Implications des niveaux de réchauffement de 1, 5◦C et 2, 0◦C dans le Bassin du Congo: Evaluation du modèle climatique régional RCA4

Abstract

Plausible climate information on future weather are needed to help policy makers in the adoption of resilience measures so as to constrain disastrous effects of changing climate. In regions with few available observed data like Congo Basin (CB), and where models feature the largest range of uncertainties, the confidence placed on a climate model projections is subject to its capacity to reproduce processes responsible for the present climate and its future changes. This study investigates the atmospheric circulation processes and land-atmosphere interactions (through the recycling ratio) associated with climate model biases in the CB total and recycled rainfall, and explores drivers of projected rainfall changes. Here we use an ensemble of simulations from the Swedish Regional Climate Model (RCM) RCA4, driven by eight General Circulation Models (GCMs) from the Coupled Model Intercomparison Project Phase 5 (CMIP5), for the 1.5◦C and 2◦C global warming levels (GWLs), and under the representative concentration pathways (RCPs) 4.5 and 8.5. RCA4 captures satisfactorily the observed patterns of CB total and recycled rainfall seasonality, but shows dry biases independent of seasons and large scale driving atmospheric conditions. While simulations mimic observed peaks in transition seasons (March–May and September–November), the rain band is misplaced southward (northward) in December-February (June-August), reducing the latitudinal extent of rainfall. Moreover, ERA-Interim reanalysis driven RCM simulation and RCM–GCM combinations show similar results, indicating the dominance of systematic biases. Modelled dry biases are associated with dry upper-tropospheric layers, resulting from a western outflow stronger than the eastern inflow and related to the northern component of African Easterly Jet. Biases in modeled recycled precipitation are due to overestimation and/or underestimation in soil moisture and/or solar radiation. North of the CB, soil moisture is found controlling the recycling of rainfall whilst solar radiation does so in the South. RCA4 simulates well land-atmosphere interactions in the CB in capturing well the spatial extents and the seasonality of the recycling ratio (RR), coupled to those of the soil moisture and surface solar radiation. From the analysis of the climate change signal, we found that regional scale responses to anthropogenic forcings vary across GWLs and seasons. Changes of rainfall and moisture divergence are correlated, with values higher in March–May than in September–November, and larger for global warming of 2◦C than at 1.5◦C . There is an increase of zonal mois ture divergence fluxes in upper atmospheric layers (> 700 hPa) under RCP8.5 compared to RCP4.5. The RR is projected to robustly decrease in the most part of the CB, promoted by the intensification in the moisture flux convergence. Moreover, it is found that additio nal warming of 0.5◦C will change the hydrological cycle and water availability in the CB, with potential to cause challenges to water resource management, agriculture, hydro-power generation, sanitation and ecosystems.