This research work focused on the thorny problem of the representation of rainfall over West Africa and particularly in the Gulf of Guinea and its surroundings by Regional Climate Models (RCMs). The first part of the study assessed the ability of RCMs from a set of simulations from AMMA-ENSEMBLES project in their representation of the rainfall pattern over West Africa and specifically over Cote d’Ivoire. The skills of the RCMs in the simulations for the periods 1990-2005 was evaluated using meteorological stations data from the National Meteorology Office of Côte d’Ivoire. Time series and statistical scores are analysed.

A second period (2010-2013) was used to evaluate the ability of these AMMA-ENSEMBLES simulations to predict the near future using the same ground-based observations. Furthermore, the sensitivities of Weather Research and Forecasting (WRF) Model are tested for changes in horizontal resolution (convective permitting versus parameterized) on the replication of West African Climate in year 2014 and also changes in microphysics (MP) and planetary boundary layer (PBL) schemes on June 2014. The results indicate that the skills of the RCMs vary from one station to another and from one season to another. None of the models considered presents a good performance over the entire country and during all the seasons. Generally, the ensemble mean of all the models presents better results when compared to the observations. These results suggest that the choice of any model to be used for precipitation outputs over the country may depend on the focus of interest: intensity or variability of the rain and also on the area of interest.

The future climate simulated by the same RCMs for 2020-2040 over West Africa indicates an unrealistic modification of the seasonal cycle over Guinea Coast and surroundings. However, the absence of common period for the simulations driven by ERA-Interim (1990-2005) and by GCMs (from 2010) did not allow to determine the origin of this change (climate change signal or unsuitable GCMs information). The sensitivity to horizontal resolution study show Both runs at 24km and 4km (explicit convection) resolution fairly replicate the general distribution of the rainfall over West African region. The analysis also reveals a good replication of the dynamical features of West African monsoon system including Tropical Easterly Jet (TEJ), African Easterly Jet (AEJ), monsoon flow and the West African Heat Low(WAHL). Some differences have been noticed between WRF and ERA-interim outputs irrespective to the spectral nudging used in the experiment which then suggest strong interactions between scales.

The link between the seasonal displacement of the WAHL and the spatial distribution of the rainfall and the Sahelian onset is confirmed in this study. The results also show an improvement on the replication of rainfall with the very high resolution run observed at daily scale over the Sahel while a dry bias is observed in WRF simulations of the rainfall over Ivorian Coast and in the Gulf of Guinea. Generally, over the Guinean coast the high resolution run did not provide subsequent improvement on the replication of rainfall. The sensitivity of WRF to MP and PBL on rainfall replication study reveals that the most significant added value over the Guinean coast and surroundings area is provided by the configurations that used the PBL Asymmetric Convective Model V2 (ACM2) suggesting more influence of the PBL compared to MP.

The change on microphysics and planetary boundary layer schemes in general, seems to have less effect on the explicit runs into the replication of the rainfall over the Gulf of Guinea and the surroundings seaboard.