Application of multidimensional statistics methods in soil research

The article discusses methods of mathematical statistics based on the example of dynamic (hydrothermal, agrochemical, and biological) properties of sod-brown-podzolic soils of the Southern Cis-Baikal region. They are developed in conditions of hummocky-hollw paleocryogenic microrelief and are in a virgin, arable, and fallow state. Using the Past 4.13 program, cluster analysis (to create dendrograms that allow combining close groups) and the principal component method (PCA), multidimensional analysis, used to simplify a complex data structure, were performed. The use of modern methods of mathematical statistics made it possible to establish links and dependencies between various soil parameters, which made it possible to identify the indicators that most affect soil fertility and health, as well as to identify problem areas that require special attention.