International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 11 Issue: 04 | Apr 2024
p-ISSN: 2395-0072
www.irjet.net
Development of A Meteorological Drought Index Using Fuzzy Logic Sindhu K N 1, Dr Rekha H B 2 1 MTech student, Department of Civil Engineering, UVCE, Bangalore University, Bengaluru
2Associate Professor Department of Civil Engineering, UVCE, Bangalore University, Bengaluru
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Drought is a natural disaster which impact lives
The study was to develop a Meteorological Drought Index (MDI) utilizing the fuzzy logic. Fuzzy logic, a computational paradigm inspired by human reasoning, has gained interest in various scientific domains for its ability to handle uncertainty and imprecision inherent in real-world data. By incorporating fuzzy logic into the realm of meteorological drought assessment, this research aims to offer a more nuanced perspective on the complex interactions between atmospheric variables that contribute to drought monitoring, assessment, and forecasting.
in various ways and leads to great harm. Droughts are of four categories namely agricultural, hydrological, socioeconomic, and meteorological droughts. There are various types of indices that are being used worldwide to effectively monitor and assess droughts. The study addresses the difficulties posed by droughts through the development of a meteorological drought index using fuzzy logic and adaptive neuro-fuzzy inference model. Various combinations of input variables, including maximum temperature, mean temperature, precipitation, and potential evapotranspiration, for the Ramanagara district of Karnataka state was used for model development. Results revealed that the Rainfall Anomaly Index (RAI) performed best among traditional indices, showing the highest correlation (0.862) with upper soil moisture, a key drought indicator. On the other hand, the average of the topperforming fuzzy logic (FL3) model surpassed all traditional indices, exhibiting a correlation of 0.983 with upper soil moisture. Notably, when the average output of the topperforming FL models has been utilized for training, the optimal ANFIS model achieved a correlation of 0.926 with upper soil moisture. To determine the developed models, drought assessments were conducted in four taluks across different seasons. The validation results indicated that the developed models performed comparably to the bestperforming traditional drought index (RAI) in most cases. In summary, the soft computing drought indices developed in this study, on the basis of fuzzy logic and ANFIS, outperformed traditional techniques, contributing significantly to more accurate drought prediction and mitigation actions.
Conventional drought indices, including PDSI (Palmer Drought Severity Index) and SPI (Standardized Precipitation Index), have played crucial roles in drought monitoring and prediction (Mhamd Saifaldeen Oyounalsoud et.al. 2022). However, these indices often have limitations in capturing the multifaceted nature of meteorological droughts, especially when faced with irregularities and uncertainties in climatic data. The proposed MDI seeks to overcome these challenges by harnessing the flexibility and adaptability of fuzzy logic, enabling a more dynamic representation of drought conditions that considers the inherent vagueness and ambiguity within meteorological data.
2. Literature Survey Numerous researches have embraced soft computing techniques for drought monitoring and forecasting. For example, Abbasi et al. (2019) used the GEP (Gene Expression Programming) model in conjunction with the Standardized Precipitation Evapotranspiration Index (SPEI) to forecast drought over a range of periods. The findings revealed an improvement in accuracy of model from 60.1% at SPEI1 (one-month scale) to 92.3 percent at SPEI48 (48-month scale), underscoring the enhancement in overall accuracy with increasing SPEI scale. Similarly, Keskin et al. (2009) employed the Standardized Precipitation Index (SPI) alongside an advanced drought analysis model incorporating FL (Fuzzy Logic) and Adaptive neuro fuzzy inference system (ANFIS) techniques for meteorological drought assessment across nine stations in Turkey at varying time scales. Their findings highlighted the high efficacy of ANFIS in assessment of drought. Furthermore, Malik et al. (2020) presented the CANFIS (Co-active Neuro-Fuzzy Inference System), a contemporary FL model designed to forecast SPI at six sites in the Indian state of Uttarakhand over several time scales. Conventional artificial intelligence models and regression were compared, and the outcomes demonstrated that the
Key Words: Drought, Fuzzy logic, ANFIS, Pearson correlation, meteorological drought, conventional drought incices, soft-computing models.
1.INTRODUCTION In recent years, the impact of climate change has gained a huge need for more sophisticated tools and methodologies to assess and manage water resources. One of the critical effects of the climate change is drought, which pose a significant threat to agriculture, ecosystems, and human populations worldwide. Traditional meteorological drought indices have proven valuable, yet there remains a compelling need for innovative approaches which enhances the capacity to characterize and give response to drought conditions with greater precision.
© 2024, IRJET
|
Impact Factor value: 8.226
|
ISO 9001:2008 Certified Journal
|
Page 631