Spatializing Water Tension in Heterogeneous Sandy Soils with Surface ERT During Rain-Evaporation Cycles

Dunes are almost exclusively the support of herbaceous vegetation in the Sahel. There is a need to better understand their hydrological functioning during the rain cycles. This paper evaluates the interest of using surface Electrical Resistivity Tomography (ERT) for spatializing water tension. A representative micro-dune has been studied in the field using 3 artificial rain cycles. ERT inversion was corrected from temperature and optimized using a comparison with in-situ resistivity measurements, giving an estimate of ERT accuracy. A porosity-independent relation between resistivity ratio (final / initial state) and water tension differences (final – initial) is found for the first 2 rains with demineralised water. When using salted tracer (third rain), the slope of this relationship changes. We found that the ERT image smoothing doesn’t allow us to track accurately some localized phenomena. Moreover, the interest of spatializating water tension using non-destructive surface ERT is limited because inversion has to be constrained using a priori information given by destructive measurements at depth. We conclude that surface ERT imaging reconstruction needs improvements to reliably spatialize resistivity variations linked with processes in heterogeneous sandy soils.