Abstract:
The objective of this study is to evaluate the
ability of seven CORDEX regional climate models
(RCMs), driven by ERA-Interim reanalysis dataset to
simulate the observed rainfall characteristics over West
Africa during the period of 1990–2008. The seasonal climatology,
annual rainfall cycles, interannual variability,
850 hPa specific humidity, and wind fields of the RCMs
outputs were assessed over a number of spatial scales
covering three climatically homogenous subregions (Guinea
Coast, Savannah, and Sahel) and the entire West Africa
domain. The ability of the RCMs to simulate the response
to El Nino and La Nina events were further assessed.
Results indicate that the RCMs captured the spatial pattern
of rainfall and the three distinctive phases of the West
African monsoon reasonably. It is worth noting that RCA
and CRCM5 failed to distinctively reproduce the monsoon
jump while CCLM, HIRHAM, and REMO largely overestimated
the amount of the pre- and the post-monsoon
rainfall. The analysis also showed significant biases in
individual models depending on the subregion and season
under consideration. These biases appear to be linked to the
model’s failure to resolve convective processes and
topography accurately. The majority of the RCMs used
were consistent with the ground observation in capturing
the dry (wet) conditions associated with the El Nino (La
Nina) events. Statistical analysis conclusively revealed that
the RCMs performance varies over the subregions and seasons, implying that no single model is best at all time. In
general, REGCM3 was found to be the most outstanding of
all the RCMs and is therefore recommended for use in
rainfall assessment over West Africa.
