Measuring soil hydraulic properties and stochastic analysis of water movement using line sources
New analytical solutions to measure soil hydraulic properties below a line source using multi-purpose time domain reflectometry (TDR) probes were developed. We use the new solutions in an inverse procedure to estimate inverse macroscopic capillary length, [alpha], and saturated hydraulic conductivity, Ks from pressure head ([Psi]), water storage (W), and solute travel time (T) measured by multi-purpose TDR probes placed below a line source with constant flux of water. The saturated soil water content, [theta]s, can also be estimated if prior information for it is available. The parameter and spatial sensitivities of each solution were calculated using sensitivity coefficients. Seven possible combinations of measurements can be used to estimate [alpha], Ks and [theta]s. The uniqueness of each approach was tested by using two-dimensional response surfaces. A field method was designed to measure unsaturated soil hydraulic properties using multi-purpose TDR probes below a surface line source with constant flux of water. The surface line source was produced inside a greenhouse using a moving irrigation system. Two hundred multi-purpose TDR probes were vertically installed in the soil beneath the line source to measure [Psi], W, and T. The soil hydraulic parameters were estimated using inverse procedures with new analytical solutions. Approximate confidence contours in the [alpha]-K s plane are used to show the accuracy of the parameter estimates. The information needed for parameter estimation is discussed. By approximating the solute travel solution, the three parameters can be obtained by optimizing the three sets of measurements sequentially. Hydraulic parameters were also measured using Guelph Permeameter (GP) and modified Guelph Pressure Infiltrometer (GPI) system. Analytical solutions of the variance of pressure head, Var[Psi] , of the variance of soil water content, Var[theta], and of the variance of water storage, Varw, were developed for steady-state water movement from a surface line source using the small-perturbation, spectral, and probability density function (PDF) methods. The solutions were tested using data from field experiments. The model predictions of both Var [Psi] and Varw were very similar to the field data except Var[Psi] at the depth 0.8 m. The existing analytical solution (Raats, 1970) for water infiltration from multiple surface line sources was, extended to water infiltration from multiple shallow trenches. Three-trench field experiments were carried out. The measured and the predicted water storage have the same horizontal variation pattern. In addition the well factor for the Guelph Permeameter method in measuring soil hydraulic parameters was expressed using empirical formulae. A software, GP-Cal, was written for processing the Guelph Permeameter data.