A critical problem at arid sites is the characterization of the properties that determine unsaturated flow. The accepted unsaturated flow theory and models of the relevant K and water-retention relations were developed for relatively wet media.Nimmo (1991) explained the problems of using these formulations in the dry range. For example, widely used representations of water retention become physically impossible as the water content approaches zero.
We created a new analytical-formula model of the water retention relation (Rossi and Nimmo, 1994), which covers the entire range from saturation to oven-dryness in a realistic way. The model makes use of features common to all porous media in order to more accurately determine dry-range water retention, even in the absence of measurements. The theoretical basis of this model makes it possible to use computed oven-dry water pressure values in addition to measured soil-moisture retention data in modeling a complete moisture retention curve. The result is a substantial improvement in accuracy for dry-range modeling because it turns the extrapolation problem (predicting water retention points in a range drier than any measurements) into an interpolation problem (predicting water retention points in a range between the measurements and the known oven-dry point). This model has also found application in representing water relations in frozen soils. We further developed the theoretical understanding of near-zero-water-content conditions in porous media to eliminate ambiguities in the presently accepted definition of oven-dryness (Rossi and Nimmo, 1996). The improvements that these developments make possible are important in arid-region unsaturated-zone hydrology, where much of the liquid, vapor, and solute transport occurs at low water contents.
More recently, we adapted the Rossi-Nimmo (1994) retention model to relate it to similar models and to compute the property known as capillary drive (Morel-Seytoux and Nimmo, 1999) which determines infiltration rates. This development includes a new equivalence procedure between various analytical formulas commonly fit to retention and unsaturated K data, preserving (a) the capillary drive and (b) the amount of water retained in a soil layer at a given depth under capillary equilibrium.
In 2000, we are conducting experiments relating specific surface area to dry-end water retention. So far, the measurements of water retention under vapor-dominated conditions show promising correlations with specific-surface areas (which are more easily measured) for samples from semiarid regions of California and Idaho.
Other investigations related to flow at low water content are described in other sections. In the study by Nimmo and Akstin (1988), described under Quantification of soil structure, we demonstrated the usefulness of the relation of water retention to unsaturated hydraulic conductivity through capillary constructs. We also have conducted experimental tests of Darcy's law and Richards' equation under low-water-content conditions.
| National Research Program | USGS | Water Resources | UZ Flow Homepage |