Large-scale multi-configuration electromagnetic induction: a promising tool to improve hydrological models

Large-scale multi-configuration electromagnetic induction (EMI) use different coil configurations, i.e. coil offsetsand coil orientations, to sense coil specific depth volumes. The obtained apparent electrical conductivity (ECa)maps can be related to some soil properties such as clay content, soil water content, and pore water conductivity,which are important characteristics that influence hydrological processes. Here, we use large-scale EMI measurementsto investigate changes in soil texture that drive the available water supply causing crop developmentpatterns that were observed in leaf area index (LAI) maps obtained from RapidEye satellite images taken after adrought period. The 20 ha test site is situated within the Ellebach catchment (Germany) and consists of a sandand-gravel dominated upper terrace (UT) and a loamy lower terrace (LT). The large-scale multi-configuration EMImeasurements were calibrated using electrical resistivity tomography (ERT) measurements at selected transectsand soil samples were taken at representative locations where changes in the electrical conductivity were observedand therefore changing soil properties were expected. By analyzing all the data, the observed LAI patterns couldbe attributed to buried paleo-river channel systems that contained a higher silt and clay content and provided ahigher water holding capacity than the surrounding coarser material. Moreover, the measured EMI data showedhighest correlation with LAI for the deepest sensing coil offset (up to 1.9 m), which indicates that the deepersubsoil is responsible for root water uptake especially under drought conditions. To obtain a layered subsurfaceelectrical conductivity model that shows the subsurface structures more clearly, a novel EMI inversion schemewas applied to the field data. The obtained electrical conductivity distributions were validated with soil probesand ERT transects that confirmed the inverted lateral and vertical large-scale electrical conductivity model. Theseresults show that multi-configuration EMI data provide detailed subsurface information up to several meters depththat can be used to improve hydrological process understanding