Footprints of Climate Predictabilityin Multi-year Malaria Risk over Africa

Climate predictability offers an opportunity to anticipate malaria risk, yet the sources of multi-year forecast skill remain poorly understood. We evaluate malaria prediction skill across Africa by forcing a mathematical–dynamical malaria transmission model (VECTRI) with CESM2-MP multi-year climate hindcasts for 1991–2020. Five major African subregions—accounting for more than half of the continent’s malaria burden—show consistently high predictive skill across lead years 1–5, although detrended skill exhibits substantial regional differences.The dominant sources of predictability vary by region and lead time. In Sub-Saharan Africa, including Malai, Burkina Faso, and South Sudan, malaria prediction skill is higher at longer lead times (LY1–5), resulting from the long-lived oceanic memory in the North Atlantic. In contrast, Central African regions such as the Democratic Republic of the Congo and Angola reveal peak skill at short lead time (LY1-2), reflecting a stronger dependence on El Niño-Southern Oscillation-related climate variability. Across all regions, surface temperature and precipitation emerge as the primary drivers of malaria predicability. These results demonstrate that distinct oceanic modes govern short- and long-lead malaria predictability across Africa, providing a physically grounded basis for climate-informed malaria early warning.