Soil Hydraulic FunctionsDetermined from Measurementsof Air Permeability, CapillaryModeling, and High-DimensionalParameter Estimation

摘要

Prediction of fl ow and transport through unsaturated porous media requires knowledgeof the water retention and unsaturated hydraulic conductivity functions. In the past fewdecades many diff erent laboratory procedures have been developed to estimate thesehydraulic properties. Most of these procedures are time consuming and require signifi -cant human commitment. Furthermore, multiple measurement techniques are typicallyrequired to yield an accurate characterization of the retention and hydraulic conductivityfunction between full and residual saturation. We present a more effi cient and robustapproach to estimating the hydraulic properties of porous media. Our method derives anoptimized pore-size distribution from measurements of air permeability and using recentadvances in capillary modeling and high-dimensional parameter estimation. The sectiondiameters of diff erent parallel capillaries representing the pore structure of a porousmedium are optimized with a multi-algorithm optimization method by comparing measuredand modeled air permeability values. The optimized pore model is subsequently usedto predict the water retention and water permeability functions. The predicted soil hydraulicproperties are shown to be in excellent agreement with experimentally determined datafrom an unconsolidated porous medium column. Our approach does not impose any func-tional form of the hydraulic properties and requires only measurements of air permeability.