ABSTRACT
Rainfall in the Sahel is extremely variable on daily to multidecadal time scales, challenging climate models to realistically simulate its past and future evolution and questioning their relevance for defining suitable climate change adaptation strategies. Improving confidence in climate models may be achieved by (i) evaluating their capacity for reproducing observed climatic evolution and (ii) attributing these evolutions. Moreover, there is a need to consider relevant climatic indicators, from an end-user point of view. Fully coupled (CMIP6-AOGCM) models with idealized detection and attribution forcings (DAMIP) as well as atmosphere-only simulations (AMIP) are used to investigate the respective roles of external forcing factors and internal climate variability in the observed intensification of the Sahelian rainfall regime. We show that CMIP6 models contain signs of the intensification of the rainfall regime as detected over the past 35 years from a regional daily observations network. Both the increase in intensity and occurrence of wet days, as well as that of extreme daily rainfall, are remarkably well reproduced by historical simulations incorporating anthropogenic forcing factors, with anthropogenic aerosols contributing the largest share of this trend. Though more strongly affected by model structure uncertainty, the greenhouse gas forcing also displays noticeably robust features. Models are shown to fail at simulating the observed dry extreme evolution. These findings give incentive for further investigating the underlying physical mechanisms that drive the Sahelian rainfall regime evolution at regional to subregional scales. Furthermore, future hydroclimatic trajectories in the Sahel should be explored, though particular caution is required as to which rainfall indicator to consider.
