Stochastic modeling of water movement in the saturated-unsaturated zone
Date
Authors
Major Professor
Advisor
Committee Member
Journal Title
Journal ISSN
Volume Title
Publisher
Altmetrics
Authors
Research Projects
Organizational Units
Journal Issue
Is Version Of
Versions
Series
Department
Abstract
A model for the one-dimensional transient water movement in the incorporated saturated-unsaturated zone was developed. Hysteresis in the soil water retention and hydraulic conductivity relationships was considered. The model considered layered geologic formations. The Monte Carlo method was used to incorporate the stochastic nature of hydraulic conductivity into the flow model. Outputs of the flow model include pressure head, water content at various depths, and the water table elevation. Outputs from the Monte Carlo simulation were used to predict the mean and standard deviation of those output variables;The major components of the flow model were the soil water flow, precipitation, infiltration, plant system, hydraulic conductivity, and soil water retention. The first order nearest neighbor model was used to handle the stochastic property of saturated hydraulic conductivity with the assumption that hydraulic conductivity was log normally distributed. Mualem's conceptual model, with modification for higher order scanning curves, was used for the hysteretic soil water retention. The modified Holtan's equation with Bailey's iteration method was used for the infiltration component. Van Genuchten's models for the relative hydraulic conductivity and soil water retention were used. An implicit method was used to solve for the pressure head based on the finite difference equation;Two Monte Carlo runs, consisting of 100 simulations each, were made varying the hydraulic conductivity distribution. The flow system consisted of 160 cm of soil divided into two sublayers. The spatial and temporal step sizes of 10 cm and 0.2 hour were used in the simulation. Input standard deviations of log saturated hydraulic conductivity used were 20% and 40% of the mean of log saturated conductivity. The Monte Carlo method was shown to be satisfactory when applied to study the stochastic flow problems.