Prediction of unsaturated soil hydraulic conductivity with electrical conductivity

Soil hydraulic conductivity (K) varies greatly with matric potential (h) and exhibits a high variability at the field scale. However, this key property for estimating water flux in soils is difficult to measure. In contrast, soil electrical conductivity (sigma) is easier to measure and is influenced by the same parameters affecting K. We derive a simple relationship between sigma and K(h) and test it against laboratory and literature data. Importantly, we show that parameters of this sigma-K(h) relationship can be completely determined with accessible measurements of saturated hydraulic conductivity, electrical conductivity of the soil solution, and clay content. This results in K(h) estimation with a RMSE ranging between 0.4 and 0.5 for log K, i.e., of the order of most experimental determinations of K. A further test of the sigma-K(h) relationship on the large UNSODA hydraulic database shows good agreement and the robustness of the relationship. Such a relation could be useful in the spatial monitoring of water fluxes at the field scale using electrical resistivity tomography if the sigma(h) relationship can be obtained.