Landslide hazard zonation and evaluation using GIS and remote sensing: A Review

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 11 | Nov 2022 www.irjet.net p-ISSN:2395-0072

Landslide hazard zonation and evaluation using GIS and remote sensing: A Review

¹Dep. of Environmental Engineering, Civil College, UP, India Sharda University, Greater Noida, Knowledge Park 3 ***

Abstract - Landslide is a significant natural hazard in mountainous terrain countries all around the world. Such disasters cause for the loss of hundreds of millions of dollars and thousands of death to all record each year. Scientific studies will help to minimize risks posed by sudden landslide hazard occurring in different corners of the world. The objective of this review focus on the study of methodologies used in previous research articles related to landslide hazard zonation. Geological maps were used as a source for lithology, soil and land use was processed from Landsat +ETM satellite. Slope and elevation were derivedfromDigitalElevationModel (DEM) collected from ASTER satellite. LHZ maps were generated using a raster calculation. Results categorize study areas into four compartments, based on their raster calculation processed using GIS environment. 30% of study areas under the review articles were found a no hazard area and only 21% were considered as high hazard.

Keywords; Landslide, hazard zonation, Geographical information system, Remote sensing

1. INTRODUCTION

Any disaster threat, which hamper natural process and accompaniedbydeleteriouseffectonhumansandecosystem canbeconsideredasanaturalhazard.Naturalthreatscanbe dividedbasedontheirtypesandeffectonthebiosphereand naturalecosystem.Thisincludegeological,whichishighly deteriorating the earth surface, water related disaster (hydrological) and metrological hazards are among the classifications.Manygeologicaldisastersarefeltwhenthey causeamassiveeconomicalloss.LandslideHazardZonation (LHZ)dividethegeographicallocationsintocompartments basedontheirexposuretolandslideincidentbystudying crucialdrivingfactorsforinsatiability[1]

Weather related hazards predominantly wind and heat driven disasters are the most common ones under the meteorologicaldisasters[2].Thisreviewarticleconstituent related research articles made since 1990 up to current statusi.e.2018,highlightingtheslowdevelopmentsofLHZ techniques and mechanism to map land slide vulnerable areas.

Heat waves affect the USA and south Asian countries frequentlyhurricanesandfreezingrainarealsoconsidered as metrological hazards. Landslide is the most popular

hazardous disaster which is caused by natural and anthropogenic reasons in many countries [3]. Massive deforestation, excessive overgrazing, quarrying, resource wastage are the main man made causes for a landslide hazardousdisaster[4].Itisdifficulttoforecastandpredict mostofthenaturalhazardousdisastersbutit'spossibleto study the pattern and deduce and design preventive measures. Even though is not possible to eliminate, if the hazard is well understood and appropriate measures are takenbytheconcernedbody,themagnitudeofthelossboth economicallyandnaturallycanbeminimized[5].Previous studieshavebeendoneinassessingtheriskofthehazard anddevelopmechanismstominimizetheriskandcopeup withthedisastrouseffect.Itispossibletodevelopanaction plan and minimize the risk if appropriate studies and informationcollectedbeforethedisasterhappen[6].Such risk minimization strategy is called landslide hazard zonation,whichisaccompaniedbyofdifferentmethodsand analysis.

Landslidedisasterreferstoanymasswastingincludingwide rangeofgroundmovement,deep-seatedrockfailure,rock falls,debrisflowsandmudflows.Therearedifferentdriving factorsforthisincidentlike,slopeimbalance,frommountain tocoastalareasorevensubmarinelandslides[7].Gravityisa primaryfactorforlandslideincludingotherreasonswhich cause slope instability, which expose slope to lose its maintenance. Trigging factors may be also heavy rainfall whichwashawaysoil ofthe certaingeographical location andmaketheareaspronetodisasterdecreasingtheshear strengthofslopematerialwillexacerbatethesuitabilityof theareastomorevulnerablecondition[8].Thefollowingare amongthenaturalcausativefactorsforlandslide

I. Glacierevaporationandmeltingdown

II. Availabilityofmorethanadequateground waterinacertainareas

III. Hydrostaticpressureonearthfractures

IV. Poor vegetative coverage and inadequate soilmanagement

V. Soilerosionbyriversandwavesofwater

VI. Chemicalpollutionofgroundwater

VII. Massiveearthquakesdestabilizingslopeby soilliquefaction

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 11 | Nov 2022 www.irjet.net p-ISSN:2395-0072

VIII. Volcanicdisasters

Heavyprolongedrainfalltakethelion'squotientforcausing landslide disasters passage of cyclones, tsunami, and thunderstormarealsodrivingreasons.Duration,intensity and chemical composition of the rain fall will highly determinethemagnitudeofthelandslide[9]

1.1 Land slide prevention mechanisms

It's possible to prevent the land slide hazard by following simplestrategiesincluding:

Plant vegetation:Plantingsmallshrubsandplantsonthe slope is an alternative mechanism. As plants and shrubs grownontheslopetheirrootholdtogetherandpreventthe soilerosionwhichinturnhinderlandslide.Someselected plantspecieswithagoodabilityofholdingsoiltotheirroot canbeselectivelyplantedincampaign.

Retaining Walls: Retaining walls are also a good mechanism,Properdrainagefacilitybyconstructingwooden orconcrete retaining wallswill savesoil erosion which in turnhinderlandslideincidence

Preventing soil erosion: building gutters will prevent landslidebydivertingwaterdischargewhichisapotential soilerosionfactor.Ifdivertingthewaterisnotsuitablefor theconditionitispossibletobuildsmalldamsandminimize thespeedofthewaterrunoffovertheslope.Asvelocityof water on the slope determine the amount of soil eroded, constructing small dams will help to save eroded soil by minimizingthespeedofthewaterrunoff

Altering the Slope Gradient: alteringtheslopegradientof the vulnerable areas is a simple tricky mechanism to minimizelandslidedamages.

CertaindeterminationfactorsneedtobeconsideredinLHZ process. Lithology is the main factors causing landslide, which the rock type of the susceptible area influencing tremendouslythehazardoccurrence.Thesehelpstopredict the potential effects and to reduce the probability of occurrence by mapping the hazardous area. Chauhan confirms that the occurrence of landslide happened in highlandareasandtheinfluenceofman-madeactivitiesare oneofthepotentialcausesofthesedisasters.Theeffectof hydrologicalprocessesisalsothemainprominentfactor.

Geo and Lee reported that climate, precipitation, bedrock and soil conditions along with slope highly influence landslideoccurrence,andtheareasaffectedinthepastwill haveamorequotientincausingsimilardisastroushazard [10].Intheinventorydataanalysisthepapersidentifies50 differentpastlandslidesfromfieldstudiesarecollectedand thesedataareprocessedbythehelpofGPS[11].Landslide hazard zonation is worked in different highland areas because most of the methodology of landslide hazard

zonation technology show advancement with technology. The nature of the susceptible area determines the risk assessment approach because parameters causing the landslide will not be the same in all risky geographical locations[12].Themechanismtocopeupwiththeriskmay bevaryingaccordingtothekeyfactorsthataredrivingthe disaster.Appropriatestudymechanismwillfollowdistinct methodologiesinstudyinglandslidezonation.Eventhough mostofthelandslideoccurrencesharessimilarparameters, only one determinant factors may not be applicable in landslide zonation [13]. Approaches in landslide zonation canbecategorizedunderdirectandindirectmethodologies. The direct method may consist inventory landslide and heuristic methodologies. In direct methodology, the expertise and cartographer on the areas determine the decision unlike the indirect one. Heuristic methods are focusingonquantitativeandnumberbasedanalysisbased on landslide inventory to identify the geological and geomorphologicproperties.Mostofthestudiesusedifferent methodstounderstandthepatternofpreviouslyrecorded landslide. These methods are; GIS based statistical & probabilityapproach,inventoryofpastlandslide,bivariate approach,deterministicapproaches[14].Intheseapproach all the methodologies are used based on the 6 causative factorscalculationofhazardindexequations.Asthestudies describesthatallthedataaretakenfromtopographicaland satelliteimages,metrologicalreports,digitalelevationmodel data, Google earth images and the DEM 30m resolution utilizedtoextractslope,aspect,elevationofthestudyarea byusingASTERsatellite.

2. Hazard mapping /Assessment Techniques/ 2.1 Statistical Approach

[15]KomacandZornreportsthat,theexposureofcertain area for the landslide in the past is highly influential in determining the occurrence for the future. Statically approachuseslandslidefrequencyandGISasthemaininput todeducetheoccurrence.Thebasicprincipleisthatbased onpreviouslyrecordedlandslidesanditwill behelpful to guess its occurrence scientifically. Variables causing landslides will be applied in calculating and analysis of landslide hazard zonation. There are wide ranges of multivariatetechniquesavailable,asmaybeseenfromthe different statistical method and factor analysis is done by usingdatasqueezingwhichcompresslargefilestosmaller, moremanageable,numberoffactors. [10]Thesestudyuses multivariate statistical methods of logistic regression methods and probabilistic methods, there are 14 probabilistic factors and 9 logistic regression factors are takenandcalculatingfromdifferentdatabases.Forlogistic regressionmethodsthedependentvariablesmustbeinput as0or1.Factorsanalysiswillhelptodistinguishthemost important factors from the junk data set (Ransom 1991). Thismethodsarethemostcommonapproacheswhichare usedindifferentliteratures[11] Showedthatbyusingthis

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 11 | Nov 2022 www.irjet.net p-ISSN:2395-0072

approaches it is possible to spot the areas on the map to easilyunderstandandstudytheeffectofdifferentlandslide causing factors like, slope, elevation, curvature, land-use, ground water surface traces. By using those causative factors, it uses the tool raster calculator in ArcGIS, total pixels occupied by each factor also identified. [13] These studies also identifies different methodologies that has a fruit full application. Other literatures like Landslide inventory, statistical (bivariate & multivariate statistics), geotechnicalapproach,butthestudymainlyfocusedonthe probabilisticmethodsofweightedoverlaymethod.

2.2 Deterministic approach

This approach gives a detailed and explanation about the effectandthemagnitudeofthelandslide.Itquantitatively represent the process, the severity and degree of the landslide by comparing determine factors in land slide hazard zonation. Deterministic approach utilized hydrological and slope data as input in zonation. Hydrologicalmodelsinapplicationtodaymaybeeithertwo orthreedimensional,thedatageneratedinthismodelswill be further processed by GIS operations. Slope stability modelscalculatethesafetyfactorsforaslopeintwoorthree dimensions. This approach is difficult to manage for a number of reasons, like amount and sampling density of requiredphysicalparameterswhichishighlycostly[16].

2.3 Artificial Neural Network

Artificialneuralnetworkisamathematicalmodelsimulating the biological neural networks and incorporating the four mathematical algorithms. These models has three simple rules. Multiplication, summation and activation. In this model, factors will have different weights and the mathematical model will consider their weight during the mathematicalcalculations.

Inartificialneuralnetworkinputswithdifferentweightswill bemultipliedbyindividualweightsandtheartificialneural networkwillbethesumofeachinput [17]

Artificialneuralnetworksconsistsofmorethanoneneural networks which can solve complex real life problems by processing information. A single neural network can also easily analyze and gives a detailed reading for simple problems. For a complex and complicated tasks neural networkscanbeinterconnectedinaprocesscalledtopology. Variousartificialneuralnetworksmayhavedistinctkindof architecture. McCulloch and Pitts (1943) were the first to prepareartificialneuralnetworkmodel.

2.4 Fuzzy logic methods

This concept is based on partial logic, in the classical set theorymembershipisdefinedas1=trueor0=false. Fuzzy logic defines the instability factors as members of a set reachingfrom1representsthehighestsusceptibility,to0,

represents no susceptibility of land sliding, allowing different degrees of membership. In several articles fuzzy logic based applications are used to mapping of landslide susceptibleareas[18]

2.5 Inventory past land slides

Inventorymodelsareveryusefulinpredictingandassessing theriskandhazardoflandslidebasedonhistoryofthearea. Previousstudiesandreoccurrenceofthedisasterwillgive animmensewaytostudyandunderstandthedegreeofits frequencyinthefuture.Morethantwoinventoriescan be used for the study as the pattern and the severity of the parameters vary according to the specified areas of study [19]. Landslide inventories give a clear look on the geographicalconditionanddistributionofpastmovements, mapsandsizeoflandslide,whichiskeyinassessingandrisk managementmeasures.

Role of remote sensing in Landslide Inventory

Comprehensivelandslideinventoryisaninputforlandslide hazardandriskanalysis.Inventorymapspredictthetypeof landslidealongwithtimeanddateofoccurrence.Inventory approaches may use distinct approaches for prediction including digital stereo image interpretation to automatic classification or the combination of both. Multi temporal images also applied in inventory mapping. The stereoimagesplayakeyroleforlandslideinventorymappingsasit provides a 3 dimensional visualization opportunity in parallel with derivation of height information. The role of satellite images is very useful with higher resolution. Landslides are directly affecting the ground surface, so remotesensingapplicationisverysuitedtoslopeinstability studies.

2.6 Application of remote sensing and GIS in hazard mapping

All methodologies in landslide hazard zonation share one common intersect whichis GISand remotesensingwhich are very important to specifically study and asses the occurrence of landslide in a specific areas [20]. Current trendsinLHZstudygrabsmoreattentionfromgeoscientists and engineering professionals. In old days, it was difficult andlaborerbecauseofatimeandeffortrequiredformanual handlinganddataprocessing.Computerbasedapplications will help to minimize the time and energy loss for LHZ mappinginadditiontoitsaccuracyandprecision.GISisa computer based technology designed to capture, store, manipulate,analyzeanddisplaydiversesetsofspatialand geo reference data. Upcoming new developments on LHZ focusonearlywaringandpredisastermanagement[21].

GIS is a technological framework for aiding accurate data processing, management, integration and analysis and display.CurrentlyGIS,isconsideredasidealtooltoworkon LHZ (Carrara et al). Even though studies have been

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056

Volume: 09 Issue: 11 | Nov 2022 www.irjet.net p-ISSN:2395-0072

conductedonproposingvarietyofdifferentGISbasedLHZ methods for sustainable management, systematic comparison of different modeling in GIS based landslide probability don’t get enough attention in the scientific community. LHZmappingisatooltoidentifythoseareas, whicharevulnerabletolandslidehazardbasedonimportant factors,andtheirpasthistoryofexposureforlandslide [22]

Extremely dangerous places for data collection follow differentproceduresbasedontheirspecifictopographical condition [23]. The authors uses different parameters for zonation including geological, topological, meteorological maps reported in published papers [11], [20], [24], [25]. Dataanalysisisusuallyprocessedbystatistical,probabilistic and GIS accompanied by taking Digital Elevation Model (DEM),enhancedthematicmappingtools,satelliteimages andrastercalculation.Thehighesteverrecordofextinction by land slide was recorded in the years from 1991 up to 1999.Studies alsostatesthatthesusceptibilityoftheserisk areovercomebyusingofmappingtheriskyplacebyusingof logistic regression and probabilistic method and extract those information from geological, topographical, soil & forest maps and these data are constructed based on the knowledgeofGIS,aerialphotographsandfieldsurvey[10] It is not difficult to spot a geographical location, which is highly prone to land slide. Most of the occurrence of landslide are predictable because they are related to mountainouslandscapesandalsofrequentlyseenincoastal area[26]

3. CONCLUSIONS

Studiesincludedinthisreviewarticleuseddistinctwaysto study the LHZ of specific areas. Specific factors are shortlisted as the main driving factors for land slide including, heavy rainfall, sloppy ways, and wet lands are amongothers.Amethodologyforstudyvariedbasedonthe geographicallocationanddatatypeofdatarelevantforLHZ. GISbasedstaticallyapproachwasusedalongwithinventory ofpastlandslidefordatarelatedwithtopographicalmaps, satellite images and metrological data, on the other hand logisticregressionmethodwidelyappliedforfiledsurvey, aerial photographs, and soil and forest maps. From the reviewed articles, it can be concluded that geology, land utilization also considered as determinant factors in studying LHZ of areas specially found in a sloppy and roadsidelocations.

Accordingtoliteraturereviewunderthisreviewarticle,the mechanism, cause and adequate understanding of the landslidearethepreconditionforeffectivelandslidecontrol management.Geologicalstructureoftheareaistheforemost requirement for ideal risk managements caused by land slide. Hydro geological study as well as geomorphologic attribute,whichmaydirectlyleadtoafailureoftheslope. Landslidecontrolmeasuresincludetwotypesofworks

1.Preventionworksand

2.Detentionworks

Theformerintendstostopormitigatealandslidemotionby changing the natural conditions, such as, topographical, geotechnicalandwaterconditionsatalandslidesite,while thelatteraimsatdetainingapartofthelandslidemotionor theentirelandslidemotionusingstructuralcontrolworks.

REFERENCES

[1] R. Anbalagan, “Landslide hazard evaluation and zonation mapping in mountainous terrain,” Eng. Geol.,vol.32,no.4,pp.269–277,1992.

[2] L. Ayalew, “The effect of seasonal rainfall on landslides in the highlands of Ethiopia,” Bull. Eng. Geol. Environ.,vol.58,no.1,pp.9–19,1999.

[3] B. Temesgen, M. U. Mohammed, and T. Korme, “Natural hazard assessment using GIS and remote sensing methods, with particular reference to the landslidesintheWondogenetarea,Ethiopia,” Phys. Chem. Earth, Part C Solar, Terr. Planet.Sci.,vol.26,no. 9,pp.665–675,2001.

[4] Lydia Elena Espizua and Jorge Daniel Bengochea, “Landslidehazardandriskzonationmappinginthe Rio Grande Basin, Central Andes of Mendoza, Argentina,” Mt. Res. Dev.,vol.22,no.2,pp.177–185, 2002.

[5] C.CencettiandP.Conversini,“Slopeinstabilityinthe Bastardo Basin ( Umbria , Central Italy ) – The landslideofBarattano,” Nat. Hazards Earth Syst. Sci., vol.2,no.1984,pp.561–568,2003.

[6] L.AyalewandH.Yamagishi,“TheapplicationofGISbasedlogisticregressionforlandslidesusceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan,” Geomorphology,vol.65,no.1–2,pp.15–31, 2005.

[7] N.D.S.GaniandM.G.Abdelsalam,“Remotesensing analysis of the Gorge of the Nile, Ethiopia with emphasis on Dejen-Gohatsion region,” J. African Earth Sci.,vol.44,no.2,pp.135–150,2006.

[8] P. K. C. Ray, S. Dimri, R. C. Lakhera, and S. Sati, “Fuzzy-based method for landslide hazard assessment in active seismic zone of Himalaya,” Landslides,vol.4,no.2,pp.101–111,2007.

[9] A. Pandey, P. P. Dabral, V. M. Chowdary, and N. K. Yadav, “Landslide hazard zonation using remote sensingandGIS:AcasestudyofDikrongriverbasin, ArunachalPradesh,India,” Environ. Geol.,vol.54,no. 7,pp.1517–1529,2008.

International Research Journal of Engineering and Technology (IRJET) e-ISSN:2395-0056 Volume: 09 Issue: 11 | Nov 2022 www.irjet.net p-ISSN:2395-0072

[10] K.E.Geol, S.Lee,and S.Lee, “Statistical analysisof landslide susceptibility at Statistical analysis of landslidesusceptibilityatYongin,Korea,”no.August 2001,pp.1095–1113,2016.

[11] G. Chimidi, T. K. Raghuvanshi, and K. V. Suryabhagavan, “Landslide hazard evaluation and zonation in and around Gimbi town, western Ethiopia a GIS-based statistical approach,” Appl. Geomatics,vol.9,no.4,pp.219–236,2017.

[12] T. Bibi, Y. Gul, A. Abdul Rahman, and M. Riaz, “Landslidesusceptibilityassessmentthroughfuzzy logic inference system (FLIS),” Int. Arch. Photogramm. Remote Sens. Spat. Inf.Sci. -ISPRSArch., vol.42,no.4W1,pp.355–360,2016.

[13] R. Roslee, A. C. Mickey, N. Simon, and M. N. Norhisham, “Landslide susceptibility analysis lsa using weighted overlay method wom along the genting sempah to bentong highway pahang,” Malaysian J. Geosci.,vol.1,no.2,pp.13–19,2018.

[14] T.HamzaandT.K.Raghuvanshi,“GISbasedlandslide hazardevaluationandzonation–AcasefromJeldu District,CentralEthiopia,GISbasedlandslidehazard evaluationandzonation,” J. King Saud Univ. - Sci.,vol. 29,no.2,pp.151–165,2017.

[15] B. Komac and M. Zorn, “Statistical landslide susceptibility modeling on a national scale: the exampleofSlovenia,” Rev. Roum. géographie,vol.53, no.2,pp.179–195,2009.

[16] M. Marjanović, B. Abolmasov, A. Miladinović, K. Andrejev, M. Samardžić-Petrović, and J. Krušić, “Comparison of expert, deterministic and Machine Learning approach for landslide susceptibility assessment in Ljubovija Municipality, Serbia,” Geofizika,vol.34,no.2,pp.251–273,2018.

[17] P. R. Prasanna and S. E. Nithya, “An Integrated ApproachwithGISandRemoteSensingTechnique forLandslideHazardZonation,”vol.1,no.1,pp.66–75,2010.

[18] V.Barrile,F.Cirianni,G.Leonardi,andR.Palamara, “A Fuzzy-based Methodology for Landslide Susceptibility Mapping,” Procedia - Soc. Behav. Sci., vol.223,pp.896–902,2016.

[19] T. K. Raghuvanshi, L. Negassa, and P. M. Kala, “GIS based Grid overlay method versus modeling approach-Acomparativestudyforlandslidehazard zonation(LHZ)inMetaRobiDistrictofWestShowa ZoneinEthiopia,” Egypt. J. Remote Sens. Sp. Sci.,vol. 18,no.2,pp.235–250,2015.

[20] H.ShahabiandM.Hashim,“Landslidesusceptibility mapping using GIS-based statistical models and Remote sensing data in tropical environment,” Sci. Rep.,vol.5,pp.1–15,2015.

[21] W. Huabin, L. Gangjun, X. Weiya, and W. Gonghui, “GIS-based landslide hazard assessment : an overview,”vol.4,pp.548–567,2005.

[22] C. Xu, X. Xu, F. Dai, J. Xiao, X. Tan, and R. Yuan, “LandslidehazardmappingusingGISandweightof evidencemodelinQingshuiRiverwatershedof2008 Wenchuan earthquake struck region,” J. Earth Sci., vol.23,no.1,pp.97–120,2012.

[23] S.Chauhan,M.Sharma,M.K.Arora,andN.K.Gupta, “Landslide susceptibility zonation through ratings derivedfromartificialneuralnetwork,” Int. J. Appl. Earth Obs. Geoinf.,vol.12,no.5,pp.340–350,2010.

[24] M.Meten,N.PrakashBhandary,andR.Yatabe,“Effect ofLandslideFactorCombinationsonthePrediction AccuracyofLandslideSusceptibilityMapsintheBlue Nile Gorge of Central Ethiopia,” Geoenvironmental Disasters,vol.2,no.1,p.9,2015.

[25] R. Anbalagan, R. Kumar, K. Lakshmanan, S. Parida, andS.Neethu,“Landslidehazardzonationmapping using frequency ratio and fuzzy logic approach, a case study of Lachung Valley, Sikkim,” Geoenvironmental Disasters,vol.2,no.1,p.6,2015.

[26] B.Abebe,F.Dramis,G.Fubelli,M.Umer,andA.Asrat, “LandslidesintheEthiopianhighlandsandtheRift margins,” J. African Earth Sci., vol. 56, no. 4–5, pp. 131–138,2010.

2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified