Assessment of agricultural drought risk leveraging earth observation data and Google Earth Engine for Odisha

This study assesses agricultural drought risk in Odisha, India, using Earth Observation (EO) data and Google Earth Engine (GEE) to develop a spatially explicit, district-level drought risk framework. The authors integrate hazard, exposure and vulnerability following the UNDRR disaster risk concept to generate agricultural drought risk maps. Drought hazard was assessed using rainfall, temperature, potential evapotranspiration (PET), vapour pressure deficit (VPD) and soil moisture derived from TerraClimate datasets, while cropland extent from ESA WorldCover was used to represent exposure and the Vegetation Condition Index (VCI) and irrigation data were used to characterize vulnerability. A cloud-based GEE application was also developed to visualize drought risk and support decision-making.

The analysis shows that western Odisha faces the highest agricultural drought risk. The findings indicate that high risk results from the convergence of persistent drought hazards, extensive rain-fed cropland exposure and elevated vulnerability linked to limited irrigation access and reduced vegetation condition. The authors conclude that the EO and GEE-based framework provides a scalable, transparent and near-real-time approach for drought monitoring and risk assessment. They recommend using the results to support targeted drought relief, crop insurance, micro-irrigation investment, watershed development, rainwater harvesting and drought-tolerant crop promotion, while future research should incorporate socioeconomic vulnerability indicators, crop-yield validation and CMIP6 climate projections.