Calibration and Validation of the CERES-Wheat Model: Phenology, Growth, and Yield in Eastern Oromia, Ethiopia

Abstract Accurate crop simulation models are vital for guiding agronomic decision-making and improving food security in smallholder systems. This study calibrated and validated the DSSAT CERES-Wheat model for two bread wheat cultivars, Kakaba and Danda, under the agro-ecological conditions of Eastern Oromia, Ethiopia, using Sonora 64 as a baseline. Calibration employed multi-year field trial data (2021–2023) from Haramaya University, while validation used independent datasets from 2024/2025 experiments conducted at Haramaya and Gurawa sites. Cultivar-specific genetic coefficients were adjusted to reflect local adaptation, highlighting traits such as reduced vernalization requirements, increased photoperiod sensitivity, shortened grain filling duration, and enhanced yield potential through greater grain number and kernel size. Calibration results confirmed robust predictive accuracy, with normalized RMSE values consistently below 10%. For Kakaba, anthesis (nRMSE = 4.4–5.3%), maturity (0.9–3.5%), biomass (0.8–2.7%), and yield (0.3–4.4%) were simulated with high precision, though harvest index showed moderate deviation. Danda demonstrated even stronger stability, with grain yield and phenology predictions consistently within acceptable thresholds (nRMSE < 5%). Canopy dynamics were well captured: Kakaba achieved strong LAI simulation (R² = 0.96–0.98; d-index = 0.94–0.96), while Danda showed strong agreement (R² = 0.96–0.97; d = 0.95–0.96). Validation across sites revealed distinct cultivar performance, with Danda outperforming Kakaba at Gurawa in anthesis, maturity, and yield prediction. Overall, CERES-Wheat reliably reproduced cultivar-specific growth dynamics across sites and years, confirming its robustness for phenology, canopy development, biomass, and yield. These findings emphasize the importance of cultivar-specific calibration to capture genotype × environment interactions, providing a solid foundation for yield forecasting, climate impact assessments, and adaptation planning in Ethiopian wheat production systems.