Groundwater Level Variation Forecasting in Coastal Area of Chennai Basin in India Using CMIP 6 and Deep Learning Technique

Groundwater is the most precious natural resource in modern days. The extreme extraction of groundwater is causing a decline in the groundwater level. The groundwater level mostly depends on various climatic factors i.e., rainfall and temperature. The pattern of these factors changes over the years due to climate change. These variations of climatic factors affect the groundwater level over the seasons. Compared to conventional hydrological modeling, today’s most intriguing model is built on a deep learning environment and exhibits higher interest in precise predictions. The study area of Chennai Basin is a coastal area where we apply the deep learning-based model to forecast groundwater levels. Using only two parameters and some hyper-parameters this model performs well. This study considers 4 SSP scenarios, SSP126, SSP245, SSP370, and SSP585. The forecasted groundwater level in SSP585 is approximately 12 m which is the maximum. The R2 and NSE are used to check the accuracy. This season-wise accurate forecasting will help India’s govt. And the farmers to plan their upcoming actions. The farmers will have an initial idea about the available groundwater level. So, they can irrigate their land according to the available crop-water ratio. Also, CGWB can plan its yearly plan for precious groundwater use in the locality. This action can help to avoid the extreme events of water scarcity.