Evaluating pedotransfer functions of the splintex model

Abstract

The soil water retention curve (SWRC) is an essential hydraulic function for the understanding and modelling of soil hydraulic processes. Its direct determination is time consuming and sometimes expensive because it requires extensive sampling, especially when spatial and temporal variation of soil hydraulic properties are an issue. The use of pedotransfer functions (PTFs) is a viable alternative to predict the parameters of SWRC from more easily determined soil attributes. Splintex is a physically based engineering model containing two PTFs to estimate SWRC parameters and saturated hydraulic conductivity, but its performance and functionality have not yet been fully evaluated and disseminated for soil science. We examined the functionality of the PTFs available in Splintex to estimate parameters of the Van Genuchten–Mualem (VGM) SWRC equation with a laboratory‐measured dataset containing information on particle size, bulk and particle density, saturated water content and parameters of the SWRCs. In addition, we tested the performance of both PTFs for deriving some soil physical–structural variables calculated from estimated SWRC parameters. Compared with VGM parameters fitted to the laboratory dataset, the model performed well. Its predictive performance for soil air capacity, relative field capacity and soil physical–structural quality indices was also evaluated. The performance of Splintex for estimating the VGM parameters depended on the combined effect of all input variables rather than on isolated correlations between an input variable and a model parameter.