Evaluation of rapid approaches for determining the soil water retention function and saturated hydraulic conductivity in a hydrologically complex soil

Abstract Knowledge of the soil water characteristic (SWC) and saturated hydraulic conductivity ( K sat ) are required for simulation of soil water movement in most Richards’ equation based soil-water models. This study compared van Genuchten (VG) parameter values and K sat determined by desorption with two rapid approaches, evaporative flux from intact cores, and inverse solution of tension infiltration data. Estimated and measured K sat varied up to four orders of magnitude depending on parameterisation approach and antecedent soil moisture content. The SWC differed markedly between the laboratory drying and in situ wetting approaches due to the effects of hysteresis, air entrapment, clay swelling and water repellence. The evaporative flux approach was subjected to errors associated with prior saturation of the soil core, lack of K ( ψ ) data near saturation, and extrapolation of values beyond the measurement range of the upper tensiometer. Inverse solution of cumulative infiltration data from tension infiltrometers was able to discern the effects of antecedent soil moisture on soil structure and water repellence, however the approach had difficulty in obtaining unique local minima when the residual soil water content ( θ r ) was included in the objective function. Both the evaporative flux and inverse solution of tension infiltration data approaches were improved by independent determination of the θ r . Researchers should be aware that initial moisture content and methodology may influence VG parameter values, and that laboratory analysis of intact cores may result in K sat and VG values which do not represent the hydraulic behaviour of field soils.