Projections of the characteristics of extreme precipitation in East Africa using bias-corrected CMIP6 models

In the context of ongoing global climate change, the intensification of the hydrological cycle is expected to modify the frequency, magnitude, and spatial distribution of extreme precipitation events. Understanding how precipitation extremes will evolve under climate change is critical for East Africa (EA), a region highly vulnerable to hydroclimatic hazards and strongly dependent on rainfall-driven socioeconomic systems. The region experiences two main rainfall seasons, namely the long rains (March-May, MAM) and the short rains (October-December, OND). This study investigates future projections of the characteristics of extreme precipitation across the EA using an ensemble of biascorrected CMIP6 global climate models. Extreme precipitation characteristics are quantified using selected precipitation indices representing rainfall intensity, persistence, and accumulation during both MAM and OND seasons. Projections are evaluated for the near future (2031-2065) and far future (2066-2100) under two emissions pathways, SSP2- 4.5 and SSP5-8.5. The observed climatology reveals strong spatial heterogeneity and marked seasonal contrasts, with MAM extremes dominated by rainfall persistence and OND extremes characterized by stronger spatial coherence and higher intensities. Future projections indicate a robust intensification of precipitation extremes across EA, particularly for multi-day rainfall (Rx5day) and very wet-day contributions (R95pTOT). Near-future changes show strong scenario dependence and greater uncertainty during MAM, whereas OND exhibits more spatially coherent increases even under moderate forcing. By the late 21st century, all indices showed widespread and statistically significant increases under SSP5-8.5, with OND emerging as a hotspot of extreme precipitation amplification. Overall, the results demonstrate that rather than isolated daily extremes, rainfall persistence is likely to become the dominant driver of future flood risk in EA. These findings provide robust, policy-relevant insights to support the design of climate-resilient infrastructure, improved early-warning systems, and targeted adaptation strategies across the region.