Figure 4.4 Seasonal means of rainfall in mm/day for the period 2001–2006: GPCP (a–d), ARC2 (e–h), GFC_CTR (i–l), GFC_SLAB (m–p).
Figure 4.5 Seasonal mean biases of precipitation in mm/day for the period 2001–2006: ARC2 ‐ GPCP (a–d), GFC_SLAB ‐ GPCP (e–h), GFC_CTR ‐ GPCP (i–l), GFC_SLAB ‐ GFC_CTR (m–p).
4.3.2. Taylor Diagram
We assessed the interannual variability and degree of similarity between the observed and model‐simulated temperature/rainfall using the Taylor diagram (Figure 4.6). It emerges that the experiments reproduce CC values between 0.6 and 0.9 for temperature and between 0.6 and 0.8 for rainfall in the Central African domain. MSEs are generally between 0.5 and 1.5 °C. As far as standard deviations are concerned, the results are heterogeneous. However, their values are around 1 °C in all seasons, with the slab‐ocean experiments closer to unity in the case of the whole of Central Africa and also for the NCA, EE, and SCA regions. It should be noted that in the WE region, which is closer to the Atlantic Ocean and therefore strongly influenced by its dynamic system, the two experiments in the model show particularly poor performance. This can be attributed to the fact that the parameterization of the slab‐ocean, which is only a simple approximation of the ocean dynamics, fails to reduce the bias of the SST and thus reduce the high values of MSE and variability on the Atlantic coast.
4.3.3. Seasonal Cycles of Temperature and Rainfall
Figure 4.7 shows the seasonal cycle of rainfall (mm/day; first row) and temperature (°C, second row) for the different experiments, observations (ARC2, GPCP, and CRU), and reanalyses (ERA 5 and ERA_Interim) in the four sub‐regions of the Congo Basin. The shaded area corresponds to the range of variation, corresponding to the standard deviation of the observations (GPCP, ARC2 for precipitation, and ERA 5, CRU, and ERA_Int for temperature), which makes it possible to present the interannual variability envelope. The observation data show an almost similar evolution, but with some slight differences in the intensities in the different sub‐regions. The NCA subzone is characterized by a single peak of 7 mm/day reached around August. We also observe rainfall of less than 1 mm/day from November to March, which is considered the dry season in this sub‐region. As far as temperature is concerned, the annual cycle has a minimum of around 24 °C around August and a maximum of 29 °C in May. We note that the maximum rainfall coincides with the minimum temperature and vice versa. This behavior can be explained by the fact that the cloud cover is much denser to the point of reducing the intensity of incoming radiation. In this sub‐region, the experiments manage to reproduce the minimum and maximum rainfall and temperature observed. However, the experiments peak in August, one month earlier than ARC2. The large discrepancy between observations and experiments (of the order of 3.5 mm/day) can be observed around May. As far as the temperature is concerned, the experiments tend to reproduce the maximum and minimum temperature observed with a large deviation of 3.5 °C reached around March.
Figure 4.6 Taylor diagrams for area‐averaged seasonal precipitation over North Central Africa (NCA), South Central Africa (SCA), West Equatorial (WE), East Equatorial (EE) and Central Africa (CA) for the period 2001–2006.
Figure 4.7 Seasonal cycle of rainfall (a–d) and temperatures (e–h) in the four sub‐regions for the period 2001 to 2006 for data: ARC2 (red), GPCP (blue), ERA 5 (red), CRU (cyan), ERA‐Interim (blue) and the different experiments (back and green). Shaded in grey, range of variation delimited by +/– standard deviation of the observation data.
The WE and EE sub‐regions present a bimodal rainfall pattern with a maximum in April and October, as noted by the majority of studies conducted in the region (Fotso‐Kamga et al., 2020; Fotso‐Nguemo et al., 2017; Nicholson et al., 2018; Pokam et al., 2014; Taguela et al., 2020; Tamoffo et al., 2019; Vondou & Haensler, 2017). The experiments manage to capture the maxima and minima, but with biases of 2 mm/day on average around April and October. As far as temperature is concerned, we observe a bias of 1 °C with the observation data. The experiments tend to capture the maximum but fail to capture the minimum temperature observed with a difference of 3 °C around January.
The SCA subzone shows a peak rainfall of about 6 mm/day, which is reached between November and March during the monsoon season. The experiments reasonably reproduce the seasonality of the rains in this sub‐region, but with a slight overestimation. As far as temperature is concerned, we observe a slight bias of about 1 °C between the observation data. The experiments manage to capture the modulations of the maximum at the beginning of the year but fail to represent their amplitude (with a bias of about 3 °C). Generally speaking, the experiments more or less represent the annual cycle of rainfall and temperature. It should be noted that the seasonal temperature cycles of the different experiments are outside the range of variability observed in the different sub‐zones. A comparative analysis of the different experiments reveals that the slab‐ocean experiment is best in the SCA subzone for rainfall and all subzones for temperature. It can be said that slab‐ocean parameterization or ocean–atmosphere coupling has slightly improved the performance of the RegCM model in the SCA zone for rainfall and all sub‐region about temperature, although it is underestimated.
4.3.4. Regional Circulation
The regional atmospheric circulation of Central Africa is governed by low‐level westerly winds, the northern and southern components of the African Easterly Jet (AEJ) and the Tropical Easterly Jet (TEJ). These winds are some of the most important factors related to the convective activity of precipitation through moisture transport from the Atlantic and Indian Oceans to the continent (Dyer et al., 2017; Nicholson & Grist, 2003; Pokam et al., 2014). Figure 4.8 shows the low‐level wind (at 925 hPa) for the period from 2001 to 2006. In this figure, we observe that the regional circulation over the Atlantic and Indian Oceans is well represented by ERA 5 during all seasons. The different experiments capture this regional circulation reasonably well but with an overestimation of the zonal wind at the eastern border, which is well marked during the JJA season and across the Atlantic Ocean. However, the direction and intensity of the wind are well represented by the different experiments. It is across the Atlantic Ocean that the direction was not well simulated by the experiments. Convergence towards the continent was not well represented.