Assessing the performance of WRF model in predicting high-impact weather conditions over Central and Western Africa: an ensemble-based approach

Abstract

For numerical weather prediction over a particular region, it is important to know the best combination of physical parameterizations available in the considered modelling frame work. The main objective of the study is to obtain the best combination of the weather research and forecasting (WRF) model physics for accurately simulating high-impact weather conditions, especially the rainfall over Western and Central Africa. For this purpose, performance of WRF from various simulations with specific configurations is assessed by comparing the results to the Tropical Rainfall Measurement Mission data. Each of the simulations is carried out for 30 h and initialized at 00 UTC. The spin-up time considered for the study is 6 h. A flood event (21–22 July 2010) is simulated by considering five cumulus physics schemes with multiple cloud microphysics, planetary boundary layer and land surface parameterizations. Analysis of the model results indicates that some of the physics combinations have good agreement with observations, especially the new (GFS) simplified Arakawa–Schubert and the modified Tiedtke cumulus parameterizations combined with Thompson and Morrison microphysics. However, most of the combinations over-estimated the rainfall over the study domain, while the simulations with Betts–Miller–Janjic cumulus parameterizations showed negative bias over designated regions of Africa.