Abstract

Accurate knowledge of the soil hydraulic properties is necessary to simulate water flow and solute transport in unsaturated soils. The multistep-outflow and evaporation methods are frequently used transient experimental methods for the simultaneous determination of the water retention and hydraulic conductivity functions in the laboratory. Despite the wide application of the two methods, relatively little attention has been paid to verifying the agreement of soil hydraulic properties obtained from them. Our objective was to compare the soil hydraulic properties obtained from the inversely evaluated multistep-outflow method with those received from a simplified evaporation method. We conducted outflow and evaporation experiments on two different soil types and compared the soil hydraulic functions obtained from both experiments. Parameter estimation was performed by applying parametric models of different complexity and flexibility and a free-form approach to avoid parameterization errors. We found that the two methods gave very similar results when the observed processes in the experiments could be well described by the applied models. Problems arose in cases where parametric models were not able to fit the observations of one or both experimental types. This illustrates the necessity of avoiding parameterization errors in the soil hydraulic properties when results of different experiments are compared. The measurement ranges of the two methods where the results are reliable are noticeably different. Extrapolation of hydraulic functions to the moisture range not covered by the experiments is a potential source of error for flow predictions. It is crucial to acknowledge that such errors cannot be identified by applying one of the methods alone.

You do not currently have access to this article.