Fingerprints of anthropogenic climate change in multi-season precipitation patterns across Sub-Saharan Africa

Abstract

Understanding how precipitation is changing across Sub-Saharan Africa is critical for adaptation, as trends in precipitation directly affect water availability, food security, and ecosystem health. Precipitation trend estimates, however, depend on how precipitation is aggregated in time, for example by individual seasons or by annual totals, and climate change may affect these differently. In this study, we estimate observed precipitation trends from 1981–2024 and evaluate how evidence of externally forced climate change depends on whether seasonal trends are considered individually, as annual totals, or jointly as a multi-season pattern. We assess the likelihood of observed trends and their seasonal structure using 31 climate model simulations under forced (historical + SSP5-8.5) and pre-industrial control scenarios, applying a likelihood-ratio framework to quantify evidence of external forcing. Results show that multi-season trend patterns can provide stronger statistical evidence of externally forced change relative to individual seasons or annual totals along the margins of the tropical rain belt; however, the multi-season likelihood ratios have wider confidence intervals. These findings highlight that climate change signals in precipitation may be more evident in some regions when seasonal shifts are analyzed jointly, particularly in regions with strong seasonal precipitation contrasts, and motivate future research on how changes in the seasonal distribution of precipitation cascade through the hydrologic cycle.