Observational Evidence of Persistent Convective-Scale Rainfall Patterns

This paper examines observational evidence of a positive feedback between the land surface and rainfall in semiarid conditions. The novelty of the work lies in the length scale of study, investigating interactions between soil moisture patterns and deep convection at scales of less than 20 km. The feedback mechanism was proposed in a previous study to explain the development of an anomalous rainfall gradient in the West African Sahel. The aim here is to assess whether such rainfall persistence occurs elsewhere in the region. Convective-scale rainfall patterns are examined using two years of observations from a dense rain gauge network in southwest Niger. Rainfall differences are analyzed between neighboring gauges separated by 7.5–15 km. Under certain surface conditions, a positive correlation between daily and antecedent rainfall differences is established. These circumstances arise when previous storm patterns have modified local evaporation rates. Rainfall gradients in subsequent events tend to persist, reinforcing soil moisture patterns. The effect appears to be most pronounced in mature, large-scale storms. The widespread occurrence of persistence in the dataset provides strong observational evidence of a surface feedback mechanism, with surface-induced low-level humidity anomalies locally enhancing convection in passing storms. Several rainfall patterns that persist for a month are identified. These patterns are linked to surface processes and the frequency of storm passage.