HRU-based Downscaling of GRACE-TWS to Quantify the Hydrogeological Fluxes and Specific Yield in the Lower Middle Ganga Basin

•The GRACE-TWSA-derived GWS manifest a high correlation with SWAT-HRU variables.•SWAT-HRU-based downscaling of GWS is advantageous for basin-scale surface water and groundwater management.•The specific yield has been calculated using regularised robust optimization.•The rate of groundwater decline from 2001 to 2014 has been 2.2 cm to 4.6 cm per year. Precise volumetric assessments of different hydrological variables, such as precipitation, evapotranspiration, and groundwater components, are necessary for comprehensive water resource management. This presents several challenges, including topographical complexity and economic limitations, mainly when aiming for high temporal and spatial resolution.The satellite mission Gravity Recovery and Climate Experiment (GRACE) has greatly improved the ability to quantify variations in GWS. We have described a two-stage downscaling system that integrates SWAT Hydrological Response Units (HRUs) with GRACE-derived Terrestrial Water Storage Anomalies (GRACE-TWSA) using Artificial Neural Networks (ANN). Using data from the Global Land Data Assimilation System (GLDAS), GWS was derived with GRACE-TWSA and further downscaled to the HRU level. The GWS trend in most of the study areas has not shown any significant trend from 2001 to 2014. The mean increasing trend was 2.16 mm/year and the mean decreasing was 2.28 mm/year. Using a decision-tree-based CatBoost model, the GWS has been used as an independent variable to determine distributed groundwater levels within the study area. A strong correlation between measured groundwater levels and GWS was observed, and a regularised robust optimization was used to determine the Specific yield. The surface water and groundwater budgeting indicate that most Gangetic area blocks are water-stressed. The study offers an extensive approach to integrated water resource management by providing insights into groundwater availability, aquifer storage characteristics, and budgeting with HRU scale GWS estimates.