Predicting Unsaturated Hydraulic Conductivity from Soil Texture

Abstract

A total of 14 methods to predict $K\left(h\right)$ from soil texture and/or other data were tested. Nine of the methods involved first predicting $K\left(h\right)$ from texture and/or other data, then using a method to predict $K\left(h\right)$ from $\theta \left(h\right)\text{.}$ Using each of these methods, $K\left(h\right)$ relationships of several soils were predicted. Some of the methods required matching the predicted $K\left(h\right)$ to the measured $K\left(h\right)$ at saturation. The predicted $K\left(h\right)$ was compared statistically and graphically to measured $K\left(h\right)$ data, which is given in the literature for these soils. Results showed that for the methods that did not match the predicted $K\left(h\right)$ to the measured $K\left(h\right)$ at saturation, a method by McCuen, et al., provided the best estimate of the measured $K\left(h\right)$ Additionally, matching the $K\left(h\right)$ at saturation improved the performance of the McCuen, et al., method. The resulting method provided the best estimate of the measured $K\left(h\right)$ of all the methods tested for the loamy and clayey soils. For the sandy soils, substitution of Ghosh's closedform equation for $\theta \left(h\right)$ into the Burdine equation for K resulted in the best approximation of $K\left(h\right)$ of all the methods tested.