4. Discussion and conclusion
The Congo basin is one of the most hydrologically active and pristine locations with limited understanding of how precipitation changes impacts on stream flow dynamics and variations in catchment stores. This poor understanding in the hydro-climatic processes of the basin is most often the results of gaps in hydrological information and insufficient observational networks for hydrological applications (e.g., Ndehedehe, 2019; Tshimanga and Hughes, 2014; Conway et al., 2009). Due to limited hydro-meteorological observation network in the basin, Munzimi et al. (2015) mentioned that much is still unknown about the Congo river basin’s hydrological behaviour. The comparative assessment of precipitation estimates from GPCC and CRU data over the Congo basin was therefore undertaken to improve quantitative knowledge of Congo basin precipitation and its influence on river discharge. There are still some ongoing debates regarding the uncertainties and ambiguities in precipitation estimates from various products over the Congo basin. While that is not the focus of this chapter, results indicate both products (GPCC and CRU) are generally consistent with high correlation values and statistically insignificant difference in their means. However, the poor association between
these data in the Congo Cuvette central during the period could suggest dynamics in the availability of gauged stations. The GPCC is a widely used observational reference data unlike the CRU precipitation. The distribution of gauges is expected to vary in time and space and the GPCC-based precipitation product has more gauged observations available globally compared to the CRU-based precipitation.
The spatial distribution of rainfall in the Congo basin takes the form of a north–south dipole patterns. These patterns result in wet and dry seasons with opposite phase in the south and north of the basin. That is, the wet season in the south (December-February) coincides with dry season in the north and vice versa. The dipole patterns are caused by the difference in seasons, the more reason why the northern section is wet between May and October. The south on the other hand, remains relatively dry during the same period. The dipoles and this alternating seasonal rainfall patterns have been attributed to the position of the Congo basin across the equator (Munzimi et al., 2015). Be it trends, annual, or semi-annual amplitudes, rainfall during different climatological periods (1901-1930, 1931-1960, 1961-1990 and 1991-2014) over the Congo basin has undergone significant changes. These changes were also echoed by Samba and Nganga (2012) who reported on considerable declines in Congo rainfall between 1980 and 1990. The surrounding oceans, especially the Indian Ocean have been identified as an important source of observed precipitation experienced in wet years in the Congo basin (Dyer et al., 2017). Generally, the Congo basin and much of central Africa are hydrologic hotspots or web of climatic influence. For instance, the SST anomalies along the Benguela Coast, and warming or cooling of the three oceans (Atlantic, Indian, and Pacific) and indices of oceanic variability have been identified as key factors that govern the complex changes in precipitation patterns over West Central Africa (e.g., Ndehedehe et al., 2019; Balas et al., 2007).
As opposed to West Africa where 3-4 significant modes of rainfall variability have been identified (e.g., Ndehedehe et al., 2016a; Sanogo et al., 2015), there are actually more than five statistically significant modes of orthogonal rainfall modes of variability over the Congo basin. Physical interpretation and attribution could be challenging for some of these modes. However, we found that rainfall in the Congo basin is characterised by annual, multi-annual (or bimodal), trends, short term seasonal signals and those resulting from regional factors and atmospheric-ocean interactions and large-scale processes. Locally generated moisture and the interactions of the nearby oceans (Sorí et al., 2017; Dyer et al., 2017) with land surface processes in the basin could also contribute to the magnitude of seasonal variations observed in the Congo rainfall. Sorí et al. (2017) argued that the various sources of moisture for the Congo basin makes nourishment possible during extreme climate events (e.g., flood) thus modulating water balance within the region. On the one hand, this could be the reason for significant amount of rainfall in the Congo basin all through the year except during the June-July period in the southern catchment. On the other hand, the observed multiple significant modes of the Congo basin rainfall can be attributed to signals emanating from the combined influence of climate, natural
variability, ocean interactions and the influence of land surface processes. As an example of the influence of land surface conditions on Congo precipitation, locally generated heat low and reduced precipitation were identified as the direct consequence of decreased evaporation over deforested area in Central Africa (Nogherotto et al., 2013). Arguably, this reassert the important role of land surface processes in the Congo basin (Koster et al., 2004). In this era of anthropogenic-induced climate change, the skills of global and regional climate models in reproducing the Congo basin climatology (spatial patterns, seasonality, and magnitude of precipitation) could be restricted due to uncertainties (Aloysius et al., 2016). However, the question of how the interactions of increasing anthropogenic pressure on the Congo forest and important processes of inter-annual variability impact surface water hydrology are therefore interesting future research directions.
So, has rainfall amounts varied comparatively less across the Congo basin despite known variations in the Congo river flow? In the analysis of historical space-time variability of rainfall (1901- 2014) over the Congo basin we found that previous rainfall (1931- 1990) varied more compared to the 1991-2014 period. As opposed to the two climatological periods between 1931 and 1990, rainfall varied less between 1991 and 2014. It is not clear if the time slice during this last climatological period contributed to this as it was not up to 30 years. However, there were more dry spells and drought events during this period (1991 and 2014) in the Congo basin compared to the last two climatological periods prior to 1991 (Ndehedehe et al., 2019; Hua et al., 2016). The strong positive anomalies of rainfall in extreme wet years are usually captured in the annual component and show strong spatial distributions (loadings) in the northern and southern sections. Although the analysis for the latter climatological period used only 23 years (1991-2014), the spatial distribution and amplitudes of rainfall is consistent and similar to other periods with relatively higher modes of variability. Generally, there seems to be a shift in the hydrological regimes of the Congo river, (especially after 1994). This change can be attributed to the rather pervasive influence of extreme droughts, which fluctuated between 50% (i.e., affected areas) in the early years of the century and 40% during the 1994-2014 period (Ndehedehe et al., 2019). To further support this argument, the rainfall analysis by Samba and Nganga (2012) in Congo-Brazzaville (Congo) during 1932–2007 period indicates that the largest rainfall deficits were observed in the 1980s and 1990s. And since 1985, Zhou et al. (2014) observed a consistent drying of the Congolese forest, which was attributed to gradual decline in precipitation. The change in rainfall trajectory in the post 1994 period is consistent with observed drought and drying in the region. In view of inconsistent trends between rainfall and discharge as evidenced in their cumulative departures, this suggests other key secondary drivers of hydrologic variability that are non-climatic exist in the basin.
The considerable association of discharge with rainfall in catchments characterised by strong annual and seasonal amplitudes in rainfall implies that the wetland hydrology of the basin is largely nourished by rainfall, in addition to possible exchange of fluxes within the Congo floodplain wetlands. Notably, a significant proportion of changes in the dominant rainfall patterns is still not explained by those of river discharge. This information signals the threshold of complex hydrological processes in the region. Importantly, it also suggests the influence of anthropogenic contributions (e.g., deforestation, changes in land cover states, surface water developments, etc.) and strong multi-scale ocean-atmosphere phenomena as key secondary drivers of hydrologic variability. Ultimately, it is obvious nonetheless, that several African regions have been identified as a hot spot with considerable influence of several climate teleconnections (e.g., El-Niño Southern Oscillation and Atlantic Multi-decadal Oscillation, Indian Ocean Dipole, etc.). Given the impacts of these multi-scaled climate indices
on rainfall variability and extreme climatic events (droughts and floods) in Africa (see, e.g., Ndehedehe et al., 2020; Gizaw and Gan, 2017; Diatta and Fink, 2014; Paeth et al., 2012), the combined influence of climate and human activities on hydrological variability is thus palpable and multi-faceted.