Evaluation of pedo-transfer functions for unsaturated soil hydraulic conductivity using an independent data set

The paper reviews eight well-known and accepted pedo-transfer functions used for evaluation of soil hydraulic conductivity (saturated and unsaturated) from routinely available soil data. A comprehensive data set containing detailed measurements of 63 German soil horizons, to which none of the models had been previously calibrated, were used for the evaluation of pedo-transfer functions. The paper considers only pedo-transfer functions which have shown good results in the past. The statistical analysis of the eight functions shows that the model of Wösten [J.H.M., Wösten, Pedotransfer functions to evaluate soil quality. In: E.G., Gregorich, M.R., Carter, (Eds.), Soil Quality for Crop Production and Ecosystem Health. Developments in Soils Science, vol. 25, Elsevier, (1997) 221–245.], which requires an input of measured saturated hydraulic conductivity, performs best in terms of prediction of the unsaturated hydraulic conductivity. Evaluation of the unsaturated hydraulic conductivity by the pedo-transfer functions shows, on average, better correlations if the saturated hydraulic conductivity (an input parameter) is also obtained from predictions of pedo-transfer functions rather than directly from experiments. This outcome is attributed to a dependence of the saturated hydraulic conductivity on soil structure, i.e. macropores, while the unsaturated hydraulic conductivity is generally more dependent on soil texture. The deviating results of estimated soil hydraulic conductivity, as compared to the experimental data, indicate that the use of pedo-transfer functions based only on estimated soil hydraulic parameters must be carried out with great caution.