Water retention in Australian soils. I. Description and prediction using parametric functions

Abstract

The soil water characteristic (SWC) is a fundamental property controlling soil water storage and movement. The SWC is often described by analytical functions because this is convenient in the solution of numerical flow equations as well as in implementation of closed-form methods of predicting unsaturated hydraulic conductivity. The analytical functions must adequately describe the measured SWC data for these applications to be successful. We used two large sets of SWC data from Australian soils to evaluate five widely used equations on the basis of how well they describe the measured data. In general, sigmoidal equations described the measured SWCs best, but the simpler power-law equations also gave a very good description. Most of the error with the power-law functions occurred at water contents near saturation from physically unrealistic equation discontinuity around the air entry potential. The good performance of the power-law equations enabled development of a method to predict SWC data from only two measured SWC points and a knowledge of soil bulk density. When a power-law equation was used together with parabolic smoothing and parameter values were obtained using just two measured SWC points, then a very good SWC prediction was obtained within the 0 to -150 m matric potential range. This method does increase the reliance on the accuracy of measurement of the two points that are used for interpolation or extrapolation. The 'two-point' prediction method significantly reduces the cost of obtaining SWC data, thus easing an important constraint to the widespread application of soil water simulation models based on Richards' equation.