Applying the CUSUM to a Hundred Years of Rainfall Data in Senegal

Résumé fourni par la source

The Senegalese territory is exposed to highly variable rainfall on which depend farmers in a country where crops are mainly rainfed. Previous studies on rainfall shifts have generally not placed the break years observed in the 1990s and 2000s within the context of century-long rainfall records. Such a long-term perspective is necessary to assess the significance of these recent shifts, which have been interpreted as marking the onset of a wetter period. In this study, an innovative approach was adopted to highlight the trends by minimizing the influence of erratic years. To this end, the Cumulative Sum (CUSUM) of the deviations from the mean annual rainfall was calculated using approximately a century of data (1921-2025). When plotted, the CUSUM series reveal long-term trends while mitigating the effect of interannual variability and outlying values. Consequently, beak years appear as maxima or minima on the CUSUM curves. In the arid and semiarid regions of Senegal, which account for about two-thirds of the country’s area, highly erratic rainfall often obscures long-term trends because individual years may depart markedly from the overall tendency. By analyzing the CUSUM curves, break years can be identified directly without the need for formal changepoint tests. The CUSUM method was applied to rainfall records for ten stations representing the main climatic zones of Senegal .The results indicate that the recent increase in rainfall does not constitute a sustained upward trend comparable to that which culminated in the late 1960s. Although annual rainfall totals have generally improved since the 2000s, the persistence of frequent below-average years prevent the emergence of a regular increasing trend. This high variability poses a major challenge for local authorities and government agencies seeking to implement effective measures to mitigate the impacts of both heavy rainfall events and droughts.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.