ИСТИНА

Urban environment is characterized by complex interactions with atmosphere due to a high concentration of heterogeneous surfaces. These interactions produce a specific urban climate that is characterized by so-called heat island and volatile atmospheric processes. Contemporary meso-scale meteorological models are becoming capable of assimilating information about urban land cover and geometry to produce more precise weather and climate forecasting results. One way to describe the complex mosaic of urban surfaces is to generalize their possible combinations to a limited number of common and frequently occurring types — local climate zones (LCZs). There are numerous methods to compute LCZs from different kinds of spatial data including proprietary spatial databases and imagery. In this study we expose preliminary findings on how land cover information can be extracted from OpenStreetMap data and used both to assess the quality of LCZs derived from space imagery and to obtain characteristics needed to direct computation of LCZs.