Spatio-Temporal Assessment of Vegetation Dynamics in the Bhadra Watershed Using Landsat-Derived NDVI and Change Detection Analysis

Abstract

Vegetation dynamics provide information for assessing ecological condition, hydrological response and land-use change within watershed systems. This study assessed spatio-temporal variations in vegetation cover in the Bhadra Watershed, Karnataka, India, using the Normalised Difference Vegetation Index (NDVI) derived from Landsat 8 OLI reflectance data for 2015 and 2025. Cloud-minimised images acquired on 22 December 2015 and 23 January 2025 were processed in ArcGIS 10.8 through layer stacking, watershed-based subsetting and spatial preparation. The red and near-infrared bands were used to compute NDVI, and the rasters were classified into five vegetation-density categories: water or non-vegetated surfaces, bare land, sparse vegetation, moderate vegetation and dense vegetation. Pixel-wise change detection compared vegetation conditions between the two years. NDVI values ranged from -0.14 to 0.45 in 2015 and from -0.06 to 0.46 in 2025, indicating a slight shift towards higher positive values. Dense vegetation remained dominant and increased from 3183.12 km² (93.02%) to 3266.56 km² (95.45%). Water bodies increased from 18.89 km² (0.55%) to 35.84 km² (1.05%), while bare land decreased from 33.85 km² (0.99%) to 20.42 km² (0.60%). Sparse vegetation declined from 50.02 km² (1.46%) to 28.52 km² (0.83%), and moderate vegetation declined from 136.33 km² (3.98%) to 70.88 km² (2.07%). These changes indicate an overall increase in dense vegetation, with localised decreases in lower-density categories. The results support the use of NDVI-based Landsat analysis for vegetation-change mapping.