Soil depth modelling for the Upper River Cinca catchment

Abstract

Soil depth has a major influence on hydrological processes, vegetation productivity and drought resilience in mountainous environments. However, soil depth is difficult to measure at catchment scale, due to rugged terrain and time and costly expenses of field observations. Furthermore, modelling studies are often limited to the top tens of centimetres of the soil. Together, this limits the availability of spatial soil depth information needed for ecohydrological modelling.

This study develops and validates a physical based empirical soil depth prediction model for the Upper River Cinca Catchment in the southern central Pyrenees. A 30 m resolution terrain digital elevation model (DEM) was used to derive terrain attributes. Mean curvature was combined with downscaled ERA5 rainfall data, using the CHELSA method, to predict soil depth in the Cinca catchment. This map was validated using sixty field observations spread out across combinations of elevations, slopes and slope direction. The resulting soil depth map was subsequently applied to estimate soil water residence time, from which a weighted drought susceptibility map was developed, further including rainfall and land cover data.

Predicted spatial soil depth patterns are closely related to topography, with shallow soils on ridges and deeper soils in surrounding valleys, consistent with general soil production and transport theory. However, no relationship was found between soil depth and elevation, slope, aspect, rainfall, or lithology and soil depth. This suggests that the current model does not account well for localized variations in soil depth. Furthermore, absolute soil depth predictions remain uncertain in this area. This indicates that improving limitations in both methodology and field observations, could greatly improve the predicted soil depth outcome. Soil water residence time and drought susceptibility patterns demonstrate that ridges with shallow soils have a lower water holding capacity and are in combination with sparse vegetation cover and low rainfall rates more susceptible to drought. This illustrates how soil depth patterns can be used to determine areas with high structural drought susceptibility.

Overall, this study highlights that high resolution DEMs can be utilized to predict catchment scale spatial patterns of soil depth, providing inputs for ecohydrological models. By connecting curvature derived soil depth predictions to soil water residence time and vegetation drought susceptibility mapping, this study demonstrates the potential of terrain attribute based soil depth modelling, to improve soil depth modelling development towards a more accurate approach for ecohydrological modelling inputs.