Comparative influence of land and sea surfaces on the Sahelian drought: A numerical study

The aim of this work is to compare the relative impact of land and sea surface anomalies on Sahel rainfall and to describe the associated anomalies in the atmospheric general circulation. This sensitivity study was done with the Meteo-France climate model: AR-PEGE. The sensitivity to land surface conditions consists of changes in the management of water and heat exchanges by vegetation cover and bare soil. The sensitivity to ocean surfaces consists in forcing the lower boundary of the model with worldwide composite sea surface temperature (SST) anomalies obtained from the difference between 4 dry Sahel years and 4 wet Sahel years observed since 1970. For each case, the spatiotemporal variability of the simulated rainfall anomaly and changes in the modelled tropical easterly jet (TEJ) and African easterly jet (AEJ) are discussed. The global changes in land surface evaporation have caused a rainfall deficit over the Sahel and over the Guinea Coast. No significant changes in the simulated TEJ and an enhancement of the AEJ are found; at the surface, the energy budget and the hydrological cycle are substantially modified. On the other hand, SST anomalies induce a negative rainfall anomaly over the Sahel and a positive rainfall anomaly to the south of this area. The rainfall deficit due to those anomalies is consistent with previous diagnostic and sensitivity studies. The TEJ is weaker and the AEJ is stronger than in the reference. The composite impact of SST and land surfaces anomalies is also analyzed: the simulated rainfall anomaly is similar to the observed mean African drought patterns. This work suggests that large-scale variations of surface conditions may have a substantial influence on Sahel rainfall and shows the importance of land surface parameterization in climate change modelling. In addition, it points out the interest in accurately considering the land and sea surfaces conditions in sensitivity studies on Sahel rainfall.