Integrated Remote Sensing and Hydrological Modelling for classification of Small Water Bodies in the Dwarkeswar River Basin

Small water bodies play a vital role in maintaining hydrological balance and ecological functions, particularly in rural and climate-sensitive regions. However, their spatial and hydrological dynamics remain poorly understood due to limited monitoring data. This study presents an integrated approach to detect, map, and classify small ponds in the Dwarkeswar River Basin, using satellite-based water detection, statistical analysis and hydrological modelling. Among the indices tested, the Advanced Normalized Difference Water Index (ANDWI) showed the highest accuracy (95.8%) in identifying small, turbid ponds. Multi seasonal analyses (2018–2022) revealed significant dry season variability, while wet season areas remained stable. K-means clustering, combined with morphometric and hydrological attributes, identified four pond types: Groundwater-Supported, Recharge-Supported, Mixed-Response, and Runoff-Supported. These classes reflect distinct hydrological behaviours based on remote sensing observations and cluster-scale hydrological indicators. The findings support targeted management strategies, such as protecting perennial sources and implementing Managed Aquifer Recharge (MAR), offering a replicable framework for data-scarce regions.