New European prediction algorithms developed to compute soil hydraulic properties under the leadership of the ATK Institute for Soil Sciences


Under the leadership of the ELKH ATK Institute for Soil Sciences, the methods for estimating the soil hydraulic properties have been updated. Soil hydraulic properties, computed by prediction functions from easy to measure or already available soil properties, provide essential information for climate-conscious and water-conserving land use planning and environmental modeling. The research was led by Brigitta Szabó (Tóth), Senior Research Fellow, and the results have been published in Geoscientific Model Development, a highly prestigeous journal in the field of earth sciences and modeling.

The research was carried out under the leading of the Institute for Soil Sciences at the ELKH Centre for Agricultural Research (ATK TAKI), with the collaboration of the European Commission's Joint Research Centre (EC JRC) and the University of Hohenheim.

As soil hydraulic properties affect several soil functions, knowledge of these properties is essential for crop simulation and hydrological modeling and indirectly helps adapting the planning of land use and water management to extreme weather conditions. Soil hydraulic data required for simulation studies are either not available in most cases or are expensive to measure. However, it is possible to compute them indirectly from easy to measure soil properties or soil map information. Indirect methods use prediction functions to describe the relationship between soil hydraulic parameters and easily available soil properties such as clay, silt and sand content, soil organic carbon content, and bulk density.

One of the properties that determines the water balance of the soil is the water retention capacity, which enables the water to be stored in the soil even for a longer period of time. In addition, it is important to identify the direction and speed at which water moves in the soil, which is most often characterized by the saturated and unsaturated hydraulic conductivity. The European prediction algorithms have been updated to compute the most commonly used soil hydraulic parameters – water content at saturation, field capacity, wilting point, plant available water, and saturated hydraulic conductivity, the Mualem–van Genuchten model parameters of the moisture retention and hydraulic conductivity curve –based on the European Hydropedological Data Inventory (EU-HYDI). The new methods have the advantage of providing information about prediction uncertainty, are significantly more accurate than the previous European predictions, considers a variety of input information, thus allows more flexibility in their application. The analysis was carried out in the ELKH Cloud infrastructure.

The estimation algorithms are integrated into a user-friendly web application available in both Hungarian and English, and in an open source are accessible as well.


Scatter plot of the measured versus median predicted water retention (VG) and hydraulic conductivity (MVG) values for methods with the worst and best prediction performance. PSD: particle size distribution; DEPTH_M: mean soil depth; OC: soil organic carbon content; BD: bulk density; PH_H2O: pH in water; CEC: cation exchange capacity; Count: the number of values in the area covered by the hexagons.


Szabó, B., Weynants, M., and Weber, T. K. D. 2021. Updated European hydraulic pedotransfer functions with communicated uncertainties in the predicted variables (euptfv2), Geosci. Model Dev., 14, 151–175.