Drought-flood abrupt alternations in East Africa during boreal spring: spatial-temporal patterns and dynamical drivers
摘要
Drought-flood abrupt alternation (DFAA) events during the boreal spring (February-May) in East Africa pose significant risks to agriculture, water resources, and livelihoods. While droughts and floods have been widely studied, the mechanisms driving their rapid subseasonal transitions remain insufficiently understood. Here, we diagnose these transitions using the Long-cycle Drought-Flood Abrupt Alternation Index (LDFAI) applied to CHIRPS precipitation data (1981–2023) together with upper-tropospheric diagnostics. Results show heightened variability and notable clustering of events after 2010, underscoring increased hydroclimatic instability. A distinct spatial asymmetry emerges: drought-to-flood (DTF) events dominate the southwest (Tanzania, Uganda, Burundi), whereas flood-to-drought (FTD) events prevail in the southeast (Somalia, southern Ethiopia). Wavelet coherence analysis identifies significant 2-8-year periodicities in DFAA variability, closely linked to ENSO and the Indian Ocean Dipole. During DTF events, enhanced wave-activity flux convergence and a strengthened Rossby wave source promote upper-level divergence and convective activation, while FTD events are associated with suppressed wave activity and negative Rossby wave forcing, inhibiting ascent and weakening rainfall. These findings provide a novel process-based characterization of subseasonal hydroclimatic reversals in East Africa, offering new insight into how large-scale climate modes translate into rapid hydrological regime shifts and suggesting pathways to strengthen seasonal forecasting and early-warning systems in this climate-vulnerable region.