The remarkable complexity of soil and its importance to a wide range of ecosystem services presents major challenges to the modeling of soil processes. Although major progress in soil models has occurred in the last decades, models of soil processes remain disjointed between disciplines or ecosystem services, with considerable uncertainty remaining in the quality of predictions and several challenges that remain yet to be addressed. First, there is a need to improve exchange of knowledge and experience among the different disciplines in soil science and to reach out to other Earth science communities. Second, the community needs to develop a new generation of soil models based on a systemic approach comprising relevant physical, chemical, and biological processes to address critical knowledge gaps in our understanding of soil processes and their interactions. Overcoming these challenges will facilitate exchanges between soil modeling and climate, plant, and social science modeling communities. It will allow us to contribute to preserve and improve our assessment of ecosystem services and advance our understanding of climate-change feedback mechanisms, among others, thereby facilitating and strengthening communication among scientific disciplines and society. We review the role of modeling soil processes in quantifying key soil processes that shape ecosystem services, with a focus on provisioning and regulating services. We then identify key challenges in modeling soil processes, including the systematic incorporation of heterogeneity and uncertainty, the integration of data and models, and strategies for effective integration of knowledge on physical, chemical, and biological soil processes. We discuss how the soil modeling community could best interface with modern modeling activities in other disciplines, such as climate, ecology, and plant research, and how to weave novel observation and measurement techniques into soil models. We propose the establishment of an international soil modeling consortium to coherently advance soil modeling activities and foster communication with other Earth science disciplines. Such a consortium should promote soil modeling platforms and data repository for model development, calibration and intercomparison essential for addressing contemporary challenges.

No abstract

No abstract

Class pedotransfer functions were used to generate average hydraulic characteristics for distinct soil texture classes. Continuous pedotransfer functions were used to generate soil hydraulic characteristics from the actually measured median of the sand particle size, bulk density and percentages clay, silt and organic matter. Both approaches were used to predict the soil physical input data to calculate five different functional aspects of soil behaviour. The functional aspects were: number of workable days, number of days with adequate soil aeration, elapsed time until 10% breakthrough of chloride, amount of cadmium leached after one year and amount of Isoproturon leached after one year.Simulations of water and solute transport were made for 88 profiles which form a statistically representative set of profiles for cover sands in the northeastern part of the Netherlands. The calculated number of workable days did not depend on the type of pedotransfer used. However, the differences between the class and continuous pedotransfer function approach were significant for the other four functional aspects of soil behaviour. For adsorbing cadmium and adsorbing and degradable Isoproturon, differences between the two approaches were statistically significant because they were systematic. However, these differences were so small that they were irrelevant in practice. When to prefer which approach was ambiguous and depended on the functional aspect under consideration. When differences were not significant or irrelevant in practice, the cheaper and easier to use class pedotransfer function approach is preferred over the continuous pedotransfer function approach.

No abstract

No abstract

No abstract

Existing groundwater table (GWT) class maps, available at full coverage for the Netherlands at 1:50,000 scale, no longer satisfy user demands. Groundwater levels have changed due to strong human impact, so the maps are partially outdated. Furthermore, a more dynamic description of groundwater table dynamics representative for the current climate is needed. A mapping method to obtain a large set of parameters describing groundwater table dynamics was developed. The method uses time series analysis and well-timed phreatic head measurements to obtain a data set at point support. This point data set is correlated to groups of exhaustive high-resolution ancillary data by stratified multiple linear regression. Finally, simple kriging is applied to interpolate the residuals of the regression model. The method was applied in a 1,790,000 ha area and its performance was measured in 10,000 and 179,000 ha test areas. The relation between higher sampling density, mapping cost and map quality was explored. Validation results show that reasonable to good quality maps of various aspects of groundwater dynamics can be obtained by this method, at much lower cost than traditional survey-based mapping methods. The method includes the quantification of uncertainty at the actual sampling density and allows the a priori estimation of uncertainty at other sampling densities. Future research aims at identification of the effect of sources of error in ancillary data and how to diminish these. D