Assessing the performance of the CMIP6 multi model mean in simulating precipitation and temperature across Africa

Abstract This study examines the capability of the Coupled Model Intercomparison Project Phase 6 to replicate temperature and precipitation across eight African sub-regions, as well as their correlation with three large-scale climate indices. Qualitative estimations indicate that reference datasets (ERA5, CRU, CHIRPS) and the MMM exhibit similar patterns, albeit with slightly different intensities in many sub-regions and seasons. For precipitation and temperature, several models exhibit good performance across various sub-regions and seasons, with PCCs of 0.98. The MMM effectively captures the signs of the observed trend in many African sub-regions but fails to capture its magnitude. The MMM exhibits better performance than individual models across various sub-regions. Models like MMM and GFDL-ESM4 consistently outperform others, particularly in precipitation. However, significant regional disparities are observed, with SWAF, CAF, and WAF (for temperature during DJF) being the most challenging areas, where few models meet the performance thresholds. In MAM, models exhibit strong performance in NAF, WAF, CAF, NEAF, and SWAF but fail to meet the cutoff in SAH and CEAF. The models exhibit robust performance across many regions during JJA and SON. The MMM also captures the increasing (but with relatively lower value) temperature depicted by ERA5 and CRU. The MMM shows a limitation in reproducing the response to some large-scale climate indices.