全球变化知识资源中心

This study explores the potential response of the seasonal cycle of extreme rainfall indices over Central Africa (CA) to the global warming for both the middle (2029-2058) and late twenty-first century (2069-2098), based on analysis of multi-model ensembles mean of fifteen regional climate models (RCMs) simulations. Although few dry/wet biases are still evident, for the present day climate, the RCMs ensemble mostly outperforms the driving global climate models, with a better representation of the seasonal cycle of various rainfall indices over two key sub-regions of CA chosen according to their particular rainfall patterns. Both middle and late twenty-first century project a non-significant decrease in total wet-day rainfall amount over the two analysed sub-regions, with peaks found during pre-monsoon months. We also found a significant decrease in wet-day frequency which was consistent with decreases in total wet-day rainfall amount, while wet-day intensity is projected to significantly increase. These results suggest that the decrease in total wet-day rainfall amount could be associated with less frequent events and not with their intensity. The results also have shown that dry (wet) spells are projected to significantly increase (decrease) over both sub-regions with shorter (longer) dry (wet) spells projected during pre-monsoon months. Consequently, countries within these two sub-regions could experience a more extended dry season, and therefore would be exposed to high drought risk in the future under global warming. However, changes in maximum 1-day rainfall amount, maximum 5-day rainfall amount, and 95th percentile are projected to significantly increase during monsoon months, with the maximum 1-day rainfall amount recording largest increases. Additionally, the total amount of rainfall events above the 95th percentile projects a significant increase of about 10-45 % during monsoon months, while the total number of occurrence of rainfall events above the 95th percentile projects a slight significant decrease of 4-8 % during pre-monsoon months but more pronounced for the late twenty-first century. This implies that in the future, extremes rainfall events could be more intense both in terms of rainfall amount and intensity during monsoon months. Such changes are likely to amplify the probability of flood risks during monsoon months over CA, particularly the two sub-regions. This study could therefore be an important input for disaster preparedness, adaptation planning, and mitigation strategies for Central African countries.