Machine learning driven assessment of climate and land use change impacts on water resources in the Kharun River Basin, Central-Eastern India

Accurate prediction of water and sediment yield under changing climate and land use is critical for sustainable watershed management. This study assessed five machine learning models Extreme Gradient Boosting (XGBoost), Random Forest (RF), Artificial Neural Networks (ANN), Support Vector Machines (SVM), and k-Nearest Neighbours (k-NN) for simulating water yield in the Kharun River catchment. Models were evaluated using R 2 , Nash-Sutcliffe Efficiency (NSE), and percent bias (PBIAS), with XGBoost outperforming others (R 2 = 0.95, NSE = 0.89, PBIAS = 14.98%) due to its high predictive accuracy and generalization capability. XGBoost was then used to examine historical LULC changes (1992–2022) and future climate scenarios (CMIP6 SSP2-4.5 and SSP5-8.5). Findings indicate significant seasonal shifts, including reduced pre- and post-monsoon baseflow and heightened monsoonal runoff, highlighting flashier hydrological responses driven by urbanization and agricultural intensification. Future projections suggest annual water yield increases up to 134.9% with rising sediment loads, posing risks of flash floods, reservoir siltation, and dry-season water stress. The study underscores the need for green infrastructure, decentralized water storage, and regulated land use, complemented by participatory monitoring and citizen science to enhance model calibration and local decision-making. Overall, integrating machine learning with climate and LULC analysis offers a robust framework for adaptive watershed management and can guide climate impact assessments in similar environments. • Assessed climate and LULC impacts on Kharun using five ML models for water yield. • XGBoost developed with CMIP6, remote sensing, and geospatial datasets for analysis. • Simulated historical and future water yield and sediment capturing seasonal shifts. • Future projections show higher monsoon runoff and sediment, raising flood risks. • Green infrastructure, storage, and land-use rules recommended for water resilience.