Coupling Satellite and Reanalysis Data with Crop Growth Models: A Comparative Analysis of AquaCrop and SAFY-E Outputs for Estimating Irrigation Needs and Yield

This study compares two modeling approaches for estimating crop water requirements (CWR) and yield of processing tomato in Campania Region, located in Southern Italy, by using satellite multispectral imagery and reanalysis data. Both the approaches use the weather variables of the AgERA5 reanalysis databases combined with CM SAF SARAH-3 solar radiation data as weather inputs, along with satellite-derived crop parameters. The first approach employs the AquaCrop model, which is forced with Sentinel-2-derived fractional vegetation cover (FVC). The second uses the SAFY dynamic crop growth model forced by Sentinel-2 leaf area index (LAI) estimates in combination with a one-step approach for crop evapotranspiration (SAFY-E) and a local simplified soil water balance. Both approaches demonstrate good agreement with field observations of harvested yield and irrigation volumes in a high-efficient irrigation context, confirming the reliability of coupling satellite-based vegetation indices with high-resolution gridded numerical weather datasets. The SAFY-based method allows for more continuous simulation of LAI dynamics that is one of the most used and significant crop parameters, while AquaCrop provides detailed soil water balance modeling. Overall, the SAFY–E outputs show slightly improved performance in terms of yield. On the contrary, the AquaCrop outputs show better accuracy in CWR estimation. Overall, these findings support the potential of reanalysis-satellite integrations for water management and water resources planning and underscore the importance of model selection based on data availability, crop type, and application scale.