资源环境科技发展态势分析平台

A generalized modular framework for partitioning soil hydraulic property (SHP) functions into a capillary and a noncapillary part is developed. The full water retention curve (WRC) is modeled as a weighted sum of a parametric capillary saturation function and a new general model for the noncapillary saturation function. This model is directly computed from any selected capillary saturation function. With it, a physically complete, continuous, and flexible representation of the WRC is achieved, ensuring zero water content at oven dryness. In a modular and hierarchical framework, the expressions for the capillary and noncapillary saturation function are used to calculate the respective hydraulic conductivity curves (HCC). This is achieved by adopting Mualem's integral for the capillary part of the HCC only and calculating the noncapillary HCC directly from the new noncapillary saturation function. This leads to consistent descriptions of measured HCC data, including the often observed change in slope beyond -100 cm pressure head. Compared to the classical van Genuchten-Mualem approach, it requires only one additional model parameter. The SHP framework model describes both WRC and HCC adequately and coherently. We demonstrate the suitability of the SHP framework and versatility by describing measured WRC and HCC data across the full moisture range using soil samples from a wide range of textures and origins. The modular framework was implemented in the soil physics and soil hydrology (spsh) R-package, available from the Comprehensive R Archive Network. It contains several SHP models, model parameter estimation, and features options for goodness of fit statistics, and model selection.

Plain Language Summary Soils are important features of the landscape and provide many ecosystem services related to food production and clean water. Most of the terrestrial life depends in some way or the other on this biologically, chemically, and physically highly active and complex environment. Soil water is a connecting factor linking most of the soils' properties. In order to predict a soil's ability to store and conduct water, mathematical models have been developed. These are commonly referred to as (effective) soil hydraulic property functions. Most commonly, the hydraulic property functions from van Genuchten and Mualem have been used, but these models have widely acknowledged deficiencies. In this study, we develop a modular framework model that overcomes some core deficiencies. The improvements are demonstrated by comparison with measured soil hydraulic properties from soils. We expect the framework model to have a considerable impact in increasing the precision and accuracy in predictions of water movement and evaporation from soils and ultimately crop growth.