Prediction of the Interannual Rainfall Variability in the Sahel: Insights from Atmospheric Circulation Patterns

The Sahel region is characterized by significant rainfall variability and has experienced hydrological changes, including a major drought from 1968 to the 1990s, followed by a subsequent period of rainfall recovery since the 1990s. Addressing this variability, this study introduces a statistical approach for predicting interannual rainfall anomalies within the region. Initially, k-means is used to classify daily atmospheric circulation patterns over West Africa, based on key variables like the V-component of wind at 700 hPa and wind speed at 200 hPa. The two high-altitude wind fields are crucial for understanding the monsoon dynamics due to their direct linkage with African Easterly Waves and the Tropical Easterly Jet. Subsequently, the annual occurrence frequencies of the atmospheric circulation patterns, along with the annual rainfall conditions in the Sahel region, are used as inputs for a multi-class logistic regression model. This model is designed to predict dry, normal, or wet years relative to the climatology. Moreover, the research presents the spatial composites of the atmospheric circulation patterns, along with a detailed explanation of the logistic regression model, an analysis of seasonal pattern occurrences, and their meteorological interpretations. The model has shown success in predicting annual rainfall variability, achieving an average proportion correct of 0.77. This level of accuracy establishes the model as a reliable tool for predicting annual rainfall amounts in the Sahel zone, offering significant insights into the climate of the region.