Mechanisms of Rainfall Biases in Two CORDEX-CORE Regional Climate Models at Rainfall Peaks over Central Equatorial Africa

Abstract

Two regional climate models (RCMs) participating in the CORDEX–Coordinated Output for Regional Evaluations (CORDEX-CORE) project feature a dipole-type rainfall bias during March–May (MAM) and September–November (SON) over central equatorial Africa (CEA), consisting of positive bias in west central equatorial Africa (WCEA) and negative bias in east central equatorial Africa (ECEA). One is the Regional Model version 2015 (REMO2015) and the other is the fourth version of the Regional Climate Model (RegCM4-v7). RCMs are nested in three Earth system models (ESMs) from phase 5 of the Coupled Model Intercomparison Project (CMIP5), and in the reanalysis ERA-Interim, at ∼25-km spacing grid resolution. This study highlights misrepresented underlying physical processes associated with these rainfall biases through a process-based evaluation. Both RCMs produce a weaker Congo basin cell, associated with a weaker land–ocean zonal surface pressure gradient. Consequently, less water vapor enters the region, and little is transported from WCEA to ECEA, resulting in higher moisture availability in the west than in the east. This leads to an unevenly distributed moisture across the region, favoring a stronger atmospheric instability in WCEA where the moist static energy (MSE) anomalously increases through an enhanced latent static energy (LSE). Moisture arrives at a slower pace in ECEA, associated with the weak cell’s strength. The intensity of ascent motions in response to the orographic constraint is weak to destabilize atmospheric stability in the lower layers, necessary for initiating deep convection. Therefore, the convection is shallow in ECEA related to underestimating the MSE due to the reduced LSE.