Mechanistic insights into West African monsoon precipitation biases in CMIP6 models

Abstract Accurately simulating West African Monsoon (WAM) precipitation remains a major challenge for Earth system models (ESMs), largely due to deficiencies in the representation of key physical processes. In this study, we use a process-based evaluation framework to diagnose the mechanisms responsible for precipitation biases in 40 Coupled Model Intercomparison Project Phase 6 models. Our results show that while most models reasonably capture the spatial distribution of WAM precipitation relative to JRA55, Global Precipitation Climatology Centre, and Climatic Research Unit gridded observations, 24 models substantially underestimate precipitation over the Sahel. We find that although sea surface temperature biases have direct and indirect influences on Sahel precipitation, they cannot fully explain the pronounced dry bias in the models. Diagnostics of the atmospheric moisture budget and thermodynamic energy balance reveal that the persistent dry bias is driven primarily by weak simulated vertical moisture advection associated with weak diabatic heating. This suppresses deep convection and promotes shallow convective regimes, leading to reduced precipitation. Additional contributions stem from limited moisture availability, largely due to underestimated surface evaporation associated with soil moisture deficits and misrepresented turbulent heat fluxes. Furthermore, inadequate regional moisture transport at mid- to upper-tropospheric levels is associated with an overly strong African Easterly Jet that induces stronger mid-level divergence, and a weak Tropical Easterly Jet that weakens meridional circulation, both of which further reduce precipitation. Our findings underscore the need for targeted improvements in model physics, particularly in the representation of diabatic heating, land–atmosphere coupling, and large-scale circulation, to enhance the accuracy of WAM precipitation simulations in future ESMs.