Identifying Important drivers of East African October to December rainfall season

Monsoon rainfall plays a crucial part in Africa's socio-economic structure and its year-to-year variability has profound implications for agricultural, energy, and other societal sectors. The current study focuses on two of the major climate drivers of the East African rainy season during October-November-December (OND), which is when the season starts for a large portion of SE Africa (e.g. Tanzania and Malawi). Such drivers could be different in early austral summer from the rest of the year, due to the relative positioning of the Intertropical convergence zone, which passes through this region - hence regions of East Africa and OND season are the focus here. The two drivers of Monsoon viz. El Niño Southern Oscillation (ENSO) and Indian Ocean Dipole (IOD) are considered: both independently indicate strong connections with monsoon OND rain. Not only is there a strong significant positive correlation in the OND season as a simultaneous relation, but the signal is also there even with the lag of a few months. This has been tested using various data sources, detrending data beforehand, analysing either recent time period or earlier time periods - covering two decades each, and using regression analyses. To further strengthen the results, a compositing approach is applied that can additionally identify strong rainfall signals to that from years when opposite combinations of ENSO and IOD phases act as confounding factors. Results of precipitation anomaly for OND for compositing, when IOD and ENSO are both negative (positive) in July–August-September(JAS) indicate a deficit (excess) in rainfall in that region. The Walker circulation seems to play a major part via altering ascending to descending branch in two situations, when both drivers are in same phase. In the last near thirty years period, a total of 9 years matched the criteria when both drivers were negative and suggested a deficit in OND rainfall; more recently, that criteria occurred also in 2022 (JAS) and was again associated with a rainfall deficit in OND 2022. Based on this analysis, it is possible to deliver an estimation of cumulative rain in terms of median value, range and distribution, one season in advance, at a point location or average over a region. Some results of compositing are confirmed for longer record (1940–2021) too and further classifications based on threshold of drivers are tested. Rainfall (OND) variability at intra-decadal, decadal and multi-decadal scale are studied applying the method of moving average of 5-year, 11-year and 21-year respectively. Our results have implications for future planning in optimizing agricultural and energy outputs, mitigating severe consequences and losses associated with droughts and excess rain and will impact the livelihoods of millions of Africans