High-resolution modelling of seasonal soil moisture patterns in Swiss forests: disentangling the role of vegetation structure using radiative transfer modelling

Soil moisture is a key regulator of forest health, drought risk, and hydrological extremes, yet its variability in space and time remains poorly understood, particularly in forested and topographically complex terrain. This knowledge gap hinders our ability to anticipate and respond effectively to forest disruptions under climate change. We combine long-term in-situ soil moisture data from a Swiss-wide network of topsoil sensors with high-resolution remote sensing data to identify the principal drivers of fine-scale (5m) seasonal soil moisture patterns in Swiss forests, with a particular focus on quantifying the effects of vegetation structure using radiative transfer modelling. We show that vegetation characteristics are important drivers of local soil moisture dynamics, markedly outweighing the effects of topography. Topsoils under denser canopies are significantly wetter, more likely to be at field capacity and less likely to dry out completely compared to topsoils under sparse canopies. The seasonal consistency of this effect hints at the critical importance of processes outside the growing season, and at the capacity of dense forests to enhance infiltration, as well as minimize evaporation losses. We further show how tree species and their ecological preferences imprint distinct signatures on soil-moisture dynamics and discuss future soil moisture modelling avenues. Our study underscores the importance and feasibility of accounting for local-scale drivers governing forest hydrology. The predictive approach allows for generating high-resolution Swiss-wide (5m) soil moisture maps for past, present and future climatic conditions, which offer practical value for forest management, provide reference data for validating satellite-based soil moisture products in heterogeneous forested terrain, and help us better understand past disturbances and future risks to forest functioning and ecosystem service provisioning.