Compilation of a national soil-type map for Hungary by sequential classification methods

Abstract

Traditionally in Hungary the soil cover under agricultural and forestry management is typically characterized independently and just approximately identically. Soil data collection is carried out and the databases of soil features are managed irrespectively. As a consequence, nationwide soil maps cannot be considered homogeneously predictive for soils of croplands and forests, plains and hilly/mountainous regions. In order to compile a national soil type map with harmonized legend as well as with spatially relatively homogeneous predictive power and accuracy, the authors unified their resources. Soil profile data originating from the two sources (agriculture and forestry) were cleaned up and harmonized according to a common soil type classification. Various methods were tested for the compilation of the target map: segmentation of a synthesized image consisting of the predictor variables, multi stage classification by Classification and Regression Trees, Random Forests and Artificial Neural Networks. Evaluation of the results showed that the object based, multi-level mapping approach performs significantly better than the simple classification techniques. A combination of best performing classifiers, when each classifier's vote on the same object is weighted according to its confidence in the voted class, led to the final product: a unified, national, soil type map with spatially consistent predictive capabilities.

Sustainability of Forest Cover under Climate Change on the Temperate-Continental Xeric Limits

“…A GIS framework involving bioclimatic [70] and soil variables (soil type, depth, texture [71]) was set up to characterize the main variables for growing conditions in the reference period. We developed Random Forest models for yield class predictions, and separately trained the algorithm with every single subsample using climatic and soil variables as predictors.…”

Digital mapping of soil ecosystem services in Scotland using neural networks and relationship modelling—Part 1: Mapping of soil classes

“…It is difficult to assess this performance against other sources in the literature because of variability between classification systems, numbers of classes and assessment methods. However, Vincent, Lemercier, Berthier, and Walter () achieved between 41 and 72% for a regional study in France, while Pásztor, Laborczi, Bakacsi, Szabó, and Illés () achieved classification accuracies between 0.17 and 0.36 across different landscape types in Hungary using a 41‐class system. Horáček, Samec, and Minár () showed an absolute match of 26% for 22 classes in Czechia, although they also showed that a lot of the class designations per “partially accurate,” that is classification to similar classes as the target.…”

author comments

“…For the catchment spatial information on soil type, clay, silt and sand content, organic matter content, calcium carbonate content and pH in water (pH) at 100 m resolution were provided by the DOSoReMI.hu (Digital, Optimized Soil Related Maps and Information; Pásztor et al, 2018b) framework (Table 1). As soil chemical properties -organic matter content, calcium carbonate content and pH -were only available for the 0-30 cm depth, those could only be considered for the topsoil predictions.…”