Taxation-deciphering indicators of plantations as the basis for the development of machine learning models for remote monitoring of forests

The solution of strategic tasks in the field of use, protection, protection and reproduction of forests in the Russian Federation, as well as improving the efficiency of protecting forests from fires, protecting forests from pests, diseases and other adverse factors, is facilitated by remote monitoring. The basis for the implementation of remote monitoring of forests is the materials of measuring taxation with a high degree of detail of work. Plantations of different ages with different adaptive potential, having different species composition and structure, located on the territory of the Prigorodny forestry of the Voronezh region, were selected as objects of study. The results of land taxation were obtained in a volume that ensures the representativeness of the sample. The basis of the integrated approach was tree-by-tree accounting of forest resources with increased detail of work, which made it possible to compile a database for the formation of training samples and the creation of machine learning models. The need for tree-by-tree accounting with the determination of coordinates is especially relevant in plantations with high canopy density. In the course of mathematical processing of the experimental material, correlations were established between the main taxation indicators. One of the direct signs, most easily determined during remote monitoring of the state of forests, is the diameter of the crown in forest plantations, which is closely related to the state of the plantation (0.71). The revealed regularity will make it possible to use crown diameters in even-aged forest stands as an additional indicator for remote monitoring of the state of forests. A set of silvicultural and taxation features of the plantation, which are of the greatest importance for remote monitoring, has been established: in addition to the crown diameter (F = 38), these are the height of the stand (F = 24) and adjoining tree crowns (F = 13).