Sugarcane (Saccharum Officinarum L.) Yield Forecasting Using Decision Support System for Agrotechnology Transfer (DSSAT)-Canegro: A Model Calibration and Validation in Isabela, Philippines

Agricultural simulation models have become increasingly vital in modern agricultural production due to their wide range of applications. To fully harness their potential in predicting crop growth and yield under specific agro-climatic conditions, local calibration and validation are essential. This paper explores the use of the Decision Support System for Agrotechnology Transfer (DSSAT)-CANEGRO model as a tool to enhance and streamline sugarcane yield estimation within the agro-climatic context of the Isabela Mill District. The genetic coefficients of the Phil 99-1793 variety were calibrated using local agronomic, weather, and soil data from the Isabela State University Research and Development Station. Model validation was subsequently conducted across five field sites in Echague, Isabela, for two cropping seasons: newly planted fields (2023-2024) and first ratoon fields (2024-2025). Calibration results showed strong agreement between observed and simulated data, with R2= 0.92, RMSE= 0.35m, d=0.85 for stalk height, and R2= 0.81, RMSE= 8.72 TC ha-1, and d=0.89 for fresh cane yield. Validation for newly planted fields produced satisfactory predictive performance (R2= 0.83, RMSE= 9.8 TC ha-1, d=0.89). These results highlighted that the DSSAT-CANEGRO model is reliable for estimating sugarcane yield for newly planted fields in the district under normal growing conditions. However, its ability to simulate yields affected by extreme weather events remains limited due to the absence of submodules that can capture typhoon-induced damage, leading to significant overestimations during the 2024-2025 cropping year. Keywords: Genetic Coefficients, Phil 99-1793, Pooled Validation, Model Performance, Yield Simulation