Long‐Term Evolution of Rainfall Extremes and Land–Atmosphere Interactions in the Volta River Basin Using Multi‐Source Satellite Observations

Abstract

ABSTRACT Hydro‐climatic variability in tropical river basins is shaped by changes in rainfall extremes and coupled land‐surface responses. This study assesses rainfall extremes and land‐atmosphere interactions across the Volta River Basin, West Africa, using daily CHIRPS precipitation and MODIS‐derived Normalised Difference Vegetation Index (NDVI), evapotranspiration (ET), land surface temperature (LST), and Crop Water Stress Index (CWSI) for 2000–2024. ETCCDI‐style indices included consecutive dry days (CDD), consecutive wet days (CWD), Rx1day, Rx5day, and R95p. Trend inference used all 25 annual observations. Lag‐1 serial correlation was diagnosed before trend inference, and the Hamed‐Rao Modified Mann‐Kendall test was used with Sen's slope; Benjamini‐Hochberg false‐discovery‐rate adjustment was applied across the 10 tested variables. Spatial maps for six representative years were retained only for descriptive visualisation. Annual rainfall showed no significant monotonic trend (Sen's slope = 4.11 mm year −1 , modified p = 0.168). R95p, Rx1day, and Rx5day increased at nominal p < 0.05, but none remained significant after false‐discovery‐rate correction. CDD, CWD, NDVI, ET, LST, and CWSI showed no statistically significant long‐term trend under the autocorrelation‐adjusted test. Full‐series correlations were substantially weaker than estimates based on six snapshot years. Monthly validation for 2015–2024 showed close agreement between CHIRPS and GPM IMERG ( R = 0.993, R2 = 0.986, RMSE = 9.35 mm, bias = −0.96 mm; n = 120) and strong agreement with ERA5‐Land ( R = 0.958, R2 = 0.917). The results demonstrate pronounced spatial hydro‐climatic contrasts while showing that evidence for basin‐wide temporal intensification is more limited than suggested by sparse‐year analyses.