Regional Climate Modeling of the West African Monsoon Using the Weather Research and Forecasting Model

In West Africa, food security and economy are based on agricultural production that is strongly dependent on monsoon precipitation. Therefore, accurately simulating the West African Monsoon (WAM) is vital to the well-being of the region. Regional climate models are valuable tools to simulate and predict WAM and provide useful information to local water managers. Previous studies showed considerable sensitivity of WAM to different land surface, convection, and planetary boundary layer schemes in regional climate models. In this study we used the Weather Research and Forecasting (WRF) model to further explore the sensitivity of simulated WAM to physical schemes and improve our understanding of WAM predictions. Through a large number of modeling experiments, we found that the different radiation schemes embedded in WRF led to significant differences in simulated precipitation and the related dynamics of WAM. These considerable differences resulted from the differences in the simulated surface temperature gradients that are closely associated with the surface radiative budgets. Additional tests of microphysics, land surface, convection, and planetary boundary layer schemes were also performed for July 2002 when monthly precipitation pattern is closest to the 30-year climatology of the major monsoon season (June, July, and August) over 1981 through 2010. The optimal combination of the physics schemes identified through the sensitivity tests were used to simulate WAM for the period from 1981 to 2010, and the WRF model realistically reproduced the observed climatology and evolution of WAM.