Simulating Wheat (Triticum aestivum L.) Yield Under Different Sowing Dates and Nitrogen Management Using CERES-Wheat Model in Tropical Highlands of Ethiopia

This study assesses climate change impacts on wheat production in the tropical highlands of Ethiopia using the DSSAT-CERES-Wheat model. Field experiments were conducted across three locations (Bore, Kulumsa, Sinana) during 2023–2024, evaluating two wheat cultivars (Shaki, Boru) under nitrogen rates (0, 46, 69, 92 kg ha⁻¹) and sowing dates (early, normal, late). The model was calibrated and validated using phenological, growth, and yield data, showing strong agreement between simulated and observed values. For Shaki, grain yield calibration yielded an RMSE of 130 kg ha⁻¹, NRMSE of 3.4%, and a d-index of 0.91; for Boru, an RMSE of 140 kg ha⁻¹, NRMSE of 3.6%, and a d-index of 0.90. Biomass RMSE was 190 kg ha⁻¹ (Shaki) and 200 kg ha⁻¹ (Boru), with d-index >0.91. Anthesis date RMSE was 1.7–1.9 days (d-index >0.92), while LAI simulations had RMSE of 0.14–0.17. Validation results confirmed robust model performance for critical variables: anthesis date (RMSE: 1.9–2.0 days), LAI (RMSE: 0.15–0.18), biomass (RMSE: 187–197 kg ha⁻¹), and grain yield (RMSE: 127-137 kg ha⁻¹) for Shaki and Boru, respectively. Key findings indicate early planting with 92 kg N ha⁻¹ maximises yields, mitigating climate-driven losses. Confirms DSSAT’s utility in guiding adaptive wheat management for Ethiopia’s climate-vulnerable agriculture.