Sensitivity studies of the RegCM4 simulation in West and central Africa during strong and weak years of Atlantic cold tongue

Abstract

This study evaluates the ability of the Regional Climate Model (RegCM4) to reproduce the observed rainfall and wind over West and central Africa (WCA) during two contrasting years of the Atlantic cold tongue (ACT): the very cold in 1992 and very warm in 2007 depicted by the Caniaux et al.’s index. The convective schemes of Kuo, Massachusetts Institute of Technology‐Emanuel (Ema) and Grell with closure assumptions of Arakawa–Schubert (GAS) and Fritch–Chappel (GFC) were compared and the best scheme was tested to study the sensitivity of WCA climate to ACT in August of each chosen year—the month which follows that of the peak of ACT. Comparisons of simulations for 1992 and 2007 reveal that some areas are sensitive to extreme ACT events. The regional distribution of observed precipitation (wind) in August 2007 was heavier (lower intensity) than that for the 1992, particularly in West than central Africa region. From GPCP data, it is observed that the range of Cameroon highlands and the Bauchi Plateau of Nigeria are not significantly affected by the ACT. It is shown that rainfall of some subregions is very sensitive to ACT. The GFC scheme is capable of reproducing the amount and distribution of the precipitation for the 2 years in the ACT area, the west and east Guinean coasts, and failed to reproduce the rainfall in west and central Soudano–Sahel regions and East Africa. However, the Kuo scheme reproduces well in these three zones. Contrarily to rainfall, the Ema scheme better reproduces, but overestimating, the wind distribution in the Sahara region. GFC better simulates the wind speed sensitivity to ACT far from the coast in Soudano–Sahel and Nigeria–Cameroon zones. The GAS and Kuo are not well correlated with ERA‐Interim in WCA. Most of the schemes adequately mimic the African easterly jet structure.

An evaluation of COSMO‐CLM regional climate model in simulating precipitation over Central Africa

“…Modelling studies with RCMs carry out over Africa mainly focused either over West Africa (Sijikumar et al, 2006;Diallo et al, 2012Diallo et al, , 2013Gbobaniyi et al, 2014;Klutse et al, 2016;Dieng et al, 2017), East Africa (Anyah and Semazzi, 2007;Endris et al, 2013) or Southern Africa (Vigaud et al, 2012;Diallo et al, 2015;Abiodun et al, 2016;Pinto et al, 2016), where they contributed to improve GCMs simulations. Meanwhile, some studies showed that over CA, RCMs have significant deficiencies associated either to the poor parameterization of internal process or to the very complex and heterogeneous topography of the region which remains a major challenge for the climate modellers community (Tchotchou and Kamga, 2010;Fotso-Nguemo et al, 2016;Mbienda et al, 2016;Vondou and Haensler, 2017;Kenfack et al, 2018;Tamoffo et al, 2019a).…”