A DIAGNOSTIC EVALUATION OF PRECIPITATION OVER WEST AFRICA IN COORDINATED REGIONAL CLIMATE DOWNSCALING EXPERIMENT (CORDEX) MODELS

Abstract

The study evaluates the ability of seven regional climate models (RCMs) driven by ERA-Interim reanalysis, in skillfully reproducing key features of precipitation (1990-2008) over West Africa within the Coordinated Regional Climate Downscaling Experiment (CORDEX) framework. The seasonal climatology, annual rainfall cycles, interannual and interseasonal variabilities of the RCMs output is assessed over three homogeneous sub regions and validated using ground observational data from eighty-one (81) stations in West Africa. In way to address the uncertainty in the different rainfall products, five gridded and satellite rainfall dataset were also analyzed and compared alongside the RCMs. The ability of the RCMs in simulating large-scale global climate forcing signals is further assessed by compositing the El Nino–Southern Oscillation (ENSO) events. Moreover, the dynamics associated with West Africa Monsoon (WAM) was also investigated. Results from the interannual and interseasonal variations indicated that all the gridded (CRU, GPCC, UDEL, ARC2) and satellite (TRMM) data replicate the noticeable variations in the ground observation data across all the sub regions and the entire West Africa domain to a very large extent. Furthermore, the spatial and temporal nature of rainfall over the region is well captured by most of the RCMs (PRECIS, REGCM, RCA, REMO, CCLM and CRCM5) although individual models exhibit wet and dry biases over different regions notable were results from Lake Chad, Fouta Djallon and Cameroun highlands. These biases were due to model physics and setup, influence of orography, circulation anomalies and moisture transport. The Regional Climate Models (RCMs) were also observed to produce lower seasonal variability of precipitation when compared to the ground observations and the magnitude of the variability varies in space and time. The occurrence of the West African Monsoon jump, intensification and northward shift of the Saharan Heat Low (SHL), during the course of the year were shown to be realistic in most RCMs. The interannual variability of seasonal anomalies is well reproduced and at the same time analysis shows significant biases in each models depending on the sub-region and season. Moreover, the models were generally able to capture the dry and wet precipitation anomaly associated with El Nino and La Nina events respectively as well as the main dynamical large scale features associated with convective activity and rainfall over West Africa. In conclusion, the analysis herein demonstrates that the CORDEX RCMs simulates West Africa rainfall adequately and can therefore be used for the assessment of future climate projections for the region.