocarpacean pollen grains become significantly more frequent in occurrence, this being particularly the case in the post-Westboume Formation succession. According to Dettmann [(69); also see Dettmann & Playford (70)], the widespread and abundant representation of Microcachrys-likQ (podocarpacean) pollen, Microcachryidites antarcticus, (epitomising the Microcachryidites Microflora) in the latest Jurassic-Early Cretaceous of Australia lends support to cool temperate 369

conditions (based on the present-day, cool temperate, Tasmanian occurrence of the trisaccate-pollen producing podocarp, Microcachrys), although the more abundant Gleichenia-likQ spores and more diverse schizaeaceous derivatives in the Great Artesian Basin (encompassing the Surat, Eromanga and Carpentaria Basins) would seem to indicate warmer climatic conditions. That the plants which produced such trisaccate pollen favoured cooler (and humid) conditions is indicated [from information compiled by Dettmann (69)]: (a) by their absence during the Early Cretaceous from northem, low latitude areas of Gondwanaland (northem South America and northem Africa) where Corollim was particularly abundant; (b) by their decline in northem and Western Australia during late Albian-Cenomanian times where Corollina was represented by increased frequencies; and (c) by their prevalence in high latitude southeastern Australia where Corollina was sparsely represented. Megafloral data from the Westboume Formation and overlying Gubberamunda Sandstone are generally lacking, but, by the time of the Neocomian megafloras of the succeeding Orallo Formation-Mooga Sandstone-Bungil Formation interval, the Ginkgoales again had become moderately conspicuous elements of the lowland vegetation [Gould (24); Day (71); Cantrill & Webb (72); Dettmann et al (73)]. As these plants generally preferred a humid, warm temperate to cool tenqjerate climate, their common occurrence in the Late Triassic (Ipswich Coal Measures and basal strata of the ClarenceMoreton and Nambour Basins), their virtual absence through a considerable part of the Jurassic, and their reappearance in the Early Cretaceous of southeastem Queensland supports the view that (in this region) climatic temperatures were generally higher through most of the Jurassic than they were in either the preceding Late Triassic or the succeeding Early Cretaceous. Moreover, the Neocomian Surat Basin floras also embrace the bennettitaleans, Ptilophyllum and Pterophyllum, as well as numerous conifers and pteridosperms. Their assignment to Krassilov's (8) ecotonal warm temperate zone is suggested [cf. Krassilov (8, p. 214, Figure 2); assigned unspecified, Early Cretaceous floras of Queensland to his warm zone]. Furthermore, the floras documented by Day (71) from the Mooga Sandstone-Bungil Formation interval (recorded by that author as the Blythesdale Fomiation) have been correlated by Cantrill & Webb (72) with the Neocomian flora of Victoria [Zone B/Ptilophyllum-Pachypteris austropapillosa Zone of Douglas (74); renamed the Phyllopteroides laevis Zone by Cantrill & Webb], which was considered by Douglas & Williams [(60); also Douglas (75)] to be of warm temperate character. Krassilov (8, p. 214, Figure 2) has indicated that this Victorian flora may belong to his warm temperate ecotone; and this view is tentatively supported here, although ginkgoaleans and bennettitaleans, are not directly associated with each other, but are otherwise both present in Zone B assemblages [Subzones a and b; see Douglas (74)]. In summary, a moist warm temperate climate is suggested for the mid Westboume Fonnation-Bungil Formation interval in the Surat Basin. However, the lower stratigraphic limit of the Phyllopteroides laevis Zone [and of the Phyllopteroides Flora of Gould (24)] is uncertain, as megaflorasfromthe Orallo Formation and underlying Gubberamunda Sandstone are not well known [Day (71); Exon (76)]. Conclusions A humid, warm temperate climate is indicated for the ginkgoalean-bearing megafloras of the Late Triassic in southeastem Queensland. However, at the beginning of the Jurassic (Hettangian-Sinemurian), xeromorphic, Coro/toa-producing, cheirolepidiacean conifers established their dominance of the regional vegetation, suggesting that a warm, generally dry climate existed at that time. As the associated lithofacies (Ripley Road Sandstone, Precipice Sandstone) point to a braided-stream depositional environment, rainfall was presumably seasonal. The lowstand of eustatic sea level and the high-riding character of the AustraUan continent at the beginning of the Jurassic had a significant effect on the regional climate, encouraging the Pangaean monsoonal circulation to extend into eastern Australia. As relative sea level and base level rose during the Early Jurassic, the monsoonal influence was terminated and warm humid conditions prevailed during Evergreen-Formation (PUensbachian-Toarcian) times. Following this change in climate, which facilitated the expansion of the Araucariaceae and led to the suppression of the cheirolepidiacean conifers, climatic conditions appear to have been relatively stable during the late Toarcian-late Oxfordian of southeastem Queensland, as no major changes are evident in the palynofloral succession. A transitional warm to warm temperate, moist climate is suggested for the associated upper Evergreen Formation-mid Westboume Formation interval, which includes the rich megaflora of the Walloon Coal Measures. This flora, on one hand, incorporates thermophilic fem and bennettitalean remains, but, on the other, embraces occurrences of fossil wood with growth rings, as well as plants that were apparently subjected to seasonal leaf fall. Moreover, and quite significantly, the Walloon flora lacks ginkgophytes, which reappeared (as far as can be ascertained from the limited record of megafloras from Jurassic strata above the Walloon Coal Measures) in lowland depositional environments by the Neocomian, following a return to a distinctly warm temperate climate. However, the palynofloral succession in the Surat Basin suggests that this change to slightly cooler conditions commenced somewhat earlier in the late Oxfordian during deposition of the mid Westboume Formation, as there is a significant increase in the frequency of trisaccate podocarpacean pollen in palynofloras at and above that level. This increase is taken to mark the introduction of the Microcachryidites Superzone to the palynological succession and to define the conclusion of a warmer climatic interlude, represented by the Applanopsis dampieri Superzone, in the Mesozoic history of the Australian region.

370

References 1.

2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12. 13.

14.

15. 16. 17. 18. 19.

20. 21. 22.

23.

24. 25.

McKELLAR, J.L.: Late Early to Late Jurassic palynology and biostratigraphy of the Roma Shelf area, northwestern Surat Basin, Queensland; and phytogeographic-palaeoclimatic implications of the Applanopsis dampieri and Microcachryidites Superzones in the Jurassic-Early Cretaceous of Australia - an overview assessed against a background of floral change and apparent true polar wander in the preceding late Palaeozoic-early Mesozoic. Queensland Geology (in preparation). ANDERSON, D.L.: Hotspots, polar wander, Mesozoic convection and the geoid. Nature, 297,391-393 (1982). ANDERSON, D.L.: Theory of the Earth, Blackwell Scientific Pubhcations, Boston, 366 pages (1989). ERWIN, D.H.: The great Paleozoic crisis: life and death in the Permian. Columbia University Press, New York, 327 pages (1993). CHASE, C.G.: The geological significance of the geoid. Annual Review of Earth and Planetary Sciences, 13, 97117 (1985). GORDON, R.G.: Polar wandering and paleomagnetism. Annual Review ofEarth and Planetary Sciences, 15,567593 (1987). BESSE, J. & COURTILLOT, V.: Revised and synthetic apparent polar wander paths of the AMcan, Eurasian, North American and Indian Plates, and true polar wander since 200 Ma. Journal of Geophysical Research, 96(B3), 4029-4050 (1991). KRASSILOV, V.A.: Changes of Mesozoic vegetation and the extinction of dinosaurs. Palaeogeography, Palaeoclimatology Palaeoecology, 34, IQl-llA (1981). VAKHRAMEEV, V.A.: Jurassic floras of the USSR. Palaeobotanist, 14, 118-123 (1966). VAKHRAMEEV, V.A.: Range and paleoecology of Mesozoic conifers, the Cheirolepidiaceae. Paleontological Journal, 4(1), 12-25 (1970) [translated fromPaleontologicheskiy Zhumal, 1970(1), 19-34; in Russian]. VAKHRAMEEV, VA.: Pollen Classopollis: indicator of Jurassic and Cretaceous climates. The Palaeobotanist, 28-29, 301-307(1981). VAKHRAMEEV, V.A.: Climates and the distribution of some gymnosperms in Asia during the Jurassic and Cretaceous. Reviev^ of Palaeobotany and Palynology, 51,205-212 (1987). VAKHRAMEEV, V.A.: Jurassic and Cretaceous floras and climates of the Earth. Cambridge University Press, Cambridge, 318 pages (1991) [originally pubUshed in Russian in 1988: Yurskiye i melovyye flory: klimaty Zemli. Nauka, Moscow]. ZIEGLER, A.M., PARRISH, J.M., YAO, J., GYLLENHAAL, E.D., ROWLEY, D.B., PARRISH, J.T., NIE, S., BEKKER, A. & HULVER, M.L.: Early Mesozoic phytogeography and chmate. Philosophical Transactions of the Royal Society of London, 341B, 297-305 (1993). DE JERSEY, N.J.: Triassic miospores from the Blackstone Formation, Aberdare Conglomerate and Raceview Formation. Geological Survey of Queensland, Publication 348, Palaeontological Paper 22,41 pages (1970). DE JERSEY, N.J.: Palynology of samples from the Tarong Beds. Queensland Government Mining Journal, 71, 308-310 (1970). DE JERSEY, N.J.: Triassic miospores from the HvoU Formation and Kholo Sub-group. Geological Survey of Queensland, Publication 353, Palaeontological Paper 28,40 pages (1971). DE JERSEY, N.J.: Palynology and age of the Callide Coal Measures. Queensland Government Mining Journal, 75,249-252 (1974). DAY, R.W., CRANFIELD, L.C. & SCHWARZBOCK, H.: Stratigraphy and structural setting of Mesozoic basins in southeastem Queensland and northeastem New South Wales; in DENMEAD, A.K., TWEEDALE, G.W. & WILSON, A.F. (Editors): The Tasman Geosyncline - a Symposium. Geological Society of Australia, Queensland Division, Brisbane, 319-362 (1974). STEVENS, J.: Palynology of the Calhde Basin, east-central Queensland. Papers, Department of Geology, University of Queensland, 9(4), 1-35 (1981). DE JERSEY, N.J. & RAINE, J.L: Triassic and earUest Jurassic miospores from the Murihiku Supergroup, New Zealand. New Zealand Geological Survey, Paleontological Bulletin 62,164 pages (1990). DE JERSEY, N.J.: Miospore zones in the Lower Mesozoic of southeastem Queensland; in CAMPBELL, ELS.W. (Editor): Gondwana Geology. Papers presented at the Third Gondwana Symposium, Canberra, Australia, 1973. ANU Press, Canberra, 159-172 (1975). HELBY, R., MORGAN, R. & PARTRIDGE, A.D.: A palynological zonation of the Australian Mesozoic; in JELL, P.A. (Editor): Studies in Austrahan Mesozoic palynology. Memoirs of the Association of Australasian Palaeontologists, 4, 1-94 (1987). GOULD, R.E.: The succession of Australian pre-Tertiary megafossil floras. Botanical Review, 41(4), 453^83 (1975). DOLBY, J.H. & BALME, B.E.: Triassic palynology of the Carnarvon Basin, Western Australia. Review of Palaeobotany and Palynology 22, 105-168 0976).

371

26.

27. 28. 29.

30. 31. 32. 33.

34. 35. 36. 37. 38.

39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

49. 50. 51. 52. 53.

DE JERSEY, NJ. & McKELLAR, J.L.: Triassic palynology of the Warang Sandstone (northern Galilee Basin) and its phytogeographic iii^lications; in CRESSWELL, M.M. & VELLA, R (Editors): Gondwana Five. Balkema, Rotterdam, 31-37 (1981). FOSTER, C.B., BALME, B.E. & HELBY, R.: First record of Tethyan palynomorphs from the Late Triassic of East Antarctica. AGSO Journal ofAustralian Geology & Geophysics, 15,239-246 (1994). DE JERSEY, N.J. & GRANT-MACKIE, J.A.: Palynofloras from the Permian, Triassic and Jurassic of New Caledonia. New Zealand Journal of Geology and Geophysics, 32,463-476 (1989). WALKOM, A.B.: Mesozoic floras of Queensland. Part 1. - concluded. The flora of the Ipswich and Walloon Series: (d) Ginkgoales, (e) Cycadophyta, (f) Coniferales. Geological Survey of Queensland, Publication 259, 49 pages (1917). JONES, O.A. & DE JERSEY, N.J.: The flora of the Ipswich Coal Measures. Morphology and floral succession. Papers, Department of Geology, University of Queensland (New Series), 111(3), 1-88 (1947). HILL, D., PLAYFORD, G. & WOODS, J.T. (Editors): Triassic fossils of Queensland. Queensland Palaeontographical Society, 32 pages (1965). HOUSTON, B.R.: Geology of the City of Brisbane. Part II. The post-Palaeozoic sediments and volcanics. Geological Survey of Queensland, Publication 324, 7-86 (1967). FLINT, J.C.E. & GOULD, RJE.: A note on the fossil megafloras of the Nymboidea and Red Cliff Coal Measures, southem Clarence-Moreton Basin, N.S.W. Journal and Proceedings of the Royal Society of New South Wales, 108, 70-74 (1975). DE JERSEY, N.J.: Triassic spores and pollen grains from the Bundamba Group. Geological Survey of Queensland, Publication 321, 21 pages (1964). DE JERSEY, N.J.: Early Jurassic miospores from the Helidon Sandstone. Geological Survey of Queensland, Publication 351, Palaeontological Paper 25,49 pages (1971). DE JERSEY, N.J.: Palynological evidence for a facies change in the Moreton Basin. Queensland Government Mining Journal, 72,464-472 (1971). DE JERSEY, N.J.: Palynology of core samples from the Helidon, Toowoomba and Kulpi areas. Queensland Government Mining Journal, 75, 128-144 (1973). DE JERSEY, N.J.: Miospore zones in the Lower Mesozoic of southeastem Queensland; in CAMPBELL, K.S.W. (Editor): Gondwana Geology, Papers presented at the Third Gondwana Symposium, Canberra, Australia, 1973. ANU Press, Canberra, 159-172 (1975). DE JERSEY, N.J.: Palynology and time relationships in the lower Bundamba Group (Moreton Basin). Queensland Government Mining Journal, 77,460-465 (1976). McKELLAR, J.L.: Palynostratigraphy of samples from GSQ Ipswich 24 and 25. Queensland Government Mining Journal, 82,479-487 (1981). McKELLAR, J.L.: Palynostratigraphy of the Lawnton to Brighton area, Nambour Basin. Queensland Government Mining Journal, 82, 52-60 (1981). McKELLAR, J.L.: Palynostratigraphy of samples from the Narangba area, Nambour Basin. Queensland Government Mining Journal, 82,268-273 (1981). McKELLAR, J.L.: Biostratigraphy of spore-pollen floras from the Toorbul Point-Burpengary-Caboolture area, Nambour Basin. Queensland Government Mining Journal, 83,30-34 (1982). McKELLAR, J.L.: Biostratigraphy of spore-pollen floras from the Caloundra to Nambour area, Nambour Basin. Queensland Government Mining Journal, 86, 368-376 (1985). McKELLAR, J.L.: Stratigraphic relationships in the Nambour Basin, southeastem Queensland. Queensland Geology, 5, 1-17 (1993). GOULD, R.E.: The geology of the Slacks Creek area, southeast Queensland. Papers, Department of Geology, University of Queensland, 6(5), 115-144 (1967). REISER, R.F. & WILLIAMS, A.J.: Palynology of the Lower Jurassic sediments of the northem Surat Basin, Queensland. Geological Survey of Queensland, Publication 339, Palaeontological Paper 15,24 pages (1969). McKELLAR, J.L.: Jurassic miospores from the upper Evergreen Formation, Hutton Sandstone, and basal Injune Creek Group, north-eastem Surat Basin. Geological Survey of Queensland, Publication 361, Palaeontological Paper 35, 89 pages (1974). CRANFIELD, L.C., SCHWARZBOCK, H. & DAY, R.W.: Geology of the Ipswich and Brisbane 1:250 000 Sheet areas. Geological Survey of Queensland, Report 95, 176 pages (1976). MARTIN, K.R.: Deposition of the Precipice Sandstone and the evolution of the Surat Basin in the Early Jurassic. APEA Journal, 21, 16-23 (1981). BRADSHAW, M. & YEUNG, M.: The Jurassic palaeogeography of Australia. Bureau of Mineral Resources, Australia, Record 1990/76, Palaeogeography 26, 60 pages (1990). BRADSHAW, M.T. & YEUNG, M.: Palaeogeographic atlas of Australia. Volume 8 - Jurassic. Bureau ofMineral Resources, Australia (1992). LELE, K.M.: Palaeoclimatic implications of Gondwana floras. Geophytology, 6,207-229 (1976). 372

54. 55. 56. 57. 58. 59. 60. 61.

62.

63. 64. 65.

66. 67. 68. 69. 70.

71. 72. 73.

74. 75.

76.

GOULD, R. & SHIBAOKA, M.: Some aspects of the formation and petrographic features of coal members in Australia, with special reference to the Tasman Orogenic Zone. Coal Geology (Australia), 2, 1-29 (1980). GOULD, R.: The coal-forming flora of the Walloon Coal Measures. Coal Geology (Australia), 1(3), 83-105 (1980). DOUGLAS, J.G.: New bennettitalean leaves from the Mesozoic of eastem Australia. Proceedings of the Royal Society of Victoria, 77, 197-206 (1963). GOULD, R.E.: The fossil flora of the Walloon Coal Measures: a survey. Proceeding of the Royal Society of Queensland, 85(3), 33-41 (1974). McLOUGHLIN, S. & DRINNAN, A.N.: A Middle Jurassic flora from the Walloon Coal Measures, Mutdapilly, Queensland, Australia. Memoirs of the Queensland Museum, 38,257-272 (1995). JONES, O.A. & DE JERSEY, N.J.: Fertile Equisetales and other plants from the Brighton Beds. Papers, Department of Geology, University of Queensland (New Series), 111(4), 1-16 (1947). DOUGLAS, J.G. & WILLL\MS, G.E.: Southem polar forests: the Early Cretaceous floras of Victoria and their palaeoclimatic significance. Palaeogeography, Palaeoclimatology, Palaeoecology, 39, 171-185 (1982). DRINNAN, A.N. & CHAMBERS, T.C.: A reassessment of Taeniopteris daintreei from the Victorian Early Cretaceous: a member of the Pentoxylales and a significant Gondwanaland plant. Australian Journal of Botany, 33, 89-100 (1985). DRINNAN, A.N. & CHAMBERS, T.C.: Flora of the Lower Cretaceous Koonwarra Fossil Bed (Korumburra Group), South Gippsland, Victoria; in ROBERTS, J. & JELL, RA. (Editors): Plants and invertebrates from the Lower Cretaceous Koonwarra Fossil Bed, South Gippsland, Victoria. Memoirs of the Association of Australasian Palaeontologists, 3, 1-77 (1986). YEMANE, K.: Contribution of Late Permian palaeogeography in maintaining a temperate climate in Gondwana. Nature, 361, 51-54 (1993). ZIEGLER, A.M.: Models come in from the cold. Nature, 361, 16-17 (1993). KUTZBACH, J.E. & ZIEGLER, A.M.: Simulation of Late Permian climate and biomes with an atmosphere-ocean model: comparisons with observations. Philosophical Transactions of the Royal Society of London, 341B, 327340 (1993). EXON, N.F. & BURGER, D.: Sedimentary cycles in the Surat Basin and global changes of sea level. BMR Journal ofAustralian Geology & Geophysics, 6, 153-159 (1981). McKELLAR, J.L.: Palynostratigraphy of core samples from GSQ Hughenden 7. Queensland Government Mining Journal, SO, 295-302 (1979). BALFE, P.E.: Stratigraphic drilling report - GSQ Hughenden 7. Queensland Government Mining Journal, 80, 282-293 (1979). DETTMANN, M.E.: The Cretaceous flora; in KEAST, A. (Editor): Ecological biogeography ofAustralia. Dr W. Junk bv Publishers, The Hague, 357-375 (1981). DETTMANN, M.E. & PLAYFORD, G.: Palynology of the Australian Cretaceous: a review; in CAMPBELL, K.S.W. (Editor): Stratigraphy and Palaeontology: Essays in honour of Dorothy Hill ANU Press, Canberra, 174-210 (1969). DAY, R.W.: Stratigraphy of the Roma-Wallumbilla area. Geological Survey of Queensland, Publication 318, 23 pages (1964). CANTRILL, D.J. & WEBB, J.A.: A reappraisal of Phyllopteroides Medwell (Osmundaceae) and its stratigraphic significance in the Lower Cretaceous of eastem Australia. Alcheringa, 11,59-85 (1987). DETTMANN, M.E., MOLNAR, R.E., DOUGLAS, J.G., BURGER, D., FIELDING, C., CLIFFORD, H.T., FRANCIS, J., JELL, R, RICH, T, WADE, M., RICH, P.V., PLEDGE, N., KEMP, A. & ROZEFELS, A.: Australian Cretaceous terrestrial faunas and floras: biostratigraphic and biogeographic implications. Cretaceous Research, 13,207-262 (1992). DOUGLAS, J.G.: The Mesozoic floras of Victoria. Parts 1 and 2. Geological Survey of Victoria, Memoir, 28, 310 pages (1969). DOUGLAS, J.G.: The Cretaceous vegetation, and palaeoenvironment of Otway Group sediments; in GLENIE, R.C. (Editor): Second South-eastern Australia Oil Exploration Symposium. Technical papers presented at PESA (Petroleum Exploration Society of Australia) Symposium, Melboume, 14-15 November, 1985, 233-240 (1986). EXON, N.F.: Geology of the Surat Basin in Queensland. Bureau of Mineral Resources, Australia, Bulletin 166, 160 pages (1976).

373

THE RELATIONSHIP BETWEEN GROUNDWATER CHEMICAL TYPE AND JURASSIC SEDIMENTARY FORMATIONS: THE EXAMPLE OF THE SANDY CREEK CATCHMENT, LOCKYER, SOUTHEAST QUEENSLAND Gerard A. McMahon & Malcolm £. Cox Research Concentration in Environmental Earth Science, School of Geology, Queensland University of Technology, GPO Box 2454, Brisbane, Qld, 4001 SUMMARY The Sandy Creek catchment hosts a small agricultural area in which the main source of irrigation water is from the Quatemary alluvium. An investigation under drought conditions and a very low water table demonstrated that groundwater chemistry is strongly influenced by discharge from the underlying Jurassic Marburg Formation. The observed hydrochemical variations are mostly related to the lithology of the bedrock inferring that cross-formational flow into the alluvium is a significant form of recharge, especially during conditions of low stream-flow. Six hydrochemical groups have been identified which relate closely to the four members of the Jurassic Marburg Formation. The Heifer Creek Sandstone Member contains low salinity groundwater which has migrated from the recharging basaltic aquifers of the Main Range Volcanics. The Winwill Conglomerate Member, Ma Ma Creek Sandstone Member, and the upper part of the Gatton Sandstone Member contribute groundwaters of higher salinity due to the release of connate interstitial saltwater, dissolution of salts, and the leaching of clays from within the sediments; and waters are typically Na,Mg-Cl type. Groundwater in the lower part of the Gatton Sandstone Member is more complex due to contributions from the sandstones, dissolution of Mg-Cl and Na-HC03 salts, and mixing of Na-HC03 groundwaters from adjacent alluvial tracts. INTRODUCTION The Lx)ckyer Valley is one of Queensland's major vegetable producing areas and groundwater plays a significant role in crop irrigation and as a minor source of domestic supplies. Most groundwater extracted for use is from recharge by streamflow originating in the surrounding ranges but also from unconfined alluvial aquifers. This alluvium primarily receives groundwater contributions from a series of Jurassic sedimentary fomiations. The Sandy Creek catchment contains some of the most saline irrigation water in the Lockyer Valley which restricts the variety of crops that can be grown. Advantage was taken of the prevailing drought conditions to sample groundwater which is least affected by meteoric recharge. The results presented in this extended abstract are summarised from a broader study of the Sandy Creek catchment by McMahon (1). The Sandy Creek catchment covers an area of about 12000 hectares in the central eastem part of the Loclg^er Valley (Figure 1). Sandy Creek is a major tributary of Laidley Creek, both forming the southeastern drainage system of the Lockyer Valley which is a major sub-catchment of the Brisbane River Drainage Basin. Talbot & Dickson (2) were the first to relate water chemistry to the lithologic units in the area. Zahawi (3) found that groundwater from the Marburg Formation was very high in total dissolved solids, particularly in the lower members. The effect of the 1980 drought on the salinity and chemistry of alluvial groundwaters in the Lockyer Valley was reported by Talbot et al (4). The significance of their study was the assumption that cross-formational flow of groundwaters from the sandstones to the alluvium would be best represented during low stream-flow periods. Under these conditions, the contribution of the underlying sandstones would be relatively greater and therefore, easier to detect. Reeve & Jones (5) compared stream waters draining basalts west of the Great Dividing Range with stream water in the Lockyer Valley and concluded that significant amounts of Na, Mg, and CI are contributed by the Marburg Formation. Dixon & Chiswell (6) investigated two other Lockyer tributaries and found that hydrochemical sections identified saline inputs from the underlying sandstones in the southwest Lockyer Valley during wet and dry seasons. The overall extent and nature of salinity in the Lockyer Valley has been reviewed by Cox & McNeil (7).

374

Figure 1. Location of the Sandy Creek catchment (shaded) within the Lockyer Valley, southeast Queensland. GEOLOGY The broad setting of the study area is the Laidley Sub-basin in the northern part of the intracratonic Clarence-Moreton Basin. Sediments of the Clarence-Moreton Basin were deposited during thermal relaxation of the crust over a series of Triassic transtensional basins (8). The basin fill consists entirely of fluvial sandstones, siltstones, shales and some conglomerates and reaches a maximum thickness of 3.5 kilometres in the Logan Sub-basin. A type section of the Marburg Formation was estimated by McTaggart (9) to be about 400 metres thick, however, a reference section of 814 metres was measured in drillhole GSQ Ipswich 18 (10) using McTaggart's boundaries for reference. The main sedimentary sequence under discussion makes up part of the Early Jurassic Bundamba Group of the ClarenceMoreton Basin. The stratigraphic nomenclature of the Marburg Subgroup has been studied by Cameron (11), McTaggart (9), Cranfield et al (12) and WeUs et al (13) and in detail by Mallet (14), Grimes (15), Gray (10), Cranfield (16), and Wells & O'Brien (13). Cameron (11) initially applied the term Bundamba Beds to a series of sandstone units between the Ipswich Coal Measures and the Walloon Coal Measures in the Ipswich area. The upper part of the Bundamba Group was formally defined the Marburg Formation by McTaggart (9) before Wells et al (13) renamed the section the Marburg Subgroup. The Marburg Formation classification developed by McTaggart (9) and defined by Gray (10) is preferred in this study since each member is discemible as a lithologic and geomorphologic entity, identifiable in the Sandy Creek catchment (Figure 2). The Marburg Formation is essentially horizontal in the study area and is overlain by the Walloon Coal Measures which is truncated and capped by Tertiary basaltflows.Alluvial and colluvial deposits form the valley fill. The four members are summarised below from the aspect of their hydrogeologic character.

375

Figure 2 .The geology of the Sandy Creek catchment and hydrochemical groups showing typical Stiff patterns. Note the relationship between hydrochemistry and the various members of the Marburg Formation. GATTON SANDSTONE MEMBER The Gatton Sandstone Member (oldest) conformably overhes the Helidon Sandstone of McTaggart (9). The member is generally a fine to medium grained, massive, sub-lithic sandstone with an argillaceous matrix rich in carbonate cement 376

and with minor siltstone and conglomerate bands throughout. The member is non-resistant to erosion and most of Lockyer Creek follows its surface exposure. The base of the Gatton Sandstone Member is not exposed in the Sandy Creek catchment. WINWILL CONGLOMERATE MEMBER The Winwill Conglomerate Member is differentiated from the Gatton Sandstone Member by its physiographic expression, a change in slope along the ridges. McTaggart (9) described the member as comprising white flaggy sandstones with fossilwood conglomerates and sandstones similar to the Gatton Sandstone Member being typically lithic and calcareous with many siUceous clasts. The conglomerate beds and abundant calcite cement form resistant horizons which can form hydrauHc barriers to groundwater flow in the alluvium. This warrants consideration of the Winwill Conglomerate Member as a separate hydrogeologic unit to the non-resistant Gatton Sandstone Member. The Winwill Conglomerate Member is visibly porous in core although its average porosity does not vary significantly from other members. However, horizontal permeability is significantly higher in the Winwill Conglomerate Member than its neighbouring members (10). MA MA CREEK SANDSTONE MEMBER The Ma Ma Creek Sandstone Member comprises flaggy lithic sandstones, shales, siltstones, and minor fossilwood conglomerate bands (9). The base of the member is gradational with the lower Winwill Conglomerate Member and its boundary is arbitrarily located where conglomerate bands decrease in abundance. The sandstones are micaceous and clayey (14) and the top of the unit is easily defined by the change from a non-resistant calcareous lithic sandstone to resistant quartzose sandstones of the Heifer Creek Sandstone Member. The Ma Ma Creek Sandstone Member has a higher proportion of siltstone and shale beds than other members of the Marburg Formation (15) and although quite porous (up to 19%), the Ma Ma Creek Sandstone Member has virtually no horizontal permeability (10). The presence of brackish water acritarchs (8) in the shales indicate that formation waters were of elevated salinity. HEIFER CREEK SANDSTONE MEMBER The Heifer Creek Sandstone Member consists of coarse to very coarse ferruginous siliceous sandstones (9) and minor shales. The Heifer Creek Sandstone Member is well cemented with calcite, limonite, or quartz which contributes to its resistant nature. The quartzose nature (>90%) of the member makes it resistant to erosion and its e^q^osure is characterised by steep hills and cliffs. The middle section of the Heifer Creek Sandstone has a higher proportion of shales, siltstones and clay and is more prone to landslips. It commonly forms moderate slopes or benches between the upper and lower sections which have greater aggregate sandstone thicknesses. The member most Hkely had a high initial porosity due to the high energy stream environment of deposition (3 &15) which is supported by the presence of coarse, poorly sorted sandstones, regdar conglomerate bands and pebble lags. Gray (10) noted that the Heifer Creek Sandstone Member has a higher horizontal permeability than the rest of the Marburg Formation although its average porosity is marginal to the other members. WALLOON COAL MEASURES The Marburg Formation is conformably overlain by the Walloon Coal Measures which, in the Lockyer Valley, rarely reach thicknesses greater than 60 metres (9). They consist of carbonaceous shales, siltstones, clayey sandstones, minor limestones, and coal seams. Porosity in the Walloon Coal Measures is low due to a high clay matrix and abundant calcite cement (16). QUATERNARY DEPOSITS Alluvial and minor colluvial deposits occur within the main valleys in the Sandy Creek catchment. The stream alluvium consists of well graded gravel, sand, and silt in a clayey matrix. The proportions of sand and gravel to clay depend upon the nature and proximity of the source rocks. Upper Sandy Creek contains narrow tracts of alluvium with high proportions of gravel and sand with cobble-size clasts at the base of former channel deposits. Finer sediments are Ukely to be deposited as point bar accretions and minor flood sheets. The clasts are dominantly basaltic in origin, however, the proportion of sandstone clasts increase downstream. Lower Sandy Creek alluvium has a higher clay content and sediments tend to reflect flood deposits. The increase in clay and silt is due to the input of non-resistant sediments and the increased distance from the resistant, sihceous formations and Tertiary basalts. The colluvial deposits are restricted to small areas on the lower slopes flanking the main stream alluvium, and are shallow deposits of poorly sorted gravels, sands, and silts. Surface Expression The Sandy Creek catchment contains two physiographic expressions reflecting the main geological units. The upper part of the catchment contains very narrow, shallow tracts of alluvium and colluvium between steep-sided, rounded hills of the Heifer Creek Sandstone Member and the Tertiary basalts of the Main Range Volcanics. The Heifer Creek Sandstone Member commonly forms steep clifif faces and the basalt flows form the resistant caps.

377

The lower catchment contains broad, deep tracts of alluvium forming extensive plains which gradually rise up to the low undulating hills comprising the less resistant Ma Ma Creek Sandstone, Winwill Conglomerate, and Gatton Sandstone Members. GROUPING OF CHEMICAL TYPES Significant variations in groundwater salinity have been recorded in a transect along the Sandy Creek catchment. In addition, variations in the chemical type of the groundwaters have also been observed to correlate with the occurrence of the four members of the Marburg Formation. From these members, six hydrochemical groups were defined. These groups were discriminated using Stiff diagrams, trilinear diagrams, and spatial relationships. Figure 2 shows the areal distribution of the hydrochemical groups and the typical variation in ionic proportions displayed by Stiff diagrams. This spatial distribution mainly demonstrates variations in Na+K and CI content. Cation trends show relative increases in Na+K to Mg with minimal variation in Ca. Anion trends similarly show relative increases of CI to HCO3 ^^^ change in SO4 proportions. The chemical type of each group was then identified on a Piper diagram (Figure 3). This plot displays a wide range of chemical types, which are incorporated in the following summary. Group 1 (Na,Mg,Ca-Cl,HC03) waters occur over the Heifer Creek Sandstone Member and are characterised by their unique Stiff pattems and isolated position on the Piper diagram. Magnesium is greater than calcium and sodium, while bicarbonate is greater than chloride and sulphate. Group 1 waters are distinctly different to the other groups and are probably chemically related to the basalts in the headwaters. Group 2 (Na,Mg,Ca-Cl) waters are distinguished by their similar Stiff pattems althou^ their Piper plots are not as differentiated as the Group 1 samples. The Group 2 waters occur over the upper part of the Ma Ma Creek Sandstone Member and show a slight increase in sodium and chloride contents from Group 1. This may indicate these waters are transitional between Group 1 and Group 3. Group 3 (Na,Mg-Cl) waters occur over most of the Ma Ma Creek Sandstone Member and all of the Winwill Conglomerate Member. These waters show a further increase in sodium and chloride proportions to the other ions. The Group 3 waters also have significantly higher salinity values than the other groups. Group 4 (Na,Mg-Cl) waters miake up the eastem margin of the Gatton Sandstone Member. These waters are characterised by the dominance of magnesium and chloride ions and a decrease in sodium and bicarbonate ions with respect to Group 3 waters. Group 5 (Na,Mg-Cl) waters occur over the westem margin of the Gatton Sandstone Member and, converse to Group 4, show an increase in sodium and bicarbonate and a decrease in magnesium and chloride. Group 6 (Na-HC03) waters are predominantly sodium bicarbonate type and occur near the junction of Sandy Creek and the larger Laidley Creek, being strongly influenced by the latter. On the Piper diagram (Figure 3), Groups 2,3 and 4 have a similar character and fall close together, but internal variations are shown by the Stiff pattems. Also, the groups can be differentiated by their geographical position (Figure 2).

378

Cations

% meq/L

Anions

Kgure 3. Piper diagram of Sandy Creek catchment samples with their respective hydrochemical groups indicated on the diamond field. The trilinear fields indicate the groundwater type of each sample. Hydrochemical Section The dominant water type is Na,Mg-Cl. However, the relative proportions of cations and anions tend to vary downstream indicating that the groundwaters are being modified throu^ various processes. To better observe changes in water chemistry with distance, data was plotted as a hydrochemical section. This method is preferred since Sandy Creek alluvium lies within a narrow, linear, isolated catchment. The hydrochemical longitudinal section (Figure 4) shows a relatively uniform trend between the major cations in the upper half of the catchment, but divergence from this in the lower half. The highest ionic concentrations occur over the Ma Ma Creek Sandstone and Winwill Conglomerate, and also the eastern part of the alluvium over the Gatton Sandstone Member. The southernmost upstream extent of alluvium does not continue to the Main Range Volcanics or the Walloon Coal Measures. Huxley (17) indicated that groundwater from the Walloon Coal Measures had a similar chemistry to the Marburg Formation in the nearby Condamine River Valley. Therefore, it is unlikely that groundwater from the Walloon Coal Measures could be discriminated, and in addition, its minor exposure in the study area would offer negligible effects on the overall hydrochemistry.

379

7060-

CO zo 2< y

50-

40-

30-

? s®

20-

10-

"c CO > D

1 1

50-

40CO

30-

20-

10"

—^

Heifer Creek Sandstone Member

1 2

1 3

1 4

1 5

1 6

1 ^ 1 1 1 1 1 7 a 9 10 11 12 13 kilometres downstream from first sample point

Ma Ma Creek Sandstone Member

(upstream)

Winwill Conglomerate Member

1 —— I 15 14

r16

1 17

1 18

1 19

1 20

Gatton Sandstone Member

(downstream)

Figure 4. Hydrochemical section of alluvial groundwaters in the Sandy Creek catchment demonstrating the variations in concentration and proportions of ionic components. The base geology is indicated undemeath the profile. The dashed lines indicate an altemative transect over the westem side of the valley. The solid lines refer to a transect over the eastem side. This demonstrates the variabihty of hydrochemistry over the Gatton Sandstone Member. Although the Main Range Volcanics are not in direct contact with the Sandy Creek alluvium, the chemistry of Group 1 waters reflect contributionsfiroma basaltic aquifer. These waters may also have leaked into the underlying Heifer Creek Sandstone Member before being discharged into the alluvium. The presence of chloride in these samples indicates that there is at least some contribution of waters from the sedimentary units. The alluvium in the upper part of the catchment may also contribute 'basaltic' type waters since the unconsolidated material consists predominantly of basaltic clasts. The low salinity and minimal change in hydrochemistry is probably due to the lack of labile components in the member. The hydrochemical section reveals a general increase in salinity and a change in chemical type over the Ma Ma Creek Sandstone Member. Group 2 waters show a proportional increase in sodium and chloride with respect to the other major ions. This is due to greater amounts of shale and mudstone layers likely to contain Na-Cl as interstitial waters and as adsorbed ions on clay surfaces. The Winwill Conglomerate Member initially shows an increase in salinity without much change in ionic proportions. The upper part of the Winwill Conglomerate Member appears to undergo the same dissolution of minerals that occurs in the Ma Ma Creek Sandstone Member. Also, the slightly more resistant nature of the Winwill Conglomerate Member may create a hydraulic restriction to groundwaterflowdown-gradient. This "bottleneck" effect could cause a decrease in flow velocity which would tend to prolong the residence time of waters in the aquifer, which can result in enrichment of solutes in the groundwater. Groundwater salinity peaks within the Winwill Conglomerate Member then generally decreases downstream with minor variations in individual ionic concentrations. The dilution of ions is probably due to the intrusion of lower salinity waters from the lower Winwill Conglomerate Member. 380

The alluvium over the Gatton Sandstone Member is much wider and more extensive than any other part of the catchment. The hydrochemistry of waters in this area is complex as it is a confluence and a number of sources contribute to the alluvium. The westem and eastem sides of the alluvium here show a marked difference in ionic distributions. The eastem margin (Group 4) decreases in salinity with some significant increases in Mg and CL TTie westem margin (Group 5) also decreases in overall salinity, however, Na and HCO3 ^^^ ^^^^ ^ increase with reductions in Mg and CI. The variation between the two zones is suggested as reflecting differences in the magnitude of groundwater mixing with upper catchment waters. Compared to Group 4, ion ratios from Group 5 are more consistent with Group 3. This could indicate that the main flow path of alluvial groundwater is along the westem margin of the lower catchment. Group 4 waters may be more strongly influenced by contributions from the Gatton Sandstone Member. The low salinity Group 6 waters most likely reflect a strong mixing with Laidley Valley alluvial waters which have a recharge area in the ranges well to the south. Talbot & Dickson (2) suggested that the Gatton Sandstone Member does not influence groundwater chemistry to any great extent in the Lockyer Valley. However, the general decrease in salinity indicates that some dilution process is occurring which is most likely caused by some form of recharge or mixing.

CONCLUSIONS

The hydrochemistry of alluvial groundwater in the Sandy Creek catchment during a low stream-flow period closely reflects the hydrochemistry of groundwaters dischargingfromthe bedrock aquifers. Six hydrochemical groups have been discriminated using Stiff patterns and a Piper diagram with reference to geographic position. Ion relationships are also consistent with the group identifications. The hydrochemical groups mainly represent variations in Na and CI content with respect to the other major ions. Spatially, these groups generally correspond with the different members of the Marburg Formation, but also represent contributions from adjacent alluvial aquifers. Alluvial groundwaters over the Heifer Creek Sandstone Member (Group 1) have a low salinity and a character typical of basaltic aquifers of the Main Range Volcanics (i.e. Mg > Ca > Na; HCO3 > CI > SO4). The composition of the groundwater continues to evolve down-gradient in response to the discharge of saline water from the Ma Ma Creek Sandstone Member and the Winwill Conglomerate Member (Groups 2 and 3). These aquifers contribute high amounts of Na, Mg, Ca, and CI in relatively equal proportions indicating that the mineral sources are reasonably consistent throughout both members. The broader zone of alluvium overlying the Gatton Sandstone Member has been divided into three hydrochemical zones. The eastem margin (Group 4) contains groundwaters with elevated Mg and CI whilst the westem margin (Group 5) has high Na and HCO3 levels with depressed Mg and CI concentrations. Group 5 waters possibly indicate a mixing trend between Groups 3 and 4. The other hydrochemical zone (Group 6) near the outflow of Sandy Creek is characterised by low salinity, Na-HC03 type waters which are likely to be derived from the mixing of groundwaters from the adjoining Laidley Valley alluvium.

ACKNOWLEDGMENTS

This paper is derived from a study which was supported in kind by the Department of Natural Resources (formerly the Department of Primary Industries - Water Resources) in Brisbane.

REFERENCES L

2. 3. 4. 5. 6. 7. 8. 9. 10.

MCMAHON G.A. 1995. Hydrochemistry of saline groundwater in the Sandy Creek catchment, Lockyer Valley, southeast Queensland. BAppSc(Hons) thesis, Queensland University of Technology, Brisbane (unpubl.). TALBOT R.J. & DICKSON T. 1969. Irrigation quality of some stream waters in the Lockyer Valley, south-eastem Queensland. Queensland Department of Primary Industries, Division of Plant Industry Bulletin No. 510. Queensland Journal ofAgricultural and Animal Sciences 26,565-580. ZAHAWI Z. 1975. Lockyer Valley groundwater investigations hydrogeological report. Geological Surrey of Queensland Record Series 36. lALBOT R.J., ROBERTS M.H., MCMAHON C.R. & SHAW R.J. 1981. Irrigation quality of Lockyer Valley alluvia bores during the 1980 drought. Queensland Agricultural College, Department of Biology. Technical Publication No. 5. REEVE R.& JONES R.M. 1984. The relationships between salinity and geology in six creeks in south east Queensland. CSIRO Division of Soils, Divisional Report 11. DIXON W. & CHISWELL B. 1992. The use of hydrochemical sections to identify recharge areas and sahne intmsions in alluvial aquifers, southeast Queensland, Australia. Journal ofHydrology 135,259-274. COX M.E. & MCNEIL V.H. (in prep). Salinisation and its extent, character and causes in the irrigated Lockyer Valley, southeast Queensland, Australia. WELLS A.T. & O'BRIEN RE. 1993. Fluvial architecture of Triassic-Jurassic sediments of the Bundamba Group in the northem part of the Clarence-Moreton Basin, Queensland. AGSO Record 45. MCTAGGART N.R. 1963. The Mesozoic sequence in the Lockyer-Marburg area, south-east Queensland. Proceedings of the Royal Society of Queensland 73, 93-104. GRAY A.R.G. 1975. Bundamba Group - stratigraphic relationships and petroleum prospects. Queensland Government Mining Journal 76, 310-324. 381

11.

CAMERON J.B. 1907. Second report on the West Moreton (Ipswich) coalfield. Geological Survey of Queensland Publication 344. 12. CRANHELD L.C., SCHWARZBOCK H. & DAY R.W. 1976. Geology of the Brisbane and Ipswich 1:250 000 sheet areas. Geological Survey of Queensland Report 95. 13. WELLS A.T., O'BRIEN RE., WILLIS I.L. & CRANFIELD L.C. 1990. A new lithostratigraphicframeworkfor the Early Jurassic units in the Bundamba Group, Clarence-Moreton Basin, Queensland and New South Wales. BMR Journal ofAustralian Geology and Geophysics 11,397-414. 14. MALLET A. 1963. The Geology of the Laidley Valley, southeast Queensland. BSc(Hons) thesis, University of Queensland, Brisbane (unpubl.). 15. GRIMES K.G. 1968. The geology of the Lockyer Valley area, south-east Queensland. BSc(Hons) thesis. University of Queensland, Brisbane (unpubl.). 16. CRANFIELD L.C. 1981. Stratigraphici:illingreport-GSQ Ipswich 24 and 25. Queensland Government Mining Journals!, 468-477. 17. HUXLEY W.J. 1982. The hydrogeology, hydrology and hydrochemistry of the Condamine River Valley alluvium. MAppSc thesis, Queensland Institute of Technology, Brisbane (unpubl.).

382

DISCRIMINATION OF GROUNDWATERS IN THE GREAT ARTESIAN BASIN, QUEENSLAND, USING THE CL/MG RATIO Vivienne McNefl* and Malcolm Cox Research Concentration in Environmental Earth Science, School of Geology, Queensland University of Technology. * also Resources Science Centre, Department of Natural Resources, Qld Summary Major aquifers in the Great Artesian Basin (GAB) differ in depth, chemistry, hydrology and areal extent. Management of the basin would be facilitated if the extent and usage of these aquifers were better defined. A series of reliability tests, developed for large data bases of chemical analyses for natural waters, was used to select the most reUable analyses for large data sets. These data were then used to test the Cl/Mg ratio in this groundwater environment. The benefit of the ratio is that it utilises stable constituents in chemical data sets, and has the potential to differentiate between aquifers and to indicate relative temperature differences. The Cl/Mg ratio alone is not definitive, but is a useful indicator of sedimentary formation and facies, and appears to respond to the effects of depth, residence time and mineral equilibria. Introduction The GAB (Figure 1) contains a number of major aquifers which differ in depth, chemistry, hydrology and areal extent. Many aspects of the general nature, structure and geochemistry of the GAB, and history of groundwater use, have been reported. Examples are Whitehouse and Ogilvie (1), Hillier (2), Torgersen et al (3), Habermehl (4), Quarantotto (5 and 6) and Denaro (7). There are numerous other studies of the GAB by State and Federal Governments and the private sector, in particular the petroleum industry. The basin underlies an area of some 1.7 million km' of which about 69% is in Queensland. The total thickness of sediments in the central Eromanga Sub-basin is over 1800 m and the south-eastern Surat Sub-basin is over 2100 m. Each sub-basin has a number of major aquifers, defined by several authors including Quarantotto (5) and (6), and Denaro (7). Those aquifers that are most widely represented in the sample set appear in Appendix A. Many bores were drilled without adequate logs or histories, and subsequent deterioration in the bore casing has made it difficult to identify the source or sources of the water. Management of the basin would be facilitated if the extent, characteristics, and usage of the aquifers were better defined. Large data bases held by the Queensland Department of Natural Resources contain chemical analyses for surface waters, and subartesian and artesian groundwaters. The data sets comprise over 100,000 sets of major ions with some minor elements and physico-chemical parameters. About 500 sets of these analyses represent samples of artesian groundwaters from the GAB. The data, collected since the 1950s, are unevenly distributed in space and time, and although unstable parameters cannot be considered reliable due to outdated sampling and handUng procedures, Rayment and Poplawski (8) show that conductivity and the major ions (excluding bicarbonate) are acceptable. Used appropriately, these data are an irreplaceable resource. A series of4ests enabled selection of valid analyses which can then be used for assessment by utilising their bulk inorganic constituents. The chloride / magnesium (Cl/Mg) ratio has proved an effective geochemical indicator elsewhere for regional reconnaissance assessment of a large volume of data (eg Cox and Thomas, 9) to establish thermal effects and mixing. As a consequence we here test its usefulness in the GAB. Water Chemistry The deep groundwaters of the GAB would be expected to differ from other natural waters in Queensland, principally because of the higher temperatures and pressures, the long residence times, and the mineralogy of the host sedimentary formations. Tiie elevated temperatures produce unique conditions in respect to mineral solubilities as well as groundwater migration. Aquifer temperatures are commonly around 50oC but are often higher and for some bores the water boils at the well head. The GAB can therefore be considered as a low temperature geothermal system from the aspect of geochemical investigation. The geothermal gradient is around 22°C/Km. QWRC and QDM (10) indicate spot temperatures which have been contoured (Figure 1.)

383

—1 liH)

ifrs

IfZ

-I

1— ISO

Measured Temperature "C

Isotherms ^

Groundwater Flow Direction

1 1 L__ Figure 1. The Great Artesian Basin with Data Sites, Flow Directions and Regional Groundwater Temperatures. In the most widely used artesian aquifers, total dissolved ions (TDI) are around 500-1000 mgL"^, dominated by sodium bicarbonate. Higher sulphate levels are encountered towards the west and in aquifers lower in the sedimentary sequence tend to contain less saline waters. Broadly, the shallowest aquifers of the GAB contain sodium chloride waters of around 1000-3000 mgL-^ TDI

384

100000 \-J fc: f Ecc O

10000,

• •

CI Mg

1000. 100. 10.

.... ^ • 0.1

100

1000

10000

100000

TDi in mg/L

Figure 2. Plots of chloride and of magnesium with salinity in artesian waters. The Cl/Mg Ratio The application of the Cl/Mg ratio is based on the principle that the CI" ion is stable in groundwater at moderate tenperatures. In contrast, the Mg^^ can be highly variable depending on the conditions of water-aquifer interaction (Schofield, 11). Reactions likely to be relevant in the GAB include ion exchange, chemical weathering of surface rocks in the vicinity of some of the intakes, and the formation of ilhte leading to the removal of Mg^* under moderate thermal conditions. Figures 2 showing the variation of magnesium and chloride in artesian waters demonstrates that althou^ both chloride and magnesium increase with salinity, the magnesium falls below 10 mgL"^ at constant salinities in the region of 1000 mgL"' TDI. Both chloride and magnesium from stream and alluvial waters in nearby recharge areas are both shown to rise Unearly on a log-log plot (Figure 3) The parallel band of low magnesium levels in the GAB aquifers suggests a loss of magnesium from the system under some conditions. This appeared to occur in aquifers where the temperature exceeded 30°C Assessment of the Data A set of general purpose, computer-based, bulk processing reliability tests have been developed for water chemistry data, using packages such as SYSTAT. These tests are based on chemical integrity, for example balance and solubility factors described in Stumm and Morgan (12), ranges observed in natural waters on a world wide basis, such as those described in Hem (13), Eriksson (14) and Hart (15), and on statistical variability within the population itself. These tests, although they cannot guarantee that the analysis represents the sample as it was collected, will identify errors in bulk chemistry and atypical outliers.

100

1000 10000 100000

TD! In mg/L

Figure 3. Variation of chloride and magnesium with salinityfromnear recharge areas. The table in Appendix A reveals that values for both Cl/Mg and TDI vary greatly within widespread formations, indicating changes in chemical environment along flow paths in the GAB. The most reliable data were then used to determine the Cl/Mg ratios. These were related, in terms of salinity, to samples of stream waterfromthe recharge zone in the Dawson River headwaters, as well as alluvial waterfromthe Callide Valley, a tributary of the Dawson. These analyses were selected from the Queensland Department of Natural Resources data bases, and average sea water from Hem (13). 385

10000

Alluvium • Artesian ' • Stream ^ Seawater '

1000.

GAB

v . •

100.

• . -^y-.'•

. ''J t ' " . . J . ' .

10.

0.1

^^-^-frv-

100

1000

• seawater

."•

•

10000

100000

TDI in mg/'L Figure 4. Trends in Cl/Mg ratios with salinity for connate waters and GAB waters. The Cl/Mg ratio plotted against total salinity (Figure 4) shows that few of the GAB analyses fall along the expected connate water trend between the end points of stream water (Cl/Mg = 2 to 6) to seawater (Cl/Mg =15) through alluvial water (Cl/Mg = 3 to 9). Instead, two other major trends were followed by the Cl/Mg ratio in artesian water as salinity increased. One included occasional low salinity (maximum TDI around 1000 mgL-1) waters, usually near basaltic recharge areas, where the ratio was below 1.0. For these waters, the proportion of magnesium increased relative to chloride with increasing salinity. The more general trend is a rise in the Cl/Mg ratio for the GAB water by orders of magnitude above the connate water trend, with 300 being the mean at TDI levels of 1000 - 2000 mgL-^ At these salinities, the cause is a loss of magnesium by adsorption or diagenesis, rather than a gain in chloride. A plot of magnesium v chloride (Figure 5) illustrates spatial and thermal affects on the ratio. We are currently attempting to better quantify values of the ratio relative to temperature by the use of geothermometers. 100000.0 • 10000.0 - • • 1 ^ _ 1000.0 100.0

Alluvium Artesian Stream Seawater

1 Fresh .. C Stream . ' ' • Water . ^ v y - /

10,0 1.0

0.1

100

1000-

10000 100000

CL Figure 5. Chloride v magnesixmi trends for connate water and GAB water. Contours of the Cl/Mg ratio (Figure 6) indicate anomalies in the vicinity of a recorded hot zone in the central basin, with temperatures ranging between 70°C and 90°C. One frequently sampled groundwater formation in this region is the Hutton which has At moderate levels of salinity in this area as shown on Figure 7. Figure 8 is a contour plot of the Cl/Mg ratio conpared to minimum groundwater temperatures as estimated from contours (Figure 1), and TDI. It indicates that up to about 30°C, total salinity appears to determine the Cl/Mg ratio, but above 30°C, temperature has an increasing positive effect.

386

•17

19 •21

•23

27 •29 138

140

142

144

146

148

150

152

Figure 6. Contours of Cl/Mg in the GAB, with measured Temperatures and Major Formations. 17

Main Aquifers Wyandra Hooray Mooga Hutton Precipice Longsight Unknown Measured "C ^100 Contours TDI

19 -

z• ,23 25 27 29 138

140

142

144

146

148

150

152

Figure 7. Contours of TDI in mg/L in the GAB, with measured Temperatures and Major Formations. Conclusions The above assessment supports the hypothesis that the Cl/Mg ratio has applications in the GAB for differentiating aquifers on the basis of temperature related rock-water reactions. Although the Cl/Mg ratio alone is not definitive, a relationship may be developed between this, and total salinity within regions to differentiate between deep and shallow formations. The broad trends in the Cl/Mg ratio (Figure 6) are of low values in basin edge recharge zones in the east, with a general increasing Cl/Mg in the southwest. There is a low value zone in the southeast, presumably influenced by magnesium 387

leached from weathered basalts partly or previously capping recharge areas. High values also occur in the north and may be related to both temperature and mineralogy.

600 TDf in nng/L

Figure 8. Variation in Cl/Mg with TDI and temperature contours in the Hutton Sandstone. The formation with the lowest ratio is the deep, Triassic Clematis Sandstone, and the formation with the highest is the shallow and relatively saline Cretaceous Mooga Sandstone in the eastem part of the Surat Basin. Acknowledgments Thanks are due to the Department of Natural Resources, Queensland, for permission to use the water quality database, and to Dr. W. P. Poplawski for his interest in the project. References 1 Whitehouse, F. W., and Ogilvie, C. (1954). "Artesian water supplies in Queensland." Report prepared by Dept. Co-ordinator General of Public Works for Queensland Parliament, Qld Govt Printer, Brisbane, 2 Hillier, J. R. (1992). "The Great Artesian Basin - a Need to Conserve Water." Water 19 Issue 6, pp42-43 3 Torgersen, T., Habermehl, M. A., PhiUips, F. M., Ehnore, D., Kubik, R, Jones, B. G., Hemmick, T., and Gove, H. E. (1991). "Chlorine 36 Dating of very old Groundwater. 3. Further Studies in the Great Artesian Basin, Australia." Water Resources Research 27, Issue 12, pp3201-3213 4 Habermehl, M. A. (1983). "Hydrogeology and Hydrochemistry of the Great Artesian Basin, Australia." In Proc. International Conference on Groundwater and Man, (Aust. Govt. Publ Serv,, AWRC and Aust Dept. Resource. Energy), AWRC Conf. Series No. 008, Canberra, ACT, Australia., 5-9 Dec, 3 pp 83-98 5 Quarantotto, P. (1986). "Hydrology of the Southeastem Eromanga Basin, Queensland." Report prepared by Qld. Dept. ofMmts, Record 1986/38. 5iip ISSN 1030-6938, Qld 6 Quarantotto, P. (1989). "Hydrology of the Surat Basin, Queensland." Report prepared by Qld. Dept. of Mines, Record 1989/26 58p ISSN 1030-6938, Qld 1 Denaro, G. (1991). "An Investigation into Aggressive C02 Corrosion in the Flinders Sub-Basin." Report prepared by Queensland Govt, Chem. Lab. for GCL Report Series No. 8., 58p PO Box 594, Archerfield Q 4108, Brisbane 8 Rayment, G. E., and Poplawski, W. R (eds) (1992). "Training notes on sampling for water quality monitoring." Report prepared by DPI ISBN 0724251111, 112p DPI, Brisbane, Qld 9 Cox, M. E. and Thomas, D. M. (1979). "Cl/Mg ratio of Hawaiian groundwaters as a regional geothermal indicator" Geotherm.Res. Counc. Trans 4, ppl45-148 10 QWRC., and QDM. (1987). "Groundwater Resources of Queensland." Report prepared by Queensland Water Resources Commission and Queensland Department of Mines, Mines Dept. State Series Map 4., Qld Govt. Printer, Brisbane. 11 Schofield, J. C. (1956). "Methods of distinguishing sea-groundwater from hydrothermal water." N.Z. J. Sci. Technol. 37, n. 5, pp 597-602 12 Stumm, W., and Morgan, J. J. (1970). "Aquatic chemistry, an introduction emphasising chemical equilibria in natural waters.", 580pp Wiley-Interscience, NY. 13 Hem, J.D. 1989: Study and Interpretation of the Chemical Characteristics of Natural Water. 3rd Ed. U.S Geological Survey Water-Supply Paper 2254. U.S. Government Printing Office ppl20- 123. 14 Eriksson, E. (1985). "Principles and applications of hydrochemistry.". Chapman and Hall, Uppsala, Sweden. 15 Hart, B. T. (1974). "A Compilation of Australian Water quality criteria." Report prepared by Australian Water Resources Committee, Report No. Z, 350 pp Australia. 388

APPENDIX A: TABLE 1: RANGES OF TDI AND CL/MG IN THE GAB Formation Age No. TDI Range Cl/Mg Range Cl/Mg Median Area Represented Rolling Downs Cretaceous i 1800 Wallumbilla-Doncaster Cretaceous 6 313-1452 26-1650 375 Wyandra Cretaceous 103 633-73,080 0.22-1076 118 SB Eromanga Gilbert River 25-2054 Cretaceous 7 3.67-191 6.7 Longsight Jur-Cret 10 648-39,088 14.5-408 117 Far West Eromanga Ronlow Jur-Cret 1 115 Hooray Cretaceous 117 346-22,838 3.98-1700 120 Sw to Central Eromanga Bungil Cretaceous 1 265 Kumbarilla Jur-Cret 4 965-1082 75-1250 425 Mooga - Orallo Cretaceous 22 766-2632 0.25-1403 90 Mainly Surat Gubberamunda Cretaceous 5 797-2396 39-600 128 Adori - Birkhead Jurassic 6 276-1228 13-1416 60 Hutton Jurassic 28 83-34,038 1.18-1563 22.5 E Edge of Eromanga Eulo-Queen Jurassic 3 300-325 7.25-13.75 7.5 Garraway 2 Jurassic 420 4.2 Boxvale — Evergreen Jurassic 6 331-717 1.96-5.93 3.9 Precipice Jurassic 16 145-3442 0.93-306 25.5 N Edge of Surat Clematis Triassic 16 135- -680 0.51-80 0.87 Yarrol Basin

389

EXTENSIONAL AND COMPRESSIONAL REGIMES OF THE CRETACEOUS CIRCUM-PACIFIC: IMPLICATIONS FOR THE EASTERN AUSTRALIAN PLATE Elizabeth L. MUler Department of Geologic and Environmental Sciences, Stanford University, Stanford, CA 94305 Since the rifting of Laurentia from Australia-Antarctica in the Late Proterozoic (700-500 Ma) (e.g. Hoffinan (1); Moores (2))), eastern AustraUa has faced a vast paleo-Pacific ocean basin. Oceanic crust within titis basin has formed, reconfigured and been subducted probably several times since this initial rifting. Circum-Pacific margin orogenic belts are the consequence of this history of subduction and are quite long-lived compared to other orogenic belts on earth. Although long-lived, the nature and style of orogenesis in circum-Pacific margin orogenic belts vary in time and space, reflecting the nature of deformation associated with the subduction process itself. In its recent history, two basic modes of subduction-related orogenesis have been recognized along the circum-Pacific. "Andean" or "Cordilleran" style tectonism occurs when the overiding continental plate undergoes intrusion by bathohths, intemal shortening, uplift and erosion, while "SW Pacific" or "Mariana" style tectonism occurs when the overiding plate undergoes rifting and extension. Uyeda and Kanamori (3) and Uyeda (4) linked theseftindamentaldifferences in tectonic style to the absolute motion of plates (trenchward motion of continental plate versus divergent motion between trench and continental plate) and Dewey (5)ftortherelaborated how these differences in motion led to the observed variability in tectonic style. It is clear that the presently observed patterns of deformation on either side of the Pacific have persisted back into geologic time and are mainly responsible for contrasts in the Mesozoic geology and tectonic history of Australia versus the American subducting margins. For example, the Mesozoic history of the Australian plate is characterized by rifting and formation of divergent margins while true mountain building occured in the American cordilleras. Divergence along the NW and W margins of Australia began 160 and 128 Ma. Slow spreading began in the Tasman Sea 96 Ma, more rapid spreading at 82-87 Ma, propagating northward to the Coral Sea 63-57 Ma and terminating at 57 Ma (e.g. Veevers(6)). Subduction, if it occured, took place outboard of these oceanic rift basins. In contrast, increased magmatism and shortening of continental crust in North America began at about 120 Ma and continued fairly continuously into the early Tertiary (55-60 Ma) after an episode of little or no subduction spanning -150-130 Ma. The growing data base on the timing of magmatism, orogenesis and plate margin tectonics along the North Pacific margin (with which this author is most famihar) suggests that during this time span, the change in deformational style (from shortening to extension) occurs in Alaska and the Bering Strait region. The transition is not a smooth one, but Cretaceous to Early Tertiary magmatism along the NE Russian portion of the N Pacific subducting boundary is associated with neutral to extensional tectonism. Extensional tectonism appears to become more important southward through China, the SW Pacific and Australia. When the observed changes in tectonism along the N Pacific margin are compared to available plate models for Pacific ocean floor (e.g. Engebretson et al. (7)), the resulting match is poor, pointing to problems in the plate models or perhaps greater complexity (more plates) during the Cretaceous. Interestingly, there seems to have been a reversal of the above pattem during the Paleozoic, when eastem Austraha e^erienced closure of marine basins, orogenesis and batholith intrusion (Lachlan and New England fold belts) while the North American Cordillera records a history of marginal basin formation or Mariana/SW Pacific style tectonics. The ultimate causes of differing tectonic styles on either side of the Pacific are unclear, but must be coupled and ultimately related to global plate motions and mantle convection. Future studies and syntheses that emphasize the comparison of the timing of events and the times of change in tectonic style along the perimeter of the Pacific provide potentially powerftd approaches to understanding the fundamental plate driving processes responsible for these differences and could provide important contraints on Mesozoic plate tectonic models for Pacific ocean floor. References 1 2 3 4

5 6 7

Hoffinan, P.F., 1991, Did the breakout of Laurentia tum Gondwanaland inside out? Science, v. 252, p. 1409-1419. Moores, E.M., 1991, Southwest U.S.-East Antarctic (SWEAT) connection: A hypothesis. Geology v. 19, p. 425-428. Uyeda, S. and Kanamori, H., 1979, Back-arc opening and the mode of subduction: Joumal of Geophysical Research v. 84, p. 1049-1061. Uyeda, S., 1987, Chilean vs. Mariana type subduction zones with remarks on arc volcanism and coUision tectonics, in. Monger, J.W.H., and Francheteau, J., eds., Circum-Pacific Orogenic Belts and Evolution of the Pacific Ocean Basin, American Geophysical Union, Geodynamics Series v. 18, p. 1-7. Dewey, J.F., 1980, Episodicity, sequence and style at convergent plate boundaries: Geol. Association Canada Special Paper v. 20, p. 553-573. Veevers, J.J., ed., 1984, Phanerozoic Earth History of Australia, Oxford University Press, N.Y., 418 p Engebretson, D.C., Cox, A. and Gordon, R.G., 1985, Relative motions between oceanic and continental plates in the Pacific Basin: Geological Society of America Special Paper 206, 59p.

390

THE MESOZOIC BASEMENT OF NEW ZEALAND

Nick Mortimer & Andy Iblloch, Institute of Geological & Nuclear Sciences, Private Bag 1930, Dunedin, New Zealand Summary The New Zealand part of the Gondwana margin contains a diversity of Mesozoic terranes and igneous suites. The geological record preserved in these rocks reflects processes of continental growth by tectonic and magmatic addition from the Triassic to the Early Cretaceous, and processes of continental dispersal in the Late Cretaceous. Some Mesozoic units have been identified offshore on the continental shelf, and are presumed to be disposed in belts of hundreds of km strike length some of which may extend into Antarctica and Australia. The purpose of this extended abstract is to (1) highlight selected recent devel opments in NZ Mesozoic geology, and (2) provide an up-to-date bibliography. Introduction Mesozoic rocks are common throughout New Zealand and most of them can be conveniently grouped into a number of fault-bounded metasedimentary and metavolcanic units (terranes), and into plutonic suites (Figures 1, 2, 3). The former mainly occur in the so-called Eastern Province, and the latter in the Median Tectonic Zone (MTZ) and Western Province. This longstanding tripartite subdivision of New Zealand into Eastem, Median and Westem parts is probably well known to the international community through the extensive review literature on New Zealand Mesozoic geology over the last 20 years (e.g. 1, 2, 3,4, 5, 6, 7, 8).

AUSTRALIAN PLATE

M60°E 160°W Figure 1. Location map of New Zealand in the SW Pacific Ocean. DR, Dampier Ridge; LHI, Lord Howe Island; LHR, Lord Howe Rise; WNR, West Norfolk Ridge; TB, Taranaki Basin; NI, North Island; ChP, Challenger Plateau; SI, South Island; CR, Chatham Rise; CI, Chatham Islands; CaP, Campbell Plateau; BT, Bounty Trough; BI, Bounty Island; M, Matakaoas; T, Tangihuas; C, Mt. Camel. Light grey areas are shallower than 2000m.

391

Late Cretaceous - Recent cover of East Coast Torlesse Terrane CaplesTerrane Maitai Terrane Murihiku Terrane Brook Street Terrane Median Tectonic Zone and related plutons (Median Suite) Separation Point and Rahu Suite plutons Undifferentiated Western Province (Paleozoic)

Haast Schist (Jurassic - Early Cretaceous) Esk Head melange Core complex detachments. Arrows indicate extension direction.

200 km Stewart Island Figure 2. Simplified geological map of South Island Mesozoic basement. T, Tapuaenuku; M, Mt. Somers; P, Paparoa Range; S, Sams Creek; K, Kirwans Dolerite; V, Victoria Range; C, French Creek; H, Hohonu Range; F, Fiordland. Buller and Takaka terranes of Western Province are not differentiated (see Cooper & Tulloch (12)).

392

Suite/intrusion

Province

Occurrence

-Area (km^)

-Age (Ma)

IVestem French Creek

Buller Ten-ane

100-83

Rahu Suite

Buller Terrane

750

115-105

Separation Point

WP&MTZ

4000

Kirwans Dolerite Sams Creek WP Median Suite

Buller Terrane Takaka Terrane WP (-MTZ)

most 125-110 some 160-105 -175 ?226 230,160,130

MedianTectonicZone Median Suite

Eastem Province

Mt Somers Tapuaenuku Matakoa/Tangihua Mt Camel/Houhora Hikurangi Plateau

MTZ (-WP)

Torlesse Rakaia Torlesse Pahau Miocene aikxtithon Waipapa? offshore Pacific

1 <1 50

3500 40 50 2000 10 350000

230-195, 160-130

Predominant lithologies

gabbro, A-type granites, peralk rhyolite, camptonite (hnb-bio) granodiorite(2 mica) granite 2-px diorite (WFO), tonal.. granodiorite-gran. dolerite aegirine, arfvedsonite gran. gabbro, diorite

gabbro-granite, diorite predom. Minor volcanics

90-80 basalt-andesite-rhyolite 100-95 gabbro, syenite, basanite ?100-?50 gabbro, dolerite, basalt 7130-100 basalt-andesite-rhyolite 125-115 basalt, dolerite, fubarite

Geochemistry

Sr initial

alkaline

0.707

cate alkaline

0.706-0.709

sodic, alkali-calcic

0.7040-0.7053

continent tholeiite peralkaline calc alkaline

0.7103-0.7108 0.7032? 0.7057

calc alkaline

0.7035-0.7044

tholeiitic alkaline MORB tholeiites tholeiitic intraplate tholeiite

0.7085 0.7028-0.7034 0.7027-0.7032 0.703-0.704 0.7036-0.7037

Table 1. Summary of New Zealand Mesozoic igneous suites. Excludes volumetrically minor imbricate slices of oceanic basalts in Eastern Province terranes. Area is very approximate onshore area, except for Hikurangi Plateau which is entirely offshore. Compiled from references given in text, and Barley et al (15), Graham & White (16), Isaac et al (17) and Tulloch and Mortimer (unpublished). See Figures 1 and 2 for locations. The general subject matter of "New Zealand Mesozoic geology" is now probably too large to summarise in single papers (but see e.g. Korsch & Wellman (9) for an impressive effort!). We use this extended abstract, not to provide a comprehensive review of Mesozoic New Zealand geology, but to draw attention to some important studies (recent and ongoing) that have revealed new developments of which this meeting may be unaware. As such, the bibliography in this extended abstract should be useful as a supplement to recently published syntheses until such time as somebody writes another major regional geological review. Any coincidence between the "important studies" and the authors' own publication lists is entirely intentional! Western Province Gondwana sequences

Precambrian cratonic basement is not exposed in New Zealand (10,11), but various Early Paleozoic terranes, and Paleozoic and Cretaceous plutonic suites have been correlated, at least in a general sense, with similar rocks in Australia and Antarctica (e.g. 12,13, Bradshaw, this volume; Weaver, this volume). Smale et al. (this volume) summarise the recent work on isolated Permian, Triassic and Jurassic rocks in New Zealand that provide specific Western Province-Gondwana stratigraphic links during this time period.

Mesozoic plutonic suites

Nomenclature for granitoid rocks in the Westem Province of New Zealand was proposed by Tulloch (3,14). Table 1 is an up-to-date summary of the characteristics of all the Mesozoic igneous suites from throughout New Zealand. Early Mesozoic igneous rocks of the Westem Province include half a dozen plutons of the so-called Median Suite (see below), and the volumetrically trivial Kirwans Dolerite (18) and Sams Creek dike (19). Voluminous I-type magmas of the Separation Point Suite, including granulites of the Westem Fiordland Orthogneiss (20, 21,22,23, 24) largely follow the 160-130 Ma pulse of the Median Suite (see below; 25, 26). However some Separation Point Type plutons are as old as 160 Ma (Tulloch, unpubl.). The Separation Point Suite is distinctively alkali-calcic with high Na, Al, Sr and low Y (23,24). Igneous rocks of the Median Suite (see below) represent an appropriate chemical and isotopic source for generation of the Separation Point Suite (21). Tectonic mechanisms for partially melting such rocks include collision and wholesale subduction of MTZ arc under the Westem Province (24), and crustal thickening by magmatic underplating (27). The relationship of granulite facies metamorphism to tectonomagmatic evolution of Fiordland is addressed by Mattinson et al (20), Bradshaw (22) and Gibson (28). In contrast to the wide distribution of the Separation Point Suite across the entire Westem Province and the MTZ, the intermediate I/S type Rahu Suite (115-105 Ma) is restricted to the (westernmost) Buller Terrane where it is associated with both crustal thickening and core complex formation (see below). Intrusion of 83 Ma A-type French Creek Suite granites likely overlapped with Tasman Sea oceanic crust formation (29). Late Cretaceous extension-related alkaline igneous rocks are also locally present in the Eastem Province (30, 31^ Cretaceous plutonism in New Zealand thus records a 50 m.y. transition firom arc magmatism and subduction to continental extension and ultimately rifting, during the period 135-85 Ma (27). 393

Metamorphic core complexes Tulloch and Kimbrou^ (32) described a classic metamorphic core complex of 110-85 Ma age in the Paparoa Range (Figure 2). Gibson et al. (33), Gibson (28) and Hill (34) described field relations in Fiordland (Figure 2) also compatible with a Cretaceous core complex, but with less of a metamorphic contrast across the decollement. Other possible areas of extensional core complex geometry in New Zealand include the Victoria and Hohonu Ranges (35; Figure 2). The core complexes are regarded as precursors to the opening of the Tasman Sea (10), and extension directions are approximately parallel to the 84-60 Ma spreading directions in the Tasman Sea (Figures 1,2,4). The New Zealand margin of the Tasman Sea can be regarded as a lower plate margin in terms of asymmetric rifting models (36), and in this respect it complements the eastem Australian margin. Similar uphft ages of metamorphic rocks (37) characterise the other complementary Mesozoic margin in West Antarctica, and again on this margin New Zealand may form the lower plate of an asymmetric rift (38). In contrast to the extension occuring over much of New Zealand in the Late Cretaceous (6; Laird, this volume), eastem North Island was clearly undergoing compression at this time (39). Eastern Province General summaries of Eastem Province geology are listed in the introduction. More specific work on particular themes includes radiolarian biostratigraphy (40), petrochemistry (41), Nd-isotope characteristics (42), Murihiku Terrane stratigraphy and paleontology (43,44), and Waipapa Terrane subdivision and correlation (45). Torlesse Terrane Studies of the Torlesse continue to play a pivotal role in the development and evolution of the terrane concept in New Zealand (1,42,46). Since the 1980s, the Torlesse has been subdivided into two major subterranes, Rakaia and Pahau (2, 6); the intervening Esk Head melange has also been treated as a separate subterrane by some authors. Recent developments in Torlesse geology include (1) location of the cryptic suture in the Jurassic Haast Schist between the Rakaia Torlesse and the Caples Terrane (47); (2) underthrust nature of the Torlesse Terrane beneath the Caples within the schist (48); (3) dating of single zircon and muscovite grains in the Torlesse (11, 49, Adams this volume) that point provocatively to a Queensland instead of Antarctic source for the Torlesse (cf. 9, 46); (4) proposal of another tectonostratigraphic subdivision of the Torlesse Terrane: the Late Jurassic-Early Cretaceous Waioeka Subterrane in eastem North Island (50) that suggests c. 300 km of dextral Late Cretaceous-Cenozoic strike slip displacement along the Gondwana Margin (Figure 4). Median Tectonic 2Lone Bradshaw (7) described the evolution of concepts on, and the then state of knowledge of the Median Tectonic Zone (formerly Median Tectonic Line). In the last few years considerable progress has been made in clarifying the nature of the MTZ, much of it in a 5 year programme of investigation by the Institute of Geological & Nuclear Sciences and coworkers. Below we present the emerging resultsfromthis research. The Median Tectonic Zone (MTZ) consists of a 10-35km wide belt consisting mainly of 230-130 Ma diorite-dominated subduction-related calc-alkaline plutons with subordinate, petrologically associated terrestrial volcanic and sedimentary sequences. Hydrothermal alteration of some rocks produced strong depletion, indicative of near polar paleolatitudes (51). Two main episodes of MTZ igneous activity are apparent, 230-195 and 160-130 Ma, with the latter dominating volumetrically (25). The older plutons occur tend to occur on the eastem side of the MTZ. MTZ igneous rocks have typical subduction related geochemical features and no unambiguous evidence for post -130 Ma subduction is found in New Zealand (27). The MTZ plutons apparently intrude the Pennian Brook Street Terrane of the Eastem Province. Several Carboniferous granites occur within the MTZ and also probably represent older basement. At least one of these Carboniferous granites has distinct A-type characteristics (Tulloch et al., unpublished data from Stewart Island),

394

50-

I5 ^5

«5

-EASTERN PROVINCE TERRANES -TorlesseTorlesse-

^^ I I 1 tI£ I

300-

'South Island-

I I I-i Is

t L - J mainly moderate quartz feldsarenite E a mainly high quartz feldsarenite

is i

-North Island-

1 mainly low quartz volcanic litharenite DOMINANT LITHOLOGY

£5

igneous

fTTTm

Haast Schist

Figure 3. Schematic diagram showing the time ranges of, and relationships between New Zealand's Mesozoic basement rocks. Sources given in text. Within each island, units are arranged in west (left) to east (right) order. similar to Carboniferous granites of the Toropuihi Suite of the Westem Province (12). This correlation supports the hypothesis that the plutons of the main MTZ belt developed in close proximity to the Westem Province. However, the generally sharp MTZ-Westem Province contact, and local presence of orthogneisses, suggest at least some tectonism along this boundary. Further work is continuing on the westem MTZ contact. The igneous-dominated MTZ, with its many intemal and extemal intmsive contacts is difficult to treat as a tectonostratigraphic terrane (Figure 3), although the minor volcano-sedimentary parts of it have been so-interpreted (e.g. Drumduan & Largs Terranes; Blattner & Williams (51); Bradshaw (7) and references therein). If terrane affiliation is demanded, then the MTZ should probably be considered part of the Brook Street Terrane. However, we here suggest that the compositionally distinctive Triassic-Cretaceous calcalkaline igneous rocks of the MTZ are conceptually best treated as an igneous suite (cf. Tulloch (14)). We provisionally suggest the name "Median Suite" to describe the igneous rocks that occur both in the main MTZ and, to a lesser extent, in the Westem Province (Figure 2, Table 1). Granite on Bounty Island (Figure 1) is included in the Median Suite (25); rocks in the Thurston Island area have a very similar magmatic history to the MTZ and have been treated as parts of the same Gondwana-fringing arc (13, 25; Bradshaw, this volume). For discussion of extensions of the MTZ towards Australia, see "Norfolk Ridge" below. Offshore geology Some 90% of the area of New Zealand's continental thickness crust lies underwater. Direct sampling of offshore basement is limited to a few dozen dredge hauls and oil exploration wells. Two areas are yielding new results.

Hikurangi Plateau

The Hikurangi Plateau is a recently recognised, triangular shaped region of 10-15km crustal thickness, that is attached to the Pacific Plate and Chatham Rise and is currently being subducted beneath North Island (Figure 1;

395

Figure 4. Rigid plate reconstruction for c. 100 Ma showing inferred offshore extent of the Median Tectonic Zone (vertical hatching) and core complex extension directions (arrows). No attempt made to close the Bounty Trougjh or correct for Cenozoic oroclinal bending in NZ, or Late Cretaceous extension. Northward displacement of East Coast North Island after Korsch & Wellman (9) and Mortimer (50). NEO, New England Ororgen; for other abbreviations see Figure 1. Wood & Davy (52)). A single dredge haul of highly altered basalts provisionally indicates that the Hikurangi Plateau is an Early Cretaceous (c. 115-125 Ma) Pacific Ocean large igneous province (53). The plateau and Chatham Rise are apparently draped by the same "KT" reflector of c. 80 Ma age (52). The arrival/collision of the plateau with the Torlesse Terrane sometime in the middle-Late Cretaceous might have played a major role in New Zealand magmatism and tectonics, but its precise role, if any, remains obscure at present (Figure 4). Norfolk Ridge

Three dredge hauls on the West Norfolk Ridge (Figure 1) have recovered rocks correlated with Brook Street metavolcanics and Median Tectonic Zone plutons (54). Ar-Ar dates of 146 Ma and 247 Ma on homblende from plutonic rocks (Mortimer, unpublished) support this correlation. Igneous rocks of Late Permian age are reported from a dredge haul on the Dampier Ridge (McDougall et al (55), Figure 1), and of Late Permian to Triassic age from the New England Orogen (metaluminous I-type Clarence River and Moonbi suites of Shaw & Flood (56)). The trend of the MTZ in New Zealand-West Norfolk Ridge does not obviously align with the strike of the Dampier Ridge/New England Orogen (Figure 4). Similar difi&culties in reconciling New Zealand-Australia correlative formations have previously been noted by Waterhouse & SiveU (57). Conclusions The New Zealand part of the Gondwana margin contains a variety of Mesozoic geological units. The rich geological record preserved in these rocks reflects Triassic to the Early Cretaceous continental growth by tectonic and magmatic processes, and Late Cretaceous continental dispersal. Major Mesozoic events include: 1. 2. 3. 4. 5. 6. 7.

Permian to Jurassic Gondwana basin deposition in Westem Province 230-130 Ma record of subduction in the Median Tectonic Zone c. 200 Ma Collision of Torlesse with Caples Terranes to form Haast Schist (Rangitata I orogeny of Bradshaw et al(2)) c. 125-115 Ma crustal thickening/sodic magmatism of Separation Point Suite c. 105-85 Ma pre-breakup extension and core complexes (Rangitata II orogeny) pre-80 Ma arrival of Hikurangi Plateau - effects unknown 85-60 Ma Tasman Sea spreading

Acknowledgements We thank Belinda Smith-Lyttle for drafting Figure 2, Rupert Sutherland for the reconstruction on which Figure 4 is based, and Bob Spark for comments on an earlier version of the abstract. References 1.

Coombs, D.S., Landis, CA., Norris, R.J., Sinton, J.M., Boms, D.J., Craw, D. The Dun Mountain OphioHte Belt, New Zealand, its tectonic setting, constitution and origin, with special reference to the southem portion. American journal of science 276: 561-603,1976. Bradshaw, J.D., Adams, C.J. & Andrews, P.B. Carboniferous to Cretaceous on the Pacific margin of Gondwana: The Rangitata Phase of New Zealand. In M.M. Cresswell & P. Vella (eds.), Gondwana Five, Balkema, Rotterdam, p. 217-221,1981. 396

3. 4.

6. 7. 8.

10. 11. 12. 13. 14. 15.

16. 17. 18.

19. 20.

21.

22. 23. 24.

25.

26.

27.

Tulloch, AJ. Granitoid rocks of New Zealand-a brief review/Geological Society of America Memoir 159: 5-20, 1983. Bishop, D.G.; Bradshaw, J.D.; Landis, CA. Provisional terrane map of South Island, New Zealand. In D.G. Howell (ed.), Tectonostratigraphic Terranes of the Circum-Pacific Region, AAPG Circum-Pacific Council for Energy and Mineral Resources Earth Science Series, No. 1, p. 515-521,1985. Sporli, K.B. & Ballance, P.F. Mesozoic-Cenozoic oceanfloor/continentinteraction and terrane configuration, SW Pacific area. In Z. Ben-Avraham (ed.). The Evolution of Pacific Ocean Margins. Oxford Monographs on Geology & Geophysics 8, p. 176-190, Oxford University Press, 1988. Bradshaw, J.D. Cretaceous geotectonic pattems in the New Zealand region. Tectonics 8: 803-820,1989. Bradshaw, J.D. A review of the Median Tectonic Zone: terrane boundaries and terrane amalgamation near the Median Tectonic Line. New Zealand journal of geology and geophysics 36:117-125,1993. Mortimer, N. Triassic to Early Cretaceous tectonic evolution of New Zealand terranes: a summary of recent data and an integrated model. In J.L. Mauk & J.D. St George (eds.). Proceedings of the 1995 Pacific Rim Congress, Australasian Institute of Mining & Metallurgy, Carlton, Victoria, p. 401-406, 1995. Korsch, R.J. & Wellman, H.W. The geological evolution of New Zealand and the New Zealand region. In A.E.M. Naim, EG. Stehli & S. Uyeda (eds.). The Ocean Basins and Margins, volume 7B, Plenum Publishing Corporation, p. 411^82,1988. Kimbrough, D.L. & Tulloch, A.J. Early Cretaceous age of orthogneiss from the Charleston Metamorphic Group, New Zealand. Earth and Planetary Science Letters v.95, p.130-140,1989. Ireland, T.R. Crustal evolution of New Zealand: evidence from age distributions of detrital zircons in Westem Province paragneisses and Toriesse greywacke. Geochimica et Cosmochimica Acta 56: 911-920,1992. Cooper, RA. & Tulloch, A.J. Early Palaeozoic terranes in New Zealand and their relationship to the Lachlan Fold Belt. Tectonophysics 214: 129-144, 1992. Bradshaw, J.D. Weaver, S.D., Pankhurst, R.J. & Storey, B.C. New Zealand superterranes recognized in Marie Byrd Land and Thurston Island. Geological Society of New Zealand miscellaneous pubhcation 81A: 86,1995. Tulloch, A.J. Batholiths, plutons, and suites: nomenclature for granitoid rocks of Westland-Nelson, New Zealand. New Zealand journal of geology & geophysics v.31, p.505-509,1988. Barley, M.E.; Weaver, S.D.; de Laeter, J.R. Strontium isotope composition and geochronology of intermediate-silicic volcanics, Mt Somers and Banks Peninsula, New Zealand. New Zealand joumal of geology and geophysics 31:197-206,1988. Graham, L J.; White, P.J. Rb-Sr dating of Rahu Suite granitoids from the Paparoa Range, north Westland, New Zealand. New Zealand joumal of geology and geophysics 33:11-22,1990. Isaac, M.J., Herzer, R.H., Brook, F.J. & Hayward, B.W. Cretaceous and Cenozoic sedimentary basins of Northland, New Zealand. Institute of Geological & Nuclear Sciences monograph 8,1994. Mortimer, N., Parkinson, D.L., Raine, J.I., Adams, C.J., Graham, LJ., Oliver, P.J. & Pahner, K. Ferrar magmatic province rocks discovered in New Zealand: Implications for Mesozoic Gondwana Geology. Geology 23:185-188, 1995. Tulloch, A.J. Petrology of the Sams Creek peralkaline granite dike, Takaka, New Zealand. New Zealand joumal of geology and geophysics 35: 193-200,1992. Mattinson, J. M.; Kimbrough, D. L.; Bradshaw, J. Y. Westem Fiordland Orthogneiss: Early Cretaceous arc magmatism and granulite facies metamorphism. New Zealand. Contributions to mineralogy and petrology 92: 383-392, 1986. McCulloch, M. T.; Bradshaw, J. Y.; Taylor, S. R Sm-Nd and Rb-Sr isotopic and geochemical systematics in Phanerozoic granulites from Fiordland, southwest New Zealand. Contributions to mineralogy and petrology 97: 183-195, 1987. Bradshaw, J. Y. Geology of crystalline rocks of northern Fiordland: details of granulite facies Westem Fiordland Orthogneiss and associated rock units. New Zealand joumal of geology and geophysics 33:465-484,1990. Tulloch, A.J. & Rabone, S.D.C. Mo-bearing granodiorite porphyry plutons of the Early Cretaceous Separation Point Suite, west Nelson, New Zealand. New Zealand joumal of geology and geophysics 36:401-408,1993. Muir, R.L, Weaver, S.D., Bradshaw, J.D., Eby, G.N. & Evans, J.A. The Cretaceous Separation Point batholith. New Zealand: granitoid magmas formed by melting of mafic lithosphere. Joumal of the Geological Society, London 152: 689-701,1995. Kimbrough, D.L., Tulloch, A.J., Coombs, D.S., Landis, C.A., Johnston, M.R. & Mattinson, LM. Uranium-lead zircon ages from the Median Tectonic Zone, South Island, New Zealand. New Zealand joumal of geology and geophysics 37: 393-419,1994. Muir, R.J., Ireland, T.R., Weaver, S.D. & Bradshaw, J.D. Ion microprobe U-Pb zircon geochronology of granitic magmatism in the Westem Province of the South Island, New Zealand. Chemical Geology (Isotope Geoscience Section) 113: 171-189,1994. Tulloch, A.J.; Kimbrough, D.L. Mesozoic plutonism in westem New Zealand - a 50 My transition from subduction to continental rifling on the Gondwana margin. The origin of granites and related rocks. Ill Hutton Symposium, USGS Circular 1129: 153-154,1995. 397

28.

29.

30. 31. 32.

33. 34. 35. 36. 37.

38. 39.

40.

41. 42. 43.

44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

Gibson, G.M. Uplift and exhumation of middle and lower crustal rocks in an extensional tectonic setting, Fiordland, New Zealand. In M.H. Salisbury & D.M. Fountain (eds.), Exposed Cross-Sections of the Continental Crust, Kluwer Academic Publishers, Netherlands, p. 71-101,1990. TuUoch, A.J., Kimbrough, D.L. & Waight, T.E. The French Creek Granite, north Westland, New Zealand - Late Cretaceous A-type plutonism on the Tasman passive margin (extended abstract). In G.J. van der Lingen, K.M. Swanson & RJ. Muir (eds.). Evolution of the Tasman Sea Basin, A.A. Balkema, Rotterdam, p. 65-66,1994. Tulloch, A.J. Alkaline plutonic and volcanic rocks of the Late Cretaceous Mandamus Igneous Complex, north Canterbury. New Zealand Geological Survey Record 43:15-23,1991. Weaver, S.D. & Pankhurst, RJ. A precise Rb-Sr age for the Mandamus Igneous Complex, North Canterbury, and regional tectonic implications. New Zealand joumal of geology and geophysics 34: 341-345,1991. Tulloch, A.J. & Kimbrough, D.L. The Paparoa metamorphic core complex, Westland-Nelson, New Zealand: Cretaceous extension associated with fragmentation of the Pacific margin of Gondwana. Tectonics 8: 1217-1234, 1989. Gibson, G.M., McDougall, 1. & Ireland, T.R. Age constraints on metamorphism and the development of a metamorphic core complex in Fiordland, southem New Zealand. Geology 16:405-408,1988. Hill, EJ. The Anita shear zone: a major middle Cretaceous tectonic boundary in northwestem Fiordland. New Zealand joumal of geology and geophysics 38: 93-104,1995. Tulloch, A.J. Precious metal mineralisation associated with the Cretaceous Paparoa metamorphic core complex, New Zealand. Proceedings of the 1995 PACRIM Congress, 575-580,1995. Lister, G.S.; Etheridge, M.A.; Symonds, PA. Detachment models for the formation of passive continental margins. Tectonics 10: 1038-1064,1991. Richard, S.M. Stmcture and cooling history of the Fosdick metamorphic complex, Marie Byrd Land, West Antarctica. In Recent Progress in Antarctic Earth Science, ed. Yoshida,Y. Terra Scientific Publishing Company, Tokyo, 1992, p. 289-294. Tulloch, A.J. A model for 2-stage asymmetric rifting of NZ-Australia-East Antarctica Marie Byrd Land. Geological Society of New Zealand Miscellaneous Publication 79A: 143,1993. Mazengarb, C. & Harris, D.H.M. Cretaceous stratigraphic and structural relations of Raukumara Peninsula, New Zealand: stratigraphic pattems associated with the migration of a thrust system. Annales Tectonicae 8: 100-118, 1995. Aita, Y. & Sporli, K.B. Tectonic and paleobiogeographic significance of radiolarian microfaunas in the Permian-Mesozoic basement rocks of the North Island, New Zealand. Paleogeography, Paleoclimatology, Paleoecology 96:103-125,1992. Roser, B.P. & Korsch, RJ. Provenance signatures of sandstone-mudstone suites determined using discriminant function analysis of major element data. Chemical Geology v.67, p.l 19-139,1988. Frost, C.D. & Coombs, D.S. Nd isotope character of New Zealand sediments: implications for terrane concepts and crustal evolution. American Joumal of Science 289: 744-770,1989. Ballance, PJ. & Caii5)bell, JX). The Murihiku arc-related basin of New Zealand (Triassic-Jurassic). In RE Ballance (ed. series ed. KJ. Hsu), South Pacific Sedimentary Basins. Sedimentary Basins of the World, 2. Elsevier, Amsterdam, p. 21-33,1993. Coombs, D.S. Does the Murihiku occupy a fore-arc basin? Geological Society of New Zealand miscellaneous publication 79A: 54,1993. Black, RM. The "Waipapa Terrane", North Island, New Zealand: subdivision and correlation. Geoscience Reports of Shizuoka University 20: 55-62,1994. MacKinnon, T.C. Origin of the Torlesse terrane and coeval rocks, South Island, New Zealand. Geological Society of America Bulletin 94: 967-985, 1983. Mortimer, N. & Roser, B.P. Geochemical evidence for the position of the Caples-Torlesse boundary in the Otago Schist, New Zealand. Joumal of the Geological Society of London 149: 967-977,1992. Mortimer, N. Jurassic tectonic history of the Otago Schist, New Zealand. Tectonics 12: 237-244,1993. Adams, C.J. & Kelley, S. Provenance of New Zealand Permo-Triassic metagreywacke terranes firom detrital mineral age populations. Geological Society of New Zealand miscellaneous publication 81A: 69,1995. Mortimer, N. Origin of the Torlesse Terrane and coeval rocks. North Island, New Zealand. International Geology Review 36: 891-910, 1995. Blattner, P. & Williams, J.G. The Largs high latitude oxygen isotope anomaly (New Zealand) and climatic controls of oxygen isotopes in magma. Earth & Planetary Science Letters 103: 270-284,1991. Wood, R.A. & Davy, B. The Hikurangi Plateau. Marine Geology 118:153-173,1994. Mortimer, N. & Parkinson, D.L. Hikurangi Plateau: a Cretaceous large igneous province in the southwest Pacific Ocean. Joumal of Geophysical Research 101: 687-696,1996. Mortimer, N. & Herzer, R.H. New petrological resultsfiromthe Vening Meinesz escarpment, and Norfolk and Three Kings Ridge areas: terrane correlations and constraints on Neogene reconstructions. Geological Society of New Zealand miscellaneous publication 81A: 148,1995. 398

55. 56. 57.

McDougall, I., Maboko, M.A.H., Symonds, PA., McCulloch, M.T., Williams, IS. & Kudrass, H.R. Dampier Ridge, Tasman Sea, as a stranded continentalfragment.Australian Journal of Earth Sciences 41: 395-406,1994. Shaw, S.E. & Rood, R.H. The New England Batholith, Eastem Australia: geochemical variation in time and space. Journal of Geophysical Research 86:10530-10544,1981. Waterhouse, J.B. & Sivell, W.J. New England Orogen and the Tasman Sea. In J.D. Kleeman (ed.). New England Orogen - Tectonics and Metallogenesis. Department of Geology & Geophysics, University of New England, Armidale,NSW,p. 199-204,1988.

399

BURIAL HISTORY OF THE EROMANGA BASIN IN NORTHEAST SOUTH AUSTRALIA

Reza Moussavi-Harami Petroleum Division, Mines and Energy South Australia, PO Box 151, Eastwood, SA 5063 Summary The intracratonic Eromanga Basin of Central Australia was formed during the Early Jurassic, approximately 193 m.y. ago. It contains four unconformity-bounded supersequences (sequence sets) which range from Jurassic to Quatemary in age. Burial history is interpreted from a series of diagrams generated from well data in the key tectonic and stratigraphic elements of the study area. During the Jurassic, subsidence rate was higher in the north (Patchawarra Trough) than south (Tinga Tingana Trougii). During the Early Cretaceous, subsidence in the northeast created more accommodation space for marine transgression from that direction into South Australia. From Early to Late Cretaceous time, the subsidence rate was high, probably due to rapid deposition of fine-grained siliciclastic sediments of marine and nonmarine successions, as well as tectonic subsidence. During the Cainozoics, subsidence rates were low to moderate due to sediment loading rather than tectonic sinking. The Jurassic source rocks of the Eromanga Basin are likely to have reached maturity sometime in Early to Late Cretaceous, when the subsidence rate was relatively high. Introduction The Eromanga Basin covers an area of 1 million km^ in Central Australia and about 360,000 km^ of the southem part of the basin is in northem South Australia. It is an intracratonic basin formed by cmstal downwarping of central Australia during the Early Jurassic (1,2). Sediments of the Eromanga Basin are Early Jurassic to Late Cretaceous in age; they lie unconformably over the Cooper Basin or older strata and are in tum unconformably overlain by Tertiary sediments of the Lake Eyre Basin. Thickness of sediments in the Cooper Basin region ranges from less than 1200 metres in the south to more than 2200 metres in the north. The purpose of this study is to interpret the burial history of the Eromanga Basin in the Cooper Basin region of South Australia. This will provide a better understanding of the sedimentation and subsidence rates which is very important in evaluation of structural episodes and timing of hydrocarbon generation and migration. Method and material Subsurface information comprises stratigraphic data from 1020 petroleum and stratigraphic wells, as well as structural contour maps of seismic horizons. Burial history diagrams were constructed for 14 wells. Ages used for construction of these diagrams were determined by calibrating the palynological work of Price et al. (3) and the published stratigraphic information (4,5,6,7,2, 8,9) to the geologic time scale (10,11). For calculation. Basin Mod software (1994) was used. Rocks type and thickness of each stratigraphic interval were determined from well logs and published regional stratigraphic studies. Missing and erosional intervals during each upUft event, are interpreted from the restored isopach map of each sedimentary package. Stratigraphic sequences Four unconformity-bounded supersequences, or sequence sets, are identified for the Eromanga and Lake Eyre Basins in northeast South Australia. These sets rangefromEarly Jurassic (J) to Quatemary (Q) in age. Sequence set J-K consists of a stack of three higher order sequences, a lower non-marine, a middle marine and an upper non-marine sequence. A restored isopach map of sequence set J-K indicates that the thickness of sediments in the Cooper Basin region ranges from 1200 to 3000 metres. The differences in thickness between troughs (such as Patchawarra) and ridges (such as Gidgealpa) are related to differential compaction rather than tectonic activity on a regional scale. It should be noted that the isopach map reflects the pre-Jurassic structural elements that were reactivated during Late Triassic uplift in northem South Australia and formed the basement of the Eromanga Basin. Sequence set T-Q consists of three unconformity-bounded sequences that were deposited in continental environments. They range from Late Paleocene to Quatemary in age and represent the Lake Eyre Basin. Thickness of the lower sequence on the restored isopach map ranges from 30 to more than 140 metres and the major depocentres were in the southem Nappamerri and Allunga-Wooloo Troughs, separated by the Moomba high. A restored isopach map of the middle sequence shows that the depocentre was in the Allunga Trough where more than 210 metres of sediments were deposited. Thickness of the upper sequence on an isopach map rangesfromzero to more than 60 metres in the study area. Thinning of this unit over ridges and highs indicates late Tertiary structural reactivation of the basement.

400

Burial history Burial history diagrams were constructed for 14 wells, representing the key tectonic and stratigraphic elements of the study area. Interpretation is presented here in ascending order from sequence set J-K to T-Q and mostly from the northem to southem parts of study area. Sequence set J-K Downwarping of the central part of the Australian continent in the Early Jurassic (approximately 193 Ma) created the Eromanga Basin and sedimentation in this basin continued through Late Cretaceous, without any major break. The Eromanga sequence set can be divided into three parts and the burial history is discussed in ascending order from the lower non-marine, through marine to the upper non-marine sequences. During deposition of the Early to Middle Jurassic Poolowanna Formation and Button Sandstone above the basal unconformity, the subsidence rate was moderate in the Patchawarra Trough (14.2m per m.y. in Cuttapirrie-1) and decreased toward the south to 3.1m per m.y. at the margin of the Tenappera Trough (well Mulga-2). There were numerous hiatuses and much reworking of sediments in this fluvial setting. During deposition of the Birkhead Formation, the rate of subsidence decreased and was at a maximum in the Nappsamerri Trough (about 13.5m per m.y. in Bulyeroo-1). The higher subsidence rate at the initial stages of basin formation is mainly related to rapid deposition of coarse-grained siliciclastic sediments, while the lower rate can be attributed to deposition of fine-grained, largely lacustrine, siliciclastic sediments. During deposition of the coarse-grained siliciclastic sediments of the latest Jurassic to earliest Cretaceous Namur Sandstone, the subsidence rate again increased from low to moderate. However, during deposition of the Murta Formation, the subsidence rate was low to very low. In general, during deposition of the lower non-marine sequence, the rates of subsidence was higher in the north and decreased toward the south. This can also be seen on the restored isopach map of the J-K sequence set, where the major depocentre was in the Patchawarra Trough. Due to comparatively greater downwarping of the northem part of the Australian continent in the Early Cretaceous, the epicontinental sea transgressed from the northeast into South Australia. The subsidence rate was initially low (about 4 to 5m per m.y.), during deposition of the Cadna-owie Formation and increased to an average of 45m per m.y., as younger marine units were deposited. Tbis increase was mainly related to rapid deposition of fine-grained siliciclastic sediments in marine environment, as well as basement subsidence. The higher sedimentation rate of the Wallumbilla Formation can possibly be related either to slumping and mass flow processes on a mildly unstable shelf, or to rapid deposition along reactivated major faults in the study area. The differences in subsidence rate between troughs and ridges can again be attributed to sediment loading and compaction and possibly to minor tectonic subsidence, as suggested by Zhou (12). During the late Albian to early Cenomanian deposition of the Mackunda Formation, the subsidence rate was higher in the Patchawarra Trough (average 52m per m.y. in Cuttapirrie-1) than in other parts of the study area. This can be mostly related to sediment loading and compaction, as well as tectonic sinking. This period of subsidence was followed by a fall of sea level and the study area was a site of continental sedimentation of the upper non-marine succession of the Eromanga Basin. During the Late Cretaceous, the rate of subsidence was much higher in the Patchawarra Trough (average 122.7m per m.y. in Cuttapirrie-1) and the Nappamerri Trough (113.4m per m.y. in Bulyeroo-1) than the intervening GMI Ridge (average 93.4m m.y. in Merrimelia-30). This was mainly due to differential compaction and very rapid deposition of fine to medium-grained siliciclastic sediments of the Winton Formation in fluvial and lacustrine environments, as well as tectonic sinking along the old Zones of weakness, as suggested by Moore and Pitt (5). Southward, the subsidence rate decreased to an average of 60m per m.y. in Kobari-1, but increased to 85.4m per m.y. in the Tinga Tingana Trough, which resulted from greater sediment loading and basement subsidence there. Thinning of the Winton Formation toward the margin of the basin is mainly related to severe erosion during Late Cretaceous to Early Tertiary time. The high subsidence rate during the Early to Late Cretaceous, is mainly related to rapid deposition of fine-grained siliciclastic sediments in marine and non-marine environments, as also suggested by Galla^er and Lambeck (13). In addition there was tectonic subsidence along reactivated major faults. It should be noted that during the Late Cretaceous, when thick siliciclastic sediments of the Winton Formation were deposited, the source rocks in deeper parts of the Eromanga Basin (Poolowanna, Birkhead and Murta Formations) reached maturity. This is supported by evidence of a rapid subsidence rate during this period as well as sediment thickness, as shown by this and other studies. This period of sedimentation and subsidence was followed by uplift and erosion from Turonian to Late Paleocene time. This is attributed to east-west directed basement compression on regional scale, starting from the eastem margin of the continent. It is calculated from a restored isopach map that the amount of section lost in the study area, at the Top Winton Unconformity, ranges from 150 to 440 metres. Sequence set T-Q From Late Paleocene to Middle Eocene, sedimentation resumed in northeast South Australia and coarse-grained siliciclastic sediments of the Eyre Formation were deposited in a fluvial environment. During this period, the subsidence rate was very low, ranging from 0.91 to 6.45m per m.y.; it is mainly related to sediment loading and compaction rather than tectonic activity. This period of sedimentation was followed by uplift and erosion from the Late Eocene to Early 401

Oligocene, related to epeirogenic movement. During this movement, portions of the Early Tertiary and Late Cretaceous sedimentary section were eroded from the crests of anticlines. Based on a restored isopach map of the Eyre Formation, it is calculated that the amount of section lost at this unconformity was probably less than 50 metres in the study area. Sedimentation again resumed in northeast South Australia and the Moomba high and the Wooloo Trough subsided at a higher rate (about 7.1m per m.y. in both Moomba-57 and Kirralee-1) than elsewhere. This higher rate can be attributed to deposition of Namba Formation dolomite in a large shallow alkaline lake that formed during the Late Eocene to Early Oligocene. Lower subsidence rates occur in the east and southeast of the Cooper Basin region (average 2.7,2.0 and L88 metres per m.y. in Strzelecki-5, Toolachee-36 and Mulga-2, respectively). This is related to both the gradual upUft of the area, which prevented the formation of dolomite, and to the lower rate of sedimentation, as shown by isopach thinning of the Namba Formation. After this period of sedimentation, epeirogenic uplift during the Early Pliocene, caused the study area to be a site of erosion and non-deposition. It is calculated that the amount of section lost at the top Namba Formation unconformity ranges from 15 to 54 metres. However, the amount of uplift and erosion was more severe in the westem Eromanga Basin, where up to 500 metres of sediments were eroded from the crest of the Dalhousie/McDills Ridge (14) and 350 metres of uplift and erosion occurred in the northem Flinders Ranges during Miocene time (15). In the Late Pliocene to Quatemary, the basin subsided and sedimentation resumed as fine to coarse-grained siliciclastics were deposited in fluvial and aeolian environments. The rate of subsidence was higher in troughs (such as Patchawarra Trough) than ridges (such as Murteree Ridge). Thinning (up to 20 metres) of this interval over structural highs indicates renewed movement along pre-existing structural trends, which are still active today. Conclusions (a) (b) (c) (d) (e)

During the Jurassic, the subsidence rate in the north (Patchawarra Trough) was hi^er than in other parts of the Eromanga Basin. The subsidence rate during the Early to Late Cretaceous was high due to rapid deposition of fine-grained siliciclastic sediments of the marine and upper non-marine successions of the Eromanga Sequence set J-K. During the Cainozoic, the subsidence rate was low to very low and was mainly related to sediment loading and compaction rather than tectonic activity. Deeply buried Jurassic source rocks of the Eromanga Basin probably reached the hydrocarbon generation window in the Late Cretaceous, when the subsidence rate was relatively h i ^ . Based on reconstructions of restored isopach maps, the amount of section lost at top Winton unconformity ranges from 150 to 440 metres, at top Eyre unconformity, less than 50 metres, at top Namba unconformity, 15 to 54 metres, and in the Late Pliocine, up to 20 metres.

Acknowledgment This study was founded as part of the South Australian Exploration Initiative. The author wishes to acknowledge Dr David Gravestock and Ms Elinor Alexander of MESA for their review of the manuscript and constructive suggestions during the course of this study. I also would like to thank Mr Rod Boucher for preparation of data and base maps, Ms Sam Harvey for digitising isopach maps, and Mrs Jeanette Bell and Mrs Melanie Lenuzzi for wordprocessing. References 1. 2.

3. 4.

5. 6.

Veevers, J.J. (editor), Phanerozoic Earth History of Australia. Oyford Geological Sciences Series, 2, Clarendon Press, Oj^ord, 1984, 418p. Krieg, G.W., Alexander, E.M. and Rogers, P.A., Jurassic-Cretaceous epicratonic basins, Eromanga Basin: In: Drexel, J.E and Preiss, W.V. (eds). The Geology of South Australia, Volume 2 - The Phanerozoic. Geological Survey of South Australia, Department of Mines and Energy South Australia, Bulletin 54,1995, pp. 101-127. Price, P.L., Filatoff, J., Williams, A.J., Pickering, S.A. and Wood, G.R., Late Palaeozoic and Mesozoic palynostratigraphical units. CSR Oil and Gas Division, palynology facility report, 274/25,1985 (unpublished). Wopfrier, H., Freytag, IB. and Heath, GJR., Basal Jurassic-Qetaceous rocks of westem Great Artesian Basin, South Australia: Stratigraphy and Qmirormorit American Association of Petroleum Geologists Bulletin, vol. 54,1970, pp. 383416. Moore, P.S. and Pitt, G.M., Cretaceous of the Eromanga Basin - implications for hydrocarbon exploration. Australian Petroleum Exploration Association Journal, vol. 24, 1984, pp. 358-376. Moore, P.S. and Pitt, G.M., Cretaceous subsurface stratigraphy of the southwestem Eromtanga Basin - a review: In: Lindsay, J.M. (editor). Stratigraphy, Palaeontology, Malacology. Papers in honour of Dr Nell Ludbrook. Department of Mines and Energy South Australia, special publication no. 5, 1985, pp. 269-286. Wiltshire, M.J., Mesozoic stratigraphy and palaeogeography, eastern Australia. In: O'Neil, B.J. (editor), Tlie Cooper and Eromanga Basins, Australia. Proceedings of the Cooper and Eromanga Basins Conference, Adelaide, 1989. Petroleum Exploration Society of Australia, Society of Petroleum Engineers, Australian Society of Exploration Geophysicists (SA branches), 1989, pp. 279-291. Gravestock, D.I., Callen, R.A., Alexander, E.M. and Hill, A.J., STRZELECKI, South Australia, Department of Mines and Energy South Australia, 1:250,000 Geological Series-Explanatory Notes, 1995, sheet SH 54-2. 402

10. 11. 12.

13. 14. 15.

Alexander, E.M. and Sansome, A., Lithostratigraphy and environment of deposition. In: Alexander, E.M. and Hibburt, J.E. (eds), The Petroleum Geology of South Australia, Volume 2: Eromanga Basin. Petroleum Division, Department of Mines and Energy South Australia. Report Book 96/20, 1996, pp. 49-86. Harland, W.B., Armstrong, R.L., Cox, A.V., Craig, L.E., Smith, A.G. and Smith, D.G., A Geologic Time Scale 1989. Cambridge University Press, 1990, 263p. Drexel, J.E, Preiss, W.V. and Parker, A.J., The Geology of South Australia, Volume 1 - Tlie Precambrian. Geological Survey of South Australia, Department of Mines and Energy South Australia Bulletin 54,1993, 242p. Zhou, S., Subsidence history of the Eromanga Basin, Australia. In. O'Neil, B.J. (ed). The Cooper and Eromanga Basins, AustraUa. Proceedings of the Cooper and Eromanga Basins Conference, Adelaide, 1989. Petroleum Exploration Society of Australia, Society ofPetroleum Engineers, Australian Society ofExploration Geophysicists (SA branches), 1989, pp. 329-335. Gallagher, K. and Lambeck, K., Subsidence, sedimentation and sea-level changes in the Eromanga Basin, Australia. Basin Research, vol. 2, 1989, pp. 115-131. Alexander, E. M. and Jensen-Schroidt, B., Eringa Trough E^loration Opportunity. Department of Mines and Energy South Australia. Report Book, 95/36, 1995, 99p. Foster, D.A., Murphy, J.M. and Gleadow, A.J.W., Middle Tertiary hydrothermal activity and uplift of the northem Flinders Ranges, South Australia. Insights from apatite fission-track thermochronology. Australian Journal of Earth Science, vol. 41,1994, pp. 11-17.

403

GOLD MINERALIZATION IN THE GRAHAMS CREEK VOLCANICS, MARYBOROUGH BASIN

Alasdair Murray, Consulting Geologist Rockcombe Pty Ltd, Brisbane Summary The Grahams Creek Formation is an Early Cretaceous volcanic unit at the base of the Maryborough Basin, consisting of sequences of intermediate to acid continental volcanics derived from local vents. It has no history of mining and has been largely ignored by modem explorationists. However, by the early 1990s two major companies had targetted the Formation and succeeded in locating hydrothermal systems and mineralization in several localities, at the best of which gold values up to 8ppm were obtained in a zone of epithermal veining and alteration. Such prospects did not meet the requirements of these companies, but much of the Formation remains underexplored. As a generality, the late discovery of this mineralization points to the presence of blind spots in conceptual target generation. Introduction The Grahams Creek Formation runs roughly north-south for a distance of some 150 km between Baffle Creek, north of Bundaberg, and the Gunalda/Glenwood area, north of Gympie, in south east Queensland (see Figure 1). It attains outcrop thicknesses in the order of 10 km in the south near Maryborougji and in the north at Bucca near Gin Gin, but it thins or is faulted out at a number of localities. Its true thickness is reported by Cranfield (1) as 1200m. Mineralization in the region at large is indicated by the presence of numerous small gold and copper deposits in the Permian Gympie Group to the west, and gold has also been mined from small shows in the upper part of the Triassic Brooweena Formation quite close to the contact with the Grahams Creek Formation near Cordalba. Graphite has been recovered from a group of workings on Mt Bauple, between Gunalda and Tiaro, also close to the Grahams Creek Fonnation contact (Figure 2). However there are no known workings in the Grahams Creek Formation itself and there is no history of mining or serious prospecting. Despite this, as will be discussed below, recent exploration has shown that the volcanics are appropriate hosts for mineralization. Regional Setting The Formation is located within the Gympie Province, where it unconfomiably overlies older sequences to the west. Over its northem section it abuts the Early Triassic Brooweena Formation, a thick series of sandstones and minor andesitic volcanics intmded by Mid Triassic granitoids. To the south, it overlies the Late Triassic Myrtle Creek Sandstone and Early Jurassic Duckinwilla Formation (including the Tiaro Coal Measures); these sedimentary units represent the first infill of the incipient Maryborough Basin. The Grahams Creek Formation runs the full length of the Maryborough Basin, and is unconformably overlain by conglomerates, sandstones and siltstones of the Maryborough Formation. It is intruded by Early Cretaceous dioritic and syenitic stocks. Regional tectonic features affect the distribution of the Formation. To the north, the Electra Fault system converges with it, and parts of its basal contact are truncated by the Bullyard Fault. To the south, repetition of the unit is caused by folding and axial plane faulting.

404

GRAHAMS CREEK VOLCANICS LOCATION AND REGIONAL SETTING 50

100

-J km

jRockhampton

Maryborough Basin

^ ^ Grahams Creek Fm

Maryborough

26-

Brisbane

—152' J—

Figure L

405

GRAHAMS CREEK VOLCANICS SIMPLIFIED GEOLOGY & PROSPECTS 10

20 I

\Km\ Cret

MaryboroughFm.

pr]

Jur-Cret.

Acid to intermediate intrusives.

Jur-Cret.

Grahams Creek Fm.

Jur.

Duckinwilla Group, incl Tiaro Coal Measures.

Triass.

Granitoid intrusives.

I ::v :| Triass.

Brooweena Fm.

Triass.

Brooweena Fm. volcanic unit.

r ^

Perm.

Gympie Group.

[in

Reported metalliferous mines, incl Au Cu Graphite.

Glenwood

Reported prospects within Grahams Ck Fm. (Map based on GSQ Bundaberg and Maryborough geological 1: 250 000 sheets)

406

Figure 2.

The Formation is a thick sequence of intermediate to felsic volcanic flows, pyroclastics and epiclastics belonging to a calc-alkaline continental suite (Cranfield,!; Murray,2.). It includes andesite, trachyandesite and trachytic flows derived from local vents. Locally, basalt and rhyolite have been described, the latter being particularly prolific in the centre and north of the belt. Textural variants of the rhyolite include crystal lithic breccias, fiamme-bearing welded ignimbrites, and flow banded/spherulitic units. A rhyolitic lava dome has been reported in the Bucca area. Andesites are dominant in the south of the belt and continue under cover from Maryborough to the coast near Tin Can Bay. The volcanics are intruded by Late Jurassic-Early Cretaceous microdiorites, microgranodiorites and syenites. Although published maps show these localized in the south east of the Formation, exploration workers have located them also in the north, and, by inference from aeromagnetics, at depth in the south west. Alteration and Mineralization Until 1990, no mineralization or hydrothermally altered zones had been desccribed in the volcanics, and no mines or prospects had been reported. Any exploration tenements which covered them did so fortuitously, as all economic interest was centred on the Pemio-Triassic Gympie Group and Brooweena Formation, on structural features such as the Electra Fault, or else on the Tiaro or Bumim Coal Measures. The graphite mined on Mt Bauple lies in a pendant of Tiaro Coal Measures and volcanics in a granodioritic and syenitic stock. Gold has been mined from small shows adjacent to the Grahams Creek Formation near Cordalba, and at one of these (the Mt Ideal/Bonnie Jan group) the Electrolytic Zinc Company found quite widespread alteration of epithemial type in the mid 1980s. Recent identification of hydrothermal activity and mineralization within the volcanics themselves are described below under "Recent Exploration Work". Recent Exploration Work Recent workers in the general area from 1980 on have included Geopeko, Electrolytic Zinc, Cyprus Gold, Keela Wee Exploration, RGC Exploration and BHP Minerals (various authors,3). However, most of these concentrated on the Permo-Triassic sediments to the west, where there are known occurrences of gold, copper and other metals. As mentioned above. Electrolytic Zinc did detailed work at the Mt Ideal group quite close to the volcanics contact. One reason why the earher companies did not recognize any potential in the Grahams Creek Formation was that the volcanics tend to form a narrow belt partly under more recent cover, and creeks which pass through this belt are often sourced outside it. Thus any volcanic-derived gold signature in stream sediments may be diluted out or wrongly attributed to the Permo-Triassic sediments. Two companies (RGC and BHP) specifically targetted the volcanics. RGC recognized geological analogies with the Triassic Aranbanga and North Arm volcanics west and south of this region respectively, in which both epithermal and mesothermal mineralization can be found. They used high density stream sediment sampling (0.5 kg of minus 200 micron, analyzed by Bulk Cyanide Leach) plus semi-regional traversing to locate targets, and followed these with detailed rock chip sampling and limited costeaning and drilling where appropriate. They succeeded in identifying several zones of alteration, spread throughout the volcanics, indicating hydrothermal activity of both epithermal and mesothermal style. These are described below under "Prospects Located". BHP aimed for large scale epithermal targets, mostly under cover, in the southem section of the belt, near Maryborough. They stream sediment sampled, using Bulk Cyanide Leach for gold and minus 200 micron for base metals, and interpreted published Bureau of Mineral Resources aeromagnetics to identify key structures, volcanic domains and concealed intrusive centres. Priority areas were drilled by Rotary Air Blast and weathered bedrock in the bottoms of the holes was sampled by Bulk Cyanide Leach, looking for alteration or geochemical haloes which might indicate major deposits. They succeeded in identifying multiple deep magnetic intrusives, and units of magnetic volcanics, as well as a zone of weak geochemical anomalism (see below). Prospects Located Indications of hydrothermal activity and/or mineralization have been recognized at separate localities more or less throughout the length of the volcanics. There are many gaps in the exploration cover of the Formation, so these by no means constitute an exhaustive list of occurrences, but are rather considered as examples of the styles to be encountered. Their locations are shown in Figure 2.

407

CHERRY CREEK Siliceous and argillic alteration in veined and brecciated rhyolitic rocks crop out over an area of 600m by 100m. Epithermal textures including colloform and crustiform quartz, chalcedony and carbonate replacement have been found, together with finely disseminated sulphides in veins and stockworks. Alunite, jarosite and possible adularia were identified by petrography. The prospect occupies the top of the volcanic sequence, probably near the palaeosurface, and is truncated by basal conglomerates of the Maryborough Formation. Rock chip samples were consistently elevated in gold, to a maximum of 8.6 ppm. Costeaning and drilling (two diamond holes) confirmed widespread alteration but failed to adequately source the epithermal textures and retumed a best intersection of only 3m at less than 0.2 ppm Au. PINE CORNER Weakly developed epithermal textures occur in sediments and tuff over approximately 200m by 100m. Silica/pyrite alteration is present, with anomalous rock chips up to 0.67 ppm Au. The rocks here are poorly ejqjosed, and no further work has been done. KANABAR This comprises a narrow structure usually less than 3m wide, striking over 1 km in length through the homfels zone surrounding a dioritic stock. The structure contains various breccias and veins of mesothermal style, and rock chips are anomalous in copper, lead, zinc and silver, with gold up to 0.40 ppm. Again, no follow up work has been done. MULLET Vein textures and alteration interpreted as high level epithermal occur over a small area. However, rock chips are anomalous in arsenic, bismuth, molybdenum and silver, which tends to suggest a deeper level system. TWENTY MILE CREEK Mesothermal style alteration, with silicification accompanied by pyrite and haematite, is found over an area 400m by 200m. Chalcedonic silica is also present. Rock chips are anomalous in arsenic, molybdenum and silver, with gold up to 0.21.ppm. MUNGAR A stream sediment sample retumed 1.35 ppb Au, by Bulk Cyanide Leach. This was statistically anomalous, but the water course which was sampled is only 2 km long and was not considered adequate to host a target of the dimensions sought in this particular programme, so it was not followed up. TUAN FOREST Several Rotary Air Blast drillholes bottomed in andesites with kaolinite-pyrite alteration. Geochemical analyses (by Bulk Cyanide Leach) retumed weakly anomalous values from the bottom samples, with levels in the 1 ppb order. It was concluded from the trend of results that prospectivity might increase to the south-east under cover, away from the area driUed. Conclusions The prospects described above failed to meet the target criteria of the particular exploration companies involved, and it is clear that their size and tenor are currently not outstanding. Nevertheless, each of these represents only a first pass of exploration, and almost no follow up has been done. The true extent and intensity of some of them may currently be understated. Furthermore, many areas within the Grahams Creek Formation have been inadequately covered, and these would benefit from additional semi-regional and/or detailed work. Some areas of the volcanics have not been e?q)lored at all, and their potential is wholly unknown at present. The distribution of the above prospects suggests that others will most probably occur within the unexplored areas. As a general comment, mineraUzation has been demonstrated in this extensive volcanic belt which has largely been ignored by explorationists. It is possible to speculate what other geological domains elsewhere might similarly have been passed over. References 1. 2. 3.

Cranfield,L.C., Maryborough 1:250000 Geological Series - Explanatory Notes. 1994, Geological Survey of Queensland. Murray,C.G., Metallogeny And Tectonic Development Of The Tasman Fold Belt System In Queensland, February 1986, in Geological Reviews, 1 (1986), 315-400. Various Authors, Open File Company Reports, Queensland Department of Minerals and Energy: C.R. Numbers 12861, 13900, 13901,15156,15884, 17361, 20928, 21267, 22678, 22932, 23248, 23545, 23844. 408

CLUES TO EARLY MESOZOIC STRUCTURAL EVOLUTION OF SOUTHEASTERN QUEENSLAND THROUGH INTEGRATED INTERPRETATION OF REMOTE SENSING AND AEROMAGNETIC DATA Colin Nash and Mike Jones World Geoscience Corporation Limited, Perth 65 Brockway Road, FLOREAT WA 6014 Summary Imaged aeromagnetic and radiometric data over the Mount Perry 1:100,000 Sheet area in southeastem Queensland have been interpreted and results correlated with an earher detailed photogeological study using Landsat imagery and small-scale airphotos (Nash, 1987). Aeromagnetic data provide excellent definition of all major lithostructural units in the study area, and systematic sinistral offsets of lithomagnetic domain boundaries of around 3km are visible along the Mount Perry Fault, which affects a Middle-Late Triassic volcanic/subvolcanic complex. The Mount Perry Fault formis part of a regional megascopic NNW-oriented lineament zone which controls emplacement of post-tectonic Triassic volcanic centres over a strike length of over 250km. Inferred strike-slip motion along NNW structures which control adjacent early Mesozoic sedimentary basins in southeastem Queensland, however, strongly suggests dextral transtension. It is concluded therefore that both sinistral and dextral strike-slip occurred during the Triassic. Introduction Mesozoic rocks in the Mount Perry 1:100,000 Sheet area of southeastem Queensland consist of Triassic ignimbrites and associated subvolcanic intrusives. Nash (1986, 1987) has shown that Triassic volcanism in the Mount Perry area is spatially related to a major NNW oriented megascopic structure (Mount Perry Lineament Zone-MPLZ), along which a number of lithologically similar complexes are emplaced (Fig. 1). Detailed photogeological studies in the Rosedale Sheet area immediately to the north have shown that some of the Triassic intrusive complexes are associated with major WNWoriented silicic dyke swarms (Nash, 1988). Structural analysis based on photogeological interpretation suggests that the previous model of Triassic dextral shear presented by several writers (Evans and Roberts, 1980; Flood and Garces, 1986; Murray, 1986) does not fit the observed lineament pattems (Nash, 1986), althou^ no definitive altemative explanation could be formulatedfiromthe remote sensing data. An aeromagnetic survey of the Mount Perry area was flown by World Geoscience Corporation Limited in 1988. Survey specifications included N70E flight line orientation, 500m line spacing and 80m flying height. Resulting data were levelled and imaged by WGC to produce a standard suite of images for interpretation. These included high-pass filtered image data in the form of a First Vertical Derivative (IVD) greyscale image and a colour Total Magnetic Intensity plus greyscale IVD image (TMI+lVD). Images were produced at 1:100,000 scale for qualitative interpretation as the main part of the present investigation.

409

Figure 1. Location of study area and distribution of Mesozoic volcanic rocks (from photogeological study by Nash, 1986; 1987). Heavy arrows show NNW-trending Mount Perry Lineament Zone which controls emplacement of topographically-prominent Middle-Late Triassic epizonal plutons volcanic complexes, followed by sinistral transpressional faulting. Photogeological Investigation (1987) A comprehensive photogeological study of the Mount Perry region was undertaken as part of a regional research project by Nash (1987). This study attempted to analyse regional structural pattems visible in remote sensing data and to use these to explain the late Paleozoic and Mesozoic evolution of the New England Orogen in southeastem Queensland. Photogeological faults andfracture-tracedata were found to reveal very pronounced preferred orientations. In particular, it was found that most small fractures and faults were oriented WNW (Nash, 1986; 1987). These probable extensional fracture trajectories could be interpreted as indicative of transtensional strike-slip motion along the regional NNWtrending MPLZ. Supporting evidence for the assumption of regional extension was found from detailed field studies in the Rosedale area, where systems of WNW-oriented silicic (rhyolitic) dykes are spatially associated with Middle-Late Triassic volcanic centres (Nash, 1988). No clear evidence for the direction of strike-slip motion along the MPLZ, however, was forthcoming from the photogeological study. This is largely due to the paucity of recognisable stratigraphic markers in the pre-Mesozoic basement, which consists largely of Paleozoic flysch and homogenous late Paleozoic granitoids with subdued expression. The Mount Perry Lineament Zone has also been the locus for subsequent Cenozoic extension and sedimentation, which further limits observation.

410

Aeromagnetic Interpretation (1996) The release of high-resolution aeromagnetic data over southeastern Queensland as part of the Geomap 2005 project (ASEG, 1995) has revealed a wealth of new lithostructural information over the New England Orogen and prompted a re-evaluation of the Mount Perry data (Fig. 2).

—

Major thrust Linear magnetic trend

Figure 2. Qualitative interpretation of imaged aeromagnetic data acquired by World Geoscience Corporation in 1988 over Mount Perry area. Sinistral motion along the Mount Perry Fault is clearly shown by the systematic offset of lithomagnetic domains (described in Table 1).

411

Table 1. Lithomagnetic and radiometric characteristics of the Mount Perry survey area. UNIT

MAGNETIC CHARACTER

RADIOMETRIC CHARACTER

STRATIGRAPHIC CORRELATION

TRv

Moderately magnetic bodies with jointed appearance

TRg

Strongly magnetic bodies

Mid-Late Triassic epizonal Hogback Adamellite and Coeval Aranbanga Volcanics Subvolcanic intrusives marginal to Hogback pluton

Circular magnetic bodies

Weakly magnetic homogenous bodies; younger WNW dykes Strongly magnetic contact zone around Hogback Pluton Weakly layered magnetic unit Strongly magnetic unit in east of area

Moderate-strong total count anomalies associated with potassic granitoids Strong total count anomahes associated with potassic granitoids Very low total count anomahes over mafic intrusive bodies High K and moderate total count anomalies; strong K zone adjacent to Mt Perry Fault Moderate-low radiometric response Foliated appearance; some K-rich beds Very low radiometric response (due to presence of mafic rocks?) Foliated appearance; some K-rich beds

Permo-Carboniferous Goodnight Beds (mainly flysch) Accreted Permian Gympie Terrane (includes ultramafics)

CPhf

CP P

Weakly-magnetic layered sedimentary unit

Gabbroic intrusions (Permian?) Syntectonic Permian intrusives

Homfelsed Goodnight Beds

Devonian-Carboniferous Curtis Island Group flysch

Interpretation of aeromagnetic data in the Mount Perry area has as its starting point the recognition of lithomagnetic domains, as described in Table 1. The table reveals the comparatively strong magnetic susceptibilities of the Triassic volcanic and subvolcanic rocks compared to the Paleozoic 'basement' into which they are emplaced. This critical factor enables the mapping of contacts between Mesozoic and older rocks to be performed quite easily, and for any fault offsets of these contacts to be observed. In this way, it is possible to show that some 3km of sinistral motion has taken place along the Mount Perry Fault (Fig. 2). Comparison with recent re-mapping of the Mount Perry 1:100,000 Sheet area by the Geological Survey of Queensland indicates that all the lithomagnetic units observed on imagery correspond to field mapped lithotypes. The post-tectonic Hogback Adamellite and coeval Aranbanga Volanics (unit TRv in Fig. 2) form moderately-strongly magnetic domains with a characteristic joint pattem typical of plutonic rocks. Marginal intmsive bodies (unit TRg) are strongly magnetic and provide excellent indicators of fault offset. The pre-Triassic lithologies in the Mount Perry area all have rather flat magnetic expressions with the exception of (a) a zone of apparent homfelsing of the Permo-Carboniferous Goodnight Beds around the northem and eastem margins of the Hogback/Aranbanga volcanic centre, and (b) two circular gabbroic plutons of inferred Permian age (Fig. 2). The allochthonous Permian Gympie Group in the eastem extremity of the study area is strongly magnetic due to the presence of mafic and ultramafic rocks of high susceptibility. Radiometric Interpretation The results of the aeromagnetic interpretation have been combined with radiometric data to help clarify the lithological distribution. Table 1 also summarises the radiometric characteristics of the main lithological groups which have been derived from a ternary image presentation (K, Th, U, as RGB) and pseudocoloured total count presentation. Preliminary results indicate strong potassic and total count responsesfi:omTriassic volcanics and subvolcanic granites can be subdivided into lithological responses and those which may result from alteration. Strong potassium anomalies which may relate to hydrothermal alteration phenomena as they exhibit clear cross-cutting relationships are associated with a number of these intrusive/extrusive complexes. At the opposite end of the spectrum of responses mafic rocks have virtually no radiometric response and this applies to Permian gabbro/diorite intrusives as well as mafic volcanics in the accreted Gympie Terrane. Sedimentary rocks from both the Devonian Curtis Island Group and Permo-Carboniferous Goodnight Beds exhibit a foliated appearance in the radiometric data and are locally enriched in potassium. This may be the result of interbedded felspathic arkoses.

412

Tectonic Significance Aeromagnetic data over different parts of the MPLZ suggest that sinistral strike-slip occurred after emplacement of the Middle-Late Triassic Hogback Ad^ellite and related ignimbrites (Aranbanga Volcanics) in the Mount Perry area. Similar sinistral displacement of the Takilberan Adamellite to the north is visible on recent Geomap 2005 imagery (ASEG, 1995). Silicic dyke pattems at on the Rosedale 1:100,000 Sheet area provide unequivocal evidence of WNW extensional faxilting (Fig. 3(a)) during the Triassic volcanic episode (Nash, 1988), which may be related to transtension along the MPLZ. These observations suggest a possible altemation of dextral and sinistral tectonic transport along the Mount Perry Lineament Zone and parallel structural lineaments during the Triassic. It has been suggested that accretion of the allochthonous Gympie Terrane took place at the close of Early Triassic times, since the NNW-trending Early-Middle Triassic Esk Trough sequence, which was deposited in a dextral strike-slip basin (Korsch et al., 1989), is strongly deformed (Murray, 1986). Dextral strike-slip has also been involved in the evolution of the undeformed Late Triassic Tarong Basin (Flood and Garces, 1986). Most writers to date have therefore tended to regarde dextral strike-slip as the dominant mechanism in Triassic accretion and strike-slip basin formation (Evans and Roberts, 1980; Korsch et al., 1989; Flood and Garces, 1986). This contradicts our observations, possibly suggesting a reversal of tectonic transport direction in Middle-Late Triassic times. Some evidence for reversal of movement along the Mount Perry Fault is to be seen from the map view of the structure (Fig 3(b)) which suggests a right (eastward)-stepping geometry compatible with dextral rather than sinistral transport, which accords with the east-stepping pattem of basin development in southeastem Queensland noted by Korsch et al., (1989). It is possible therefore that the MPLZ evolved as a dextral transtensional feature which facilitated the emplacement of post-orogenic volcanic complexes, followed by sinistral transtentional faulting. (b) MPLZ (335®)

\ Triassic dyke and fracture direction (120®)

Figure 3. (a) Summary of photogeological lineament information in Mount Perry and Rosedale Sheet areas (Nash, 1987) and possible dextral transtensional model; (b) interpreted east-stepping plan view of Mount Perry Fault system from aeromagnetic interpretation, possibly indicative of sinistral transtension. Acknowledgements The writers wish to express their thanks to the management of World Geoscience Corporation Limited for permission to use the Mount Perry aeromagnetic survey data. References Cited ASEG, 1995. Queensland exploration initiatives. Preview, 58, 21-24. Evans, P.R. and Roberts, J., 1980. Evolution of central eastem Australia during the late Palaeozoic and early Mesozoic. J. Geol. Soc. Aust, 26, 325-340. Flood, RG. and Garces, B.L, 1986. The Tarong Basin, Queensland. In: W.F. Willmott, ed., 1986- Field Conf.-South Burnett District. Brisbane, Geol. Soc. Aust., 77-81. Geological Survey of Queensland, 1986. Mount Perry 1:100,000 Geological Sheet (preliminary). Korsch., R.J., O'Brien, P.E., Sexton, M.J., Wake-Dyster, K.D. and Wells, A.T., 1989. Development of Mesozoic transtensional basins in easternmost Australia. Aust. J. Earth Sci., 36,13-28. Murray, C.G., 1986. Metallogeny and tectonic development of the Tasman Fold Belt system in Queensland. Ore Geol. Rev., 1,315-400. Nash, C.R., 1986. Permo-Triassic evolution and metallogenesis of the Rockhampton-Maryborough area, Queensland-a photogeological investigation. Ore Geol. Rev., 1,401-412. Nash, C.R., 1987. Late Palaeozoic-Cainozoic evolution of the New England Orogen in southeastem Queensland-a photogeological investigation. Unpubl. PhD Thesis, Macq. Univ., Sydney, 277pp. Nash, C.R, 1988. Photogeological identification and significance of Mesozoic intrusions in the Rosedale 1:100,000 Sheet area, southeastem Queensland. Proc. R. Soc. Qld., 99,1-7. 413

GEOPHYSICAL TRIALS FOR FAULT LOCATION AT CALLIDE COALFIELDS' TRAP GULLY MINE CALLIDE BASIN, EAST CENTRAL QUEENSLAND Wes Nichols and Mine Geologist Callide Coalfields Pty Ltd. Biloela, Qld Keywords: Callide Coalfields, Coal Downhole Seismic, Geophysics, Mini-Sosie, Optimum Offset, Radio Imaging, RIM, Seismic Refi-action, Seismic Reflection, Tomography, TSIM, Uphole Seismic, VLF-EM, VSP Seismic Summary Geophysics is commonly used to interpret structure and detect anomalies between widely-spaced boreholes. The common techniques utilised usually present good results at a significant cost saving (compared to the cost of drilling to locate structures and anomalies). Widely-spaced exploration drilling at the Trap Gully 'B' area had confirmed that a fault with considerable throw (up to 20m, down to the east) existed in advance of the current pit. However, neither the exact location, dip direction nor dip angle on this fault were known. Due to the advance of mining, the fault was expected to become problematic for highwall location sometime in 1998. In early 1990, mini-sosie trials were conducted over the area in an attempt to define the fault, but, due to near-surface unconsolidated sandstone layers which constituted seismic velocity inversions, these trials were not successful. Even reprocessing of the original seismic data at a later date did not yield any definitive results. Further seismic trials, where sound sources (detonators) were initiated below the suspect unconsolidated sandstone layers, were slightly more conclusive. Due to the encroachment of mining closer to the fault area, in 1995, VLF-EM (TSIM) surveys were then conducted to locate the lateral manifestation of the fault at the surface/near-surface. Following this, several pairs of boreholes (~50m apart) were drlled, with one hole up-dip and the other down-dip of the indicated fault location. This drilling confirmed the TSIM results. The boreholes were left open and were capped. RIM (radio imaging) surveys were then conducted and provided vital information on the dip and structure of the fault. It was intended to conduct TEM surveys over the area, but these were not conducted in time for the publication of this paper and will be conducted later this year. Also, at the time of writing this paper, downhole dipmeter, sonic and teleview surveys were being conducted in the boreholes along TSIM lines 2 & 3. It is expected that the resultsfiromthese surveys will show evidence of faulting within these boreholes. Location Callide Coalfields is located some 105 kilometres south-west of Gladstone and 450 kilometres north-northwest of Brisbane in East Central Queensland (Figure 1). The coalfield covers an area of about 18,000 hectares and is located within a north-west to south-west trending synclinal basin 22.5 kilometres long by 8 kilometres wide. Present open-cut mining operations are located in the Dunn Creek, Trap Gully and The Hut areas in the south as part of the Callide Mine, and at Boundary Hill in the north-west of the basin.

414

Figure 1. Callide Coalfields Location Plan The Callide Basin is a fault bounded synclinal basin of Middle to Upper Triassic age. It is elongated along a north-west trending axis and asymmetric with strata dipping typically < 10 degrees. The Triassic Coal Measures form a grossly fining upward megasequence with thick conglomerate sequences at the base passing upward somewhat abruptly into coal bearing, fine grained clastic rocks. Thick, laterally extensive quartzose fluviatile "sheet" sandstones of the Precipice Sandstone were deposited over the Coal Measures by a braided river system in the Early Jurassic. The Callide Basin sequence was tilted to the south-west during the Tertiary, exposing and eroding the eastem margins and subsequently burying the westem margins under Tertiary sediments of the Tertiary Callide Basin. Volcanic activity exhibited by basalt 415

flows that blanket much of the eastern area, accompanied the tilting. Basalt dykes and stocks also intruded the Callide Coal Measures to the north at Boundary Hill [Biggs and others, 1989 (1); Nichols, 1996 (2)] The Trap Gully Mine is situated approximately halfway along the westem margin of the Callide Basin. The Callide Coal Seam Member at Trap Gully has been divided into a number of plies based mainly on downhole geophysical log signatures and recognised seam splitting. However, within the twenty year mine plan, the seam will be mined as a single mining section averaging 13 to 17m in thickness. The Trap Gully area has been divided into A and B Pits (Figure 2) due to significant stmctural disturbances through the centre of the area. The A and B Pits are separated by a series of closely spaced north north-west trending normal faults with cumulative throws of up to 15m. Seam dips in the mining area are typically 5 to 10 degrees, steepening to greater than 50 degrees beyond the mining area along the south-west basin margin. The A Pits are punctuated by a series of wide spaced low angle reverse faults with throws varying from 2 to 10m. The B Pit contains dips of 2 to 20 degrees increasingfiromsouth-east to north-west within the mining area and dips increase to greater than 50 degrees along the north-west deposit margins with associated closely space east-west trending faulting. Two east-west trending normal faults of 2 to 3m throws are encountered towards the northem end of the B Pits [Biggs and others, 1989 (1)]. Precursory Drilling at Trap Gully Drilling in advance of the mining areas prior to 1995 had located an area with anomalous dip. It was suspected that this area contained a fault with a throw of around 15m down to the east. Drillhole spacing at that time was wide (approx. 200m in the vicinity of the presumed fault) and no drilling was done specifically to locate the fault. Seismic Trials Conscientious efforts to locate the fault began in early 1990. The first major trial involved mini-sosie surveys within the target area and focussed on Line 5 (Figure 2). This was followed by shallow refraction and downhole seismic methods. The original mini-sosie data was reprocessed with more sophisticated software in 1991. However, the ground in the fault area was not generally conducive to seimic methods and these methods proved to be too inconclusive for defining the fault. A fiirther VSP survey was conducted at Trap Gully in 1992 but, again, the results did not clearly define the fault structures. However, this survey did define two low-velocity zones which appeared to coincide with the presumed faulting. Seismic Reflection Surveys a) Mini-Sosie In April 1990, mini-sosie surveys were performed [Velseis, 1990 (3)] along 7 lines at Trap Gully. The data that were collected along these Hnes remained inconclusive for seam and fault location (see Une 5 data - Figure 3). Line 5 proved to have easiest access and intersected the presumed fault location at 90°. Subsequently, line 5 was selected for further trials. The original data from line 5 were later reprocessed by ACIRL at Curtin University of Technology [Lamboume, 1991 (4)]. Even though the result showed more continuity than previous processing (mainly due to inclusion of the variable refraction static data collected by Bob Whiteley in October 1990), interpretation remained inconclusive (Figure 3).

416

MiniSosie Line 5 Showing Location of Shot Holes S0007 & S0020

TSM Line 2 Showing Location of Boreholes R1833 & R1834

= Boreholes Drilled Prior to MiniSosie Surveys

Location of Anomaly Along TSIM Line 2

Distance (metres)

Figure 2. Trap Gully Mine Location Plan MiniSosie Line 5 & TSIM Line 2

417

Figure 3; Seismic Sections Original MiniSosie Record (Velseis, 1990)

Reprocessed MiniSosie Record (Lamboume, 1991)

Optimum Offset Record (Whitely, 1991) Shear Zone? I

Downhole VSP Record (Hatherly, 1993) Shear Zones?]

418

b) Optimum Offset (High Resolution) In October, 1990, surveys were conducted [RJ. Whiteley, 1991 (5)] along line 5 using the Optimum Offset technique [Gagne et al., 1985 (6)]. Shot holes (10m spacing) were drilled to the base of weathering (25m-40m depth) and "D" size detonation boosters were used as the sound source. Only one dipping reflecting surface (presumably the upper surface of the coal seam) was identified from the results of these surveys (Figure 3). There were no obvious reflections associated with the base of the coal seam or structures within the seam. Due to the lack of clarity from the mini-sosie section, it was difficult to relate the data from the two techniques. Seismic Refraction Surveys & Uphole Seismic R.J. Whiteley (5) also conducted shallow layer refraction and uphole surveys along line 5 in October, 1990. The results from these surveys outlined the main reason for previous difficulties with collecting seismic data in the area. Several low velocity anomalies existed in the near-surface layers (Figure 4). The existence of these rapid lateral velocity variations posed problems for both data collection and processing. Downhole VSP Seismic Survey In August, 1992, a reversed VSP survey [Hatherly et al, 1993 (7)] was conducted along mini-sosie line 5 between two shot holes (S0007 & S0020) 130m apart and 80m deep. These holes were selected because of their close proximity to the expected fault and the intention was to determine whether the fault could be mapped by the VSP method. Initial results showed strong first breaks but these did not display the usual hyperbolic moveout expected under normal lithological conditions. Reflections were not immediately evident, but postprocessing yielded the possibility of two locations for faults (Figure 3). An isotropic tomographic image (Figure 4) was produced which indicated (concordant with Whiteley's refraction and uphole surveys) that a complex velocity structure existed within the near-surface along line 5. Two distinct low velocity zones were present and these were inferred to be indicative of faulted and sheared rock and, thus, were interpreted as expressions of faults. TSIMVLF-EM Surveys The Thiel Surface Impedance Method (TSIM), a VLF-EM technique, has been trialed at Callide Coalfields for fault location, blasted overburden characterisation and subcrop location. Accurate location of coal subcrop and structural features using the TSIM machine has provided significant savings in drilling costs [Nichols, 1995(8); Biggs, 1990 (9)]. Relatively large resistivity contrasts exist at structural boundaries in relatively homogenous, stratified sedimentary media . In this case, the interpreted resistivity contrast could be formed by the dry, air-filled and rubbled crush zone along the fault plane. The TSIM method can easily detect this contrast and locate structural anomalies up to 50m down from the surface. Hence, TSIM serves as a simple means of accurately targeting boreholes. TSIM surveying is a one-person operation and, consequently, is a cheap geophysical method for anomaly location. Eleven lines were surveyed in the target area using the TSIM method in July 1995 (Figure 4). The lateral location of a near-surface anaomaly (interpreted as the manifestation of a fault) was easily detected by the TSIM instrument (Figure 4). Follow-up Exploration Drilling Following the seismic surveys at Trap Gully, a cored hole was drilled adjacent to shot hole S0020 on mini-sosie line 5. Good evidence for, at least, minor faulting was logged in the core. Several sutured microfaults, both normal and reverse, with up to 5cm of movement were recorded [Jorgensen, 1993 (10)].

419

INE S ]?EFRACTaiN

LINE S REFSfC i TIOM

LINE 5 HEFRACTO IN

LINE 5 REFRACTO IW

i : r 1 ; : •: r"" •

' : -i ! t^l

LINE 5 KER

^^^ •••:• 5-.

-T 1 1 2S5

385

405

Seismic Refraction Interpretation - Lateral Velocity Variations (Whitely, 1991)

SOO'20

1.50

seismic

image

2.50

230m

3.50

Isotropic Tomographic Image of Seismic Velocities (Hatherly, 1993) Figure 4. Low-Velocity Zone Sections Following the apparent lateral location of the fault by the TSIM method in 1995, two boreholes were drilled (50m apart, one on each side of the anomaly) along several of the TSM lines. The set of holes on TSIM line 2 was selected for investigation by the radio imaging (RIM) technique to verify the TSIM method and further investigate the structural anomaly. Also, more detailed analysis of the TSIM data is currently being performed to attempt to gain the orientation (dip and throw) of the fault by the TSIM method alone. RIM Radio Imaging METS Pty. Ltd. performed the RIM surveys on TSIM line 2 in June, 1996. As expected, the tomographic results show that a good conductivity contrast exists between the overburden and coal seam (Figure 5). Further, the interpretation of the results from this tomography survey [Neil, 1996 (11)] indicate the anomalous structure to be a set of two en-echelon reverse faults which constitute a total throw of approximately 21m-23m down to the east. This correlates with the interpreted throw (15m-20m) from borehole results. Also, the lateral location of the two-fault anomaly (between 25m45m) along the survey hne from borehole R1834 correlates with the TSIM results. 420

Conclusion The seismic methods employed were highly variable in their results and lacked success in locating the target anomaly. This was largely due to the severe lateral velocity variations in the near-surface layers. In hindsi^t, it would have been more prudent to perform a seismic refraction survey, prior to the much more costly reflection surveys, to detect these velocity variations. The TSIM method was very successful in locating the lateral position of the apparent near-surface manifestation (expected to be a dry, rubbled and air-filled crush zone along the fault plane) of the anomaly and allowed drillhole targets to be located effectively. This method is a very cheap and easy means of geophysical surveying. The information from the holes drilled (eg. R1833 & R1834) following TSIM surveying indicated that a fault with 15m20m throw (down to the east) existed between them. RIM surveys, subsequently, revealed that the lateral location of the fauh (as detected by the TSIM) and vertical displacement (as shown from drilUng data) correlated well. Further, RIM was able to indicate the anomaly to be a two-step, en-echelon reverse fault structure. Overall, the fauh was far more easily, cheaply and successfully located by TSIM & RIM than by Seismic methods. The site will continue to be a test-bed for geophysical techniques and it is intended to conduct TEM surveys in the area later this year. Also, downhole dipmeter, teleview and sonic logs are being performed in the boreholes along the TSIM lines in an attempt to detect manifestations of the fault shear zone.

421

4 Source 0 Receiver

1:500

^a. o^

jCO

o JZ

d)o

10Q

Figure 5. RIM Tomographic Interpretation References (in order of appearance in the text) 1. Biggs, M.S., Broadley, R. Crawford, E. and Carr, G. 1989 "Summary of the Geology and Mining Operations of the Dunn Creek, Trap Gully and Hut Deposits, Callide Coal Measures Basin, Central Queensland" Unpublished Callide Coalfields Report. 2. Nichols, WJ.E 1996 "EPC188, Callide Coalfields, Report on Sub-Blocks ReUnquished, July 1995" CalUde Coalfields Report released to Qld. Dept. for Mines and Energy Open File (Report No. 27253) 3. Velseis 1990 "High Resolution Seismic Reflection Surveys for Callide Coalfields, Trap Gully Deposit, Central Queensland" Unpubhshed Velseis Report to Callide Coalfields Pty Ltd 422

4. 5. 6. 7. 8.

10. 11.

Lamboume, A.N. 1991 "Reprocessing of Trap Gully Seismic Line 5" Unpublished ACIRL Report to Callide Coalfields Pty Ltd Whiteley, R.J. 1991 "CalUde Coalfields, High Resolution Seismic Trial Surveys, Trap GuUy Area" UnpubUshed Report to Callide Coalfields Pty Ltd Gagney, R.M., Pullan, S.E. and Hunter, LA. 1985 "A Shallow Seismic Reflection Method for Use in Mapping Overburden Stratigraphy" in: Unknown Publication (pp 90-102) Geological Survey of Canada, Ottawa Hatherly, RJ., Yu, G., Zhao, R McKenzie, K.B. and Wenzel, F. 1993 "The Borehole Vertical Seismic Profiling Method for Detailed Coal Seam Mapping" ACIRL Final Report on NERDDC Project 1604 Nichols, W.LF. 1995 "Location of Coal Subcrop at Callide Coalfields Using the TSIM Geophysical Method" in: Bowen Basin Symposium 1995 Proceedings (pp. 257-263). Geological Soc. Aust., Queensland, ISBN 0 909869 98 7 Biggs, M.S. 1990 "Results firom Trials of the TSIM Surface Impedance Geophysical Technique at CalHde Coalfields" in: Bowen Basin Symposium 1990 Proceedings (pp. 181-184). Geological Soc. Aust., Queensland, ISBN 0-909869-74-X Jorgensen, P.J. 1993 "1993 Exploration Drilling Program - Drill Site Geologist's Report" Unpublished Calhde Coalfields Report. NeH, M. 1996 "Trap Gully Mine, Radio Imaging Method Survey" Unpublished METS Report to Calhde Coalfields Pty Ltd

423

LATE MESOZOIC TO EARLY CENOZOIC THERMOTECTONIC fflSTORY OF THE SYDNEY BASIN AND THE EASTERN LACHLAN FOLD BELT, AUSTRALL^ Paul B. O'SuUivan, Dave A. Coyle, Andrew J.W. Gleadow, Bariy P. Kohn Australian Geodynamics Cooperative Research Centre, School of Earth Sciences, La Trobe University, Bundoora, Victoria 3083, Australia Summary Apatite fission-track results indicate that the Sydney Basin and adjoining eastern Lachlan Fold Belt experienced significant cooling from Late Mesozoic to Early Cenozoic time. We suggest this cooling, which has affected various regions of the basin differently, can be attributed to the effects of rifting in the Tasman Sea. At -95 Ma rapid km-scale denudation occurred within the eastern Lachlan Fold Belt (eastern highlands), possibly in response to underplating during initial extension along the eastern margin. At this time, the fluid flow regime in the Sydney Basin changed to a restricted-flow system which resulted in the formation of authigenic illite, and possibly the precipitation of magnetite responsible for a mid-Cretaceous paleomagnetic overprint. As breakup commenced during the Late Cretaceous and continued into the Early Cenozoic, cooling was recorded throughout much of the Sydney Basin as -1.5-2.0 km of overburden was removed by erosion, possibly in response to a drop in base level when the Tasman Sea opened. The effects of this later cooling were more pronounced on rocks located along the eastem margin of the basin, probably due to both the denudation as well as the relaxation of recorded elevated geothermal gradients associated with rifting. Introduction The literature abounds with information concerning the thermal and tectonic development of both the Sydney Basin (Fig. 1; eg. 1-8), and the adjoining Lachlan Fold Belt (eg. 9,10). Furthermore, several models have also been proposed (eg. 1114) to explain the affects of Late Mesozoic to Early Cenozoic rifting on the thermotectonic history of the eastem margin of Australia. Most studies have suggested that rifting affected the near-coastal regions of the Sydney Basin either by increasing the paleogeothermal gradients or short-term km-scale denudation. However, due to the lack of relevant geologic control and incomplete geochronology, the actual effects of rifting within the Sydney Basin and adjoining eastem Lachlan Fold Belt have been difficult to constrain, except in a relative sense.

King/ Island

Figure 1. Location map showing the major basement terrains and sedimentary basins along the southeastem margin of Australia, as well as the eastem Lachlan Fold Beh and Sydney Basin regions discussed herein. The magnetic stripes (oblique to the present-day coastline) formed during Tasman Sea rifting between -80 and 60 Ma. Other locations include: BB-Bathurst Batholith; CMB-Clarence-Moreton Basin; NEFB-New England Fold Belt. In an effort to constrain the themiotectonic history of the Sydney Basin and eastem Lachlan Fold Belt (Fig. 1), a large (and continually growing) data base of apatitefission-trackanalysesfiromsedimentary and granitic rocks collected firom the region has been generated. Fission-tracks from apatites within rock sequences record the low temperature thermal history of the host rock (15,16). It has been shown that for times on the order of millions of years,fission-trackages and 424

the length of confined fission-tracks are mainly reduced by temperatures between --60 and 120°C. In addition, the distribution of the confined track lengths in apatites directly reflects the thermal history of a sample (15,16). By combining both age (including single-grain age data) and confined-length measurements, apatite fission-track results provide not only estimates of maximum temperatures and the time of coolingfirommaximum temperatures, but also allow determination of the thermal history (time-temperature path) experienced by the host rock (15,16). The aim of this paper is to summarise recent apatite fission-track data from the eastem Lachlan Fold Belt and Sydney Basin which constrain their Late Mesozoic and Early Cenozoic thermotectonic histories, and to assess these in the light of other pertinent information (vitrinite reflectance and paleomagnetic data, analyses of fluid inclusions in diagenetic quartz, and K/Ar dating of authigenic illite). Using all this information, we present an interpretation for the Late Mesozoic to Early Cenozoic thermotectonic evolution of the region. Regional Setting Sydney Basin

The Sydney Basin is a foreland basin which overlies basement rocks, in part consisting of the Lachlan Fold Belt to the west and the New England Fold Belt to the north. Deposition and subsidence in the basin commenced in the Early Permian resulting in a thick sequence of marine and non-marine sediments up until the Jurassic, with a proposed hiatus in the Late Triassic (1). In general, most researchers believe that stable conditions, probably with minor erosion, prevailed fi-om the Early Cretaceous to the mid-Cretaceous (-100 Ma) after which time the basin experienced rapid uplift and erosion in response to the commencement of sea-floor spreading in the Tasman Sea. The subsequent history of the Sydney Basin is largely unconstrained. However, several studies incorporating vitrinite reflectance and paleomagnetic data, analyses of fluid inclusions in diagenetic quartz, and K/Ar dating of authigenic illite, have been undertaken in order to determine the basin's post-Jurassic history. Throughout the basin, regional vitrinite reflectance (R^) values have been used to constrain maximum post-deposition paleotemperatures. In general, Ro values of surface samples increase fi-om <0.6% (Tmax <95°C) in the western and central parts of the basin, to values >0.7-1.0% (Tmax >115-150°C) along the eastem margin (Fig. 2; 3,6). These variable results were originally interpreted to suggest either the existence of a strong localized heat source along the eastem margin of the basin, possibly in the form of a regional deep crustal intrusion (6), or that deeper burial occurred along the eastem margin (3). However a similar trend is seen in wells drilled throughout the basin, with maximum recorded paleogeothermal gradients increasingfirom'-25°-40°C/km (> present-day values of ~25°-35°C/km) in wells in the centre of the basin to ~50°-60°C/km (>present-day values of --25°-35°C/km) in wells along the eastem margin of the basin (6,7). Since deeper burial alone (without a significant change in heat flow) could not produce significant changes in the subsurface geothermal gradient these results suggest that deeper burial was not the dominant factor determining the increasing R^ values/gradients eastward across the basin. Furthermore, recent maturation modelling of wells within the southem half of the Sydney Basin using the RQ data, heatflowand conductivity (of proposed overburden) measurements, and burial geohistory information led Faiz (7; personal communication, Feb. 1996) and Faiz and Hutton (8) to conclude that similar amounts of erosion had occurred throughout the basin (-1.5-2.0 km) and that there was no evidence to support a model attributing the higher maturation values recorded along the eastem edge of the basin to deeper burial. In fact, the amount of total apparent erosion seems to increasefirom-1.5 km near the coast, up to -2.0 km inland (Fig. 2; 7). Paleomagnetic data have also been used to constrain maximum paleotemperatures within the basin (Fig. 2). The magnetisation history of any particular rock unit depends primarily on the thermal conditions to which it has been subjected (eg. 18), however, lithology and diagenisis can also play a significant part in the formation of any recorded magnetisation (17). Schmidt and Embleton (18), Middleton and Schmidt (3) and Schmidt et al. (17) have reported paleomagnetic data indicating that magnetic "overprints" are ubiquitous throughout the Permo-Triassic sediments and Jurassic breccia diatremes of the Sydney Basin, and that the consistent magnetic polarity of the overprints suggests they formed during the mid-Cretaceous at -90 Ma (17). The paleomagnetic data were originally interpreted to indicate that rocks in the basin experienced low-temperature (-250°C) metamorphism during the mid-Cretaceous in response to deep burial, and presumably cooled in the Late Cretaceous in response to removal of km-scale overburden (eg. 3,18). However, Schmidt (personal communication Dec., 1995), suggested that there is now evidence that the supposed magnetic overprints across the basin occurred in response to oxidation of siderite and production of magnetite, rather than the simple heating and cooling proposed earher. Although the temperature window in which magnetite precipitates is not well known, it seems likely that the ubiquitous magnetic overprints within the Sydney Basin may have occurred at paleotemperatures as low as those suggested by the lowest regional vitrinite reflectance values (<95°C in the centre of the basin with a Ro of -0.6%). Invoking this model explains how the paleomagnetic data within the basin could record a mid-Cretaceous "overprint" whereas the other maximum paleotemperature information (ie. RQ and the fission-track discussed below) suggest that since deposition the rocks were never exposed to temperatures necessary to reset the magnetic polarity by purely themial means.

425

Figure 2. Contours of vitrinite reflectance values for the surface of the Sydney Basin (modified after 6), estimates of the removed overburden from four localities in the southem Sydney Basin (results from 7), and locations of rocks for which paleomagnetic data has been generated by Schmidt et al. (17). Analyses of fluid inclusions in Triassic sediments within the Sydney Basin have also been used to constrain their postdepositional thermal history (4,5). Fluid inclusions are primarily pressure-sensitive thermal indicators because of the coii5)ressibihty of fluids, thus limiting the certainty in estimating paleotemperatures. However, when additional constraints are introduced during interpretation, the temperature at the time of inclusion can be estimated (4). The results from fluid inclusions in quartz overgrowths located at a few hundred meters depth in a well drilled near tiie coast south of Sydney, reveal minimum homogenisation temperatures of up to 100°C, indicating -45°-55°C higher temperatures than currently prevail at the sampled depths (5). Modelling of these results as well as incorporating limited R^ data from the well, led Eadington et al. (4) and Bai et al. (5) to propose that the higher temperatures resulted from a combination of a maximum heat flow of 2.1 HFU (--SO^C/km gradient), higher than the present values of 1.7-1.9 HFU, together with removal of --1.8-2.1 km of overburden. In evaluating these results, Faiz (7) suggested that the estimated amount of removed overburden was closer to --1.5 km (Fig. 2) and proposed that the difference in the two estimates was due to: 1) an inadequate quantity of RQ data used; 2) lack of an acceptable correlation with the limited vitrinite data and their calculated vitrinite values, and 3) no assumption being made for the lithology of the missing section. Faiz (7) also pointed out that -2.2 km of removed overburden might be possible only if the removed section was 100% sandstone, which he suggested, based on the regional lithologies preserved throughout the basin, was most unlikely. Finally, K/Ar dates of fine authigenic illite reveal a consistent eastward decrease in age from -146 Ma to -91 Ma towards the eastem margin of the basin (4), closely paralleling the regional RQ contours (5). The classical interpretation of these ages is that areas closer to the coast e>q}erienced higher relative paleotemperatures, implying deeper burial. However, as with the vitrinite reflectance data, given the basin modelling and recognition of substantially higher paleo-heat flow along the eastem margin of the basin (eg. 6,7) a model invoking significantly deeper burial closer to the coast is not required. Bai et al. (5) suggested that the illite in the Sydney Basin formed at temperatures of-110°-135°C as thefluidflowregime changed during the mid-Cretaceous from a dynamic system with significant flow to a closed/restricted system with respect tofluidflow.Though it has been recognized that thefluidflowregime in the basin must have changed at this time (eg. 4,5), a satisfactory explanation for what caused the change has not yet been presented. However, the similarity between the youngest illite ages and the timing of the paleomagnetic overprint discussed above, led Schmidt et al. (17) to propose that the two were related. Furthermore, it is possible that fluids percolating through the rocks were dramatically influenced by regional changes brought about by initial extension and breakup of the eastem margin of the continent. Eastem Lachlan Fold Belt

The Lachlan Fold Belt of southeastem Australia comprises the southem part of the eastem Australian Tasmanides and is bounded to the west by the Mesozoic to Cenozoic Murray-Darling Basin and to the east by the Permian to Jurassic Sydney Basin and the Tasman Sea (Fig. 1). The Lachlan Fold Belt is characterized by early to middle Paleozoic rocks, including metamorphosed Cambrian through Devonian (primarily Ordovician) volcanic and cratonic-derived deepmarine sedimentary rocks, and extensive Early Silurian, Early Devonian, and Late Carboniferous granitic rocks (9,19,20). It has been proposed that deformation responsible for regional shortening (-60%, 10) and formation of structures 426

throughout the fold belt occurred along a convergent margin setting throughout the Early Silurian to middle Carboniferous (9,19,20). By the end of the middle Carboniferous, all regional deformation within the Lachlan Fold Beh had ceased (9), an idea supported by the presence of "seemingly undeformed" Late Carboniferous granites, which suggest that regional deformation associated with Lachlan orogenesis ceased prior to their intrusion (19,20). The subsequent thermotectonic history of the Lachlan Fold Belt is largely unconstrained due to the lack of cross-cutting relationships or low-temperature geochronological data. Apatite fission-track data from the fold belt in southern New South Wales by Moore et al. (11) showed that apparent ages inland from the east coast were scattered between -220 and 360 Ma with no obvious regional change across the sampled area, while within -50 km of the coastline the apparent apatite ages dropped rapidly to the youngest observed values between -80 and 120 Ma. From these results, Dumitru et al. (13) suggested that the Eastem Highlands (within the eastem Lachlan Fold Belt) had not experienced much erosion associated with rifting, while the young ages along the coast occurred in response to a combination of elevated heat flow and km-scale denudation. More recently, interpretation of fission-track data from samples collected from within the eastem Lachlan Fold Beh in the Snowy Mountains (21) and the Bathurst Batholith (22), suggested that onset of significant cooling within the region occurred at -90±10 Ma and 95±10 Ma (±2a) respectively, which was probably related to significant uplift and erosion at that time. Fission-Track Results And Interpretations From The Sydney Basin And Eastem Lachlan Fold Belt In order to constrain the Late Mesozoic to Early Cenozoic thermotectonic histories of the Sydney Basin and the adjoining eastem Lachlan Fold Belt, apatite fission-track data from both regions have now been completed (Fig. 3, Fig. 4; 23,24). These results, which show younger ages present along the coast and older ages inland are similar to those seen in the earlier fission-track studies carried out further to the south (11,13). Sydney Basin Apatite fission-track data have been obtained from more than 40 samples collected from the Sydney Basin (23; O'SuUivan unpubhshed data; Gleadow unpubUshed data). Apparent apatite fission-track ages from the centre of the Sydney Basin range between --74 to 222 Ma, with no obvious trend in relation to elevation or distance from the coast (Fig. 3). In all but one case (sample T in Fig. 3 with an apparent age of 222 Ma from a Triassic sedimentary rock), samples from the centre of the Sydney Basin do not yield primary ages for the rock unit from which they were derived. Therefore the apparent apatite ages have been reduced in response to post-depositional heating and subsequent cooling. Modelling of the fission-track resultsfromthese samples, using the procedure presented by Gallagher (25), suggests that most of the cooling, and hence removal of the estimated 1.5-2.0 km of overburden, likely occurred during the Late Cretaceous between -80 and 60 Ma (Fig. 5). Detailed analysis of the apparent ages shows that many of the samples contain large spreads in single-grain apatite ages, often ranging between -80-300 Ma (23; Fig. 3). The older grain ages in each sample are interpreted as provenance ages while the young grain ages are interpreted to have been reduced primarily by mid-Cretaceous fluid flow (related to initial extension along the margin) as well as heating associated with deeper burial prior to removal of 1.5-2.0 km of overburden.

427

Precision (1/a]

Figure 3, Apparent apatite fission track agesfiromthe Sydney Basin from Coyle (23), 0'Sullivan et al. (24), Gleadow (unpublish^ data), and O'Sullivan (unpublished data). Also shown are representative single grain age results, in the form of radial plots. Radial plots show apatite fission track ages for each grainfiroma sample. A unit standard error is assigned to each grain on the y-axis, its actual precision is indicated on the x-axis, and age is indicated by extrapolating a line from the 0-point through the plotted point. These results indicate that the composite ages for many of the samples shown represent a mixture of old and young grains as suggested by the shaded areas. Stars (*) represent those samples for which data was modelled for Figure 5. The apatite datafiromsamples collected near the eastem edge of the basin suggest a different cooling history than those from the centre. These samples generally give ages between -95-65 Ma, and a variation in apparent ages from coastal samples suggests that there is a regional younging to the northeast of the basin. Interpretation of the data from nearcoastal samples south of Sydney suggest that the rocks were exposed to maximum paleotemperatures >110°C prior to rapid cooling from paleotemperatures >110°C to <60°C between -80 and 90 Ma (Fig. 5). This interpretation is indicated by a suite of samples from a shallow drill hole located in Wollongong, where multiple samples in Triassic sediments give apparent ages of -85 Ma and the apatites contain confined track lengths >14.2 jim with very narrow distributions (Gleadow, unpublished data); all indicative of rapid cooling at the time suggested by the apparent ages. However, interpretation of the data from near-coastal samples north of Sydney suggest that the rocks were exposed to maximum paleotemperatures between -80-100°C prior to cooling by >50°C at some time between -60 and 75 Ma. This 428

interpretation is suggested by two samples in particular with ages of -59 and 109 Ma (Fig 3). In both cases: 1) the samples show a significant spread in single-grain ages from >200 Ma to -55 Ma, and 2) the young grains contain only long confined tracks while the older grain contain all the shortened tracks. These details suggest that the young grains have been reset and have recorded cooling during the Late Mesozoic/Early Cenozoic, while the older grains have only been partially reset prior to cooling. Finally, there are also variations in the data from samples collected north of Sydney perpendicular to the coast. For QxamplQ, data from samples 'c' and'd' collected from the Triassic Narrabeen Group south of Newcastle (Fig. 3), suggest they have been more hi^ly affected by e^osure to higher paleotemperatures towards the coast. Sample'd' collected - 8 km from the coast shows a spread in ages from >250 Ma (depositional ages) as well as a group of reduced ages of-120 Ma. Sample 'c' collected at the coast also shows a spread in single-grain ages with some grains still giving depositional ages, but in this case most grains have been reduced to as low as -60 Ma. The same trend is seen in samples 'a' and 'b' collected north of Newcastle (Fig. 3). Interpretation of these results suggest samples along the coast have been e?qposed to higher paleotemperatures than those located a relatively short distance inland.

5 10 15 20 ,

Length (^im)

. .

Precision (1/a)

Figure 4. Representative data from eastern Lachlan Fold Belt (modified after 24). Trends in results discussed in the text. M.L., mean confined track length (^im); S.D., standard deviation (iim); N, number of tracks measured in sample. Eastern Lachlan Fold Belt

In a study designed to determine the low-temperature tectonic history of the eastem Lachlan Fold Belt, 40 apatite fissiontrack analyses of granitic and sedimentary rocks were completed from the study area (Fig. 1; 24). The resulting apatite fission-track ages range between 81 ± 4 and 254 ± 8 Ma and the mean confined track lengths for the samples range between 14.2 ± 0.1 |im and 11.5 ± 0.2 |am with standard deviations between 1.1 and 3.4 |im.

429

UJ a: a:

LQ.U sliJ 120 -

350

280

210 175 140 105

a g e (MA) Figure 5. Preferred time/temperature histories derived using the modelling procedure presented by Gallagher (25) for samples from (A) central Sydney Basin, (B) southeastem Sydney Basin, and (C) northeastem Sydney Basin. Samples modelled are shown by stars (*) in Figure 3. Modelling suggests: 1) that sample B was totally reset due to exposure to paleotemperatures >110°C prior to rapid cooling, and 2) that samples A,C were exposed to maximum paleotemperatures <110°C, hence not totally reset, prior to cooling. Dashed lines during the cooling episodes suggest that cooling rates below are poorly constrained and different paths could have been followed. All apparent ages were much younger than the sample's depositional or emplacement ages (Ordovician through Carboniferous), indicating a significant reduction in fission-track ages has occurred in response to thermal anneahng since the rock was crystallised or deposited. Furthermore, the younger samples (apparent ages of -80-100 Ma) were all located along the eastem flank of the fold belt, while all of the oldest samples (>200 Ma) were located west of the crest of the highlands. These results suggest two distinct trends in the regional data (Fig. 4). First, samples with old apparent ages (>220 Ma) typically contain a single population of older grains, while samples with young apparent ages (<120 Ma) contain primarily younger grains. However, sair5)les with intermediate ages commonly contain grains with significant variations in age, with both old and young components. This trend in single-grain ages within granitic terrainsfromsingle components, to mixed components, back into single components with decreasing age is indicative of a trend between a regional partial- to total-resetting of apatite ages (26). Detailed interpretation of the data just from rocks exposed along the eastem flank of the Lachlan Fold Belt, suggests that these rocks were exposed to paleotemperatures between -100° and 110°C in the Early Cretaceous, prior to being rapidly denuded by a minimum of-2 km at -95±10 Ma (±2a; Fig. 5; 22,24). Late Mesozoic to Early Cenozoic Tasman Sea Rifting Prior to the Late Mesozoic, Austraha, Antarctica, and New Zealand were joined together in eastem Gondwana. During this period the eastem margin of Gondwana was dominated by collision with oceanic plates driven from the Pacific region. As a result, the fold belts and basins of eastem Australia (Fig. 1) record a series of subduction-related deformational events, including Early to Middle Paleozoic episodes in the Lachlan Fold Belt, Late Permian to Early Triassic deformation in the New England Fold Belt, Late Permian to Early Triassic compression within the Sydney Basin, and Jurassic to Early Cretaceous compression in the Clarence-Moreton Basin. Starting in the Late Jurassic, extension initiated along the southem margin between Australia and Antarctica (eg. 27). As a result of related subsidence, many kilometers of Lower Cretaceous sediments were deposited into southem-margin basins including the Otway and Gippsland Basins (Fig. 1). Rifting then initiated along the southem margin in the midCretaceous and Australia and Antarctica began to drift apart. Soon after, initial south- to north-directed continental extension started along what is now the southeastem Australian margin, followed by breakup and opening of the Tasman Sea during the Late Cretaceous to Paleocene, and later by the opening of the Coral Sea in the Paleocene to Eocene. It is believed that the structure and topography of the present-day eastem margin are predominantly controlled by this rifting event (11,12,27). Timing of sea-floor spreading along eastem Austraha is well constrained by oceanfloormagnetic anomalies and seismic data (eg. 27,28). The geometry of the magnetic anomaly pattern in the Tasman Sea is unusual compared to that seen in typical "Atlantic style" passive margins, including the southem margin of Australia (28). Lineations along typical passive margins run almost parallel to the margin while those along southeastem Australia are oblique to the margin and appear truncated by it (Fig. 1). Veevers et al. (28), when reviewing the patterns of seafloor spreading around Australia, proposed that the first signs of extension in the Tasman Sea region occurred at -96 Ma, approximately the same time that rifting initiated along the southem margin between Australia and Antarctica (95±5 Ma). Based on the magnetic anomaly data. 430

rifting of the Lord Howe Rise and New Zealand away from eastern Australia commenced at --80-85 Ma and continued into the early Cenozoic at -60 Ma at which time it ceased. Recognising that traditional models for passive margin rifting did not explain the geological relationships seen along the eastem margin of Australia and the Lord Howe Rise, Lister and Etheridge (12) suggested an asymmetric detachment model for the eastem Australian margin and proposed that eastem Australia acted as an upper-plate margin during rifting. Models for continental extension based on the existence of detachment faults and shallow-dipping shear zones predict that extension is asymmetric, resulting in the formation of conjugate passive margins that are quite different in character (12). The upper-plate margin occupies the hanging wall of the master detachments, and the lower-plate margin consists of the detachment footwall, commonly including highly faulted and extended segments of the upper plate (12). Essentially, as the lower plate, including the lower cmst and underlying lithosphere, is pulled out from beneath the margin without significant extension of the upper crust, the upper-plate margin is uplifted vertically in response to igneous underplating as relatively cool, more dense lithosphere is replaced by hotter, less dense lithosphere (eg. 12). Igneous underplating results in the formation of a broad upwarp (passive margin mountain chain) inward of and parallel to the rift. Discussion Based on the close approximation between the timing of the recorded cooling episodes within the Sydney Basin and adjoining eastem Lachlan Fold Belt with events associated with rifting, we suggest that the fission track results record different episodes, or at least stages, associated with the rifting process during opening of the Tasman Sea. We suggest that the initial episode at --95 Ma recorded within the eastem flank of the Lachlan Fold Belt west of Sydney was probably related to denudation of a 'passive margin mountain chain' at the onset of continental extension. If so, then possibly the proposed underplating inward of the rift under the upper plate margin of Australia resulted in localized km-scale denudation. Altematively, Lambeck and Stephenson (29) suggested that the highlands represent deeply eroded remnants of a Paleozoic orogenic mountain chain and that erosion of the old orogenic mountain range has resulted in isostatic rebound since -250 Ma. While the fission-track data from the eastem flank of the fold belt do support the idea that uplift and erosion of the highlands began at -250 Ma, they also indicate that a more recent episode of rapid cooling occurred at -95 Ma (22,24). In the Sydney Basin, cooling on the eastem margin appears to have been time-transgressive south-to-north during the Late Cretaceous to Early Cenozoic, which we believe was related to denudation near the present-day coastal margin, accompanied by locally high heat flow resulting from continental breakup. Since rifting was oblique to the present-day margin and time-transgressive from south to north, this might explain why the apparent apatite ages from rocks along the coast seemingly decrease northward. Furthermore, the effects of this episode were variable in different areas so that the fission track ages were reset to different degrees. Some record the time of cooling, while others have been significantly reduced but not totally overprinted. The thermal effects inland from the coast are also variable along the margin; in general the results suggest that total to near-total overprinting of apatite ages was locally restricted to the coastal areas and that samples more than -40 km away from the coast were not severely affected (Fig. 3). Acknowledgments Earlier components of this work were supported by the Australian Research Council. Continued support has been provided by the Australian Geodynamics Cooperative Research Center (AGCRC) and La Trobe University. Sample irradiations performed at the Australian Atomic Energy Commission HIFAR Reactor were supported by fimding from the Australian Institute of Nuclear Science and Engineering. The interpretations presented here were assisted by discussions with John Webb. This publication is released with the permission of the Director of the Australian Geodynamics Cooperative Research Center. References 1. 2. 3. 4. 5.

Mayne, S.J., Nicholas, E., Bigg-Wither, A.L., Rasidi, J.S., and Raine, M.J., Geology of the Sydney Basin—A review: Bureau of Mineral Resources, Australia, Bulletin 149,1974,229 p. Branagan, D.F., The Sydney Basin and its vanished sequence: Journal of the Geological Society of Australia, v. 30, 1983, p. 75-84. Middleton, M.F., and Schmidt, R, Paleothermometry of the Sydney Basin: Journal of Geophysical Research, v. 87, 1982, p. 5351-5359. Eadington, RJ., Hamilton, RJ., and Bai, G.R, Fluid history analysis-a new concept for prospect evaluation: APEA Joumal,v. 10,1991, p. 282-294 Bai, G., Keene, J., Eadington, P.J., Hamilton, P.J., and MacDougall, I., Fluid flow history of E. Triassic Narrabeen Group sandstones of the southern Sydney Basin: Australian Society of Exploration Geophysics 8th Conference, Feb. 1991, p. 122. Middleton, M.F., Thermo-tectonic influences on the Sydney Basin during the breakup of Gondwana, in Findlay, RH., Unrun, R., Banks, M.R., and Veevers, J.J., eds., Gondwana Eight Assembly, evolution and dispersal, A.A. Balkema, Rotterdam, 1993, p. 613-622. 431

7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Faiz, M.M., Thermal history and geological controls on the distribution of coal seam gases in the southern Sydney Basin, Australia: Ph.D. thesis dissertation, University of Wollongong, Austraha, 1993,215 p. Faiz, M.M., and Hutton, A.C., Two kilometres of post-Permian sediment-did it exist?: 27th Newcastle Symposium - "Advances in the Study of the Sydney Basin", Newcastle, NSW, April, 1993, p. 221-227. Powell, C.McA., Uluru regime, in Veevers, J. J., ed., Phanerozoic Earth History of Austraha: Oxford, United Kingdom, Oxford University Press, 1984, p. 290-340. Fergusson, C.L., and Coney, P.J., Convergence and intraplate deformation in the Lachlan Fold Belt of southeastem Australia: Tectonophysics, v. 214,1992, p. 417-439. Moore, M.E., Gleadow, A.J.W., and Lovering, J.F., Thermal evolution of rifted continental margins: new evidence from fission tracks in basement apatites from southeastem Australia: Earth and Planetary Science Letters, v. 78, 1986, p. 255-270. Lister, G.S., and Etheridge, M.A., Detachment model for the uplift and volcanism of the Eastem Highlands, in Johnson, W., ed., Intraplate volcanism in Eastem Australia and New Zealand: Cambridge University Press, New York, 1989, p. 297-312. Dumitru, TA., Hill, K.C., Coyle, DA., Duddy, LR., Foster, D.A., Gleadow, A.J.W., Green, PA., Kohn, B.P., Laslett, G.M., and O'Sullivan, A.J., Fission track thermochronology: Application to continental rifting of southeastem Australia: APEA Journal, v. 10,1991, p. 131-142. O'Sullivan, P.B., Foster, DA., Kohn, B.P., Gleadow, A.J.W., and Raza, A., Constraints on the dynamics of rifting and denudation on the eastem margin of Austraha: Fission track evidence for two discrete causes of rock cooling, in Mauk, J.L. and St. George, J.D., eds.. Pacific Congress 1995: Proceedings of the 1995 PACRIM Congress, 1995, p. 441-446. Gleadow, A. J. W., Duddy, I. R., Green, P. F., and Lovering, J. F., Confined fission track lengths in apatite—a diagnostic tool for thermal history analysis: Contributions to Mineral Petrology, v. 94,1986, p. 405-415. Green, R F., Duddy, L R, Laslett, G. M., Hegarty, K. A., Gleadow, A. J. W., and Lovering, J. F., Thermal annealing of fission tracks in apatite, 4—Qualitative modelling techniques and extensions to geological timescales: Chemical Geology, v. 79,1989, p. 155-182. Schmidt, P.W., Lackie, MA., and Anderson, J.C., Palaeomagnetic evidence for the age of the Lapstone Monocline, NSW: Austrahan Coal Geology, April, 1995, p. 13-22. Schmidt, P.W., and Embleton, B.J.J., Magnetic overprinting in southeastem Australia and the thermal history of its rifted margin: Journal of Geophysical Research, v. 86,1981, p. 3998-4008. Chappell, B.W., White, A.J.R., and Hine, R., Granite provinces and basement terranes in the Lachlan Fold Belt, southeastem Australia: Australian Joumal of Earth Sciences, v. 35, 1988, p. 505-521. Glen, RA., Thrust, extensional and strike-slip tectonics in an evolving Palaeozoic orogen—a structural synthesis of the Lachlan Orogen of southeastem Australia: Tectonophysics, v. 214,1992, p. 341-380. Kohn, B.P., and Gleadow, A.J.W., The rise of the Snowy Mountains: when and how much: evidencefiromapatite fission track thermochronology (ab.): Geological Society of Austraha Abstracts, v. 37,1994, p. 226. O'Sullivan, RB., Kohn, B.R, Foster, D.A., and Gleadow, A.J.W., Fission track datafiromthe Bathurst Batholith: evidence for rapid middle Cretaceous uplift and erosion within the eastem highlands of Australia: Austrahan Joumal of Earth Sciences, v. 42,1995, p. 597-607. Coyle, D.A., The apphcation of apatite fission track analysis to problems in tectonics: Ph.D. thesis dissertation. La Trobe University, Australia, 1994, 258 p. O'Sullivan, P.B., Foster, D.A., Kohn, B.P., and Gleadow, A.J.W., Tectonic iii5)lications of Early Triassic, and middle Cretaceous denudation in the eastem Lachlan Fold Belt, NSW, Australia: Geology, in press. Gallagher, K,, Evolving temperature historiesfromapatitefission-trackdata: Earth and Planetary Science Letters, V. 136,1995, p. 421-435. O'Sullivan, P.B., and Parrish, R.R., The importance of apatite composition and single-grain ages when interpreting fission track datafiromplutonic rocks: A case studyfiromthe Coast Ranges, British Columbia: Earth and Planetary Science Letters, v. 132,1995, p. 213-224. Johnson, B.D., and Veevers, J.J., Oceanic palaeomagnetism, in Veevers, J.J., ed., Phanerozoic Earth History of Australia, Clarendon Press, Oxford, 1984, p. 17-38. Veevers, J. J., Powell, C.McA, and Roots, S.R., Review of seafioor spreading around Australia. L Synthesis of the pattems of spreading: Australian Joumal of Earth Sciences, v. 38,1991, p. 373-389. Lambeck, K., and Stephenson, R., The post-Palaeozoic uphft history of south-eastern Australia: Australian Joumal of Earth Sciences, v. 33,1986, p. 253-270.

432

AUSTRALIA'S PREMIER ONSHORE OIL PROVINCE - THE EROMANGA BASIN Virginia Passmore and Len Pain, Bureau of Resource Sciences, Canberra Summary The Eromanga Basin contains over 60% of the initial oil reserves in Australia's onshore basins. The basin is the largest in Australia and overlies gas-rich infra basins such as the Cooper Basin which help source the Eromanga hydrocarbons and controlled some of the structures that form Eromanga traps. Perceptions of the hydrocarbon prospectivity of the basin waxed and waned between the first oil recovery in the late 1920s and the first commercial oil discovery in 1978. Extensive exploration, particularly above and near the Cooper Basin, has discovered oil in 11 Eromanga units, the most important of which is the Button Sandstone that holds 66% of the Eromanga region oil. To date 350 accumulations have been found in this region. More that half of the Eromanga accumulation are multi-reservoired. The presence of stacked hydrocarbon-rich basins and the potential for dual basin and multiple reservoired discoveries make the Eromanga a unique and prospective area to e^qplore. Introduction Significant amounts of Austraha's petroleum reserves are contained in both onshore and offshore Mesozoic basins. The most important of the onshore hydrocarbon basins is the Jurassic and Cretaceous Eromanga Basin. Located in the central eastem half of the continent, the Eromanga Basin has an aerial extent of over one million km^ that covers parts of Queensland, South Australia, New South Wales and the Northem Territory (Figure 1). It is the largest basin in Australia. Its exploration area is more than half the size of the North West Shelf. EROMANGA BASIN OUTLINE

•

Uneconomic Oil accumulations

TASMANIA

TASMANIA ^ ^ ^

Figure 1. Location map of the Eromanga Basin and comparison of its size relative to the North West Shelf.

433

Early-Middle Palaeozoic

A NW

Late Palaeozoic-Early

Mesozoic

EROMANGA BASIN PEDIRKA BASIN

SIMPSON BASIN

BIRSVILLE TRACK RIDGE

Mesozoic/Cainozoic

G-M-l M-N TREND TREND COOPER BASIN

S.L-

PEDIRKA BASIN

0 1

20 km I

^=15 H

SIMPSON ^^ BASIN

EROMANGA BASIN COOPER BASIN

GALILEE BASIN

20 km —I K Cretaceous J Jurassic •R Triassic

P Permian CI Late Carboniferous D Devonian

• O

Oil discovery Gas discovery Oil and gas discovery

^=15 H

Figure 2, Location map of the Eromanga Basin and its infra basins. Schematic cross-sections showing the relationship between the Eromanga and underlying basins and location of hydrocarbons accumulations. The Eromanga Basin is a very broad intracratonic downwarp that developed in a relatively quiescent tectonic environment. Up to 3000m of Early Jurassic to Late Cretaceous clastic sediment (sandstone, siltstone, mudstone) and minor coal, sourced largelyfromthe west and south, were deposited in broad depositional systems that spread over large 434

areas of the basin. The thickest sequences are above the Cooper and Simpson Basins. The Eromanga Basin is underlain by a number of Paleozoic and Triassic infra basins (Figure 2), several of which are rich in gas. The most important of these is the Cooper Basin. Deformation in the Eromanga Basin is relatively mild, mainly confined to syndepositional and post-depositional movement along reactivated older faults and to Tertiary compression that produced broad, low amplitude folds and rollover structures. The tectonically more active Early to Middle Paleozoic and Carboniferous to Triassic basins that underlie the Eromanga Basin controlled its sedimentation pattems and structure to varying degrees. Many of the Eromanga traps formed as a result of sediment drape over the stmctural highs of the underlying basins. Exploration History The Eromanga Basin forms part of the Great Artesian Basin hydrological system. It was the oil and gas shows in water wells of this system that encouraged petroleum e}q)loration of the Eromanga Basin sediments in the mid 1920s. The first measurable amount of oil in the basin was recovered at the eastem end when the Longreach No. 1, well drilled in 192831 by Longreach Oil Wells Ltd, (Figure 1), recovered several gallons of oil. Althougji an uneconomic accumulation and insignificant compared to most of the rest of the Eromanga Basin discoveries, the next oil accumulation would not be found until nearly 50 years later. The sediments of the Eromanga Basin declined as exploration targets in that period due to the lack of oil shows in most of the e?q)loration wells drilled and in most of the water wells across the basin. Attention was focused on the infra basins when deep wells drilled in the 1950s and early 1960s encountered the thick Permian reservoir and source rocks of the Cooper Basin and seismic surveys indicated structures below the Eromanga Basin that had potential to trap oil. Commercial gas discoveries in the Cooper and Adavale Basins in 1964 concentrated attention on Paleozoic sediments for the next 12 years. In 1976, economic quantities of gas were found in the Namur Sandstone Member of the Eromanga Basin in Namur No.l, a well targeted in Cooper Basin sediments. The following year, Poolowanna No.l, drilled above the Simpson Basin, flowed oil in the Poolowanna Formation . This second accumulation was uneconomic and at the end of 1977, the Eromanga Basin still lacked commercial quantities of oil. The first commercial Eromanga oil discovery was made in 1978 when the Hutton Sandstone flowed 2400 BOPD in Strzelecki No. 3, a Cooper Basin gas appraisal well. From that period, the Eromanga Basin sediments went from being considered largely overburden above the hydrocarbon targets back to being perceived as potential hydrocarbon reservoirs. Hydrocarbon Potential of the Eromanga Region The Eromanga region is Australia's prime onshore hydrocarbon province. More than 60% of the initial oil reserves in the onshore basins are contained in the Eromanga Basin (Figure 3); the second highest reserves are in the Cooper Basin, althou^ the Cooper Basin contains less than 6% of the total oil of the Eromanga region. The Eromanga Basin is primarily an oil province, unlike the underlying infra basins which are gas prone. Below the Eromanga Basin, economic reserves of gas occur in the Cooper, Adavale and Warburton Basins (Figure 2). The Cooper Basin has the largest onshore gas reserves in Australia. Oil in the Eromanga is sourced by both Cooper and Eromanga Basin source rocks. The depositional history of the Eromanga Basin created a highly favorable habitat for hydrocarbons. Nearly every major formation of Jurassic age and several of Cretaceous age have fair to good reservoir and /or source and seal potential. Oil has been discovered in 11 formations or members of the Eromanga Basin and gas and condensate in seven of them (Figure 4).

435

Australian Onshore Basins - Initial Oil Reserves 70.0 60.0

50.0

c0)

40.0 30.0 20.0

10.0 0.0 Amadeus

Bowen

Canning

Surat

Figure 3. Percentage plot of onshore oil reserves held in each of Australia's onshore basins. The best reservoirs are in the coarse sandstones of thefluvialsystems such as the Hutton Sandstone and Namur Sandstone Member. The Hutton Sandstone has good porosity and permeability, commonly up to 20% and 1 darcy, respectively. It is the most productive of the Eromanga Basin formations, containing 66% of the Eromanga region reserves (Figure 5). Oil occurs in the Hutton Formation in 51 of the Eromanga Basin accumulation. The point bar or lacustrine sandstones, like the Murta Member, are thinner and commonly of poorer quality. Porosity is more variable in them than in the bedload fluvial sandstones of the Hutton and net pay is often thin. Accumulations Since the start of e?q)loration of the Eromanga region, 350 petroleum discoveries have been made. Accumulations extend as far west as the Poolowanna accumulation and as far east as the Longreach accumulation (Figure 1). Most of the accumulations, however, are in the Cooper and Eromanga Basins near the central part of the Eromanga region; clustered near the Queensland/South Australia boundary, above and just beyond the limits of the Cooper Basin. Accumulations in this area are shown on Figure 6. Of the 350 accumulation, 166 are Eromanga Basin discoveries, 117 of which are considered economic. All but eight of the Eromanga Basin accumulations were oil discoveries, and a few, like the Chookoofield,produce both oil and gas.

436

AGE

HYDROCARBON POOLS

EROMANGA BASIN STRATIGRAPHIC NOMENCLATURE D-

Winton Formation

D o DC O

Mackunda Formation Oodnadatta Formation

Coorikiana Fm

lU o< IUJ

CC o

Allaru Mudstone

Toolebuc Formation Wallumbilla Formation

Bulldog Shale (0 UJ

z^ o Q O

Mt Anna Sst Mbr

Wyandra Sst Member C a d n a - o w i e Formation Murta Member McKinlay beds Namur Sandstone Member

Westbourne Formation

Algebuckina Sandstone

Adori Sandstone

0) < oc -3

INJUNE CREEK GROUP

Birkhead Formation

o -a

•D

(0 LU

Hooray Sst

Hutton Sandstone Poolowanna Formation •

Windorah Formation

'Basal Jurassic' Oil

Gas

F i g u r e 4. Eromanga Basin stratigraphy and hydrocarbons reservoirs.

437

Distribution of Oil by Formation

Basal Jurassic/ Poolowanna 5% Triassic 0% Permian 5%

Birkhead 3% Adori 1% Westbourne 6% Namur 7%

Wyandra 1% Murta 6% McKinlay 0%

Figure 5. Distribution of oil in the Eromanga region by formation. Eromanga oil and gas are mainly trapped within anticlines with four-way dip. The traps were formed by the drape of Jurassic and Cretaceous sediments over pre-Jurassic structures that were enhanced by the reactivation of old fault systems and Tertiary compression, or by the creation of new Tertiary reverse-faulted rollover structures. More than 50% of the accumulations are multi-reservoired within the Eromanga Basin such as the Dullingari and Strzelecki fields (Figures 7 and 8) Some 48 of the 166 Eromanga Basin accumulations are dual-basin accumulations which encountered hydrocarbons in both the Eromanga and Cooper Basins such as the Naccowlah, Dullingari and Strzelecki fields illustrated on Figures 7 and 8. Conclusions While the accumulations of the Eromanga Basin and its infira basins are small compared to offshore accxmiulations, the multiple stacked basins and the multi-reservoired petroleum-rich occurrences in many of the accumulations increases the chances of successfiil drilling and makes the Eromanga Basin a hi^ly prospective part of Australia in which to carry out exploration.

438

—

1 142°00'

—

141°00'

I 143^00'

'Inland

Moothandella

"Cuddapan ° Chandos

Horse Creek —

-I

—

26"'00' -

uow

SA Marama

Solitaire

Marengo

Bodalla South. Black Stump

Mt Howitt

Kenmore

Cook Wareena

Barrolka

Tintaburra\

^Talgeberry

iKeleaiyl Daer

.Cuttapirrie

27°00' Keilor

Telopea Pepita

Beanbush

Tartulla ""

Challum

Judga °Naccowlah

Charo^ 1

Nulla

Yanda

^Wackett f V " • Naccowlah West' / iChookoo^ . . . Baryulah

0 d

ISturtI

• V .

28°00' • Watson South

•

Ashby

^ ^

• •

100 km I

Moolion Kudrieke -

Coonatie iBookabourdiel

Innaminka

Brolg^ Fly Lake ^

Maxwell

Thoar 1 I 1

SEE INSET

Jacks( ^

J"irrawarra

Pirraminta

Wantana

Spectre^ Merimelia

EROMANGA BASIN ^

Packsaddle/Pondrinie

Kirby

Burley e>

Oil accumulation Gas accumulation

^ ^

Oil and gas accumulation

Uneconomic accumulationn

|Meranjl| ^ Tindilpie ^cLe. ^ Nappacoongee East _

|Mawson|^c::\

I I I

COOPER BASIN Orientos

JDullingari] ^ Burke

Kirralee Spencer

I Big Lake I Daralingie

Wolgolla

Toolachee

I Stokes Wills j Crumby TIckalara

Munkarie Narcoonowie Limestone Creek^iala j

Wancoocha

Oil and gas accumulation

Uneconomic accumulation

Rheims.

Kerinna^

Yiiki ^ Jena «

Oil accumulation Gas accumulation

EpsilonI

I Lowanna

•

Accumulations named in BOLD are situated in the Eromanga Basin, those not so are in the Cooper Basin. Those with a box around the name are dual basin accumulations.

I INSET

^Kobari

^ Figure 6. Location of Eromanga region accumulations.

439

DULLINGARI FIELD

Figure 7. Examples of stacked multi-reservoired and dual -basin accumulations.

440

NACCOWLAH SOUTH FIELD

CHOOKOO FIELD

Birkhead Formation

Hutton Sandstone

Toolachee Formation

STRZELECKI FIELD

Cadna-owie Formation Murta Member

Namur Sandstone Member

Figure 8. Stacked reservoirs in the Dullingari accumulation.

441

SAPPHIRE-PRODUCING CRETACEOUS/TERTIARY VOLCANOFLUVIAL DEPOSITS IN EASTERN AUSTRALIA Simon R. Pecover, Lecturer in Economic Geology Department ofApplied Geology, University of Technology, Sydney, Australia Summary Sapphires are coimionly found associated with basaltic rocks within numerous Cretaceous/Tertiary volcanic provinces that straddle the Great Dividing Range in Eastern Australia. Sapphire is typically found as xenocrysts in both lava and pyroclastic flow deposits, and as placer accumulations within a variety of fluvial deposits. Concentrations of sapphire in the lavas are typically low. However, concentrations in pyroclastic and related epiclastic sediments, can be particularly higji. This reflects both the initial sampling of sapphire-bearing parental rocks from lower crustal depths during early eruption front development, and crystal/lithic segregation processes during subsequent volcanic flow. As volcanic flows t5T)ically exploit existing drainage, then high concentrations can occur in palaeochannel structures that are in both subvolcanic and intravolcanic locations. Intravolcanic and post-volcanic sapphire-bearing alluvial systems may range from narrow, moderately sinuous channel deposits within broad valleys (Central Volcanic Province, New England Gemfields), to extensive distal alluvial fan and braidplain deposits (Hoy Volcanic Province, Central Queensland Gemfields). Introduction There are approximately 40 separate locations in eastem Australia where sapphires are associated with basaltic volcanic rocks. They include locations as far south as Tasmania, to locations in far north Queensland. In virtually all of these volcanic provinces, sapphire occurs as rare xenocrysts in basaltic lava. However, the bulk of sapphire appears to have been delivered to the Earth's surface in reddish coloured highly ferruginous pyroclastic rocks. The stratigraphic relationships evident in many of these provinces indicates a volcano-fluvial history involving explosive Tertiary basaltic volcanism, contemporaneous with palaeodrainage development. These processes infilled Palaeozoic basement topographic lows with sapphire-bearing volcanogenic deposits and alluvium. Typically, deposition in these systems was terminated by the eruption of basaltic lava which filled palaeochannel structures. In eastem Australia, large scale production of gem sapphire is presently confined to only two volcanic provinces; the Central Volcanic Province in nortiieastem NSW (the New England Gemfields), and the Hoy Volcanic Province in Central Queensland (the Central Queensland Gemfields). This paper will concentrate on the sapphire geology of the better preserved Central Volcanic Province (CVP), with a brief comparison of the more eroded Hoy Volcanic Province (HVP). New England Gemfields - Central Volcanic Province Regional Geological Setting The regional geological setting of the New England Gemfields, comprises Tertiary basaltic volcanics, intrusives, and minor sediments of the Central Province (Johnson 1989) (1) (figure 1). These Tertiary rocks in tum locally overlie and intrude Devonian to Triassic volcanics, metasediments, and plutonics of the Central Block of the New England Fold Belt.

442

VolCMnic Pluffs Btu« Nobby Mtn. Mt. RusmII BaMoors P M k Gragin P««k Th«Nob SwiPsak Wit*rloo SuoarkMf Spring Mtn. 9 Gough Sugarloif 10 Mt. Mitch«ll 11 Balbair Sugadoaf 12 Chandlers P««k

Figure 1. Geology of the New England Gemfields showing the distribution of sapphire-bearing drainage relative to basaltic volcanic rocks. Volcanic Structure of the Central Volcanic Province

The volcanic pile of the CVP lying between Inverell and Glen Innes in northeastern NSW is dominated by a large vent complex (named the Maybole Volcano by Pecover (1987) (2)), located approximately 20 km southeast of Glen Innes, and immediately west of Glencoe. This structure consists of a central dome of basaltic rocks approximately 6 km in diameter. The dome rises to a height of 1,405 m asl and its circumference is defined by a pronounced circular drainage system comprising Maybole Creek and Grahams Valley Creek, which coalesce to form the head-waters of Beardy Waters. The central dome is surrounded by a well developed system of radially draining streams and basalt flow ridges which extends the overall diameter of the structure to in excess of 50 km. Numerous plugs, small domes and several breccia-filled diatremes occur scattered in and around the confines of this volcano; a feature that strongly suggests that the Maybole Volcano is an eroded shield volcano, similar to the Tweed, Focal Peak, Nandewar, and Canobolas volcanic centres. The Maybole Volcano is typical of shield volcanoesfromplains-basalt provinces, where eruptions take place from central vents. The Maybole Volcano also appears to have been a long lived polygenetic volcanic centre, which produced a diversity of products, including low-aspect-ratio, shield-forming lavas, and high-aspect-ratio domes. Lava types associated with this volcano are also quite diverse in both composition, grainsize and texture. The Maybole Volcano lies on the intersection of at least two major, continental-scale lineaments. The first is a NNW-SSE trending Uneament defined by Wellingrove Creek (termed the Wellingrove Creek Lineament by Pecover (1988) (3). The second, is a much longer, NNE - SSW trending lineament, that also passes through the Nandewar Shield Volcano, to the southwest. Several other northeast-southwest and near east-west trending fracture systems also intersect the Maybole Volcano. Virtually all the streams that drain from the Maybole Volcano are sapphire-bearing (figure 1). A second major vent complex is centred on the middle to upper Swan Brook and Kingsland area. This has been termed the Swan Brook - Kings Plains Vent Complex (Pecover 1992) (4), Unlike the Maybole Volcano, this vent complex has a more subdued topography. It has no clearly defined central dome, althou^ a possible in-vent domal structure composed 443

of very coarse-grained teschenitic ankaramite, partly overlain by fine-grained alkali basalts, is centred at the headwaters of the western feeder drainage of Kings Plains Creek (figures 1 and 2). Corundum/sapphire, zircon, ilmenite, pleonaste, and clinopyroxene, are abundant in the small streams draining off these rocks. Two prominent volcanic plugs are evident within the confines of this vent complex. These include Swan Peak (phonolite) and Stony Nob (hawaiite). A third major vent complex lies midway between the Maybole Volcano and the Swan Brook - Kings Plains Vent Complex, and is centred on the volcanic plugs of Spring Mountain (analcimite). White Rock Mountain and Waterloo Sugarloaf. As with the other two vent complexes, sapphire, zircon, ilmenite, and pleonaste occur in streams draining the rocks surrounding these plugs. Basaltic Lava Types The CVP is dominated by mildly to strongly undersaturated volcanic rocks, including alkali basalt and basanite, and minor hawaiite, mugearite, nephelinite, analcimite, and some pyroxene-rich types approaching ankaramite. Associated high-level intrusive rocks are teschenite and alkali dolerite. The greatest development of strongly undersaturated rocks appears to be in the Inverell area, where they form the uppermost lavas in the sequence, and are underlain by alkali basalt. Tholeiitic basalts form a minor but important coii5X)nent of the succession in the Inverell area, where they form the lowermost flows (Duggan 1972) (5). Lava flows range in thicknessfiromless than a metre, to in excess of 10 m. Lava flows throughout the province are flat-lying and form a relatively thin veneer, typically less than 100 m, but reaching a maximum of about 300 m. The predominant alkaline basaltic rocks in the Central Volcanic Province are very similar chemically and texturally to basalts in lava-field provinces elsewhere in eastem Australia. Pyroclastic and Epiclastic Rock Types Sapphire-bearing basaltic pyroclastic/epiclastic rocks occur at many widely spaced locations throughout the CVP, with the highest concentration occurring in a corridor that runs from the southern end of the Swan Brook - Kings Plains Vent Complex to the northem end of Kings Plains Creek, north of Inverell (figure 1) Numerous, fine and coarse-grained varieties exist, with angular fragments rangingfiromPalaeozoic metasedimentary and granitic types to Tertiary basaltic volcanic types. In general, these rocks display a distinctive red to reddish brown colour, with coarse angular fragments supported in a matrix of finer-grained, hi^ly ferruginous silty clay. They can occur as intrusive dykes and stock-works within Palaeozoic basement rocks (typically showing well developed jigsaw-fit textures resultingfiromhydraulic fracturing), or as flows, less than 100 mm to over 10 m thick, draping pre-volcanic topography (Pecover and Coenraads 1989) (6). In palaeochannel settings, they can occur as intra-volcanic deposits interbedded with weathered basaltic lava and alluvial deposits. In semi-dry, weathered exposures, these sapphire-bearing pyroclastic/epiclastic rocks commonly form nodular ferricrete. In locations where these rocks have been subjected to prolonged water saturation, such as in ancient drainage channels, iron oxides are largely removed through leaching, leaving grey or off-white clayey sediments. When these volcanic rocks are deposited into pre-existing drainage channels, separation of crystals and rock fragments from finer-grained material during volcanic flow, can lead to high concentrations of sapphire in the basal parts of these deposits. Deposits of this type typically show a high concentration of volcanic lithic clasts (in contrast to high concentrations of comminuted basement lithologies) that are frequently well rounded. Many of these volcanic clasts are composed of mixtures of smaller clasts forming composite clastic textures. Subsequent reworking of these volcanic sediments by stream action can produce exceptionally rich deposits of sapphire, with grades typically greater than 200 grams per bank cubic metre (some grades can be as high as several kilograms per bank cubic metre). These rocks typically consist of a mixture of clayey volcanogenic sediments, and alluvially processed basaltic ash, breccia and lava, together with various amounts of Palaeozoic basement lithologies. These rocks may rest unconformably on Palaeozoic strata, or basalt, and may be overlain by basalt. Sapphire-bearing clayey sediments of this type, are currently being mined along the valley floor of Kings Plains Creek, at Kingsland (see further discussion of these deposits below). The fragmental, and intrusive/extrusive nature of these rocks, and their concentration near the base of the volcanic pile, suggests that they have primarily been derived from extremely explosive volcanic eruptions that took place early in the eruptive histories of the vents that produced them (Pecover 1993) (7). The high concentration of sapphire in these rocks appears to reflect a process in which sapphire-bearing parental rocks at depth have been sampled by rising columns of basaltic magma as eruption fronts proceeded from mantle depths during periods of extensional tectonics. Landform Evolution and Post Volcanic Alluvial Deposits Regional scale reconstruction of the palaeodrainage over the New England Gemfields has been carried out by Coenraads (1991) (8). This work has shown that the modem drainage pattem is largely coincidental with Tertiary and pre-Tertiary drainage pattems. This work has estabhshed that lavasflowedinto palaeovalleys and progressively filled them. The areas of basement rocks exposed today are generally of equal or higher elevation than the iimiediately adjacent volcanic rocks. These acted as highs around which the lava initially flowed. In some cases, it is clear that the basement was covered and 444

has now been exhumed. Very Uttle basement rock penetrates the thickest and highest part of the volcanic pile in the vicinity of the Maybole Volcano. Radial palaeodrainage around the Maybole Volcano suggests pre-volcanic doming, and is therefore similar in this regard to other shield volcanos in eastern Australia. This doming may have also facilitated the opening of faults and fractures in Palaeozoic basement rocks As the pre-volcanic surface was buried by basalts throughout the New England Gemfields, the drainage appears to have been severely disrupted, with stream thalwegs being reset in different positions, and with different local base levels. The palaeodrainage reconstruction work carried out to date over the New England Gemfields has shown the following distinctive features:(a) Lateral and twin lateral development. (b) Streams following their original pre-basaltic course. (c) Streams that have reversed their flow directions (d) Cross-cutting and stream capture by headward eroding pirate streams. Many of the above features occur together and at varying levels within the volcanic pile, and thus may overprint one another in any given area. While intravolcanic volcanogenic sapphire deposits have produced the richest individual deposits in the CVP, postvolcanic alluvial systems have supported numerous mining operations over the years. Figure 1 ^ows the major sapphirebearing catchments and drainages mined for sapphire in the New England Gemfields. The pattern of erosion and deposition in these valleys has been characterised by lateral and vertical accretion, with the development of sapphirebearing gravel lag deposits. Sapphire crystals in strictly alluvial deposits are typically abraded, with crystal surfaces having a low sheen due to the presence of numerous tiny pits. In contrast to the heavily abraded nature of these grains, the amount of gem sapphire can be relatively high, due to the tendency of intemally flawed, weaker stones, to be destroyed by prolonged reworking. Valley stmcture, and morphology, is perhaps the most important controlling factor in the formation of alluvial placer deposits in the Central Volcanic Province. Features such as rock bars, valley pinch-outs, changes in stream gradient, and small to large scale undulations in the stream bed of these systems, are all important, in localising concentrations of sapphire (Pecover 1992) (4). The Kings land Sapphire Field—Example of a Volcano-Fluvial Depositional System Kingsland is part of the Swan Brook - Kings Plains Vent Complex, and is located within a large upland valley drainage system at the headwaters of Kings Plains Creek, approximately 45 km northeast of Inverell (figures 1 and 2). The valley con5)rises a roughly bowl-shaped, raised plateau-like landform, that is surrounded by more deeply incised sapphirebearing drainage systems, and inverted topography. This upland area hes on a prominent north—south trending lineament basement fracture zone, which is defined by the drainage of Arrawatta Creek, to the north of Kingsland.

445

Figure 2. Geology of the Kingsland Sapphire Field (Swan Brook - Kings Plains vent complex) Central Volcanic Province, Northeastern N.S.W. (modified after Pecover 1992a) Prior to Tertiary basaltic volcanism, the fracture controlled valley drainage system of Arrawatta Creek, appears to have been deeply incised in Permian felsic volcanic rocks. During the Tertiary, basaltic volcanism produced a variety of lava types which filled the original valley of Arrawatta Creek, creating a new, and raised, land surface. Valley formation, coupled with the establishment of a Tertiary drainage system proceeded while volcanism was still active in the area. After volcanism had ceased, a Quatemary drainage system established itself above the earlier Tertiary system. Today, Kingsland is drained by a dendritic pattem of small streams which feed into two larger central channels (the westem and eastem feeders of Kings Plains Creek). These larger channels coalesce farther down the valley to form the main axial channel of Kings Plains Creek. During the past 15 years, sapphire-bearing gravels have been mined from several palaeochannel systems at Kingsland, adjacent to the modem drainages of both the westem and eastem feeders of ICings Plains Creek. These palaeodrainages comprise channels, that are part of straight to slightly sinuous Tertiary drainage regimes. To date, mining has e?q)loited approximately 60% of the known and inferred channelled areas (figure 2). Mining within these channels has revealed a complex, multi-component, volcano-fluvial depositional system comprising three distinctive sapphire-bearing gravel layers (Pecover 1993) (7). The uppermost layer comprises Quatemary brownish coloured clayey alluvial gravels. The pebbles in this layer are predominantly composed of fine-grained alkali basalt, with the fabric of the sediment ranging fi-om clast to matrix supported. The middle layer comprises greyish-coloured clayey gravels of predominantly debri-flow volcanic origin. The pebbles in this layer are composed of coii:5X)site clasts of basaltic volcaniclastics, altered basalt, and minor Palaeozoic sihcic volcanics and metasediments. The clasts in this layer are predominantly matrix supported. In some parts of the deposit, this unit exhibits textures that strongly suggest that crystal/lithic settling processes took place during volcanic flow (including beds of fine-grained grey clay enclosing bands of heavy minerals, containing rounded and polished grains of sapphire, zircon, pleonaste and ilmenite). In other locations, the grey colour of these clastic rocks was observed to grade into reddish-brown breccias, indicating that these sediments are similar to the red pyroclastic/epiclastic rocks occurring elsewhere in the New England Gemfields.

446

The lowermost layer comprises orange-tan coloured clayey gravels of initially volcanic origin (similar to the middle layer), that have undergone prolonged alluvial processing (developed along the westem feeder palaeochannels). The pebbles in this layer are also primarily composed of composite clasts of basaltic volcaniclastics, altered basalt, and minor Palaeozoic silicic volcanics and metasediments, but are predominantly clast supported. Along the lower eastem feeder (figure 2), fluvial processing of similar volcanogenic material does not appear to have taken place. Instead, two sapphire-bearing layers that are very similar to the middle level gravels described from the westem feeder, occur as grey clayey upward-fining units stacked one on top of the other. This suggests that reestablishment offluvialactivity after thefirsteruption only occurred in the more narrow westem feeder palaeodrainage, and not in the eastem feeder palaeodrainage. As most oftiieeastem feeder remains to be ejq)lored, further work in this palaeodrainage system will help to unravel the complex volcano-fluvial history of the Kingsland deposits. In terms of concentration, the highest sapphire grades typically occur in the lowermost gravel layer of the westem feeder, in pockets in the weathered basalt floor of the palaeochannel system. Sapphire concentrations in this layer have been found in recent mining operations to exceed several thousand grams per bank cubic metre in some areas (Great Northern Mining Corporation N.L. 1992) (9). However, economic concentrations of sapphire also occur in the other two gravel units. Basalt lava has been found in several locations along the westem feeder palaeochannel system, cross-cutting the two lowermost gravel layers, with the uppermost gravel layer overlying this basalt. Basalt flows have also been found to overlie palaeochannel deposits in the eastem feeder palaeodrainage system. This stratigraphy indicates that the Kingsland sapphire deposits were formed within an intra-volcanic, volcano-fluvial depositional setting, and show that rich sapphire accumulations can result when sapphire-laden pyroclastic/epiclastic rocks are deposited into active alluvial channel systems. The sapphire deposits at Kingsland are probably the richest yet mined from a basaltic volcanic environment, anywhere in the world. In parts of the channel system currently being mined at Strathdarr (figure 2), some exceptionally h i ^ concentrations of sapphire have been found. Initial deposit evaluations in this area indicated that more than 50% (18,700 kg - worth approximately $30 million) of the total estimated sapphire/corundum resource (25,300 kg - worth approximately $45 million) occurred within an area conprising approximately 4% of the total area delineated. This demonstrates, that in some parts of volcano-fluvial palaeochannel systems, mega concentrations of sapphire can occur. Large areas of Kingsland remain to be adequately explored, with every geological indication that the rich palaeochannels described above, continue farther up the valley on both the westem and eastem feeder palaeodrainages of Kings Plains Creek. While the source vent(s) for these deposits is still uncertain, palaeodrainage reconstmction indicates a source to the south east of Kingsland, with inferred palaeochannel extensions for both the westem and eastem feeder palaeodrainages trending in that direction (figure 2). Comparison with the Central Queensland Gemfields The Central Volcanic Province and the Hoy Volcanic Province represent near opposite ends of a spectrum of volcanofluvial terrain development in eastem Australia. In general the CVP is larger and better preserved that the HVP. As a result, the volcanic piles surrounding vents and vent complexes in the CVP are thicker, more areally extensive and contain a greater diversity of lavas and pyroclastic rock types. Intravolcanic palaeochannels hosting both volcanogenic and alluvial sapphire deposits are better preserved in the CVP, as is the development of inverted topography. Palaeochannels in the CVP are typically long and narrow (0.5 km to >10 Km long x 20m to 100m wide) and are dominated by granular to pebbly gravel bodies that range in thickness from <100 mm to >3m. As discussed, some exceptionally rich sapphire deposits have been formed in the CVP where pyroclastic rocks have been deposited into, and reworked within some palaeochannel systems. In contrast to the CVP, the HVP contains a much smaller preserved volume of basaltic volcanic rocks (figure 3). Extensive erosion has removed most of the lava and pyroclastic rocks from around individual vents. As a result, the modem landscape is dominated by numerous isolated plugs. Intravolcanic palaeochannels or inverted topography (of volcanic origin) are rare, with the area being dominated by thick (0.5m to >5m) and areally extensive (1 km^ to >15 km^) Cainozoic gravel deposits. The Cainozoic gravels comprise a mixture of facies types, including sandy and clayey, pebbly to cobbly, braided fluvial deposits, distributary alluvial fan deposits and debri-flow deposits. These deposits are typically elevated and sub-parallel to the modem Quatemary drainage.

447

Figure 3. Geology of the Central Queensland Gemfields showing the distribution of sapphire-bearing volcanogenic and fluvial lithologies in part of the Hoy Volcanic Province (modified and re-interpreted after the Rubyvale 1:100,000 Geological Sheet) The concentration of sapphire in host gravels is typically much lower in the Central Queensland Gemfields than in the New England Gemfields. However, where gravel deposits in the HVP display a basaltic volcaniclastic component, the sapphire concentration is typically higher than in those gravels that show evidence of prolonged high energy reworking. Where gravel deposits in both provinces show evidence of prolonged, high energy alluvial working, the gem content is typically higher, even though the number of grams per bank cubic metre may be low. Exploration and deposit modelling of economic sapphire deposits is considerably harder to achieve in the HVP than in the CVP. This is because the sapphire contained within the braided alluvial gravels of the Central Queensland Gemfields are commonly confined to braid-bar islands that occur as discreet packets of gravel within more sandy units. Even closespaced drilling (ie <50m) may not be able to provide geostatistically meaningfiil data on sapphire type, distribution or grade in this type of deposit. This is particularly true of gravel deposits occurring to the south east of Rubyvale, where the Scrub Lead forms a widening distributary alluvial fan con:5)lex, east of the hamlet of Sapphire (figure 3). In this part of the gem field, complex braiding of sand and gravel deposits across the face of the fan have resulted in a very complex facies architecture. Recent attempts by an AustraUan public company to commercially exploit these deposits have failed, leading to a significant loss of shareholders funds. This failure is the direct result of a lack of understanding of the complex fluvial nature of the depositional system, and a belief that the deposit could be mined, before adequate geological work was carried out. In the future, any large scale mining of sapphire-bearing gravels in the Central Queensland Gemfields will need to more adequately address the facies architectures of the various deposit types, from both a geological process/product, and grade control perspective. Concluding Remarks Volcano-fluvial depositional systems containing basaltic pyroclastic and epiclastic sediments in Qetaceous/Tertiary volcanic terrains in eastem Australia represent major exploration targets for gem sapphire. Similar potential may also exist in alkali basaltic terrains in parts of Southeast Asia, China and westem Africa. Acknowledgments The research guidance of Dr Eric Middlemost, Department of Geology and Geophysics, Sydney University, is gratefully acknowledged. The assistance of the Department of Applied Geology, University of Technology, Sydney and the Australian Sapphire Corporation Pty Ltd, in the presentation of this paper, is also gratefully acknowledged. References 1.

Johnson, R.W., 1989. INTRAPLATE VOLCANISM IN EASTERN AUSTRALIA AND NEW ZEALAND, 408pp. Cambridge University Press, Cambridge.

448

8. 9.

Pecover, S.R., 1987. Tertiary maar volcanism and the origin of sapphires in northeastern New South Wales, in Extended abstracts from seminar on Tertiary volcanics and sapphires in New England district, 13-21. New South Wales Geological Survey - Report GS1987/058 (unpubL). Pecover, S.R., 1988. Cainozoic raaar volcanism and the origin of sapphire and possibly diamond in eastem Australia - achievements in Australian geoscience. Australian Geological Convention, 9th, Brisbane -Abstracts 21, 314-315. Pecover, S.R., 1992. A guide to the geology of economic sapphire deposits in the Central Volcanic Province, northeastem N.S.W. In Wilhnott W.F. ed. Regional and economic geology of the New England district, 1992, Field Conference, pp. 55-73. Geological Society of Austraha, Queensland Division, Brisbane. Duggan, N.T., 1972. Tertiary volcanics of the Inverell area. A study of the mineralogy, petrology and chemistry of basaltic rocks near Inverell, northem New South Wales. University of New England (Armidale) - BSc Hons Thesis, (unpubl.). Pecover, S.R., and Coenraads, R.R., 1989. Tertiary volcanism, alluvial processes, and the origin of sapphire deposits at "Braemar", near Elsmore, northeastem New South Wales. New South Wales Geological Survey Quarterly Notes 77, 1-23. Pecover, S.R., 1993. The geology and mining of the Strathdarr Sapphire Deposit, New England Gem Fields, northeastem New South Wales. In Flood, P.G., and Aitchison, J.C. eds, New England Orogen, eastem Australia, 493-503. Department of Geology and Geophysics, University of New England, Armidale. Coenraads, R.R., 1991. Alluvial sapphires and diamonds of the New England Gemfields, New South Wales, Austraha. PhD Thesis, Macquarie University, Sydney, Austraha, (unpubl.). Great Northem Mining Corporation N.L., 1992. Prospectus for $6 milhon issue of 30,000,000 fully paid ordinary shares of $1.00 par value at an issue price of $0.20 per share. Underwritten by Resource Finance Corporation Ltd, (unpubl.).

449

A NEW GENETIC MODEL FOR THE ORIGIN OF OPAL IN THE GREAT AUSTRALIAN BASIN Simon R Pecover, Lecturer in Economic Geology Department of Applied Geology, University of Technology, Sydney, Australia Summary Opal deposits occurring in Cretaceous sediments of the Great Australian Basin are considered by most workers to be the result of deep and intense weathering during the Tertiary. The opal is thought to have been deposited by silica-laden ground waters, which percolated to depth through sandstones, aided by enhanced permeability pathways, such as faults and fractures. When these silica-laden ground waters reached permeability barriers such as clay-rich sediments, and were trapped within shrinkage cracks and voids in the upper parts of these facies, opal is thought to have then been precipitated. A new model of opal genesis is presented here. This model advocates a syntectonic depositional process for opal fomiation involving the generation of fault controlled, cyclic, fluid pressurised systems resulting from the antiformal defomiation of interbedded Cretaceous sandstones and claystones. This deformation has produced veins that show textures of incremental opal deposition associated with multiple episodes of cracking and self sealing. Precious opal occurring in these veins shows well developed fibre growth in response to local incremental longitudinal strain rates acting at the time of rupture initiation and fissure dilation. In some veins multiple cracking and fibre growth self-sealing has lead to several layers of precious opal being formed. These layers are referred to by miners as colour bars. Introduction Opal is found in many widely spaced locations within Cretaceous marine to terrestrial claystones and clayey sandstones of the Great Australian Basin (figure 1). The opal typically occurs as common potch and much rarer precious opal in veins, cavity infiUings, and as replacements of ^ell and plant fossil remains in clay-rich facies rocks. Opal-bearing sedimentary rocks are commonly overlain by hard caps of silicified Tertiary sand and gravel. The opalbearing Cretaceous sediments, and their opalised silcrete caps, commonly form low ridges. In eastem Austraha, these ridges exhibit a pronounced NE-SW elongation.

Figure 1. Relationship of lineaments and fracture zones. Figure shows the major lineaments that dominate the structural fabric of opal-bearing Cretaceous sediments in northwestern N.S.W. and southwestem Queensland (modified after Scheibner 1979) The Existing 'Weathering" Model of Opal Genesis Numerous workers have long ascribed to a genetic model for the formation of opal in the Great Australian Basin, involving the deep and intense weathering of Cretaceous sediments during the Tertiary (Darragh et al 1966 (1) and 1976 (2), Senior 1975 (3), MacNevin and Holmes 1980 (4), Carr et al 1979 (5), Bames and Townsend 1982 (6), Watkins 1985 (7), Robertson and Scott 1987 (8), Bames and Townsend 1990 (9)). In this model, large amounts of opaline silica are postulated to have been derived from the breakdown of feldspar to kaolin within permeable sandstones, during periods of localised high ground waterflow.The resultant silica-laden ground 450

waters, are said to have percolated to depth through these sandstones, aided by enhanced permeability pathways, such as faults and fractures. The deposition of opal is then thought to have occurred when these silica-laden ground waters reached permeability barriers such as clay-rich sediments, and were trapped within shrinkage cracks and voids in the upper parts of these facies. Precipitation of the opaline silica from colloidal suspensions in these spaces is thought to have been controlled by the rate at which evaporation of the sihca-laden ground waters took place in a slowly drying climatic environment (up to 5 million years to produce 1 cm of opal at a depth of 40m has been postulated by some workers (Darragh et al 1966) (1). This process is considered to have produced sihca gels, which slowly hardened through the compaction of silica spheres during dehydration. Inherent in the weathering model of opal genesis is the belief that faults and fractures have acted primarily as passive local permeability pathways for the movement of meteoric waters into the upper parts of the Cretaceous sedimentary pile (Watkins 1985) (7). These structures are not considered by the majority of previous workers to have any syntectonic relationship to the opal, and are therefore generally considered to predate the formation of opal (Carr et al 1979) (5), (Bames and Townsend 1982) (6). A New "Syntectonic" Model For The Formation Opal In The GAB Many opal miners believe that there is a more direct relationship between the location of opal and faults. In fact, most miners livelihoods depend on their being able to recognise faults, and in particular, areas of brecciation, as a means of locating and following opal mineralisation. Field studies by the author during the last 6 years in the Lightning Ridge, Coocoran, and Glengarry opal fields (figure 2) has revealed a direct, though complex, relationship between faulting, brecciation and vein opal mineralisation.

i^ f - ' T ^ ^" ^ 0 10 20 30 40 >. J . ^O" •• ! 2 /' •••• / / • 4' 8-El Figure 2. Relationship between the Regional Structural Setting and Outcropping Cretaceous Sediments in Northwestem N.S.W. (modified after Scheibner 1973) Opal Vein Textures and Formation Both potch and precious opal occur as horizontal, sub-horizontal and near vertical discordant vein infillings associated with thrust faults and normal faults. Horizontal veins typically occur as vein arrays which form extensive sheets linked to nearby high and low angle reverse or oblique-reverse faults. Along many of these thrust faults, infillings of opal can occur, particularly at dilational jogs. Vertical veins and isolated blebs appear to be associated with nearby normal faults. Extensional vein arrays in both settings, commonly show composite vein textures comprising both potch and precious opal, that record histories of incremental opal deposition associated with multiple episodes of cracking and self sealing. Potch opal is typically deposited first, with precious opal being deposited later as cracking of the original potch-filled vein, and subsequent dilation, provides space for precious opal to form. However, veins filled entirely with precious opal are also common.

451

The precious opal typically occurs as parallel fibres that have filled the newly created space in the cracked and dilated opal potch vein. In extensional fissures, thesefibresare oriented normal to the walls of the vein, and show extreme length to width ratios. Fibres of precious opal can be straight or curved, and may vary in width along the vein. The width of individual fibres can also vary along their length. Some opal veins examined by the author show characteristics typical of syntaxial fibre veins. In these veins, fibres of precious opal occur in two distinct groups attached to either wall of the vein respectively. The contact between the two groups of veinfibresis a more or less centrally located suture. Some fibres may become wider towards the central suture at the expense of nei^bouring fibres. This change in fibre width can be interpreted in terms of progressive growth sensefi-omthe wall to the centre of the vein, as a resuh of certain fibres being more favourably oriented for growth than their neighbours. Fibre growth implies a crystallisation process, and this is considered to be the case for those minerals that typically form fibre veins in extensional fissure environments (eg quartz, calcite, gypsum and chlorite etc). However, precious opal firom Lightning Ridge is amorphous (Jones et al 1964 (10), and Sanders 1964 (11)). Given that precious opal is composed of an orderly stacking of equidimensional-sized-spheres, then this stacking may be the result of a pseudo-crystallisation process operating in response to the local incremental longitudinal strain acting at the time of rupture initiation. Because the fissure is a site of low pressure, and because the fluid phase infilling the developing fissure contains silica in solution, which can be deposited in a low pressure zone, the fissure is filled with opal at the same time as it is progressively opening, forming the vein. Multiple cracking and fibre growth self-sealing by precious opal can build-up a layering in opal veins. This layering is commonly referred to by opal miners as colour bars. When precious opal is cut, the polished surface is typically at right angles to the long axis of the fibres, with the viewer seeing the polished ends of individual fibres (ie similar to looking at the ends of a bundle of rods of different shapes and sizes). This gives precious opal its classic mosaic pattem of colour. An almost unlimited number of pattems of colour can result firom different combinations of fibre widths and shapes. In some cases, fibre veins of precious opal can form in shear vein systems, with fibre growth being sub-parallel to thefissurewalls (typical of some "boulder" type opal - ie opal occurring with limonitefiromsome Queensland gem fields). The extensional nature and crack-seal textures exhibited many opal veins, shows that they were formed as hydraulic extensionfiractures,when fluid pressures exceeded lithostatic loads. These hydrauhcfiracturesare likely to have formed perpendicular to the least principal compressive stress of the near-stress field, when these stresses were exceeded by the fluid pressure of opaline silica-rich intraformational waters during periods of active faulting. Furthermore, the multiple vein-wall-normal and vein-wall-parallel crack-seal textures exhibited by these veins suggest deposition of opal at different stages in a repeating cycle of pressurisation followed by rupture induced decompression and rapid precipitation of opaline silica. This precipitation would have lead to self-sealing within the rupture zone (ie vein). As compressional stresses reaccumulated, fluid pressures would have again built-up as a prerequisite to the next episode of extensional failure and opal precipitation. Brecdation and SUicification In all the opal producing areas studied, brecciation is pervasive and widespread. Brecciation ranges firom crackle brecciation and net veinfiracturing,to more openfirameworkbreccias. Textural variation rangesfiromclast supported and jigsaw-fit textures to matrix supported breccias showing clast rotation. Breccias showing well developed rotation and rounding of clasts as a result of autogenous milling are also common. Many of these breccias occur as subvertical to vertical pipe-like structures, that cut across several layers of sandstone and claystone, or may be restricted to a single layer of either facies. They range in diameterfi-omseveral centimetres to several metres, and can be several tens of metres in length, commonly breaking through to the surface. These structures are typically heavily silicified by opaline silica, and occur close to areas of well developed extensional and thrust faulting. They commonly contain numerous broken fragments of vein opal, or may be intersected by veins of opal that cut across the clastic fabric of the pipe. Horizontal to near vertical vein systems of potch and precious opal commonly occur adjacent to these structures, thou^ predominantly in surrounding claystones or clay-richfine-grainedsandstones. The existence of extensive zones of brecciation, and in particular breccia pipes that cross-cut stratigraphy, is considered by the author to be further evidence of intense hydraulicfiracturingassociated with overpressurised fluid systems, that developed during periods of episodic faulting. These breccia pipes are commonly associated with normal faults, with some pipes showing a transition from non-brecciated fault plane surfaces at their bases to contorted cylindrical zones of crushed rock. Some brecciation may be developed at sites of dilational fault jogs, where the sudden creation of cavities and intense fluid-pressure differentials would result in the formation of implosion breccias during rupture propagation (Sibson 1987) (12). The close association with juxtaposed and cross-cutting opal vein arrays, and the intense opaline silicification of many of these breccia pipes, strongly suggests a genetic link between the two features. Structural Setting and Sources of Fluids For Opal Formation Earlier researchers have cited the intense weathering of Cretaceous rocks in the opal-producing areas as evidence of significant ground water movement. However, intense firacturing and brecciation of these rocks coupled with hydrothermal alteration and silicification associated with fault controlled fluid pressurised systems, would also produce 452

rocks that appeared to be intensively weathered. Moreover, any subsequent weathering processes are likely to be greatly aided by syntectonically induced faulting,fracturing,brecciation and hydrothermal alteration, which would have openedup the fabric of the rock to later ground water movement. However, any subsequent overprinting effects due to wea&ering are likely to be minor compared to the changes in rock fabric and mineralogy brougtit about by the effects of extensive and pervasive faulting, hydrothermal alteration and silicification. In eastern Australia, many of the opal fields are associated with low north east trending ridges. Landform and structural analysis by the author, utilising Landsat imagery and extensivefield-basedmapping over ridges in the Lightning Ridge, Coocoran, and Glengarry opal fields, shows that they are antifomial structures. Lineament and fault analysis over these ridges is also in accord with them being antiforms, with many opal-producing districts being associated with areas of intense fault clustering within particular parts of these antiforms (figure 2). As antiformal structures are generally formed in compressive stress regimes, then overpressured fluidised systems can easily develop in interbedded sandstones and claystones, that are undergoing compressional dewatering. In this environment, amorphous opaline silica is more likely to be stripped locallyfromclaystones and clayey sandstones as they underwent dewatering. Claystones would also act as barriers tofluidflow,thereby aiding the confinement offluidsunder pressure. Many of the Cretaceous claystones in the opal areas, have an andesitic volcanogenic detrital component, which includes volcanic glass. Furthermore, these rocks contain h i ^ concentrations of alkali mineral salts. Dissolution of an andesitic vitric component under conditions of high alkalinity may provide a viable source of opaline silica for precipitation in vein systems in these rocks. However, for this type of reaction to occur in the presence of buffering phyllosilicate minerals such as clays, hot f200°C- 300°Q, highly alkaline solutions may have been necessary (Foumier 1985) (13). Altematively, if large amounts of acid waters were added to the rock mass from an outside source (ie not from the dewatering of the claystones hosting the opal deposits), then acid attack upon feldspars and other silicates in both claystones and sandstones may release silica to solution, causing supersaturation with respect to amorphous silica. A potential source of non-local silica-richfluidsmay have been palaeoartesian watersfromgeopressured reservoirs at depth within the Great Australian Basin. Thesefluidsmay have travelled towards the surface during periods of postfailure fluid discharge along regional and local scale fault controlled rupture zones. Fluids migrating from locally compacted claystones, orfromgreater depths, along normal and reverse shears, may have entered numerous, near surface, dilational fault structures over large areas, or penetrated the surface, leading to abrupt drops influidpressures towards hydrostatic values. Such drops influidpressures would lead to rapid cooling as a result of decompressional boiling, and the formation of hypersaturated silica solutions. As large degrees of silica supersaturation are required for amorphous silica to precipitate (Foumier 1985) (13), then this mechanism is considered highly likely as a means of precipitating opal in near surface vein arrays, such as those found in most opal fields. Fibrous vein systems are generally most abundant in the more competent layers of a lithologically stratified succession of rocks. In sedimentary assemblages they usually form preferentially in competent sandstone layers (Ramsay and Huber 1983) (14). However, at Lightning Ridge and outlying opal fields, the opal mineralisation tends to be located in less competent claystones, close to the boundaries with overlying and underlying sandstone units. Ramsay and Huber (1983) (14) consider that vein systems are only developed strongly in incompetent materials where these materials have suffered especially high local strains (or strain rates). An example would be differential slip between competent layers (ie the sandstones at Ligjitning Ridge) in flexural shp folds, localised in the incompetent strata (ie the claystones at Lightning Ridge), leading to a higher frequency of veins in the claystones. Therefore, given the competency differences between sandstones and claystones undergoing folding and faulting in broad antifomial structures, the opal vein systems at Lightning Ridge are considered more likely to have developed preferentially in the less competent claystones and clay-rich sedimentary rocks, where local strain rates are high, than in the more competent sandstones. As strain rates are likely to be especially high along zones of preferential slip near the boundary of competent and incompetent layers, then these zones are likely to show a greater development of opal veining. This is supported by the fact that where several layers of claystone exist at different levels in the stratigraphy of an area, opal is commonly found at both the top and bottom of each layer. Thus opal miners are able to exploit the roofs and floors of multiple claystone "levels" in some claims. However, extension associated with nearby faulting is likely to play a more dominant role in the localisation of vein arrays that showfibregrowth of precious opal nonnal to vein walls. Conclusions Recent research by the author involving local and regional scale structural analysis, coupled with studies of opal-bearing vein systems and associated faults and silicified breccia pipes in opal-producing areas, has lead to the formulation of a new genetic model for the fomiation of opal in the Great Australian Basin. This new model advocates a fault controlled, cycUc,fluidpressurised, hydrofracturing origin for the deposition of opal in the near-surface folded Cretaceous sediments of the basm. 453

As such, this model stands in stark contrast to the weathering model put forward by most researchers since the 1960s, and provides the basis for a new approach to opal exploration and mining in the future. Acknowledgments Associate Professor Brian Marshall, Department of Applied Geology, University of Technology, Sydney, provided valuable comment on the content of this extended abstract. The assistance of the Department of Applied Geology, UTS and Opal Ventures ML, in the presentation of this paper is also gratefully acknowledged. References 1 2. 3. 4. 5.

6. 7. 8.

10. 11. 12. 13.

14.

Darragh, P.J., Gaskin, A.J., Terrell, B.C., and Sanders, J.V., 1966. Origin of precious opal. Nature vol 209, no. 5018, London, pp 13-16. Darragh, P.J., Gaskin, A.J., and Sanders, J.V., 1976. Opals. Scientific Amencan, 234(4), pp 84-95. Senior, B.R., 1975. Precious opal in Queensland. Australian Gemmology, 11(2), p 155. MacNevin, A.A. and Hoknes, G.G., 1980. TheMinerallndustry ofNew South Wales, No 18; Gemstones, 2nd Ed. Geological Survey of New South Wales: Sydney. Carr, S.G., OUiver, J.G., Conor, C.H.H., and Scott, D.C., 1979. Andamooka Opal Fields. The geology of the Precious Stones Field and the results of the subsidised mining program. South Australia Geological Survey, Report ofInvestigations no. 51. Bames, L.C. and Townsend, LJ., 1982. Opal: South Australia's gemstone. South Australia Department of Mines and Energy, Handbook 5. Watkins, J.J., 1985. Future prospects for opal mining in the Lightning Ridge Region. New South Wales Geological Survey Report GS1985/119. Robertson, R.S., and Scott, D.C., 1987. Coober Pedy Opal Fields, Geology of the Precious Stones Field and results of the subsidised exploration program 1981. South Australia Geological Survey, Report of Investigations no. 56. Bames, L.C. and Townsend, LJ., 1990 Opal deposits in Australia, in Geology of the Mineral Deposits of AustraUa and Papua New Guinea (Ed. F.E. Hughes), pp. 77-84 (The Australasian Institute of Mining and Metallurgy: Melboume). Jones, LB., Sanders, J.V., and Segnit, E.R., 1964. Structure of opal. Nature vol 204, London, pp 990-991. Sanders, J.V., 1964. Colour of precious opal. Nature vol 204, London, pp 1151-1153. Sibson, R.H., 1987. Earthquake rupturing as a mineralizing agent in hydrothemial systems. Geology, Vol 15, pp 701704. Foumier, R.O., 1985. The behaviour of silica in hydrothermal solutions, in Berger, B.R., and Bethke, P.M., eds, Geology and geochemistry of epithermal systems: Society of Economic Geologists, Reviews in Economic Geology, vol 2, pp 45-72. Ramsay, J.G., and Huber, M.L, 1983. The techniques of modem structural geology. Volume 1: Strain Analysis, Session 13, Measurement ofprogressive deformation, pp 235-263. Academic Press, London.

454

CRETACEOUS AQUIFERS OF THE NORTHERN TERRITORY'S TOP END

G.W. Prowse, RB. Jolly, D.A. Yin Foo and L Matthews Water Resources Division, Power and Water Authority, Northern Territory, Australia Summary Across the Top End of the Northern Territory there are widespread remnants of Cretaceous aged sedimentary platform rocks. The Cretaceous sea transgressed a low relief landscape formed on rocks of Archaean to Cambrian age on the North Australian shield. The region is characterised by tectonic stability since the Cambrian with intense weathering in the Tertiary producing a deep lateritic profile. The Mesozoic sediments cover approximately 50% (?) of the Northern Territory's Top End and consist of claystone, mudstone andfinegrained sandstone with a maximum thickness of 950 metres overlying a basal unit which has a maximum thickness of 200 metres. This basal unit is the regionally significant aquifer in the top end of the Northern Territory. Where it consists primarily of poorly cemented sandstone, sustainable yields in excess of 50 litres per second (L/s) can be obtained from sin^e production bore sites. The thickest areas of sandstone coincide with topographical lows in the pre-Mesozoic surface in close proximity to Proterozoic granite or gneiss outcrops. Water sourced from the Cretaceous aquifers is excellent with low TDS. However, it is potentially extremely corrosive due to low pH caused by high dissolved carbon dioxide levels. Community water supplies currently based on these aquifers are generally considered to be stable and reUable with scope for development. Background Currently all communities in the Top End rely on groundwater as their primary source of supply, with many relying on the groundwater resources in the Cretaceous sediments (refer Figure 1). Originally many communities relied on natural springs and wells until development reached a point where significant problems were being encountered maintaining either the quality or quantity of the water source, usually late in the year. Due to the long dry season and relatively small requirements for most communities, groundwater has been considered to offer the most manageable option in the majority of cases. Cape

yen Diemen

Coburg Peninsula

Gove Peninsula

NHULUNBUY

Yirrkala

umba Iroote Eylandt

Figure LOCATION MAP CUmate The climate is monsoonal, with a short summer "wet" season of 5 months extendingfromDecember to April. The mean annual rainfall ranges from 1500 mm at Nguiu to 1010 mm at Numbulwar. Flooding is common during this period and regional road access is usually limited to the major sealed highways. The mean monthly rainfall templates in Figure 2 represent the regional centres of Darwin and Nhulunbuy. Shallow groundwater temperatures throughout the region are constant at approximately 310°C which is indicative of the very warm average annual air temperatures.

455

400 I 300 1 200

I 100

Darwin

IlL F

400 -r

-4 0

Nhuiunbuy

i^lIlL J

J D

Figure 2. MEAN MONTHLY RAINFALL - DARWIN AND NHULUNBUY Geology and Geomorphology The Top End of the Northern Territory can be divided into three similar geomorphological environments ~ coastal and estuarine plains, undulating sandy plains and tablelands and plateaus. The region primarily consists of basement metamorphics and intrusives of Archaean and Lower Proterozoic age, highly folded Lower Proterozoic geosynclinal rocks and flat lying and gently folded Upper Proterozoic, Cambrian and Mesozoic platform sedimentary rocks. The region has been tectonically stable since probably the Cambrian. Intense subaerial weathering in the early Tertiary resulted in the development of a deep lateritic profile. Since uplift and minor faulting in the mid-Tertiary, an erosional regime has dominated. Soil profiles indicate wide climatic variations in the Pleistocene, associated with fluctuating sea levels. Holocene sediments underlying the coastal and estuarine plains were deposited following the last marine incursion. Noakes (1) described and named the Cretaceous Sediments in the Top End of the NT as the Mullaman Group. Later workers downgraded the Mullaman Group to the Mullaman Beds: a name which was then used to describe all the Mesozoic Sediments of the Top End. In the Darwin Region Hughes and Senior (2) renamed the Mullaman Beds the Bathurst Island Formation during their remapping of the areas covered by Bathurst Island, Melville Island and Coburg Peninsula. The description of the Bathurst Island Formation was enhanced by datafi-omoil wells that had been drilled in that area (Hugjies (3)). Krassay (4) has remapped the Cretaceous geology of East Amhem Land. He proposed two new names, the Walker River Formation and the Yirrkala Formation. Over the remainder of the Top End the Cretaceous Sediments are still known as the Mullaman Beds. The combined areal extent of the Bathurst Island Formation and the Mullaman Beds is shown on Figure 3. Melville Island

TERTIARY

fAr^/A

Sandstone

MESOZOIC f l - = ~ Z - : K CRETACEOUS

PERMIAN

Mullaman Beds and Bathurst Island Formation Sandstone, mudstone Sandstone, shale

CAMBRIAN

PROTEROZOIC

km 100

€

Dolomite, sandstone, shale

Dolomite, sandstone, shale, grey wacke, dolerite. granite 100

200 km

Figure 3. GEOLOGY MAP The Bathurst Island Formation unconformably overiies Proterozoic basement over some of this area. Regional oil exploration work indicates that the Formation dips to the north at an angle of less than 0.5 degrees. It has been subdivided into four units — Marligur Member, Darwin Member, Wangarlu Mudstone and Moonkinu Member — by Hughes and Senior (2). The basal arenaceous unit - the Marligur Member of Aptian age- has a thickness of less than 100 metres in the mapped area. It consists primarily of fine to very coarse sandstone. In localities where the lithology changes from an 456

arenaceous unit to a more argillaceous unit (eg in the vicinity of Darwin) the basal member is known as the Darwin Member. Both members have infilled a low relief pre-Cretaceous topography that gently sloped to the north. They are conformably overlain by a sequence of up to 550 metres of massive dark grey mudstone of Albian to Cenomanian age known as the Wangarlu Mudstone. The Wangarlu Mudstone is conformably overlain by a deltaic sequence of interbedded fine mudstone, siltstone and fine sublabile sandstone - the Moonkinu Member of Cenomanian age. The Moonkinu Member has a maximum thickness of 400 metres in the mapped area. Skwarko (5) defined an inland belt of sediments and a coastal belt of sediments. The inland belt occurs as relatively thin (less than 100 metres) and scattered remnants of what were once more continuous and more extensive sheets. These remnants are generally flat lying and consist of sandstone, siltstone and claystone. The coastal belt appears to be similar in lithology to the Bathurst Island Formation with a basal arenaceous unit being the only remnant in most areas. The remnant coastal belt of sediments are unconformably underlain by Proterozoic basement over most of it's area shown on Figure 3. Regionally these sediments infilled a variable relief pre-Cretaceous topography that gently sloped away firom the continent. Investigation drilling for water supply development in East Amhem Land has identified the existence of a dark grey mudstone (Chin (6)) that conformably overlies the basal arenaceous unit. This may be the equivalent of the Wangarlu Mudstone. The recent work by Krassay (4) has indicated that these sediments are of late Aptian to early Cenomanian age (ie similar to the age range for the Bathurst Island Formation). During the early Tertiary the Van Diemen Sandstone was laid down across the submerging northwest margin. Hydrogeological Characteristics The water supply potential of the groundwater resources contained in the Bathurst Island Formation and the Mullaman Beds have been investigated at many locations. Some of those locations are shown on Figure 1. The following statements regarding the hydrogeological characteristics of the various components of these Cretaceous sediments is based on the data acquiredfiromthese investigations. 1. Bathurst Island Formation As stated in the previous section, the Bathurst Island Formation has been subdivided into four units with differing lithologies. The MarUgur Member, primarily composed of sandstone, is a very high yielding aquifer. The thickest most permeable areas of sandstone appear to coincide with topographical lows in the pre-Mesozoic surface in close proximity to Proterozoic granite or gneiss outcrops. Yields of 90 L/s have been obtainedfi-omthis unit in the Wildman River area where loosely cemented coarse sandstone appears to have infilled river valleys. Transmissivities in excess of 1000 square metres per day have been calculated and specific yields have been estimated to range between 5 and 20% (AGC (7) and Pidsley (8)). The more argillaceous Darwin Member is a low yielding aquifer. Its hydrogeological significance has resulted as a consequence of it overlying a high yielding aquifer developed in the Proterozoic aged Koolpinyah Dolomite in the Darwin Rural Area and the adjacent Gunn Point Peninsula. Yields of 60 L/s have been obtainedfiromthe semi-confined aquifer that has been developed in the dolomite. The clayey sandstones and sandy claystones of the Darwin Member provide the storage that sustains the aquifer developed in the dolomite in the Darwin Rural Area (refer Figure 4) while on the adjacent Gunn Point Peninsxila the aquifer is confined by the Wangarlu Mudstone (refer Figure 5). The Wangarlu Mudstone primarily acts as a confining layer for aquifers developed in either the Marligur or Darwin Members. At Murgenella a potable artesian water supply is exploitedfiromthe MarUgur Member beneath 250 metres of Wangarlu Mudstone. Occasionally a low yielding aquifer has in some places developed in the laterised top of the Mudstone and has been exploited as a water supply source where no other economic water supply options have existed (eg Warruwi community). This type of aquifer usually e}q)eriences difficulties sustaining water supply requirements at the end of the dry season due to the low specific yield of this type of aquifer. Where investigated the Moonkinu Member has indicated that it has the potential to yield significant suppHesfiroma very fine grained sublabile sandstone. However due to the deltaic nature of the sediments, investigation of its potential as a water supply source has proven to be very complicated. Currently ground water is being extractedfiromthe sandstone for water supply purposes at Milikapiti and Wurankuwu (refer Figure 6). The aquifer at both locations has behaved as confmed with a transmissivity of approximately 300 square metres per day and a storage coefficient of 1x10^.

457

WEST

EAST

li N >s,,.

i <

cc UJ > cr o

^ \\\ 1\ ^\

w LU O z

\ \ \

GC S cc LU

i <

5 OQ

Ppk

LEGEND Kid

DARWIN MEMBER BATHURST ISLAND FORMATION

SIMPLIFIED FORMATION & UNIT CONTACTS

Ppw

WHITES FORMATION

TYPICAL END OF DRY SEASON SWL

Ppk

KOOLPINYAH DOLOMITE

SOUTH ALLIGATOR GROUP

Figure 4. DARWIN RURAL AREA CROSS-SECTION SOUTH o

Kla

10L/S

r 15

DISTANCE (kilometres) LEGEND Kuw

WANGARLU MUDSTONE

Kid

DARWIN MEMBER

Kla

MARLIGUR MEMBER

Ppk

KOOLPINYAH DOLOMITE

|MB 22295 TD T

GEOLOGICAL BOUNDARY MARKER BED REGISTERED NUMBER OF BORE TOTAL DEPTH TRANSPOSED

Figure 5. GUNN POINT PENINSULA CROSS-SECTION

458

T 30

80-

60-

Lateritic soil

VAN DEMEN SANDSTONE"

•SANDY'CLAYrfine^very'fTne: ~ yellow-brown with occasional • poorly iron cemenferi hanris

- -

40Clay: white with purple, yellow and brown mottle

20Q

"S -20^ -40MUD: dark grey, with 30-40% clayey sand, fine-very fine, white-yellow

-60-

2000 m

-80-

LEGEND LITHOLOGICAL BOUNDARY

— 0 — - INFERRED LITHOLOGICAL BOUNDARY JSZ_

STANDING WATER LEVEL-SHALLOW

AQUIFER

POTENTIOMETRIC LEVEL-DEEP AQUIFER

Figure 6. WURANKUWU CROSS-SECTION 2. Mullaman Beds The inland belt occurs as relatively thin (less than 100 metres) and scattered remnants of what were once more continuous and more extensive sheets. These remnants are generally flat lying and consist of sandstone, siltstone and claystone. Where these sediments are thickest, they unconformably overly the Cambrian dolomites of the Daly, Wiso and Georgina Basins. In these situations water tables usually occur beneath the unconformity in the dolomite. Where they are underlain by Proterozoic basement rocks they usually occur as a thin (less than 30 metres) veneer of laterised sandstones. These sandstones form an extensive aquifer and are the source of water for many springs in the Katherine region (eg the spring that is the source of Barunga's water supply). This sandstone aquifer also sustains the baseflow for Seventeen Mile Creek which maintains a constant water level during the dry season in the waterhole at the mouth of Katherine Gorge as well as providing a significant proportion of Katherine's town water supply. The sandstone aquifer also provides a large proportion of the recharge to the aquifer in the Tindall Limestone that sustains the baseflow in the Katherine River at a flow rate in excess of one cumec during the dry season. The coastal belt of sediments primarily unconformably overlie Proterozoic basement. Regionally these sediments infilled a low relief pre-Cretaceous topography that gently sloped towards the Gulf of Carpentaria. For the coastal belt, as for the Marligur Member of the Bathurst Island Formation, the thickest most permeable areas of sandstone appear to coincide with topographical lows in the pre-Mesozoic surface in close proximity to Proterozoic granite or gneiss outcrops. This has resulted in the development of extensive groundwater resources in East Amhem Land adjacent to where Lower Proterozoic granite and gneiss outcrop or subcrop over a large area. They have been most extensively investigated on the Gove Peninsula where a water supply has been developed to meet the needs of the Nabalco bauxite mining operation. The borefield is capable of supplying 11 000 megalitres (ML) per year. Investigation work commenced at Gove in 1965 and has delineated an unconfmed aquifer in Cretaceous sandstone and sandy clays overlying Proterozoic granites. The aquifer has been deUneated over an 80 square kilometre area with water supply bores capable of producing 50 L/s. The Cretaceous sediments have a maximum thickness of approximately 200 metres. Poorly cemented sandstones are the main source of groundwater. Water level data acquired for the Gove aquifer is the longest record of water level data that exists for any Cretaceous aquifer in the NT. Water level data has been collected by Nabalco since commencement of operations as a condition of their extraction Ucence. The data gives a good indication of the seasonal and long term variations in water levels for this type of aquifer. Water level data for monitoring bore RN 7980, mean daily streamflow for Yirrkala Creek which is one of the creeks fed by the aquifer and cumulative rainfall is presented in Figure 7. Superimposed on seasonal water level and steam flow variations is a longer retum period event which resulted in minimum annual water levels rising over 7 m between 1973 and 1979, adding an estimated 1 x 10^ ML to storage. The mean annual rainfall for this period was 500 459

mm above the longer temi mean. This increase in storage resulted in the minimum annual baseflow in Yirrkala Creek increasing by a factor of two. Monitoring Bore RN 7980 26

? 2422

1 20 sCO IS 16 14 p£>12 10

/ \r

M ^

f—«

M i

M ^

M i

M k

M i

M M M M

lAean Daily Streamflow - Yirrkala Creek (G8260054)

0.1 71 I 72 I 73 I 74 I 75 I 76 I 77 I 78 I 79 I 80 I 81 I 82 I 83 I 84 I 85 I 86 I 87 I 88 I 89 I 90 I 91 I 92 I 93 I

Cumulative Rainfall 1972 -1993 (Water Year Sep to Aug) 2500

n i

£ s; s s s; s: a gj Figure 7. GOVE PENINSULA AQUIFER HYDROGRAPHS Water supply investigation work at Baniyala (Chin (6)), and in the region to the north of Baniyala, has identified a regionally extensive high yielding sandstone that underlies an area of approximately 3500 square kilometres. This sandstone aquifer is thought to have been developed in the same unit as the aquifer at Gove and have similar hydrogeological characteristics. The only difference is that over much of its regional extent it is overlain by a thick dark grey mudstone aquitard (refer Figure 8). Similar sandstone aquifers exist at other locations (eg Angurugu).

460

SOUTH BLUE MUD rBAY BANIYALA

NORTH MAIN ARNHEM ROAD TO GOVE

LOCAL S C A R P n RETREAT

GURRUMURU

ARNHEM B^

-LATERITE CAPPING

AQUIFER

Figure 8. EAST ARNHEM LAND CROSS-SECTION NORTH SOUTH (Schematic only) Environmental Significance The aquifers developed in the Cretaceous sediments are important sources of water for reticulated water supplies throughout the central and eastern regions of the Top End of the NT. Investigations to date have been focussed primarily on this aspect. These aquifers also play a major role in sustaining both the flora and fauna of these regions. They do this by: 1. Sustaining baseflows in many river and creek systems during the dry season. Several examples have already been mentioned and basic flow data is held for other river and creek systems. 2. Sustaining patches of rain forest that have developed around point spring discharges. More disperse spring discharges also sustain corridors of feed during the dry season, the extent of which determines the abundance of certain marsupials (Braithwaite and GrifiSths (9)). 3. Feeding creeks that drain the uplands and thus sustaining water levels in lagoons on wetlands such as the Mary River floodplains (Chin et al (10)). Much work is still required to develop even a basic understanding of the environmental significance of these aquifers. Water Quality Water obtained from sandstone aquifers within the MuUaman Beds usually have a water quality similar to the sandstone aquifers in the Bathurst Island Formation. The general view of water supplies derived from these sediments is that it is excellent to taste and is of good quality. They are characterised by a low total dissolved solids constituent (TDS) of less than 50 miUigrams per litre The only significant components being Si02, NaCl and dissolved CO2. The high dissolved CO2 results in a pH of approximately 4 at the borehead. With its low buffering capacity due to low alkalinity (usually <10 mg/L as CaC03), the water is potentially very corrosive. This quality is consistent except where affected by recharge of sea water. Marks and Jolly (11) note that such aquifers provide soft acidic groundwaters which are highly corrosive to copper, brass, mild steel, galvanised iron and cement lined pipes and fittings. They explain that "... dissolved CO2 is beheved to originate from the high respiration rate of trees and plant roots in the aquifer recharge areas during the wet season". The corrosive nature of these waters necessitates the use of inert material, such as PVC, ABS, fibreglass and stainless steel for bore construction, bore pumps and fixtures throughout the reticulation system. Corrosion of domestic copper pipes has occurred in the water reticulation systems of most communities causing leakages, although the majority of problems have now been rectified through the replacement of these pipes using polybutadiene pipework. The low pH does not impact on public health issues directly and all waters provided for community water supply meet National Health and Medical Research Council guidelines (12). Treatment processes primarily involve pH correction. Aeration by washboarding or cascading and spraying into storage tanks will generally raise the pH adequately, by release of dissolved CO2 gas. In some cases, the addition of soda ash, or a direct contact with marble chips is used as a secondary treatment. Further treatment may include chlorination where necessary. 461

Water Supply Water supplies developed range from 5 L/s for the Baniyala community (pop. <150) to mine process supply at the Nabalco plant at Nhulunbuy, Gove Peninsula requiring approximately 30 ML per day (350 L/s). The level of service delivered to communities in the Top End is dependent on community population. Major communities with populations in excess of200 target to service average and peak demands of 0.8 and 1.2 kilolitres/capita/day respectively while smaller communities have a water supply service capable of delivering between 0.2 and 0.6 kilolitres/capita/day. The developed water supply capacity for a number of communities in relation to the minimum level of service required is shown on Figure 9. This indicates that all such communities currently have an acceptable water supply status. 20000

3 'o a O^ 0)3 a> 5 o 0 1Q0)

+1 15000

Q. Q.

+ 2

10000

Peak Demand (based on 1.2 kL/c/d^

Q-

5000

2000

4000

6000

Population

1 Katherine 2 Nhulunbuy 3 Alyangula 4 Angurugu 5 Yirrkala 6 Milikapiti 7 Gapuwiyak 8 Barunga 9 Wamiwi 10 Baniyala 11 Wurankuwu

8000

10000

Figure 9. POPULATION VERSUS DEVELOPED WATER SUPPLY CAPACH Y Summary Mesozoic Cretaceous sediments cover approximately 50% (?) of the Northem Territory's Top End and consist of claystone, mudstone and fine grained sandstone with a maximum thickness of 950 metres overlying a basal unit which has a maximum thickness of 200 metres. This basal unit is the regionally significant aquifer. Where it consists primarily of poorly cemented sandstone sustainable yields in excess of 50 L/s can be obtained from single production bore sites. The thickest most permeable areas of sandstone coincide with topographical lows in the pre-Mesozoic surface in close proximity to Proterozoic granite or gneiss outcrops. For these higher yielding poorly cemented sandstone aquifers, transmissivities in excess of 1000 square metres per day have been calculated and specific yields have been estimated to range between 5 and 20%. Water obtained from these sandstone aquifers usually have a TDS less than 50 mg/L with the only significant components being Si02, NaCl and dissolved CO2. Consequently the water is potentially very corrosive. This quality is consistent for all four units of the Bathurst Island Formation and for the Mullaman Beds, except where affected by recharge of sea water. These sandstone aquifers play a major role in sustaining both the flora and fauna of these regions by sustaining baseflows in many river and creek systems during the dry season. Some of these creeks sustain water levels in lagoons on important wetlands such as the Mary River floodplains. They also feed springs that sustain patches of rain forest and that provide a feed source for the fauna of the region during the dry season. However much work is still required to develop even a basic understanding of the environmental significance of these aquifers. References 1. NOAKES, L.C., 1949. A Geological Reconnaissance of the Katherine-Darwin Region, Northern Territory, Bulletin 45, Bureau of Mineral Resources, Canberra. 2. HUGHES, R.J., & SENIOR, B.R., 1974. New Stratigraphic Names for Cretaceous and Cainozoic Units of Bathurst and Melville Islands and Cobourg Peninsula, Northem Territory, AustraHan Oil and Gas Review, 20(2). 3. HUGHES, R.J., 1978. The Geology and Mineral Occurrences of Bathurst Island, Melville Island and Cobourg Peninsula, Northem Territory, Bulletin 177, Bureau of Mineral Resources, Geology and Geophysics, Department of Natural Resources, Canberra,. 4. KRASSAY, A.A., 1994. The Cretaceous Geology of North-eastern Amhem Land, Northem Territory, Record 1994/40, Austrahan Geological Survey Organisation, Canberra. 5. SKWARKO, S.K., 1966. Cretaceous Stratigraphy and Palaeontology of the Northem Territory, Bulletin 73, Bureau of Mineral Resources, Geology and Geophysics, Department of National Development, Canberra. 462

6. 7. 8. 9. 10. 11. 12.

CHIN, D., 1990. Baniyala Groundwater Resource Assessment -1990, Water Resources Division Internal Report 58/1991 Water Resources Division, PAWA, Darwin. AUSTRALIAN GROUNDWATER CONSULTANTS (AGC), 1986. Wlldman River Station Water Source Investigation, Report for Water Resources Division, Department of Mines and Energy, Darwin. PIDSLEY, D., 1990. Jabiru Groundwater Supply Assessment, Water Resources Division Internal Report 32/1990, PAWA, Darwin. BRAITHWAITE, R.W. and GRIFFITHS, A.D., 1996. The Paradox ofRattus Tunneyi: Endangerment of a Native Pest, CSIRO Division of Wildlife and Ecology, Darwin, Wildlife Research, 1996,23, pp 1-21. CHIN, D., et al, 1992. Mary River Coastal Plain Subsurface Hydrology Study, 1992, Water Resources Division Internal Report 20/93, Water Resources Division, PAWA, Darwin. MARKS, A.R., & JOLLY, P.B., 1987. Extreme Corrosivity of Northern Territory Coastal Groundwater Supplies — Origin, Effects and Materials of Construction, Institution of Engineers Australia, Engineering Conference, Darwin, May 11-15,1987, Developing Remote Areas, pp 334-342. NATIONAL HEALTH AND MEDICAL RESEARCH COUNCIL, 1987. Guidelines for Drinking Water Quality in Australia, Australian Government Publishing Service, Canberra, 1987.

463

MESOZOIC DENUDATION AND TECTONICS OF THE CENTRAL EASTERN MARGIN OF AUSTRALIA: FISSION TRACK THERMOCHRONOLOGY Asaf RazaS Kevin C. HiUS Russell J. KorschS Roderick W. Brown^ 1 Australian Geodynamics Cooperative Research Centre, School of Earth Sciences, La Trobe University, Melbourne, Victoria, 3083. 2 Australian Geodynamics Cooperative Research Centre, AGSO, GPO Box 378, Canberra, ACT 2601. 3 School of Earth Sciences, La Trobe University, Melbourne, Victoria, 3083. Summary: New apatite fission track data from the central eastern Australian continental margin provide evidence for three different episodes of cooling during the Mesozoic caused by denudation initiated by tectonic events of regional significance. The first cooling event recorded by the data occurred during the Triassic and was related to contractional defonnation caused by the Hunter-Bowen Orogeny. The subsequent two events were in the Cretaceous: the first, between 100-80 Ma, affected both the margin and the interior basins, while the second, occurred at ca. 60 Ma, and was restricted to a <50 km wide zone along the present coastline. The cooling at 100-80 Ma is regional in nature and was caused by widespread denudation associated with intracontinental tectonism. This, in tum, was driven by major plate motion/geometry changes around the Australian plate at --95 Ma. The younger episode of cooling is roughly coeval with the onset of seafloor spreading in the northem Tasman Sea. The coincidence in timing, the restricted nature of this later episode and the geometry of magnetic lineations can possibly be explained by a northward propagating rift system. Introduction: The southem New England Orogen (SNEO) and the adjoining Gunnedah-Bowen and overlying Surat basins to the west are some of the distinct tectonostratigraphic units that constitute the central eastem margin of Australia (Figure 1). The tectonic evolution of the region firom Palaeozoic to Early Cretaceous is reasonably well known from the preserved geological record and has been documented by various studies [1-4]. However, there is less knowledge of the geological events over the last 100 Ma as ahnost no stratigraphic record is preserved. Yet, the interpretation of identified magnetic anomaly data from the Tasman Sea shows that during this time the Lord Howe Rise-New Zealand separated from Australia [5-7]. Apatite fission track (AFT) analysis has the potential to elucidate the time-temperature history of rocks that have been affected by episodes of regional significance such as rifting. A number of previous AFT studies indicates substantial cooHng and associated denudation along the eastem and southeastem margin of Austraha due to Late Cretaceous Tasman rifting [8-10]. Objectives The purpose of this study is to identify the timing of cooling events along the central eastem margin of Australia to enhance our understanding of the processes of continental rifling along the margin and its distal effects on the interior west of the margin. This study specifically addresses the following questions: 1. 2. 3.

What was the magnitude and extent of thermal effects (both surface and subsurface) of Cretaceous rifling along the central eastem margin of Austraha? Does the chronology of the thermal history parallel to the rifled margin vary with the asynchronous opening of the Tasman Sea from south to north? Do the rocks from the region preserve evidence of thermal events associated with the Hunter-Bowen orogenic deformational episode?

Tectonic Setting Prior to Late Cretaceous rifling. New Zealand, Lord Howe Rise, Austraha and Antarctica together formed eastem Gondwana. Many tectonic models that describe the evolution of the palaeo-Pacific margin of Australia during the Palaeozoic and Early Mesozoic envisage a convergent plate boundary associated with the west dipping subduction system [e.g. 11-12]. Other models infer continuation of this tectonic setting up to the mid-Cretaceous [12-13]. The major deformational episode that affected a large part of eastem Australia including SNEO and the adjoining Sydney-Gunnedah-Bowen basin system occurred during the period 260 -230 Ma and is termed the Hunter Bowen Orogeny (HBO) [11]. It was a phase during which metamorphism, uphfl and erosion occurred along the eastem margin. The HBO was followed by a renewed phase of sedimentation with a series of new basins developing during the Late Triassic-Early Jurassic. The Qarence-Moreton and Maryborough Basins were the first to accumulate sediments followed 464

by the Surat and Eromanga Basins. Subsidence at this time was thermally driven [14]. The youngest preserved sediments in the region are Cenomanian in age, -90 Ma in the Eromanga Basin and --99 Ma in the Surat Basin, and any stratigraphic record of the subsequent geological history is missing. Sedimentation rates prior to the termination of the basin system were as high as 100 m/Ma [15]. Korsch and Totterdell [16] believe that increased subsidence was related to an extensional event that affected the eastem Australia during the Early Cretaceous and led to thefragmentationof the Gondwanaland. The sedimentary sequences in these basins are only mildly deformed, the degree of deformation being greater in basins close to the present-day continental margin than further inland. ElUot, [12] suggested a mid-Cretaceous regional contractional episode responsible for this deformation [see also 16].

New England Orogen Mesozoic Basins

Well Location

Gunnedah Basin Lachlan Fold Belt

Fault

65 ^

Concealed boundary of Permo-Triassic sediments DF = Demon Fault

MF = Mooki Fault

Isobath (m)

Sample location and its AFT age PFS = Peel Fault System

MGF = Moonie-Goondiwindi Fault

•

Town

Figure 1. Location map of four east-west transects (A-D) along which fission track data are presented. The transects run from the coast through the southem New England Orogen to the inland sedimentary basins and are perpendicular to the structural grain of the study area. Interpretation of a convergent plate setting along the eastem Australian margin during the Early Cretaceous is based on the volcanogenic nature of Aptian to Cenomanian sediments of the Surat and Eromanga Basins and the Aptian-Albian sedimentary sequence of the Otway, Gippsland and Bass basin system [17]. In both basin systems, palaeo current directions indicate a source to the east [15,18]. Other models, however, envisage that during the Early Cretaceous eastem AustraUa was experiencing extension, as the nature of magmatism at this time is characteristic of a divergent plate boundary [19-22]. Hamilton [23] points out, however, that even in convergent margin settings the volcanism is associated with extension above the downgoing slab. In fact, both models are not incompatible, with extension in the back-arc region occurring simultaneously with convergence in the fore-arc and subduction of oceanic crust. From Permian to mid-Cretaceous, the New Zealand sector of Gondwana was associated with a convergent plate boundary along its eastem margin. Subduction ceased at --105 Ma when the Pacific-Phoenix spreading ridge colUded with the trench [24] resulting in a major change in the geometry of the tectonic regime. Many lines of evidence from New Zealand, Marie Byrd Land and Australia suggest extension followed soon after. It is possible that the extension propagated northward at that time. Thermochronology Fission track analysis was performed on apatites obtained from 31 outcrop samples from the SNEO and 56 samples from 13 boreholes in the adjacent Gunnedah-Bowen basins. Lithologies preferred during the sample collection were 465

sandstones and granitic rocks as they were both suitable for extraction of apatites. Samplesfromthe SNEO were obtained along four transects (A-D) that run perpendicular to the coastline, whereas boreholes chosen for analysis are aligned in a north-south direction parallel to the Mooki-Moonie-Goondiwindi Fault System (Figure 1). Apatite fission track analysis of samples was carried out according to the procedure described by Naeser [25] and Green [26] using the external detector method [27]. Principles used for interpretation of data have been comprehensively summarised by Naeser [25]; Gleadow et al. [28]; Green et al. [29-30] and Brown [31]. Results discussed in this paper are largely from the SNEO. Some of the results from the Gunnedah and Bowen basins have been previously presented by Raza et al. [32-33]. Results and Interpretation Figure 1 shows the sample locations and AFT ages obtained for each sample along four E-W transects running from the locations of some of the boreholes analysed coast through the SNEO to the inland sedimentary basins. Also shown are by AFT analysis from the Gunnedah and Bowen basins. The AFT datafromthe Clarence-Moreton Basin is from Gleadow and O'Brien [34]. All the AFT ages for samples from the SNEO are younger than the stratigraphic age of the host rocks. This indicates that samples have cooled from elevated palaeotemperatures. Interpretation of AFT parameters measured i.e. age and track l e n ^ data, suggest that samples have cooled from palaeotemperatures between '-60°C to >110°C. AFT ages are generally younger close to the continental margin and tend to get older with distance away from it. The oldest age, 214 Ma, is found along the Transect 'A' near to the westem boundary of the SNEO east of the Mooki Fault (Figure 1). The younger ages and long mean track lengths of samples from coastal regions suggest that these samples cooled from palaeotemperatures >110°C. The regional thermal history can be summarised by displaying the relationship between the fission track age, mean track length and standard distribution for each sample (Figure 2). Figure 2A shows a systematic variation of fission track ages with mean track lengths. Some of the representative confined track length distributions which are characteristic of different parts of the plot are also shown. For a simple case with a provenance age and subsequent heating/cooling event, most of the fission track ages fall between the two end member ages and follow a 'U' shape trend, often called a 'boomerang trend'. This suggests that all the samples have experienced a common style of thermal history. The end member ages, called here the initial age and the total reset age, record the times of cooling from elevated temperatures (>110°C). The interpretation is more complicated for the SNEO because three events have been inferred. The oldest age from the SNEO of 214 Ma, corrects to -248 Ma allowing for track length reduction [35] and corresponds, well with the peak (255 Ma) of the HBO suggesting that the region cooled rapidly during the Triassic. The total reset age on the left hand side of the boomerang plot is clustered around 80-60 Ma (Figure 2A). The expanded version for the interval 40 - 1 1 0 Ma, reveals the possibility of two separate events (Figure 2B). Samples with ages similar to the two inferred coohng ages of 80 Ma and at 60 Ma have long mean track lengths of-14.5 ^un (8322-200, NE-6, NE12), indicative of rapid cooling from temperatures >110°C. The track lengths are unimodal and have 'volcanic type' [28] distributions, suggesting that virtually all the tracks formed and then remained below -'60°C. It is important, however, to note that almost all the samples he within 2a error of a median value around 70 Ma (only l a is shown on Figure 2B). Thus a valid altemative explanation is for continuous slow cooling through the Late Cretaceous on a regional basis. In order to ascertain whether or not the AFT ages between 110 Ma-40 Ma from the SNEO and Clarence Moreton Basin are part of two discrete populations, the mixture modelling

466

INFERRED PALAEOTEMPERATURES

>nox

>ii(rc

^ncrc

-9(rc

^GSX

12.75±a24 Z08 74

250

300

Apatite Fission Track Age (Ma)

110 Fission Track Age (Ma)

Figure 2. Plots showing the relationship between AFT ages and mean track lengths from the SNEO (new data) and Clarence-Moreton Basin [34]. A: Plot displays all the data from the SNEO and Clarence-Moreton Basin (CMB). The AFT ages (±1 a error) with respect to the mean track lengths (±1 a error) show a boomerang trend typical of samples which have cooled from different temperatures in response to the same cooling event. Note that all ages are younger than the Hunter Bowen Orogenic event. See Figure 4 for data description shown with the length histograms. B: Plot shows an expanded view of the AFT data in 'A for those samples which yielded ages younger than 100 Ma. The relationship between the ages and mean track lengths is consistent (within the l a error) with two major episodes of cooling at -60 Ma and --80 Ma. technique [described by Sambridge and Compston, 36] was applied. The results are shown in Figure 3 and demonstrate that the individual grain ages can be separated into statistically discrete populations which have mean ages of ~63±1 and Ma. Although the mean ages of the two populations are similar to the time of proposed cooling events, this is fortuitous. Because the track lengdi data for some of these samples indicate they are 'mixed' ages. Nonetheless, this does support the hypothesis that the rocks of the SNEO and the adjoining Clarence Moreton Basin have been affected by two discrete cooling events during the Cretaceous.

467

10.00

5.00

3.33 45 "NC 0SO

Error in Age

Figure 3: Estimation of maximum likelihood set of age populations and their mean ages using 22 apatite age measurements ftom the SNEO and the Clarence Moreton Basin. These data are the same as shown in Figure 2B. The mean age of each population was determined assuming Gaussian error and by applying the mixture modelling approach described by Sambridge and Con:5)ston, [36]. The radial plot [37] and histogram clearly show two populations, the younger with a mean age of--63 Ma and the older with a mean age of--87 Ma. A similar fission track age-length relationship has been documented from the southeastern margin of Australia [8, 9] but there are also some differences compared to the central eastem margin. The oldest ages from the southeastern margin are -350 Ma and have long mean lengflis, indicating minimal heating since the Palaeozoic. In contrast the oldest AFT age from the central eastem margin indicates Triassic cooling. The effects of mid Cretaceous denudation can be traced as far as 130 km inland along the southeastem margin but -350 km from the coast along the central eastem margin. Opening of the Tasman Sea The youngest apatite ages observed along the coastal region appear to young from -80 to -60 Ma over a distance of 350 km. But at this stage there is insufficient data to confirm the existence of d^ progressive decrease in apatite fission track age towards the north. However, the broad correlation between the time of seafloor spreading adjacent to the margin and the range of observed apatite ages suggests a genetic link between the opening of the Tasman Sea and the thermal history of presently outcropping rocks along the margin. The younger apatite ages are -80 Ma in the south of the study area and are concordant with the apatite ages found further south along the coastal regions [8-10]. North of 31°30'S, however, the youngest AFT ages are -60 Ma (Figure 4). The magnetic anomaly pattem indicates that seafloor spreading in the Tasman Sea was asynchronous along its length and asymmetric across its width [5]. The orientation of magnetic anomaUes obliquely abut the margin, being oldest in the south and progressively younging northward. The most commonly cited plate tectonic model, however, envisages opening of the Tasman Sea along its full length at 96 Ma [e.g. 38, Fig. 6; 17, Fig. 132]. One of the critical assumptions of this model is that the Dampier Ridge is considered to be oceanic. Other models, however, have proposed separate stages of seafloor spreading within the southem and northem Tasman Sea basins. These models have considered the Dampier Ridge to be a rifled continental fragment [e.g. 7], which has been confirmed by a recent study of dredged samples [39]. Discussion Three cooling events have been inferred for the SNEO, all in the Mesozoic: 1) Triassic cooling from the HBO that involving low grade-metamorphism, uplift and erosion; 2) mid-Cretaceous cooling in the SNEO Highlands; and 3) Latest Cretaceous, -60 Ma, cooling along the coast.

468

Figure 4. Sketch shows northward younging offissiontrack ages along the coastal region. Also shown are the extrapolated magnetic anomahes adjacent to the 4000 m isobath assuming uniform rate of spreading. Anomalies are projected from Shaw, [6, Fig. 2.5] and their ages are after Harland et al., 1982. Thefissiontrack ages are rougjily concordant with the onset of seafioor spreading. Deformation, metamorphism, and erosion that accompanied the Hunter-Bowen Orogeny is well documented [e.g. 3,11, 40]. It is considered to be one of the major events in the geological history of the SNEO. EvidencefromAFT data that the SNEO, which forms the central part of the Eastern Highlands,firstcooled during the Triassic has an important bearing on the timing of erosion of this part of the Eastem Highlands. Many studies that deal with the evolution of the Eastem Highlands infer that they were uplifted during the mid-Cretaceous [41^3]. It appears that different parts of the Highlands have different uplift histories, but have experienced a common phase of denudation during the mid-Cretaceous. This conclusion partly supports the Lambeck and Stephenson [44] model that proposes initial uplift of the Highlands between 200-300 Ma and imphes that the Highlands cannot be treated as a single unit [45]. It is important to note, however, that the AFT data do not directly constrain the uplift history, but rather provide indirect evidence through estimates of erosion [31,46]. 469

AFT data, however, provide compelling evidence for two discrete cooling episodes during the Cretaceous that we believe are related to the Tasman Sea rifling. The first episode at -100-80 Ma is identical to the one found from inland sedimentary basins [32, 33] suggesting cooling at this time was regional in nature. Some specific questions that can be asked about this mid-Cretaceous cooling episode are: 1. 2. 3. 4.

How does the timing of cooling corresponds with other studies along the eastem margin? Was this cooling caused by the decay of an elevated palaeogeothemial gradient or related to widespread denudation with constant gradients? Where did the denuded material go, supplying thick deposits of Late Cretaceous sediments? What are the possible tectonic causes of such regional denudation?

Mid-Cretaceous denudation along the eastem margin All of the AFT studies along the eastem and southeastem margin, whether from sedimentary basins or basement rocks, have documented a discrete cooling event during the mid-Cretaceous [8-10,32,43,47-50]. AFT results from Mesozoic sedimentary basins suggest maximum temperatures were achieved due to rapid burial in the Early Cretaceous followed by the onset of mid-Cretaceous denudation [33,49-50]. Cooling due to a drop in palaeogeothermal gradient or denudation The SNEO is bounded on all sides by a series of Late Palaeozoic to Mesozoic sedimentary basins except towards the east where it forms the continental margin. Palaeomagnetic data , fluid inclusion data and K/Ar dates of fine authigenic illite from the Sydney Basin [51-52, 53 in 52] indicates that the eastem part of the Sydney Basin was exposed to high palaeotemperatures during the mid-Cretaceous, ca. 90 Ma. The published surface VR datafromthe Sydney and ClarenceMoreton basins [54-55] indicate that isoreflectance contours are more closely spaced adjacent to the margin than further west, with the hi^est values, >2%, in the Clarence-Moreton Basin. These high VR values with little or no change in stratigraphy, and their tight distribution, suggest that rocks along the margin have been cooled from elevated gradients. Using the VR values from the deep wells, Middleton and Schmidt [54] reported palaeogeothermal gradient of ~4560°C/km from the Sydney Basin. In contrast, further inland, the Gunnedah and southem Bowen Basins appear to have palaeogeothermal gradients calculated from VR and AFT datafromboreholes of between --25-35°C [33]. lliese gradients are normal for sedimentary basins and are similar to the present day geothermal gradient of ~25-30°C [56]. In summary, the surface and borehole coal maturation studies indicate that the high ranks of VR in the vicinity of the coastal areas are due to high heat flow and high palaeogeothermal gradients. However, inland, basins are less affected and maintained 'normal' gradients prior to the onset of denudation at -95 Ma. Similarly, the variation in coal maturity with depth in the Gunnedah-Bowen and overlying Surat basins suggests that the increase in coal rank was due to deep burial under normal palaeogeothennal gradients. Extrapolating from the basins to the SNEO 'basement' implies that the regional mid-Cretaceous cooling was under a normal palaeogeothermal 'gradient' and therefore involved a significant amount of denudation. The maximum palaeotemperature from which each sample has cooled was determined using 'Monte Trax' software [57], which applies the forward modelling approach of Laslett et al. [58]. The depth of denudation has been estimated for each sample along four profiles, using palaeogeothermal gradients of 20°C/km, 30°C/km and 40°C/km and assuming surface temperature of 20°C (Figure 5). In general the increased annealing of fission tracks with depth means that younger fission track ages indicate exhumation from a greater depth. Generally, the youngest ages are obtained from rocks that are near to the margin, and therefore the estimated depth of denudation reaches a maximum close to the margin. The distribution of the -60 Ma cooling ages along a narrow coastal strip (Figure 4) is coincident with the area of high palaeogeothermal gradients indicated by the VR data. This is consistent with the suggestion of a transient thermal anomaly

470

Springfield-1

Tamworth

Walcha

Wauchope

Surat ^ ' Y Mooki Faulty Gunmaati

Basin ^ ^ •TD=858. 1m

Narrabri-2

Macintyre-1

Barraba

Armidale

Dorrigo

Bingara

Goondiwindi

Texas

Tenterfield

Tabulum

Casino

DISTANCE FROM COAST (km)

Figure 5- Estimated amounts of post mid-Cretaceous denudation along the transects A-D. Estimates are based on geothermal gradients of 20°C/km, 30°C/km and 40°C/km shown by solid lines X, Y, Z respectively. The geothermal gradients are consistent with gradients in the adjacent basins. Depth of denudation was determined from the maximum temperatures inferred by the thermal modelling approach described by Laslett, et al. [58] for each sample along the transects. The surface temperature was assumed to be ~20°C. Samples which were totally reset by the inferred transient heat anomaly generated by the seafloor spreading at --60 Ma, were not included. Note that the depth of denudation increases towards the present day margin. generated during initiation of sea-floor spreading. Although it is possible that the elevated gradients recorded by the VR data existed earlier in the basin's history, i.e. Triassic-Jurassic [53 in 52] and that Cretaceous gradients were more like the present day values of 25-30°C/km [56]. Even if the elevated gradients (45-60°C) did exist during the Cretaceous, the peak palaeotemperatures of >110°C inferredfromthe AFT data, would still require at least ~1.5-2 km of denudation at those localities where apatite ages are -60 Ma and have mean track length of 14.5 |im. 471

Possible depocentre of denuded sediments AFT data clearly provide evidence of large scale denudation during the mid-Cretaceous, and therefore there should be a preserved record of Late Cretaceous clastic sediments in those basins which were rapidly subsiding at this time. There are no Late Cretaceous depocentres present in the proximity of the study area, although major depocentres exist to the north in the Papuan Basin, and further south in the Gippsland, Otway and Great Australian B i ^ t basins. Prior to the opening of Tasman Sea, the Lord Howe Rise was attached to Australia and it is possible that denuded sediments were transported to the east. Among all the known depocentres, however, the Ceduna Terrace contains the greatest thickness, km, of Late Cretaceous clastics [17, 59] which implies that most of sediments were transported to the west. This is consistent with asymmetric uplift increasing towards the margin. Tectonic causes for mid-Cretaceous regional cooling

The concept of rigid plates suggests that tectonic processes occurring along one part of a plate boundary could induce considerable disturbance in the stress regime many thousand of kilometres away [60-62]. Many studies have shown that such effects are not often restricted to the plate, and in fact extend across the plate boundaries suggesting that plate motions are interconnected [60-64]. Therefore, changes in plate boundary conditions will result in deformation mainly along plate boundaries but also incite realignment of relative plate motions, including the creation of new ocean floor. The mid-Cretaceous cooling and associated denudation has been recorded along the eastem Gondwana such as AustraUa and New Zealand. A regional event of this kind requires a cause that could affect a large area. The timing of this event corresponds well with a number of changes that occurred along the Australian plate boundaries. These include the initiation of seafloor spreading in the Southern Ocean and relocation of the spreading centre to the south of Tasmania, leaving Bass Strait as a failed rift arm; and the reorganisation of the ridge system within the southeast Indian Ocean when the spreading direction changed from NE to N-NW so that India began to move rapidly northward with respect to AustraUa and Antarctica. In addition, rifting of New Zealand from westem Antarctica took place at 95 Ma [65-66]. Mid-Cretaceous cooling followed incipient Aptian-Albian extension along the Tasman Sea margin of New Zealand as suggested by the development of core complexes, magmatism and the formation of fault-bounded sedimentary basins [67-70]. It is also the time when the nature of magmatism, considered to be related with mantle plume activity changed in Marie Byrd Land [71]. The sudden cessation of Cretaceous magmatic activity along the northeast coast of Queensland, as evidenced from the youngest K-Ar age of 95 Ma [21-22] was accompanied by a major unconformity in Mesozoic basins of eastem Gondwanaland. Taken together, this evidence leads us to conclude that the temporal and spatial changes in the tectonic configuration of Australian plate during the mid-Cretaceous, the synchronous magmatism, the deformation and regional denudation are genetically linked. It is evident that the --95 Ma plate re-organisation initiated true rifting along the entire palaeo Pacific margin, but the specific cause of kilometre scale denudation is so far unknown. Different possibilities can be considered. With the onset of continental extension, a change in base level will occur along the incipient rift margin caused by subsidence [46, 72]. This will induce rapid denudation of the adjacent margins. Many models recognise mantle upwelling as a cause of the crustal thinning associated with the extension and during this process the emplacement of large quantities of mantle derived mafic material takes place within the cmst or uppermost mantle (or both). The mid-Cretaceous denudation of the central eastem margin of Austraha perhaps signifies the geomorphic response to this episode of underplating [41] which domed the entire area during the early stages of continental extension. Acknowledgments: Funding for this research was provided by the AGCRC, AGSO and Australian Institute of Nuclear Science and Engineering. Many thanks to the members of the Fission Track Research Group at La Trobe University for stimulating discussions on the Mesozoic tectonics of eastem Australia. Barry Kohn provided helpful comments on the manuscript. This paper is published with the permission of the Director of the Austrahan Geodynamics Cooperative Research Centre. References 1. 2. 3. 4.

Murray, C.G., Fergusson, C.L., Flood, RG., Whitaker, W.G. & Korsch, R.J., Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Joumal of Earth Sciences, 1987, 34, 213-236. Harrington, H.J., & Korsch, R.J., Tectonic model for the Devonian to middle Permian of the New England Orogen. Austrahan Joumal of Earth Sciences, 1985, 32,163-179. Harrington, H.J., & Korsch, R.J., Late Permian to Cainozoic tectonics of the New England Orogen. Australian Joumal of Earth Sciences, 1985, 32,181-203. Veevers, J.J., Conaghan, P.J. & Shaw, S.E., Permian and Triassic New England Orogen/Bowen- Gunnedah-Sydney Basin in the context of Gondwanaland and Pangea. In: Flood, P.G. & Aitchison, J.C., (eds.). New England Orogen, eastem Australia. University of New England, Armidale, 1993, 31-51. Weissel, J.K. & Hayes, D.E., Evolution of the Tasman Sea reappraised. Earth and Planetary Sciences Letters, 1977, 36, 77-84. 472

6. 7. 8.

10. 11.

12. 13.

14. 15. 16. 17. 18. 19.

20.

21.

22.

23. 24. 25. 26. 27. 28. 29.

30.

31.

Shaw, R.D., On the Evolution of the Tasman Sea and adjacent continental margin. Ph.D. Thesis, University of Sydney (unpublished), 1979. Shaw, R.D., Development of the Tasman Sea and easternmost Australian continental margin-a review. AGSO Bulletin, 1990,232, 53-66. Moore, M.E., Gleadow, A. J.W. & Lovering, J.F., Thermal evolution of rifted continental margins: new evidence from fission tracks in basement apatites from southeastem AustraUa. Earth and Planetary Sciences Letters, 1986,78,255270. Dumitru, T.A., Hill, K.C., Coyle, D.A., Duddy, I.R., Foster, DA., Gleadow, A.J.W., Green, P.F., Kohn, B.P., Laslett, G.M. & 0'Sullivan, A.J., Fission track thermochronology: appUcation to continental rifling of southeastem Australia. APEAJoumal, 1991, 31,131-142. Coyle, DA., The application of apatite fission track analysis to problems in tectonics, Ph.D. Thesis, La Trobe University (unpublished), 1994. Fergusson, C.L. & Leitch, E.G., Late Carboniferous to Early Triassic tectonics of the New England Fold Belt, eastem Australia. In: Flood, P.G. & Aitchison, J.C., (eds.). New England Qrogen, eastem Australia. University of New England, Armidale, 1993, 53-59. EUiott, L.G., Post- Carboniferous tectonic evolution of eastem Austraha. APEA Journal, 1993, 33, 215-236. Veevers, J.J., Mid-Cretaceous tectonic climax. Late Cretaceous recovery, and Cainozoic relaxation in the Australian region. In: M.A.J. Williams, P.De Deckker and A.P. Kershaw, (eds.). The Cainozoic in Australia: A reappraisal of the evidence, 1991,1-14. Gallagher, K., Permian to Cretaceous subsidence history along the Eromanga-Brisbane geoscience transect. AGSO BuUetin, 1990, 232,133-142. Exon, N.F., Senior, B.R., The Cretaceous of the Eromanga and Surat Basins. BMR Journal of Australian Geology and Geophysics, 1976,1, 33-50. Korsch, R.J. & Totterdell, J.M., Mesozoic deformational events in Eastem Australia and their impact on onshore sedimentary Basins, (this volume) Veevers, J.J., Phanerozoic Earth History of Australia (Second Edition). Qarendon Press, Oxford, 1986,1-418. Constantine, A.E. & Holdgate, G.R., Selwyn Symposium, Gippsland Basin Excursion Guide: October 1993, GSV, unpaginated. Ewart, A., Schon, R.W. & Chappell, B.W., The Cretaceous volcanic-plutonic province of the central Queensland (Australia) coast - a rift related 'calc-alkaUne ' province. Transactions of the Royal Society of Edinburgh: Earth Sciences, 1992, 83, 327-345. Allen, C.M. & Chappell, B.W., Contrasting Carboniferous-Permian and Cretaceous plutonism in the Urannah Batholith, northern New England Fold Belt. In: Flood, RG. & Aitchison, J.C., (eds.). New England Orogen, eastem Austraha. University of New England, Armidale, 1993, 573-579. Stephens, C.J., Schon, R.W., & Ewart, A., Mesozoic cmstal extension in the northem New England Orogen: Geochemical and isotopic evidence from large silicic magmatism. In: Flood, P.G. & Aitchison, J.C., (eds.). New England Orogen, eastem Australia. University of New England, Armidale, 1993, 637-642. Parianos, J., Bryan, S., Ewart, A., Schon, R.W., Early Cretaceous rift volcanics of the central Queensland coast. In: Flood, P.G. & Aitchison, J.C., (eds.). New England Orogen, eastem Australia. University of New England, Armidale, 1993, 655-663. Hamilton, W.B., Subduction systems and magmatism. In: Smellie, J.L. (ed.), Volcanism associated with extension at consuming plate margins. Geological Society London Special Pubhcation No. 81,1995, 3-28. Bradshaw, J.D., Cretaceous geotectonic patterns in the New Zealand region. Tectonics, 1989, 8, 803-820. Naeser, C.W., Thermal history of sedimentary basins from fission track dating of subsurface rocks. In: Aspects of Diagenesis. Society of Economic Paleontologists and Mineralogists. Special Publication, 1979,26,109-112. Green, RE, Duddy, LR., Gleadow, A.J.W., Tingate, RT. & Laslett, G.M., Thermal annealing of fission tracks in apatite: 1- a qualitative description: Isotope Geoscience, 1986,59,237-253. Gleadow, A.LW., Fission track dating methods: what are the real alternatives? Nuclear Tracks, 1981, 5,1-14. Gleadow, A.J.W., Duddy, LR., Green, RE & Lovering, J.F., Confined fission track lengths in apatite: a diagnostic tool for thermal history analysis Contribution to Mineral Petrology, 1986, 94,405-415. Green, RE, Duddy, LR., Gleadow, A.J.W. & Lovering, J.F., Apatite fission track analysis as a palaeotemperature indication for hydrocarbon exploration. In: N.D. Naeser (ed.). Thermal history of sedimentary basins-methods and case histories. New York, Springer-Verlag, 1989,181-195. Green, RE, Duddy, LR., Laslett, G.M., Hegarty, KA., Gleadow, A.J.W. & Lovering, J.F., Themial annealing of fission tracks in apatite 4. Qualitative modelling techniques and extensions to geological timescales. Chemical Geology (Isotope Geoscience Section), 1989, 79,155-182. Brown, R.W., Summerfield, M.A., and Gleadow, A.J.W., Apatite fission track analysis: its potential for the estimation of denudation rates and implications for models of long-term landscape development. In: Kirkby, M.J. (ed.). Process Models and Theoretical Geomorphology. John Wiley & Sons. 1994,23-53.

473

32.

33.

34. 35.

36. 37. 38. 39. 40. 41. 42.

43.

44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Raza, A, Korsch, R.J. & Hill, K.C., Apatite fission track thermochronology of selected wellsfiromthe Gunnedah Basin, eastern Australia. In: Flood, RG. & Aitchison, J.C., (eds.), New England Orogen, eastern Australia. University of New England, Amiidale, 1993,137-145. Raza, A, Hill, K.C. & Korsch, R.J., Mid-Cretaceous regional uplift and denudation of the Bowen-Surat Basins, Queensland and its relation to Tasman Sea rifting. In: Supplement to Follington, I.L., Beeston, J.W. & Hamilton, L.H., (eds.), Bowen Basin Symposium 1995 Proceedings. Geological Society of Australia, 1995,1-8. Gleadow, A.J.W. & O'Brien, P.E., Apatite fission track thermochronology and tectonics in the Clarence-Moreton Basin of eastem Australia. AGSO Bulletin, 1994,189-194. Green, P.F., The relationship between track shortening and fission track age reduction in apatite: combined influences of inherent instability, annealing anisotropy, length bias and system calibration. Earth and Planetary Science Letters, 1988,128, 335-352. Sambridge, M.S., & Compston, W., Mixture modelling of multi-component data sets with apphcation to ion-probe zircon ages. Earth and Planetary Science Letters, 1994, 128, 373-390. Galbraith, R.F., The radial plot: graphical assessment of spread in ages. Nuclear Tracks and Radiation Measurements, 1990, 5, 3-14. Veevers, J.L, Powell, C. McA. & Roots, S.R., Review of seafloor spreading around Australia. I. Synthesis of the pattern of spreading. Australian Joumal of Earth Sciences, 1991, 38, 373-389. McDougaU, I, Maboko, M.A.H., Symonds, P.A., McCulloch, M.T., WiUiams, LS. & Kudrass, H.R., Dampier Ridge, Tasman Sea, as a stranded continentalfiragment.Australian Joumal of Earth Sciences, 1994,41, 395-406. Collins, W.J., A reassessment of the 'Hunter-Bowen Orogeny': Tectonic implications for the southem New England Fold Belt. Australian Joumal of Earth Sciences, 1991, 38,409-423. Welknan, P., Eastem Highlands of Australia; their uplift and erosion. BMR Joumal of Australian Geology and Geophysics, 1987,10, 277-286. Lister, G.S. & Etheridge, M.A., Detachment model for the uplift and volcanism of the Eastem Highlands. In: R.W. Johnson, (ed.), Interplate Volcanism in Eastem Australia and New Zealand, Cambridge University Press, New York, 1989, 297-312. O'Sullivan, P.B., Kohn, B.P., Foster, DA. & Gleadow, A.J.W., Fission track data from the Bathurst Batholith: evidence for rapid mid-Cretaceous uplift and erosion within the eastem highlands of Australia. Australian Joumal of Earth Sciences, 1995,42, 597-607. Lambeck, KL and Stephenson, T., The post-Palaeozoic uphft history of south-eastern Australia. Australian Joumal of Earth Sciences, 1986, 33, 253-270. Bishop, P., Geomorphology and evolution of the eastem highlands. In: Johnson, W. (ed.). Intraplate Volcanism in Eastem Australia and New Zealand, Cambridge University Press, New York, 1989, 21-26. Summerfield, M.A., Global Geomorphology. Longman, London and Wiley, New York, 1991, 537pp Marshallsea, S., Thermal History of the Bowen Basin Queensland: an apatite fission track study, Ph.D. Thesis, University of Melbourne (unpublished), 1985. Duddy, I. & Green, P.F., Tectonic development of the Gippsland Basin and Environs: Identification of key episodes using apatitefissiontrack analysis. Gippsland Basin Symposixim, Melbourne, 22-23 June 1992,111-120. Gallagher, K., Dumitru, T.A. & Gleadow, A.J.W., Constraints on the vertical motion of eastem Australia during the Mesozoic. Basin Research, 1994, 6, 77-94. Hill, K.C., Hill, KA., Cooper, G., O'Sullivan, A.J., O'Sullivan, RB. & Richardson, M.J., Inversion around the Bass Basin, SE Australia. Geological Society of London Special Publication. 1995, 525-547. Schmidt, P.W. & Embleton, B.J.J., Magnetic overprinting in southeastern Australia and the thermal history of its rifted margin. Joumal of Geophysical Research, 1981, 86, 3998-4008. Schmidt, P.W., Lackie, MA. & Anderson, J.C., Palaeomagnetic evidence for the age of the Lapstone Monocline, NSW. Australian Coal Geology, 1995,10,13-22. Bai, G.P., Eadington, P.J. & Hamilton, P.J., Fluid history investigation of Permo-Triassic sandstones in the Sydney Basin. CSIRO Division of Exploration Geoscience, North Ryde, Restricted Report 417R, 1993. Middleton, M.F. & Schmidt, P.W., Paleothermometry of the Sydney Basin. Joumal of Geophysical Research, 1982, 87, 5351-5359. Russell, N.J., A Palaeogeothermal study of the Clarence-Moreton Basin. AGSO Bulletin, 241,1994,237-276. Cull, J.P. & Conley, D., Geothermal gradients and heat flow in Australian sedimentary basins. BMR Joumal of Australian Geology and Geophysics, 1983, 8, 329-337. Gallagher, K., Evolving temperature historiesfi-omapatitefission-trackdata. Earth and Planetary Science Letters, 1995, 136,421-435. Laslett, G.M., Green, RE, Duddy, LR., & Gleadow, A.J.W., Thermal annealing of fission tracks in apatite, 2, A quantitative analysis. Chemical Geology (Isotope Geoscience Section), 1987, 65,1-13. Eraser, A.R. & Tilbury, L.A., Stmcture and stratigraphy of the Ceduna Terrace region, Great Australian Bight Basin. APEA Joumal, 1979,19, 53-65. Cloetingh, S., Interplate Stress: A new element in basin analysis. In: Kleinspehn, K.L. & Pola, Chris., (eds.). New Prospectives in Basin Analysis. Springer-Verlag, 1988,205-230. 474

61. 62.

63. 64.

65. 66. 67. 68.

69.

70.

71.

72.

Ziegler, P.A., Evolution of the Arctic-North Atlantic and the western Tethys. American Association of Petroleum Geologist Memoir, 1988,43,198pp. Etheridge, M., McQueen, H. & Lambeck, K., The role of interplate stress in Tertiary (and Mesozoic) deformation of the Australian continent and its margins: a key factor in petroleum trap formation. E?q)loration Geophysics, 1991,22, 123-128. Scotese, Ch. R., Cahagan, L.M. & Larson, R.L., Plate tectonic reconstruction of Cretaceous and Cenozoic ocean basins. Tectonophysics, 1988,155, 27-48. Janssen, M.E., Stephenson, RA. & Qoetingh, S., Temporal and spatial correlations between changes in plate motions and the evolution of rifled basins in Africa. Geological Society of American Bulletin, 1995, 107, 11, 1317-1332. Kamp, P.J., Late Cretaceous-Cenozoic tectonic development of the Southwest Pacific Region. Tectonophysics, 1986,121,225-251. Stock, J. & Molnar, P., Revised history of the early Tertiary plate motion in the southwest Pacific. Nature, 1987, 325,495^99. Tulloch, A.J., Kimbrough, D.L., The Paparoa Metamorphic Core Complex, New Zealand: Cretaceous extension associated withfragmentationof the Pacific margin of Gondwana. Tectonics, 1989, 8, 6,1217-1234. Spell, T.L., McDougall, L, Thermochronology of a metamorphic core conplex associated with the breakup of Gondwana: Paparoa Range, South Island, New Zealand. Abstracts of the Eighth Intemational Conference on Geochronology, Cosmochronology and Isotope Geology, U.S. Geological Survey Circular 1107. 1994, 300p. Muir, R.J, Bradshaw, J.D. & Weaver, S.D., Crustal extension prior to the opening of the Tasman Sea Basin: Evidence from New Zealand granites. In: Gerrit J. Van Der Lingen, Kerry M. Swanson & Roderick J. Muir, (eds.) Evolution of the Tasman Sea Basin. A A. Balkema/Rotterdam/ Brookfield, 1994, 55-64. Laird, M.G., Geological aspects of the opening of the Tasman Sea. In: Gerrit J. Van Der Lingen, Kerry M. Swanson & Roderick J. Muir, (eds.) Evolution of the Tasman Sea Basin. A.A. Balkema/Rotterdam/ Brookfield. 1994,1-17. Weaver, S.D., Storey, B.C., Pankhurst, R.J., Mukasa, S.B., DiVenere, V.J., & Bradshaw, J.D., Antarctica - New Zealand rifting and Marie Byrd Land lithospheric magmatism linked to ridge subduction and mantle plume activity. Geology, 1994, 22, 811-814. Gilchrist, A.R & Summerfield, M.A., Tectonic models of passive margin evolution and their implications for theories of long-term landscape development. In: Kirkby, M.J. (ed.). Process Models and Theoretical Geomorphology. John Wiley & Sons. 1994, 55-84.

475

LATE TRIASSIC VOLCANISM OF THE IPSWICH BASIN Andrew Roach Macquarie University Summary The initiation of the Late Triassic intra-montane Ipswich Basin in south-east Queensland is marked by the eruption of intra-cratonic, rift-related mafic and felsic volcanic sequences. Within the Ipswich area, the Sugars Basalt and Weirs Basalt, both basaltic-andesites, locally form the lowest unit in the basin. Younger, pyroclastic rocks occur interbedded within the overlying epiclastic rocks. The most significant of these are the ignimbrites and air-fall tuffs of the rhyodacitic Hector Tuff, and the rhyolitic tuffs of the Mount Crosby Formation. In the Brisbane area to the east, the rhyolitic Brisbane Tuff forms the base of the basin. The multiple ignimbrites that comprise most of the Brisbane Tuff were probably erupted from a single vent over a short time. To the south, the Chillingham Volcanics crop out on either side of the NSW - Queensland border, and include rhyolitic lava flows and domes, pyroclastic rocks, and epiclastic sedimentary rocks. Other surface occurrences of volcanic rocks within the Ipswich Basin include the rhyolitic and andesitic lavas of Moreton Island and the rhyolitic lavas of Stradbroke Island. In DDH GSQ26, the basal section comprises seven volcanic sub-units (1) ranging in composition from basaltic-andesite to rhyolite, and occurring as pyroclastic rocks and lava flows. Introduction The Ipswich Basin is one of several intermontane basins which formed within the New England Foldbelt during the Late Triassic. The basin is characterised by the thick accumulations of coal measures which were predominantly deposited on an alluvial plain (2). The commencement of the basin however is marked by the eruption of a basal sequence of mafic and felsic volcanic rocks. These volcanic rocks form part of a major silicic igneous province of Late Triassic age occurring throughout Southeast Queensland which also includes: the Agnes Water Volcanics (3); the Arrambanga Volcanics (3,4); the Mount Byron Volcanics (3); the Muncon Volcanics (3); the North Arm Volcanics (5, 3); and numerous plutons occurring on the eastem side of the New England Batholith. This paper focuses on the volcanic stratigraphy of the early basin forming volcanic rocks of the Ipswich Basin: the Weirs Basalt; Sugars Basalt; Tuff within the Mt Crosby Formation; Hector Tuff; Brisbane Tuff; Chillin^am Volcanics; Stradbroke Island rhyolites; Moreton Island volcanics; and a suite of volcanic rocks form DDH GSQ 26. Mafic Rocks of the Ipswich Area -The Weirs Basalt and Sugars Basalt The Sugars Basalt and Weirs Basalt crop out in the Moggill and Mt Crosby areas respectively (see Figure 2). Both formations are of basaltic-andesite composition and, despite stratigraphic separation (6), are probable equivalents. Geochemically the basalts of the two units are indistinguishable. The Weirs Basalt, which attains a maximum thickness of 30m in DDH NS295 (6) and 11m at it's type section (7), is defined as an altered porphyritic basalt (7). Underlying the unit both at the type section, and within DDH NS295, are members of the Blackwall Breccia, which is inferred to be a scree deposit derived from the local Palaeozoic basement. Although closure of the Weirs Basalt is limited, at least four lavaflows,including amygdaloidal and non-amygdaloidal, porphyritic and aphyric basalts comprise the formation. The Sugars Basalt's maximum recorded thickness of 105 m is in DDH IC 256 (6), and has a thickness of approximately 37m at it's type section (7). The formation is defined by Cranfield (7) as an amygdaloidal basalt. At the type section of the Sugars Basalt, at least seven separate lava flows and an air-Ml tuff occur. Houston's (6) drill log for DDH IC256 indicates the presence of two additional lava flows occurring above those at the type section. The lithologies of the flows within the type section include essentially non-porphyritic lavas, porphyritic lavas, and amygdaloidal versions of the aforementioned types. The development of cooling columns which are restricted to individual flows indicates that the thermal history of individual flows is unique and that at least small time gaps separated the individual flows. Given the comparatively small thickness of the flows, these breaks in activity may not have been significant. The air-fall tuff within the Sugars Basalt at the type section is light green-grey coloured, and contains abundant feldspar and pyroxene phenocrysts. The mineralogy of the tuff is consistent with it being of basaltic-andesite composition and directly relate to the lava flows within which it is interbedded. Basaltic-andesite vents are not usually prone to extensive magmatic pyroclastic eruptions, the results commonly being a proximal splatter cone derived from fire-fountaining. Therefore the presence of the air-fall tuff within the upper sequence of the Sugars Basalt probably represents the product of a phreato-magmatic eruption. The extensiveness of this tuff indicates that the tuff is not the result of a secondary eruption caused by the lavas entering water. Instead, it imphes that molten lava was emplaced directly into water. lUjff within the Mt. Crosby Formation The Mount Crosby Formation (see Figure 2) conformably overlies the Blackwall Breccia, disconformably overlies the Weirs Basalt and Sugars Basalt (7), and consists of polymictic conglomerates with comparatively minor quantities of 476

mudstone and arenite. Minor amounts of lithic tuff have also been reported within the Mt. Crosby Formation (7). Surface occurrences of tuff are hmited, however two tuffaceous sections separated by a small sequence of interbedded tufifaceous arenite and mudstone were observed near Pine Mountain. Within both DDH NS295 and DDH NS256 Houston (6) recorded the presence of Mt. Crosby formation tuffs. In both examples, two tuff horizons were present, once again separated by fine grained sedimentary rocks. Almond (1) reported the presence of a tuff within the Mt. Crosby Formation in DDH GSQ26, however he did not differentiate how many separate horizons of tuff were encountered. Hawkins (8) described a 25m thick tuffaceous horizon within NS 93. This section has been subsequently reinterpreted as a lateral equivalent of the Mt. Crosby Formation, and has the two characteristic tuff horizons separated by shale beds. The phenocryst content of both tuffs was made up of plagioclase feldspars, quartz, alkali-feldspars, and possible trace biotite. The approximate thickness of the upper tuff is approximately 1.3 m , while the lower tuff's thickness varies between 3.7m (DDH 295) and 5.3m (DDH 256). The upper, slightly coarser grained tuff contains abundant dense lithic clasts which were composed exclusively of black chert (average size 3mm). Small, comparatively unflattened pumice clasts were also observed within the upper tuff. This, together with the foHated appearance (parallel to bedding) of the bed when weathered, suggests that the upper tuff is ignimbritic in origin. The lower tuff contains chloritically altered glass shards, pumice clasts, and clay clots psudomorphing pumice. The lower sequence possesses an ignimbrite like texture and parallel clasts, indicating that it too may be ignimbritic in nature. Commonly informally regarded as a lateral equivalent of the Brisbane Tuff, the two tuffs of the Mt. Crosby Formation, both of which are probably ignimbritic in origin, have no conclusive correlation with the Brisbane Tuff. Chemically the rocks are more dacitic than the Brisbane Tuff, and their dense lithic clast content is different from that of the Brisbane Tuff, being both depleted in comparative volume and dominated exclusively by black cherts where as the Brisbane Tuff contains chert and phyllitic clasts. Additionally, the Brisbane tuff shows evidence of a greater degree of crystal enrichment than the Mt. Crosby Formation. Potentially the Mt. Crosby formation could be representative of a non-fines depleted, or possibly fines enhanced Brisbane Tuff, however chemical differences suggest a different eruption. Hector Ibff The Hector Tuff (See Figure 2) consists of air-fall tuffs, ignimbrites, mudstone, arenite, and minor amounts of conglomerate shale and coal. The formation, which is a member of the Kholo Subgroup, conformably overlies the Colleges Conglomerate and is conformably overlain by the Cribb Conglomerate. The formation occurs extensively throughout the Ipswich Area, however good outcrops occur only near the type section, where the unit is 22m thick (9,7). The unit progressively thins eastward, and this thinning is accompanied by an apparent reduction in the amount of primary volcanic material preserved within the formation. The tuff is of rhyodacitic composition. Most exposures of the formation consist of reworked, tuffaceous arenite or mudstone. Primary volcanic rocks observed within the formation include thin ignimbrite units, air-fall tuffs, and an accretionary lapilli tuff. The ignimbrites, which are usually grey-green in colour, show no evidence of welding. The clast content of the ignimbrites was difiRcult to determine due to the weathering of most outcrops, however small pumice clasts (up to 3 cm) and rare chert clasts were identified at some localities. The thickest individual ignimbrites observed in outcrop were only approximately 1.5m thick. Air-fall tuffs are the volumetrically most abundant of the pyroclastic rocks within the formation. Although primary airfall tuffs themselves are not as volumetrically significant as the ignimbrites, when the volumes of re-worked tuff (preserved in the form of tuffaceous arenites) are included, they take on greater importance. The air-fall tuffs are typically green in colour, are moderately fine grained, and rarely porphyritic. The thickness of observed primary occurrences of tuff seldom exceed 1.5 m, however re-worked tuffs (in the form of tuffaceous arenite) accumulated in thicknesses up to 5m. Brisbane T^ff The Brisbane Tuff comprises rhyolitic ignimbrite and minor air-fall tuff, conglomerate, breccia, and volcano lithic arenite. It locally forms the base of the Ipswich Basin around Brisbane, but elsewhere is commonly deposited on thin breccia or rarely on a few metres of fluviatile sediment. The unit outcrops over a distance of approximately 50km, in a northwesterly striking band extendingfiromNarangabah north of Brisbane, through to Woodridge in the south (see Figure 3a). The best outcrops of the unit are within the vicinity of the city of Brisbane, and it's inner city suburbs. The Brisbane Tuff's thickness ranges from less than a metre to more than 50 metres. Maximum thickness of the unit has been variously quoted as 300 feet (96 metres) (10), 700 feet (225 metres) (11), and 250 metres (9). The maximum observable thickness however is only approximately 40m, and occurs in the Kangaroo Point CUffs. Ground surges arefi-equentlypreserved at the disconfformal contacts with the underlying basement rocks. The surge layers are typically comparatively crystal rich, and contain abundant charcoal, but are otherwise essentially deficient in lithic clasts. The ground surge horizon is normally comparatively thin with poorly defined cross beds, but at a disused quarry at Windsor, prominent cross beds occurring in a thicker horizon indicate a local flow direction with a strike of 150-330". Air-fall tuff deposits, often exhibiting mantle bedding, also underlay the main ignimbrite body. These air-fall tuffs are usually crystal rich, and are commonly silicified or stratified. 477

Multiple Ignimbrite Theory Both direct and indirect evidence exists to support the theory that the Brisbane Tuff is constituted of multiple ignimbrites. The presence of interbeded fluviatile sediments (12, 9, 13), layers of accretionary lapilli air-fall tuff (14), weathered layers (13), and air-fall tuff within the ignimbrite provides the best direct evidence for multiple ignimbrites. Additionally three distinct types of pumice clasts have been recognised within different samples of the Brisbane Tuff: 1. 2. 3.

A non-porphyritic pumice clast; A porphyritic quartz and felspar rich pumice clast; and A porphyritic quartz and biotite rich pumice clast.

The most common type of clast is the non-porphyritic clast. In any one outcrop only one of the above types of pumice clast is present. The presence of these three distinct types of pumice clasts indicates differing material being erupted at vent, and hence provides further evidence of multiple ignimbrite flows constituting the Brisbane Tuff. Weathered horizons (13) within the Brisbane Tuff imply that a significant time break may have occurred between eruptions. However all outcrops studied indicate that the Brisbane Tuff occurs as a single cooling unit, therefore the break between eruptions must have been short enough for retention of heat within the ignimbrite. Additionally charcoal is conspicuous within the lower portions of the ignimbrite and surge layers, indicating the mass destruction of vegetation during the early eruptions. No conspicuous charcoal is visible within the upper ignimbrites suggesting that no significant re-vegetation occurred between eruptions. Fossil Fumaroles Fossil fimiaroles are common within the Brisbane Tuff, and range in sizefiromsmall pipes and joints of less than 1cm in width to large fumaroles of several metres in width. The smaller fiimaroles are typically infilled with opal or goethite whereas the pronounced alteration halos of the larger fumaroles are composed of illite and muscovite. Where it is possible to establish stratigraphic control on the position of the fumaroles withbci the ignimbrite, they are usually restricted to the lower sequences of the ignimbrite. Some of the fiimaroles are apparently rootless, but for many of the roots are not exposed. Deposition Environment and Eruptive History Much of the preserved extent of the Brisbane Tuff was believed to have been deposited as a valley fill ignimbrite. Most preserved contacts between the basement rocks and the Brisbane Tuff are steep (some almost sub-vertical) when compared to the orientation of bedding within the ignimbrite. Some localities have fossil scree slopes of brecciated basement material preserved under the overlying ignimbrite, which is consistent with there being steep valley walls. Prominent near horizontal cooling columns preserved near the contact with the basement at Windsor also indicate that the cooUng surface (a valley wall) was near vertical. In addition to the flow lineations parallelling or sub parallelling the outcrop orientation, the presence of many fumaroles near the base of the ignimbrite showing evidence of water/ignimbrite interaction is also consistent with the ignimbrite being emplaced into a river valley. Chillingham Volcanics The Chillingham Volcanics crop out in a 2-5km wide strip extending for approximately 100km, lying about 25km inland firom the south-east Queensland/ north-eastem New South Wales coastline (see Figure 4). They crop out as a discontinuous north-north westerly striking belt of gently westward dipping rocks. Both pyroclastic and efiusive rhyohtic rocks are preserved within the Chillingham Volcanics, and most sites ^ow evidence of two phases of activity - an early pyroclastic phase and a later lava effiision phase. Pyroclastic Phase Pyroclastic rocks within the Chillingjiam Volcanics occur at the base of the Chillingjiam Volcanics and mark the initiation of volcanic activity and a depositional environment where ever the unit is preserved. Pyroclastic rocks are volumetrically most significant in the northem occurrences of the Chillingham Volcanics where only one eruptive cycle is preserved. Where two cycles of activity are preserved, the initial pyroclastic rock deposits (prior to lava effusion) are the thickest, with comparatively thin deposits occurring between successive lava effusion cycles. The pyroclastic rocks preserved within the Chillingham Volcanics include ignimbrites, air-fall tuffs (some with accretionary lapilli), and pyroclastic surges, with the volumetrically most significant being the ignimbrites and air-fall tuffs. The thickest individual ignimbrites observed within the Chillingham Volcanics are in the order of 200m thick, while the thinnest are less than Im thick. Most are several tens of metres thick. The ignimbrites of the Chillingham Volcanics have a tremendous range in the lithologies of dense lithic clasts. Unlike the ignimbrites of the Brisbane Tuff, which essentially only have fragments of Palaeozoic Basement rocks as the common dense lithic component, most dense lithic clasts in the ignimbrites of the Chillingjiam Volcanics are rhyolite. Commonly most of these clasts are of flow banded lavas, with small proportions of spherulitic lavas and re-worked ignimbrites.

478

Perhaps because the preservation of air-fall tuffs within the Chillingham Volcanics would have been more dependant on the palaeo-environment than the ignimbrites, the ignimbrite fades are quite coimion throughout the ChilUngham Volcanics, while air-fall tuffs are restricted to sections which also have epiclastic sedimentary rocks, or in rare cases where air fall tuffs have been preserved between ignimbrites. No air-fall tuffs have been observed preserved in isolation from either sedimentary rocks or ignimbrites. Lava Flow Phase Lava flows volumetrically constitute most of the Chillingham Volcanics, and probably mark the termination of volcanic activity within the unit, as no evidence of late intrusions that might have fed still younger Triassic volcanism has been identified within the outcropping examples of the Chillingham Volcanics. All areas of the Chillingham Volcanics have at least one cycle of effusive rhyolitic activity preserved. Typically the lava flows of the Chillingham Volcanics possess a pervasive flow banding fabric, which range in scale from the microscopic through to flow bands (or domains) several metres in width. Commonly the larger bands of decimetre scale or larger, segregate domains of finely flow banded rhyolitic lava from more coarsely flow banded lava, spheruhtic lava, or apparently massive lava. The areas of near horizontal flow banding, which constitute most flows, have been interpreted as being the distal parts of theflows/domes,while the near vertical portions have been interpreted as either being the surface manifestation of feeder dykes or the feeder dykes themselves. Studies of the orientation of the originally vertical and near vertical flow band zones have revealed that there is a consistent north-westerly strike possessed by the majority of the flows. This strike is pervasive in all flows except for some south of Mt. Waming which strike east-west. The pattem of the flow fabrics is strongly suggestive of a fissure type eruption mechanism for the effusion of the lava flows. The strike of these vertically oriented zones also closely parallels the outcrop orientation of the Chillingham Volcanics, and the orientation of the Ipswich Basin itself, inferring that the eruptive fissures may be related to deep structures associated with the basin's formation. Volcanics on Moreton Island Moreton Island is essentially a sand island of Quatemary age except for the rocks which form the north-eastem point on the island (see Figure 3b). These rocks belong to three groups; undifferentiated Late Triassic volcanic rocks; Late Triassic Ipswich Basin sediments; and sediments of the Early Jurassic Woogooroo Subgroup of the Clarence Moreton Basin. Within the sequence of volcanic rocks, two compositional types are present - rhyolitic lavas in the northem most outcrops and a series of sub-aqueously emplaced mafic lavas on the eastem most outcrops on the island. At least two tuffaceous arenites, probably representing re-worked tuffs, were observed within the Ipswich Basin sediments overlying the mafic lava flows. The most common facies preserved within the rhyolite lavas on Moreton Island is that of flow banded lava. The flow bands observed on Moreton Island are exclusively within the mm domain size, commonly highly contorted, and typically steeply dipping. Spheurilte growth within the lavas was largely restricted to the sub-mm scale, except for one narrow zone near North Point where they are up to 5 cm. Emplacement History At least two episodes of lava eruption are inferred as having occurred on Moreton Island. The first phase, which was a lava dome phase, is responsible for most of the rhyolitic lavas on the island. It is characterised by steeply dipping monotonousflowbanded lava, with no observed spherulitic patches and little autobreccia. The younger lava, whose mode of emplacement could not be determined, contains more varied textures, including spherulitic and autobreccias. A significant time break occurred between the emplacement of the first lava dome, and subsequent emplacement of the second lava. At Honeymoon Bay, steeply inclined flow banded rhyolitic lava of the earlier dome is abruptly truncated by an erosive contact, and subsequently overlain by crumble breccias of the later lava. The break between flows was long enough to allow for the de-vitrification of lavas of the first dome, and for significant erosion to expose the (originally) glassy centre of the dome. Rhyolite on Stradbroke Island Stradbroke Island, like Moreton Island, is essentially a sand island. Quatemary sand deposits have built up behind a rock barrier composed of undifferentiated rhyoUtic lavas on the north-eastem comer of the island (see Figure 3c). The rhyolites which crop out on Stradbroke Island are typically flow banded on the mm to cm scale, have porphyritic textures, and contain abundant quartz phenocrysts. Most of the rocks which comprise Stradbroke Island are quartz phenocryst rich rhyolitic lavas and autobreccias. Some rocks occurring at Adder Rocks and Point Lookout have an ignimbritic looking texture; however such occurrences are isolated to small zones, and the absence of recognisable fiamme or pyroclastic flow structures probably precludes the possibility that they are truly ignimbritic. These ignimbrite like rocks at Point Lookout contain clasts of rhyolite, and are overlain by quartz rich porphyritic flow banded rhyolitic lavas identical to those Qxpostd elsewhere on the island. The rhyolite clasts are altered, probably metamorphosed, dark grey coloured, non-porphyritic, and (unlike their host) are

479

quartz phenocryst poor. It is likely these rocks represent lenticular zones within the lava flows, containing small blocks of older lava flows incorporated as xenolith material. Eruptive History At least three eruptive periods are represented on Stradbroke Island. The earliest phase of effusive activity produced a non-porphyritic, flow banded rhyolite which is only preserved as metamorphosed xenoliths within other flows on the island. At least three generations of porphyritic lavas were then emplaced from at least two different domes located at Adder Rock and South Headland. Age relationships between the two domes were impossible to determine as no crosscutting relationships were observed. Texturally these rocks bear a close resemblance to the ignimbrites of the Brisbane Tuff, and given the similar stratigraphic horizon the units occupy, it is possible that the two units may be related. Volcanic Rocks in GSQ 26 Thick accumulations of volcanic rocks of assumed Late Triassic age have been intersected in many drill holes. The thickest suite uncovered so far were those in DDH GSQ 26 (see Figure 1). This drill hold intersected 350 m of undifferentiated volcanic rocks which ranged in composition from rhyolites and dacites through to andesites. Almond(l) identified a total of seven previously undescribed sequences of volcanic rocks. Ahnond (1) correlated two of these undescribed units with the Chillingham Volcanics (Subunit E) and the Sugars basalt (Subunit F), however neither of these units show any chemical affinity with either of these units, suggesting no genetic link. Additionally, none of the volcanic rocks of GSQ 26 show chemical affinities with any other volcanic or plutonic rocks occurring in the Ipswich Basin, suggesting they foraiedfiroma unique vent (15). Conclusions The volcanic history preserved within the Ipswich Basin suggests two distinct phases of activity, an older phase of mafic volcanism followed later by a phase of felsic volcanism. The mafic volcanic rocks within the basin are basaltic-andesites and andesites. Although only stratigraphically recognised as the Weirs Basalt and Sugars Basalt in the Ipswich area, other mafic rocks have been identified within drill holes and on Moreton Island. These rocks are highly altered and a small disconformity exists between these rocks and other Ipswich Basin rocks, suggesting a brief pause between their efiusion and the deposition of the rest of the basin. Chemically these basalts have signatures of rifl-related basalts and andesites (15), which is in character with the extensional environment inferred for the Ipswich Basin in the Late Triassic. The second phase of volcanic activity, which is better preserved, apparently occurred sporadically though much of the Late Triassic. The volcanic rocks produced during this phase were dominantly rhyolitic in composition, as evident from the Chillingham Volcanics, Brisbane Tuff, and the Moreton and Stradbroke Island rhyolitic lavas. Rhyodacitic rocks were also produced, although they are volumetrically insignificant compared to the rhyolites, and occurred predominantly within the Ipswich Area. Multiple centres of activity have been inferred for this period, with most units possessing distinctive chemical signatures (15). All rocks, however, are continental and rift-related. References 1.

ALMOND C.S., 1982. Stratigraphic drilling report - GSQ Ipswich 26, Queensland Government Mining Journal V83,p 514-523

FALKNER, A.J., FIELDING C.R., SAUNDERS B.J., 1988. The Ipswich and Walloon Coal Measures, in, HAMILTON, L.H., ed., 1988. Field Excursions Handbook for the ninth Australian Geological Convention, Geological Society ofAustralia, Queensland Division, p 81-94

DAY R.W, WHITAKER W.G., MURRAY C.G., WILSON I.H., and GRIMES K.G. 1983. Queensland Geology, Geological Survey of Queensland Publication 383

STEPHENS C., 1986. Late Triassic Volcanism near Gayndah, In, WILLMOTT W.F. ed., 1986 Field Conference, South Burnett District, Geological Society ofAustralia, Queensland Division, p 32-38

ASHLEY RM. & DICKIE G.J., 1988. North Arm Volcanics and Associated Epithermal Gold Mineralization at North Arm Prospect, in, MURRAY C.G. & WATERHOUSE J.B. eds, 1987. Field Conference Gympie District, Geological Society ofAustralia, Queensland Division, p 60-69

HOUSTON, B.R., 1965. Triassic Volcanics From The Base of the Ipswich Coal Measures South-east Queensland, Geological Survey of Queensland, Publication 221

CRANFIELD L.C., HUTTON L.L, GREEN P.M., 1989. Ipswich 1:100 000 Geological Map Commentary, Queensland Department ofMines

HAWKINS, b.w., 1956. Ipswich Borehole N.S. 93 Cooneana estate, Queensland Government Mining Journal, V57,p 214-219

CRANFIELD L.C., SWHARZBOCK, H., & DAY R. W., 1976. Geology of the Ipswich and Brisbane 1:250 000 Sheet areas Geological Survey of Queensland, Report 95 480

10. 11. 12. 13. 14. 15.

BRIGGS C., 1928. The Brisbane Tuff, Proceedings of the Royal Society of Queensland. V40, p 147-163 BRYAN W.H. & JONES O.A. 1960. Brisbane and South East Moreton. Journal of the Geological Society of Australia. V7,p262-263 HIGGINSON 1942. A general study of the Mesozoic sediments in the Brisbane area east of Oxley, University of Queensland Unpublished Honours thesis HOUSTON B.R. 1967. Geology of the City of Brisbane, Part H - The post- Palaeozoic sediments and volcanics. Geological Survey of Queensland, PubUcation 324 RICHARDS H.C . & BRYAN W.H. 1927. Volcanic mud baUs from the Brisbane Tuff. Proceedings of the Royal Society of Queensland, V37, p 55-60. ROACH A,F. (in prep.) Volcansim in the Late Triassic Ipswich Basin of Eastem Australia, Macquarie University Unpublished PhD Thesis W \ N FIGURE 3 \ C>f W ^ B INSET S

) Moreton Island

SCALE km 0

V BRISBANE _

IPSWICH

INSET 3c V

^ S FIGURE 2 \

cT North Stradbroke 1 Island /

( ^/ V ^^

FIGURE GOLD COAST

Canungra

IA] QLD

/ MURWILLUMBAH

NSW Figure 1. Locality map showing the positions of the detailed maps, and the location of DDH GSQ26.

481

\ \ \ \ \ \ \ \ \ S

\ \

f'V^vl (basaltic-andesite flows and tuffs)

Mt. Crosby Formation

Weirs Basalt

\ V\ \ \ \ H \ Xg w w w w w v

\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ^ \ \ \ \

77771 Sugars Basalt

Tuff (rhyolitic air-fall tuff)

(basaltic-andesite flows)

Palaeozoic

(phyllite, chert, arenite, basalt & serpentinite)

Diamond Drill Hole

v . \

\ ' \

\ V v v

\ \

. . .

r'- V V\ V\ -LT. Caboolture • ' -'^^' • • • I. V. . . . '. O • .•••' '.KO'-' A \ ~ 70000^ \ W

70110 ^

North Point

\ \

PACIFIC OCEAN

....

\ \ • '' V-;:- \ . •

MORETON

\ \ \ \ V, V V Vo , V V V , ^-.^^Narangabah.-o. -69900 V V X

v.-'g-.-'r^

: • ••« -

ISLAND :

200 m-

\ \

-'sssoor^T-'ri^.s,!.-^ \

\ \ ^ \ \ \ \

Xf Slrathpine O". ; -70090 S.-:

•• o

\ ' ' — \

w \

hl-:-;;'

-69600 \

^ BRISBANE

IPSWICH^ - 69400.

•. 00 .

- • o ' o

—L-J

\ ••••• • i-

' H

^ T ^ . Cainozoic

(Ipswich Basin and Clarence Moreton Basin) -t

I 10 km

BrisbaneTufF

0 1

Moreton Island Andesites

• • • • 1 (alluvium, sedimentary and volcanic rocks)

Mesozoic sediments

•• . A v.-rrf

\ \

Mid-Late Triassic igneous intrusives (granite, granodiorite, & diorite)

Palaeozoic Basement (greywacke. chert, meta-basalts. phyllite, and serpentinite)

Rhyolitic Lava

Figure 3. Distribution of volcanic rocks of Triassic age in the Brisbane area, (a) Shows the distribution of the Brisbane tuff and of some Late Triassic plutons. Inset (b) shows the distribution of the undifferentiated rhyolitic and mafic lava flows on the north-eastem tip of Moreton Island. Inset ( c) shows the distribution of rhyolitic lavaflowson the northeastem tip of North Stradbroke Island. (All grid references are in AMG)

483

:;69200 s - . v - K

v - ^ c R

V ^ i

SOUTH STRADBROKE ISLAND

\ ^

\ \ \ \

•^.O.Coomera-

W ^ A

(-r

\ \ \

\ W

r ar a .: J^ -. .> i i \ •CCa an nu un no g

v s ^

•

x \

W / i , Southport J

Surfers

\/A \ \

\ \

\ V ^ " . -

G O L D C O A S T

>< 'f:;-, \ \ \ \ \ \ \ \ \

% >< ^

W i-.

;;Valley .N N -68^0/- ^

W \

\ X

W \

\ \ \ \

^ X ^ < ^ A <

r \ \

'N^'^SpringbrookxT u

\ \ \ \ \ \ \ \ V,

V/a

\ \

Coolangatta

\ \

y• —

vS^^^ \

Tyalgum VcUV^

'.sseoo;;:.^ -i. -7 I ^ N >< ^ V, ^ A < ^

•• • i M t V\feming H

\ X W

A <;:• z. •

\ \ \ \ L. ^

'. '

N^-.v:;-.z. ^ i N ^ M ^ A ^

\v \v \v \ -l \ v \ -l \ v \v

-c,

A N

•: V '

w w \ \ \

•^Z.'

•cyv

q Cainozoic Cover -

(Alluvium and sedimentary rocks)

< Cainozoic Voicanics (ML Warning suite)

Chillingham Voicanics \ ^ Palaeozoic Basement

Mesozoic Cover

(Ipswich and Clarence Moreton Basins)

Figure 4. Distribution of the Chillingham Voicanics. (Grid references are in AMG) 484

DIAMONDS IN QUEENSLAND Allan D. Robertson Department of Mines and Energy Summary The present knowledge of diamond occurrences in Queensland indicates that diamonds are restricted to two regions - one located in the northwest of the State and a second extending from above Cooktown to the Queensland - New South Wales border along the Great Divide. The diamonds in the northwestem zone appear to have a kimberlite and/or lamproite association. The origins of the diamonds along the Great Divide (eastem zone) are problematical but they appear to be associated with Late Cretaceous to Cainozoic alkali volcanism. Diamond exploration in Queensland Systematic, broad-based exploration for diamonds did not commence in eastem Australia until the late 1960s, when Stockdale E^qjloration Limited and Australian Selection Trust, began what was apparently an Australia wide regional mineral search. Diamond exploration has continued intermittently since that time with a number of other explorers becoming involved. Since the late 1960s, 37 different companies and/or individuals have held Exploration Permits for the purpose of diamond exploration. The number of diamond explorers and permits have fluctuated from year to year generally following overall economic trends, but also reflecting developments in diamond exploration elsewhere. Until the mid-1980s, explorers tended to concentrate their efforts in areas where diamonds had been previously reported, usually during tin, gold or sapphire mining. Given that most companies exploring in Queensland have had their grounding in classical diamond techniques, it is not surprising that the approach to diamond exploration has been initially dominated by concepts and techniques originating in South Africa. Not one was prepared to accept another possible source for diamond. After diamonds were discovered in lamproite in Western Australia, a number of companies modified their prospecting methods. Few explorers, however, have considered seriously the possibility of primary sources other than kimberlite or lamproite in spite of the accumulating body of evidence that other primary sources exist (MacNevin (1); Robertson & Robertson (2); Janse (3); Sutherland et al (4); Meyer et al (5)). Many diamond explorers consider that economic primary deposits are restricted to ancient cratons and flanking mobile zones (Cliffords Rule; Janse (3)) This view is based on diamonds from South Africa and Russia (kimberlites) and from Argyle (lamproite) have been shown to be very old and that a thick lithosphere and associated low geothermal gradients are necessary to allow diamonds to be preserved and later brought to the surfece. By these criteria, diamonds should not occur in eastem Australia where the lithosphere is relatively thin and geothermal gradients comparatively high. This has led some workers to suggest that the alluvial diamonds of eastem Australia have a remote source. However data from recent studies (Sutherland et al (3): Barron et. al (6, 7)) indicate that the geothermal gradients have not always been unfavourable and that "windows" of suitable conditions have existed in the geological past. Distribution of diamond in Queensland Known diamond occurrences in Queensland can be grouped into two regions • northwest Queensland; and • eastem Queensland Northwest Queensland This region is located mainly to the north and northwest of Mount Isa and west of Mount Oxide where more than 160 diamonds have been recovered from alluvial deposits. Although no classical diamond indicator minerals have been recorded and companies consider that diamond pipes are not close by, these diamonds appear to be derived from deposits associated with the ancient craton. On the Northern Territory side of the border, diamond bearing kimberlitic pipes (Emu and Coanjula) have been located in the McArthur Basin. This zone also lies at the eastem end of a microdiamond bearing belt that extends from the Victoria River in the Northern Territory into Queensland. Eastem Queensland The eastem zone extends from the Queensland-New South Wales border to above Cooktown and appears to follow the position of the Great Divide. In many parts of this zone diamonds are found where Late Cretaceous to Cainozoic volcanicity has occurred and at several locations diamonds have been recovered from deposits containing sapphire and zircon. Within this east Queensland zone, there are three areas where more than 50 diamonds have been recovered. These are Stanthorpe, Anakie Sapphire Fields and the Elizabeth Creek area northwest of Mount Surprise.

485

Stanthorpe Within the Stanthorpe region, diamonds were recovered during alluvial tin mining operations. Reports in the news papers of last century indicate that more than 100 diamonds were recovered from sluice boxes during the early period of mining,. However, there are no obvious sources for the diamonds in the vicinity of Stanthorpe. Diamond distribution is confined to the headwaters of one drainage system-the Sevem River and its tributaries. Diamond is not recorded on the eastem fall on the New South Wales side of the border close to this headwaters system or on the westem fall in the Texas area. Sapphire, zircon, garnet, ilmenite, black tourmaline, and spinel are accessory minerals in the stanniferous alluvials. The presence of low chrome pyropic gamet of basaltic afi&nity suggests an input from a basaltic source. Diamonds firom Stanthorpe show similar surface characteristics to those found in the Copeton area and may have a similar origin to the Copeton stones (age approximately 300 Ma and suspected subduction origin). Sapphire Fields The Sapphire fields of central Queensland (Anakie region) have produced a number of diamonds, some exceeding 5 ct in wei^t,firomthe alluvial deposits in the Rubyvale-Sapphire-Keilambete area, Tomahawk Creek and the Willows. The larger stones appear to be more prevalent in the Willows district southwest of Anakie. To the northwest of Rubyvale, small diamonds (less than 1 ct) have been recoveredfiromgravels in the Tomahawk Creek area by amateur prospectors. These diamonds have been recovered during the search for sapphire (and zircon) (Robertson & Sutherland (8)) but classical diamond indicator minerals have not been found. Elizabeth Creek Diamonds in the Mount Surprise area are confined mainly to the Elizabeth Creek drainage system downstream of O'Brien's Creek and extending downstream to below the junction with Big Sandy Creek. Since the 1950s, at least 100 diamonds, one to 5 ct weight, have been found in hi^-level cassiterite-bearing deposits and in alluvial deposits accompanied by prolific topaz and less common aquamarine. Many of the diamonds were recovered during tin mining operations in what are considered to be Cretaceous alluvial deposits. No substantive evidence has been offered as to the validity of this age other than upstream these deposits are overlain by basalts of the McBride Province. Company e?q)loration recovered diamonds but classic diamond indicator minerals were not recognised. Gamet, chromite and ilmenite are present in the alluvials but the compositions of these minerals indicate a basaltic origin. There are several areas within eastem Queensland where diamonds have been found either as isolated occurrences or in small numbers. Brigooda Possibly the most important single occurrence is the maar at Gamet Gully, Brigooda. Diamond has been found associated with this volcanic structure and is the only place in Queensland, at present, where diamond can be linked directly with an alkali basalt and its pyroclastic derivatives. Volcanic emptions within the Brigooda area have been dated, ranging from 16 Ma to 0.46 Ma. The age of the maar emption from which the diamond was recovered is 0.46 - 4.9 Ma (K-Ar on anorthoclase feldspar). The absence of a stable ancient craton, the presence of a thin lithosphere and proposed high thermal gradients as well as the lack of associated classic indicator minerals for either kimberlite or lamproite indicates a transporting medium other than kimberlite and lamproite (Robertson & Robertson (2)). Lakeland Downs Lakeland Downs ,an area southwest of Cooktown, has produced a small number of diamonds from a number of scattered locations. Diamonds have been recovered firom alluvial deposits in the headwaters of the Laura River, Palmer River, Spear Creek (tributary of the Pabner River) and Normanby River. In the Normanby River, Palmer River and Spear Creek, diamond is associated with alluvial gold. In Spear Creek it is also accompanied by sapphire and zircon. In all instances, classical diamond indicator minerals are absent. Within the region there is one maar and one possible diatreme. The maar at Tom's Hollow and the diatreme at Bull Hollow carry a mineral assemblage (gamet and chromite) with chemical con5)ositions approaching those minerals found in kimberlite. However company exploration has failed to locate diamond in these structures. Several kilometres north of Lakeland Downs, leucitite occurs along with basaltic pyroclastics at Hosking's vent (Barron et al (9)). This rock type is of deep seated origin and is considered capable of hosting diamond. Atherton Tableland The Atherton Tableland and surrounds also present a dilemma in relation to diamond occurrences. Diamonds has been recovered from the Herberton "Deep Lead", the Russell River Terraces, the headwaters of the Mulgrave River and from the Beattrice River. Diamonds also have been recovered from drainage systems cutting the basaltic terrain on the Atherton Tableland. In all instances, classic diamond indicator minerals are absent. The diamond in the Herberton deep lead was associated with alluvial cassiterite. This deposit is partly covered with basalt approximately 7 Ma old. The diamonds from the Russell River Terraces and the headwaters of the Mulgrave River were recovered during gold mining operations early this century. The deposits are mostly covered by basalts of the Atherton Province and contain besides gold, minor 486

cassiterite, zircon and sapphire. The diamonds in the Beattrice River are associated with zircon, sapphire and platinum as well as rare ruby. Other areas where diamonds have been found include the headwaters of Murphy's Creek , Googa Googa Creek in the Blackbutt area, Cania Goldfield northwest of Monto and the Kangaroo Hills Tin field. Possible origins of diamond in eastern Queensland For the Stanthorpe, Brigooda and Anakie areas, there are now strong indications that the sources of the diamonds may not be kimberlite or lamproite. Stanthorpe diamonds bear strong resemblances to the Copeton diamonds and their relative close proximity to one another may in^cate the same or a similar source. Diamonds from the Brigooda (Robertson & Robertson (2)) and Anakie (Robertson & Sutherland (8)) areas appear to exhibit a relationship with the volcaniclastic rocks found in the two areas. In the Anakie area, the relationship between diamond, nephelinitic pyroclastic deposits and nephelinite appears to be more than just coincidence. The presence of diamond with sapphire and zircon in fluviatile sediments raises several questions, one of which is the source(s) of these minerals. Low uranium zircon and diamond are known to be associated with kimberlite, lamproite and carbonatite. These rocks are yet recorded from the sapphire fields. Basaltic breccia pipes occur in the Rubyvale area. Circular depressions have been recorded between Reward and the Drummond Range as well as within the Drummond Range. From available data (Robertson & Sutherland (8)), it would appear that diamond may have been brought to the earth's surface in this region between the Late Jurassic and earliest Tertiary during volcanic activity that was the precursor to the eruption of the Cainozoic Hoy Basalt. The Elizabeth Creek area, the McBride and Chudleigh Provinces, the Atherton Tableland region as well as the Lakeland Downs area are compUcated in as much as they are adjacent to assumed stable craton. Whether basaltic volcanism or kimberlite/lamproite eruptions were responsible for diamond emplacement is yet to be determined. There appears to be a link between the presence of diamond, nephelinitic pyroclastic deposits and nephelinite. The headwaters of Elizabeth Creek drain the basaltic terrain of the McBride Province, hence the presence of basaltic indicator minerals is readily explained. The origin of the diamond is not so simple. To the west and south lies the Proterozoic rocks of the Georgetown Inlier on which diamond from unspecified locations have been reputedly found. Diamond also has been found occasionally associated with sapphire and zircon in the McBride and Chudleigh (volcanic) Provinces to the southeast and south respectively. If they have a similar origin to those in the 'Deep Lead" at Herberton and in the Russell River Terraces, then they would be older than the majority of the basaltic outpourings in the McBride, Chudleigh and Atherton Provinces and could be Early Tertiary or older. Before uplift of the Great Divide in the Late Cretaceous the ancestral drainage of the Elizabeth Creek-Einasleigh River system may have extended ftarther to the southeast, having its headwaters about the position of the Cardwell and Seaview Ranges. If the source of the diamond was at the headwaters of this ancient system, then the source will now be on the eastem fall of the great divide and possibly buried beneath the basalts of the McBride Province. The lack of widespread distribution of diamond in the Lakeland Downs Region points to sources of restricted areal extent and that it is unlikely the diamonds have been generated from reworked sedimentary deposits (eg Mesozoic sandstones of the Laura Basin). The presence of h i ^ pressure minerals associated with several eruptive phases of deep seated origin may indicate an S-Diamond type origin for the diamonds. Within the Atherton region, most of the diamond occurrences appear to be older than the main outpourings of the Atherton Basalt. The presence of diamonds in gold - bearing alluvials interbedded with pyroclastic deposits and overlain by basalt in the headwaters of the Russell and Mulgrave Rivers suggest a possible relationship between the diamonds and the basaltic pyroclastics. Diamonds within the drainage systems developed on the Atherton Basalt even poses a more complex problem. Are they derived directly from a primary source or has volcanic activity sampled the diamond bearing alluvial deposits beneath the basalt?. A number of maars and one diatreme occur within the Atherton Province. Each is capable of transporting diamond. However classic diamond indicator minerals are absent in their ejecta products. The maar at Lake Eacham contain minerals that approach the composition of kimberlite indicator minerals but are still considered by diamond exploration companies to be of "no consequence". It is likely that the maars and the diatreme may have sampled the underlying diamond bearing alluvials. One theory put forward for the occurrence of the diamonds is that the diamonds were transported during the initial stages of eruption of the Atherton Province basalts and reworked by later eruptions. This has been based on the apparent association of diamond - pyroclastic deposits in the headwaters of the Mulgrave and Russell Rivers. The lack of microdiamond, ancient craton and thin crust below as well as a high geothermal gradients indicate that the diamonds may not have been transported by kimberlite or lamproite. There is no direct evidence to support transportation by alkali basaltic eruptions other than the apparent association mentioned above.

487

Exploration potential in Queensland In the northwestern part of the State (Mount Isa block), the potential for diamond discovery is considered to be good. The geological setting has similarities to the Kimberley Diamond Province of Westem Australia and potential diamond sources have been located on the Northem Territory side of the border. A much more determined approach to prospecting will have to be exercised if rewards are to be reaped. The presence of diamonds cannot be treated lightly and, in the case of the northwestem Queensland diamonds, the lack of suitable indicator minerals begs the question - what is the source?. Few exploration companies appear to have taken up this challenge. In eastem Queensland, the tried and tested kimberlite and lamproite exploration techniques have failed to provide sources for the alluvial diamonds. If meaningful results are to be achieved, broad tectonic applications such as Cliffords Rule (Janes (3)) will have to be ignored. Basaltic involvement and unusual subduction settings for generating diamonds and indicator minerals diflferingfromthose of kimberlite and lamproite should provide new prospectivity potential. Areas in eastem Queensland where a more radical approach to diamond prospecting is needed are: • Stanthorpe; • Brigooda; • Anakie sapphire fields; • Mount Surprise and McBride Province; • Atherton Tableland; and • Lakeland Downs Region. References 1. 2. 3.

5. 6. 7.

MacNevin A.A. Diamonds in New South Wales. Department of Mines Geological Survey of New South Wales Mineral Resources 42 1977. Robertson A.D. & Robertson C.M. The Brigooda diamond enigma. Queensland Government Mining Journal 95, 1994,pp.32-33. Janse A.J A. Reviews of supposedly non-kimberlitic and non-lamproitic diamond host rocks. In Meyer, H.O.A. and Leonardos, O.H. eds. Proceedings of the fifth International Kimberlite Conference, Araxa, Brazil, 1991, Volume!. Diamonds: Characterization, Genesis and Exploration, CPRM Special Publication IB, 1994, pp. 144159. Companhia de Pesquisa de Recursos Minerals, Sutherland F.L., Temby P., Raynor L.R. & Hollis J.D. A review of the East Australian diamond province. In Meyer, H.O.A. and Leonardos, O.H. eds. Proceedings of the fifth International Kimberlite Conference, Araxa, Brazil, 1991, Volume!. Diamonds: Characterization, Genesis and Exploration, CPRM Special Publication IB, 1994 pp. 170-184. Companhia de Pesquisa de Recursos Minerals, Brasilia. Meyer H.O., Milledge H.J. & Sutherland F.L. Unusual diamonds and unique inclusions from New South Wales, Australia. Sixth International Kimberlite Conference, Novosibirsk, Russia, 1995, Extended Abstracts 1995, pp. 379-381. Barron L.M., Lishmund S.R., Oakes G.M. & Barron B.L Subduction diamonds in New South Wales: implications for exploration in eastem Australia. Geological Survey of New South Wales, Quarterly Notes 94,1994, pp. 1-23. Barron L.M., Lishmund S.R., Oakes G.M., Barron B.J. & Sutherland F.L. Subduction model for the origin of some diamonds in the Phanerzoic of eastem New South Wales. Australian Journal of Earth Sciences 43, 1996, pp. 257-267. Brasilia. Robertson A.D.C. & Sutherland F.L. Possible origins and ages for sapphire and diamond from the central Queensland gemfields. In Sutherland 1.1 ed. R.O. Chalmers Commemorative Papers (Mineralogy Meteoritics, Geology), 1992, pp. 45-54. Records of the Australian Museum Supplement 15. Barron B.J., Robertson A.D. & Sutherland F.L. Olivine 'leucitites', their xenolith and megacryst suits, Hoskings Peaks, north Queensland. Australian Journal ofEarth Sciences 43,1996, pp. 231-244.

488

MATURATION OF QUEENSLAND MESOZOIC COALS : THERMAL AND MICROSCOPIC ANALYSIS OF PETROLEUM AND GAS GENERATION J.D. SaxbyS M. Glikson^ and L.S. Szabo^ 1 CSIRO Division of Coal and Energy Technology, North Ryde, NSW 2 Department of Earth Sciences, University of Queensland Summary Although there are many difficulties in simulating geochemical reactions, laboratory e?q)eriments can provide valuable insigjits. This is particularly true when quantitative chemical data is combined with microscopic observations of structures having dimensions of either micrometers or nanometers. Maturation of Callide and Tarong coals has been carried out by heating at 10°C/minute to maximum ten^eratures up to 800°C in order to simulate the generation of gas and oil during relatively brief periods of hydrothermal heating. Thermogravimetric analysis and thermomechanical analysis provide data on weight and volume changes that accompany the release of volatiles. Transmission electron microscopy of the residues shows features characteristic of the rank increase accompanying hydrocarbon e^q>ulsion. Gas that is generated, but not readily released, can be considered as coalseam gas and results indicate a significant role for bitumen in its generation and retention. Introduction Coals can undergo maturation to higher rank either through slow heating resulting from increasing burial in a subsiding basin or through relatively rapid heating to higher temperatures as a result of regional hydrothermal events. Permian and Triassic coals of the Bowen-Sydney Basin have been subjected in places to magmatic heating during the Mesozoic and Tertiary following the break-up of Gondwana. Simoneit (1) has described how generation of oil and gas can occur in hydrothermal systems in only brief geological time periods (years to hundreds of years). Continental hydrothermal systems of this type include failed or dormant rifts and regions around piercement volcanoes. Recent isotopic evidence from Collerson et al. (2) supports a significant role for hydrothermal heating at 300°C to greater than 400°C in the maturing of some Queensland coals. In the present study laboratory simulation experiments have been undertaken on Triassic Callide Basin and Tarong Basin coals with a view to understanding how rapid rank-increasing processes relate to petroleum generation and coal-seam methane accumulation. Experimental Bright and dull lithotypes of Callide mine and Meandu mine (Tarong) coals were cut or cored to provide samples (4.5 x 1.5 mm) suitable for heating in thermal analysis equipment. Samples were heated at 10°C/minute to maximum temperatures between 400 and 800°C under inert (oxygen-free) conditions. This heating rate is not appropriate for slow burial in a subsiding basin but may not be too unreasonable as a simulation for some hydrothermal events. Thermogravimetric analysis (TGA) enables changes in sample weight to be recorded under effectively open anhydrous conditions. Thermomechanical analysis (TMA) allows ch^ges in sample volume to be determined as contraction, swelUng or coking occur during loss of volatiles. By combining TGA and TMA data apparent density changes can be quantified. Gases and oils were collected during TGA for chromatographic analysis. Solid residues from samples heated to a range of maximum ten5)eratures were subjected to elemental microanalysis, light microscopy, vitrinite reflectance and transmission electron microscopy (TEM). Results and Discussion TGA. Typical weight loss curves for vitrite and inertite from Callide coal are shown in Figure 1. An initial drop due to loss of mainly water is followed by a second significant loss of generated hydrocarbons. A higher yield of volatiles from vitrinite (compared with inertinite) is always apparent. Quantitative data is given in Table 1. Under the conditions used, the oil window threshold temperature (Tt) and the temperature of maximum rate of hydrocarbon generation (Tim) are most accurately determined from the first derivative of the weight loss curve. Another usefiil parameter is the weight loss from degassing at Tim which is determined by observing the loss of

489

400

600

Temperature (T) Figure 1. TGA Curves Showing Weight Loss from Bright and Dull Callide Coal Samples Cores (-100 mg) heated at 10°C/minute. Table 1. TGA Data from Simulated Maturation of Vitrite from Callide and Tarong Coals Mtial vitrinite reflectance (%) Weight loss (300-550(C) (%) Weight loss (300-600(C) (%) Oil window threshold teri:5)erature (Tt) ((C) Temperature of maximum rate of oil generation (Tim) ((Q Weight loss from degassing at Tim (%)

Callide

Tarong

0.5 23 26 300 450 6

0.6 23 25 345 450 7

trapped volatiles when the sample is rapidly cooled from the temperature of maximum rate of hydrocarbon generation (Tbn). Significantly, these and other results show that the tendency to retain generated hydrocarbons within the maturing coal increases with increasing rank. TMA. Figure 2 shows typical curves for Callide samples. Derivative curves enable temperatures at which the volume of the coal core is changing at maximum rates to be determined accurately. This temperature is 150°C higher for inertinite than for vitrinite. Overall inertinite shows virtually no volume change up to 600°C and minimal decreases thereafter compared with vitrinite. These sub-bituminous Triassic coals differ greatly from higher rank bituminous coals which often contract at the hydrocarbon generation stage before rapidly expanding into a coke-type structure which then contracts again at higjier temperatures. Vitrinite Reflectance. Measurements carried out on residues from the heating e?q)eriments are shown in Figure 3. Also given are results from earher very slow heating experiments carried out by Saxby et al. (3) on brown coal and torbanite having initial reflectance values of 0.3 and 0.7% respectively. As expected, faster heating

490

200

400

Temperature CC)

800

600

Figure 2. TMA and First Derivative Curves of Vitrite and Inertite CbresfromCallide Coal. Curves show how sample contraction accon^j^es loss of volatiles and degassing at a heating rate of 10°C/minute.

% a> CQ

1 •mm

300

400

500

600

Maximum Temperature CO

700

Figure 3. Vitrinite Reflectance of ResiduesfromArtificially Matured Coals. Callide and Tarong: 10°C/minute; brown coal and torbanite l°C/week. results in higjier temperatures being needed to increase the degree of coalification of the residue after the generation and expulsion of hydrocarbons. Chromatographic analysis of the productsfromvitrinite shows that the oil/gas ratio tends to decrease as the temperature increases. Light Microscopy. The Callide vitrite residue at 500°C shows evidence of flow and mineral aUgnment along bedding planes. By 600°C coking is evident. Callide vitrite gives an ash yield of 14% and there are indications that fluidity/plasticity is retarded by the presence of minerals, particularly clays. Tarong vitrite develops large pores during coking at 550°C and, afterftirtherdevolatilisation, mosaic textures are apparent in residues at 700°C. TEM. Callide vitrite at 500°C gives evidence of oil generation and movement of tar (or bitumen) into cleats and inertinite cell cavities (Figure 4). At 600°C char formation has commenced and organisation into molecular orientation domains (Figure 5). At 700-800°C charring and formation of mosaic texture is complete, giving at h i ^ magnifications an ordering 491

that closely resembles naturally matured coals (Figure 6). By observing sections at different orientations and with varying thicknesses, a three dimensional model of a typical mosaic texture can be reconstructed (Figure 7). Tarong vitrite behaves similarly with evidence of cracking of bitumen into pyrobitumen within char cavities at >600°C. At the same time minerals are no longer flow aligned but are concentrated into micro-lenses within pyrobitumen.

Figure 4. TEM of Callide Vitrite Heated to 400(C, Mineral matter (M) in vitrinite Gas Outbursting. TGA experiments on bright coal containing abundant bitumen (from Calhde and Tarong as well as other mines) have revealed an unexpected phenomenon. A typical example is shown in Figure 8. A remarkable rapid loss in weight occurs at about 500-600°C and is accompanied by cooling of the sample. The increase in weight that occurs after the sudden drop is difficult to explain but may be due to adsorption of purge gases on freshly exposed coal surfaces. It is possible this mini-outbursting is caused by the pressure of trapped gas increasing to the point of explosive release even in a small TGA sample. It is clear that coal rank and bitumen content are key factors in rapid gas release of this type. This observation may have significant implications for the prediction of gas outbursts in underground mines.

100 nm Figure 5. TEM of Callide Vitrite Heated to 600°C. Mineral matter dispersed through the molecule domains of a mosaic texture. 492

Figure 6. TEM of Callide Vitrite Heated to 800°C. Under high magnification nano-sized mineral particles irregularly dispersed in the mosaic.

493

Figure 7. Reconstruction of a Typical Mosaic Texture Developed by Callide and Tarong Coals During Artificial Maturation to High Rank.

494

400

600

Temperature (T)

Figure 8. Typical Example of Unusual Gas Release Event in TGA of a Bitumen-rich Bright Coal Core. Heating rate 10°C/minute (with cooling at 400°C to enable release of generated hydrocarbons). Conclusions Simulated maturation e?q)eriments in thermal analysis equipment provide insights into how the generation and expulsion of oil and gas could have occurred from Queensland coals as a result of hydrothermal heating events. Callide and Tarong coals are typical of those that may have undergone, in parts, rapid Mesozoic (or Cainozoic) heating. Bitumen can be mobilised (or cracked to pyrobitumen) and appears to be important in leading to accumulations of gas under pressure in coal. TEM reveals mosaic structures which provide sites where nano-sized hydrocarbon inclusions may accumulate. Acknowledgements The authors gratefully acknowledge the support of ERDC and the Queensland Electricity Commission in this continuing research. References 1. Simoneit, B. R. T. (1993). Aqueous hi^-temperature and high-pressure organic geochemistry of hydrothermal vent systems. Geochim. Cosmochim. Acta, 57, 3231-3243. 2. Collerson, K. D., Zhu Jing, Glikson, M. and Golding, S. (1996). Gondwana rifting, magmatism and the generation of gas in Australian coals, submitted to Nature. 3. Saxby, J. D., Bennett, A. J. R., Corcoran, J. F., Lambert, D. E. and Riley, K. W. (1986). Petroleum generation : Simulation over six years of hydrocarbon formation from torbanite and brown coal in a subsiding basin. Org. Geochem., 9, 69-81.

495

COAL SEAM METHANE POTENTIAL OF THE WALLOON COALS

Steven Scott Geologist Department ofMines and Energy, Queensland Summary Within the Surat and Moreton Basins of south-eastern Queensland extensive resources of very high volatile, bituminous, low rank, non-coking, perhydrous coals occur in the Walloon Coal Measures. The coal resources, both measured and indicated, total 5 000 million tonnes. The coals within the Walloon SubGroupAValloon Coal Measures occur in thick banded intervals where individual coal bands are separated by lenticular beds of carbonaceous shale, mudstone, siltstone and sandstone. The coals occur within the appropriate depth range for coal seam methane production, but are at the lower end of the vitrinite reflectance range. While this would appear to be a disadvantage, Walloon coals have a high (between 70 and 90%) vitrinite content that should increase the potential of these coals for commercial gas production. Evidencefrompetroleum drilling indicates that the Walloon coals contain gas, but there has been limited measurement of the gas volumes involved. Gas analysesfromthe few seams that have been tested, show that the methane content of the coal seam gas is at least 50%, with little or no CO^. Introduction The Walloon Coal Measures are found in the Mesozoic Surat and Moreton Basins, formedfromintracratonic depressions in the Early Jurassic. Both basins occupy 222 OOOkm^ in south-eastem Queensland, Presently the Permian coal measures of the Bowen and Galilee Basins are being targeted for coal seam methane. Commercial production has recently begun from the Late Permian Baralaba Coal Measures in the south-eastem Bowen Basin at Moura. The Surat and Moreton Basins contain very large measured and inferred coal resources in the Walloon Coal Measures at shallow depths. Very little exploration for coal seam methane has been undertaken with only one stratigraphic test well having been drilled in the eastem Surat Basin. Walloon Coal Measures The Walloon Coal Measures extend throughout both the Surat and Moreton Basins (Figure 1). Coals within the measures are described as very high volatile, bituminous, low rank, non-coking and perhydrous (Coxhead (1)). The Walloon Coal Measures were deposited as coal swamps in an environment of waning energy. A number of depositional settings have been suggested; McLean-Hodgson & Kempton (2) concluded that the sequence was deposited in a high sinuosity fluviatile environment dominated by meandering streams; Clark & Cooper (3) interpreted it to be afine-grainedmeanderbelt river system. In the north-eastem Surat Basin Jones & Patrick (4) upgraded the Walloon Coal Measures to subgroup status and subdivided the measures into two coal bearing successions separated by a sandstone unit. This subdivision, into die upper Juandah Coal Measures, the Tangalooma Sandstone and the lower Taroom Coal Measures, has not been recognised elsewhere. The Walloon Coal Measures range in thickness in the Surat Basin between 200m and greater than 400m (Exon (5)). The unit reaches its greatest thickness in the north-east and thins to the south-west and south-east. In the westem part of the Surat Basin, the percentage of coal within the Walloon Coal Measures decreases and the unit grades into the more sandy and less coaly Birkhead Formation (Exon (5)). In the Moreton Basin, the coal measures attain a maximum thickness of 600m in the axis of the Logan River Syncline in northem New South Wales (Goscombe & Coxhead (6)). However, much of the unit has been partly or wholly eroded, especially along anticlinal crests and on the upthrown sides of faults.

496

NiV,

Surat Basin

Mx)i-eton Basin ^

•^Brisbane

Figure 1. Location Surat and Moreton Basins. SURAT BASIN WESTERN

CENIRAL

NORTH-EASTERN

MORETON BASIN

Juandah Coal Measures Birkhead Formation

Walloon Coal Measures

Walloon Subgroup

Tangalooma Sandstone

Walloon Coal Measures

Taroom Coal Measures Table 1. Walloon Coal Measures stratigraphy - Surat and Moreton Basins. Coal seams of the Walloon Coal Measures are lenticular, variable in thickness and limited in areal extent (Goscombe & Coxhead (6)). The main coal deposits are typically ovoid to crescent shaped and extend for up to 10km' in area. The seams thin progressively towards the margins of the deposits and then split and either lens out or grade laterally into carbonaceous shale and non-coal strata. Coal seams range in thickness from a few centimetres to more than 10m in the Brigalow area in the eastem Surat Basin. In the areas of maximum development, coal makes up about 4% of the coal measures (Swarbrick (7)). Coals of the Walloon Coal Measures typically contain between 70 to 90% vitrinite and 10 to 20% liptinite, 5 to 15% of which is suberinite (Goscombe & Coxhead (6)). These results arefromthe shallow, potential coal mining areas along the northem and north-eastem basin margins. There has been little or no petrography undertaken on the deeper coals, especially in the Moreton Basin where even vitrinite reflectance work has not been undertaken on seams intersected by petroleum exploration wells. The Walloon Coal Measures have the potential to be a very large contributor of thermal coal or coal for conversion to liquid fuel. Measured and indicated open cut resources in the Surat Basin are 2 650Mt (Coxhead (1)). Measured and indicated resources for both open cut and underground deposits for the Moreton Basin are 2 214Mt. Petroleum/Coal Seam Methane Exploration Over 800 petroleum wells drilled in the Surat Basin have intersected the Walloon Coal Measures. Only 15 petroleum wells have intersected the Walloon Coal Measures in the Moreton Basin. Coal seam methane e}q)loration in the Surat and Moreton Basins has been much more limited than in the Bowen and Galilee Basins, with only one coal seam methane well, ANU South-East Teatree 1, having been drilled. This well was drilled as a stratigraphic test of the Walloon Coal Measures in the eastem Surat Basin. The coal measures were cored (from near the top of the unit, 336.87m) to 508.95m where drilling was abandoned due to mechanical difficulties. Twenty coal seams were intersected, with an average thickness of 0.75m (Nguyen, Renison & Gumey (8)). Only 3 seams were thicker than Im with the thickest being 4.6m. Locally the coal showed little mineralisation in the cleats and most of the coal and carbonaceous mudstone exhibited bleeding gas when the core was opened on the surface. The coalfromthe seam intersected at 347m had a mean maximum vitrinite reflectance of 0.52%.

497

Eight coal samples were taken for determination of total gas content. The results ranged from 3.5 to 7.7g/cc (dried airfree basis). The maximum result was recorded at the thickest seam. The desorbed gas comprised between 85 and 93% methane on a calculated air-free basis. All samples except one were methane saturated (Nguyen, Renison & Gumey (8)). Numerous petroleum e?q)loration wells encountered high mud gas readings while drilling through the Walloon Coal Measures. ANU Miles 1, located 2km south of the township of Miles in the eastem Surat Basin, intersected the coal measures between 263.5 and 748.0m (Anulka NL (9)). Three significant coal seams between 2 and 4m thick were intersected in the upper half (Juandah Coal Measures). These seams can be correlated with the uppermost seams in ANU South-East Teatree 1. One seam in ANU Miles 1, at 360m, had gas peaks reaching over 1100 units (228 800ppm) during drilling breaks. All gas encountered consisted of methane without any accompanying heavier hydrocarbons. Coal seam methane indications have not only come from petroleum wells. During the drilling of a water bore, R.N. 17322, (Holt's water bore) in March 1967 gas blew out while the bore was drilling through the Injune Creek Group (Gray (10)). The Injune Creek Group was named by Exon (11) to describe the mudstone, siltstone, labile sandstone and coal of the (in ascending order) Birkhead and Westboume Formations in the eastem Eromanga and westem Surat Basins. Swarbrick (7) used the term throughout the Surat Basin to describe the (in ascending order) Eurombah Formation, Walloon Coal Measures, Springbok Sandstone and Westboume Formation. The blowout, in Holt's water bore, reached a maximum height of 10m and lasted for approximately 40 hours before dying out. The gas analyses is shown in Table 2. Volume Gas 54.9 moles% Methane 35.3 moles% Nitrogen 9.7 moles% Oxygen Carbon Dioxide 0.11 moles% Table 2. Gas analyses - Holt's water bore. The nitrogen and oxygen were probably contamination making the gas essentially methane. Gray (10) documents other occurrences of methane from the Injune Creek Group or its equivalent in wells at Bymount near Injune, Millmerran and at Mutdapilly near Rosewood in the Moreton Basin. In the Wandoan Town Bore, methane flowed continuously from October 1947 until May 1948. These gas occurrences have been detailed by the Geological Survey of Queensland (12). Many of the reported occurrences have analyses, with most having greater than 70% methane. Coal Seam Methane Potential The coal seam methane potential of the Walloon Coal Measures is difiScult to accurately predict because of the very small number of tests conducted within the Surat and Moreton Basins. This lack of data across the basins means that to gain an indication of the coal seam methane potential, other indicators instead of gas content, permeability and porosity must be used. The best indicators are vitrinite reflectance, vitrinite content and coal thickness. This data can be collected from conventional petroleum e?q)loration wells. Vitrinite reflectance determinations from 63 petrolexim e:?q)loration and stratigraphic wells were used to show the relationship between vitrinite reflectance and present day depth of the Walloon Coal Measures and their equivalents in the Surat Basin (Figure 2). There is a straigjit line relationdiip between the two, though this may have been skewed slightly by the number of shallow (and low) reflectance points. Ward (13) relates American Society for Testing and Materials (A.S.T.M.) coal rank classes to vitrinite reflectance limits (Table 3). By setting the parameters as 0.45 and 0.75% for vitrinite reflectance as the range for the Walloon Coal Measures then the coals would mostly be classified as high volatile bituminous C to high volatile bituminous B. Using these classifications and depths between 500 and 1 000m, then the estimated maximum producible methane content for die Walloon Coal Measure coals can be predicted from the chart produced by Eddy, Rightmire & Byrer (14) (Figure 3).

498

2,000 1,750 1,500 1,250 £ oa Q

1,000 750 500 250 0 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 Vitrinite Reflectance (Rv max%)

Figure 2. Correlation between reflectance and present day depth, Surat Basin. Rank

Maximum Reflectance (%)

sub-bituminous high-volatile bituminous C high-volatile bituminous B high-volatile bituminous A medium volatile bituminous low volatile bituminous semi-anthracite anthracite

<0.47 0.47-0.57 0.57-0.71 0.71-1.10 1.10-1.50 1.50-2.05 2.05-3.00 (approx.) >3.00 (approx.)

Table 3. Vitrinite reflectance limits and A.S.T.M. coal rank classes (After Ward (13)). The parameters were selected on the basis that between the depths of 500 and 1 000m the vitrinite reflectance for coals of the Walloon Coal Measures is mostly likely to range from 0.45 to 0.75%. These depths are the most economic for drilling and production costs. The reflectance values are mostly on the low side of the suggested optimum rank range for methane extraction. Even with these quahfications from Figure 3 it appears that the lost and desorbed gas content for the Walloon coals should range between 4 and 12mVtonne. These figures agree favourably with the results of a range between 3.5 and 7.7m3/tonne from ANU South-East Teatree 1. Two types of porosity, fracture and matrix, are present in coal. The most significant when considering methane retention potential is fracture porosity (Rightmire (15)). Mostfracturesin coals are cleats and cleats are more common in vitrinite. With such a high vitrinite content, (70 to 90%) the Walloon coals should have a high fracture porosity and thus a high methane deliverability when intersected.

100

200

300

400

500

600

700

800

900

1,000 1,100 1,200 1,300 1,400

Depth (m)

Figure 3. Estimated maximum producible methane, Walloon Coal Measures (Modified after Eddy, Rightmire & Byrer (14)).

499

Conclusions The potential for coal seam methane productionfromcoals of the Walloon Coal Measures appears to be very poor at first glance. The coals are relatively low in rank, (mostly between high volatile C bituminous to high volatile B bituminous) and the seams tend to be very lenticular in nature and mostly high in intra-seam strata. Finally there has been limited drilling for this resource in the Surat Basin and none in the Moreton Basin. These negative parameters can be offset by positives. The Walloon coals are very higt in vitrinite content inferring that the fracture porosity should be high, enabling good deliverability. Although the seams tend to be lenticular, they can be thick, with seams in the Brigalow area being up to 10.7m thick (Exon (16)). Another positive is that the Walloon Coal Measures are relatively shallow over large areas, making drilling costs to recover any methane low. Another positive is the numerous occurrences where significant gas readings have been recorded during drilling of the Walloon Coal Measures. It should be noted that the Surat and Moreton Basins are crossed by the Roma to Brisbane gas pipeline and both basins are also relatively close to the populous east coast and a ready market. Both the Surat and Moreton Basins contain very large coal resources. The coals also may contain very large methane resources that could be drained economically to produce a clean burning fiiel for electricity generation or as a raw commodity for the production of methanol. Finally the Walloon Coal Measures should not be totally disregarded as a potential coal seam methane producer even if there has been only one test conducted. All the parameters can be compared and contrasted but the only true test to evaluate the potential of the Walloon's is to drill wells, test the coal seams and carry out analyses on the gas recovered. REFERENCES 1. COXHEAD, B.A. Queensland Coals 10th Edition. Queensland Coal Board. 1995. 2. MCLEAN-HODGSON, J. & KEMPTON, H. The Oakey-Dalby region. Darling Downs Coalfield: Stratigraphy and depositional environments. Coal Geology, 1 (4). 1981.165-177. 3. CLARK, W.J. & COOPER, D.M. Sedimentological and wireUne log aspects of the Walloon Coal Measures in GSQ Dalby 1 and GSQ Chinchilla 3, Surat Basin, Queensland. Queensland Government Mining Journal 86, 1985. 386-394. 4. JONES, G.D. & PATRICK, R.B. Stratigraphy and coal exploration geology of the northeastem Surat Basin. Coal Geology, 1 (4). 1981. 153-163. 5. EXON, N.F. Geology of the Surat Basin in Queensland. Bulletin of the Bureau of Mineral Resources, Geology and Geophysics Australia, 166.1976. 6. GOSCOMBE, P.W. & COXHEAD, B A. Clarence-Moreton, Surat, Eromanga, Nambour and Mulgildie Basins. Geology of Australian Coal Basins. 1995.489-511. 7. SWARBRICK, C.F.J. Stratigraphy and economic potential of the Injiine Creek Group in the Surat Basin. Geological Survey of Queensland Report, 79.1973. 8. NGUYEN, D.L., RENISON, M. & GURNEY, E.H. Southeast Teatree CBM-1 Well completion report Vol I & H. Unpublished company report held by the Department of Mines and Energy. 1995. 9. ANULKA N.L. Well completion report ATP 374P Block B ANU Miles 1. Unpublished company report held by the Department of Mines and Energy as CR 24345. 1993. 10. GRAY, A.R.G. Natural gas occurrence in the Brigalow area, March, 1967. Queensland Government Mining Journal, 68, 1967. 394-398. 11. EXON, N.F. Revised Jurassic to Lower Cretaceous stratigraphy in the south-east Eromanga Basin. Queensland Government Mining Journal, 67, 1966. 232-238. 12. GEOLOGICAL SURVEY OF QLTEENSLAND. Occurrence of petroleum and natural gas in Queensland. Geological Survey of Queensland Publication 299.1960. 13. WARD, C.R. (Ed). Coal geology and coal technology. Blackwell Scientific Publications. 1984. 95-96. 14. EDDY, G.E., RIGHTMIRE, C.T. & BYRER, C. Relationship of methane content of coal, rank and depth. Proceedings of the SPE/DOE Unconventional gas recovery symposium. 1982. 117-122. 15. RIGHTMIRE, C.T. Coalbed methane resource. IN RIGHTMIRE, C.T, EDDY, G.E. & KIRR, J.N.(Eds) Coalbed methane resources of the United States. AAPG Studies in Geology Series #17. 1984. 1-13. 16. EXON, N.F. The stratigraphy of the Surat Basin, with special reference to coal deposits. Coal Geology, 1 (3). 1980. 57-69.

500

EXPLORATION FOR PRECIOUS OPAL WITHIN THE EROMANGA AND SURAT BASINS, QUEENSLAND AND NEW SOUTH WALES Brian R. Senior, Geological Consultant, Senior & Associates Pty Ltd, Canberra, ACT., Robert E. Besley, Managing Director, Redfire Resources KL, Sydney NSW, Summary Australia has produced over 90% of the world's precious opal for the past 100 years and is likely to remain as the world's leading producer for the foreseeable future. Annual production is approximately $111 million per annum. Cumulative production from each of the major fields amounts to $1 bilUon on current value. Opal mining is essentially a cottage industry but recent changes to mining legislation in S A, Qld and NSW have opened large tracts of potentially opal bearing country for systematic e?q)loration. The deposits lie within weathered rocks which have probably undergone two phases of deep weathering. Silicaladen groundwater wasfiltered,and formed silica gels within specific entrapment zones which accord with the structure and bedding geometry of the weathered profile. Exploration techniques which are currently being used in the search for new opal fields within the Eromanga and Surat Basins are discussed. Introduction Precious opal is an hydrated form of sihca which contains a variable proportion of water which is chemically bonded within the mineral structure (Si02.nH20). It is amorphous and consists of sub-microscopic sihca spheres packed in regular array. In the precious form the layers of silica spheres diffract white light and break it up into the colours of the spectrum. Darragh and other (1) showed that an infinite array of colours may be produced according to the diameter and arrangement of these spheres. The body colour or background may be milky white, pale to dark grey, brown, blue, black or colourless. Black opal is the variety which commands the highest prices as the dark background enhances the play of colours. The largest precious opal deposits are found around the southem margin of the Great Artesian Basin at Lightning Ridge and White Cliffs in New South Wales and, in or in proximity to, the west and southwest margin at Mintabie, Coober Pedy and Andamooka in South Australia. According to Oliver and Townsend (2) current production levels are around $111 million per year at the field buying prices. Li^tning Ridge and Coober Pedy have been the dominant fields for the past 50 years and are consistently being extended. Estimates of their past production, based on present day values, is of the order of $1 billion from each field. The Coocoran Field in the Lightning Ridge district, which was discovered in 1989 and has an area of about 1.8km^ has produced over $250 million. The old fields at White Chffs, NSW and Andamooka, SA. are now essentially mined out. Boulder opal, a type of opal that occurs in ironstone concretions, is produced from about 100 localities in southwest Queensland. Current production is in the order of $2 million per year. At today's prices about twelve of these deposits would have each produced over $10 milhon worth of precious opal. Exploration It can be seen from the extensive distribution of precious opal deposits (Figure 1) that it is likely that other large deposits remain to be found. Most, if not all, of naturally eroded surface opal-float has been evaluated by prospectors but vast areas, where the host weathered profile is poorly exposed or covered with surficial sediments, remain to be evaluated. Investment and adoption of modem exploration techniques are the major requirements necessary for discovery of one of these glittering prizes. However, until recently. State legislation prevented acquisition of Exploration Licences of a suitable size for systematic exploration for these deposits. In Queensland, since 1990, precious opal is treated in the same way as any other mineral commodity and exploration of large areas is pemfiitted. A similar situation exists in NSW, except that the entire opal province of about 4,500km^ comprising the 80km long Lightning Ridge Reserve is excluded. In S.A. legislation was passed early 1996 which, for the first time, permits E^q^loration Licences for opal in that State. Most existing fields have been preserved in all States for exploration by individual miners and prospectors.

501

Lightning Ridge

WLpMEniES

10 \

v^Jiiib

"—^^^ /

Rcfcrcncc Major opal fields •

Opa) deposits

Fig 1. Locality map.

Figure 1. Locality map The exploration targets for precious opal mostly lie within 25m of the ground surface, and are very variable in size. For example, Coober Pedy occupies about 6.351mf and the Coocoran Field (Lightning Ridge district) occupied about L8km^ in 1990 but is still being extended (Fig. 2). Aerial photographs at 1:20 000 scale, acquired by Hasmay-Redfire joint venture partners, show a sinuous, divergent and convergent path of crowded workings at Coocoran which is a pattem common to several of the known opal fields of the Lightning Ridge district. The distribution appears to reflect a primary stratigraphic control, and probably indicates that the deposits follow a palaeochannel within the estuarine Cretaceous Griman Creek Formation. Similar features are noted by Senior and others (3) in the fluvial and lacustrine sedimentary rocks of the Lower to Upper Cretaceous Winton Formation, and form a primary control over some of the ironstone-hosted (boulder) opal deposits. Drilling by Hasmay Pty Ltd within Prospect Licence areas located about 18km southwest of the Coocoran Field, revealed similar-sized and similarly orientated palaeochannels (Fig. 3). Fragments of non-precious and traces of precious opal were recovered from several drill holes within this feature. Whether this is a potential opalfield remains to be evaluated. The gentle topographic feature known as the Lightning Ridge Trend extends northwards through the State border into the Hebel-Dirranbandi area of south-central Queensland. Here, Redfire Resources N.L (4) have successfully recovered Lightning Ridge type black opal from drill holes and from trial mining of bulk samples. This was the first recorded discovery of black opal in Queensland. E}q)loration here is based on grid RAB drilling and careful samphng of potentially opaline horizons. Samples are processed and occurrences of both macro (>5mm) and micro (.05-5mm) opal are quantitatively assessed. Contour maps based on these data assist in identifying the distribution of potential deposits. Following positive results large diameter (1 metre) Calweld holes are drilled to facihtate bulk sampling and trial mining. Geophysical techniques have been used experimentally to locate stratigraphic or structural features in which deposits of precious opal may occur. Senior and others (3) used detailed ground magnetic surveys in an attempt to locate remanently magnetised, channel-like structures in westem Queensland. This technique led to the discovery of several small deposits, but perhaps is more useful in identifying the distribution of potential host rocks, and separating these from nonprospective areas. Watkins (5), in discussing the Lightning Ridge fields suggested SIROTEM, resistivity and seismicrefraction as possible useful tools. Ground impulse radar was also trialed by the authors across some westem Queensland opal occurrences but the rocks are apparently too resistive, and the results were inconclusive. Low level aerial photography at about 1:10 000 scale has proved useful in photogeological mapping of opal deposits and has established the presence of hitherto unknown extensions to previously worked mines. It is emphasised however, that the techniques discussed have only been trailed ejq)erimentally on small areas, funded privately with minuscule budgets.

502

1km -J

Fig. 2 Distribution of crowded workings as seen in aerial photographs acquired in 1991, of the Coocoran opal field. The sinuous divergent pattern of these workings indicate that these deposits follow a Cretaceous palaeochannel. Similar channel-like features are interpreted through clos^jjspaced RAB drilling in Hasmay's E.L. areas, located to the southeast of the Coocoran field.

Age of Opalisation Age determination of the weathered profile and thus an indirect age of opalisation contained therein was undertaken by Idnurm & Senior (6). Using palaeomagnetism to establish the age of the stable remanently magnetised components, they demonstrated that the western Queensland deposits occur within a profile that developed during two weathering events, the first in the late Cretaceous and Palaeocene (Momey profile) and the second in the late Oligocene-early Miocene (Canaway profile). The second event involved reweathering of the partly truncated older profile. Bird and Chivas (7) undertook oxygen isotope studies of sanqjles from the Momey and Canaway profiles and confirmed that two distinct weathering events had occurred. Natural Gamma-ray Logging During exploration of the Prospecting Licence areas in the Lightning Ridge district, Hasmay Pty Ltd adopted a program of natural gamma-ray logging of their RAB drill holes. The purpose was to provide better lithological control for correlation of rock units between drill holes. Reference sections were obtained by drilling and gamma-ray logging in proximity to currently producing mines. Relative high radioactive anomalies were found associated with the sandstone and underlying claystone contact (Fig. 3). These anomalies diminished in intensity upwards over a distance of about 5 metres. The sandstone coincident with this anomaly was invariably cemented by opaline silica with increasing induration towards a hard layer located just above the contact with the claystone. This hard layer is known as the 'steel band' by the local miners. It seems hkely that the downward movement of sihca-laden groundwater has been accompanied by transportation and accumulation of trace amounts of radioactive elements. If this is the case then gamma-ray logging promises to be a usefiil tool for indicating the probable presence of precious opal. Further research is required, and there is potential for development of a purpose-designed, spectral down-hole logging tool which may greatly assist exploration for this gemstone. Market Outlook If a commercial opal deposit is discovered by this new exploration effort, then it would be the first opal field to be controlled and mined by a single entity. This would result in a more consistent supply of precious opal based upon measured resources determined prior to mining. This in tum would lead to orderly marketing and provide opportunities for expansion established upon known resources and scheduled production rates. Efficiencies of scale could be brought to bear in mining, and comprehensive development of the resource would be possible under a single management/mine planning strategy. At the completion of mining, comprehensive environmental rehabilitation would be undertaken in accordance with practices currently adopted in all other sectors of the mining industry. The generous size of existing reserves would ensure that traditional opal mining would continue, with the added benefit to miners of stabilised prices and better developed markets. Conclusions Systematic geological data pertaining to precious opal have only slowly been acquired due to the former restrictive nature of State legislation. Since 1991, these impediments have progressively been removed, and with the recent availability of exploration hcences in South Australia, almost the entire Great Artesian Basin is available for exploration. If current techniques and investment are sustained, the opal industry is likely to follow a re-development phase akin to the diamond and sapphire industries.

504

X< QC LU § < QC O Z) o

LUJ IL

0DC (L

0Q QC O ^

UlI I I

0 cco

z ^ I

0 J

Steel band

LLI

^ P i I LL

—I< o

Fi^. 3 Dislrihulion of natural gamma-ray radioactivity recorded in an opal exploration drill hole. Note the increase in radioactivity towards the silicified base of the sandstone, which indicates that increasing silicification is accompanied by a progressive increase in radioactive elements.

The problems of exploring for this elusive gemstone within weathered regolith, often under shallow cover, are many and further experimentation and research are required. The most promising techniques are those which have the potential to delineate stratigraphic and structural features within the weathered sequence. Existing State and Commonwealth aerial photographs and satellite imagery can be used in a geomorphological evaluation of weathered profiles across large areas. Once target areas are selected, colour aerial photographs at about 1:20,000 scale could be photogeologically evaluated to identify local structural and stratigraphic targets, and to design a drilling program. 505

Quantitative assessment of macro and micro opal recovered from drill holes is a new exploration technique pioneered by Redfire Resources N.L. The relative high radioactive anomalies associated with Li^tning Ridge and perhaps other deposits, offer potential for further refinement in diagnosing the likely presence of precious opal in drillholes. Other geophysical techniques, particularly ground magnetic surveys, have met with limited success but have not been apphed widely enough for full evaluation. References 1.

DARRAGH, RJ., GASKIN, A.J., TERRELL, B.C., & SANDERS, J.V., 1966 - Origin of precious opal. Nature 209,13-16.

OLLIVER, LG., & TOWNSEND, LJ., 1993 - GEMSTONES IN AUSTRALIA. Australian Government Publishing Service. Canberra.

SENIOR, B.R, McCOLL, D.H., LONG, B.E., & WHITELEY, R.J., 1977 - Geology and magnetic characteristics of precious opal deposits, southwest Queensland, Bureau of Mineral Resources Journal ofAustralian Geology & Geophysics. 2, 241-251.

REDHRE RESOURCES, 1995 -Annual Report for Redfire Resources, N.L.

WATKINS, J.L, 1985 - Future prospects for opal mining on the Lightning Ridge region. Department of Mineral Resources, New South Wales geological Survey Report GS 1985/119.

IDNURM, M., & SENIOR, B.R., 1978 - Palaeomagnetic ages of weathered profiles in the Eromanga Basin, Qld. Palaeogeography, Palaeoclimatology, Palaeoecology. 24, 263-277.

BIRD, M.J., & CHTVAS, A.R, 1993 - Geomorphic and palaeoclimatic implications of an oxygen-isotope chronology for Austrahan deeply weathered profiles. Australian Journal of Earth Sciences. 40, 345-358.

506

GRANITE-HOSTED, DISSEMINATED GOLD MINERALISATION AT TIMBARRA, NEW SOUTH WALES. H.W. Simmons, PJ. PoUard, J.L Stewart*, LA. Taylor and R.G. Taylor National Key Centre in Economic Geology, Department ofEarth Sciences, James Cook University, Townsville, Qld 4811. *Homestake Gold ofAustralia Pty. Ltd, 226 Great Eastern Highway, Belmont, W.A. 6104. Summary A new class of gold deposit has been identified at Timbarra, New South Wales, in the southern New England Fold Belt. Disseminated granite-hosted gold mineralisation occurs within extensive zones of pervasively altered granite in the upper levels of highly-fractionated plutons, stocks and dykes of the Stanthorpe Adamellite suite. Pegmatites and rare narrow quartz veins indigenous to the granites also host gold. Gold mineralization at Timbarra appears to have evolved through magmatic-hydrothermal processes similar to those which account for disseminated granite-hosted Sn-W deposits. Introduction Gold was discovered at Timbarra in 1853, prompting a rush which peaked during the period 1858-1866. Initial mining focussed on rich, easily-worked alluvial and eluvial deposits. Early miners discovered that the source of the gold was disseminated mineraUzation within altered coarser grained granite and crosscutting microgranite dykes. These primary deposits were mined where weathering had rendered them soft enough for alluvial-style mining techniques to be effective. Incomplete records indicate that between 1858-1900 more than 100,000 oz. of gold may have been produced from the Timbarra goldfields (Wilkinson, 1980). Important deposits include Horton's, Poverty Point, RMT, Big Hill. During the 1980s, a renewed interest in Timbarra resulted from exploration for large, low-grade gold deposits. Examination and sampling of the old workings followed by drill testing indicated the presence of extensive zones of generally low-grade gold mineralization within granites on Timbarra Tableland. The combined resources of several disseminated gold deposits are currently estimated at greater than 400,000oz, most of which comes from the Poverty Combine/Big Hill area which has an indicated resouce of 13.54 million tonnes @ 0.81 g/t Au (Ross Mining NL Annual Report, 1995). Regional Geology Timbarra is located in the southem portion of the New England Fold Belt which extends from Newcastle in New South Wales to Townsville in Queensland. The southem New England Fold Belt consists of a gently folded Devonian-Permian fore-arc basin in the west and more strongly deformed subduction-related accretionary terranes in the east, separated by the Peel Fault system. Two extensive suites of granitic rocks, which together form the New England Batholith, have intruded the eastem accretionary terranes: (a) syn-orogenic Carboniferous-Permian S-type granites, and (b) post-orogenic PermianTriassic I-type granites (Shaw and Flood, 1981; Gilligan and Bames, 1990). The post-orogenic I-type granites and their associated volcanics form a chemically-related supersuite (Butler, 1974). The largest member of this supersuite is the Stanthorpe Adamellite which was emplaced at approximately 222 Ma (Shaw, 1964). The Stanthorpe Adamellite crops out in two parts - the (northem) Stanthorpe mass and the (southem) Timbarra mass. The Timbarra mass is situated east of Tenterfield on Timbarra Tableland, an isolated plateau approximately 16 km long and 6 km wide and averaging over 900 m in elevation that contains a sequence of felsic,fractionatedgranites, some of which host disseminated gold mineralisation. Stanthorpe Adamellite The Stanthorpe Adamellite is composed of a suite of magnetite-series, metaluminous to peraluminous granites. The metaluminous Stanthorpe mass is composed mostly of coarse-grained, equigranular granite which commonly contains phenocrysts of K-feldspar and plagioclase. Accessory minerals include biotite, homblende, zircon and titanite. Granites consisting of coarser crystals contained in afine-grainedmatrix of the same minerals crop out in patches throughout the Stanthorpe mass. Their heterogenous texture suggests an abrupt change in physical conditions occurred during crystaUisation, causing an increase in crystal nucleation. Within the Stanthorpe mass, the Ruby Creek Granite hosts Mo, W and Sn mineralisation. On Timbarra Tableland, porphyritic fine- to medium-grained homblende-biotite granite (Monty's granite) is intmded by felsic, coarse-grained biotite granite (Surface Hill granite). The contact between Monty's granite and the Surface Hill granite is characterised by minor brecciation of Monty's granite and by the presence of pegmatite, fine grained granite and microgranite facies in the roof zone and margin of the Surface Hill granite. Monty's granite and the Surface Hill granite have both been intruded by felsic microgranite dykes, some of which contain gold mineralisation. The Surface Hill granite is composed mainly of quartz, alkali feldspar, plagioclase and biotite with accessory zircon and apatite. Within mineralized areas the Surface Hill granite exhibits pervasive hydrothermal alteration. In hand specimen this appears as chloritization of biotite and greenish, sericitic alteration of feldspars, but this is difficult to distinguish from similar alteration in unmineralized granite. In thin section the alteration is observed to consist of sericite-albite 507

alteration of K-feldspar, sericite-albite-clay-carbonate alteration of plagioclase and sericite-carbonate-chlorite alteration of primary biotite. Towards the margins of mineralized zones the alteration appears to be less intense and is commonly more obviously localized by grain boundaries. Whole rock analysis of the granites indicates a continuous evolutionfromthe least evolved coarse grained granites in the Stanthorpe mass to the highly evolved microgranites of the Timbarra mass. However, in the hi^ly evolved granites of the Timbarra mass there has clearly been a redistribution of elements during subsolidus hydrothermal alteration. Si02 in the granites ranges from approximately 72-79 wt.% and is positively correlated with alumina saturation index (0.8-1.2). The high-Si02, peraluminous granites contain higher abundances of Rb and HREE, and lower abundances of Ti02, MgO, CaO, Ba, Sr and LREE compared to the less evolved phases. Horton's Prospect At Horton's prospect the Surface Hill granite crops out as an elongate stock within Monty's granite. Gold mineralisation occurs only within the Surface Hill granite and is localized within several distinctive facies in the roof zone of the granite (Simmons, 1993). The upper parts of the stock have a vertical sequence of crystallisation textures which are interpreted to relate to the volatile content and degree of undercooling of the melt. From the top downward, this sequence includes: 1) miarolitic granophyric microgranite, 2) microgranite grading tofinegrained granite, 3)finegrained granite grading to medium grained granite, and 4) medium grained granite. Within these facies the growth habit of quartz varies from granophyric intergrowths within the miarolitic textured and granophyric granites, to skeletal textures in the fine grained granite and anhedral texture within the coarse grained granite. This sequence of quartz textures is interpreted as a downward crystallisation sequence: initial crystallisation occurred adjacent to the upper contact, with the melt evolving from bemg undercooled (e.g. <50°C) and saturated in volatiles (miarolitic textured granite), to a higjier degree of undercooling (e.g. >100°C) and saturation in volatiles (granophyric granite), to severely undercooled and undersaturated in volatiles (fine grained granite) to a lower degree of undercooling (e.g. <50°C) but undersaturated in volatiles (medium grained granite). The miarolitic granophyric microgranite facies contains elevated Rb and Yb, and low La, compared to the underlying facies. Plots of these elements against depth have sharp inflexions at the interface between the miarolitic granophyric microgranite and the underlying microgranite-fine grained granite facies. This impUes that the material from which the miarolitic granophyric microgranite facies crystallized was enriched (depleted) in these components prior to crystallization. Gold mineralization (3-5 g/t Au over 10s metres vertically) occurs principally within thefineand medium grained granite facies beneath the miarolitic granophyric microgranite (Simmons, 1993). Gold Mineralisation Gold mineralisation at Timbarra occurs in the highlyfractionatedgranites and is associated with elevated abundances of Sb, Bi and Mo. Mineralized samples are commonly enriched in Si02 and Sb relative to unmineralized samples and have a higher alumina saturation index. Within the granites, gold (<lmm) occurs in primary miarolitic cavities, interstitial spaces between primary granite minerals, and in dissolution cavities within K-feldspar. Very fine gold particles (<0.05mm) occur on hydrothermal illite formed within miarolitic cavities. Pegmatite sheets in the upper portion of the Surface Hill granite have also been mined extensively for gold which occurs predominantly along microfracture arrays within pegmatite quartz. Rare veins within the granite (e.g. at Poverty Point) contain quartz, molybdenite, gold, pyrite, calcite, albite, fluorite and illite, and are interpreted to represent limited fracturing during development of mineralized zones within the granites. Disseminated gold in the granites contains 5-15 wt% Ag and has a close spatial association with bismuthinite and molybdenite. Within mineralised zones pure gold occurs as free grains in soil and weathered granite. Arsenopyrite and pyrite are commonly observed in gold-mineralised rocks as interstitial phases or as infill in miarolitic cavities and holes in altered feldspar, but do not share the same close association with Au as noted for bismuthinite and molybdenite. Primaryfluidinclusions in quartz that is texturally associated with gold (e.g. interstitial quartz or quartz infill in niiarolitic cavities) consist of C02-rich and low-salinity (5-7 % NaCl equivalent) H20-rich types. Discussion Gold deposits similar to those at Timbarra have not been described elsewhere in the literature and they represent a new style of large, low-grade gold target. The association of mineralization with miarolitic and granophyric facies varieties in the upper levels of fractionated granites is similar to the characteristics of disseminated tin-tungsten and rare-metal deposits (e.g. Pollard, 1989; Pollard et al., 1991a, b). The magnitude of metal enrichment during magmatic fractionation and hydrothermal alteration is similar to that observed in disseminated tin systems, i.e. ppm to lOOO's ppm for tin, ppb to lOOO's ppb for gold. Detailed fluid inclusion and stable isotope evidence from the disseminated tin deposit at Zaaiplaats, South Africa (Pollard et al., 1991a) indicates that mineraUzation formed during interaction between the granite and magmatic-hydrothermal fluids released during cooling of the pluton. Mineralization occurs at different levels within the granite and is related to textural variation including the development of granophyric and miarolitic textures (Pollard et al., 1991b). A similar model appears to be broadly appUcable to Timbarra-style gold deposits. Microgranite dykes and minor veins at Timbarra probably reflect minor fracturing during crystallization and fluid evolution. 508

The chemistry of the Stanthoipe Adamellite and the widespread association with W-Mo-Bi-Au mineralization in the New England Fold Belt has many parallels with the Carboniferous-Permian granites and associated W-Mo-Bi-Au mineralization in north Queensland (e.g. Bamford Hill, Kidston). Other provinces with many similar characteristics include the Telfer district (Westem Australia) and the Fairbanks district (Alaska). The differences in mineralization style probably reflect mainly differences in depths of emplacement of the magmas, initial volatile contents (e.g. Bumham and Ohmoto, 1980), and structural/lithological features of the host rocks. This suggests an opportunity for further discoveries of Timbarra-style gold deposits, particularly in the deeper intmsions within these and similar metallogenic provinces. Acknowledgements: We thank the management of Ross Mining N.L for their permission to pubhsh this paper. The provision of research funding and field support from Ross Mining N.L, Homestake Gold AustraUa Pty Ltd and James Cook University is also greatly appreciated. References Bumham, C.W. and Ohmoto, H., 1980. Late-stage processes of felsic magmatism. Mining Geology Special Issue no. 8, p. 1-12. Butler, H.R., 1974. The geology of the Stanthorpe - Leybum mineral district, Queensland and New South Wales. Unpublished BSc (Hons) thesis. University of New England, Armidale, N.S.W., Australia. GiUigan, L.B. and Bames, R.G., 1990. New England Fold Belt, New South Wales - regional geology and mineralisation, in Hughes, EE. (ed.) Geology of the Mineral Deposits of Australia and Papua New Guinea, Australasian Institute of Mining and Metallurgy, Melboume p. 1417-1423. Pollard, RJ., 1989. Geologic characteristics and genetic problems associated with the development of granite-hosted deposits of tantalum and niobium. In Moller, R, Cemy, R and Saupe, F. (eds.): Lanthanides, tantalum and niobium (Berlin, Springer-Verlag), 237-253. Pollard P.J., Andrew, A.S. and Taylor, R.G., 1991a. Huid inclusion and stable isotope evidence for interaction between granites and magmatic-hydrothermal fluids during formation of disseminated and pipe-style mineralization at the Zaaiplaats tin mine. Economic Geology 86,121-141. Pollard, RJ., Taylor, R.G., Taylor, R.P. and Groves, D.I., 1991b. Petrographic and geochemical evolution of pervasively altered Bushveld granites at the Zaaiplaats tin mine. Economic Geology 86,1401-1433. Ross Mining N.L. Annual Report, 1995. Shaw, S.E., 1964. The petrology of portion of the New England Batholith, New South Wales. Unpublished PhD Thesis, University of New England. Shaw, S.E. and Flood, R.H., 1981. The New England Batholith, geochemical variations in time and space. Journal of Geophysical Research 86B, p.10530-10544. Simmons, H.W., 1993. Textural and geochemical variations within Horton's Granite, Timbarra, NSW, with respect to disseminated gold mineralisation. Unpublished BSc (Hons) Thesis, James Cook University of North Queensland. Taylor, lA., 1992. Petrology and geochemistry of the Stanthorpe Adamellite. Unpublished BSc (Hons) Thesis, James Cook University of North Queensland. Wilkinson, L, 1980. Forgotten Country - the story of the Upper Clarence gold fields, Northem Rivers College of Advanced Education, Lismore, 292pp.

509

PERMIAN/TRIASSIC SEDIMENTARY SEQUENCES IN NORTHWESTERN SOUTH ISLAND, NEW ZEALAND, AND CORRELATION WITH AUSTRALIAN GONDWANALAND SEQUENCES. David SmaleS Hamish CampbelP, Rodney Grapes^ Nick Mortimer^, Ian Raine^ 1. Institute of Geological and Nuclear Sciences, P.O. Box 30368, Lower Hutt, NZ 2. Research School of Earth Sciences, Victoria University of Wellington, P.O. Box 600, Wellington, NZ 3. Institute of Geological and Nuclear Sciences, Private Bag 1930, Dunedin, NZ Summary New Zealand's Western Province, divided from the Eastern Province by the Median Tectonic Zone, contains two small (2-3km^) late Paleozoic and early Mesozoic sedimentary sequences that are quite distinct from the overwhelmingly predominant rocks of similar age in the Eastem Province. The sequence near Parapara Peak (northwest Nelson) consists of black slates, quartzofeldspathic sandstones and mudstones, and quartzose sandstones. Despite almandine-grade metamorphism, an early Late Permian (Ufimian to early Kazanian) fauna near the middle of the sequence has long-recognised affinities with Tasmanian and eastem Australian faunas. The Topfer volcaniclastic sequence near Reeflon contains a Middle or Late Triassic palynoflora resembling others included in the southem high-latitude "Ipswich Microflora", and is intruded by distinctive Jurassic dolerite of Ferrar afiSnity. Together with petrographic and geochemical characteristics, the association of these features provides strong evidence of Westem Province correlation with Gondwanaland in the Late Permian to Middle Triassic interval. Introduction New Zealand is divided into a Westem Province and an Eastem Province by the Median Tectonic Zone (Fig.l). Late Paleozoic and Triassic sediments are very abundant in the Eastem Province, but coeval sediments in the Westem Province are of very limited extent, and of quite different character. Resemblances of some Westem Province Permian/Triassic sequences to others in Tasmania and Eastem Australia have been noted (1,2,3), and lead appropriately to further comparisons (presented here in summary) of the petrography, geochemistry and fauna and flora of the Pennian Parapara Group in Northwest Nelson (4), and the Triassic Topfer Formation 150km to the SSW near Reefton (5). The Topfer Formation is intmded by the Kirwans Dolerite, a rock distinct from any other basic igneous rocks in New Zealand, but showing close aflSnities with flood basalts of the Jurassic Ferrar magmatic province (6). This work is part of a 5-year study of the Median Tectonic Zone and related rocks by geologists of the Institute of Geological and Nuclear Sciences and co-workers. Parapara Group The Parapara Group consists of Flowers and Walker Formations, and covers Skm^ (4), but a further 200+m of underlying carbonaceous siliceous pelitic schist with sparse pebbly layers (Bay Schist) is now thought to be part of the sequence. The whole sequence shows metamorphic foliation at a low angle to bedding. The thin (20m) Pupu Conglomerate, at the base of the Flowers Formation and overlying Bay Schist, contains dominantly quartzose mudstone and sandstone pebbles, with minor feldsarenite, volcanic, and trace amounts of granitic pebbles. All have been tectonically elongated during metamorphism. The bulk of the Flowers Fonnation consists of metamorphosed poorly sorted feldsarenites and lithic feldsarenites, some of which are fossiliferous, interbedded with variably carbonaceous mudstones. Metamorphism has recrystallised the finer (<0.1mm) parts of the rocks to a homfelsic mosaic of quartz and albite-oligoclase with varying amounts of conspicuously aligned sericite. Porphyroblasts of gamet, biotite, chlorite, and ilmenite occur in all lithologies, but varying concentrations, particularly of gamet and chloritoid, probably reflect sedimentary compositional layering. Recognisably detrital grains, mostly quartz and albite, form only 10-20% of the rock, and composition of the sandstones on the QFR diagram (Fig. 2) are no more than indicative. Coarse angular dropstones and poorly sorted sub-angular debris flow sediments related to suspected glacigenic processes occur widely in the Flowers Formation and Pupu Conglomerate. Black slate beds similar to the Bay Schist become more abundant towards the top of the formation, and ultimately form the Pariwhakaoho Slate member, a 30m layer at the top of the Flowers Formation. The Bay Schist and the Flowers Formation have been intruded by thin (l-2m) porphyry and granitoid sheets (7) most of which are likely to be of Early Cretaceous age. The Walker Formation is at least 200m thick, and is predominantly more quartzose than the underlying Flowers Formation. It consists of metamorphosed, thick-bedded, frequently cross-bedded feldsarenites and subfeldsarenites with scattered lenses rich in quartz pebbles. Metamorphic fohation continues through the Walker Formation, but gamet and rare chloritoid are present only at the base, and the grade of metamorphism appears to decrease upwards. Petrographically the Walker Formation is highly distinctive, and is characterised by a matrix containing little but sericite.

510

Geochemistry A plot of K20/Na20 vs Si02 (Fig.Sa) allows a distinction between Walker Formation and Flowers Formation sediments. Walker Formation (and Pariwhakaoho Slate) plot as Passive Margin, but the Flowers Formation has a lower K20/Na20 ratio and extends into Active Continental and Oceanic Island Arc Margin fields (8). Flowers Formation rocks have low K2O contents (<3%), reflecting an absence of potash feldspar, while low Na20 content (<0.5%) of Pariwhakaoho Slate reflects a dominance of sericite and silica. Ti/Zr vs La/Sc plots (Fig.3b) show most samples in a Continental Island Arc setting, with some (particularly Walker Formation) extending into a Passive Margin (9). Fauna Parapara Group fossils, first described by Waterhouse and Vella (10), are more or less restricted to a deformed siltstone-shellbed succession within the lower 35m of Flowers Formation. The fauna is a richly fossiliferous and moderately diverse bryozoan-brachiopod-moUuscan association. It represents a mid-shelf, cool water fauna. Our work, based on new collections, has identified a revised list of about 40 taxa. The common brachiopods include Terrakea brachythaera, Glendonia duodecimcostata, Tomiopsis ingelarensis and T. mantuanensis. These taxa, along with other elements of the fauna, enable correlation with the Terrakea brachythaera Zone of other Austrazean sequences and indicate a Ufimian to early Kazanian age range (11). Strophalosiid brachiopods, which offer greater biostratigraphic resolution than any of the above-mentioned brachiopods, are rare. Waterhouse and Vella (10) record Wyndhamia clarkei, but our investigations have not confirmed this identification and hence there is reason to doubt the narrow age assessment discussed by Waterhouse (2) and Waterhouse and Sivell (12). Further collecting is necessary in order to better estabUsh species of some key brachiopod taxa. Topfer Formation The Topfer is poorly e?q)osed in a 2km^ area near Reefton. It is dominated by near-shore or fluvial brownish grey to pale grey medium-grained, moderately indurated massive sandstones with associated plantfiragmentsand coal. It is divided into a westem Waitahu petrofacies dominated by litharenites, and an eastem Boatmans petrofacies dominated by feldspathic litharenites (5). The proportion of dacitic and rhyolitic fragments is very high (60-80%). Traces of granitic debris (feldspar and quartz with granitic texture, myrmekitic intergrowths, and microchne grains) are present, and rare sedimentary clasts resemble older Westem Province rocks of the Greenland (Ordovician) or Reefton (Devonian) Groups. Parts of the Boatmians petrofacies have conspicuous grain coatings of iron oxide and laumontite cement. The Wait^u petrofacies contains widely varying amounts of calcite cement, in some areas forming more than 50% of the rock. Geochemistry Analyses of 11 samples (7) show 64-73% silica, reflecting the dominantly dacitic and rhyolitic composition of the detritus, as confirmed by analyses of several individual clasts. Proportions of K2O tend to be low in relation to Na20 (Fig.3a). Other trace element contents are consistent with sandstones derived firom continental (i.e. petrologically evolved) magmatic arcs (5). Flora The only identifiable fossils so far recovered firom the Topfer Formation are palynomorphs, but the stratigraphic palynology of the Topfer Formation has not yet been fiilly investigated; poor exposure has Umited collection of a comprehensive sample sequence. At present, distinct palynofloral assemblages of different ages within the Triassic are knownfiromoutcrop at two localities. The fossil assemblages are generally well-preserved and entirely nonmarine; they are dominated by pteridosperm pollen, especially Alisporites spp. A palynoflorafiromthe Boatmans petrofaciesfirombelow the main sill-like intrusion of Kirwans Dolerite, contains, inter alia, Aratrisporites spp., Limbosporites denmeadii, Annulispora folliculosa, Rogalskaisporites cicatricosus and Polycingulatisporites crenulatus, suggesting correlation with the lower part of the Polycingulatisporites crenulatus Zone of De Jersey & Raine (13). This is approximately equivalent to the Otamitan Stage of the New Zealand marine sequence, i.e. mid-Norian (Late Triassic). The assemblage is similar to that of the upper part of the Beacon Supergroup of southem Victoria Land (Subzone D of Kyle (14)), and upper Unit 4, the uppermost part of the Upper Parmeener Supergroup of Tasmania (15). Palynoflorasfromthe Waitahu petrofacies contain Aratrisporites spp., but lack Limbosporites denmeadii, Annulispora or Polycingulatisporites and also differ in containing elements such as Lundbladispora cf. willmottii and common striate bisaccate pollen (Lunatisporites, Protohaploxypinus). Further work is required to correlate this assemblage confidently, but it is broadly similar to Early to Middle Triassic palynofloras widely reported from Eastem Australia and Antarctica; in the Tasmanian sequence it appears to be most similar to that of upper Unit 2 or lower Unit 3 of the Upper Parmeener Supergroup (16). These results indicate that the Waitahu petrofacies is thus stratigraphically below the Boatmans petrofacies. Palynological analysis of float material suggests that the sequence is more or less continuous between these horizons.

511

Kirwans Dolerite The Ikm^ Kirwans Dolerite is a distinctive intrusion that has no known petrological correlative in New Zealand. Mineralogy consists of augite+ plagioclase+ pigeonite+ titanomagnetite+ Fe augite/hedenbergite± enstatite. The dolerite intrudes the Topfer Formation as two steeply-dipping sills, approximately 150 and 300m thick, joined by a thick dike or plug at their northem end. On the basis of similar ages, major and trace element concentrations, and Sr, Nd, and Pb isotopic ratios, Mortimer et al. (6) have correlated the Kirwans dolerite with Jurassic low-Ti tholeiites of the well-known Ferrar magmatic province of Gondwanaland. Field relations and preliminary paleomagnetic data also support this correlation. Although the Kirwans Dolerite is the first, and so far only, Ferrar correlative to be reportedfiromNew Zealand, it considerably increases the known shape and areal extent of the Ferrar province to close to the inferred paleo-Gondwanaland margin. This margin is generally thought to have been a convergent margin throughout much of the Phanerozoic, so the presence of the intraplate Kirwans Dolerite in New Zealand raises the possibility that the south Gondwanaland margin either (1) underwent drastic post-Ferrar tectonic truncation to remove Middle Jurassic subduction-related magmatic arcs or (2) was not a subduction zone but perhaps temporarily a passive or strike-slip margin in the Middle Jurassic. Relation to New Zealand Eastern Province and Australian sequences Several lithological elements of the Parapara sequence are comparable with parts of Eastem Australia (16), but the sequence of carbonaceous mudstone, fossiliferous sandstone overlain by quartzose sandstones, is more Uke that known from Parmeener Supergroup sequence in Tasmania (17). At the generic level the Parapara fauna is close to correlative faunas in Southland (Productus Creek Group (18,19)), Tasmania (Lymingtonian faunas of the Parmeener Supergroup (20)), and Queensland (Bowen Basin (21,22)). However, there are some differences in presence or absence of taxa. Perhaps most notable is the apparent absence in the Parapara fauna of atomodesmatinid shell, whose presence is a hall-mark of Productus Creek Group faunas. The Triassic Upper Parmeener sediments, with abundant acid volcanic lithic detritus, common iron oxide grain coatings and laumontite cement, lithologically and geochemically parallel Topfer sediments. The palynofloras so far recovered from the Topfer Formation are clearly comparable with those of the Triassic southem high latitude Ipswich Microfloral Province (13,23), also known from Victoria Land, Tasmania, and Eastem Austraha. They share this biogeographic province with Triassic floras of the Murihiku Supergroup of New Zealand and New Caledonia. While these have a majority of species in common, in the present state of knowledge it is not clear whether slight differences are due to local ecologic factors, or regional floral variation. Overall, the stratigraphy and regional association of Parapara Group, Topfer Formation, and Kirwans Dolerite with its Ferrar affiliation, are similar to Eastem Australia and particularly Tasmania (Fig.4). Individually the similarities may be coincidental; in combination the comparison is compelling. References 1. 2. 3. 4. 5. 6. 7.

8. 9. 10. 11.

Clark, R.H., Vella, P., Waterhouse, J.B. The Permian at Parapara Peak, north-west Nelson. New Zealand Journal of Geology and Geophysics 10: 232-246,1965. Waterhouse, J.B. Permian Pectinacea and Limacea (Bivalvia)fromNew Zealand. New Zealand Geological Survey Paleontological Bulletin 49,1982. Mortimer, N., & Campbell, H. J. Devonian to Jurassic rocks of New Zealand: classification content and Gondwana context. In Proceedings of the Ninth International Gondwana Symposium. Geological Survey of India Special Publication in press 1996. Grindley, G.W. Sheet S8 Takaka (1st ed.) "Geological Map of New Zealand 1:63,360. Department of Scientific and Industrial Research, Wellington, 1971. Mortimer, N., & Smale, D. Petrology of the Topfer Formation: first Triassic Gondwana sequence from New Zealand. Australian Joumal of Earth Sciences 43(4) in press 1996.. Mortimer, N., Parkinson, D.L., Raine, J.I., Adams, C.J., Graham, I.J., Oliver, P.J., & Palmer, EL Ferrar magmatic province rocks discovered in New Zealand: imphcations for Mesozoic Gondwana Geology. Geology 23:185-188, 1995. Grapes, R., Smale, D., Gibson, G., Landis, C., Kawachi, Y., Campbell, H., Mortimer, N., Pakner, K., and Ren, D. X-ray fluorescence and electron microprobe analyses of Westem Province Permian and Triassic sediments and associated intrusives, northwest South Island of New Zealand and Tasmania. Victoria University of Wellington Analytical Facility PubUcation 18,49p. 1996. Roser, B.P., & Korsch, R.J. Determination of tectonic setting of sandstone-mudstone suites using Si02 content and K20/Na20 ratio. Joumal of Geology 94(5): 635-650,1986. Bhatia, M.R., & Crook, ELA.W. Trace element characteristics of greywackes and tectonic setting discrimination of sedimentary basins. Contributions to Mineralogy and Petrology 92:181-193, 1986. Waterhouse J.B; Vella P. A Permian fauna from north-west Nelson, New Zealand. Transactions of the Royal Society of New Zealand 3: 57-84,1965. Archbold, N.W., Dickins, J.M. Australian Phanerozoic Timescales: 6. A standard for the Permian System in Australia. Bureau of Mineral Resources, Australia, Record 1989/36, 17p. 1991. 512

12.

13. 14. 15. 16.

17. 18.

19. 20. 21. 22.

23. 24. 25.

Waterhouse J.B.; Sivell W.J. New England Orogen and the Tasman Sea. In J.D. Kleeman (ed.), New England Qrogen. Tectonics and Metallogenesis. Department of Geology and Geophysics, University of New England, Armidale, p.199-204,1988. De Jersey, N.J. & Raine, J.L Triassic and earliest Jurassic miosporesfromthe Murihiku Supergroup, New Zealand. New Zealand Geological Survey paleontological bulletin 62.164p. 1990. Kyle, RA. Palynostratigraphy of the Victoria Group of South Victoria Land, Antarctica. New Zealand joumal of geology and geophysics 20:1081-1102,1977. Forsyth, S.M. Upper Parmeener Supergroup. In Burrett, C.F. & Martin, E.L. (Eds), Geology and mineral resources of Tasmania. Geological Society of Australia special publication 15, pp309-333,1989. Veevers, J.J., Conaghan, RJ., and Powell, C McA. Eastern Australia. In Veevers, J.J. & Powell, C.McA. (eds.) Permiari'Triassic Pangean Basins and Foldbelts along the Panthalassan Margin of Gondwanaland, pp.11-171. Geological Society of America Memoir 184,1994. Burrett, C.F., Martin, E.L. "Geology and Mineral Resources of Tasmania". Geological Society ofAustralia Special Publication 15, 574p. 1989. Waterhouse, J.B. Permian Stratigraphy and Faunas of New Zealand. New Zealand Geological Survey Bulletin 72, 1964. Waterhouse, J.B. Aspects of the Permian Tethys: its definition, interface with Gondwana, original disposition, and subsequent deformation. In K.G. McKenzie (ed.) Shallow Tethys 2. Proceedings of the International Symposium on Shallow Tethys 2. A.A. Balkema, Rotterdam, p.l31-148,1987. Clarke, M.J. Late Permian (Late Lymingtonian=?Kazanian) brachiopodsfromTasmania. Alcheringa 11:261-289, 1987. Waterhouse, J.B. Late Paleozoic Brachiopoda (Athyrida, Spiriferida, and Terebratulida) from the southeast Bowen Basin, east Australia. Palaeontographica A 196: 1-56,1987. Waterhouse J.B.; Briggs D.J.C. Late Palaeozoic Scyphozoa and Brachiopoda (Inarticulata, Strophomenida, Productida and Rhynchonellida) from the southeast Bowen Basin, Australia. Palaeontographica A 193: 1-76, 1986. Dolby, J.H. & Bahne, B.E. Triassic palynology of the Camarvon Basin, Westem Australia. Review of paleobotany and palynology 22: 105-168,1976. MacKinnon, T.C. Origin of the Torlesse terrane and coeval rocks. South Island, New Zealand. Geological Society of America Bulletin 94: 967-985,1983. Dickinson, W.R., Beard, S.L., Brakenridge, G.R., Eqavec, J.L., Ferguson, R.C., Inman, K.F., Knepp, R.A., Lindberg, F.A., & Ryberg, P.T. Provenance of North American sandstones in relation to tectonic setting. Geological Society of America Bulletin 94: 222-235, 1983.

Figure 1. Reconstruction of the southern part of Gondwanaland (dot pattem) during the Jurassic, showing location of Permian and Triassic sequences. SB= Sydney Basin; NC= New Caledonia; TAS= Tasmania; MBL= Marie Byrd Land. Map of South Island shows Parapara Peak and Reeflon in the Westem Province (w), the Median Tectonic Zone (MTZ) and Alpine Fault, and Eastem Province (e) terranes. (After Mortimer & Smale (5)).

513

Figure 2. Ternary QFR plot showing Parapara sandstones (upright triangles — Walker Formation; inverted triangles — Rowers Formation) and field of Topfer sandstones (dotted hexagon) in relation to selected New Zealand and Gondwanaland suites. Arrows point from mean composition of older to younger suites: TAS: Tasmania (Late Permian to Middle Triassic); EA: East Australia (Early Permian to Late Permian); Torlesse (Permian to Late Triassic MacKinnon (24)). Interpreted fields after Dickinson et al. (25); c.b.= continental block; t.arc = transitional arc; d.arc = dissected arc; u.arc = undissected arc; r.o.= recycled orogen. Adapted from Mortimer & Smale (5).

514

(a) 100.0

10.0

1.0

0.1

0.0 60.0

50.0

70.0

80.0

90.0

Si02 (wt%)

(b) eo 70

Parapara: A Walker Formation V Flowers Formation 0 Bay Schist

Ii; V ji 11

Parmeenen X Triassic lithic sandstone + Triassic quartzose sandstone ® Permian sandstone H Permian mudstone

Continental arc-derived sandstones (Bhatia& Crook 1986)

W*. I

0.0

1.0

2.0

3.0

4.0

5.0

6.0

La/Sc

Figure 3. Geochemistry of Parapara and Parmeener samples and fields of samplesfromother New Zealand suites plotted on (a) K20/Na20 vs Si02 diagram (b) La/Sc vs Ti/Zr diagram. Fields for Maitai, Caples, Murihiku and Torlesse sandstones from Roser (pers. comm) and Roser & Korsch (8), and for Topfer Formation from Mortimer & Smale (5). Dotted field = continental arc-derived sandstones of Bhatia & Crook (9) PM = passive margin; ACM = active continental margin; ARC = oceanic island arc margin. Samples are dominantly sandstones, except for Bay Schist samples (which are pelitic), the Permian Parmeener mudstone, and four other mudstones distinguished by "m" above the symbol.

515

EASTERN AUSTRALIA

REEFTON/ PARAPARA AREA

TASMANIA

, Kirwans Dolerite

Ferrar Supergroup

U. Parmeener [vVvVv'X^^^^^^ Supergroup

Topfer Formation

L Parmeener Supergroup

Parapara sequence

CL Fremouw g Formation u.

I Tasmanian dolerites Tr "vViisisi'siJ

EAST ANTARCTICA

.'^.'f.'^.'f.':

A <D

•J 3 CO

o 0 D

low Ti continental tholeiites

Reefton Group

f volcaniclastic petrofacies dominant

quartzose petrofacies dominant

cartx)naceous mudstone

1 Taylor Group

Pleurothyrella brachiopod fauna

Figure 4. Devonian to Jurassic lithostratigraphy of the Reefton/Parapara area of New Zealand compared with Gondwanaland cover sequences in Tasmania, Eastem Australia and Antarctica (after Mortimer & Smale (5)).

516

TECTONIC EVOLUTION OF THE EASTERN GONDWANALAND MARGIN DURING THE LATE PALAEOZOIC AND EARLY MESOZOIC

C J . StephensS RJ. HolcombeS C,R. FieldingS DA. Gusf 1 Central Norseman Gold Corporation Ltd. PO Box 56, Norseman WA 6443 2 Department ofEarth Sciences, The University of Queensland, Qld 4072 3 School of Geology, Queensland University of Technology, GPO Box 2434, Brisbane Qld 4001 The eastern Australian continental margin has been much modified from the Gondwanaland margin by middle to late Mesozoic and Tertiary tectonics. These tectonics were, perhaps arguably, dominated by extensional processes leading to the development of the modem passive continental margin (see Bryan et al.. Fielding et al., this volume). The New England Fold Belt (NEFB) and adjacent sedimentary basins of eastem Australia contain the record of Palaeozoic to Mesozoic rocks that developed at, or at least near, this margin, albeit much modified structurally by the subsequent deformations that have effected the region. The event that had the greatest influence on the distribution of those rocks preserved within the orogen is the Late Permian and Early Mesozoic contractional event that Holcombe et al. (b, in press) broadly termed the 'Hunter-Bowen event'. This paper contains a brief summary of the evolution of the NEFB during the Late Palaeozoic and Early Mesozoic, and describes the changes in style of the deformation along the orogen that contribute to the complex distribution of successions within the orogen. The Hunter Bowen event was a protracted (-35 Ma) contractional deformation that coincided with the westward migration of a convergent margin volcanic arc. The character of the deformation appears to have varied significantly with time, and in character, along the orogen. The earliest identifiable, basin-wide stratigraphic breaks that can be correlated with the onset of contraction within the Bowen Basin occurred as early as --270 Ma, while the last record is the terminal foreland-fill phase of the basin evolution at 235-230 Ma. This latter phase, whilst accounting for less than half of the absolute time of basin evolution, accounts for greater than 80% of the stratigraphic thickness preserved within the eastem basin. The chronological evolution of the deformation is not well constrained, but may have included several pulses that correspond to stratigraphic breaks within the developing foreland basin. The thrust front advanced westward with time, uplifting the earlier, relatively meagre, basin successions and re-working them into developing foreland. It is also probable that the onset of deformation varied in time from north to south, such that the timing of events in the Sydney Basin, to the south, or north of the present structural boundaries of the Bowen Basin, may have been significantly different. Thrusting appears to have propagated westward into the basin throughout its development, although the predominant thrust structures within the present basin are probably of Middle Triassic age. The NEFB contains four distinct stmctural domains characterised by varying styles of response to the contractional deformation. The northem-most domain is characterised by open fold structures both within the basement and supracrustal sequences. Basement in this region comprises Late Carboniferous-Early Permian co-magmatic granitoids and volcanics, and thrust deformation is concentrated along the westem margin of these sequences. To the east, the supracrustal rocks are folded into broad anticlines and synclines east to the present coastline. South of this domain is the Gogango Overfolded Zone (GOZ), a complex fold-thrust terrain characterised by variable cleavage development, indicative of moderate degrees of shortening, within supracrustal sequences. The GOZ is separated from the northern domain by a southwest-trending tear fault zone. Within the GOZ, there is widespread involvement of basement within the thrusting, typically as thrust-bounded windows within the supracrustal rocks but also as cores to nappes ahead of propagating thrusts. Basement in this region is the Late Carboniferous-Early Permian magmatic suite in the west, but earlier Siluro-Devonian volcanic and volcaniclastic sequences that are widely regarded as forming elements of a subduction complex, in the east. Surface structural and seismic data suggest that thrusts within this region sole out onto a basal structure at depths of 3-5 km. A sheet of serpentinite and structurally-incorporated high grade metamorphic rocks lies across the eastem part of the region and is bounded at its base by a near-horizontal thrust fault This sheet must have been emplaced very late in the structural evolution of the region, presumably corresponding to the Middle Triassic terminal foreland basin phase in the adjacent Bowen Basin. South of the GOZ, the NEFB is dominated by basement terrains that are elements of the Siluro-Devonian subduction complex in the east, but predominantly granitoid in the west. The nature of the structural transition between the GOZ and this domain is not known. Early Triassic volcanic and volcaniclastic sequences overlie basement within this domain, and are partly preserved within basins that may have developed over earlier Permian rifts. The age of the westem granitoids is not well constrained, but is regarded on regional correlations to be Late Carboniferous. Preliminary 40Ar/39Ar dating of these granitoids suggests that they were uplifted through the Ar blocking temperature for biotite at 250-240 Ma. The southem-most structural domain, the New England Province, is separatedfromthe northem domains by the Jurassic Clarence-Moreton Basin. The structure of this terrane is dominated by a broad megafold within the pre-Permian accretionary complex rocks that is separatedfromgently folded and faulted supracrustal sequences to the west by a major fault zone. This fault contains abundant serpentinite, and is widely regarded as a major fault along which the orogen was 517

thrust. Seismic data, however, shows that the major seismic feature is a west-dipping structure, suggesting that the orogen in this region was, at least in its later phases, underthrusting the basin to the west. If the interpretation from seismic that the New England Province is underthmsting the basin holds, then the Clarence-Moreton Basin may overlie a significant tear fault. The two southern structural domains have been affected by voluminous calc-alkaline magmatism, characterised by extensive intermediate-composition granitoids and, toward the north, widely preserved andesitic volcanic sequences. The age of these rocks predominantly fall within the range 250-230 Ma, ie the later half of the Hunter-Bowen event, although there is a suggestion that granitoids in the New England Province are mostly at the older end of this range. Analysis of the distribution of Triassic and later successions in the northem of these domains suggests that the volcanics formed a widespread, but not necessarily thick, blanket throughout the region. Geochemical data is consistent with these rocks having been developed in association with a convergent margin environment. The differences in character of the deformation along the orogen, and of the greater abundance of magmatism in the southem two domains, can be related to the along-orogen variations seen along modem convergent continental margins such as the south American Andes. Changes in dip of the subducting slab, clearly recorded in earthquake epicentre data, are correlated with significantly different upper cmstal structural and magmatic provinces. Nonetheless, isotopic data clearly show that these provinces also have significantly different lower crustal compositions such that, even in this apparently well-constrained environment, it is not necessarily clear whether the structural development of topographically prominent regions, or the character and distribution of volcanic rocks as few as 3-5 my old, can be related to the present slab geometry and plate convergence vectors. It may, thus, be unrealistic to expect that changes in the upper-crustal architecture of an orogen of Permo-Triassic age, such as the NEFB, can be uniquely related, both in time and space, to modem tectonic scenarios.

518

THE ROCKHAMPTON PLUME AND ITS LATE MESOZOIC TRACE? FX. SutherlandS A.D. RobertsonS B.J. Barron^ and RE. Pogson^ L Division of Earth and Environmental Sciences, Australian Museum, Sydney 2, Geological Survey of Queensland, Brisbane Summary Eroded central volcanoes east of Rockhampton give 75-79Ma uranium-lead isotope dates for zircons. Rhyolites and trachytes form a bimodal association with alkali and transitional basalts that have evolved to mugearite. Rhyolites form three distinct groups on Zr and Ba contents, a low Zr and Ba (<250ppm) group a moderate Zr (400-700ppm) and variable Ba (up to 2860ppm) group and a high Zr (<1500ppm) and low Ba (<200ppm) group. Nepheline syenite and basanitic intrusions dated at 69-72Ma (K-Ar dating) to the south suggest a potential plume trace. The Ba/Nb and La/Nb ratios for Rockhampton basalts overlap values for east Australian and Hawaiian hotspot basalts. These ratios are distinct from those of basalts associated with Early Cretaceous sihcic volcanism along the Whitsunday rift. Late Cretaceous volcanism is scarce in eastem Australia, so a potential Rockhampton plume trace bears on Australian plate motions. Such a plume may have originated with the mid-Cretaceous Cato and Capricorn rift events. Introduction Late Cretaceous volcanic rocks outcrop northeast of Rockhampton within the Jim Crow Basin and along the coast south of Yeppoon as volcanic plugs and remnant flows (Figure 1A,B,C); as described by Willmott et al. (1) and Duggan et al. (2). Wellman (3) noted that the Rockhan^Dton province marks an early episode (70Ma) in late Cretaceous-Cainozoic igneous activity in eastem Australia. An extended interval exists between Rockhampton activity and later east AustraUan felsic volcanism shown by Sutherland (4) to date from 46Ma. Wellman considered it important to confirm the Late Cretaceous age for Rockhampton volcanic rocks and compare their chemistry with the Cainozoic rocks. A Late Cretaceous hotspot trail was proposed by Sutherland et al. (5) to e^qjlain the Rockhampton volcanism and related rocks to the west and south. Joint studies by Australian Museum and Geological Survey of Queensland personnel provide some initial results that are reported here. Physiography and geology The Mt Hedlow trachyte plugs, as described by Wilhnott et al. (1) and Robertson and Sutherland (6), rise from the alluvial plain of Hedlow Creek from around 40m elevation to form sharp peaks up to 200 m in hei^t. Similar plugs up to 393 m high intrude undulating to hilly country of the Cawarral Serpentinite Belt. South of Yeppoon several bodies up to 100m high in coastal exposures intmde sedimentary beds of the early Carboniferous Wandilla Formation. During the Cretaceous and Cainozoic minor downwarping and localised faulting, often on pre-existing faults, occurred. Within the Jim Crow Basin, the emplacement of plugs appears to be confined by suspected bounding faults (Figure IB). Outside the basin, the structure controlling emplacement is not obvious. West of Rockhampton, more extensive basalt flows and felsic rocks include pyroclastic deposits and flows and extend for 50 km, northwest to Mt Sahnon and southwest to Mt Sebastapol. However, early Cretaceous volcanics are also present, so that the Late Cretaceous components are of uncertain extent. Some basalt was dated as Late Cretaceous by Wellman (3) and a basalt dyke intrudes rhyolites at Mt Hay (authors' observations). Mt Ramsay, a nepheline syenite intrusion, near Baralaba and basanitic plugs carrying mantle inclusions at Mt Runsome extend the Late Cretaceous rocks to 200 km south of Yeppoon (Figure 1A,D).

519

Figure 1. Distribution of Cretaceous volcanic rocks in the Yeppoon - Rockhampton - Baralaba area, including dated Late Cretaceous volcanic rocks and offshore structures. A Distribution of basalts (stippled areas), felsic volcanics (enclosed areas) and felsic intrusives (enclosed dots), with volcanic regions linked by dashed trend line. B Detailed distribution of basalts and silicic intrusives and bounding faults, northeast Rockhampton-Yeppoon area. MM Mt Mungawappa, MH Mt Hedlow, RC Rocky Cone Mt, MC Mt Coberra, CH Camp ffill Rock, PM Pine Mt, JC Jim Crow Mt, IP Iron Pot Mt, MW Mt Wheeler, C Cawarral, DH Double Head, BP Bluff Point, PP Pinnacle Point. C General offshore structures, Queensland east coast in relation to Rockhampton area. D Age distribution of dated Late Cretaceous volcanic rocks, Rockhampton province, showing apparent migration trend and extrapolation of trace to time of Capricorn Basin rifling. Symtols represent features named in C. E Late Cretaceous — Early Tertiary spreading rift systems, east Australian margin, in relation to South Tasman seamount trace (arrowed lines in stippled basaltic areas) and proposed Rockhampton volcanic trace R. Numbers are ages in Ma. CS Coral Sea spreading, CB Cato Basin rifting, TS Tasman Sea spreading, LHR Lord Howe Rise. Midocean spreading axes (double lines), offset by transform faults (lines).

520

Sampling procedures Known outcrops of Mount Hedlow trachyte were sampled, including observed textural and/or mineralogical variations. Basalts were sampled from boulder outcrops or cobbles in basaltic soil and showed greater textural variation than the felsic plugs. Basaltfroma plug at Eulogie Park, 50 km SWS of Rockhampton provedfreshenough for whole rock dating, although it carries abundant country gabbro and mantle inclusions. The variable weathering of the felsic rocks made them unsuitable for K-Ar age determinations. However, the weathering products release zircon, which was recovered where possible forfissiontrack and U-Pb isotope dating. Analytical methods The Eulogie Park basalt was dated using convention K-Ar methods, at AMDEL Laboratories, Adelaide (Table lA). Zircons from the felsic rocks were submitted forfissiontrack analyses at Geotrack Laboratories, Melboume. However, due to high uranium contents, a h i ^ density of tracks weakened the structure which generally did not survive the etching process. One low U-grain (U 50ppm) from Camp Hill East survived for an unsupported age measurement. To overcome the dating problems, zircon grains were dated through U-Pb isotope methods, using the SHRIMP ion microprobe facility at Australian National University. The zircon euhedra are considered late-accessory minerals crystallising in the rock, so should provide formation ages for the host volcanic rocks. These geologically young zircons require special treatment of the Pb-U isotopes, as described in Coenraads et al. (7). The ^^Tb provides an efficient means of monitoring the concentrations of''common" Pb. In this case the normalised to a standard provides an initial can be calculated. This value, using a modelled common Pb age estimate. From this the e^qjected radiogenetic composition, then yields an accurate proportion of non-radiogenic in the total It provides a revised estimate of radiogenic ^^Tb/^^U and hence a corresponding age (Table IB). Major and trace elements for the rocks were detemiined at the Queensland Government Chemical Laboratories mostly using a Phillips P.W. 1401 sequential spectrometer. Full matrix corrections were employed. Chlorine measurements used pressed powder discs and heliumflushing.A Leco induction ftimace was used for carbon dioxide (by infra red) and fluorine (by specific ionfluorideelectrode) determinations. Cu, Co, Cr and Ni were measured by Atomic Absorption Spectrophotometer (AAS). The limit of detection of trace elements determined using XRF methods is 1 part per million (ppm), with an accuracy of ± 5 percent. Elements determined by AAS have detection limits of lOppm for Cr, Co and Ni and 5ppm for Cu. The precision of determination at the single standard deviation for major elements is ± 0.2 for Si02, ±0.1 for AI2O3, MnO, MgO, CaO and Na20 and ± 0.05 for Ti02, K2O and P2O5. The rock nomenclature used was based on the TAS classification system of Le Maitre (8) as most rocks were too finegrained to assess accurate mineral modes. Some analysed basalts gave over 0.5% CO2 indicating weathering. Chemical analyses, related indices and CIPW norms are presented for representative rocks of known late Cretaceous age (Table 2A, B) and provide data forfiirtherdivisions of rock types. Results The Eulogie Park basalt date of 69Ma (Table lA) shows Late Cretaceous volcanism extends south of Mt Morgan. A lone fission track measurement of a zircon from Camp Hill East rhyolite gave 50.5±4.4Ma (Geotrack Report 61, Australian Museum) suggesting a probable association with nearby plugs that gave Late Cretaceous U-Pb zircon ages. Zircon from Black Mountain trachyte, Cawarral rhyolite and Jim Crow Mountain rhyolite gave 73-80±3Ma, apartfromone Jim Crow Mountain grain which gave 242±5Ma and is presumably a xenocrystfromthe underlying basement. Mt Ramisay zircons also gave significantly older ages (260-510±5-10Ma) than the Late Cretaceous K-Ar age for the intrusion, also suggesting they are basement xenocrysts. This is compatible with the peralkaline nature of the rock, as Linthout (9) showed such conditions inhibit zircon crystallisation. Major element analyses reveal a trimodal distribution for basalts, saturated felsic rocks and undersaturated felsic rocks (Figure 2A). The basalts range from primary basanite/ne hawaiite (Mg# 0.72) that carries mantle xenoliths into mugearites (Mg#0.52). Some basalts appear transitional to olivine tholeiites (over 10% Hy), but contain 2 wt% CO2 so this may reflect an alteration effect. Qz trachytes (up to 10% Qz) range into rhyolites (over 10% Qz), using the classification of Johnson and Duggan (10). Rhyolites are more common (c 60%) and are mostly low Si-types with Ab exceeding Or, with rare high Si-types (Qz over 30%), following the usage of Ewart (11). The term Mount Hedlow Trachyte for the east Rockhampton felsic rocks clearly does not convey thefiillpetrologic range of these rocks. The Mt Ramsay nepheline syenite represents a mildly sodic (Na/K 1.5) and peralkaline (Ac 2%) undersaturated end member (Ne 7%). In trace elements the felsic rocks show strong decrease in Sr and enrichment in Rb compared to the basalts. Zr and Nb are enriched in most felsic rocks as expected from crystalfractionationprocesses, as found by Ewart (11) for Cainozoic suites. However the high Si-rhyolites show abnormally low Zr, below levels in most basalts, and low Nb, La and Ce compared to other rhyolites, which Ewart (12) suggested may reflect a different genesis involving crustal melts. Zr-Ba plots (Figure 2B) show trachytic (and ne syenite) rocks form a coherent group showing moderately high Zr. Rhyolites^ 521

however, split into three distinctfields,high Zr (over ISOOppm) and low Ba (<200ppm), moderate Zr (400-700ppm) and variable Ba (up to 2860ppm) and low Zr and Ba (<250ppm). Ba/Nb and La/Nb plots for basalts show a fairly low Ba/Nb range and a more extreme La/Nb range (Figure 2C). Discussion The 73-79Ma zircon ages for east Rockhampton volcanoes confirm the previous Late Cretaceous datingfiromK-Ar determinations and suggest an apparent decrease in age and volume in the volcanic activity fijrther south (Figure lA, B, D). The full extent of Late Cretaceous activity west of Rockhampton awaits detailed delineation of the Early Cretaceous volcanic centres there. The older Palaeozoic 'xenocryst' ages for zircons from Mt Ramsay indicate that U-Pb isotope dating alone is not always sufficient to establish host rock ages, unless checked with other dating methods. The essential bimodal nature of Late Cretaceous Rockhampton volcanism is observed in several east Australian Cainozoic migratory central volcanoes, including central Queensland provinces, based on pattems compiled in Knutson (13) and found in a 21Ma central volcano at Mt Belmore, New South Wales (authors' unpubhshed data. Figure 2B). This bimodal separation is not seen in the mafic-silicic early Cretaceous Whitsunday rift volcanic suites described by Stephens et al. (14) north of the Rockhampton province. Trace element data indicates that the Rockhampton rhyolites were evolved along different paths, possiblyfiromdifferent basaltic parents and trachytic intermediaries, or in the case of the high Si rhyoUtes with involvement of crustal melts. Plots of Ba/La against La/Nb for Rockhampton province basalts are compared with east Australian and oceanic island plume related basalts, such as Hawaii (Figure 2C), following Sun et al. (15). The Rockhampton data partly overlaps the fields for east Australian central volcano basalts, the newer Victorian basalts and the Hawaiian Koolau basalts. In contrast, basaltsfiromthe early Cretaceous Whitsunday rift volcanism studied by Stephens et al. (14) show significantly higher Ba/Nb and fall well outside the Rockhampton and other plume related basalt fields. The new 75-79Ma dating for east Rockhampton volcanism in relation to known 71-72Ma activity south of Baralaba suggests a potential west to southwesterly migration of volcanic activity (Figure ID). The estimated migration rate of around 2-4 cm/yr is comparable with a southwesterly migration rate shown in Sutherland (16) for a Tasman seamount plume trace datedfi-om71-43Ma (Figure IE). Extrapolation of the Rockhampton plume migration back in time places the plume below the Capricom Basin around 80-85Ma, close to the time indicated for inception of this rift structure by Hill (17). Further eastward extrapolation would place the plume near the Cato Trough margin, on a mid-Cretaceous thermal rift line that Sutherland (4,16) considered formed a locus for an extended series of plumes along the Tasman margin. Late Cretaceous volcanism between 70-80Ma is relatively sparse in Australia outside the Rockhampton province, apart firom basaltic activity associated with the Ballan graben, as shown by Sutherland (4). Thus, the proposed Rockhaii5)ton plume trace provides additional information on Australia's plate motion beyond that availablefiromthe Tasman spreading zone. Conclusions 1. A Late Cretaceous age for the Rockhampton volcanic province is confirmed onfiirtherK-Ar basalt and zircon UPb isotope dating. 2. The province exhibits bimodal basaltic and silicic activity and the basalts show afiSnities to plume related Australian and oceanic island basalts. 3. A proposed plume trace suggests northeast to eastward Australian plate motion between 70-80Ma and possible origin with mid-Cretaceous rifting. Acknowledgements C.M. Fanning, Geochronology Laboratories, Research School of Earth Sciences, Australian National University for UPb isotopic analyses of zircons through the SHRIMP ion microprobe. J.D. HoUis, Australian Museum, Research Associate for field assistance. Department of Resource Industries Laboratories, C^eensland, for assistance with preparation of samples and chemical analyses. Australian Museum Trust for support fimding. S. Folwell and R. Springthorpe helped in the preparation of the paper. References 1. Wilhnott, W.F., O'Flynn, M.L. & Trezise, D.L. 1986. Rockhampton Region Queensland, Geological Survey of Queensland 1:100000 Geological Map Commentary. 2. Duggan, M.B., Sutherland, EL. & Martin, D J. 1989. Mesozoic intraplate volcanism and related intmsions. In Johnson R.W. (ed.) Intraplate Volcanism in Eastern Australia and New Zealand, pp. 149-150. Cambridge University Press, Cambridge. 3. Welknan, P. 1986. Potassium-Argon ages of Cainozoic rocksfiromBundaberg, Rockhampton and Clermont areas of eastem Queensland. Proceedings of The Royal Society of Queensland 89, pp.59-64. 4. Sutherland, F.L. 1991. Cainozoic volcanism, Eastem Australia: a predictive model based on migration over multiple 'hotspot' magma sources. In WiUiams, MA.J., De Decker, P. & Kershaw, A.P. (eds.) The Cainozoic in Australia: A re-appraisal of the evidence. Geological Society of Australia Special Publication 18, pp. 15-43. 522

5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Sutherland, FX., Robertson, A.D. & Mollis, J.D, 1988. The Rockhampton Province - a Cretaceous central volcano migration? Geological Society of Australia Abstract Series 21, pp. 389-390. Robertson, A.D. & Sutherland, F.L. 1994. Mount Hedlow Trachyte. In Holcombe R.J., Stephens, C.J. & Fielding, C.R. (eds.) Capricorn region, central coastal Queensland. 1994 Field Conference, pp.118-121. Geological Society of Australia, (^eensland Division, Brisbane. Coenraads, R.R, Sutherland, EL. & Kinny, RD. 1990. The origin of sapphires: U-Pb dating of zircon inclusions sheds new light. Mineralogical Magazine 54, pp.113-122. Le Maitre, R.W. 1989. A Qassification of Igneous Rocks and Glossary of Terms. Blackwell Scientific Publications, Oxford. Linthout, K. 1984. Alkali zirconosilicates in peralkaline rocks. Contributions to Mineralogy and Petrology 86, pp.155-168. Johnson, R.W. & Duggan, M.B. 1989. Rock classification and analytical data bases. In Johnson, R.W. (ed.) Intraplate Volcanism in Eastem Australia and New Zealand, pp.12-13. Cambridge University Press, Cambridge. Ewart, A. 1989. East Australian petrology and geochemistry. In Johnson, RW. (ed.) Intraplate Volcanism in Eastem Australia and New Zealand, pp. 189-246. Cambridge University Press, Cambridge. Ewart, A. 1989. Fractionation, assimilation and source melting: a petrogenetic overview. In Johnson, R.W. (ed.) Intraplate Volcanism in Eastem Australia and New Zealand, pp.324-333. Cambridge University Press, Cambridge. Knutson, J. 1989. East Australian volcanic geology. In Johnson, R.W. (ed.) Intraplate Volcanism in Eastem Australia and New Zealand, pp.89-155. Stephens, C.J. Schon, R.W. & Ewart, A. 1993. Mesozoic crustal extension in the northern New England Orogen: Geochemical and isotopic evidencefi-omlarge scale silicic magmatism. In Flood, P.G. & Aitchinson, J.C. (eds.) New England Orogen, eastem AustraUa, pp.63 7-642. Department of Geology and Geophysics, University of New England, Armidale. Sun, S.S., McDonough, W.F. & Ewart, A. 1989. Four con^nent model for East Australian basalts. In Johnson, R.W. (ed.) Intraplate Volcanism in Eastem Australia and New Zealand, pp.333-347. Cambridge University Press, Cambridge. Sutherland, F.L. 1994. Tasman Sea evolution and hotspot trails. In Van der Lingen, G.J., Swanson, K.M. & Muir, R.J. (eds.) Evolution of the Tasman Sea Basin, pp.35-51. A.A. Balkema, Rotterdam. Hill, P.J. 1992. Capricorn and Northern Tasman Basins: Structure and depositional systems. Exploration Geophysics 23, pp.153-162.

523

felsic rocks D.I. (qz. or, ab. ne. Ic. ks) 80 Z benmoreite

60 phonolite nepheline benmoreite

mugearite ne-mugearite

40 Hawaiite ne-hawaiite

basalt basanite

2000

An An+Ab% 30

50 Zr (ppm) •

B 27

Ba/Nb

R•

1500

^pNS

... ^ \ 1000

19 WV

V •

•J

BV 1

X RV

500

K. *

^ /

/T® / 4 1

2860 ppmBa

HB •

Ba (ppm) 500

•

1000

U/Nb

0.6

1.0

Figure 2, Chemical variation diagrams, Rockhampton province. A Differentiation Index (DI) versus noraiative plagioclase ratios. Arrowed lines suggest potential fractionation paths. B Zirconium versus barium plots for Rockhampton province (dots) and comparative bimodal Cainozoic Belmore province (crosses, authors' unpublished data). Enclosedfieldsinclude B basalt, T trachyte, Rrhyolite, S ne syenite C Ba/Nb versus La/Nb plots, (dots) for Rockhampton basalts province (RV)in relation to basalts from other fields. CV Cainozoiccentral volcanoes, BV Cainozoic Belmore volcano, W Late Cainozoic Victorian volcanics, HB Hawaiian Koolau basalts, WV Early Cretaceous Whitsunday volcanics, CC crustal contaminated melts. 524

Table lA: K-Ar results for basalt plug, Eulogie Park, Qld Sample %K# moles/g) ^Ar^AT'Total Eulogie Park 1.923 23.304 0.971 1.921 Total Rock # The mean K value is used in the age calculation * Denotes radiogenic Ar + Age in Ma with error limits given for the analytical uncertainty at one standard deviation. Constants ''K = 0.01167 atom % ^3=4.962x10-^^ X^=0.581xl0-y

Age^ 68.6±0.5

Table IB: U-Pb isotope dating of zircons, Rockhampton Province Grain spot Uppm Thppm Th/U Pb* ppm 204/206 206/238 Age Black Mountain 1.1 710 929 1.31 11 0.001256 0.0123±0.0004 79±3 2.1 159 80 0.50 2 0.0125±0.0004 80±3 3.1 365 276 0.76 5 0.000311 0.0123±0.0004 79±3 Cawarral 1.1 1197 1640 0.73 22 0.001374 0.0120±0.0005 77±3 2.1 1344 637 0.47 16 0.000014 0.0113±0.0004 73±3 1374 3.1 1798 24 0.76 0.000383 0.0118±0.0004 75±3 Jim Crow Mountain 264 11.1 142 0.54 3 0.0116±0.0003 74±2 12.1 277 221 4 0.80 0.0117±0.0003 75±2 14.1 104 0.62 65 4 0.000904 0.0383±0.0008 242±5 Mount Ramsay 121.1 72 275 0.26 22 0.000321 0.0822±0.0017 509±10 122.1 343 39 0.11 26 0.000008 0.0796±0.0016 494±10 123.1 811 456 0.56 35 0.000088 0.0415±0.0008 262±5 (1) All 3 scan reconnaissance data. (2) denotes no detectable '^'Pb. (3) Uncertainties given at one sigma level. (4) Correction for common Pb made on the basis of extrapolation to concorida along a mixing line with common Pb. Black Mountain and Cawarral grains from mount Z1311 and Jim Crow Mountain and Mount Ramsay grains from mount Z1554. Analyst C.M. Fanning, SHRIMP Facihty, Geochronology Laboratory, AustraUan National University. -

525

Table 2A: Representative analyses, Rockhan:5)ton Province basaltic rocks 4

Si02 Ti02 AI2O3 Fe203 FeO MnO MgO CaO Na20 K2O P2O5 CO2 H2O+ H2O-

46.30 48.20 1.60 2.30 13.80 17.50 2.80 1.70 7.80 7.80 0.16 0.16 11.10 7.30 7.70 9.90 3.50 3.90 0.70 1.90 0.23 0.58 0.10 <0.10 0.36 1.03 0.34 0.30

Total

99.14

99.92 100.48

99.66

99.68

11.40 18.01 14.76 8.39 15.73

4.20 26.91 30.40 1.69 14.56

5.45 30.35 27.03

4.25 31.28 25.27

10.32 40.00 12.58

22.99 2.60 4.43 1.39 0.23

16.28 2.33 3.08 0.55

9.37 4.10 14.79 2.74 4.67 1.53

2.34 10.21 14.82 2.35 3.51 1.34 4.67

CIPWNorm Or Ab An Ne Di Hy 01 Mt 11 Ap Cc C 1 2 3 4 5

Wt % oxide

47.50 46.40 48.10 1.90 1.80 2.40 16.40 15.70 16.70 2.80 2.80 4.10 6.80 7.10 7.20 0.17 0.16 0.18 8.20 4.90 6.60 8.80 6.40 8.40 4.60 3.60 3.50 1.70 0.90 0.70 0.76 0.63 0.55 2.30 2.00 1.55 1.76 1.79 0.49 0.60 0.88

13.29 7.60 2.43 3.71 1.85 5.38 2.88

ppm

Ba Ce CI Co Cr Cu Dy F Ga Gd Hf La Nb Nd Ni Pb Pr Rb Sc Sm Sr Th U V Y Yb Zr

300 65 365 30 640 105 1 580 18 3 6 32 48 31 460 6 6 30 20 7 725 6 1 180 21 2 255

145 39 305 20 190 43 3 230 19 3 5 19 11 18 60 4 4 10 32 6 470 2 0 0 26 2 190

175 60 270 20 180 45 3 490 22 7 4 40 20 31 70 4 35 13 26 7 550 3 0 0 29 3 210

220 61 325 20 310 51 0 530 17 5 4 27 28 31 60 4 5 13 25 6 705 3 0 0 26 1 250

445 80 280 20 120 30 3 520 16 8 8 37 37 39 50 3 9 22 15 7 960 5 0 0 29 3 395

Ne hawaiite (Mg#0.72, An%49.9, D.I.32.8), Mt Runsome, Mundubbera 1:250000 (362,886) Alkali basalt (Mg#0.62, An%53.0, D.I.32.8), East Lake Mary, Rockhaii?)ton 1:100000 (536,414) Hawaiite (Mg#0.56, An%47.1, D.I.35.8), 5 km W of Alton Downs, Ridgelands 1:100000 (280,207) Transitional liawaiite(Mg#0.64, An%44.7, D.L35.5).West Mt Cobbera,Rockliairptonl: 100000(542,207) Mugearite (Mg#0.52, An%23.9, D.I.50.3), Mt Hedlow, Rockhampton 1:1000000 (549,312)

526

Table 2B: Representative analyses, Rockhanpton Province felsic rocks Wt % oxide 6 7 8 9 10 ppm 6 7 8 9 61.80 68.50 70.80 72.40 74.80 Si02 Ba 120 215 13 18 0.12 0.18 0.01 0.11 0.05 Ti02 Ce 270 190 125 190 AI2O3 18.20 15.80 13.90 13.20 14.00 CI 240 140 150 50 Fe203 3.10 2.80 2.80 2.50 0.45 Co 0 0 0 30 FeO 1.10 0.12 0.26 0.40 0.08 Cr <10 20 0 1 0.14 0.06 0.00 0.05 0.00 MnO Cu <5 0 0 <5 MgO 0.10 0.27 0.15 0.20 0.25 8 16 Dy 17 CaO 0.50 0.27 0.09 0.10 0.14 F 230 20 260 410 Na20 8.30 5.30 5.80 5.70 4.00 Ga 41 42 28 37 K2O 4.90 5.10 4.30 4.20 4.30 Gd 16 9 15 9 0.01 0.02 <0.02 Hf 18 14 42 26 P2O5 <0.10 <0.10 CO2 La 141 110 30 79 H2O+ 1.21 0.64 0.60 0.56 1.01 Nb 210 59 125 180 H2O0.38 0.45 0.39 0.55 0.56 Nd 94 73 45 76 Ni <10 0 0 <10 Total 99.96 99.51 99.10 100.09 99.64 Pb 27 11 31 15 Pr 32 28 34 20 CIPW Norm Rb 195 105 155 200 15.90 19.08 22.63 34.81 Sc 0 3 0 <10 Q Or 29.43 30.69 25.96 25.10 25.92 Sm 17 12 9 16 Ab 54.31 45.67 48.59 45.01 34.52 Sr 12 18 6 6 Ne 7.10 Anl.23 0.71 Th 30 15 21 32 Di 1.61 0.40 0.44 U 11 0 0 6 Ac 2.06 1.37 1.41 V 2 0 0 <10 4.30 4.57 4.65 1.25 Hy Y 125 43 125 130 01 4.60 C1.14 2.56 Yb 5 6 15 6 Mt 0.66 0.01 0.12 Zr 1485 585 1850 1255 11 0.23 0.35 0.02 0.21 0.10 Ap 0.05 0.02 Ns 0.38 0.50 6 Ne syenite (Mg#0.05, An%0.0, D.I.61.4), Mt Ramsay, Baralaba 1:100000 (930,130) 7 RhyoUte (Mg#0.18, An%2.6, D.L92.3), Black Mt, Rockhampton 1:100000 (512,359) 8 Rhyolite (Mg#0.10, An%0.0, D.I.93.63), E. Camp HiU, Rockhampton 1:100000 (534,348) 9 Rhyolite (Mg#0.13, An%0.0, D.I.92.74), Jim Crow Mt, Rockhampton 1:100000 (574,304) 10 Rhyolite (Mg#0.52, An%2.0, D.I.95.25), Cawarral, Rockhampton 1:100000 (609,257)

10 195 32 180 0 20 0 5 230 28 3 3 9 61 18 0 12 3 91 0 5 14 10 0 0 38 5 235

527

MESOZOIC RIFT BASIN DEVELOPMENT OFF EASTERN AUSTRALIA P.A. Symonds, J.B. ColweU, H,LM. Struckmeyer, J.B. WiUcox and P.J. Hfll Australian Geological Survey Organisation, GPO Box 378, Canberra, ACT, 2601, Australia Summary During the Late Jurassic-Palaeogene a vast extensional terrane developed througji eastern Australia in response to the fragmentation and dispersal of Gondwana. The earliest (Late Jurassic-Early Cretaceous) episode of rifting occurred between Australia and Antarctica, and probably onto the Lord Howe Rise, in a convergent tectonic setting more than 1000 km behind the proto-Pacific plate boundary. The transition from convergence to divergence was characterised by extensive Barremian-Cenomanian, extensional/transtensional magmatismnear the locus of future Tasman breakup. Areas of intense volcanism in the northeast and southeast shed large volumes of volcanogenic sediment westwards into the developing basin systems. In the Cenomanian-Campanian a second major rift system developed along eastem Austraha in association with slow seafloor spreading south of Australia. This extensional terrane was broken up and dispersed during Campanian to Eocene seafloor spreading. Along-strike variations in extensional style and the location of breakup have produced a complex present-day distribution of plateaus, troughs, rift systems and ocean basins off eastem Australia. Introduction Following a prolonged episode of convergent tectonism that ended in the Middle-Late Triassic in eastem Australia [1] and probably in the mid-Cretaceous in the New Zealand region [2], a Late Jurassic-Palaeogene extensional system developed along the eastem margin of the Australia plate. This system extended from New Guinea in the north to Antarctica in the south, and was probably up to 1000 km in width prior to continental breakup and seafloor spreading. Research programmes over the past decade by the Australian Geological Survey Organisation (AGSO) have provided new information on the Mesozoic extensional regime. These programmes have confirmed that large sedimentary basins, some of considerable thickness and structural complexity, occur in deep water (200-2000 m) beyond the frontier of conventional offshore e^loration activity along the eastem Australian margin and its conjugate (Fig.l). The most prospective rift basin systems lie adjacent to the Gulf of Papua, where they have now been incorporated into a distal foreland basin setting, and also beneath the Queensland and Townsville Troughs off northeastem Australia, where they are fully intact; the flanks of the Lord Howe Rise and West Norfolk Ridge, where they were breached during breakup; the southem New Caledonia Basin; the flanks of the South Tasman Rise; and in the deep-water part of the Gippsland Basin. Apart from the system of basins in Bass Strait (Otway, Bass and Gippsland Basins) and the Taranaki Basin off westem New Zealand, the margin and its conjugate remains largely une?5)lored for petroleum. This paper presents a brief overview of Mesozoic rift basin development off eastem Australia. It draws on regional seismic data tied to sparse exploration wells and Deep Sea Drilling Program (DSDP) holes, as well as seafloor samphng and onshore mapping, to document deformation, magmatic and depositional events that controlled rift basin and passive continental margin development. Tectonic Setting and Events In spite of a lack of detailed well control and the sparsity of seismic data in many areas, it is possible to document some widespread deformational and magmatic episodes (Table 1), and define dramatic changes in tectonic style throughout the region. During most of the Palaeozoic to early Mesozoic eastem Austraha was a convergent margin, with periods of oblique subduction represented by northwest trending, parallel belts of volcanic arc, forearc basin and subduction complex successions [e.g. 1,3,4]. 'Basement' underlying the present-day margin and its conjugates is thus likely to comprise an amalgamation of these provinces. The Maryborough Basin, and similar coastal basins in northeast Australia, developed in the latest Triassic to Middle Jurassic as epicratonic downwarps within a foreland setting [5] following docking of the Gympie Province (terrane) with the New England Orogen [6]. To the north, in New Guinea, Triassic-Early Jurassic rifling lead to breakup, margin development and seafloor spreading in the Middle-Late Jurassic [7]. In the Middle Jurassic further south, a short-lived tholeiitic magmatic event formed the Tasmanian dolerite province (175±18

528

23/OA/724

Figure 1. Generalised bathymetry and seafloor spreading anomalies off eastem Australia (modified from Symonds & Colwell, [9]). Also shows the outlines of major basins (large dots) and other areas of Mesozoic rift basin development (stipple). Tasman Basin magnetic lineations after Shaw [11,12]; others after Scheibner et al. [12]. Ma; Hergt et al. [8]). This was part of a long, linear flood basaltic belt within Gondwana that stretched from southem Africa (Karoo province) through Antarctica (Ferrar province) to AustraUa(Tasman province)[13, 14]. This magmatism probably represents the initial stages of Gondwana breakup through eastem Australia, perhaps associated with a broad zone of mantle melting related to a reduction in subduction-plate boundary forces along the convergent, proto-Pacific Gondwanan margin, which lay some thousands of kilometres to the east [14]. The Middle Jurassic magmatic belt is normally shown as ending in the Tasmania area [13, 14], but it is plausible that it could have extended fijrther along eastem Australia, and its remnants may now lie beneath Lord Howe Rise. In the Late Jurassic to Early Cretaceous, despite initial rifting in places, the overall tectonic setting of the margin was still largely convergent, with the plate boundary lying to the east of the Lord Howe Rise/Norfolk Ridge system [e.g. 15], which was still attached to the AustraUan continent. In northeast Australia large volumes of volcanogenic sediment were shed west into the Surat and Maryborough Basins in pulses that began in the Middle Jurassic (<165 Ma), and, although their chemical character and tectonic setting remain unclear, they have been inferred to be the products of a magmatic arc lying east of the present coastline [16]. The latest phase of this magmatism has been associated with the silicicintermediate Whitsunday-Proserpine-Grahams Creek volcanic belt, which extends for about 900 km along the northeasi 529

Australian coast; however, this 135-95 Ma episode of volcanism is now thought to have formed in an extensional/transtensional setting as a precursor to breakup along eastern Australia [17]. NW-SE directed oblique extension may have commenced in the Queensland and Townsville Basins at this time [18,19]; however, there is no direct evidence for the age of the syn-rifl section. Further south, a major phase of intracontinental extension was occurring in the Late Jurassic-Early Cretaceous (Tithonian-Barremian) along southem Australia resulting in block rotation, synrifl/late-rifl sedimentation and some basic volcanism. This complex rift system (the Southem Rift System; [20]) extended from the Great Austrahan Bight (Eyre Sub-basin) through the Otway (Casterton/Pretty Hill Fomiations) and Bass Basins into the Gippsland Basin (lower Strzelecki Group) [20,21,22], and perhaps to the west and south of Tasmania onto the South Tasman Rise The direction of initial extension throughout the rift system is still under debate, and directions varying from NW-SE to NE-SW have been assigned to various segments [see 22]. There is no direct evidence for rift basins of this age in New Zealand or beneath Lord Howe Rise. In fact, New Zealand and New Caledonia were active convergent margins at this time [2] driven by the oblique subduction of the Phoenix Plate. The Southem Rift System formed on a trend that was roughly perpendicular to this convergent plate boundary, and initial rifting therefore occurred in a distal backarc setting with respect to the subducting proto-Pacific plate. In the Early Cretaceous (Aptian) an abrupt change occurred in the tectonic setting of the eastem Australian plate from a compressional to an extensional regime. The culmination of the convergent regime is represented by the widespread deformation, metamorphism, uplift, erosion and calc-alkaline volcanism of the Rangitata II Orogeny (-110-105 Ma) in New Zealand [2], and the Neo-Cimmerian Orogeny in New Caledonia [23], where it appears to have continued for a ftirther 10 million years. This change in tectonic style has been attributed to the collision of the spreading ridge between the Pacific and Phoenix Plates with the subduction zone [2]. Laird [24] proposed that the earliest extensional tectonism in the New Zealand region occurred in the Albian (105-100 Ma), but there is evidence indicating it may be as old as 110 Ma [e.g. 25], It is represented by: graben development along the west coast, including the Taranaki Basin and its possible extension beneath the southem New Caledonia Basin [26], parts of the Campbell Plateau, Bounty Trough, Chatham Rise, and probably the Challenger Plateau of the southem Lord Howe Rise [24], and throughout the West Norfolk Ridge complex [27,28,29]; and by ductile deformation associated with metamorphic core complexes on the northwestem part of the South Island [25]. On the basis of evidence from the South Island, such as thermal overprinting, fission trackderived uplift/cooling histories and intraplate magmatism. Laird [24] suggested a second Albian-Cenomanian (100-95 Ma) extensional episode. All the indicators point to NNE-NE directed extension in the mid-Cretaceous. From the Barremian-Cenomanian (120-95 Ma) the southeast Australian basin system (Otway, Bass and Gippsland) was filling with late-rift to sag-phase, slightly alkaUne [30], volcanogenic sediments of the Eumeralla Formation, the Otway Group and the upper Strezlecki Group (Wonthaggi Formation), which are all thou^t to have an eastem provenance [22]. The composition and provenance of these sediments has led many workers to conclude that they were derived from an Aptian-Albian volcanic arc to the east of the Gippsland Basin, presumably now beneath Lord Howe Rise [e.g. 16, 22, 31]. However, as they were being deposited during the time that convergence was ceasing in New Zealand (110-105 Ma) and widespread extensional tectonism was begmning (105-95 Ma), and as the source volcanics would have been 5001000 km behind the subduction front, we believe that the volcanogenic sediments are most likely the result of transitional, extensional/transtensional magmatism along the line of incipient Tasman Sea breakup (see also Bryan et al. [32]). Other evidence for such an Aptian-Albian extensional/transtensional magmatic province off eastem Australia includes: 100-90 Ma age mafic to intermediate intmsions alongtihiesoutheast Australian coast [16]; subaerial to very shallow marine 94 Ma (96 Ma as recalculated by Laird, [24]) rhyolites and tuffsfromDSDP Site 207 on the southwestern Lord Howe Rise [33]; and the main phase (120-100 Ma) of the Whitsunday-Proserpine volcanics and concurrent left-lateral faulting [17] off northeast Australia, and associated volcanogenic sediment of the Aptian-Cenomanian Rolling Downs Group of the Surat and Eromanga Basins [16]. This province may now lie largely on the Lord Howe rise. The southem part of the province and the volume of volcaniclastics shed east into the Bass Strait basins may have been influenced by tiie Balleny

530

Age

Papua New Guinea

Northeast Australia Southeast Australia Lord Howe Rise New Caledonia/ New Zealand n Accretion/collision related ?Back-arc extension; mafic Arc & accretion related Maryborough Basin; silicic & intrusives of Tasmania & tectonism/magmatism mafic plutons/volcanics (210- Transantarctic Mts. (155-185 230 Ma) Ma) L Jurassic-E. Breakup & sag; Rifting - Ttranstension; Southern margin rifting; Otway Arc & accretion related Cretaceous tholeiites. andesites & acid-intermediate volcanics &. - Casterton Fm. basaltic tectonism/magmatism (120-150 Ma) shoshonites - ?arc or intrusives - Graham's Ck. volcanics (Tithonian) & Pretty extension Fm. (136-145 Ma); Hill Fm. (Berriasianvolcanogenic sedin:ient into Hauterivian); Gippsland Surat Basin; ?start of rifting in lower Strzelecki Gp. Queensland & Townsville (Tithonian-Hauterivian); ?-130 Basins Ma breakup south of Australia BarremianMargin sag Intracratonic sag - extension; Late synrift-sag; non-marine, Start of extension on Deformation, uplift & erosion, Cenomanian silicic & niinor intermediate slightly alkaline volcanogenic west Lord Howe Rise metamorphism - end of (95-120 Ma) & mafic pyroclastics, lavas & sediment (97-120 Ma)(?Albian) convergent tectonism intrusives - Whitsunday Eumeralla Fm. (Otway). undiff. (Rangitata II Orogeny - -110 volcanics (mainly 1(X>-120 Otway Gp. (Bass) & Upper' Ma); start of extensional Ma); distal volcanogenic Strzelecki Gp. (Gippsland); tectonism (100-105 Ma); sediment into Surat/Eromanga 95-100 Ma intrusives/ ?seafloor spreading east &. Basins (Rolling Downs extrusives in Boobyalla Subwest of Norfolk Ridge - ? 102 Group) & rapid basin basin & NE Tasmania; maficMa oceanic basement now subsidence; rifting in intermediate intrusion along SE ophiolites of Northland (Queensland/Towns vi 1 le coast (90-100 Ma) Allochthon, New Zealand basins Cenomanian Margin sag Contraction? - folding, uplift, Start southern margin slow Subaerial/shallow New Zealand - extension & (90-100 Ma;-95 erosion in Bowen, Surat & spreading (96 Ma); uplift, marine rhyolite niffs & alkaline basaltic, intraplate Ma) Maryborough Basins (90-95 erosion & some compression flows (96 .Ma); volcanism/ plutonism (95-100 Ma) (90-100 Ma) brecciation & Ma); also uplift and cooling in hydrothermal places; end Neo-Cimmerian alteration of granite Orogeny in New Caledonia (95 Ma); ?extension CenomanianUplift & rifting along ?Exiension in Queensland. Extension in Gippsland; NE- ?Rift basin New Zealand extension Campanian sinistral strike-slip Townsville & Capricorn SW extension in SE Bass development intrusion, uplift & fault (80-95 Ma) fault system in eastem Basins (Boobyalla Sub-basin); reactivation (80-85 Ma) Papuan Basin ?extension South Tasman Rise CampanianRegional uplift & Breakup volcanism & Breakup & stan of seafloor Gradual subsidence ?Opening of New Caledonia Paleocene erosion seafloor spreading in Coral spreading in Tasman Basin Basin; rifting & basaltic (60-80 Ma) Sea Basin & Cato Trough chron 33 (-80 Ma); ridge jump volcanism along west coast of (-65 Ma) to west of Dampier Ridge at New Zealand (65-75 Ma) chron 31 (-70Ma); ?compression & uplift in Otway (-75 Ma) PaleoceneFrom ?Early Eocene End seafloor spreading in End seafloor spreading in Increase in subsidence Tectonism ceased - erosion; Late Eocene oblique convergence Coral Sea Basin chron 24 Tasman Basin chron 24 (-54 in Early Eocene; Late Paleocene-Early Eocene developed to NE of (-54 Ma); early sag phase Ma); some Late Paleocene Paleocene-Early unconformity in New Caledonia PNG between deposition in continental inversion in Boobyalla Sub- Eocene unconformity Basin Australian & Pacific niargin basins; wrench basin & uplift/cooling from 50- & compression on N ?compression/reactivation in Plates reactivation in Cato Trough 60 Ma; Middle Eocene (-45- Lord Howe Rise; Late basin 49 Ma) change in pole and Eocene volcanism faster spreadine Late Eocene- Mid Oligocene & Mid-Late Eocene Compressional fault Subsidence; Early Convergence Miocene younger collision in compression in Capricorn reactivation in MioceneMiocene volcanism (collision/subduction) along PNG related to Basin - fault reactivation, Pliocene due intraplate stress Norfolk Ridge system; New terrane docking uplift & erosion; minor associated with collision along Caledonia - Late Eocene compressional OUgocene-Miocene northern Australian margin ophiolite obduction and foreland reactivation, reactivation in Townsville basin to west; New Zealand fold/thrust belt & Basin; Early Miocene obduction of foreland basin Northland Allochthon & middevelopment Late Miocene backarc spreading in Norfolk Basins L Triassic-M. Jurassic

Rifting & basic volcanism

Table 1. Late Mesozoic - Cainozoic tectonic event summary for the region off eastem Australia based on AGSO timescale. The various events are referenced in the text. Plume (hotspot), which is thought to have lay beneath the southwest margin of Lord Howe Rise, to the east of Bass Strait [34,35,36], leading up to breakup at about 80 Ma. A reduction in plate boundary forces associated with a global change in plate boundary configurations [14] in combination with mantles plumes, may have produced the extensive magmatic phase during the transition to the final stages of Gondwana breakup. The plumes themselves may haveultimately controlled the timing and/or location of breakup [14, 35, 37]. In the Cenomanian (-95 Ma), widespread tectonism of variable style occurred throughout the eastem Australian and New Zealand/New Caledonia region. In the northeast, it is represented by: regional uplift and compression in onshore basins [1]; folding in the Maryborough Basin [5]; the start of another syn-rift pulse in the Townsville Basin [19]; and the end of convergence in New Caledonia [23]. Further south,fissiontrack data indicate uplift and denudation in a zone parallel to the southeastem Australian margin at about 100-80 Ma [38], and on the margins of Bass Strait at 95-90 Ma [e.g. 31,39], Also, a brecciation and hydrothermal alteration event at 95 Ma has been noted on the southwest margin of Lord Howe Rise [40]. The Cenomanian tectonism was accompanied by: breakup and the start of slow NNW-SSE spreading between AustraUa and Antarctica (96 Ma; [15]), although others beheve that this may have commenced in pre-Valanginian time 531

[41]; major subsidence of the southern Otway Basin [22]; wrench-style basin development along the west Tasmanian margin and westem South Tasman Rise [42]; the start of NE-SW extension in the southeast Bass Basin (Boobyalla Subbasin; [21]); both extension [31] and inversion [21] in the Gippsland Basin; and the start of the main episode of rifting on the eastem South Tasman Rise [42] and along the east Australian margin and the margins of Lord Howe Rise [28,43]. The Cenomanian event is also documented by tectonism and magmatism on the South Island of New Zealand [24], and by widespread uplift and erosion leading to the start of peneplanation throughout New Zealand [44]. The widespread and variable effects of the Cenomanian event indicate that it is likely the product of major global plate adjustments. The adjustments resulted in slow NNW-SSE seafloor spreading south of Australia following southern margin rifting, and the NE-SW directed extension of the Tasman rift system, preceding Tasman Sea opening. Breakup of the Tasman-Coral Sea extensional terrane commenced in the Campanian (>80 Ma, chron 33) in the central Tasman Basin [10,11,45,46], although it may have been somewhat earher if initial spreading was very slow [15]. Off New South Wales, early seafloor spreading probably lay to the east of the Dampier Ridge, which has now been confirmed as a continental fragment [47]. A ridge jump to the west, at about 69 Ma (chron 31), started seafloor spreading in the northem Tasman Basin, with some margin segments initially having a significant strike-slip component of separation [10, 11]. Seafloor spreading adjacent to the Capricom Basin did not commence until around 63 Ma (chron 27, Early Paleocene), about the same time as in the Coral Sea Basin. The Coral/Tasman spreading ridges were presunoably connected by a series of transform/ridge segments [11] through the Cato Trough (Fig. 1), althou^ the spreading pattem has not yet been defined in this area. This single spreading system continued until --54 Ma (chron 24, earliest Eocene), when the entire ridge system became inactive. Although this was clearly the main phase of seafloor spreading, there is some evidence of older Gondwana breakup in the eastem-most part of the region: ?Albian (--102 Ma) oceanic cmst forms the ophiolites of the Northland Allochthon of New Zealand (K Herzer pers. comm.); the New Caledonian ophiolites (presumed older than 80100 Ma; [48]); the MORB-type basaltic province of the west coast of New Caledonia that probably formed in the Coniacian-Campanian (<89 Ma) during initiation of the New Caledonia Basin [49]; seismic evidencefi-omthe New Caledonia Basin [50]; and the regional considerations and reconstructions of Korsch & Wellman [44], which also indicate pre-Tasman Basin opening in the New Caledonia Basin, as well as the Bounty and Bellona Trougjis. Rifting (accompanied by magmatism) occurred along the west coast of New Zealand during and following Tasman Sea breakup [24]. This Campanian-Maastrichtian age rifting formed the West Coast and Taranaki Rift systems by oblique NNE-NE extension, which utilised the transfer faults of the earlier Albian extension [24]. This transtensional rift system may have linked divergence in the New Caledonia Basin to the Tasman Basin spreading ridge [24]. Following breakup, most of the rift basin systems off eastem Australia underwent sag-phase deposition; however, many of the basins were affected by significant episodes of compressional reactivation (Table 1), which were probably the local manifestation of global plate boundary adjustments and associated changes in intraplate stress. For example, probable Late Paleocene-Early Eocene compressional reactivation and inversion in the Cato Trough [19], on the northeast Lord Howe Rise [51], in the southem New Caledonia Basin [50], and in the Boobyalla Sub-basin of Bass Strait [31] is probably related to the plate boundary change that lead to the cessation of spreading along the Tasman-Coral Sea system in the earliest Eocene. Later tectonism was associated with the Middle Eocene 45Ma) pole change and increased spreading rate between Antarctica and Australia [e.g. 15]; and the Late Eocene-mid Oligocene commencement of convergent tectonism (collision and subduction) along the whole eastem margin of the east Australian plate from New Guinea, through New Caledonia to New Zealand. This latter episode marks a major change in the style of tectonism off eastem Australiafi-omdivergent to convergent, and resulted in foreland basin development and ophiolite obductionfi-omNew Guinea [52], through New Caledonia [49], to northem New Zealand [29]. A variety of tectonic events associated with the new convergent regime caused basin inversion and reactivation throu^ the Mesozoic extensional terrane off eastem Australia [see 31].

532

Figure 2. Major structural features of the Townsville Basin (after Struckmeyer et al. [19]). East Australian Mesozoic Rift System - Structural and Basin Fill Style It is beyond the scope of this paper to present a description of the tectonostratigraphic development of every element of the rift system. The following section attempts to summarise the style of structuring and basin fill for key segments of the system by focussing on the Townsville Basin (northeast province). Lord Howe Rise and its conjugate southeast Australian margin (central Tasman province), the Gippsland Basin (southeast Tasman province) and the South Tasman Rise (south Tasman province). Northeast province - Townsville Basin The northeastem Australian extensional terrane extends over a distance of about 2000 km from the Capricom Basin in the south to Papua New Guinea in the north, and prior to breakup would have been over 700 km in width (Fig. 1). The terrane is underlain by a northeastem extension of the Tasman Fold Belt [53, 54,55] that was modified during Mesozoic rifting to produce the present continental margin configuration of large marginal plateaus, such as the Queensland and Marion Plateaus, separated by bathymetric troughs (e.g. Townsville and Queensland Troughs). The Townsville Basin [19] is an oblique extensional basin underlying the Townsville Trough [18], an east-west trending bathymetric feature separating the Marion and Queensland Plateaus (Fig. 2). With the exception of several ODP holes which intersected Late Miocene to Recent sediments, there is no direct control on the stratigraphy of the Townsville Basin, and thus the timing of basin formation and filhng can only be interpreted within a regional tectonostratigraphic context. The maximum sediment thickness in the basin reaches approximately 6.5 km, and thefillcan be subdivided into two main seismic megasequences - a synrift and a sag-phase megasequence (Fig. 3). Underlying ?pre-rift or earhest synrift section may correspond to equivalents of the Jurassic-Early Cretaceous foreland to transtensional coastal basins such as the Maryborougji Basin, which formed in a convergent tectonic setting. The mainly Cretaceous synrift megasequence has a maximum thickness of up to --2 km and occurs in fault-controlled depocentres. It probably ranges from basal coarse terrigenous and volcaniclasticfiuviatilesediments to marginal marine clastic sediments at the top. The Tertiary sag-phase megasequence occurs as drape fill and reaches a thickness of up to -3.8 kmL The early sag-phase sequences are Paleocene-Eocene terrestrial shallow marine (shelf) clastics and ooze. Overlying sequences consist of terrigenous and calcareous sediments, including carbonate platforms, of Neogene age, and are separated from the underlying section by the ?mid-01igocene regional unconformity.

533

? pre-rift sediments Synrift sediments

Basement

10km

Figure 3. Line drawing of seismic profile across the western Lihou Sub-basin (see Fig. 2) of the Townsville Basin (after Struckmeyer et al. [19]). The structural style of the Townsville Basin is characterised by a half-graben morphology (Figs 2 & 3). The half-graben are bounded by major, often rotational, normal faults and are typically composed of a number of tilt blocks. Depth to basement, total sediment thickness, synrift isopach and gravity data all indicate that the basin is compartmentalised into distinct sub-basins by major NNW- to NW-trending transverse structural zones (Fig. 2). These transverse structures are associated with distinct changes in structural trend and style, and may be controlled by major pre-existing, crustal-scale, Tasman Fold Belt basement structures or 'hard links'. Such 'hard links' can strongly segment the developing rift by creating loci for fault relaying, polarity flips and reactivation, as described in the Timor Sea by O'Brien et al. [56]. Local thickening of late rift sediments in the opposite direction to that of the early rift sediments probably reflects at least two significant extensional episodes during basin formation. These rift events are separated by a period of uplift and erosion which may represent the regional Cenomanian (-95 Ma) plate adjustment (Table 1). The Townsville Basin formed part of a complex rift system that probably began to form in the ?Late Jurassic to Early Cretaceous. This system may have been initiated by transtension [18] in a largely convergent setting. However, the main phase of basin development occurred through oblique NW-NNW directed extension [18,19], which utihsed pre-existing Palaeozoic structural trends. Comparison with interpreted structural trends of the adjacent Queensland Basin (Queensland Trou^; [57]) supports the suggestion that formation of both basins was independent of the NE-SW divergence related to seafloor spreading in the Tasman and Coral Sea Basins. The overall form of the northeast Australian extensional terrane is controlled by variations in the partitioning of upper crustal and lower crustal/upper mantle extension throughout the province. The subsided 'basement' platforms beneath the marginal plateaus are associated with deep extension, whereas the rift troughs largely represent areas of both shallow and deep extension [18]. Central Tasman province - Lord Howe Rise and southeast Australian margin The central Tasman extensional terrane encompasses an enoraious area of complex seabed lying 300-1500 km east of Australia, and extending for over 2000 km from the latitude of New Caledonia to New Zealand. It includes relatively shallow water elongate plateaus and ridges such as the Lord Howe Rise (LHR), and the Norfolk, West Norfolk and Dampier Ridges, and intervening deeper water basins such as the Lord Howe, Middleton and New Caledonia Basins (Figs 1 & 4). The conjugate southeast Australian margin formed the westem edge of the central Tasman extensional terrane prior to Campanian breakup and seafloor spreading in the Tasman Basin (Fig. 4). The conjugate consists of a steep, narrow, largely basin-free margin in the nor&, which would have lay adjacent to the Dampier Ridge and the N-trending part of the LHR; and the complex basin-system of the Gippsland-Bass basins and South Tasman Rise to the south, which would have lay adjacent to the NNW-trending part of the LHR. Although major basin development is absent beneath the narrow part of the southeast Australian margin, small, mid-slope graben and half graben are present (Fig. 5; [58]). These contain unsampled ?Cretaceous syn-rift section which thickens into major westerly- (landward) dipping faults on the eastern side of the half-graben (Fig. 5). Major rift basin development occurs throughout the LHR/Norfolk Ridge region (Figs 1 & 4), but the synrift section has not been sampled in any of the DSDP holes throughout the region. However, dredging has sampled CampanianMaastrichtian oil-bearing shalefromthe southeastem Norfolk Ridge [29] and Cenomanian-Turonian coal measures from the West Norfolk Ridge [59]. The rift system beneath the westem LHR consists of a 200 km wide zone of NW-trending

534

TLG

Votcanics,

intrusions

Palaeozoic

basement

Top Latrobe Group

? B/u

Probable Tasman Margin breakup

COB

Continent-Ocean

boundary

unconformity 23/0A/602

Figure 4. Reconstructed schematic section across the Gippsland Basin and Lord Howe Rise showing the nature and distribution of the Cretaceous rift system adjacent to the Tasman Basin (after Symonds & Willcox, [43]). horst and graben structures in water depths of 1000-2000 m (Fig. 4; [50]). Elsewhere, particularly beneath the eastem third of the LHR, relatively thin sediments overlie basement, which is commonly planated. Individual graben are up to 50 km wide, several tens of kilometres long, and are best developed north of Lord Howe Island, where sediment fill is up to 4500 m in places [60]. Diapir-like structures have been recognised in several of the grabens (Fig. 4). The NNWtrending southem LHR/Challenger Plateau appears to be more volcanic than in the north, particularly its western flank north of the Bellona Trough (i.e. Monowai Spur area; see 50). The extensional basins appear to be less complex and are commonly deeply eroded [61]. Some horst blocks appear to contain dipping pre-rift strata of ?Mesozoic age [28]. The sedimentfillwithin the LHR basins is generally assumed to be of late Mesozoic age, but correlation with the older, convergent-related sedimentary basins of eastem Australia (e.g the Esk Trough) cannot be ruled out in all cases. The earliest fill in the rift basins is likely to be fault controlled terrestrial clastics - perhaps similar to the mid-Cretaceous Hawkes Crag Breccia and Late Cretaceous Pakawau Group coal measures of New Zealand. It seems Ukely that the Late Jurassic-Early Cretaceous Southem Rift System of Australia extended onto the southwestem LHR; however, much of the synrift section (lower Strzelecki Group equivalent) was probably eroded out or altered by later volcanism, and is now only recognisable at the base of isolated graben/half-graben. During the BarremianCenomanian the southem (NNE-trending) part of the LHR, particularly its southwest margin, was probably an extensional/transtensional magmatic province, and the -96 Ma rhyolites at DSDP Site 207 may represent the youngest phase of this volcanism. The magmatism was probably the result of rifting above a mantle plume, and was accompanied by underplating and uplift of the southem LHR resulting in erosion and intmsion of the Early Cretaceous rift basins and the pre-existing convergent terrane. The volcanic province shed volcanogenic sediment west into the Southem Rift System depositing the Eumarella and upper Strzelecki sequences throughout the Otway, Bass and Gippsland Basins. The shallower Late Cretaceous depositional environment at DSDP 207 on the southem LHR compared to the more northern DSDP Site 208 [62], may reflect a permanent positive buoyancy related to underplating in the south.

535

Figure 5. Line drawing of W-E seismic profile across the southern NSW margin SE of Jervis Bay. The narrow midslope half-graben is about 15 km wide - synrifl section shown by stipple. ModifiedfromColwell et al. [58]. The commencement of slow spreading began between Australia and Antarctica and NE-SW directed CenomanianCampanian extension in the Tasman Sea region (with associated left-lateral movement between Australia and LHR), produced the Boobyalla Sub-basin beneath southeast Bass Strait [21], and a broad but variable extensional terrane throughout the LHR-Norfolk Ridge-New Zealand region. It is interesting to note that there are significant similarities between the structural and seismic sequence characteristics of half-graben in the Boobyalla Sub-basin and beneath the southwestem LHR (Fig. 6). The structural differences between the southeast Australia and LHR margins have been accounted for by asymmetric Campanian breakup along the westem edge of the extension terrane leaving the main rift structures preserved to the east beneath the LHR [W]). Consequently, the LHR region is underpinned by extended lower crust and upper mantle, overlain by zones of upper cmstal extension and rift basin development, particularly beneath the westem part of the Rise (Fig. 4). In an attempt to integrate the variety of tectonic features across the Tasman Sea, Etheridge et al. [65] invoked a detachment model. The southeast Australia margin was interpreted to be an "upper plate" characterised by few rift structures, landward-dipping faults and long-lived uplift resulting in passive-margin mountains (Australia's Eastern Highlands) related to thermal buoyancy caused by rise of the asthenosphere as well as igneous underplating of mantlederived melts. The LHR forms a "lower-plate" margin characterised by extensive rift development on its westem side. Durroon

AGSO 88/306

5 km

AGSO 114/02

5km

Figure 6. A comparison of seismic datafromthe Boobyalla Sub-basin, Bass Strait (left) and southwestem Lord Howe Rise (right). TS - top of Aptian-Albian Otway Group sequence (upper Strzelecki equivalent); CA - Tasman breakup unconformity at top of Cenomanian-Campanian sequence; TL — Top of Campanian-Eocene Eastem View Coal Measures. Note the similar characteristics of the pre-TS sequence and the ?Cenomanian erosion event at its top.

536

BMR

90/2 90/7

90/15

BMR

82/3

68/14

WNW 3-

ESE

H BASEMENT RIDGE

23/0V/163

40 km Vertical exaggeration approx 5 at seafloor

GIPPSLAND BASIN

Santonian & Campanian dredge hauls (Marshall, 1988 & 1990)

TERTIARY -QUATERNARY

Seaspray Group (Lakes Entrance Fm & Gippsland Lst)

RFS

Rosedale Fault System

L CRETACEOUS - E TERTIARY

Latrobe Group

FFS

Foster Fault System

L CRETACEOUS

Golden Beach Group

Top Latrobe

?L JURASSIC - E CRETACEOUS

Strzelecki Group

Con tinen tal basemen t Well-layered

crust

TASMAN BASIN L CRETACEOUS -QUATERNARY

Tasman Basin sediments

Late Palaeocene

Maastrichtian

Campanian, Top Golden Beach

Top Strzelecki

Intra Strzelecki

Top basement

COB

Continental/ocean Basin-forming

Oceanic

basement

Prominent

boundary

detachment

crustalreflectors

Figure 7. Line drawing of seismic "strike" line down the axis of the Gippsland Basin across the continent-ocean boundary into the Tasman Basin (after Willcox et al. [21]). Although such a model can ejq)lain the general relationship between the southeast Australian margin and its northem LHR conjugate, it does not explain the more poorly developed rift basins with generally east-dipping faults and more deeply eroded/planated nature of the southem LHR. However, this can be accounted for if the upper/lower plate characteristics switch at the Gippsland Basin such that the Gippsland to South Tasman Rise rifted margin becomes the lower plate, and its southem LHR conjugate becomes the upper plate. This e?q)lanation also fits with the change in character and trend of the Eastem Highlands to the north of Bass Strait. Southeast Tasman province - Gippsland Basin The Gippsland Basin, which lies at the eastem end of Bass Strait (Fig. 1), has been Australia's major oil-producing province for 30 years. It formed at the confluence of the Southem Rift System and the Tasman rift system, with sedimentation beginning in the Late Jurassic in a narrow rift related to incipient breakup along the southem margin, and continued through a Late Cretaceous period of rifting associated with the opening of the Tasman Sea, and Tertiary postbreakup subsidence. The Gippsland Basin contains up to 16 km of sediments (Fig. 7) in an ESE-trending depocentre bounded on its north and northwestem margins by a detachment ramp, and on its southem side by a relatively linear, ?listric fault system [21]. The basin, together with the adjacent Bass and Otway basins, is thought to have formed part of a linked, oblique extensional/transtensional system near the eastem end of the Southem Rift System. The earliest extensional phase phase may have had a NW-SE sense [21], although there is still disagreement about this with others favouring a more N-S to 537

NE-SW extensional direction [22]. The sediments in the basin can be divided into five major tectono-stratigraphic units: the ?Late Jurassic - Early Cretaceous non-marine, syn-rift (southem margin rifting) lower Strzelecki Group; the AptianAlbian volcanogenic, late-rift upper Strzelecki Group; the Late Cretaceous non-marine, syn-rift (Tasman rifting) Golden Beach Group; the Late Cretaceous-Eocene mainly non-marine, sag-Latrobe Group (the main petroleum producer in the basin); and 4 e overlying, marine, Oligocene and younger Seaspray Group. Only Seaspray Group and the upper part of the Latrobe Group are well knownfi:omdrilling in the basin. The geometry of units and the relationship of the eastem part of the Gippsland Basin to the adjacent Tasmian Basin are illustrated in Figure 7. The early syn-rift sediments of the lower Strzelecki Group (below sequence boundary IS' in Fig. 7) appear to thin to the west, whereas the volcanogenic upper Strzelecki thickens and onlaps a large, high-standing basement block beneath the outer shelf. Further east, below the continental slope and rise, the upper Strzelecki section appears to change seismic character, possibly representing an increase in volcaniclastics and lava flows. This section onlaps a broad basement ridge lying adjacent to the continent/ocean boundary. The basement high beneath the shelf-break (Fig. 7) may be a southem continuation of the granitic blocks of the southem Dampier Ridge [21]. The outer basement ridge is probably a volcanic remnant of the southwest LHR magmatic province that sourced the Aptian-Albian volcanogenic sediment of the southeast Australian basins. In the Gippsland Basin, the mid-Cretaceous (Cenomanian) uplift and compressional event is expressed by high-angle faulting extending to the top Strzelecki level erosional surface; and by apparent overthmsting of the lower Strzelecki Group and the underlying ramp surface towards the west (Fig. 7). Later adjustments and reactivation, largely in response to Tasman Basin rifting [21] and to younger, far-field intraplate stress changes [31], produced wrench- and compressional-related inversion structures which form the major petroleum traps (usually at top Latrobe level) in the Gippsland Basin. South Tasman province - South Tasman Rise The most southerly part of the Mesozoic rift system off eastem Australia occurs south of Tasmania on the South Tasman Rise (STR; Fig. 1). The structural setting of the rift basins on the Tasmanian margin and beneath the East Tasman Plateau (ETP) and STR is still debated. It has been known for some time from DSDP drilling [66] and dredging that the STR is a Palaeozoic continental fragment. Recent dredging on the margins of the ETP [67] has also proved its continental

Figure 8. Stmcture and sediment thickness map for offshore west Tasmania and the South Tasman Rise (after Hill et al. [42]).

538

^ 500 r

[;.•.•/] Late Oligocene and Neogene (largely carbonate) Late Eocene — earliest Oligocene (largely terrigenous) Cretaceous ? Early Cretaceous Intrusives and rift volcanics Palaeozoic basement

4500 Figure 9. Line drawing of seismic profile across the central South Tasman Rise. Also shows the withfi-eeair gravity profile (after Willcox et al. [69]). origins, although it has been modified by volcanism associated with the central Soela Seamount, which formed on the Balleny Plume trail [35]. Willcox & Stagg [20] suggested that the whole of the STR may have lay on the Antarctic plate adjacent to the Otway/Sorell Basins during the Late Jurassic-Early Cretaceous, southem margin syn-rift phase, and perhaps during the initial slow spreading phase. It became attached to the Australian plate during the rapid N-S spreading beginning in the Middle Eocene. More recently it has been suggested that only the westem, N-S trending part of the STR (Fig. 8) moved in this way, with the eastem province remaining approximately in its present location, being essentially a part of the Tasmanian Palaeozoic block [68]. The STR consists of a triangular core of Palaeozoic basement,flankedby rift basins containing up to 6 km of sediments (Fig. 8). Although there is no direct evidence for their age, the seismic sequences in these basins appear similar in character to those of the Otway Basin [69]. The westem basin terrane lies between two major north-trending strike-slip fault systems, and is characterised by rotated basement blocks and intervening E-W trending basins (Figs 8 & 9). Although it is likely that Late Jurassic-Early Cretaceous syn-rift section of the Southem Rift System underlies this terrane, its structural style is clearly transtensional associated with lateralmovement between Antarctica and Austraha until the continents cleared in the Early Oligocene [42]. A similar mechanism was proposed for the origin for the Sorell Basin depocentres (Figs 1 & 8) off west Tasmania [70]. The eastem basin terrane has a major NW-trending boundary fault and WNW-trending normal faults and basins (Figs 8 & 9). This has been interpreted as a largely extensional basin province [69], and its main phase of structuring was probably during NE-SW Cenomanian-Campanian extension associated wift the Tasman rift system. Similar rift basins may be present beneath the margins of the ETP. Conclusions The Mesozoic rift basin system along eastem Australia extends for a distance of about 5000 kmfiromthe Gulf of Papua in the north to the southem end of the South Tasman Rise in the south, and was probably up to 1000 km in width prior to continental breakup and seafloor spreading. It developed in response to two major extensional episodes within eastem Gondwana that formed nearly perpendicular to each other, and in quite different regional tectonic settings. The rifting followed a prolonged period of convergent tectonism that ended in the Middle-Late Triassic in eastem Australia. The Southem Margin Rift system began to fomi in the Late Jurassic-Early Cretaceous by oblique extension along a trend between Australia and Antarctica. Concurrent passive margin development to the north in New Guinea may have been linked to the Southem Rift System and/or the proto-Pacific plate boundary via a strike-slip fault system along the northeast AustraUan margin, which may have included subsidiary fault systems and transtensional basin development along the site of the present Queensland-Townsville-Capricom basin system. A change fi-om compression to extension in eastem Gondwana was marked by the development of an extensive Barremian-Cenomanian (120-95 Ma) intraplate magmatic province along eastem Australia. An abrupt change fi-om convergence to extension in New Zealand in the Albian (110-105 Ma), in combination with other evidence, suggest that this magmatism was not arc-related, and probably had an extensional/transtensional setting. A widespread Cenomanian uplift and compression event, presumably related to a major global plate adjustment, was followed by slow seafloor spreading between Australia and Antarctica and a second major extensional episode in the Cenomanian-Campanian. Rifting occurred throu^out the whole of the eastem Australia plate, forming the Tasman-Coral sea extensional terrane. The divergent tectonism became inactive in the Eocene, and by the mid-Oligocene the whole region had again reverted to a largely convergent regime. 539

The overall fonii of the east Australian Mesozoic extensional terrane was controlled by a combination of old pre-existing structural trends, the extensional transport direction, and variations in the partitioning of upper crustal and lower crustal/upper mantle extension throughout the region. ITie subsided 'basement' platforms of the marginal plateaus represent areas of deep extension, the rift troughs are areas of both deep and shallow extension, and the shelf rift basins are areas of shallow extension. Along-strike variations in both the extensional style and the locus of breakup, have produced a complex present-day arrangement of marginal plateaus and rift troughs off northeast Australia, and a narrow, largely basin-free margin in the southeast, where much of the rift system was probably detachedfromAustraUa and now lies beneath the margins of LHR and the Norfolk Ridge system. Most of the rift basins of the east Australian extensional terrane are poorly known as they lie in deep water (500-2000 m) beyond the frontier of conventional hydrocarbon exploration. However, their variety of structural styles and considerable thicknesses of sediment offer long-term opportunities for petroleum discovery. Acknowledgements We acknowledge the help of many of our colleagues in AGSO's Continental Margins Program (CMP) whose work over many years has contributed to the evolution of our ideas on the development of the region off eastem Australia. We also thank R.J. Korsch (AGSO), and R. Herzer and R. Wood (IGNS, New Zealand) for usefiil recent discussions. We gratefiilly acknowledge the seamen, and data acquisition and processing staff associated with Rig Seismic, without whose efforts many of the areas discussed in this paper would remain largely unsurveyed. We thank the Cartographic Services Unit, AGSO, for drafting the figures. Published with permission of the Executive Director, AGSO. References 1. Korsch, R.J. & Totterdell, J.M., 1996. Mesozoic deforaiational events in eastem Australia and their impact on onshore sedimentary basins. Mesozoic Geology of the Eastem Australia Plate, Ext. Abstract, this volume. 2. Bradshaw, J.D., 1989. Cretaceous geotectonic pattems in the New Zealand region. Tectonics, 8, 803-820. 3. Harrington, H.J. & Korsch, R.J., 1985b. Late Permian to Cainozoic tectonics of the New England Orogen. Australian Journal of Earth Sciences, 32,181-203. 4. Murray, C.G., Fergusson, C.L., Hood, RG., Whitaker, W.G. & Korsch, R.J., 1987. Plate tectonic model for the Carboniferous evolution of the New England Fold Belt. Australian Joumal of Earth Sciences, 34 (2), 213-236. 5. Hill, P.J., 1994. Geology and geophysics of the offshore Maryborough, Capricom and northem Tasman Basins: Results of AGSO Survey 91. Australian Geological Survey Organisation, Record 1994/1. 6. Harrington, H.J. & Korsch, R.J., 1985. Deformation associated with accretion of the Gympie terrane in eastem Austraha. Geological Society of Australia, Abstracts, 14,104-108. 7. Struckmeyer, H.I.M., Yeung, M. & Pigram, C.J., 1993. Mesozoic to Cainozoic plate tectonic and palaeogeographic evolution of the New Guinea region. In: Carman, G.J. & Z. (Editors), Petroleum E]q)loration and Development in Papua New Guinea. Second PNG Petroleum Convention, Port Moresby, 31st May - 2nd June 1993, Proceedings, 261-290. 8. Hergt, J.M., Chappell, B.W., Faure, G. & Mensing, T.M., 1989. The geochemistry of Jurassic doleritesfromPortal Peak, Antarctica. Contributions to Mineralogy and Petrology, 102,298-305. 9. Symonds. P.a. & Colwell, J.B., 1992. Geologicalframeworkof the southern Lord Howe RiseAVest Norfolk Ridge region-cruise proposal. Australian Geological Survey Organisation, Record 1992/97,44pp. 10. Shaw, R.D., 1978. Seafloor spreading in the Tasman Sea: a Lord Howe Rise - eastem Australia reconstruction. Australian Society of Exploration Geophysicists, 9 (3), 75-81. 11. Shaw, R.D.,1979. On the evolution of the Tasman Sea & adjacent margins. PhD Thesis University of Sydney (unpub.). 12. Scheibner, E., sato, T. & Craddock, C., 1991. Tectonic Map of the Circum-Pacific region, Southwest Quadrant, 1:10,000,000. Circum-Pacific Council for Energy and Mineral Resources, Houston. 13. Elliot, DJH., 1992. Jurassic magmatism and tectonism associated with Gondwanaland break-up: an Antarctic perspective. In: Storey, B.C., Alabaster, T. & Pankhurst, R.J. (Eds), Magmatism and the causes of continental break-up. Geological Society Special Publication, 68,165-184. 14. Storey, B.C., 1995. The role of mantle plumes in continental breakup: case historiesfromGondwanaland. Nature, 377, 301-308. 15. Veevers, J.L, Powell, C.McA. & Roots, S.R., 1991. Review of seafloor spreading around Australia. L Synthesis of the pattems of spreading. Australian Joumal of Earth Sciences, 38, 373-389. 16. Jones, J.G. & Veevers, J.J., 1983. Mesozoic origins and antecedents of Australia's Eastem Highlands. Joumal of the Geological Society of Australia, 30, 305-322. 17. Ewart, A., Schon, R.W. & Chappell, B.W., 1992. The Cretaceous volcanic-plutonic province of the central Queensland (Austraha) coast - a rift related 'calc-alkaline' province. Transactions of the Royal Society of Edinburgh: Earth Sciences, 83, 327-345. 18. Symonds, P.A., Capon, D., Davies, P.J., Pigram, C.J. & Feary, 1987. Structural style of the Townsville Trougji and its implications for the development of the northeast Australian margin. BMR Research Symposium: Applied Extension Tectonics, Bureau of Mineral Resources, Geology & Geophysics, Record 1987/51,165-172. 540

19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.

Struckmeyer, H.I.M., Symonds, P.A., Fellows, M.E. & Scott, D.L., 1994. Structural and stratigraphic evolution of the Townsville Basin, Townsville Trough, offshore northeastern AustraUa. Australian Geological Survey Organisation, Record 1994/50. Willcox, J.B. & Stagg, H.M.J., 1990 - Australia's southern margin: a product of oblique extension. Tectonophysics, 173,269-281. Willcox, J.B., Colwell, J.B. & Constantine, A.E., 1992 - New ideas on Gippsland Basin regional tectonics. In Energy, Economics and Environment Gippsland Basin Symposium, Joint PESA/AusIMM Symposium, 22-23 June, 1992, Melbourne, Australasian Institute of Mining and Metallurgy, 93-110. Hill, KA., Finlayson, D.M., Hill, K.C. & Cooper, G.T., 1995. Mesozoic tectonics of the Otway Basin region: the legacy of Gondwana and the active Pacific margin - a review and ongoing research. The APEA Journal, 35,467-493. Paris, J-P., 1981. Geologic de la Nouvelle-Caledonie: un essai de synthese. Bureau de Recherches Geologiques et Minieres, Memoire 113. Laird, M.G., 1994. Geological aspects of the opening of the Tasman Sea. In: Van der Lingen, G.J., Swanson, K.M. & Muir, R.J. (Eds), Evolution of the Tasman Basin, AA. Balkema, Rotterdam, 1-17. Tulloch, A.J. & Kimbrough, D.L., 1989. The Paparoa metamorphic core coiiq)lex. New Zealand: Cretaceous extension associated with fragmentation of the pacific margin of Gondwana. Tectonics, 8(6), 1217-1234. Uruski, C. & Wood, R., 1991. A new look at the New Caledonia Basin, an extension of the Taranaki Basin, offshore North Island, New Zealand. Marine and Petroleum geology, 8, 379- 391. Eade, J.V., 1988. The Norfolk Ridge system and its margins. In: Naim, A.E.M., Stehli E.G. & Uyeda, S.(Eds), The Ocean Basins and Margins, Vol. 7B, The Pacific Ocean, Plenun Press, N.Y. and London, 303-322. Symonds, P.A., and Willcox, J.B., 1989. Australia's petroleum potential beyond an Exclusive Economic Zone. BMR Joumal of Australian Geology and Geophysics, 11,11-36. Herzer, R.H. & Mascle, J., in press. Anatomy of a continent-backarc transfomi - the Vening Meinesz fracture Zone northwest of New Zealand. Marine Geophysical Researches. Gleadow, A.J.W. & Duddy, I.R., 1981. Early Cretaceous volcanism and the early breakup history of southeastem Austraha: Evidence from fission track dating of volcaniclastic sediments. In: Cresswell, M.M. & Vella, P. (Eds), Fifth International Gondwana Symposium, WelUngton, New Zealand, A A. Balkema, Rotterdam, 295-300. Hill, K.C.,ffill,KA., Cooper, G.T., 0'Sullivan, A.J., O'SuUivan, RB. & Richardson, M.J., 1995. Inversion around the Bass Basin, SE Australia. In: Buchanan, J.G. & Buchanan, P.G. (Eds), Basin inversion. Geological society Special PubUcation, 88, 525-547. Bryan, S., Constantine, A., Stephens, C., Ewart, T, Schon, R. & Parianos, J., 1996. The Whitsunday Volcanic Province (central Queensland) and the Gippsland/Otway basins (Victoria): a conparison of Early Cretaceous riftrelated volcano-sedimentary successions. Mesozoic Geology of the Eastem Australia Plate, Extended Abstracts, this volume. Mc Dougall, I. & van der Lingen, G.J., 1974. Age of rhyolites on the Lord Howe Rise and the evolution of the southwest Pacific Ocean. Earth and Planetaiy Science Letters, 21,117-126. Duncan, RA. & McDougall, L, 1989. Volcanic time-space relationships. In: Johnson, R.W. (Ed), Intraplate Volcanism in Eastem Australia and New Zealand , Cambridge University Press, Cambridge, 43-54. Lanyon, R., Vame, R. & Crawford, A.J., 1993. Tasmania Tertiary basalts, the Balleny plume, and opening of the Tasman Sea (southwest Pacific Ocean). Geology, 21, 555-558. Sutherland, F.L.,1994. Tasman Sea evolution and hotspot trails. In: Van der Lingen, G.J., Swanson, KLM. & Muir, R.J. (Eds), Evolution of the Tasman Basin, A.A. Balkema, Rotterdam, 35-51. Weaver, S.D., Storey, B.C., Pankhurst, R.J., Mukasa, S.B., DiVenere, VJ. & Bradshaw, J.D., 1994. Antarctic-New Zealand rifting and Marie Byrd land lithosphere magmatism Unked to ridge subduction and mantle plume activity. Geology, 22, 811-814. Moore, E.M., Gleadow, A.J.W. & Lovering, Balleny., 1986. Thermal evolution of rifled continental margins: new evidence from fission tracks in basement apatites from southeastem AustraUa. Earth and Planetary Science Letters, 78,255-270. Foster, DA. & Gleadow, A.J.W., 1992. Reactivated tectonic boundaries and implications for the reconstrction of of southeastem Australia and northem Victoria land, Antractica. Geology, 13,471^88. Tulloch, A.J., Kimbrough, D.L. & Wood, RA., 1991. Carboniferous granite basement dredged from a site on the southwest margin of the Challenger Plateau, Tasman Sea. New Zealand Joumal of Geology and Geophysics, 34, 121-126.

Stagg, H.M.J. & Willcox, J.B., 1992 - A case for Australia-Antarctica separation in the Neocomian (ca. 125 Ma). Tectonophysics, 210,21-32. Hill, P.J., Exon, N.E & Royer, J.Y., 1995. Swath-mapping the Australian continental margin: resultsfromoffshore Tasmania. E^q)loration Geophysics, 26,403-411. Symonds, P.A. & Willcox, J.B., 1993. Deep water petroleum prospects in the Lord Howe Rise region. In: Second AGSO Petroleum group Seminar, 17-19 November 1993, Key Abstracts, AGSO Record 1993/83, 12-13. Korsch, R.J. & WeUman, H.W., 1988. The geological evolution of New Zealand and the New Zealand region. In: Naim, A.E.M., Stehli, EG. & Uyeda, S. (Eds), The Ocean Basins and Margins, Vol. 7B, Plenum, New York, 411^82. 541

45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70.

Hayes, D.E., and Ringis, J., 1973. Seafloor spreading in the Tasman Sea. Nature, 243,454-58. Weissel, J.K., and Hayes, D.E., 1977. Evolution of the Tasman Sea reappraised. Earth and Planetary Science Letters, 36, 77-84. Symonds, P.A., Willcox, J.B. & Kudrass, H.R., 1988. Dampier Ridge in the Tasman Sea: a continental fragment. Geological Society of Australia Abstracts, 21, 393-394. CoUot, J.Y., Rigolot, P. & Missegue, F., 1988. Geologic structure of the northem New Caledonia Ridge, as inferred from magnetic and gravity anomalies. Tectonics, 7(5), 991-1013. Kroenke, L.W., 1984. Cenozoic tectonic development of the southwest Pacific. United Nations Economic and Social Commission for Asia and the Pacific, Committee for Co-ordination of Joint Prospecting for Mineral Resources in the South Pacific Offshore areas. Technical Bulletin, 6,122p. Willcox, J.B., Symonds, PA., Hinz, K. & Bennett, D., 1980 - Lord Howe Rise, Tasman Sea - preliminary geophysical results and petroleum prospects. BMR Journal of Australian Geology and Geophysics, 5,225- 36. Lafoy, Y, Pelletier, B., Auzende, J-M, Missegue, F. & Mollard, L., 1994. Tectonique compressive cmozoique sur les rides de Fairway et Lord Howe, entre Nouvelle-Caledonie et Australie. C.R. Acad. Sci. Paris, t.319, serie n, 1063-1069. Pigram, C.J. & Symonds, P.A., 1991. A review of the timing of the major tectonic events in the New Guinea Orogen. Joumal of Southeast Asian Earth sciences, 6(3/4), 307-318. Taylor, L.W.H. & Falvey, D., 1977. Queensland Plateau and Coral Sea Basin: stratigraphy, structure and tectonics. Tte APEA Joumal, 17(1), 13-29. Mutter, J.C. & Kamer, G.D., 1980. The continental margin off northeast Australia. In: Henderson, RA. & Stephenson, P.J. (Eds), The geology and geophysics of northeast Australia. Geological Society of Australia (Queensland Division), Brisbane, 47-49. Symonds, P.A., Fritsch, J., Schluter, H.U., 1984. Continental margin around the westem Coral Sea Basin: structural elements, seismic sequences, and petroleum geological aspects. In: Watson, S.T. (Ed), 3rd CircumPacific Energy and Mineral Resources Conference, Hawaii, Transactions, 243-252, American Association of Petroleum Geologists, Tulsa. O'Brien, G.W., Higgins, R., Symonds, P., Quaife, P., Colwell, J. & Blevin, J., 1996 - Basement control on the development of extensional systems in Australia's Timor Sea: an example of hybrid hard linked/soft linked faulting? APPEA Joumal, 36 (1), 161-200. Scott, D.L., 1993. Architecture of the Queensland Trough: implications for the structure and tectonics of the northeastem Australian margin. AGSO Joumal of Australian Geology & Geophysics, 14,21-34. Colwell, J.B., CoflBn, M.F. & Spencer, R.A., 1993. Structure of the southern New South Wales continental margin, southeastem Australia. BMR joumal of Australian Geology & Geophysics, 13, 333-343. Zhu, H. & Symonds, P.A., 1995. Seismic interpretation, gravity modelling and petroleum potential of the southem Lord Howe Rise region. In: 1994 New Zealand Petroleum Conference Proceedings, Ministry of Commerce, WeUington, 223-230. Roeser, H. and Shipboard Party, 1985. Geophysical, geological and geochemical studies on Lord Howe Rise. Bundesanstalt fur Geowissenschaflen und Rohstoffe Report, Cruise S036 (1). Wood, RA., 1993. The Challenger Plateau. In: Ballance, RF. (Ed), South Pacific Sedimentary Basins, Sedimentary basins of the World, 2, Elsevier, Amsterdam, 351-364. Bums, R.E., Andrews, J.E., and others, 1973. Initial reports of the Deep Sea Drilling Project, v.21, U.S. Government Printing Office, Washington, D.C.. Walley, A.M., 1992. Cretaceous-Cainozoic palaeogeography of the New Zealand- New Caledonia region. Bureau of Mineral Resources, Geology & Geophysics, Record 1992/11. Jongsma, D. & Mutter, J.C., 1978. Non-axial breaching of a rift valley: evidencefiromthe Lord Howe Rise and the southeastem Australian margin. Earth and Planetary Science Letters, 39, 226-234. Etheridge, M.A., Symonds, P.A. & Lister, G.S., 1990 - Application of detachment models to reconstruction of conjugate passive margins. In: Tankard, A.J. & Balkwill, H.R. (Eds), Extensional Tectonics and Stratigraphy of the North Atlantic margins, AAPG Memoir, 46, 23^0. Kennett, J.P., Houtz, R.E. & others, 1975. Initital Reports of the Deep Sea Drilling Project, vol. 29. Washington, D.C.: U.S. Government Printing OflSce. Exon, N.F., Marshall, J.F. & others, 1995. AGSO Cruise 147 report - Tasman Rises geological sampling cruise of Rig Seismic: stratigraphy, tectonic history and palaeoclimate of the offshore Tasmanian region. Austrahan Geological Survey Organisation, Record 1995/56. Rollet, N., Royer, J.-Y, Exon, N.E & Hill, P. J., in press. Collage de deux fragments du Gondwana Oriental au sein du Plateau Sud-Tasman (sud de la Tasmanie). Comptes Rendu de TAcademic des Sciences (Oceanographie/Geophysiques Marine). Willcox, J.B., Baillie, P., Exon, N.E, Lee, C.S. & Thomas, B., 1989. The geology of westem Tasmania and its continental margin - with particular reference to petroleum potential. Bureau of Mineral Resources, Australia, Record 1989/13. Moore, A.M.G., Willcox, J.B., Exon, N.F. & O'Brien, G.W., 1992. Continental shelf basins on the west Tasmanian margin. APEA Joumal, 32 (1), 231-250. 542

GEOCHRONOLOGICAL EVIDENCE FOR THE PRESERVATION OF CRETACEOUS WEATHERING PROFILES IN NORTHWESTERN QUEENSLAND Paulo Vasconcelos Department of Earth Sciences, University of Queensland, Brisbane, Qld 4072 Australia The history of weathering in the Australian continent during the Mesozoic and Cenozoic suggests a polycycUc evolution, with alternating weathering-prone and erosion-prone conditions (King, 1950; Twidale, 1956; Watkins, 1966; Quilty, 1977; Grimes, 1979; Fairbridge and Finkl, 1980; Oilier, 1992). King (1950) and Twidale (1956) identified several weathering surfaces in Australia and estabhshed their chronological evolution; this approach was subsequently extended by Grimes (1979) in Queensland. Until recently, chronology in weathering and landscape evolution was established from field observations, and land surfaces were recognised based on their their stratigraphic and topographic relationships, their soils, their surface morphology, and the nature of their duricmst or weathering profiles (Grimes, 1979). This stratigraphic approach to landscape evolution was very useful in the past, when other means of measuring absolute and relative ages of soils, weathering profiles, and landforms were not available. However, thesefieldbased geochronological studies have become insufficient to address current problems in geomorphology and geochemistry. Dating of volcanic rocks overlying weathered sequences in eastem Australia has been used to delimit the ages of the weathering profiles (Wyatt and Webb, 1970; Wellman and McDougall, 1974). The application of paleomagnetic dating to laterite studies (Idnum and Senior, 1978) added some quantitative constraint on the ages of selected Queensland weathering surfaces. Elsewhere in Australia, K-Ar dating of supergene alunite has been successfully used to further delimit the ages of Cenozoic weathering processes (Bird et al., 1990). Despite these advances, absolute ages of the weathering surfaces and the chronology of weathering reactions in Queensland weathering profiles has remained largely undetermined. Recently, K-Ar and analyses of samples derived from different horizons in weathering profiles were used to date the advance of weathering fronts in Queensland. The K-Ar analysis of supergene K-bearing manganese oxide samples from the Overhang Mn Deposit, near Cloncurry, Northwestem (^eensland has yielded the following results: K-Ar Results: Mineral

Sample

K20(AA) Wt %

Weight g

^"Ar* cc/g

ClonOl ClonOS Clon04

0.53 2.14 0.82

0.40360 0.50500 0.90650

1.373E-06 5.296E-06 2.163E-06

Hollandite

31.6 72.0 53.5

Age Ma

±lo Ma

80 77 82

6 1 2

Several samples from the same weathering profile were also analysed by the 40Ar/39Ar method, but only one sample yielded a well defined plateau. The results are listed below: "Ar/"Ar Results: Sample 10011-02A CLON94-05cA CLON94-05cB CLON94-05cC CLON94-05cD CLON94-05cE CLON94-05cF CLON94-05cG CLON94-05cH CLON94-O5cl CLON94-O5cj CLON94-05cK Plateau Age Integrated Age

*Ar Moles

^•Ar Moles

«Ar* Moles

%Rad

Age Ma

±la Ma

4.10E-14 5.06E-14 2.17E-14 1.69E-14 2.37E-14 3.00E-14 2.81E-14 1.08E-14 1.44E-15 9.45E-16 -9.46E-18 64.7 63

6.36E-17 3.29E-16 4.91E-16 6.77E-16 1.05E-15 1.57E-15 1.56E-15 6.26E-16 8.36E-17 4.28E-17 1.40E-18 ±0.5 Ma ±2 Ma

5.62E-16 2.81E-16 3.72E-16 4.85E-16 7.63E-16 1.17E-15 1.55E-15 2.88E-15 1.74E-15 6.61E-16 O.OOE+00

9.3 5.1 32.2 53.5 65 76 79.9 71.8 67.8 26.8 %1854.6

254.81 35.31 63.65 59.57 65.69 64.96 64.08 55.75 52.31 26.76 -686.77

81.76 6.93 2.07 1.17 0.90 0.51 0.50 0.91 5.48 10.75 587.93

The discrepancy between the K-Ar and the ""Ar/^'Ar results is attributed to the presence of several generations of Mn-oxides, identified petrographically, in the samples analysed. Additional high-resolution ^Ar/^^Ar analysis of each of the generations of manganese oxide present in the samples is necessary to unravel the complete history of weathering for the region. 543

Despite the apparent discrepancy between the K-Ar and the ^ÂiPÂi analyses, dating manganese oxide samples formed in the weathering profiles suggest that some weathering profiles in Northwestem Queensland were already exposed to surface weathering conditions at the end of the Mesozoic. It is difficult at present to determine the maximum ages of the weathering profiles due to the scarcity of dates. Further appUcation of K-Ar and Ar/Ar dating of weathering minerals will certainly improve the database available for evaluating the role of Mesozoic weathering on the evolution of weathering profiles preserved in the Queensland landscape today. References Bird M., Chivas A. R. and McDougall 1. (1990) An isotopic study of surficial alunite in Australia: 2. Potassium-argon geochronology. Chem, Geol 80, 133-145. Day R. W., Whitaker W. G., Murray C. G., Wilson I. H. and Grimes K. G. (1983) Queensland Geology. Geological Survey of Queensland PubUcation 383. Brisbane, Geological Survey of (êensland, 194 p. Fairbridge R. and Finkl C. (1980) Cratonic Erosional Unconformities and Peneplains: J. of Geology 88, 69-86. Grimes K. (1979) The stratigraphic sequence of old land surfaces in northem Queensland. BMR1 Aust. Geol Geopk, 4,33^6. Idnurm M. and Senior B. R. (1978) Palaeomagnetic ages of late Cretaceous and Tertiary weathered profiles in the Eromanga Basin, Queensland. Palaeogeogn, Palaeoclimatol, Palaeoecol 24, 263-277. King L. C. (1950) The cyclic landsurfaces of Australia. Proc. Roy Soc. Victoria, 62, 79-95. Oilier C. D. (1992) Global change and long-term geomorphology. Terra Nova 4, 312-319. Twidale C. R. (1956) Chronology of denudation in northwest Queensland. Geol. Soc. America Bull 67, 867-882. Watkins J. R. (1966) The relationship between cUmate and the development of landforms in the Cainozoic rocks of Queensland. J. geol Soc. Aust, 14, 153-168. Wellman P. and McDougall I. (1974) K/Ar ages in the Cainozoic volcanic rocks of new South Wales. J. Geol. Soc. Aust., 15,247-272. Wyatt D.H. and Webb A.W. (1970) K/Ar ages of some north Queensland basalts and an interpretation of the late Cainozoic history. J. Geol. Soc. Aust., 17,39-51.

544

THE CRYSTAL CHEMISTRY AND GENESIS OF CHRYSOPRASE Paulo VasconcelosS Balbir Singh^ 1 Department of Earth Sciences, University of Queensland, Brisbane, Qld 4072 Australia 2 Centre for Microscopy and Microanalyses, University of Queensland, Brisbane, Qld 4072 Australia Abstract Spectroscopic, X-ray diffraction, and high resolution microscopic investigation of chrysoprase samples from the Gumigil Mine, Marlborough, Queensland, indicates that this apple green variety of quartz owes its colour to platelets of Ni-talc, willemseite, which are intergrown with cryptocrystalline supergene silica. When the intergrown willemseite platelets are randomly distributed throughout the sample, the chrysoprase colour is homogeneous and stable. Electron microprobe and X-ray fluorescence (XRF) analyses indicate that stable colour chrysoprase contains between 1.5 and 2.5 wt% NiO. Fourier transform infrared spectroscopy (FUR) and the XRF results also indicate that stable colour chrysoprase contains less than 1 wt% structural and molecular H2O. If the Ni-content of chrysoprase exceeds 4 wt% NiO, the willemseite platelets form large booklets and randomly oriented masses of Ni-silicates intergrown with cryptocrystalline silica. High Ni-contents (> 4 wt% NiO) correlates with relatively high water contents C 1.8 wt%) and with unstable colour centres. These samples undergo dehydration a few hours to a few days after mining, changing from a vitreous to a dull lustre, becoming useless for omamental purposes. In addition to chrysoprase, the Gumigil weathering profiles also host apple green and orange yellow varieties of opal. The apple green variety (prase opal) may have the exactly same colour as chrysoprase (as determined by visual comparison and optical absorption spectroscopy), but it differs from chrysoprase by the higher vitreous lustre, greater transparency, much larger water contents (> 4 wt%) and the crystallinity of the silica hosting phase. In these opaline varieties, the green colour is also attributed to dispersed lOA Ni-bearing phyllosilicates. However, the silica phases are primarily crystoballite and tridymite, as opposed to chalcedony in chrysoprase. The yellow opal variety shows completely different optical absorption spectra from the apple green variety, lacking the broad absorption band at 650 nm characteristic of Ni in octahedral coordination. Transmission electron microscopy indicates that various Fe-, Mg-, and some Ni-bearing phyllosilicates are present as finely dispersed platelets throughout the sample. Further high-resolution electron microscopy is necessary to positively identify these phases and to determine the mineral(s) responsible for the orange yellow colour. Chrysoprase occurs throughout the weathering profile in the Gumigil Mine. It is present in parts of the ferruginous duricrust, in the upper saprolite horizons, and it is particularly abundant in the lower saprolite and saprock horizons. Field interpretation suggests that chrysoprase veins form as a result of silica replacement of magnesite veins during lateritic weathering of serpentinites. Silica released from serpentine and clay minerals during formation of the lateritic duricrust is redeposited, together with Ni, within the weathering profile. No particular control on the distribution of prase opal has been identified to date in the Gumigil weathering profiles. However, in the weathering profiles associated with the Brolga Ni-laterite deposit, also in the Marlborough Region, prase opal occurs predominantly at the boundary between the ferruginous duricrust and the upper saprolite. The yellow opal variety, however, only occurs within the ferruginous duricrust in the Gumigil profiles, and in the ferruginous saprohtes associated with mafic dykes in the Brolga profile, suggesting a close correlation between the orange yellow colour and high iron contents in the weathering solutions.

545

GEOCHEMISTRY, ENVIRONMENT AND FORMATION OF OPHIOLITIC ROCKS IN PAPUA NEW GUINEA: IMPLICATIONS FOR RELATIONSfflP OF INDIAN PLATE AND EASTERN AUSTRALIAN PLATE Khin Maung Wai, J Foden, M J Abbott and A E Grady School of Earth Sciences, Flinders University of South Australia, Australia, Department of Geology and Geophysics, Adelaide University, Australia SUMMARY Tholeiitic and transitional tholeiites are observed in ophiolitic rocks from the Himalayas and southwest Pacific regions. Multi-stage melting, on- and off- axis metamorphism are considered in these ophiolites. Probably they are HIMU sources related and originated from a fixed, unstable lower mantle plume. INTRODUCTION Cretaceous to Eocene ophiolitic rocks firom Indian plate (e.g. ophiolites from eastern India and Burma) and eastem Australian plate (e.g. ophiolites from Papua New Guinea, New Zealand and New Caledonia) may contain at least one belt. These ophiolites may be related with Ninetyeast ridge, and Australia-Pacific plate boundary transform faults. Tholaiitic and transitional tholeiites are observed in these ophiolites. Comparison of the Nd isotopic composition of the Sadowa Igneous Complex (SIC) (^Nd = + 1-02 to + 6.3) and Papuan Ultramafic Belt (E^^ = + 7.4) reveals a much larger degree of heterogeneity in the SIC. The SIC data plot within the MORB - GIB array and a wide range of 143 Nd/144Nd ratios (0.51270-0.51294) are qxiite comparable with those of oceanic island basalts. Probably it is a mixture of EM, DMM and HIMU sources, and mixing processes occur at shallow depths. HIMU sources m i ^ t originate from a fixed, unstable lower mantle plume (probably it is only major plume sourced in southwest Australia). Archean, Proterozoic, Cambro-Grdovician, Mesozoic to Eocene and Miocene ophiolites elsewhere could be related with multi-branded rifting of micro continents and originated from only that major plume (here it is named 'the Australian plume'). Mineral deposits (e.g. Uranium, gold diamond, ruby and sapphire) m i ^ t be related with that major plume.

546

PLATE TECTONIC EVOLUTION OF EASTERN PAPUA NEW GUINEA: MODEL OF EMPLACEMENT OF OPHIOLITIC ROCKS IN PAPUA NEW GUINEA Khin Maung Wai, * C J Pigram, * M J Abbott & * A E Grady School of Earth Sciences, Flinders University of South Australia, Australia, Australian Geological Survey Organisation (AGSO), Australia SUMMARY Ophiolitic rocks in Papua New Guinea are supposed to floor in small ocean basins between the continental fragments due to multi-branched rifting of a microcontinent. The Sadowa Ignens Complex is probably detached from hot zone and obducted onto the Owen Stanley metamorphics after the emplacement of the Papuan Ultramafic Belt. INTRODUCTION Cretaceous to Eocene ophiolitic rocks in Papua New Guinea (Sadowa, Kutu complexes, Papuan Ultramafic Belt) wrap around the Owen Stanley terrane. Probably they are fomied in small ocean basins between the continental fragments due to multi-branched rifting of a micro continent after a Permo-Triassic compression event following the southwest subduction. The Sadowa Igneous Complex (SIC) is supposed to floor a basin between the continental fragments of the Eastem, Papua and proto-Owen Stanley terranes. During the late Eocene to Middle OUgocene the SIC is probably detached from the hot zone and obducted onto the Owen Stanley metamorphics after the emplacement of the Papuan Ultramafic Belt (PUB) to form East Papua Composite Terrane (EPCT). During the Miocene the EPCT coUided with Eastem and Papuan plateaus. Subduction related Miocene to Pliocene the Astrolabe agglomerate was unconformably exposed on the SIC. Due to branch closing time of the ophiolitic complex, age of emplacement may be different. These ophiolitic complexes may form in narrow Red Sea-type basins or marginal basins. Geochemical data confirm that the SIC is largely heterogeneity in sources and could be originated in incipient and episodic spreading center, propagating rift tips or ridge-transform intersection which are supplied by small magma chambers. Possibly their emplacement is related with change in plate motion.

547

CRETACEOUS MAGMATISM, MAKE-UP AND BREAK-UP OF THE SW PACIFIC GONDWANA MARGIN S.D. Weavers J.D. BradshawS R.J. Pankhurst% R.J. Muir^, B.C. StoreyS T.E. Waight^' and XR. Ireland' 1 Department of Geological Sciences, University of Canterbury, Christchurch , New Zealand 2 British Antarctic Survey Cambridge CB3 OET, UK 3 Isotope Geology Unit, Scottish Universities Research and Reactor Centre, Glasgow G75 OQU, UK 4 Department of Geology, La Trobe University, Bundoora, Victoria 3083, Australia 5 Research School of Earth Sciences, Australian National University, Canberra ACT 0200, Australia Summary Early Cretaceous calc-alkalic I-type magmatism, related to subduction along the South-West Pacific margin of Gondwana, is a major component of the Median Tectonic Zone of New Zealand and the equivalent Amundsen Province in Marie Byrd Land and Thurston Island, West Antarctica. The active margin was characterized by oblique subduction, large scale strike-slip faulting, diachronous tectonic and magmatic events and significant compositional variation in the magmas produced. Major transcurrent faulting between adjacent superterranes in Marie Byrd Land may be linked in New Zealand to arc collision and the melting of mafic lithosphere followed by significant dextral rotation. In New Zealand, a rapid change to extensional tectonics in the mid-Cretaceous was recorded by the formation of metamorphic core conqjlexes and rift-bounded sedimentary basins, and in westem Marie Byrd Land by voluminous A-type and associated mafic magmatism. This was followed by the rifting of the Campbell Plateau from Marie Byrd Land representing the separation of the New Zealand continental block from West Antarctica and its transfer to the Pacific plate. Introduction The SW Pacific sector of Gondwana, New Zealand, Marie Byrd Land and Thurston Island, was an episodically active miargin thoughout the Mesozoic. In the currently accepted tripartite division of New Zealand geology into Westem Province, Median Tectonic Zone and Eastem Province, each province represents a grouping of tectonic terranes or a single superterrane. Two of New Zealand's superterranes, Westem Province and Median Tectonic Zone, are now recognized to extend into Marie Byrd Land and Thurston Island (Bradshaw et al., (1,2)).

Figure 1. SW Pacific continental configuration at present and at 100 Ma. Stipple indicates thin continental crust, dotted Unes are selected terrane boundaries. Areas mentioned in text are CP = Campbell Plateau, CR = Chatham Rise, EP = Edward Vll Peninsula, RHC = Ruppert-Hobbs Coast, TI = Thurston Island, MBL = Marie Byrd Land. Superterrane amalgamation in the region was the result of convergent tectonics and strike slip faulting, and was completed in New Zealand by about 118 Ma. This was followed rapidly by a period of extensional tectonics corresponding to the rifting of New ZealandfiromWest Antarctica and Australia in response to significant plate boundary reorganization in the South West Pacific (eg. Lawver & Gahagan (3)) (Fig. 1). Thus the Cretaceous geological history of the New Zealand - West Antarctica region records the change firom a long-lived convergent margin to continental extension and the development of rifled margins. Igneous rocks in the region provide a record of these changes and indicate that there was significant variation along the margin in the nature and timing of magmatic events and processes. 548

This paper is an opportunity to summarize briefly new data on the geochronology and geochemistry of New Zealand and Marie Byrd Land magmatism arising from collaboration between New Zealand, United Kingdom, United States and Australian geoscientists. New Zealand The Westem Province is largely made up of Lower Paleozoic quartzose metasedimentary rocks cut by DevonianCarboniferous granitoids (eg. Cooper (4), Muir et al, (5,6)). It had attained continental thickness and its gross structure by the end of the Carboniferous and represents a fragment of the Gondwana Paleozoic margin. The Eastern Province developed as a result of convergent margin processes in Permian to Cretaceous times and is dominated by accretionary complex, forearc and arc rocks. Separating Aese two provinces is a narrow belt of mainly magmatic rocks known as the Median Tectonic Zone (Bradshaw (7)) (Fig. 2). Mesozolc - Cenozolc cover rocks Charleston Metomorphlc Group Rahu Suite Separation Point Suite

Cretaceous granites

Koramea Suite - Palaeozoic granites Riwoko Complex - Devonian Buller terrane mainly low-grade metasedimentary rocks Takaka terrane Anatoki Ttirust

South Island Nelson/ Westlond

Fiordlond

MTZ

Eastern Province

Western ^ Province V^^MTZ

fffH Western Fiord land Orthogneiss - I - j Central Fiordland metasediments and granitic orthogneiss Eastern Fiordland igneous belt S W Fiordland block 0

lOOkm

Figure 2. Basement geology of Westem Province of New Zealand (after Muir et al. (5)). Median Tectonic Zone The Median Tectonic Zone comprises a narrow belt of largely plutonic rocks running through the South Island from Nelson in the north to eastem Fiordland in the south (interrupted by the Alpine Fault) and continuing into Stewart Island. Mesozoic ages in the range 130-160 Ma predominate (Kimbrough et al. (8,9), Muir et al. (10)). Triassic plutons 220230 Ma) tend to occur on the east side of the zone and there are isolated occurrences of Late 305 Ma) and Early 345 Ma) Carboniferous granitoids. Mesozoic rocks range from gabbroic to granitic varieties but mafic and intermediate lithologies are most abundant. These rocks are calc-alkalic, I-type compositions, presumed to be the result of long lived episodic subduction processes. No comprehensive petrogenetic study has been published to date but Early Cretaceous values of leucogabbros of the Darren Complex, eastem Fiordland, have initial ®'Sr/®^Sr ratios of 0.7037-0.7039 and +3.9 to +4.6 (McCulloch et al. (11)) indicative of a depleted mantle source. The intemal structure of the Median Tectonic

549

Zone is complex, arcs having been attenuated and dismembered by large scale transcurrent faulting in the Early Cretaceous (Kimbrough et al. (8)). Western Province The Early Cretaceous Separation Point Batholith 750 km^) of northem South Island is the easternmost belt of granitoids in the Western Province (Fig. 2). It intruded Lower Paleozoic rocks of the Takaka terrane. Western Province, (Cooper (4)) and Late Jurassic - Early Cretaceous mafic plutonic rocks of the Rotoroa Igneous Complex, Median Tectonic Zone, (Kimbrough et al. (8)), stitching the two superterranes at the ~ 118 Ma time of emplacement (Muir et al. (5)). Na-rich, alkali-calcic diorites, granodiorites and monzogranites make up the batholith (Muir et al. (12)). These differ from typical calc-alkalic subduction-related granitoids but are closely comparable to adakites (eg. Kay (13)), and Archean trondhjemite-tonalite-dacite suites. Geochemical characteristics include very high Sr concentrations ( 1000 ppm Sr at 70% Si02) and SrA^ ratios (Fig. 3), initial ''Sr/^Sr ratios of - 0.7042 and e^d ^lues of+1.2 to +1.8 (Muir et al. (12)). These features are consistent with melting of a mafic lithospheric underplate of gamet-amphibohte mineralogy beneath a thickened (> 40 km) arc. Because the trench at the Pacific margin of the Eastem Province was too remote (>600 km), it is necessary to postulate a subduction zone in the region of the Median Tectonic Zone itself. One possibility is closure of a back-arc basin between the Median Tectonic Zone arc and the Westem Province margin, although remnants of such a basin have yet to be identified. •1000 •

MBL

A SPB

Sr/Y SPB NZ

153-2W NZ

NZ20

500-

i Ev_

-100

MBL

XV I N Adakite field

-10

-L

100- ^^

island Arc field

.. -

30 20 10 Figure 3a. Multielement plot comparing typical I-type granodiorites of Marie Byrd Land and Separation Point • Batholith, New Zealand. Note large positive Sr anomaly and low Y of latter. Figure 3b. SrA^ versus Y plot showing Separation Point rocks plotting in adakite field and Marie Byrd Land I-types in normal island arc field. Within the Westem Province occur numerous isolated small plutons having the same chemistry as the main batholith. An ion microprobe U/Pb zircon age of 111 Ma has been obtained from one of these, the 01yii5)us pluton (Muir et al. (5)) (Fig. 2). Several of these bodies are low grade porphyry Mo prospects (Christie & Brathwaite (14). South of the Alpine Fault, undeformed granitoids in the region of lakes Te Anau and Manapouri, eastem Fiordland, are petrologically identical to Separation Point rocks but give ion microprobe ages of 123-124 Ma (Muir et al. (10)). Westem Fiordland Orthogneiss (Fig. 2) has the same distinctive chemistry, yields ages of ~ 125 Ma and is considered to be a lower crustal equivalent of Separation Point granitoids. Why Separation Point-type magmas south of the Alpine Fault were emplaced 5-7 Ma earlier than in northem South Island is as yet unclear but indicates along strike diachroneity. Granitoids of more typical calc-alkalic composition also occur in the Westem Province forming most of the small Paparoa 380 km^) and Hohonu 320 km^) batholiths and a small portion of the dominantly Devonian Karamea Batholith. Tulloch (15,16) designated these Rahu Suite and identified them as having intermediate I/S petrological features. The Buckland Granite of the Paparoa Batholith (Fig. 2) has a crystallization age of 110 Ma inferredfiromion microprobe zircon meausurements (Muir et al. (5)) and seven plutons of the Hohonu Batholith (Fig. 2) have ion microprobe ages in the range 109-114 Ma (Waight et al. (17)). Compositions rangefiromtonalite througii granodiorite to monzogranite with initial ®'Sr/®®Sr ratios of 0.7062-0.7085 and sj^^j values of-4.4 to -6.1. Waight (18) has demonstrated that these granites are the products of various degrees of mixing of Separation Point-type magmas with Lower Paleozoic metasediments. This magmatism may be the last vestige of subduction-related activity in the Westem Province or the result of extension and cmstal thinning immediately postdating the colhsion of the Median Tectonic Zone arc and emplacement of the Separation Point Batholith. In the Paparoa Batholith, ductile deformation associated with the formation of a metamorphic core complex (Tulloch and Kimbrougji (19)) postdates the 110 Ma Buckland Granite (Muir et al. (20)). We have recently obtained a pooled ion microprobe age of 101.2 ± 2.0 for melt precipitated zircons in two samples of silicic tuffs (Stitts Tuff Member, Pororari Group) at the base of the coarse infill in a half graben formed as a result of movement on one of the major detachment faults. These data confirm the conclusion of Laird (21,22) that the change to extensional tectonics following a long history of convergence is recorded in the Westem Province at --105 ± 5 Ma. In the Hohonu Batholith, Late Cretaceous magmatic activity is represented by the emplacement of the peralkaline A-type French Creek Granite at 83 Ma contemporaneous with the intmsion of an alkaline mafic dike swarm (Tulloch et al. (23), -i-

-L.

Pb Rb Ba Th

550

Nb La Ce Sr Nd P

Zr Ti

Waight (18)). These events coincide with a second period of extensional tectonics involving graben formation (Laird (21)) and correlate with the first appearance of oceanic crust in the Tasman Sea. Eastern Province

By contrast with the Western Province and Median Tectonic Zone, magmatic rocks of Cretaceous age are scarce in the Eastern Province. Alkaline basalts of the Gridiron and Lookout Volcanics give K/Ar ages of 95-98 Ma (Laird (21)) and may be the volcanic correlatives of the alkaline pyroxenite-gabbro-syenite complexes of Tapuaenuku and Blue Mountain in Marlborough (Grapes (24), Baker et al. (25)) and Mandamus in North Canterbury. The Mandamus Complex has given a Rb/Sr age of 97 ± 0.5 Ma (Weaver and Pankhurst (26)) and has an initial "Sr/®^Sr ratio of 0.70342 and value of +4.0 (Weaver et al (27)). Tapuaenuku rocks have ''SrP'Sr ratios of 0.7028-0.7030, e^d ^^^^^^ of+3.5 to +5 and '^Tb/'^Pb ratios of 20.1. Both Baker et al. (25) and Weaver et al. (27) have emphasised the HIMU-type OIB characteristics of these rocks, features often associated with mantle plumes. These rocks are thought to have been emplaced during crustal extension in the thick Torlesse accretionary wedge. Interestingly, the ~ 99 Ma alkalic pyroxenite-monzonite-syenite complexes of Mt Dromedary, south-east New South Wales (Smith et al. (28)) appear to petrologically similar to the Tapuaenuku-Mandamus suite and were also presumably emplaced in an extensional intraplate setting. Gamet-bearing rhyolites, formed by melting of Torlesse protoliths (Barley (29)), and andesites of the Mount Somers Volcanics of central Canterbury give slightly younger ages of -- 89 Ma (Barley et al. (30)), and are also regarded as representing extensional tectonics. Marie Byrd Land and Thurston Island Recent work in Marie Byrd Land has shown that two of the New Zealand superterranes may be recognized there (Bradshaw et al. (2), DiVenere et al. (31)). The terms Ross Province (equivalent to Westem Province) and Amundsen Province (equivalent to Median Tectonic Zone) are proposed by Pankhurst et al. (32). The approximate boundary between the two provinces is shown in Fig. 4. Isotopic data for Marie Byrd Land rocks quoted here are derived mostly from Rb/Sr geochronology and Sr-Nd data of R.J. Pankhurst, and U/Pb geochronology of S.B. Mukasa, which will be formally published elsewhere. Early Cretaceous subduction-related magmatism

Cretaceous subduction-related igneous rocks are restricted to the Amundsen Province of Marie Byrd Land. On the Hobbs Coast, calc-alkaUc, I-type plutons occur in the vicinity of McDonald Heights on Mt Sinha and Mt Steinfeld. These vary from diorite to monzogranite with homblende-titanite granodiorites predominating, have ages in the range 108-124 Ma and initial ®'Sr/^Sr ratios of 0.7054-0.7056 and e^d 0 (Weaver et al. (27)). In Pine Island Bay, eastem MBL (Fig. 4), we recognize calc-alkalic granodiorite-monzogranite plutons of similar age (114-125 Ma) but having slightly more primitive isotopic compositions («'Sr/^Sr 0.7044-0.7048, s^d 0 to +2). On eastem Thurston Island (Fig. 4), calc-alkalic gabbro-granodiorite complexes give Rb/Sr ages of 110-125 Ma and have initial ®'Sr/^Sr ratios of 0.7055-0.7070 and 0 to +1 (Leat et al. (33), Pankhurst et al. (34)). In the Kohler Range and Pine Island Bay other granodiorites with ^'Sr/^'^Sr ratios of 0.7055-0.7070 and ej^d of 0 to - 3 occur but these yield ages of 95-105 Ma. Correlating with this group are calcalkaUc volcanic rocks of the USAS Escarpment and probably those of the Jones Mountains, Eigihts Coast, east of Thurston Island which may have ages as young as 90 Ma (Pankhurst et al. (34)).

551

Figure 4. Sketch map of Marie Byrd Land and Thurston Island showing distribution of Cretaceous I-type and A-type granitoids, and Ross and Amundsen Provinces. A-types are limited to rifted margin along Edward Vll Peninsula Ruppert-Hobbs coast. All the above mentioned rocks have h i ^ Ca0/Na20 ratios, low SrA^ ratios and negative Nb anomaUes, typical of calcalkalic magmas of Andean margins (Fig. 3). The age pattem suggests that subduction ceased at - 110 Ma along the westem section and persisted until ~ 90 Ma along the eastem section of this margin (Mukasa (35), Weaver et al. (36)) (Fig. 5). This pattem is in accord with the strongly oblique nature of Phoenix ridge convergence (eg. Bradshaw (37), Lawver & Gahagan (3)). Although the Marie Byrd Land granodioritic rocks are partly coeval with the Separation Point Suite of New Zealand 125-110 Ma), they lack the distinctive positive Sr anomalies, high Sr/Y ratios and alkali-calcic composition of the latter. Qearly, different conditions of magmagenesis applied along the New Zealand sector where melting at the base of an exceptionally thick arc, involving back-arc basin subduction and arc-continent collision, has been postulated (Muir et al. (12)). Mid-Cretaceous rift-related magmatism Along the Ruppert and Hobbs coasts of Marie Byrd Land, a diverse suite ofA-type granitoids was emplaced at about 95105 Ma. Lithologies include syenites and monzogranites to alkali feldspar granites, including peralkaline varieties. Atype granitoids are predominant in the Ickes Mountains and McDonald Heights, Hobbs Coast, are subordinate to Paleozoic granitoids in the Ford Ranges and again predominant on Edward Vll Peninsula adjacent to the Ross Sea (Fig. 4) (Weaver et al. (27)). Granitoids on the Ruppert-Hobbs coasts have initial ®'Sr/«^Sr ratios of 0.7041-0.7045 and s^^ values of 0 to +4. On Edward Vll Peninsula however, a distinct group of A-type monzogranite-syenogranite plutons in the Rockefeller and Alexandra montains gives Rb/Sr ages of 95-105 Ma (Adams et al. (38)) and have ®^Sr/®^Sr ratios of 0.7116-0.7206 and values of -5.5 to -7.7, clearly indicating a significant crustal component in their petrogenesis. An origin by partial melting of Devonian I-type diorite-granodiorite basement has been suggested (Weaver et al. (39)). Mafic rocks in the form of an extensive coast-parallel dike swarm and stacked sheets occur along the Ruppert-Hobbs coast. Magma mingling relationships are seen in the field between early mafic dikes and the host granite at Mt Prince, the youngest dated I-type subduction-related pluton with an U/Pb age of 110 ± 1 Ma. A late mafic dike at Mt Prince has given an age of 100.6 ± 0.7 and a layered alkahc gabbro at Cape Burks has given an ^^ArP^Ar age of 99.9 ± 0.1 Ma (Mukasa (35)). A few mafic dikes cut the youngest dated A-type granites, the age span of mafic magmatism is therefore inferred to be ~ 110-95 Ma. Isotopic compositions of most A-type granitoids (excepting those at Edward Vll Peninsula) suggest that they are related by partial melting orfiractionationprocesses to coeval mafic magmas with little or no crustal contamination. Mafic rocks resemble some continental tholeiites and are considered to represent melting of subcontinental lithosphere during extension (Weaver et al. (27)). In the Fosdick Mountains, Ruppert Coast, and Alexandra Mountains, Edward Vll Peninsula, crustal extension has exposed metamorphic complexes in which mafic magmas have been injected and migmatization has occurred (Adams and Weaver (38), Richard et al. (40)). 552

W-MBL

E-MBL

Ma xxxxxx

Spreading 90

Passive

- Extension Rifting -

95 MAN - g

XXXXXXX

.Tl JONES

100

•Kkohl

105

XXXXXXX

' i 1

PIB

110

— PAP — HOH 115 — SPB -PIB

120

--WFO 125

Figure 5. Diagram showing west to east younging of the cessation of subduction-related magmatism (black) and timing of rift magmatism (stippled) in New Zealand, Marie Byrd Land and Thurston Island, localities or units mentioned in text are MAN = Mandamus Complex, PAP = Paparoa Batholith, HOH = Hohonu Batholith, SPB = Separation Point Batholith, WFO = Westem Fiordland Orthogneiss, KOHL = Kohler Range, PIB = Pine Island Bay, Tl = Thurston Island, JONES = Jones Mountains. The age, location, composition and bimodality of the mid-Cretaceous magmatism point strongly to a causal relationship with continental rifting. A narrow continental shelf with the lOOQm bathymetric contour close to the coast between Edward Vll Peninsula, and the Hobbs Coast suggests an intact rifted margin (Fig. 4). By contrast, eastern Marie Byrd Land to Thurston Island has a wide shelf suggesting a collapsed margin. Rift-related magmatism occurs througiiout the Ross Province and crosses into the Amundsen Province (Fig. 4), effectively stitching the two cmstal blocks together. The 100 Ma New Zealand -West Antarctica reconstruction of Lawver and Gahagan (3) has the Campbell Plateau tight up against this section of Marie Byrd Land. There is no evidence of rift-related magmatism fiirther east in the Kohler Range, Pine Island Bay or Thurston Island. In Pine Island Bay, mafic dikes that cut Early Cretaceous granodiorites have initial "Sr/®^Sr ratios of 0.7057-0.7070 and s^^ values of -2 to -3, are of calc-alkalic composition and appear to represent the last vestiges of subduction-related magmatism. On Edward Vll Peninsula, K/Ar biotite and zircon fission track ages are concordant with Rb/Sr whole rock ages, indicating rapid cooling after granitoid emplacement (Adams et al. (38)) and immediate uplift of the rift margin. The oldest identified magnetic anomalies in the south-west Pacific (Chron 34) suggest that sea floor spreading began at about 84 Ma (Mayes et al. (41)), whereas associated continental rifting in Marie Byrd Land and the initiation of sedimentary basins on the Campbell Plateau (Field et al. (42)) occurred at least 20 Ma earlier. In westem Marie Byrd Land, there is no rift magmatism younger than 95 Ma. It seems likely therefore that the continental margin had formed at this time and that continental separation occurred ~ 10 Ma prior to the time suggested by the marine magnetic anomaly record. Wider Tectonic Interpretations On paleomagnetic grounds, DiVenere et al. (31) propose that a major sinistral displacement of the order of 1000 km occurred between Eastern Marie Byrd Land (Amundsen Province) plus Thurston Island and Westem Marie Byrd Land (Ross Province) after ~ 117 Ma and before final amalgamation at ~ 100 Ma. Whereas in New Zealand, amalgamation of the Median Tectonic Zone {= Amundsen Province) and Westem Province (= Ross Province) was achieved prior to the 118 Ma emplacement of the stitching Separation Point Batholith. There is therefore conflict concerning the timing of

553

superterrane amalgamation. However, Bradshaw et al. (43) suggest that the New Zealand section of the Gondwana margin suffered oroclinal bending involving 90° dextral rotation post-118 Ma and prior to the onset of extensional tectonics at ~ 101 Ma. This is the striking "Z" shaped swing in the pre-Cenozoic tectonic grain of New Zealand. It is possible therefore that the post-117 Ma sinistral stike-slip motion in Marie Byrd Land was accommodated in New Zealand by the significant dextral rotation. The driving force for rotation may have been collision of the Median Tectonic Zone with the Westem Province incorporating an attempt to subduct the former beneath the latter. Such a tectonic scenariofitsnicely the petrogenetic requirements of Separation Point Suite magmas. If this interpretation is correct, the production of Separation Point-type magmas in Westem Fiordland at ~ 125 Ma indicates that coUision and significant strike-slip motion along the superterrane boundary occurred prior to 117 Ma. Cessation of Early Cretaceous magmatism and convergent tectonics in the New Zealand - Marie Byrd Land region has been ascribed to final stages ox subduction of the Phoenix plate along the Gondwana margin, either involving ridge subduction (Bradshaw (37), Lawver & Gahagan (3)) or subducted slab capture (Luyendyk (44)). Whatever the detailed mechanism, the demise of the Phoenix plate necessitated the development of a new extensional plate boundary in the region resulting in the ultimate separation of the Campbell Plateau from Marie Byrd Land and the transfer of New Zealand to the Pacific plate. The locus of rifting initially followed the junction between the Chatham Rise accretionary prism and the Marie Byrd Land arc, that is between the New Zealand Eastern Province and the Median Tectonic Zone. It then follows the line separating Campbell Plateau from Marie Byrd Land, cutting across the arc and through the Median Tectonic Zone and Westem Province. We have speculated that this curious path may have been influenced in part by the initiation of a mantle plume in mid-Cretaceoustimesand which is now manifest as the Erebus hot spot in the Ross Sea (Weaver et al. (27)). Acknowledgements Work in New Zealand was done under Foundation for Research Science & Technology contract UOC 313 "New Zealand Granites and Crustal Evolution". Research in Marie Byrd Land derives largely rom NZ-UK-US joint e?q)editions in 1990-92. We acknowledge the significant scientinc contributions to our data anc nterpretations of our US colleagues, Ian Dalziel, Vic DiVenere, Anne Grunow, Sam Mukasa and David Palais. Woric in Antarctica could not have been accomplished without the logistic support of the US National Science Foundatic , US Navy, US Coast Guard, British Antarctic Survey and New Zealand Antarctic Programme. This paper is dedicated to Andy Harris, our Marie Byrd Land field leader, who disappeared on May 10th 1996, descending Mt Everest. References 1. Bradshaw, J.D., Pankhurst, R.J., Weaver, S.D., Storey, B.C., Muir, R.J. & Ireland, T.R. 1996. The Mesozoic continental margin: Carboniferous-Mesozoic arc terranes in West Antarctica, New Zealand and Australia. In: Mesozoic Geology of the Eastern Australia Plate. 2. Bradshaw, J.D., Pankhurst, R.J., Weaver, S.D., Storey, B.C., Muir, R.J. & Ireland, T.R. (in press). New Zealand superterranes recognized in Marie Byrd Land and Thurston Island. Terra Antartica. 3. Lawver, L.A. & Gahagan, L.M. 1994. Constraints ontimingof extension in the Ross Sea region. Terra Antartica, 1, 545-552. 4. Cooper, R.A. 1989. Early Paleozoic terranes of New Zealand. J. Royal Society ofNew Zealand, 19, 73-112. 5. Muir, R.J., Ireland, T.R., Weaver, S.D. & Bradshaw, ID. 1994. Ion microprobe U-Pb zircon geochronology of granitic magmatism in the Westem Province of the South Island, New Zealand. Chemical Geology, 113,171-189. 6. Muir, R.J., Ireland, T.R., Weaver, S.D. & Bradshaw, J.D. 1996. Ion microprobe dating of Paleozoic granitoids: Devonian magmatism in New Zealand and its correlation with Australia and Antarctica. Chemical Geology, 127, 191-210. 7. Bradshaw, J.D. 1993. A review of the Median Tectonic Zone: terrane boundaries and terrane amalgamation near the Median Tectonic Line. N.Z. J. Geology & Geophysics, 36, 117-125. 8. Kimbrough, D.L., Tulloch, A.J., Geary, E., Coombs, D.S. & Landis, C.A. 1993. Isotopic ages from the Nelson region of South Island, New Zealand: cmstal structure and the definition of the Median Tectonic Zone. Tectonophysics, 225, 433-488. 9. Kimbrough, D.L., Tulloch, A.J., Coombs, D.S., Landis, CA., Johnston, M.R. & Mattinson, J.M. 1994. Uraniumlead zircon agesfiromthe Median Tectonic Zone, New Zealand. N,Z, J, Geology & Geophysics, 393-419. 10. Muir, R.J., Ireland, T.R., Weaver, S.D., Bradshaw, J.D. & Shelley, D. 1994. Geochronology of Eastem Fiordland, South Island, New Zealand. Geological Society ofN.Z. Misc. Publication, 80A, 140. 11. McCulloch, M.T., Bradshaw, J.Y. & Taylor, S.R. 1987. Sm-Nd and Rb-Sr isotopic and geochemical systematics in Phanerozoic granulitesfiromFiordland, southwest New Zealand. Contributions to Mineralogy and Petrology, 97, 183-195. 12. Muir, R.J., Weaver, S.D., Bradshaw, J.D., Eby, G.N. & Evans, JA. 1995. Geochemistry of the Cretaceous Separation Point Batholith, New Zealand: granitoid magmas formed by melting of mafic lithosphere. J. Geological Society ofLondon, 152, 689-701. 13. Kay, R.W. 1978. Aleutian magnesian andesites: meltsfiromsubducted Pacific ocean crust. J. Volcanology & Geothermal Research, 4, 117-132. 554

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Christie, A.B. & Brathwaite, R.L. 1995. Mineral Commodity Report 9-Molybdenum. New Zealand Mining 18,2232. Tulloch, A.J. 1983. New Zealand granitoid rocks: a brief review. Geological Society ofAmerica Memoir 159, 5-

20.

Tulloch, A.J. 1988. Batholiths, plutons and suites: nomenclature for granitoid rocks of Westland-Nelson, New Zealand. N.Z. J. Geology & Geophysics. 31, 505-509. Waight, T.E., Weaver, S.D., Ireland, T.R., Maas, R.., Muir, R.J. & Shelley, D. (in press). Field characteristics and geochronology of the Hohonu Batholith, North Westland, New Zealand. RZ. J. Geology & Geophysics. Waight, T.E. 1995. The Geology and Geochemistry of the Hohonu Batholith and adjacent rocks. North Westland, New Zealand. Unpublished Ph.D. thesis. University of Canterbury, New Zealand. Tulloch, A.J. & Kimbrough, D.L. 1989. The Paparoa metamorphic core complex: Cretaceous extension associated with fragmentation of the Pacific margin of Gondwana. Tectonics, 8,1217-1234. Muir, R.J., Bradshaw, J.D., Weaver, S.D. & Ireland, T.R. 1994. Crustal extension prior to the opening of the Tasman Sea Basin: evidence from New Zealand granites. In: Evolution of the Tasman Sea Basin, van der Lingen, G.J., Swanson, K.M. & Muir, R.J. (eds.), A.A. Balkema, Rotterdam, 55-64. Laird, M.G. 1993. Cretaceous Continental Rifts: New Zealand Region. In: South Pacific Sedimentary Basins. Sedimentary Basins of the World, 2, Ballance, P.F. (ed.) Elsevier, Amsterdam, 37-49. Laird, M.G. 1996. Mid Cretaceous break-up basins of the South Island, New Zealand. In: Mesozoic Geology of the Eastern Australia Plate. Tulloch, A.J., Kimbrough, D.L. & Waight, T.E. 1994. The French Creek Granite, North Westland, New Zealand Late Cretaceous A-type plutonism on the Tasman passive margin. In: Evolution ofthe Tasman Sea, van der Lingen, G.J., Swanson, K.M. & Muir, R.J. (eds.), AA. Balkema, Rotterdam, 65-67. Grapes, R.H. 1975. Petrology of the Blue Mountain Complex, Marlborough, New Zealand. J. Petrology, 16,371428. Baker, J.A., Gamble, J.A. & Graham, I.J. 1994. The age, geology and geochemistry of the Tapuaenuku Igneous Complex, Marlborough, New Zealand. N.Z. J. Geology & Geophysics, 37,249-268. Weaver, S.D. & Pankhurst, R.J. 1991. A precise Rb-Sr age for the Mandamus Igneous Complex, North Canterbury, and regional tectonic implications. N.Z. J. Geology & Geophysics, 34,341-345. Weaver, S.D., Storey, B.C., Pankhurst, R.L, Mukasa, S.B., Di Venere, V.J. & Bradshaw, J.D. 1994. Antarctica-New Zealand rifling and Marie Byrd Land lithospheric magmatism linked to ridge subduction and mantle plume activity. Geology, 22, 811-814. Smith, I.E.M., White, A.J.R., Chappell, B.W. & Eggleton, RA. 1988. Fractionation in a zoned monzonite pluton : Mount Dromedary, southeastern Australia. Geological Magazine, 125, 273-284. Barley, M.E. 1987. Origin and evolution of mid-Cretaceous garnet-bearing, intermediate and sihcic volcanics from Canterbury, New Zealand. J. Volcanology & Geothermal Research, 32, lAl-lGl, Barley, M,E., Weaver, S.D. & de Laeter, J.R. 1988. Strontium isotope composition and geochronology of intermediate-silicic volcanics, Mt Somers and Banks Peninsula, New Zealand. N.Z. J. Geology & Geophysics, 31, 197-206. DiVenere, V., Kent, D.V. & Dalziel, I.W.D. 1995. Early Cretaceous paleomagnetic results from Marie Byrd Land, West Antarctica: imphcations for the WeddeUia collage of crustal blocks. J. Geophysical Research, 100, 81338151. Pankhurst, R.J., Weaver, S.D., Bradshaw, J.D., Storey, B.C. & Ireland, T.R. The pre-Mesozoic geology of Marie Byrd Land, Antarctica, submitted to J. Geophysical Research. Leat, P.T., Storey, B.C. & Pankhurst, R.J. 1993. Geochemistry of Palaeozoic-Mesozoic Pacific rim orogenic magmatism, Thurston Island, West Antarctica. Antarctic Science, 5,281-296. Pankhurst, R.J., Millar, I.L., Grunow, A.M. & Storey, B.C. 1993. The Pre-Cenozoic magmatic history of the Thurston Island crustal block. West Antarctica. Geophysical Research, 98,11835-11849. Mukasa, S.B. 1995. U-Pb, Rb-Sr and ^Ar/^'Ar age constraints on the development and tectonic evolution of microplates in West Antarctica. Abstracts VIIInternational Symposium on Antarctic Earth Sciences, Siena, Italy, 278. Weaver, S.D., Pankhurst, R.J., Storey, B.C., Bradshaw, J.D. & Muir, R.J. 1995. Cretaceous magmatism along the SW Pacific Gondwana margin. Abstracts VII International Symposium on Antarctic Earth Sciences, Siena, Italy 401. Bradshaw, LD. 1989. Cretaceous geotectonic pattems in the New Zealand region. Tectonics, 8, 803-820. Adams, C.J., Seward, D. & Weaver, S.D. 1995. Geochronology of Cretaceous granites and metasedimentary basement on Edward VII Peninsula, Marie Byrd Land, West Antarctica. Antarctic Science, 7, 265-277. Weaver, S.D., Adams, C.J., Pankhurst, R.J. & Gibson, LL. 1992. Granites of Edward VH Peninsula, Marie Byrd Land: anorogenic magmatism related to Antarctic-New Zealand rifling. Geological Society ofAmerica, Special Paper, 272, 281-290. Richard, S.M., Smith, C.H., Kimbrough, D.L., Fitzgerald, RG. & Luyendyk, B.P. 1994. Cooling history of the northern Ford Ranges. Tectonics, 13, 837-857. 555

41. 42.

43. 44.

Mayes, C.L., Lawver, LA. & Sandwell, D.T. 1990. Tectonic history and new isochron chart of the South Pacific. 1 Geophysical Research, 95, 8543-8567. Field, B.D., Browne, G.H., Davy, B., Herzer, R.H., Hoskins, R.H., Raine, J.J., Wilson, G.J., Sewell, RJ., Smale, D. & Watters, W.A. 1989. Cretaceous and Cenozoic Sedimentary Basins and Geological Evolution of the Canterbury Region, South Island, New Zealand. N.Z. Geological Survey Basin Studies 2, NZGS, Lower Hutt, 94pp. Bradshaw, J.D., Weaver, S.D. & Muir, RJ. (in press) Mid-Cretaceous oroclinal bending of New Zealand terranes. N.Z, J. Geology & Geophysics. Luyendyk, B.R 1995. Hypothesis for Cretaceous rifling of east Gondwana caused by subducted slab capture. G^o/ogy, 23,373-376.

556

THE WOOGAROO SUBGROUP IN THE HELIDON HILLS (SE QLD): STRATIGRAPHIC APPRAISAL AND TECTONIC IMPLICATIONS

Edwin C WiUey, University of Southern Queensland, TOOWOOMBA, Qld ABSTRACT Reevaluation of the stratigraphy of the Woogaroo Subgroup (WSG) (Laidley Sub-Basin, Clarence-Moreton Basin) in the Helidon Hills area, Lockyer Valley (SE Qld) has resulted through integrated study of the sequence in GSQ Stratigraphic Borehole - Ipswich 9 and surface e?q)osure in its vicinity. Correlation with the stratigraphic scheme of Wells et al (1990)(13) and with formations (Aberdare Conglomerate, Raceview Formation and Ripley Road Sandstone) of the WSG is discussed. Petrographic studies indicate interdigitation of sands of two different provenance - a dominant quartz-lithic provenance and a quartzrich provenance in the west at the top of the WSG. These provenances are reflected in porosity and permeability control of groundwater distribution. Faults have been previously described, and structure contours developed on the top of the WSG have helped identify other faults. These faults together with other structural features along the margins of the Yarraman and Cressbrook-Buaraba Blocks with the Laidley Sub-Basin give indication of a dominantly strike-slip extensional tectonic regime before, during and after the deposition of the WSG. INTRODUCTION This presentation contributes to existing stratigraphic and structural, and thus to the tectonic, knowledge of the relationship between the Laidley Sub-Basin of the Clarence-Moreton Basin and the Yarraman/Cressbrook-Buaraba Blocks in the Helidon Hills and adjacent areas. A framework for this study is provided by the detailed summary and the review of previous work on the ClarenceMoreton Basin presented by Wells & O'Brien (1) (lithostratigraphy), O'Brien & Wells (2) (sedimentology) and O'Brien et al (3) (tectonics). Work of other authors will also be referred to. Besides reUance on published materials, much of the data en:q)loyed derives from: (a) unpubhshed reports, including project reports prepared by third year students at the Darling Downs Institute of Advanced Education (DDL^E) now the University of Southem C^eensland (USQ), (b)fieldand laboratory observations collected specifically or eclectically over many years particularly in relation to student projects but also in relation to recent consultancies, and (c) water borehole information obtained from drillers, property owners and reports to government agencies. This work was supported by USQ departmental, faculty and university research funds, but the main ideas arose from a consultancy undertaken for the Grantham Progress Association. Grateful thanks are also due to USQ students, colleagues generally, drillers and property owners. While stratigraphy embraces lithological, environmental, biostratigraphic, petrographic, provenance, palaeocurrent and other studies, concentrated attention has been given to lithological description particularly of sandstones in the sequence. Lithological study has depended chiefly on hand specimen study with some supportfromthin section studies. Availabihty of 1:25,000 topographic maps with 5m contour intervals aided the structure contouring of the stratigraphic marker near the Woogaroo/Marburg Subgroup boundary. Locations are indicated by 'AMG' followed by a six or eight digit grid reference which refer to 9343 Esk and 9342 Helidon 1:100,000 Topographic Maps. STRATIGRAPHY OF THE HELIDON HILLS AND ADJACENT AREAS Introduction The Bundamba Group (Mesozoic) of the Helidon Hills area (see Figure 1) rest unconformably on diverse units chiefly of Palaeozoic age and are unconformably overlain by basic volcanics and associated sediments and soils of the Main Range Volcanics capping higher ground in the north of the area (4, also 5,6,7,8,9). Basement Palaeozoic rocks outcrop in the north of and as inliers within the Helidon Hills area, and have been penetrated in boreholes. Basement includes a basic complex (the core of the Mt Cross Inlier (AMG200625)), regional metamorphics, and sedimentary and volcanic sequences of Devonian to Triassic age. Greissenised granite outcrops along Sandy Creek (AMG20305450), and alkali granites are penetrated in GSQ-Ipswich 9 (AMG14535296? (precise location uncertain, see Appendix) at 304-309m), the dry water bore at Boxmoor (AMG21285061 at 155m) and in a water bore (AMG20375685 at 97-152m). Correlation with local granites would suggest a latest Permian to early Triassic age (4,10).

557

Figure 1. Location of Helidon Hills area. A. Regional tectonic setting (based onfigure4, Cranfield et al 1976). Ipswich 4, 9 and 18 and NS272: Geological Survey of Queensland Boreholes. (Devonian-Permian: D'Aguilar Block (DAB), Beenleigh Block (BB), Yarraman Block (YB) and Texas Block (TeB). Permian: Cressbrook Creek Block (CCG) and Northbrook Block (NB). M. Triassic: Esk Trough (ET). U. Triassic: Tarong Basin (TaB). Uppermost Triassic - Jurassic: Clarence-Moreton and Nambour Basins. B Geological map (adapted from Ipswich 1:250,000 Geological Map 1973). Tertiary: Main Range Volcanics (horizontal lines); Mesozoic: Bundamba Group and younger (stippled), Toogoolawah Group (dashed lines); Palaeozoic: Plutonic rocks (crosses), Cressbrook Creek Group (vertical lines), Sugarloaf Metamorphics (squiggly lines). Bundamba Group The Bundamba Group, which includes all Mesozoic strata in the Helidon Hills, is extensively developed in the ClarenceMoreton Basin, Great Artesian Basin and associated basins. The stratigraphic terminology appUed to the Bundamba Group by McTaggart (11) was adapted by the Geological Survey of Queensland (4,12). However, these schemes have proved inadequate, particularly regarding the definition of the Helidon Sandstone. Nomenclature for the ClarenceMoreton Basin strata as recently revised by Wells et al (13) is followed here. The approximate equivalence between the schemes of McTaggart, GSQ and Wells et al (13) are shown in Table 1.

558

Table 1 Stratigraphic terminology in the Bundamba Group McTaggart (1963)

GSQ(1974,1976) Upper Marburg Formation

Wells et al.(1990) Koukandowie Formation

Marburg Formation Gatton Sandstone Calamia Member Ripley Road Sandstone Helidon Sandstone Raceview Fomiation Woogaroo Sub-Group Aberdare Conglomerate The Bundamba Group rocks present in the main study area belong to the Woogaroo Subgroup (WSG) and lowermost Marburg Subgroup and include the Aberdare Conglomerate and equivalents, the Raceview Formation (RVF), the Ripley Road Sandstone (RRS) (defined in 14)) and the basal part of the Gatton Sandstone, including the Calamia Member (defined in 13). Lower Marburg Formation

Geological Survey of Queensland Stratigraphic Borehole Ipswich No. 9 (GSQ Ipswich 9; AMG14535296? (see Appendix)) provides an uninterrupted sequence of Bundamba Group strata. This sequence and its correlation with other sections in the main study area are summarised in Figure 2. Sandy Creek Catchment

GSQ Ipswich 9

I Boxmoor Water Bore

Paradise Creek Section

^X fi! CO '

Its' isil illl

ffi C7|

150m_

/ /

Figure 2. Representative stratigraphic sections in the Helidon Hill area (see Figure 4 for location of sections. Paradise Creek Section (based chiefly on Jacob staff section 12/12/73), GSQ Ipswich 9 (core study Feb. 1994), Sandy Creek area (based on field study Mar-Jun 1994, and water bore data), Boxmoor Bore (driller's log). The basal Woogaroo Subgroup varies considerably across the areas; some variants are discussed later. The RVF (type section in NS272 (see Figure 1)) comprises quartzose fine to medium sandstones interbedded with siltstones, shales, carbonaceous shales and thin coals. The RRS (type section also in NS272), by contrast, contains massive fine to very coarse sandstones containing predominately clear or milky quartz. (14). The overlying Gatton Sandstone is composed of medium to very coarse quartz-lithic to lithic sandstones, with minor lithic conglomerates and interbedded siltstones, shales and very thin coals (13). The Calamia Member is developed at the base of the Gatton Sandstone; this unit is encountered in the main study area, such as (a) interbedded fine to very fine sandstones

559

(about 80 to 150 mm thick) and shales at AMG17805185, or (b) shales and siltstones intercalated between sandstones in GSQ Ipswich 9. Discussion of stratigraphy Some basal strata of the WSG may be readily referred to the Aberdare Conglomerate (see (a) below), but others force the definition of the Aberdare Conglomerate ((b) and (c) below). For example: (a) in GSQ Ipswich 9 and the Boxmoor Waterbore (AMG20375685), and southwest of Perseverance Reservoir (AMG120790) and also between AMG290700 and AMG340713, conglomerates (5 to 25 metres) referrable to the Aberdare Conglomerate overlie pre-Bundamba rocks. (b) at AMG20305450, AMG12126347 and AMG 08207062, breccias (0.2 to 0.5 m) occur containing greissenised granite, Permian sedimentary rocks and Permian intemiediate volcanics respectively, and (c) at AMG076839, consolidated granite grus (about 10 metres) overlies the Crows Nest Granite.

Figure 3, QFR diagram of sandstonesfiromRaceview Formation (solid) and Ripley Road Sandstone (open - quartzlithic; crossed - upper quartz-rich). Data from Hawkins (1975; circlesfi-omNS272 (type sections), triangles from GSQ Ipswich 18, diamonds from GSQ Ipswich 4) and O'Keeflfe (1995; squaresfi-omGSQ Ipswich 9). More detailed study of this part of the succession may require either redefinition of the Aberdare Conglomerate or the definition of new local lithostratigraphic units. The definitions of the RVF and RRS (14) indicate two differences - broad lithology and sandstone composition. Using these two criteria the two major units of the Woogaroo shoxild be readily differentiated. However, sandstone petrography of these two formations suggests otherwise. Petrographic data reported by Hawkins (15 (fig 1)) and by O'Keeffe (9 (App. 1)) has been recalculated and plotted in Figure 3. This plot indicates that compositional character of the sandstones does not serve to differentiate between the quartz-lithic character of RRS and RVF as defined in NS272 or as recognised elsewhere, with the exception of the quartz-rich horizons in the upper RRS. Wells and O'Brien (1) said: The boundary between the Ripley Road Sandstone and Raceview Formation can be defined primarily on the presence or absence of interbedded siltstone, shale and coal and partly on sandstone composition and sedimentary characteristics.... (page 8) However, the ambiguity introduced by the petrographic data would encourage the use of broad lithology as the sole rather than the primary consideration in differentiating the units. Given this, placing the boundary between RVF and RRS in GSQ Ipswich 9 at 134.9 metres below surface (see Fig. 2) seem justified. The quartz-lithic and the quartz-rich sandstones can be differentiated both petrographically (9,15) and in hand specimen. In GSQ Ipswich 9, the quartz-rich sandstones coincide with visible porosity (David Carmichael, personal communication). Also, permeability of GSQ Ipswich 9 core (16) showed high permeabilities coinciding with quartz-rich sandstones, while generally very low or undetectable permeabilities characterised the quartz-lithic sandstones. Information of groundwater distribution gained from various sources conforms with the stratigraphic distribution of composition, porosity and permeabihty described above. However, the interpretation of sandstone composition of the dry Boxmoor waterbore depends on the hydrological character of the strata encountered. The distribution of these two sandstone compositions is controlled by primary depositional features, although secondary diagenetic modification is not ruled out entirely.

560

Interdigitation of quartz-rich and quartz lithic sandstones in the RRS (see Figure 4) implies two coexisting sources. The roundness and mineralogical maturity of the quartz-rich sandstones suggests long transport. However, the thickening of the quartz-rich sandstones westward does not necessarily prove a westem source, as similar quartz-rich sandstones occur at the top of the RRS in its type section in NS272 (15). The sandstones of the Gatton Sandstone are quartz lithic to lithic (13) and the basal sandstones of the Gatton Sandstone (Calamia Member) in GSQ Ipswich 9 have been described petrologically as Q 64%, F 4% and R 2% (9). Thus they differ

Figure 4. Structure contour map on Top Woogaroo Marker in the Helidon Hills area. This marker is well developed in the westem third, more indistinct in the centre and northeast, and absent in the southeast. little from the dominant sandstones of the WSG, however, they do contrast markedly with the quartz-rich sandstones at the top of the WSG. Because of this (and, locally, the development of the Calamia Member), the boundary between the outcrops of the top Woogaroo and basal Marburg Subgroups is often readily recognisable. The Woogaroo outcrop has poor sandy soils, and is only rarely cleared of timber for grazing or cultivation, while the Marburg outcrop has sandy loams and is often cleared at least for grazing. The permeable character of the quartz-rich sandstones at the top of the WSG encourages percolation of rain rather than run-off. As a consequence the outcrop of the top of the RRS resists erosion and is marked by mesa/dip slope development. The top of this feature (Top Woogaroo Marker) can be readily recognised in aerial photographs and is easily mapped on the ground. Wright's Quarry (AMGl 70556) and several other quarries recovering building stone (Helidon Sandstone) are sited above the Top Woogaroo Marker. STRUCTURE OF THE HELIDON HH^LS AREA Structure contours on the Top Woogaroo Marker reveals several faults. One of these (#4 in Fig. 5) is represented on the Ipswich 1:250,000 Sheet, and has been identified in the field at AMGl 17585 (David Carmichael, personal communication), AMG14645273 and AMG40795190. One (#3 in Fig. 5) is aligned with a major fault associated with serpentinite in the Mount Cross Inlier separating basic volcanics from rocks of the Cressbrook Creek Group. The three eastem faults (#7, 8 and 9) with others ftirther north are elements of the 'Westem Border Fault', which with east-west striking unconformities mark the boundary between Palaeozoic and Mesozoic rocks (17). The large cutting on the Warrego Highway (AMGl 87517) exposes the RRS lying between two splays of Fault #5; and the westernmost splay is exposed in the Highway cutting at AMGl 8075185; several surface e?q)ressions of this group of faults can be seen in the 561

field. Along line #2, the sequence at the base of the Bundamba Group (AMG081707) shows a westerly dipping unconformity surface overlain by a basal scree breccia against which fluvial strata of the main Bundamba Group abut; this relationship may be controlled by faulting. The straight trend of base of the Bundamba Group in the northwest (#1, Fig.5) is probably fault controlled.

Figure 5. Summary of structural relationship between Laidley Sub-Basin and basement (area shown is same as shown in Fig lA). Faults (tick on downthrow side): solid lines; inferred fault controlled relationship: broken lines. Basement (see Fig 2.); Bundamba Group (stippled); post-Bundamba Group rocks (not shown). The main trend exhibited by these faults (about 1601I]) is the same as that of subvertical joints locally developed in Bundamba Group rocks through the area. Similar trend and attitude has been noted for mylonitic shear zones identified in recent mapping of the Maronghi Creek Beds. Faults with similar trends could account for the structure within the Cressbrook Creek Group. Assuming that at least the faults cutting Bundamba Group rocks have a common tectonic origin, the nature of the principal stress must have had a significant strike-slip component. Strike-slip would account for the lack of consistency of downthrows across the area and for the downthrow along Fault #5 changing side (see Fig. 5). Were the area affected by significant east-west extension, then the sense of displacement of the faults (unless reactivated later) would be consistently on one side. The detail of the faulting shows that the Gatton Arch is in effect a low plunging horst, rather than an anticline. In summary, these observations point to fault activity occurring both before and after the deposition of the WSG; incomplete mapping northwest of Crows Nest suggests movement during deposition of the WSG. Although observations indicating sense of any strike-slip movement have not been noted, nothing contradicts the dextral strike-slip movement recognised for the region through out the deposition of the WSG (3). CONCLUSIONS While there are considerable details which bear on the matter of the relationship between the basal strata of the Laidley Sub-Basin and older rocks, this preliminary review permits the following conclusions: • The RVF and RRS sandstones are dominated by quartz-lithic material, with quartz-rich sandstones occurring at, and generally marking, the top of the WSG. • The RVF and RRS should be differentiated on the basis of broad lithology alone; sandstone composition should not be used. • The Top Woogaroo Marker is a useful stratigraphic marker in the Helidon Hills area. • Structure contours on the Top Woogaroo Marker delineates structure including faults, which indicate movement before and after, and possibly during, deposition of the WSG. • Interpretation of the fault movements supports and provides detail for the strike-slip tectonic regime previously described (see (3)), although sense of movement is obscure.

562

Understanding of the depositional and tectonic evolution of the Clarence-Moreton Basin would benefit fi-om comprehensive studies of the provenance and dispersal of sand material deposited in the Clarence-Moreton Basin. General stratigraphic and structural study of the relationship between Bundamba Group strata and basement will also yield data contributing to developing a more detailed tectonic model. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

14. 15. 16. 17.

Wells AT & O'Brien PE 1994. Lithostratigraphic fi-amework of the Clarence-Moreton Basin. Australian Geological Survey Organisation Bulletin, 241,4-41. O'Brien PE & Wells AT 1994. Sedimentology of the Bundamba Group. Australian Geological Survey Organisation Bulletin, 241, 72-136. O'Brien PE, Korsch RJ, Wells AT, Sexton MJ, & Wake-Dyster K 1994. Structure and Tectonics of the ClarenceMoreton Basin. Australian Geological Survey Organisation Bulletin, 241, 195-216. Cranfield LC, Schwarzbock H and Day RW 1976 Geology of the Ipswich and Brisbane 1:250,000 Sheet Areas. Report of the Geological Survey of Queensland, 95. Smith PE 1974 The geology of the Mount Cross area (Parish of Murphy) South-east Queensland. Third year Geology Project, DDIAE (now USQ) (unpublished). Missen DD 1975 The geology of the Alice Creek—Mount Cross area. South-east Queensland Third year Geology Project, DDIAE (now USQ) (unpublished). Sowerby RD 1986 Stratigraphy of the Basal Tre-Helidon " members of the Clarence-Moreton Basin. Third year Geology Project, DDIAE (now USQ) (unpublished). Hamerh K 1992 Geological observations in the Buaraba Creek-Mount Perseverance, South-east Queensland. Third year Geology Project, USQ, (unpubUshed). O'Keeffe AJ 1995 Stratigraphy of the Sandy Creek Area Third Year Geology Project Report, USQ (unpubUshed). Murray CG 1994 Basement cores from the Tasman Fold Belt System Weath the Great Artesian Basin in Queensland. Queensland Geological Record, 1994/10. McTaggart NR1963 The Mesozoic sequence of the Lockyer-Marburg Area, South-east Queensland. Proceedings of the Royal Society of Queensland, 73, 93-104. Cranfield LC and SchwarzbockH 1974 New and revised stratigraphic names in the Ipswich 1:250,000 Sheet area. Queensland Government Mining Journal, 75,322-323. Wells AT, O'Brien PE, Willis IL and Cranfield LC 1990 A new stratigraphic framework for the Early Jurassic units in the Bundamba Group, Clarence-Moreton Basin, Queensland and New South Wales. BMR Journal of Australian Geology and Geophysics, 11,397-414. Staines HRE 1964 Stratigraphic Nomenclature of Bundamba Group in the Ipswich area. Queensland Government Mining Journal, 65,33-35. Hawkins PJ 1975 Appendix- Petrological Analysis of Selected Core Samples from NS272 and GSQ Ipswich 18. Queensland Government Mining Journal, 76,325-328. Thompson PS 1987 Petrological and petrophysical data from Mesozoic sandstones of the Bundamba Group, Qarence-Moreton Basin. Record of the Bureau of Mineral Resources, Geology and Geophysics, 1987/8. Ross LA 1988 An Investigation of the Geology and Fault relations of the Western Border Fault of the South Western Esk Trough. Third year Geology Project, DDIAE (now USQ) (unpubhshed).

APPENDIX Note on the Location of GSQ Ipswich 9. The location of GSQ Ipswich 9 is uncertain. Zahawi (1975) showed it at AMG14535296, but Queensland Water Resources Commission (Toowoomba office) place it almost 400 metres to the SW at AMG 14205275. Pubhshed information cannot discriminate between these two locations, and file details on precise position have not been found. Its position is not without interest as the two reported locations he on either side of a fault known from surface exposure and interpretedfiromstructure contours on the Top Woogaroo Marker. Reference: Zahawi Z 1975 Lockyer Valley groundwater investigations. Hydrogeological report. Record of the Geological Survey of Queensland, 1975/36.

563

THE MESOZOIC GEOLOGY OF THE EASTERN AUSTRALIA PLATE Neil Williams and Russell Korsch Australian Geological Survey Organisation, GPO Box 378, CANBERRA ACT2601 Summary During the Early Mesozoic, the Eastern Australian Plate formed part of the eastern Gondwanan margin of the Pangaean supercontinent, and was dominantly subduction-related. In the Late Mesozoic, Pangaea started to fragment and disperse, removingfromeastem Australia former components such as New Zealand and New Caledonia, and the region was predominantly influenced by divergent tectonics. Thus sedimentary basins and magmatic activity developed within this framework. In terms of earth resources, Mesozoic sedimentary rocks are important reservoirs of groundwater in Austraha, and of oil and gas in Australia and New Zealand. Most metalliferous deposits in the Mesozoic of the Eastem Australian Plate are related to calc-alkaline igneous activity, but appear not to be as economically important at this time as elsewhere in the world. Interestingly, the most economically important Mesozoic metalliferous deposit is unrelated to igneous activity. This is the Groote Eylandt manganese deposit, one of the world's largest stratiform manganese deposits. Introduction This paper provides a broad introduction to the geology of the Eastem Australia Plate (EAP) during the Mesozoic Era which embraces the Triassic, Jurassic and Cretaceous Periods, and extends in absolute age from about 250 to 65 millions of years ago (Young and Laurie, 1996). The paper focuses on the global character of the Mesozoic, the geological evolution of the EAP during the Mesozoic, and on the earth resources occurring within the Mesozoic of the EAP. Global Setting of the EAP At the start of the Mesozoic, the EAP formed part of Pangaea, a single supercontinent that comprised most of the earth's continental cmst. At this time, most continental fragments had coalesced and Pangaea was ahnost at its maximum size. Pangaea started to fragment in the Late Mesozoic, fomiing Laurasia in the north and recreating the supercontinent of Gondwana in the south. The world's cUmate was generally warm and wet in the Triassic at the beginning of the Mesozoic. This climate supported the appearance on land of many new plants, including conifers, ginkgoes, cycads, ferns, and seed fems. Worldwide warm to hot and wet conditions during the Jurassic saw these new plants continue to flourish, a situation that remained until early Cretaceous times when the world began to cool, and Pangaea began fragmenting (White, 1988). Conditions on land also supported the appearance of the first dinosaurs in the Triassic, and these reptiles flourished and diversified through the Jurassic and Cretaceous until their sudden demise at the end of the Mesozoic. Both mammals and birds appeared on earth by the Jurassic (Vickers-Rich et al., 1991). Globally, the Mesozoic was a period when vast quantities of earth resources were formed. The Mesozoic was particularly important for the formation of hydrocarbons, with some 54% of the world's oil, and 44% of the world's natural gas occurring in Mesozoic sediments, primarily in the Jurassic and/or Cretaceous of the Middle East, Western Siberia, and the Gulf of Mexico. By contrast the oil and natural gas percentages for the Palaeozoic are 14% and 29% respectively, and for the Tertiary, 32% and 27% respectively (Tissot & Welte, 1984). In terms of metallogeny, the Mesozoic marked the beginning of the formation of porphyry copper deposits in the Cordillera of North America, a major global metallogenic event that peaked in the Tertiary when it expanded to include the Cordillera of South America (Meyer, 1981). The EAP in the Eariy Mesozoic As part of Pangaea at the beginning of the Mesozoic, the EAP included New Caledonia, New Zealand, Marie Byrd Land and the Transantarctic Mountains. The EAP also contained continental crustal fragments that have been thinned by later extension and are now submerged, such as the Dampier Ridge, Lord Howe Rise, Norfolk Ridge, Campbell Plateau and Chatham Rise (see reconstructions by Walley & Ross, 1991; Powell & Li, 1994). All of these units together formed the eastem Gondwanan margin of Pangaea. The final convergence of crustal fragments occurred in the Middle Triassic, by which time incipient rifling in Laurasia and westem Gondwana had commenced (Veevers, 1989). This marked the initial phase of the breakup of Pangaea, althougii it would not be until the Late Jurassic that the stretching of continental crust, eventually leading to breakup, would commence in the EAP. In the Early Triassic, the margin of the EAP was a convergent one, with active subduction being west-dipping. The best evidence for this is now preserved in New Caledonia and New Zealand, where remnants of a convergent margin, such as an accretionary wedge, forearc basin and continental margin magmatic arc, are preserved (Korsch & Wellman, 1988; 564

Mortimer, 1995; Cluzel et al., 1995). Thus, most of eastem Australia was in a back-arc to intracratonic plate tectonic setting. Several tectonic events in eastem Australia, as described by Totterdell & Korsch (this volume), were possibly related to collisional events, such as the accretion of terranes, at this active plate margin. This scenario continued, althou^ subduction was probably not continuous over this period (e.g. Mortimer, 1995), until late in the Early Cretaceous, when cessation of subduction along this margin occurred at about 100 Ma. The EAP in the Late Mesozoic Pangaea started to fragment and disperse in the Late Mesozoic, and one of the first effects observed in the Eastem Australia Plate was the intrusion of Middle Jurassic dolerites in the Transantarctic Mountains and Tasmania as a result of back-arc extension of the continental cmst. This was followed by further stretching of the continental cmst in the Late Jurassic, with the development of half-graben geometries in basins such as those in Bass Strait (e.g. Hill et al., 1995). This led to the development of significant basin systems along the entire marginfiromthe South Tasman Rise to the Gulf of Papua. The formation and distribution of these basins are discussed by Symonds et al. (this volume). The breakup and dispersal of the Eastem Australia Plate, and generation of new oceanic crust, did not commence until the Late Cretaceous. Sea-floor spreading commenced in the Tasman Sea at about 80 Ma and was time transgressive, with the spreading ridge propagating northwards, and progressive separation of the Lord Howe Rise and New Zealand firom Austraha. The spreading commenced later in the Coral Sea and continued in both the Tasman and Coral Sea basins until the Early Cainozoic, when the ridge system became inactive at about 56 Ma. Thus, the geological development of the Eastem Australia Plate in the Late Mesozoic is related to rifling and drifting mechanisms associated with the breakup of the eastem Gondwanan margin of Pangaea. Mesozoic palaeogeography A series of palaeogeographic maps representing 27 time shces through the Mesozoic were compiled for the whole of onshore Australia (BMR Palaeogeographic Group, 1990). These maps show the distribution of volcanic rocks and sedimentary environments in Australia at the appropriate time, and thus provide a concise summary for the evolution of eastem Australia, even though they show the present-day distributions without palinspastic restoration. The Mesozoic of onshore eastem Australia was dominated by principally non-marine sedimentary basins, containing mainly fluvial to lacustrine deposits, which increased in surface areafiromthe Early Triassic to the early Late Cretaceous. The offshore part of eastem Australia saw the development of extensive rift basins in the Early Cretaceous (Symonds et al., this volume). By contrast, during most of the Mesozoic, New Zealand and New Caledonia were situated at the active plate margin and their sedimentary basins were dominated by clastic marine deposits, including abundant turbidites, derived from the magmatic arc. Magmatic activity In the Triassic, magmatism in eastem Austraha was relatively extensive, but mainly confined to the New England Orogen, where it is expressed as predominantly felsic volcanic and plutonic rocks that are calc-alkaline in character. The magmatic rocks are mostly related to the active subduction margin further to the east, althougji they were emplaced into both extensional and compressional tectonic settings. Variations in the character of the magmatism are a response to variations in the slab geometry and convergence rate (e.g. Gust et al., 1993; Stephens et al, 1993). In New Zealand, the roots of the magmatic arc are preserved in the Median Tectonic Zone (Kimbrough et al., 1994). In the Jurassic, magmatism was principally located to the east of the present Australian coastline, where felsic calcalkaline magmatism was related to a continental margin magmatic arc (firagments of which are preserved in New Zealand). This resulted periodically in a high input of volcanic detritus into the Jurassic sedimentary basins of eastem Australia. Further south, the initiation of continental extension prior to the breakup of Pangaea resulted in the intrusion of tholeiitic dolerite in Tasmania and Antarctica, and rare occurrences in New Zealand. In the Cretaceous, magmatism in eastem Australia was principally confined to the Queensland coastal zone, mainly in the Maryborough Basin and the Whitsunday Islands, although some felsic Cretaceous plutons intruded into the Bowen Basin. The magmatism is considered to be related to the continental extension that was occurring along this margin of the Eastem Australia Plate (Allen & Chappell, 1993). In New Zealand, Early Cretaceous plutons in the Median Tectonic Zone, however, are considered to be part of a continental margin magmatic arc, as is the shghtly younger Separation Point batholith (Kimbrough et al., 1994). Magmatism associated with continental extension commenced in New Zealand relatively late in the Early Cretaceous, apparently shghtly later than in Australia. Distribution and character of sedimentary basins The character of the sedimentary basins in eastem Australia changed markedly through the Mesozoic. In the Early Triassic, the Sydney-Gunnedah-Bowen basin was the dominant depositional system, consisting mainly of fluvial and lacustrine sediments. Tte system developed in an back-arc tectonic setting, with subsidence being driven by foreland loading, due to active thmsting in the New England Orogen to the east. This produced an asymmetric basin that was 565

thickest at its eastern margin and which thinned dramatically towards the west (see Korsch & Totterdell, this volume). Further east, in the continental fragments now rifted from Australia, sedimentation was occurring in various convergent plate margin settings such as in the trench, forearc basin and intra-arc basins, a pattem which continued there until the Late Cretaceous. The westward-propagating thrust front in the New England Orogen reached the present position of the SydneyGunnedah-Bowen basin system in the Middle Triassic, effectively causing upUft and the cessation of sedimentation. An exception to this scenario in eastem Australia is the deposition of marine Early Triassic sediments in the Gympie area, which probably occurred well to the east of the main basin system, because the Gympie Block, at least in part, is considered an exotic terrane that was possibly accreted to the Australian margin in the Middle Triassic. The Late Triassic saw the development of a new depositional system, commencing in the Clarence-Moreton Basin, and in the Early Jurassic eventually extending northwards to form the Maryborough Basin and westwards to form the Surat and Eromanga basins. The principal driving mechanism for the subsidence was most likely thermal relaxation of the lithosphere, resulting in relatively low subsidence rates, in contrast to the extremely high subsidence rates in the Triassic due to foreland loading (see Korsch & Totterdell, this volume). The depositional environments were dominantly fluvial to lacustrine from the Late Triassic throu^ to the Early Cretaceous. Eastem Australia was slowly flooded in the Early Cretaceous, eventually resulting in one of the most extensive marine floodings of the continent seen in the Phanerozoic, with sediments being deposited in a relatively shallow marine environment. In onshore eastem Australia, cessation of sedimentation occurred in the early Late Cretaceous, with uplift and minor contractional deformation occurring within a tectonic regime related to the continental extension prior to sea-floor spreading. These 'mid'-Cretaceous events were important for the maturation, generation and migration of petroleum in the Bowen and Surat basins (see Boreham et al., this volume; Loutit et al., this volume). The continental extension led to the development of extensive Early Cretaceous and younger rift-related basins in the offshore region (see Symonds et al., this volume). In New Zealand, the Triassic to late Early Cretaceous sedimentary basins are principally related to the active plate margin, with arc-derived clastic sediment being deposited in the trench, in trench slope basins ponded above the accretionary wedge, in the forearc basin and in intra-arc basins. The early Late Cretaceous saw the development of narrow continental extensional basins prior to breakup later in the Cretaceous. Non-metallic resources in the Mesozoic of the £AP The sedimentary basins which comprise the bulk of the Mesozoic geology of the EAP host a range of economically important non-metallic earth resources. In Australia, probably the most important of these non-metallic resources is groundwater, particularly that occurring in the Early Cretaceous-Jurassic confined aquifers of the Great Artesian Basin (Habermehl, this volxime). This groundwater, which underUes vast arid and semi-arid regions, underpins an important pastoral industry, as well as supplying many towns and several mining and oil and gas production operations. Most of Australia's commercial oil reserves, up until now, have been produced from the Bass Strait basins, predominantly the offshore Gippsland Basin (Bureau of Resource Sciences, 1996). These basins also produce a significant amount of natural gas, and the hydrocarbons are sourced from, and reservoir in Late Cretaceous and Early Tertiary rocks. Commercial oil and gas accumulations also occur onshore in the Eromanga and Cooper Basins, as well as in the Bowen and Surat Basins. Of historical importance is the Moonie Oil Field in the Surat Basin, which was the first commercial oil field to be discovered on the Australian mainland, in 1961, and has been producing since 1964. In New Zealand, oil and natural gas are produced commercially from Cainozoic reservoirs in both the onshore and offshore Taranaki Basin, although most of the hydrocarbons appear to be sourced from Late Cretaceous coal measures (Geosearch, 1991). Up until now, all of New Zealand's hydrocarbon production comes from the Taranaki Basin. Economically important Mesozoic coal deposits occur in several small basins in both Australia and New Zealand, althou^ the deposits are significantly smaller than the voluminous deposits of Permian coal in the Sydney-GunnedahBowen Basin system. In Queensland, Triassic coal is mined in the CaUide, Tarong and Ipswich basins, whereas Jurassic coal is mined in the Clarence-Moreton Basin, and Cretaceous coal is mined in the Maryborough Basin (Huleatt, 1991). Coal of Triassic age is also mined in Tasmania. In New Zealand, most of the Mesozoic coal deposits occur on the South Island and are Late Cretaceous in age. Significant coalfields include Greymouth (Beamish, this volume), Kaitangata and Ohai (Barry et al., 1994). Important deposits of precious opal are hosted by Cretaceous sediments of the Eromanga and Surat Basins, although they are though by many to have formed during the Tertiary (BMR Palaeogeographic Group, 1990). Metallic resources in the Mesozoic of the EAP Metalliferous deposits occurring in the Mesozoic are almost all hydrothermal in origin and related to calc-alkaline igneous activity in the New England Orogen. Horton (1978) has documented two periods of porphyry copper and related style mineralisation in Queensland, the older Permo-Triassic event extendingfromRockhampton southwards to the NSW 566

border, and the younger Early Cretaceous event situated in central-east Queensland. Of particular interest is the Gympie Goldfield, which produced 116 tonnes of gold, and historically was one of Australia's largest gold producers. It is a Middle Triassic mesothermal gold-quartz vein system (Cunneen, 1994). In the New England Orogen of NSW and Southern Queensland, significant hydrothermal mineralisation is associated with a suite of I-type granitoids of Middle Permian to Early Triassic age (Gilligan & Barnes, 1990). This mineralisation occurs in a variety of forms, including veins, stockworks, pipes, disseminations, greisens and skams, with the main metals being tin, molybdenum, bismuth, tungsten, gold, silver, lead, copper, and zinc. Significant tin production has occurred in the Stanthorpe-Amosfield area (NSW and Qld), the Mole Granite (NSW), and the Tingha-Stannifer-Gilgai areas (NSW). In New Zealand, metalliferous mineralisation similar in style and nature to that of the New England Orogen is associated with extensive I-type Early Cretaceous granitoids of the Westem Province (Brathwaite & Pirajno, 1993). Several goldsilver-polymetallic vein-type deposits are associated with the Separation Point Suite plutons in West Nelson and Westland. At Sams Creek, west Nelson, gold-arsenopyrite-quartz veins and disseminations occur in an altered A-type granite dyke, while small porphyry-type molybdenum deposits occur in Separation Point Suite stocks in west Nelson. Interestingly, the most economically important metalliferous deposit known in the Mesozoic of the EAP is not related to igneous activity. It is the Groote Eylandt manganese deposit in the Gulf of Carpentaria, one of the world's largest know stratiform manganese deposits. It is hosted by Cretaceous sediments of the C^entaria Basin and Bolton et al. (1990) suggest that its formation occurred through the concentration of dissolved manganese during a marine transgression in the late Albian. References Allen, C.M. & Chappell, B.W., 1993. Contrasting Carboniferous-Permian and Cretaceous plutonism in the Urannah Batholith, northem New England Fold Belt. In: Flood, P.G. & Aitchison, JA. (editors). New England Orogen, eastem Austraha, University of New England, Armidale, 573-579. Barry, J.M., Dufif, S.W. & MacFarlan, D.A.B., 1994. Coal resources of New Zealand. New Zealand Ministry of Commerce, Energy & Resources Division, Resource Information Report, 16. BMR Palaeogeographic Group, 1990. Australia: evolution of a continent. Australian Government Pubhshing Service, Canberra, 97 pp. Bolton, B.R., Berents, H.W. & Frakes, L.A., 1990. Groote Eylandt manganese deposit. In: Hughes, F.E., (editor). Geology of the mineral deposits of Austraha and Papua New Guinea. Australasian Institute of Mining & MetaUurgy, Melboume, 1575-1579. Braithwaite, R.L. & Pirajno, F., 1993, Metallogenic map of New Zealand, Institute of Geological and Nuclear Sciences, Monograph 3, 215 pp. Bureau of Resource Sciences, 1996. Oil and gas resources of Australia 1995. Bureau of Resource Sciences, Canberra, 185 pp. Cluzel, D., Clarke, G. & Aitchison, J., 1995. Northem New Caledonia high-pressure metamorphic core complex. From continental subduction to extensional exhumation. In: Mauk, J.L. & St George, J.D., (editors). Proceedings of the 1995 PACRIM Congress. Australasian Institute of Mining & Metallurgy, Publication No. 9/95,129-134. Cunneen, R., 1994. New insights into the Gympie Goldfield. Queensland Department of Minerals and Energy Symposium, Queensland Exploration Potential 1994, Handbook, p. 26. Geosearch, 1991. Petroleum resources of New Zealand. New Zealand Ministry of Commerce, Energy & Resources Division, Resource Information Report, 10, 69 pp. Gilligan, L.B. & Bames, RG., 1990, New England Fold Belt, New South Wales - regional geology and mineralisation. In: Hughes, F.E., (editor). Geology of the mineral deposits of Australia and Papua New Guinea. Australasian Institute of Mining & Metallurgy, Melboume, 1417-1424. Gust, D.A., Stephens, C.J. & GrenfeU, A.T., 1993. Granitoids of the northem NEO: their distribution in time and space and their tectonic implications. In: Flood, P.G. & Aitchison, J.A. (editors), New England Orogen, eastem Austraha, University of New England, Armidale, 565-571. Hill, KA., Finlayson, D.M., HiU, K.C. & Cooper, G.T., 1995. Mesozoic tectonics of the Otway Basin region: the legacy of Gondwana and the active Pacific margin - a review and ongoing research. APEA Joumal, 35,467-493. Horton, D.J., 1978. Porphyry-type copper-molybdenum mineralization belts in Eastem Queensland, Australia. Economic Geology, 73, 904-921. Huleatt, M.B., 1991. Handbook of Australian black coals: Geology, resources, seam properties, and product specifications. Bureau of Mineral Resources, Resource Report, 7,116 pp. Kimbrough, D.L., Tulloch, A.J., Coombs, D.S., Landis, C.A., Johnston, M.R & Mattinson, J.M., 1994. Uranium-lead zircon ares from the Median Tectonic Zone, New Zealand. New Zealand Joumal of Geology and Geophysics, 37, 393-419. Korsch, R.J. & Wellman, H.W., 1988. The geological evolution of New Zealand and the New Zealand Region. In: Nairn, A.E.M., Stehli, EG. & Uyeda, S., (editors). The Ocean Basins and Margins, Vol. 7B, Plenum, New York, 411-482. Meyer, C., 1981. Ore-forming processes in geologic history. Economic Geology, 75th anniversary volume, 6-41. 567

Mortimer, N., 1995. Triassic to Early Cretaceous tectonic evolution of New Zealand terranes: a summary of recent data and an integrated model. In: Mauk, J.L. & St George, J.D., (editors). Proceedings of the 1995 PACRIM Congress. Australasian Institute of Mining & Metallurgy, Publication No. 9/95,401-406. Powell, C.McA. & Li, ZX,, 1994. Reconstruction of the Panthalassan Margin of Gondwanaland. Geological Society of America, Memoir, 184, 5-9. Stephens, C.J., Schon, R.W. & Ewart, A., 1993. Mesozoic crustal extension in the northem New England Orogen: geochemical and isotopic evidence from large scale silicic magmatism. In: Flood, P.G. & Aitchison, J.A. (editors). New England Orogen, eastem Australia, University of New England, Armidale, 637-642. Tissot, B.P. & Welte, D.H., 1984. Petroleum formation and occurrence. Springer-Verlag, Berlin, 699 pp. Veevers, J.J., 1989. Middle/Late Triassic (230±5 Ma) singularity in the stratigraphic and magmatic history of the Pangean heat anomaly. Geology, 17, 784-787. Vickers-Rich, P., Monaghan, J.M., Baird, R.F. & Rich, T.H., 1991. Vertebrate palaeontology of Australasia. Pioneer Design Studio, Melboume, 1437 pp. Walley, A.M. & Ross, M.I., 1991. Preliminary reconstructions for the Cretaceous to Cainozoic of the New Zealand-New Caledonia region. Bureau of Mineral Resources, Australia, Record, 1991/12,43 pp. White, M.E., 1988. The greening of Gondwana. Reed Books, French's Forest, 256 pp. Young, G.C. & Laurie, J.R., 1996. An AustraUan Phanerozoic timescale. Oxford University Press, Melboume, 279 pp.

568

MAJOR EXTENSIONAL EVENTS RECORDED BY ZIRCON XENOCRYSTS FROM THE CENTRAL QUEENSLAND GEMFIELDS J.M. WordenS H. BaadsgaardS D.N, CrackneUS D Krstic^ 1 University of Southern Queensland, Toowoomba, Australia 2 University of Alberta, Edmonton, Canada. SUMMARY Four discrete eruptive periods of volcanicity have been identified in the Central Queensland gemfields based on U-Pb geochronology of xenocrystic zircon accompanying corundum (sapphire). These periods correlate with extensional tectonic events affecting this portion of the Australian plate. Each period is repetitively documented by initial pyroclastic eruptions followed by associated alkali basalts along long established deep crustal fractures. The oldest event centred on 105 Ma is correlated with the cessation of subduction and compressional tectonics along the eastem margin of the Australian plate. A second volcanicity peak is recorded around 88 Ma, the beginning of sea floor spreading in the Tasman Sea. The third period of volcanic activity occurs at the end of the Cretaceous around 66 Ma marking yet another period of extensional tectonics. The youngest volcanic episode at 20 Ma is identified with widespread Tertiary volcanism along the eastem coast of Australia. Only the central portions of xenocrysts were analysed by micro isotopic techniques ensuring that nearly all zircons were concordant, recording no inherited age or significant disturbance of their isotopic systems due to weathering and depositional processes. INTRODUCTION Extensive basaltic lava plain and localised central volcanic activity form the Eastem Highlands of Australia extending from Cape York to Tasmania (Wellman and McDougall, [1]; Johnson et al., [2]). At particular localities ie. the RubyvaleAnakie area of central Queensland and Glenn Innes-Inverell, northem New South Wales, the initial phases of basaltic volcanism were preceded by maar-type pyroclastics which delivered corundum-zircon xenocrysts to the surface (Lishmund and Oakes, [3], Robertson and Sutherland, [4]; Pecover, [5]; and Oakes, Barron and Lishmund, [6]). The corundum-zircon xenocrysts have been concentrated by younger alluvial processes locally generating economic concentrations of sapphire (corundum) that have been worked for decades (Robertson, [7],[8]). The early K-Ar dating by Wellman and McDougall [1] clarified the eruptive ages of the basalts and established that they ranged from approximately 70 Ma to Recent. Subsequently, Sutherland, ([9], [10], and [11]), proposed that this volcanism resulted from the northward migration of the Australian plate awayfiromAntarctica and its passage over mantle hot spots during the Cenozoic. Stephenson, [12] has also confirmed the ages of certain volcanic features in the Hoy volcanic province previously described by Veevers et al., [13]: Stephenson, [14] Robertson, [8]; Stephenson et al., [15]; Robertson and Sutherland, [4]; Duffy, [16]; Grimes and Withnall, [17] and Withnall et al., [18]. Numerous volcanic plugs and associated small basalt flows occur throughout the area but generally within a north-northeast trending corridor (Stephenson et al., [15], Robertson and Sutherland, [4] These features have been collectively referred to as the Hoy basalt (Veevers, et al.,[13]). While some form prominent cone-shaped hills, others outcrop as low mounds that have been reduced by prolonged erosion to barely perceptible circular to elliptical shaped outcrops (Robertson, [8]). The majority of the volcanic plugs are comprised of basanite with lesser occurrences of alkali basalt and hawaiite (Stephenson et al.,[15], Stephenson [12]). They range from 56 Ma (Policeman Knob near Rubyvale) to 17.9 Ma (Stephenson, [12]) and the concentration of plugs along the north-northeasterly lineaments indicates the existence of deep crustal structures (Stephenson et al., [15]). The existence of xenocrysts within the plugs was first noted by Dunstan [19] who recorded pleonaste 'embedded in the basalt of Policeman Knob' and at both Mt Hoy and Mt Leura where it and other xenocrystic phases 'show the effects of corrosion'. The source of the xenocrystic zircon and corundum has been proposed as lower crustal or upper mantle (Stephenson, [14], Quo et al., [20]) and at all crustal levels recently by Oakes et al., [6]. Zircon is a preferred mineral for geochronology since it possesses a high closure or "blocking" temperature of >900°C (Chemiak et al., [21]; Mezger et al., [22]; Ames et al., [23], Vry et al., [24] and Zhang and Scharer, [25]). Many instances of zircon inheritance have been reported, however at basaltic magma temperatures of between 1100-1200°C, zircons will be reset to the emptive event. Zircon xenocrysts which are intimately associated with corundum (Coenraads et al., [26]), will indicate the age of the associated alkaline basaltic volcanism that delivered the xenocrystic phases to the earths surface. This paper reports the results of U-Pb isotopic investigations of seventy seven (77) zircon xenocrystsfiromthe RubyvaleSapphire area of Central Queensland and discusses their possible tectonic significance.

569

METHODOLOGY Zircon xenocrysts are recovered together with sapphires (corundum) during mining in the central Queensland gemfields. The bulk of the recovered corundum and zircon is extracted from re-worked alluvials termed 'washes' (Robertson, [8]) along the existing and paleo-drainages of Policeman Creek and Retreat Creek. Very rarely, 'in situ' weathered pyroclastics are encountered during mining activities, however, none were accessible during this study. Individual miners and companies generously provided the authors with access to their production and donated collections of representative zircons to this study. The zircon xenocrysts were initially classified by colour into six (6) categories namely; white or colourless, champagne, yellow, brown, pink and red. Not all colours were represented at each locality, however, every effort was made to ensure that each bulk zircon sample was 'colour inclusive' and representative. A single zircon xenocryst was selected from each colour group at all localities that met the criteria of beingfreefromvisible cracks andflaws,inclusions, visible cores and overgrowths under microscopic observation. Most zircon xenocrysts were less than 1 cm in diameter, displayed subdued crystal forms and magmatic resorption, and not infrequently re-entrants from other primary phases. Rarely so-termed 'vent pohshed' zircon xenocrysts were encountered and analysed. In the laboratory, single zircon xenocrysts were gently broken in a HN03 cleaned agate mortar in such a manner that permitted access to a thin sliver of "inside" zircon including the core of the xenocryst. The extracted "sliver" (=30 mgms) was then ground to an impalpable powder in the agate mortar. Between 10-20 mgms was then processed following the analytical method of Manton [27] and Duke [28]. Thefinalspiked separation was loaded onto silica gel on an outgassed rhenium single filament. Pb blanks varied from 12-29 pg with the common range between 14-18 pg while the U-blank is <lpg. Stacey-Kramers common lead of the appropriate age was employed for the common lead corrections. RESULTS U-Pb age data for seventy-seven zircon xenocrysts are summarised in Table 1. Complete analytical data will be reported separately. TABLE 1. Simimarised U-PB ages for Zircon Xenocrystsfromthe Rubyvale-Sapphire area. Central Queensland Zircon xenocrystic colour variants U-PbAges Brown Pink Centr White/Colourless Champagne Yellow Red Interval Tendency _ 2 13 102-1 lOMa 3 6 105Ma 1 8 7 83-90Ma 88Ma 2 1 5 3 3 61-70Ma 66Ma 14 3 3 3 17-28Ma 20Ma 24 28 3 5 6 11 Data points are virtually concordant for practically all zircon xenocrysts and clearly define four discrete age periods or intervals. These age intervals are 102-110 Ma, 83-90 Ma, 61-70 Ma and 17-28 Ma, of which the early three are Mesozoic. The oldest interval (102-110 Ma) includes all colour variants, however, the predominant colour is champagne (13 xenocrysts), followed by brown (6), white (3) and red (2). This oldest group occur in both Policeman and Retreat Creeks paleo-drainages and blanket the region, although they constitute a greater proportion of analysed xenocrysts within the Reward and Rubyvale designated mining areas. This is possibly a reflection of sampling bias directed towards primary washes and remnant weathered pyroclastics which directly mantle basement lithologies and host corundum (sapphire) deposits. Their widespread distribution is due to prolonged erosion and fluvial transport along eastward directed drainages. Altematively, the original maar-related pyroclastic eruptions were both violent and empted from multiple vents throughout the area. Recent aeromagnetic surveys of the region have confirmed the existence of many more basic vents which are structurally controlled than formerly recognised (M.Duffy - pers. comm. 1995). Robertson [8] noted that corundum (sapphire) xenocrysts are coarser grained in the Reward designated area and that their size diminishes further eastwards. Volcanism was clearly episodic spanning the 99-116 Ma interval with a peak at approximately 105 Ma. A resurgence of explosive volcanism occurred between 78-96 Ma indicating continuous eruptions over a prolonged interval or geological event. White and Champagne colour varieties dominate the zircon xenocrysts of this interval, and are more abundant along the Policeman Creek paleo-drainage indicating more northerly vents. Very low U concentrations (<30 ppm) are indicative of kimberlite-like sources accessed during this volcanic episode (Heaman et al, [29]). A third zircon xenocrystic group clusters around 66 Ma and the close of the Mesozoic, covering an interval from 61-70 Ma. Zircon xenocrysts are more tightly grouped around 66 Ma than in the prior grouping and are noticeably better represented in Policeman Creek paleo-drainage. While white zircon is conspicuously absent, all other more highly coloured types are almost equally represented. 570 —

—

The youngest group cluster around 20 Ma but range from 12 Ma to approximately 28 Ma. These zircon xenocrysts are highly coloured varyingfrompink through brown to red, however, the majority are brown. They are areally widespread and noticeably more common in the lower reaches of Policeman Creek and the Subera mining leases. Since these zircon xenocrysts were the last erupted, they are found in predominantly shallow 'washes' in low-lying areas. A very large percentage of the analysed zircon xenocrysts have low U contents and strongly suggest a mantle or kimberlitic origin for the zircon (Heaman et al., [29]). As the eruptive sequence or interval progressed, zircon xenocrysts become progressively more coloured and higher in uranium. This trend is very evident in the two youngest zircon xenocrystic groupings and still discemible within the older zircon groupings. With the waning of eruptive episodes, the possibilities for cmstal contamination are enhanced. DISCUSSION A unique continental record of episodic basaltic volcanism is preserved in zircon xenocrysts despite continuous erosion and fluvial processes removing the contemporaneous basalts and pyroclastics. The resistant zircon xenocrysts have accumulated locally in surficial 'washes' and along paleo-drainage channels (Robertson, [8]). Thus, xenocrystic zircons have preserved a 'tectonic footprint' retaining evidence of eroded volcanic episodes resultingfromthe evolution of the Australian plate stress field (Pilger, [30]). Extensional periods relaxed deep crustalfractureswhich were then used as conduits for the eruption of intraplate alkali basalts (Stephenson et al., [15]). Zircons which had crystallised from C02 rich, silica undersaturated partial melts within subduction enriched mantle (i.e. fertile lithosphere) were caught up by rising magmas and rapidly transported through the crust to be erupted as xenocrysts (Guo et al., [20]). The very low uranium contents of the xenocrystic zircons indicate a kimberlitic afSnity rather than a crustal derivation, a conclusion additionally supported by their depleted trace element character (Worden, Cumming, Harlow and Krstic-in preparation). By 105-110 Ma ago, plate convergence and subduction which characterised the eastem Austrahan margm, ceased in westem New Zealand and was replaced by an extensional regime (Bradshaw, [31]; Gibson and Ireland, [32]; and Luyendyk, [33]). Closer to Queensland, the Whitsunday volcanics were erupted between 95-132 Ma (with the main activity between 105-120 Ma) during extension of the eastem Australian plate margin prior to the breakup and opening of the Tasman Sea (Ewart et al., [34], Parianos et al., [35] and Stephens et al., [36]). We propose that the oldest zircon xenocryst grouping marks the onset of basaltic volcanismfromdeep-seated fractures that developed in response to magmatic underplating and the onset of extensional rifting. Cretaceous volcanism of this age is also present in the Maryborough basin of South Queensland, the Lord Howe Rise and New Caledonia, confirming widespread volcanism during the dismemberment of the Eastem Australian continental margin (Ewart et al., [34], Robertson, [37]). The initial opening of the Tasman Sea began around 95 Ma and continued to 82 Ma, followed by rapid sea floor spreading to 57.5 Ma (Jones and Veevers, [38]). Spreading was accompanied by the formation of the rift valleys along the northeastem margin of the Australian plate (Veevers, [39]). The second grouping (Table 1) of zircon xenocryst ages centred on 88 Ma and rangingfrom83-90 Ma, has a poorly developed maximum and is suggestive of a continuous period of volcanism in accord with an extensional regime during opening of the Tasman Sea. Coral Sea spreading was of shorter duration, taking place between 63.5-57.5 Ma and terminating sea floor spreading in the Tasman Sea-Cato Trough-Coral Sea regions (Veevers, [39]). The third zircon grouping (Table 1) correlates closely with this event with a pronounced peak at 66 Ma. The relatively narrow age range displayed by these zircons lends support to the theory that they are a consequence of a brief geological/tectonic event. The time period between the third and final youngest zircon groupings is remarkable for its quiescence with only one xenocryst recording an age of 34 Ma. Widespread Cenozoic uplift that accompanied the formation of the Eastem Australian Highlands was accompanied by widespread basaltic volcanism (Welknan and McDougall 1; Veevers, [39]). The youngest zircon xenocryst grouping spans an interval of 17-28 Ma with a solitary xenocryst recording 12 Ma. The early Miocene is well known as a period of widespread basaltic volcanism throughout Central and Southem Queensland and is clearly documented by zircon xenocrysts. The repetitive basaltic volcanism and associated pyroclastics of the Rubyvale area of Central Queensland are correlated with deep crustal fracture zones that respond to the prevailing stress regime in the Australian plate. That volcanism occurred during four essentially discrete periods within the same area does not tend to support a 'hot spot model' but rather deep crustalfracturesthat respond to the prevailing tectonic regime. CONCLUSIONS Zircon xenocrystsfromthe Rubyvale-Sapphire area of Central Queensland record four separate and discrete intervals of basaltic volcanism. The oldest group reflect the change from convergent tectonics to an extensional regime along the eastem Australian margin. The second and third xenocrystic zircon groupings document sea floor spreading in the Tasman and Coral Seas. The youngest zircon grouping are a response to early Miocene uplift and magmatic underplating 571

along the eastern Highlands region. The repetitive character of volcanism lends support to the concept of deep crustal fractures responding to the prevailing tectonic regime during extensional periods. Xenocrystic zircons have potential as 'tectonic footprints' for basaltic volcanism induced by plate-wide tectonic processes, long after the basalts have been removed by erosion. ACKNOWLEDGMENTS The authors gratefully acknowledge the friendly and generous donation of both time and recovered zircon xenocrysts by the many individual miners and companies active in the central Queensland gemfields. Without their assistance this investigation would not have been possible. Michael Duffy of Great Northem Mining Corporation contributed to this study with his local expertise and exploration samples from the Subera leases. We tha^ Great Northem Mining Corporation for their cooperation. Dr A.D. Robertson also donated san:5)lesfromFreehold and the Scrub Lead and his discussions on the gemfields are gratefully acknowledged. The authors wish to acknowledge support by the Universities of Alberta and Southem Queensland. REFERENCES L Wellman, P., and McDougall, I., Cainozoic Igneous Activity in Eastem Australia, 1974, Tectonophysics, [23], p.49-65. 2. Johnson, R.W., Knutson, J. and Taylor, S.R., eds., Intraplate Volcanism in Eastem Australia and New Zealand, Cambridge University Press, Cambridge, 1989,408 p. 3. Lishmund, S.R. and Oakes, G.M., Gemstones. In Johnson R.W., Knutson, J. and Taylor, S.R., eds., Intraplate Volcanism in Eastem Australia and New Zealand, Cambridge University Press, Cambridge, 1989, p. 152-155. 4. Robertson, A.D.C. and Sutherland, F.L., Possible Origins and Ages for Sapphire and Diamond from the Central Queensland Gem Fields. In Sutherland, F.L. and Chahners, R.O., eds. Commemorative Papers (Mineralogy, Meteoritics, Geology), Records of the Australian Museum Supplement, [15], 1992, p.45-54. 5. Pecover, S.R., The Geology and Mining of the Strathdarr Sapphire Deposit, New England Gem Fields, North Eastem New South Wales, In Flood, P.G. and Aitchison, J.C., eds. New England Origin, Eastem Australia, Department of Geology and Geophysics, University of New England, Armidale, 1993, p. 493-503. 6. Oakes, G.M., Barron, L.M. and Lishmund, S.R., Alkali Basalts and Associated Volcaniclastic Rocks as a Source of Sapphire in Eastem Australia. Australian Journal of Earth Sciences, 1996, [43], p.289-298. 7. Robertson, A.D., The Geology and Sapphire Deposits of the Anakie Mineral Field, Geo logical Survey of Queensland (unpubhshed report) 1975. 8. Robertson, A.D., Notes on the Geology of the Central Queensland Sapphire Fields, Geological Survey of Queensland, 1983, Record 1983/51. 9. Sutherland, F.L., Cainozoic Volcanism, Eastem Australia: A predictive model based on Migration over Multiple 'Hot Spot' Magma Sources, Geological Society of Australia special Publication, [18], 1991, p. 15-43. 10. Sutherland, F.L., Late Thermal events based on Zircon Fission Track Ages in Northeastem New South Wales and Southeastem Queensland: Links to Sydney Basin Seismicity?, AustraUan Journal of Earth Sciences, 1993, [40], p.461470. 11. SutherlandJF.L., Alkaline Rocks and Gemstones, Australia: A review and synthesis, Australian Journal of Earth Sciences, 1996,43, p.323-343. 12. Stephenson, P.J., The Geological Context of Sapphire Occurrences in the Anakie Region, Central Queensland, Geological Society of Australia Abstracts Series [25], 1990,p.232-233. 13. Veevers, J.J., Mollan, R.G., Olgers, F. and Kirkegaard, A.G., The Geology of the Emerald 1:250,000 Sheet Area Queensland , Bureau of Mineral Resources, Australia, 1964, Report 68 14. Stephenson, P.J., Sapphire and Zircon in some Basaltic Rocks from Queensland, Australia, Abstracts of the 25th International Geological Congress, Sydney, 1976, p. 602-603. 15. Stephenson, P.J., Sutherland, F.L., Robertson, A.D. and HoUis, J.D., Hoy, In Johnson R.W., Knutson, J. and Taylor, S.R., eds. Intraplate Volcanism in Eastem Australia and New Zealand, Cambridge University Press, Cambridge, 1989, p. 101-103. 16. Duffy, M., Subera Mine, An Alluvial Sapphire Deposit, Central Queensland Gemfields, In Withnall, I.W., ed. Clermont-Anakie Region Central Queensland, 1995, Field Conference Report, Geological Society of Australia, Queensland Division, 1995, p.102-103. 17. Grimes, K.G. and Withnall, I.W., Cainozoic Geology of the Anakie-Rubyvale Area, In Withnall, I.W.,ed. Qermont-Anakie Region of Central Queensland, 1995 Field Conference Report, Geological Society of Australia, Queensland Division, 1995, p.64-69. 18. Withnall, I.W., Blake, P.R., Crouch, S.B.S., Tennison-Woods, K., Grimes, K.G., Hayward, M.A., Lam, J.S., Garrad, P., and Rees, I.D., Geology of the Southem part of the Anakie Inlier, Central Queensland, Queensland Geology, 1995, [7]. 19. Dunstan, B., On the Sapphire Fields of Anakie, Geological Survey of Queensland, 1902, Publication 172. 572

20. 21. 22.

23.

24.

25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

35.

36.

37. 38.

39.

Guo, J., O'Reilly, S.Y. and Griffin, W.L., Corundum from Basaltic Terrains: A Mineral Inclusion Approach to the Enigma, Contributions to Mineralogy and Petrology, 1996, [122], p.368-386. Chemiak, DJ., Lanford, W.A. and Ryerson, F. J., Lead Diffusion in Apatite and Zircon using Ion Implantation and Rutherford Backscattering Techniques, Geochimica et Cosmochimica Acta, 1991 [55], p. 1663-1673. Mezger, K., Rawnsley, C.M., Bohlen, S.R. and Hanson, G.N., U-Pb Garnet, Sphene, Monazite and Rutile Ages: Implications for the Duration of High-grade Metamorphism and Cooling Histories, Adirondack Mts. New York, Journal of Geology, 1991, [99], p.415428 Ames, L., Zhou, G. and Xiang, B., Geochronology and Isotopic Character of Ultra Hi^-Pressure Metamorphism with Implications for Collision of the Sino-Korean and Yangtze Cratons, Central China, Tectonics, 1996, [15], p.472-489. Vry, J., Compston, W., Cartwright, I., Shrimp II, Dating of Zircons and Monazites: Reassessing the Timing of High-grade Metamorphism and Fluid Flow in The Reynolds Range, Northem Arunta Block, Australia, Journal of Metamorphic Geology, 1996, [14], p.335-350. Zhang, L-S., Scharer, U., Inherited Pb Components in Magmatic Titanite and their Consequence for the Interpretation of U-Pb Ages, Earth and Planetary Science Letters, 1996, [138], p.57-65 Coenraads, R.R, Sutherland, F.L., and Kinney, E.D., The Origin of Sapphires: U-Pb Dating of Zircon Inclusions Shed New Light, Mineralogical Magazine, 1990, [54], p.l 13-122. Manton, W.L, Separation of Pb from young Zircons by Single-Bead Ion Exchange, Chemical Geology (Isotope Geoscience Section) 1988, [73], p.147-152. Duke, M.J.M., The Geochronology and Isotope Geology of the Type-Nuk Gneisses of the AikaTerrane, Southem West Greenland, Unpublished PhD Thesis, University of Alberta, Edmonton, Canada, 1993,286 p. Heaman, L.M., Bowins, R. and Crocket, J., The Chemical Composition of Igneous Zircon Suites: hnphcations for Geochemical Tracer Studies, Geochimica et Cosmochimica Acta, 1990, [54], p.1597-1607. Pilger, RH., The Origin of Hot Spot Traces: Evidence from Eastem Australia, Journal of Geophysical Research, 1982, [87],B3,p.l825-1834. Bradshaw, J.D., Cretaceous Geotectonic Patterns in The New Zealand Region. Tectonics, 1989, [8], p. 803-820. Gibson, G.M. and Ireland, T.R., Granulite formation during Continental Extension in Fiordland, New Zealand, Nature, 1995, [375], p. 479-482. Luyendyk, BJP., Hypothesis for Cretaceous Rifting of East Gondwana caused by Subducted Slab Capture, Geology, 1995, [23], p. 373-376. Ewart, A., Schon, RW. and Chappell, B.W., The Cretaceous Volcanic-Plutonic Province of the Central Queensland (Australia) Coast - a Rift Related "Calc Alkaline" Province. Transactions of the Royal Society of Edinburgh; Earth Sciences, 1992,83, p.327-345. Parianos, J., Bryan, S., Ewart, A., and Schon, R.W., Early Cretaceous Rift Volcanics of the Central Queensland Coast, In Flood, P.G. and Aitchison, J.C., eds. New England Orogen, Eastem Australia, Department of Geology and Geophysics, University of New England, Armidale, 1993, p. 655-663. Stephens, C.J., Schon, RW. and Ewart, A., Mesozoic Crustal Extension in the Northem NewEngland Orogen: Geochemical and Isotopic Evidence from Large Scale Silica Magmatism, In Flood, P.G. and Aitchison, J.C., eds. New England Orogen, Eastem Australia, Department of Geology and Geophysics, University of New England, Amiidale, 1993, p. 637-642. Robertson, A.D.C., The Bundaberg Volcanic Province: Field Relations, Petrochemistry and Tectonic Setting, Unpublished PhD Thesis, University of Queensland, 1992,310 p. Jones, J.G., and Veevers, J. J., Morphotectonics of the Platform Regions, focused on the Highlands, In Veevers, ed. Phanerozoic Earth History of Australia, Oxford Geological Sciences Series, Clarendon Press, Oxford, 1984, [2], p. 115-142. Veevers, J.J., Morphotectonics of the Divergent or Rifted Margins In Veevers, ed. Phanerozoic Earth History of Austraha, Oxford Geological Sciences Series, Clarendon Press, Oxford, 1984,[2] p. 201-210.

573

SEDIMENTOLOGY OF THE MIDDLE JURASSIC WALLOON COAL MEASURES IN THE GREAT ARTESIAN BASIN, EASTERN AUSTRALIA Joel V.IL Yago and Christopher R. Fielding Department of Earth Sciences, University of Queensland, Qld 4072, Australia Summary The Middle Jurassic Walloon Coal Measures were deposited on a vast alluvial plain crossed by mainly meandering rivers of varying dimensions, which were separated by extensive shallow lakes and wetlands. The formation as a whole, andfivecomponent members defined on lithological criteria, can be traced firom the Clarence-Moreton Basin as far west as the Eromanga Basin. The unit is interpreted to have formed by mainly passive, thermal subsidence, with a north-south depocentre in the Clarence-Moreton Basin possibly recording a failed (incipient) rift. Introduction The Middle Jurassic Walloon Coal Measures and their stratigraphic equivalents the Mulgildie Coal Measures, lower Injune Creek Beds, Birkhead Formation occur in the Clarence-Moreton, Surat, Mulgildie, and Eromanga Basins, all of which are part of the Great Artesian Basin system. Further likely stratigraphic equivalents are preserved in the upper part of the Tiaro Coal Measures in the Maryborou^ Basin of coastal SE Queensland. The coal measures comprise an erosionally truncated succession of interbedded clastic sedimentary rocks and coals totalling up to 920 m in thickness. Using data firom examination of natural outcrops, open pit mines and boreholes, a basin analysis of the Walloon Coal Measures has been undertaken. This paper summarises the majorfindingsof the study, documented in full in Yago (1). Results A total of fifteen lithofacies has been recognisedfiromthe data set as a whole. Analysis of surface exposures has allowed recognition of nine architectural elements. The facies and element assemblage is interpreted as deposits of an extensive alluvial plain which was crossed by mainly meandering streams of varying dimensions. These rivers were separated by shallow floodbasin and discontinuous peat-forming wetland environments that were continually affected by overbank sand deposition and reworking by channel migration and avulsion. Resulting peat (coal) deposits exhibit a sheet-like architecture locally, but a complex geometry on a regional scale. Periodic volcanic eruptions showered ash across much of the basin. Ash was preferentially preserved in mire and distal floodbasin environments. This supports the notion that volcanic activity was widespread in eastem Australia during the Jurassic. Volcanic centres may have been intra-basinal (Fig. 1). Palaeocurrent data indicate the dominant sediment dispersal direction during accumulation of the Walloon Coal Measures was westerly,firomthe Clarence-Moreton Basin towards the Surat and Eromanga Basins (Fig. 1). Palaeocurrent pattems and the lack of lithological variation towards present outcrop edges indicate that the Walloons originally covered a considerably greater area than their preserved extent. Outliers in the Mulgildie and Tiaro regions of SE Queensland are interpreted to have been contiguous with the Great Artesian Basin, and interpreted sediment dispersal pattems in the NSW portion of the Clarence-Moreton Basin indicate original lateral continuity with the southem Surat Basin. The coal measures and their stratigraphic equivalents accumulated in a slowly, passively subsiding intracratonic basin considered to have been formed by thermal relaxation of crust attenuated by extension during the Late Triassic. Lithostratigraphic correlation indicates the lateral persistence of five lithologically distinct members (Intervals 1-5: Fig. 2) which are stacked in a "layer-cake" arrangement over the entire Great Artesian Basin. It is suggested that the principal controls on stratigraphic architecture were the passive style of subsidence, and variable rates of coarse sediment supply. A north-south elongate depocentre in the east that covers much of the Clarence-Moreton Basin coincides approximately with a zone that encompasses the greatest concentration of volcanic ash deposits, possible coeval volcanic centres, eruptive and intrusive rocks. This, and the bimodal mafic/felsic composition of the igneous rocks, suggest that the northsouth belt may represent the site of a failed, incipient rift. No evidence was found to support previous models that interpret the Clarence-Moreton and Surat Basins as a foreland system, into which sediment was delivered fi'om an active volcanic arc. Reference 1.

Yago, J.V.R., Basin Analysis of the Middle Jurassic Walloon Coal Measures in the Great Artesian Basin, Australia, PhD Thesis, University of Queensland, June 1996,274pp.

574

Limits of present day Walloon Coal Measures outcrop belt (eroslonal truncation)

V C - Volcanic centres (known and suspected) Arrows indicate sediment dispersal direction

Figure 1. Schematic reconstruction of the Great Artesian Basin during accumulation of the Walloon Coal Measures. WEST

EAST e u M C M c a - MOWTOM M S M

L o c a t i o n Map

Figure 2. Lithostratigraphic correlation of the Walloon Coal Measures (Intervals 1-5) across the Great Artesian Basin. Wells A and B lie within a major, north-south elongate depocentre.

575

Learn more about exploration in Queensland.... ....consult the Department of Mines and Energy for: geoscientific data commodity information mineral and energy resources land availability mining safety and environmental requirements and everything for minerals exploration and development The Department of Mines and Energy, Queensland QMEC Building 61 Mary Street Brisbane 4000 PO Box 194 Brisbane 4001 telephone (07) 3237 1436 - facsimile (07) 3221 9517