Yield Estimation of Paddy in Bhandara District, Maharashtra, India, Using a Semi-Physical Remote Sensing Approach

Abstract

Accurate and early estimation of crop yield is essential for food security planning, market regulation, and sustainable resource management. This study applies a semi-physical remote sensing approach to estimate paddy yield in Bhandara District, Maharashtra, India, for the 2018-2022 kharif seasons. Satellite-derived Photosynthetically Active Radiation (PAR), Fraction of Absorbed PAR (fAPAR), Radiation Use Efficiency (RUE), temperature stress, and water stress were integrated into a Net Primary Productivity (NPP) model to generate spatially explicit yield maps. A paddy crop mask was prepared from Sentinel-2 data, while MODIS and INSAT-3D products provided biophysical inputs. Grain yield was obtained by applying a harvest index to NPP estimates. Validation against five-year crop-cutting statistics showed an overall mean deviation of -1%, indicating close agreement between remote sensing estimates and official records. The results confirm that combining simplified process-based modeling with satellite data can provide reliable and timely yield forecasts over heterogeneous agricultural landscapes. This methodology supports decision-making for farmers, planners, and policymakers, and offers potential for operational yield monitoring in other rice-growing regions.