Étude des caractéristiques pluviométriques et projections des impacts du réchauffement global à partir des modèles climatiques couplés «CMIP5» en Afrique Centrale

Abstract

In this work, we used daily rainfall data from 20 General Circulation Model (GCM) simula tions, participating in Phase 5 of the Coupled Coupled Models Intercomparison Project (CMIP5) and eight daily rainfall indices, defined by the Expert Team on Climate Change Detection and Indices (ETCCDI), to study changes in extreme weather condition under RCP2.6, RCP4.5 and RCP8.5 scenarios. We carried out a comparative study between the 20 CMIP5 models and six (06) rain observations data and two (02) reanalyses from 1980 to 2005 for the rainfall climatology and the dynamic system that regulates the rainfall. In order to analyze the inter-annual variability, the Niño3.4 index was used to evaluate the influence of the ENSO phenomenon on rains in Cen tral Africa. The results show that the CMIP5 models succeed in reproducing the spatio-temporal characteristics observed, but with different amplitudes. Some CMIP5 models take into account dynamic parameters such as African jets during rainfall simulation. It can also be seen that im proving the speed of the West African jet plays an important role in the simulation of heavy rains. Moreover, there is a good agreement between the strong upward movement and the amplitude of the spatial rains. As a result, the performance of the multi-model ensemble (EMM), which actually refers to the set of best-performing models selected using Taylor’s diagram analysis, was evalua ted. We found that the MME is more efficient than the models taken individually, as regards the simulation of the climatology of the different rainfall indices. Except in the case of the dry spell where the high variability of the models resulted in the degradation of the overall performance of the MME. At the end, future changes in rainfall indices indicate significant declines in the total rainfall amount in southern central Africa during the season September-October-November. But, an increase has been recorded in the south (north) during the December-January-February (September-October-November) season at the end of the XXIe century. While the increase in to tal rainfall is associated with the increase in extreme rainfall, the decline is rather related to the decrease in rainfall frequency.