Evaluating the performance of selected CMIP6 GCMs for simulations of historical temperature over Ethiopia

This study evaluates the performance of seven Global Climate Models (GCMs) from the Coupled Model Intercomparison Project Phase 6 (CMIP6) and their Ensemble mean in reproducing historical maximum (Tmax) and minimum (Tmin) temperatures across five agro-ecological zones (AEZs) of Ethiopia during 1995–2014. The assessment covers daily to annual time scales using observational and GCM datasets. Model performance was evaluated using Percent Bias (PBIAS), Root Mean Square Error (RMSE), and correlation coefficient (r), while the Comprehensive Rating Index (CRI) was applied for ranking. Results show substantial spatial and temporal variability in model performance. For Tmax, the Ensemble mean and EC-Earth3-veg performed best at the daily scale, while EC-Earth3-veg, MPI-ESM1-2-LR and BCC-CSM2-MR excelled at the monthly scale across most AEZs. Seasonally, top-performing models included Ensemble mean, MPI-ESM1-2-LR and MRI-ESM2-0 during Bega (October-January), EC-Earth3-veg, CNRM-CM6-1 and Ensemble mean during Belg (February-May) and CNRM-CM6-1 and MPI-ESM1-2-LR during Kiremt (June-September). For Tmin, CNRM-CM6-1, BCC-CSM2-MR and Ensemble mean ranked highest at both daily and monthly scales in most AEZs. At the annual scale, MRI-ESM2-0, Ensemble mean, MPI-ESM1-2-LR and CNRM-CM6-1 excel for Tmax, while EC-Earth3-veg, BCC-CSM2-MR and Ensemble mean lead for Tmin across most AEZs. MIROC6 consistently exhibited the weakest performance for both Tmax and Tmin across most AEZs and time periods. Thus, the Ensemble mean of all evaluated models does not consistently rank among the top three performers across all AEZs and time scales. The identified best-performing CMIP6 models provide valuable tools for assessing climate change impacts and developing region-and time-specific adaptation strategies in Ethiopia. Given the variability in GCM performance across AEZs and time scales, national or large-scale studies would benefit from using the Ensemble mean derived from the best-performing models. This study provides crucial insights into the strengths and weaknesses of different GCMs, supporting evidence-based decision-making and enhancing efforts to build climate resilience and adaptation strategies in the region.