CLUSTER ANALYSIS FOR STRUCTURING SPATIAL DATA

The article discusses cluster analysis algorithms using various sets of source spatial data and software capabilities for processing them in accordance with the task. The work result is the selection of the most accurate cluster procedure and the construction of a map of the research area with selected homogeneous hydrological areas. The possibility of using cluster analysis methods to detect structures inherent in complex data sets has been substantiated. The estimation of spatial interpolation of runoff characteristics for the regionalization of hydrological data has been performed. The quality and character of hydrometeorological data is of great importance in hydrology; therefore, all characteristics must comply not only with quantitative criteria, but also be subject to accurate spatial analysis.