Visual Analytics of Urban Environments using High-Resolution Geographic Data

High-resolution urban data at house level are essential for understanding the relationship between objects of the urban built environment (e.g. streets, housing types, public resources and open spaces). However, it is rather difficult to analyze such data due to the huge amount of urban objects, their multidimensional character and the complex spatial relation between them. In this paper we propose a methodology for assessing the spatial relation between geo-referenced urban environmental variables, in order to identify typical or significant spatial configurations as well as to characterize their geographical distribution. Configuration in this sense refers to the unique combination of different urban environmental variables. We structure the analytic process by defining spatial configurations, multidimensional clustering of the individual configurations, and identifying emerging patterns of interesting configurations. This process is based on the tight combination of interactive visualization methods with automatic analysis techniques. We demonstrate the usefulness of the proposed methods and methodology in an application example on the relation between street network topology and distribution of land uses in a city.