EGU24-5880, updated on 08 Mar 2024
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Multiday mesoscale soil moisture persistence and atmospheric predictability – an illustration from the Sahel

Christopher Taylor1,2, Cornelia Klein1, and Bethan Harris1,2
Christopher Taylor et al.
  • 1UKCEH, Wallingford, Oxon, United Kingdom of Great Britain – England, Scotland, Wales (
  • 2National Centre for Earth Observation, Wallingford, Oxon, UK

The hydro-climate of the Sahel is dominated by organised Mesoscale Convective Systems (MCSs), which typically bring intense rain every few days during the West African Monsoon season. MCSs leave a swath of wet soil often hundreds of kilometres across, which in turn create strong spatial patterns of surface fluxes of heat and water back into the atmosphere. Previous studies have shown that soil moisture patterns exert a strong control on the initiation and propagation of MCSs, significantly enhancing the predictability of convection on scales of 10 – 100s km. Here, we use satellite observations to examine how this strong, locally negative, soil moisture-precipitation feedback evolves and impacts rainfall patterns over a series of storms.

We track the response of the surface and atmosphere to over 5,000 MCS events from the period 2004-2020, using a combination of satellite-derived products (Land Surface Temperature; LST, soil moisture, Vegetation Optical Depth, rainfall, cloud-top temperature). Initial anomalies in LST and soil moisture weaken rapidly in the 3-4 days after the MCS, particularly in climatologically wetter regions. However, a statistically significant memory of the original MCS event still remains in surface anomalies out to 20 days. In terms of rainfall, we see a strong suppression of convection in the first 48 hours after the MCS in areas which initially received heavy rain. There is also some evidence of enhanced MCS activity around the edges of the original swath in the first 4 days. The persistence over several days of mesoscale rainfall patterns anti-correlated with the original MCS point to an important role for surface-atmosphere feedbacks. Synoptic forcing cannot explain the finer scale rainfall response, whilst post-MCS cold pool effects are too short-lived. On longer time scales (5-20 days) in climatologically drier areas, we also find a weak but statistically significant enhancement of rainfall around the original initiation zone.

These results have important implications for rainfall forecasting on scales of tens to several hundred kilometres. Pre-existing soil moisture heterogeneity provides strong predictability of where future convection will occur under favourable synoptic conditions. This provides skill out to 2-4 days, but strongly depends on regional rainfall frequencies. Because new MCSs create new soil moisture patterns, the combination of storms every few days and a strong negative land feedback at the mesoscale actively degrades longer term predictability within the rainy season, effectively limiting intra-seasonal to seasonal forecast skill for severe weather.

How to cite: Taylor, C., Klein, C., and Harris, B.: Multiday mesoscale soil moisture persistence and atmospheric predictability – an illustration from the Sahel, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5880,, 2024.