Relative spatial variability in building heights and its spatial association: Application for the spatial clustering of harmonious and inharmonious building heights in Tokyo

Abstract

Whether or not a streetscape skeleton – the 3D spaces of streets defined by the arrangement of surrounding buildings – is vertically harmonious depends to a large extent on the degree of difference between the heights of buildings adjacent to one another, known as the relative spatial variability in building heights. Surprisingly, this subject has been overlooked in previous studies examining the harmony of streetscapes. Data on precise building heights are indispensable for evaluating the relative spatial variability in building heights and its spatial association. The recent relaxation of data limitations on precise building heights in Tokyo enabled us to identify the relative spatial variability in building heights and quantify its spatial association. Therefore, in this paper we aim to answer the following question: where are harmonious or inharmonious building heights locally clustered? To this end, we computed the spatial association of the relative spatial variability in building heights as a set of edges whose indices enabled us to evaluate the local indicator of spatial association (LISA). Subsequently, we statistically demarcated locally harmonious and inharmonious building heights without having to set predetermined basic spatial units. In this respect, our methods and findings are novel and can contribute to establishing a new method for measuring the variability in vertical streetscape skeletons, which is important for developing urban design policies.