Analysis of the West African Monsoon system in the regional climate model COSMO-CLM

摘要

The West African Monsoon (WAM) is a major component of the global monsoon system. The contrast between land surface temperature (LST) (in the Sahel and Sahara) and sea surface temperature (SST) dominates the WAM formation. This study investigated the WAM representation, and the impact of surface temperature uncertainties in three regional climate simulations with the model COnsortium for Small-scale MOdelling-Climate Limited-area Model (COSMO-CLM) (grid-spacing: 0.44?). The regional simulations were driven by present-day climate simulations with the global climate model ECHAM5 (grid-spacing: ～1.9?), and by the re-analysis data ERA-Interim (～0.7?). The WAM dynamics were quantified using the WAM wind shear index (WAMI). In addition, indices for outgoing long-wave radiation (OLR) (an indicator for convective clouds), and total precipitation were used to assess monsoon characteristics. The largescale patterns of precipitation were adequately reproduced by COSMO-CLM compared to observations, but there were significant uncertainties at regional scales, such as a strong overestimation of precipitation in the Sahel. The model also significantly overestimated convective activity and simulated a too intense monsoon circulation as indicated by WAMI. The impact of bare soil albedo on LST was investigated by implementing an MODIS-observation based bare-soil albedo parameterization, which led to a reduction of the simulated warm bias in the Sahara region during the monsoon season by up to 3K, and an improvement of simulated Sahel precipitation. However, the simulated monsoon circulation was not improved. Using either ERA-Interim or ECHAM5 at the lateral boundaries showed that the COSMO-CLM results were very sensitive to the driving data. And, on coarse grid-scales (of the order of the grid-spacing of the driving datasets) the regional climate model was not able to perform substantially better than the forcing data. The differently driven COSMOCLM simulations also indicated that the land-sea temperature gradient, and its impact on WAM, is dominated by SST and less by LST (in Sahara and Sahel).