Cluster differentiation of municipal districts of Orenburg oblast by features of rural settlement pattern

Abstract

The article presents a study of the rural settlement pattern by cluster analysis methods, which were used to typify the districts of Orenburg oblast, construct models, and reveal population distribution patterns in conjunction with nature management features. According to the results of expert estimates and correlation analysis, ten socioeconomic and landscape-ecological indices were distinguished for further clustering by different methods. To determine latent groupings of municipalities, two hierarchical algorithms and one iterative algorithm (K-averages method) were applied. The multidimensional statistical analysis, performed using the Statistica 10 program, made it possible to distinguish seven types of rural settlement patterns, combining districts with five or more common features. Four groups of districts are determined by the association of rural settlements with river valleys, two clusters are determined by the population density and the density of human settlements, and the last two, by economic and geographical location relative to transport highways. These districts are also characterized by a certain degree of uniformity in the distribution of rural people by settlement. Nearly half of all districts of the oblast are represented by three groups of districts, whose settlements tend toward watershed-type settlements; they are also distinguished by a different level of settlement density and their location along main transport highways, at a distance of 5 km from regional and federal borders. The dynamics of the rural population size is an important parameter for all clusters. Therefore, such indices as the share of settlements located within a radius of 25 km from the district center and the number of inhabitants of settlements had the lowest influence on distinguishing types of rural settlement patterns.