Differential Signal of Regional Rainfall Change Among Rainfall Components in West Africa

Abstract:

Rainfall components likely differ in the magnitude and direction of their long-term changes for any given location, and some rainfall components may carry a greater signal of regional rainfall change than rainfall totals—which has been the main focus in tropical rainfall studies. This study, therefore, sought to evaluate the magnitude of change of each disaggregated rainfall component relative to all other components, and the greatest locations of change (independent of sign) across all the rainfall components in West Africa. Hourly rainfall data from the ERA5 reanalysis dataset was used to derive twelve rainfall components. Long-term means, interannual variability and long-term changes were estimated for each rainfall components. The spatio-temporal magnitude of changes among the components was estimated using absolute z-score values of the slopes of each component and the count of significant grid-point trends. Rainfall components were broadly similar in the spatial patterns of long-term means and interannual variability, but considerable differences in long-term trends exist among them. Across all components, decreasing trends generally predominate, particularly in the northwestern Congo Basin, experiencing significant drying trends and a 3% per year decrease in annual rainfall. Another important signal of change is the contrasting trends between the central Sahel (increasing) and the western Sahel (decreasing) for heavy rainfall frequency and rainfall totals. This dichotomy has been widely reported in the literature. Rainfall frequency (rain-days, moderate rainfall frequency and mean daily hours of rainfall per year) account for 53% of all combined significant trends for the whole domain. Rainfall totals and rainfall intensity account for 26% and 7% of all combined significant trends across the domain, respectively, while rainfall timing and rainfall seasonality account for the only 6%. The study finds evidence that rainfall frequency is more sensitive to the regional rainfall signal of change compared to rainfall totals and should be used as a rainfall input in climate models to potentially improve regional predictions of future rainfall.