Soil maps : Predictive mapping of soil properties for the evaluation of soil functions at regional scale

In the “Soil maps” project, researchers used methods of digital soil mapping and assessment to generate spatially highly resolved soil property maps and spatially assess soil functions.

  • Background (completed research project)

    Dropdown Icon

    Soils play an important role in ecosystems and provide important services for humans, for example by retaining water after heavy rainfall or by filtering and clearing water. Most soil functions cannot be measured directly but must be deduced from soil properties by modelling. Spatial information on soil properties is however only available for about 30 per cent of the agricultural area in Switzerland.

  • Aim

    Dropdown Icon

    The project aimed to develop methods for creating high-resolution maps of soil properties using digital soil mapping and for converting such information to soil function maps.

  • Results

    Dropdown Icon

    The team mapped properties and functions of soils in three study regions in the cantons of Berne and Zurich. First they spatially modelled soil texture, content of stones and humus, hydromorphic features, soil depth, pH, cation exchange capacity, exchangeable cations and bulk density. To this end they fitted geo-additive models to harmonised legacy soil data and environmental covariates that characterise soil formation factors (topography, geology, climate, vegetation, land use). Using established assessment methods from Germany and Switzerland, maps of soil function scores were computed for agricultural production, regulation of nutrient and water cycles, and filtering of contaminants from the soil property maps.

  • Implications for research

    Dropdown Icon

    The project advanced methods for digital soil mapping and assessment: the team developed a workflow for harmonising legacy soil data, explored how to best account for topography in spatial modelling of soil properties, developed methods for optimal use of airborne imaging spectroscopy data and implemented a new machine learning method for digital soil mapping.

  • Implications for practice

    Dropdown Icon

    Spatially explicit information on soil functions is a prerequisite for the sustainable use of soils. Missing spatial soil information is a major obstacle to the inclusion of soil quality in spatial planning and in agricultural policies. The pilot study shows how spatial soil information can be created from legacy soil data and spatial information on soil formation factors.

  • Original title

    Dropdown Icon

    PMSoil: Predictive mapping of soil properties for the evaluation of soil functions at regional scale