Spatial rice yield estimation using semi-physical and crop simulation models: A comparative approach

This study presents a comparative assessment of two modelling approaches for estimating spatial rice yields across the Cauvery delta zone (CDZ) in Tamil Nadu, viz., (a) a process-based crop simulation model (DSSAT-CERES Rice) and (b) a semi-physical model (SPM). The DSSAT model was calibrated using field data, cultivar-specific genetic coefficients and further refined through the integration of leaf area index (LAI) derived from Sentinel-1 synthetic aperture radar (SAR) imagery. In contrast, the SPM utilized remote sensing-derived inputs such as photosynthetically active radiation (PAR), fraction of absorbed PAR (FPAR), temperature and water stress indices and radiation use efficiency (RUE) to compute yield via net primary productivity (NPP). Results revealed that DSSAT achieved higher alignment with crop cutting experiment (CCE) data (88.6 %) than SPM (83.3 %), attributed to its capability to simulate complex crop-soil-climate interactions. However, the SPM demonstrated greater scalability and ease of implementation, particularly in regions with limited field data. The study highlights the strengths and limitations of each approach, offering insights into model selection based on accuracy, data availability and operational feasibility. These findings suggest that integration of remote sensing and crop modelling could serve as a valuable strategy forimproving regional yield forecasting and enhancing agricultural decision-making.