Paper No. 21
Presentation Time: 9:00 AM-6:30 PM


ZHANG, Zhuanfang F., Earth Systems Science Division/Hydrology Group, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, WA 99352,

Common conceptual models for unsaturated flow assume that the matric potential is attributed to the capillary force only. These models are successful at high and medium water contents but often give poor results at low water contents. The lower bound of existing water-retention functions and conductivity models was extended from residual water content to the oven-dry condition (i.e., zero water content) by defining a state-dependent residual-water content for a soil drier than a critical value. The advantages of the extended water-retention functions include not refitting the retention parameters for the un-extended model, its reduction to the un-extended form when the soil is wetter than the critical value, and its compatibility with existing relative permeability models. In addition, a hydraulic conductivity model for film flow in a medium of smooth uniform spheres was modified by introducing a correction factor to describe the film flow-induced hydraulic conductivity for natural porous media. The total unsaturated hydraulic conductivity is the sum of those due to capillary and film flow; it smoothly transits between capillary-dominated flow and film-dominated flow over the full range of water content. The film flow is insignificant when the soil is wetter than the critical water content and, vice versa, the capillary flow is insignificant when the soil is drier than the critical water content. The extended retention and conductivity models were tested with measurements. Results show that, when the soil is at high and intermediate water content, there is no difference between the un-extended and the extended models as defined by the theory; when the soil is at low water content, the un-extended models overestimate the water content but underestimate the conductivity. The mathematical models and tests with experiment data will be presented.