Congo Basin Hydrology, Climate, and Biogeochemistry. Группа авторов

Чтение книги онлайн.

Figure 5.6 Assessing climate influence on surface water hydrology over the Congo Basin. (a) Relationship between river discharge and SPEI, and (b) relationship of Congo River discharge with TWS and surface water storage. TWS here is the GRACE‐hydrological signal in the second orthogonal mode (TWS‐2, Figure 5.5).

Temporal variability in discharge is expected to be driven by changes in precipitation patterns and other land surface conditions, including land cover change. Considerable variability in the Congo River’s discharge between 1960 and 1995 was reported by Alsdorf et al. (2016), consistent with a 21% increase in the Congo River discharge during the same period. Ultimately, this would imply that increased rainfall led to a rise in the Congo River discharge. But the time series of SPEI and SRI were largely inconsistent during most parts of the 1990s when extreme drought was observed (Figure 5.6a). For instance, SRI indicated wet episodes for most of the period after 1995 until 2000 (except 1998) and even during the post‐2000 period, while SPEI was largely characterized by drought episodes in between these periods (Figure 5.6b). Further, it is shown here that the surface water of the Congo Basin is a key component of the GRACE water column, indicating significant association with river discharge and SWS (Figure 5.6b). The multi‐annual variations of TWS (Figure 5.5a,b) observed around the Congo Cuvette centrale is dominated by the Congo River discharge (r = 0.88 at α = 0.05) (Figure 5.6b). The response of the Congo River discharge to climate variations was predicted using the leading modes of SST anomalies of the surrounding oceans (Atlantic, Indian, and Pacific) as predictands in an SVR scheme. The output of the linear SVMR show that global climate through SST anomalies of the three oceans are associated with fluctuations in the Congo river discharge (Figure 5.7a–c). Given the moderately strong correlation (r = 0.79, p = 0.0000) between the observed and predicted (Figure 5.7a,b), SST of the Atlantic and Pacific are relatively stronger predictors of river discharge compared to SST of the Indian ocean, which indicated a moderately strong correlation (r = 0.74, p = 0.0000) (Figure 5.7c). From the SVMR model, the first SST mode (annual) from the Pacific and Indian oceans had the strongest coefficients (second mode of Atlantic SST had the highest coefficients out of the five predictors). However, while the first and second SST modes of the Indian ocean showed strong coefficients, the fifth mode of the Pacific SST showed the second highest coefficients. Overall, the weight of coefficients of the predictands in the SVMR model confirm the importance of slow oceanic and climate signals (e.g., ENSO) from global SST anomaly on hydrological changes and surface water hydrology in the Congo Basin. Furthermore, there is significant difference in the spatial distribution of SWS during extreme drought (2004) and wet (2007) periods in the basin (Figure 5.8a–h, cf. Figure 5.1). Generally, strong spatial patterns of SWS and total inundation are restricted to the Congo River channel with values reaching 200 mm in the September–October period (Figure 5.8a–h). With a gradual rise in rainfall during the November–December period, surface water storage extends to the Cuvette centrale and is perhaps stored as floodplain waters. During the 2004 drought period (Figure 5.8e and g), the floodplain waters around the Cuvette centrale area of the Congo Basin in the November–December period are not as noticeable as the wet period in 2007 (Figure 5.8f and h). There is a significant difference in the SWS spatial and temporal patterns shown for the wet and dry periods (Figure 5.8a–h) and a wider distribution of surface water during the former is observed. This is expected for the Congo Basin, as diminished flow under limited rainfall conditions would be normal. Additional analysis based on observed spatial trends in SWS were also undertaken. These short‐term trends of SWS were estimated for specific drought (e.g., 2005–2005) and wet (e.g., 2006–2007) periods and they are consistent with the aforementioned results.

5.4. DISCUSSION AND CONCLUSIONS

      5.4.1. Understanding Drought Variabilities, Intensities, Characteristics and Drivers

Although the Congo Basin is one of the most humid regions of the world, similar to the Amazon Basin, droughts and its impacts are unavoidable. Drought variability and frequency tend to be higher in the southern part of the Congo Basin, where seasonal rainfall amount is highest during the December–March period in the basin areas. Although extreme droughts affected more than 40% of the basin between 1992 and 2001, drought episodes and their intensities diminished over the Congo Basin after late 2006 when the basin became extremely wet because of strong changes in rainfall. Generally, there is consistency between the results here and the global scale analysis by Spinoni et al. (2014), who showed prolonged and severe droughts during the same period (1991–2010) over the Congo Basin. While the degree of intensity or impacts of extreme droughts might be different due to catchment characteristics, land cover change, topography, and land surface conditions, water deficits caused by prolonged climate‐induced, below average rainfall could have implications on freshwater variability and availability. For example, evolutionary patterns of standardized precipitation index and discharge show that these variables have considerable linear relationships in the Congo Basin (e.g., Ndehedehe et al., 2018c, 2019). Consistent with this study, we have noticed a rise in SWS of the basin in areas below the equator during wet periods. Similarly, a fall in SWS was observed during the 2004 drought period, confirming the critical role of climate variability on changes in surface water hydrology.