Decoupling Scale Effects and Natural Variability to Develop Pedotransfer Functions for Soil Saturated Hydraulic Conductivity

Abstract:

The USDA-ARS database on soil saturated hydraulic conductivities included around 1000 datasets and was previously analyzed to estimate quartiles of Ksat statistical distributions for each of common 12 textural classes subdivided into two subclasses by soil porosity. Inspection of the database sources showed that samples of different sizes were used in different works. Soil hydraulic conductivity is a scale-dependent property, and can increase by half to one order of magnitude when support area increases one order of magnitude in laboratory conditions. The objective of this work was to reanalyze this database accounting for differences in sizes and to develop pedotransfer relationships for the dominant support area should such area be found. The range of support area from 20 cm² to 50 cm² was well represented in the database and was selected to analyze statistical distributions of Ksat for each of 24 abovementioned subclasses. The distributions were close to lognormal, and an excellent dependence of both geometric means and standard deviations on the average clay content of subclasses was established. The new table of Ksat values for textural classes in two porosity ranges provides substantially less variability in estimates. Decoupling scale effects and natural variability allows one to focus on derivation of scaling relationships to obtain Ksat estimates appropriate for the scale of a specific soil hydrology project.