This study tests modern approaches to spatial modeling of forest communities at the regional level based on a supervised classification. The study is conducted by the example of mapping the composition of forest communities in a large urbanized region (the Moscow Region, area 4.69 million hectares). A database of 1684 field descriptions is used as sample plots. As environmental variables, Landsat spectral reflectances, vegetation indices (5 images), digital elevation model and morphometric parameters of the relief, 54 layers in total, are used. Additionally, the Palsar-2 radar dataset is included. The main mapped units are formations and groups of associations identified on the basis of the ecological-phytocoenotic classification. Formations and groups of associations are similar in semantics and principles of allocation to units of forest typology. It is shown that the maximum entropy method has a wide range of applications, in particular, for mapping the typological diversity of forest cover. The method is used in combination with geographically structured spatial jack-knifing, spatial rarefication of occurrence data and independent testing of model feature classes and regularization parameters. Spatial rarefication is a critical technique when points are not evenly distributed in space. The resulting model of the spatial structure of forest cover is based on the integration of the best models of each thematic class of different types of forest cover into a single cartographic layer. It is shown that under conditions of uneven and sparse distribution of points, it is possible to provide an average point matching level of 0.45 for formations and 0.29 for association groups. Herewith, the spatial structure and the ratio of the formation’s composition correspond to the official data of the forest inventory. An attempt is made to identify and evaluate the distribution of more detailed syntaxonomic units: association groups. The necessary requirements for improving the quality of the forest cover model of the study area for 2 hierarchical typological units of forest cover are formulated. These include the additional sampling in order to equalize their spatial density, as well as to achieve equality of samples based on stratification according to the resulting map.

East European forests dominated by Norway spruce (Picea abies (L.) H. Karst.) in the broad-leaved–coniferous zone should be considered as secondary communities formed under the influence of centuries-long activities (logging, fires and planting) and extended outside their natural range. The study raises an issue—how stable is the current state of Norway spruce forests in the center of the Russian plain and what is the effect of silviculture on the forest cover of the large urban agglomeration—the Moscow Region? Current study is based on multidisciplinary research and consequently concerns the age dynamics of spruce plantation forests, the species and typological diversity of mature spruce forests and spatial pattern of spruce forests along the south edge of their range. The composition and structure of Norway spruce plantations have been studied for various age classes and compared with mature natural spruce forests and pine plantations on the basis of field data. Remote sensing data and modeling approach were applied to estimate the spatial structure of spruce forests. It is found that mature plantations (over 80 years) become similar to natural forests in terms of structure and composition. The relationship between the distribution of spruce formations and the climatic and geomorphological conditions are confirmed. The proportion of spruce and spruce–aspen/birch communities follows the pattern of zones—the transition from the coniferous and broad-leaved forest zone to the broad-leaved forest zone. Despite the significant anthropogenic impact and the high proportion of plantations in the composition of Norway spruce forests (about 60–80%), their floristic and typological diversities correspond to such properties of zonal broad-leaved–coniferous communities. Over-matured plantations can provide valuable habitats for the re-establishment of native typological diversity. This makes it possible to use silviculture stands as an accelerated alternative to the natural recovery of disturbed habitats.

The development of the composition and structure of northern taiga spruce forests on the Kola Peninsula under regeneration succession with a reduction in emissions from the metallurgical combine are described. Parameters of species diversity and ecological-coenotic structure of plant communities, including algosinusia of soils, are assessed under different technogenic impacts. Significant indices of stable transformations of plants from different layers during the regeneration succession were revealed. Strongly transformed communities in the vicinity of the combine are shown to remain in a critical state due to the preserved high level of soil pollution and demutation changes of the taiga components in the course of allogenic succession.

Trends of changes in the successional status of northern taiga forests exposed to technogenic pollution in the Kola Peninsula are discussed. Main studies of digression in spruce and pine forests are characterized on the basis of the dynamics of species composition of phytocenoses and chemical composition of atmospheric fallout, soils, and dominant plants.

Forests with predominance of Norway spruce (Picea abies (L.) H. Karst.) and Scots pine (Pinus sylvestris L.) within the hemiboreal zone are considered as secondary communities formed under long-term human activity (logging, plowing, fires and silviculture). This study raises the question—how stable is current state of coniferous forests on the southern border of their natural distribution in the center of Eastern Europe using the example of the Moscow region (MR)? The object of the study are spruce and pine forests in different periods of Soviet and post-Soviet history within the Moscow Region (MR). The current proportion of spruce forests is 21.7%, and the proportion of pine forests is 18.5% from total forest area according to our estimates. The direction and rate of forest succession were analyzed based on current composition of populations of the main forest-forming species (spruce, pine, birch, aspen, oak, linden, and ash) based on ground-based research materials collected in 2006–2019. This allowed to develop the dynamic model (DM) of forest communities with the participation of Norway spruce and Scots pine for several decades. Assessment of the spatial distribution of coniferous communities is based on field data and spatial modeling using remote sensing data—Landsat 8 mosaic for 2020. In parallel, a retrospective model (RM) of the spatial-temporal organization of spruce and pine forests for a 30-year period was developed using two Landsat 5 mosaics. For this, nine different algorithms were tested and the best one for this task was found—random forest. Geobotanical relevés were used as a training sample combined with the 2006–2012 mosaic; the obtained spectral signatures were used for modeling based on the 1984–1990 mosaic. Thus, two multi-temporal spatial models of coniferous formations have been developed. Detailed analysis of the structure of spruce and pine forests based on field data made it possible to track trends of successional dynamics for the first time, considering the origin of communities and the ecological conditions of habitats. As a result, ideas about the viability of spruce and pine cenopopulations in different types of communities were formulated, which made possible to develop a dynamic model (DM) of changes in forest communities for future. Comparison of the areas and nature of changes in the spatial structure of coniferous formations made possible to develop the RM. Comparison of two different-time models of succession dynamics (DM and RM) makes possible to correct the main trends in the transformation of coniferous forests of natural and artificial origin under the existing regime of forestry. A set of features was identified that indicates risk factors for coniferous forests in the region. A further decrease of the spruce and pine plantations and increase of the spruce-small-leaved and deciduous formations are expected in the study area. The proportion of pine-spruce forests does not exceed 3% of the area and can be considered as the most vulnerable type of forest.