Evaluation of CMIP6 models for projecting climate trends and variability in the Didessa sub-basin, Ethiopia: Implications for the Blue Nile basin

Global warming intensifies extreme events, requiring bias-corrected climate models for effective adaptation and mitigation. This study utilizes data from nine climate models in the sixth phase of Coupled Model Intercomparison Project (CMIP6) to evaluate projected changes in temperature and precipitation over Didessa sub-basin. Projections are made for two Shared Socioeconomic Pathways (SSPs), SSP2–4.5 and SSP5–8.5, for the periods of 2030–2059 and 2070–2099. CMIP6 outputs are bias-corrected using a distribution-mapping method based on daily observations from 14 meteorological stations for the baseline period of 1985–2014. Model performance is assessed using the Comprehensive Rating Metric (CRM), while climate trends are analyzed using the Modified Mann–Kendall test. Among individual models, EC-Earth3 most accurately simulates observed precipitation, while GFDL-CM4 and MRI-ESM2–0 best capture maximum and minimum temperatures, respectively. The multi-model mean ensemble (MMME) outperforms all individual models across all variables. Conversely, INM-CM5–0 and BCC-CSM2-MR perform poorly in replicating observed precipitation and temperature. Spatial projections indicate increasing precipitation and temperature across both scenarios. The greatest changes occur under SSP5–8.5 in the 2080 s, with precipitation changes of 16.4–32.9% and temperature changes of 2.5 ℃ to 3.7 ℃ and 3 ℃ to 7.3 ℃ for maximum and minimum, respectively. These findings highlight substantial spatial variability in climate variables, providing policymakers with essential insights to develop site-specific mitigation strategies.