Remote Sensing‑Based Assessment of Surface Water Bodies in Satara District, Maharashtra

Abstract This study evaluates the spatio-temporal dynamics of surface waterbodies in Satara district, Maharashtra, using the Normalized Difference Water Index (NDWI) derived from multi-temporal Landsat imagery for the years 1994, 2004, 2014, and 2024. NDWI-based classification was employed to delineate three land categories: Non-Water (NDWI ≤ 0.0), Moist Areas (0.0–0.2), and Waterbodies (NDWI > 0.2). Spatial analysis revealed that while waterbody coverage remained low throughout the study period, it exhibited a steady increase from 1.59% in 1994 to 2.56% in 2024. The highest extent of non-water areas was observed in 2014 (49.12%), coinciding with possible drought conditions, whereas moist areas remained dominant overall but showed a temporary decline in 2014. Temporal analysis supported these findings, highlighting significant inter-decadal variations in surface moisture availability and surface water extent. The steady increase in waterbody areas, along with the partial recovery of moist areas by 2024, reflects improved rainfall patterns and effective watershed management practices. Accuracy assessment using confusion matrices and classification metrics indicated a progressive enhancement in classification reliability, with overall accuracies ranging from 88.67% to 92.00%, and waterbody classes consistently achieving user’s and producer’s accuracies above 90%. The study demonstrates the robustness of NDWI in long-term waterbody monitoring, particularly when integrated with advancements in Landsat sensor technology and cloud-based platforms like Google Earth Engine. The methodology proves efficient for regional-scale hydrological assessments and provides valuable insights for sustainable water resource management in drought-prone regions such as Satara district.