An Investigation of the Hydroclimate Variability of Eastern Africa

East Africa, spanning from 22.5°E to 52.5°E and 13.5°S to 12.5°N, has been experiencing a persistent decline of the Long Rains (March-May) for multiple decades. However, the Intergovernmental Panel on Climate Change (IPCC) projects wetter conditions. This apparent contradiction between the recent multi-decadal rainfall decline and the projected wetter climatic conditions is referred to as the East African (Monsoon) Climate Change Paradox. The purpose of this study is to investigate the climate variability associated with the East African Climate Change Paradox for the recent decades. More specifically, it addresses the following the objectives: (i) understanding the climatic factors responsible for decline in precipitation for the Long Rains over Eastern Africa, and (ii) investigating the hydrological factors involved in the decline of Lake Victoria levels in context of the decline of the Long Rains.;Although the connection between the decline and the Indo-Pacific Ocean has received much attention, the role of the Atlantic Ocean has not been recognized. Our analysis begins by showing a previously unrecognized stationary atmospheric wave form that links the northern Atlantic Ocean basin source region and the cessation of the East African Long Rains. The Atlantic Multi-decadal Oscillation (AMO) variability dominates the variability during the cessation of the Long Rains in May. The negative phase of the AMO is associated with enhanced rainfall during the cessation. In contrast, reduced rainfall occurs during the positive phase of AMO, and it has contributed to the ongoing multi-decadal decline. The projected continuation of the positive phase of AMO for several more decades by recent studies implies the likelihood of the Atlantic Oceans' potential contribution to prolong the ongoing drought conditions over East Africa. It is shown that CORDEX regional models downscaling ERA-interim reanalysis is able to capture the relationship between May precipitation over the region and AMO. However, the decline in the Long Rains is not as pronounced in CORDEX as it is in observations.;We further investigate whether the decline in precipitation in the Long Rains (MAM) is also present in other seasons and annually. We find that annual precipitation is slowly increasing, which indicates that an increase in rainfall during one or more of the remaining seasons (January-February, July-September and October-December) is compensating for the decline in the Long Rains rainfall.;The corresponding annual Lake Victoria levels modeled using observed rainfall do not decline in the recent decades, except when the Long Rains seasonal variability is considered without variability from other seasons. However, lake levels modeled using an ensemble of bias corrected CORDEX models show a decline in lake levels from 1989 to 2008 using annual precipitation, which is the opposite trend to the lake levels modeled using observed rainfall.