Performance of the SWAP–WOFOST Model for Simulating Cotton Growth and Yield Under Tropical Cerrado Conditions
摘要
Process-based agro-hydrological models are essential tools for assessing crop growth and yield under contrasting management and environmental conditions. This study assessed the performance of the coupled Soil-Water-Atmosphere-Plant and World Food Studies Simulation Model (SWAP–WOFOST) to simulate cotton (Gossypium hirsutum L., cv. FiberMax 975 Wild Strike®) growth and yield under tropical Cerrado conditions in western Bahia, Brazil. Field experiments were carried out during the 2014/2015 growing season under three nitrogen management treatments: ammonium sulfate®, prilled urea®, and a control without nitrogen fertilization. The SWAP component was calibrated to simulate soil water dynamics, whereas WOFOST was used to represent crop growth and yield. Model performance was evaluated using the coefficient of determination (R²), Willmott’s agreement index (d-index), Root Mean Square Error (RMSE), Nash–Sutcliffe efficiency (NSE), and the performance index (c-index). The SWAP model showed limited performance in simulating soil volumetric water content, with low precision (R²), agreement (d-index) and efficiency (NSE), indicating limitations in representing near-surface soil water dynamics, particularly after rainfall events. In contrast, the WOFOST component showed excellent performance in simulating leaf area index, reproductive structure biomass, and total aboveground biomass across all treatments. These results indicate that, despite limitations in soil water content representation, the coupled SWAP–WOFOST framework was robust for simulating cotton growth and yield responses under contrasting nitrogen management strategies. The model may support assessments of crop performance and nitrogen fertilization strategies under tropical Cerrado conditions.