Large-scale climate drivers of drought-to-flood events in Sub-Saharan Africa: insight from CMIP6 large-ensembles

Abstract

Sub-Saharan Africa (SSA) is increasingly experiencing unprecedented drought-to-flood events, posing critical challenges to water and food security. These rapid or seasonal transitions between extreme hydroclimatic conditions underline the urgency of advancing climate adaptation strategies and enhancing risk management frameworks in the region. However, the role of large-scale climate variability, such as the El Niño-Southern Oscillation (ENSO), Atlantic Multidecadal Variability (AMV), and Indian Ocean Dipole (IOD), in influencing decadal trends in these events across SSA remains inadequately understood.

This study aims to address this gap by evaluating how well eight single-model initial-condition large ensembles (SMILEs) from the sixth phase of the Coupled Model Intercomparison Project (CMIP6) simulate the spatiotemporal patterns of drought-to-flood events in SSA. ERA5-Land data is used as the observational reference. We also investigate potential seasonal links between the probability of drought-to-flood events and large-scale modes of climate variability.

Drought-to-flood events are defined as the sequential occurrence of a flood following a drought. To capture these events, we employ a variable threshold approach for identifying droughts, while floods are characterized using absolute thresholds (50th to 90th percentiles). To assess potential differences between meteorological and hydrological definitions of drought and flood, we compare results derived from precipitation, soil moisture, and runoff datasets.