Assessing soil water content spatio-temporal variability at the hillslope scale in a headwater catchment using a multi variable interpolation model based on EMI surveys (Draix, South Alps, France)

International audience

英语. To improve flood prediction in headwater catchments, hydrologists need to know initial soil moisture conditions that precede rain events. In torrential hydrology, soil moisture mapping provides a valuable tool for investigating surface runoff generation processes. In these mountainous environments, soil moisture prediction is challenging because of highly heterogeneous land cover and soil properties. This survey propose a methodology to study spatial soil moisture variations in the mountainous and torrential environment of the Draix Bléone experimental site—Laval 0.86 km2. This approach associates water content measurements at the plot scale with spatialized soil bulk electrical conductivity (ECa) measurements combined in a multivariate statistical analysis based on topographical parameters. Between the summer of 2015 and winter of 2016, four geophysical surveys were conducted under various moisture conditions and along the same pathway, using the Slingram electromagnetic induction (EMI) technique (EM31 device) in horizontal dipole to identify changes in soil properties to a depth of 3 m. These results were analyzed to determine water dynamics in this mountainous catchment. Temporal variations of ECa vary among land cover types (forest, grassland, and black marl). A significant relationship was observed between ECa and soil water content (SWC) measured with capacitive sensors in forest and grassland. A multiple linear regression produced using the spatial interpolation code LISDQS shows a significant correlation between ECa and landform units depicted on a high-resolution DEM. ECa variations decrease with distance to talwegs. Riparian zones appear as potential hydrological contributing areas with patterns varying according to moisture status. This study shows that multiple linear regression analysis and EMI make it possible to fill gaps between SWC plot measurements, over wide areas that are steep and that present numerous obstacles due to vegetation cover.