[A-TT29] Machine Learning Techniques in Weather, Climate, Ocean, Hydrology and Disease Predictions

[ATT29-P01] AI-ML techniques application in groundwater level forecasting: A review

*SUMIT SHOUMYA¹, Pankaj Kumar Gupta^2,3 (1.AMITY UNIVERSITY, 2.Indian Institute of Technology Delhi, 3.University of Waterloo Canada)

Keywords:Artificial intelligence , Machine learning, Groundwater, Hydrology

Groundwater resources have been declined significantly due to over exploitation, urbanization and intensive farming in the global space and time. Artificial intelligence and machine learning (AI-ML) based modelling approaches like Genetic Algorithm (GA), Artificial Neural Network (ANN), Support Vector Machine (SVM), etc. can be efficient tools to understand the role of hydrological drivers of groundwater level dynamics. AI based models, like ORELM model, can be helpful in knowing the groundwater fluctuation. Groundwater table dynamics can be predicted using historical data input of groundwater levels, rainfall, temperature, NOI, SOI, NIÑO3 and monthly population growth rate. The objective of this study is to review the applicability of AI-ML for the groundwater level forecasting under varying climatic conditions. This study evaluated 112 articles using the Scopus database by providing the search keywords- Genetic Algorithm, Artificial Neural Network, Support Vector Machine, ORELM, hybrid models, groundwater level, and so on. The results suggest that the machine learning based approach with data pre-processing predict groundwater levels accurately and the result is about R > 85%. The study also suggests that the hybrid models (HANN and HSVM) perform better than the original models (ANN and SVM) while predicting groundwater level fluctuations, particularly in urban areas. It was also found that prediction accuracy decreases if any increase in the time lead for both original and hybrid models. ORELM models were predicted groundwater level changes in rainy and non-rainy years without the use of complex and time-consuming numerical models with high accuracy. This review work based presentation will help the field managers who are responsible for the groundwater management to apply an appropiate AI-ML tools.