Aims of the PTF working group
Aims of the PTF working group
The working group ‘Pedotransfer functions and land surface parameterization’ aims to bring together international experts working on pedotransfer functions and land surface parametrization in different disciplines such as soil sciences, climate, and crop modelling. Hereby, the focus will be in a first step on pedotransfer functions (PTF) to estimate soil hydraulic parameters. In addition, also thermal and biogeochemical pedotransfer functions will be tackled. Within the working group urgent needs in pedotransfer and land surface parameterization development and validation will be identified.
To point 1 and 2. Currently, there are no pedotransfer functions available which directly account for soil structural effects, albeit that the effect of soil structure on the soil hydraulic properties is well recognized in the soil hydrological community. Different conceptual models to account for soil structure effects on soil hydraulic properties are available such as the Dual porosity concept of Durner (1994) or the Dual-Permeability Models of Gerke and van Genuchten [1993a, 1996]. Especially, the Durner model can be easily used to develop PTFs, as there is a direct link of the parameters to observable data in the retention and unsaturated hydraulic conductivity function. A first trial to establish Durner based PTFs is currently performed by Vereecken and Weihermüller.
To point 3. Recently, Lehmann et al. (2020) showed that PTFs report a and n parameter (in van Genuchten (1980) combinations that may be in disagreement with physical soil properties. To overcome this problem, Lehmann et al. (2020) proposed to use a physical constraint, namely, the soil specific characteristic length of evaporation, to reduce unphysical combinations of the estimated parameters. There are alternative physical constraints that may be exploited here in future studies, e.g. the characteristic infiltration time. Furthermore, revisiting existing databases and directly including dependencies between the parameters in the PTF development is a promising approach that is worth trying. Machine Learning techniques such as regression trees and physically constrained artificial neural network (ANN) may be useful in this respect
To point 4. In most cases, PTFs have been validated either against parts of the data reported in the databases from which the PTFs has been developed or against independent measurements. Many publications have already compared PTFs with respect to measured hydraulic properties. However, PTFs have been rarely analysed in terms of water flow, as there are only limited data, which can be used for such purposes such as lysimeter studies. As an alternative, functional sensitivity studies using model predictions can be used, whereby the reality in this case remains unknown. A first study in the last direction is currently performed by Weihermüller.
To point 5. Recent advances in both observational and computational techniques make it now possible to obtain the three-dimensional structure of the pore-space and conduct simulations of fluid flow at the pore-scale, allowing us to derive soil hydraulic parameters numerically. An alternative to investigate the effects of soil structure on the soil hydraulic properties is to apply pore-scale modelling to different soil structures and then simulate the soil hydraulic parameters by considering the effects with and without soil structure. The index used to quantifying soil structure includes, but is not limited to the following, such as pore size distribution, surface density, mean curvature, pore connectivity, pore distance, tortuosity, and hydraulic radius. The optimal index will be selected to quantify soil structures, which is likely to be used as input for the next generation of soil pedotransfer functions.
To point 6. Soil hydraulic information in Land Surface and Earth System models is derived from pedotransfer functions. Correct quantification of soil hydraulic information affects the simulation of infiltration, runoff, groundwater recharge, etc. After changing the PTF with e.g. ones accounting for soil structural effects or physical constraint parameters, the new models should be validated against large-scale observations of soil moisture, evapotranspiration, and water and energy flux between land and atmosphere. A first trial study is currently performed by Dani Or's group.
References
Durner, W. (1994). Hydraulic conductivity estimation for soils with heterogeneous pore structure. Water Resources Research. 32(9): 211-223.
Gerke, H.H. and M. Th. van Genuchten. (1993). A dual-porosity model for simulating the preferential movement of water and solutes in structured porous media. Water Resources Research. 29: 305-319.
Gerke, H.H. and M. Th. van Genuchten. (1996). Macroscopic representation of structural geometry for simulating water and solute movement in dual-porosity media. Adv. Water Resour. 19: 343-357.
Lehmann, P., S. Bickel, Z. Wei, and D. Or. (2020). Physical constraints for improved soil hydraulic parameter estimation by pedotransfer functions. Water Resources Research. 56. e2019WR025963: https://doi.org/10.1029/2019WR025963.
van Genuchten, M. T. (1980). A Closed-form Equation for Predicting the Hydraulic Conductivity of Unsaturated Soils. Soil Science Society of America Journal, 44(5), 892–898.