Mapping urban form and function at city block level using spatial metrics

Abstract This paper focuses on the potential of urban metrics describing the presence and the configuration of built-up and open space areas for mapping distinct types of urban form and function at city block level. Next to traditional, patch-based metrics used in landscape ecology, alternative metrics are proposed, measuring the presence and the spatial arrangement of built-up and open space areas along a set of radial transects, along contours parallel to the urban block boundary and along the block's perimeter, as well as metrics describing the internal composition of the built-up area. Use of the proposed metrics for identifying different types of urban form and function was tested on the Brussels Capital Region. Large-scale vector data was used to define built-up structures and to analyse the morphological properties of the built-up area at block level. Decision tree classification was applied in conjunction with bootstrap aggregation to gain insight in the distinctive character of the defined metrics, and the robustness of land use and urban form classification based on these metrics. Our study points out the shortcomings of traditional landscape ecological metrics for mapping urban form and emphasizes the need for alternative approaches for analysing urban landscapes, more explicitly describing the morphological characteristics of the urban fabric.