West African monsoon dynamics and precipitation: the competition between global SST warming and CO2 increase in CMIP5 idealized simulations

Climate variability associated with the West African monsoon (WAM) has important environmental and socio-economic impacts in the region. However, state-of-the-art climate models still struggle in producing reliable climate predictions. An important cause of this low predictive skill is the sensitivity of climate models to different forcings. In this study, the mechanisms linking the WAM dynamics to the CO2 forcing are investigated, by comparing the effect of the CO2 direct radiative effect with its indirect effect mediated by the global sea surface warming. The July-to-September WAM variability is studied in climate simulations extracted from the Coupled Model Intercomparison Project Phase 5 archive, driven by prescribed sea surface temperature (SST). The individual roles of global SST warming and CO2 atmospheric concentration increase are investigated through idealized experiments simulating a 4 K warmer SST and a quadrupled CO2 concentration, respectively. Results show opposite and competing responses in the WAM dynamics and precipitation. A dry response (-0.6 mm/day) to the SST warming is simulated in the Sahel, with dryer conditions over western Sahel (-0.8 mm/day). Conversely, the CO2 increase produces wet conditions (+0.5 mm/day) in the Sahel, with the strongest response over central-eastern Sahel (+0.7 mm/day). The associated responses in the atmospheric dynamics are also analysed, showing that the SST warming affects the Sahelian precipitation through modifications in the global tropical atmospheric dynamics, reducing the importance of the regional drivers, while the CO2 increase reinforces the coupling between precipitation and regional dynamics. A general agreement in model responses demonstrates the robustness of the identified mechanisms linking the WAM dynamics to the CO2 direct and indirect forcing, and indicates that these primary mechanisms are captured by climate models. Results also suggest that the spread in future projections may be caused by unbalanced model responses to the CO2 direct and indirect forcing.