Multiscale Pedotransfer Functions for Soil Water Retention

Soil hydraulic properties are needed in globaland regionalscale circulation models for hydrologic and climate forecasting. They are also necessary in pointand nonpoint-source contaminant transport models. Prediction accuracy of these models is highly dependent on the quality of the model parameters. Collecting the required soil hydraulic parameters by direct measurement at any model grid scale is expensive and time consuming. Also, to capture the variability of these spatially distributed parameters, a large number of samples needs to be collected. All of these factors make it highly impractical to match direct measurements of the soil hydraulic parameters to the invoked model grids. For these reasons, pedotransfer functions (PTFs) have been advocated to estimate the required soil hydraulic parameters from other available or more easily measurable soil data. A large number of studies have been performed in the recent past to develop such PTFs and test them against available soil properties databases (e.g., Rawls et al., 1991; van Genuchten and Leij, 1992; Schaap et al., 1998; Pachepsky et al., 1999; Wösten et al., 2001; Sharma et al., 2006). “Point regression” PTFs use empirically derived regression equations to predict the soil water content at fi xed soil water potentials (e.g., Rawls et al.., 1982; Ahuja et al., 1985; Tomasella et al., 2000). On the other hand, “function parameter” PTFs predict the parameters of the water retention and hydraulic conductivity functions (e.g., Vereecken et al., 1989; Schaap et al., 1998; Wösten et al., 2001), such as those given by Brooks and Corey (1964), Campbell (1974), and van Genuchten (1980). Both approaches have been widely utilized for various soil databases. Arguably, function parameter PTFs were preferred to point regression PTFs since they generate the complete function of the relationship between the water content and the pressure head. This made it easy to construct the entire soil water retention curve useful for modeling studies. It was recently shown, however, that point regression PTFs perform better than function parameter PTFs (Tomasella et al., 2003). Since the relationship between soil physical properties and soil water retention parameters is rather complicated, the variability in the retention parameters is controlled by different subsets of soil physical properties at different ranges of soil water pressure. Tomasella et al. (2003) suggested that the reason for the relatively poor performance of function parameter PTFs is due to their inability to accurately describe these complex physical relationships at all pressure heads from wet to dry conditions. Soil texture (sand, silt, and clay percentages) has been popularly used for predicting soil hydraulic properties. It has been shown that the use of detailed particle-size distributions can increase the accuracy of soil hydraulic parameter predictions Multiscale Pedotransfer Functions for Soil Water Retention