Creating a modern concept for a large-scale digital soil map and updating the existing soil map

The existing large-scale Estonian soil map is a unique, widely used and has a high value, but needs to be partially updated in terms of soil units distribution, spatial accuracy, topology, as well as the related database. The main goal of the project is to develop a road map and modern methodological approach for modernizing the soil map. In order to achieve the objective, the most important activities are: - the existing vector map database is updated and new parameters are added (incl. references are established with the international WRB classification and texture distributions); - a matrix of soil change risks and a map layer reflecting probability for changes in soil will be developed, which will make it possible to distinguish between areas and data sets that need to be updated as a matter of priority; - to develop and validate a method for specifying the distribution of peat soils and eroded soils; - to add new soil classes missing in current vector map to cover distribution and attribute data in urban, and mining areas and to represent sealed soils. Classical soil field exploration and classification, laboratory analysis, remote sensing and machine learning methods and other novel technologies are combined to carry out the work. Modelling is based on combination of pedotransfer functions as well as machine learning supported by expert assessments. Previous datasets are used to develop the methodology, update and validate the map layers and additional fieldwork is carried out.