Abstract
The unsaturated hydraulic conductivity of a soil-like material required in flow models is usually quantified with the aid of predictive models. In this study, an optimum set of input parameters for Campbell's hydraulic conductivity function was estimated. The resulting hydraulic conductivity relationship was compared to that obtained using the van Genuchten model. A flow cell was developed in which transient water contents and matric suctions were monitored while the soil underwent drying and wetting cycles. A one-dimensional numerical model was used to simulate the experimental flow problem for a wide range of input parameters. The sum of the square of the differences (SSD) between the observed and simulated water contents was calculated. The parameters corresponding to the minimum SSD were used to establish three hydraulic conductivity relationships for drying, wetting, and redrying conditions. For wetting conditions, the relative hydraulic conductivity function calibrated against the estimated optimum saturated hydraulic conductivity compared well with the measured unsaturated hydraulic conductivity. A lower-saturated hydraulic conductivity was required to calibrate the hydraulic conductivity function for drying conditions.
