Boundary layer schemes in the regional climate model RegCM4.6 over Central Africa

This study examines the performance of two Planetary Boundary Layer (PBL) parameterizations in the RegCM4.6 regional climate model for the Central African domain. These are the Holtslag scheme and the University of Washington (UW) scheme. The evaluation is made by performing two experiments over 5 years (from January 1, 2002 to December 31, 2006) with a horizontal resolution of 0.35o×0.35o\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$0.35^{o}\times 0.35^{o}$$\end{document}. The UW-PBL scheme only takes into account the gradients within adjacent vertical levels in the model whereas the Holtslag PBL scheme takes into account the vertical transport over a deeper layer covering several levels in the model. The analysis extends over December–January–February (DJF), March–April–May (MAM), June–July–August (JJA), and September–October–November (SON). For more specific analysis, the study domain is divided into five zones. The results show that the Holtslag scheme is favorable for simulating rainfall in Central Africa mostly during JJA season in zone 3. As far as the wind is concerned, both schemes have a more or less reasonable approach with the positioning of the jets and the observed monsoon flow but with a slight difference. The patterns show a much greater diversity in the way the turbulent mix is taken into account in PBL. The inclusion of evaporation in our analysis shows that it may be the origin of the absence of water vapor in clouds which considerably reduces the amount of rainfall at the ground surface. Results also show that the simulated total cloud cover can explain the better performance of UW PBL scheme than Holtslag scheme to reproduce surface temperature.