Optimizing Cocoa Pod Harvest Timing Using the SIMPLE Model: Calibration and Performance Evaluation Across Colombian Regions

This study evaluates the performance of the SIMPLE model for predicting cocoa pod harvest timing across five regions in Colombia. Utilizing daily weather data from January 1, 2018, to December 31, 2020, and transforming it into .In WHT files, the model incorporates key input variables, including flowering date, solar radiation ( SRAD), maximum and minimum temperatures (TMAX, TMIN), and rainfall. The research focused on three regional parameters—thermal time required for pod harvest (Tsum), Radiation Use Efficiency (RUE), and yield—while common physiological parameters were used across regions. Calibration was performed for cultivars ICS95 and CCN51, with a crop cycle set to 200 days from flowering to harvest, though farmers traditionally harvest at 180 days. Model performance was assessed using the relative root mean square error (RRMSE) by comparing simulated yields with field data from Fedecacao. Results indicate that the SIMPLE model effectively predicts cocoa yield with an overall RRMSE of 7.2%, outperforming other crop models. The model demonstrated variations in thermal time and yield across regions, highlighting that harvest timing can be optimized based on regional temperature and solar radiation patterns. The study also identified the potential for improved cocoa quality and yield predictions by considering environmental effects on fruit development. Future work will focus on incorporating additional factors such as diseases and nutritional deficiencies and developing an accessible app for farmers to enhance decision-making.