Defining the Agricultural Wet Season in Africa Using Soil Moisture From the Soil Moisture Active-Passive Satellite

Abstract

The wet season is commonly defined based on daily precipitation accumulation, which represents water inputs but does not account for losses from evaporation, infiltration, and runoff. Here, we estimate root-zone soil moisture using observations from the Soil Moisture Active Passive (SMAP) satellite to capture year-to-year variations in seasonal soil moisture availability across Africa from 2016 to 2023 using a cumulative anomaly algorithm. Our analysis shows that seasonal soil moisture timing correlates more strongly (p (Formula presented.) 0.01) with seasonal vegetation timing than precipitation across African croplands with over 30% crop cover. Additionally, soil moisture-based onsets capture small early season rainfall events that precipitation-based methods misclassify as false onsets. However, in Southern Hemisphere woodlands, neither soil moisture nor precipitation fully explains vegetation variability, likely due to deep-rooted trees accessing moisture beyond SMAP's detection limits. These findings highlight soil moisture as a valuable indicator for refining wet season definitions, particularly in agricultural regions.