A mechanistic model for electrical conduction in soil–root continuum: a virtual rhizotron study

Abstract. Electrical Resistivity Tomography (ERT) has become an important tool to study soil water fluxes in cropped field. ERT results translates to water content via empirical pedophysical relations that take soil physical properties into account, usually ignoring the impact of roots. Studies shows high root dense soils behaves quite differently than less root dense soils in terms of bulk electrical conductivity. Yet, we do not completely understand the impact of root segments on the ERT measurements. In this numerical study, we coupled an electrical model with a plant-soil water flow model to investigate the impact of plant root growth and water uptake on the ERT virtual experiment. The electrical properties of roots were explicitly accounted in the finite element mesh and we obtained the electrical conductivities of root segments by conducting specific experiments on real maize plants. The contrast between electrical conductivity of roots and soil depends on factors such as root density, irrigation, root age, and root water uptake pattern. Root growth and water uptake processes thus affect this contrast together with the soil electrical properties. Model results indicate a non-negligible anisotropy in bulk electrical conductivity induced by root processes. We see a greater anisotropy in a sandy medium when compared to a loamy medium. We find that the water uptake process dominates the bulk electrical properties. The Gauss-Newton type ERT inversion of virtual rhizotron data demonstrate that, when root-soil electrical conductivity contrasts are high, it can lead to error in water content estimates since the electrical conductivity is partly due to root. Thus, incorporating the impact of root in the pedophysical relations is very important to interpret ERT results directly as water content.