Mechanisms Driving Drought–Flood Abrupt Alternation in East Africa during Boreal Spring

Subseasonal drought–flood alternations during the boreal spring (February–May) in East Africa pose a significant threat to water resources, agriculture, and livelihoods. Although droughts and floods have been widely examined for the entire long rainy season over East Africa, the mechanisms driving their rapid subseasonal transitions remain unclear. Here, we quantify these transitions using the subseasonal drought–flood abrupt alternation index (DFAI) applied to Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS; 0.05° × 0.05°) precipitation data and large-scale oceanic and atmospheric reanalysis datasets for the period 1981–2023. The DFAI is derived from standardized precipitation anomalies between February–March and April–May and is combined with composite anomaly analysis to investigate associated sea surface temperature and atmospheric circulation patterns. During drought-to-flood (DTF) events, the early phase (February–March) is characterized by suppressed convection [positive outgoing longwave radiation (OLR) anomalies], subsidence, low-level divergence, and upper-level convergence, followed by a rapid shift in April–May to enhanced convection, negative OLR anomalies, ascent, moisture inflow, lower-level convergence, and upper-tropospheric divergence. Our analysis shows that these reversals are driven by changes in sea surface temperature anomalies (SSTAs) over the central–eastern Pacific and western Indian Ocean, transitioning from negative SSTAs in February–March to positive SSTAs in April–May. These SSTA changes strengthen the Walker circulation during February–March (inducing easterlies and moisture divergence over East Africa) and weaken it during April–May (inducing westerlies and moisture convergence). Conversely, flood-to-drought (FTD) events display the opposite evolution. The SSTA transition over the central–eastern Pacific corresponds to the onset phase of El Niño (La Niña) for DTF (FTD) events. These insights provide a process-based framework to enhance early warning systems and climate risk management for the region.