Previous slide
Next slide

High-resolution soil particle size distribution mapping using machine learning for improved soil characterization

Researchers at the Institute for Soil Sciences have developed a new method for creating large-scale soil particle size distribution maps that show the detailed spatial distribution of sand, silt, and clay content at a resolution of 25 meters for six soil layers (down to a depth of 2 meters). The maps were generated using a geostatistical method combined with machine learning, with input data sourced from the Profile-Level Database of the Hungarian Large-Scale Soil Mapping (Hungarian acronym: NATASA). This database contains data from previous large-scale soil surveys. The new method estimates soil particle size distribution based on the upper limit of soil plasticity according to Arany, soil pH, calcium carbonate content, organic matter content, and soil type. The spatial extension of sand, silt, and clay content was further refined using topographic, geological, and land cover data. The first set of high-resolution maps was produced for the Lenti District of Zala County as a test area. However, the method is fully applicable to other areas covered by the expanding nationwide NATASA database.

These high-resolution particle size distribution maps will support a more accurate understanding of soil hydrology and water management conditions in the mapped regions. This, in turn, will aid agricultural and water management planning, helping to adapt to increasingly frequent extreme climatic events.

The study has been published in Geoderma and is available at the following link:

Kassai, P, Kocsis M, Szatmári G, Makó A, Mészáros J, Laborczi A, Magyar Z, Takács K, Pásztor L, Szabó B. 2025. Large-scale mapping of soil particle size distribution using legacy data and machine learning-based pedotransfer functions. Geoderma 454, 117178.

© Copyright ATK Minden jog fenntartva. 2023