To date, numerous studies have explored the connection between water stored in the soil, also called soil moisture, and precipitation. Using coarse-resolution global climate models, these studies have consistently found a positive feedback between soil moisture and precipitation. In other words, wet soils favor rain. And as rain itself wets the soil, a positive feedback between rain and soil moisture is maintained: soil moisture matters!
In a pioneer study, Hohenegger et al. (2009), looked for the first time at this same soil moisture-precipitation feedback in km-scale simulations conducted over the Alpine region and found the opposite result: dry soils favor rain. Hohenegger et al. delved into the reasons behind these conflicting findings and related them to how different models represent convection. Coarse-resolution models rely on simplified statistical representations, known as parametrization, to describe convective processes. In contrast, kilometer-scale models represent convection explicitly by solving the underlying fluid dynamical equation (a mathematical description that explain how liquids and gases move and behave). Yet, due to computational constraints, the study of Hohenegger et al. (2009) could only simulate the climate over a small region. It could not be excluded that the lateral boundary conditions (LBCs) required at the border of that region, which are taken from a coarse-resolution global model, may spuriously affect the sign of the soil moisture-precipitation feedback (What are LBCs? Read: Davies, 2013).
Building on these insights, Lee and Hohenegger, in their 2024 study, sought to overcome the limitations of Hohenegger et al. (2009) and following studies. They employed a global, coupled climate model with explicit convection and a 5km resolution. Using the storm-resolving version of the ICON model allowed Lee and Hohenegger to represent the feedback more accurately than coarse-resolution models by representing convection explicitly, while allowing the large-scale circulation to freely evolve and interact with convection by using a global domain (a simulation over the full Earth). Their remarkable findings suggest that precipitation is less influenced by soil moisture and evapotranspiration (the combined effect of evaporation, the transport of moisture from the earth surface directly to the air, and transpiration, the transport of moisture from the soil to the air via plants) than coarse-resolution climate models have led us to believe.
The study revealed several key points:
When compared with observational data, the global, coupled storm-resolving model provided more accurate representations of the strength of the correlation between soil moisture and precipitation for over 80% of the locations, suggesting that this type of model may be better suited for global precipitation modeling.
These findings indicate that coarse-resolution climate models may overestimate the role of land cover change and of the land surface in general for precipitation. They challenge our understanding of climate over land and may indicate that precipitation patterns may be more robust than previously thought.
For more on this topic, find the entire publication here.
References:
Hohenegger, C., Brockhaus, P., Bretherton, C. S., & Schär, C. (2009). The Soil Moisture–Precipitation Feedback in Simulations with Explicit and Parameterized Convection. Journal of Climate, 22(19), 5003-5020. DOI: 10.1175/2009JCLI2604.1
Lee, J. & Hohenegger, C. (2024). Weaker land–atmosphere coupling in global storm-resolving simulation. Proceedings of the National Academy of Sciences (PNAS), 21(12). DOI: 10.1073/pnas.2314265121