Evaluating the performance of coupled model intercomparison project phase 6 (CMIP6) models in simulating rainfall characteristics over West Africa

Abstract This study evaluates the seasonal and annual precipitation patterns in West Africa using daily observation data from Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) and Tropical Applications of Meteorology using Satellite and ground-based observation (TAMSAT), and the multi-model ensemble mean (hereafter “GCM_Ensemble”) derived from 30 Coupled Model Intercomparison Project Phase 6 (CMIP6) models, for the period from 1985 to 2014, with all the data at resolution of (0.25° × 0.25°), using the correlation, mean bias error, and mean absolute error . During the dry season (DJF), there is low precipitation with monthly mean value less than 20 mm from 6.5 o N and from 6 o S of the coastal area. The study shows that some models replicate the observed pattern while others underestimate the precipitation spatial distribution pattern. During the pre-monsoon period (MAM), all models were able to simulate the local maximum observed over the Cameroon Highlands and surrounding elevated regions, but with varying magnitudes. During the monsoon period (JJA), there is a maximum amount of rainfall up to 300 mm to 400 mm at the region of maximum rainfall amounts localized over the highlands, and all models show similar patterns with the observational data. Compared to JJA, there is a reduction in rainfall over West Africa in the post-monsoon (SON) season with a southward retreat of the region. The study also discusses the comparison of observed precipitation patterns with GCM ensemble simulations for three regions in West Africa: Guinea coast, Savannah, and Sahel. The performance of different climate models in terms of their biases, correlations, and mean absolute error is also discussed. Models show varying accuracy across subregions, with GFDL-ESM4 consistently performing best overall. The utilization of the multi-model ensemble mean (GCM_Ensemble) further enhances the depiction of precipitation features across the study area.