Spatial yield estimation in rice using spectral indices derived from satellite datasets and DSSAT crop simulation model

An inter-comparison of 2 rice yield estimation procedures namely, the Semi-Physical model and the Crop Simulation model, Decision Support System for Agrotechnology Transfer (DSSAT), was conducted for the Samba Season (August 2023 to January 2024). This study focused on the Mayiladuthurai and Nagapattinam districts of Tamil Nadu. The Semi-Physical model utilized satellite-derived remote sensing data, including Photosynthetically Active Radiation (PAR), Fraction of Photosynthetically Active Radiation (FPAR), Water stress, Temperature stress and crop physiological parameters such as Radiation Use Efficiency (RUE) and Harvest Index (HI), to estimate crop yield. The DSSAT crop simulation model estimated rice yield using inputs such as soil properties, weather data, genetic coefficients and crop management practices. The simulated yield was further integrated with Synthetic Aperture Radar (SAR) data to generate spatial yield maps. The generated yields were validated with Crop Cutting Experiment (CCE) based yield estimates. The DSSAT model demonstrated superior performance, achieving an agreement of 88.2% in the Mayiladuthurai district and 86.8% in the Nagapattinam district.