Agricultural drought in semi-arid India threatens food security, particularly during the Rabi season when crops rely on limited irrigation. This study presents an integrated geospatial framework for village-scale crop water stress and drought monitoring, demonstrated in Sangamner Taluka, Maharashtra. The framework combines Landsat 8/9 thermal and multispectral imagery with ERA5-Land soil moisture, CHIRPS rainfall, and MODIS evapotranspiration. An empirical Crop Water Stress Index (ES-CWSI) is computed using NDVI-derived emissivity and scene-specific percentile normalisation of land surface temperature, enabling spatially relative stress assessment. A composite Drought Severity Index (DSI) conditionally integrates ES-CWSI, root-zone soil moisture, and rainfall deficits. Processing occurs across three layers: Google Earth Engine for quality-controlled data acquisition, QGIS/Python for index computation, and a web dashboard for visualisation and internal validation. Analysis of the 2025–2026 Rabi season reveals substantial intra-taluka variability, with 58.9% of villages experiencing persistent high stress and peak stress occurring on 13 March 2026 (mean ES-CWSI=0.579). Strong correlations between ES-CWSI and NDVI (𝜌 = −0.81), soil moisture (𝜌 = −0.68), and DSI (𝜌 = 0.96) confirm physical consistency. The framework provides an operationally robust, uncertainty-aware system suitable for regional agricultural advisory applications.
Dighe et al. (Thu,) studied this question.