Street Networks - Alternative models, measures & their merits

Other conference contribution, 2018

Cities concentrate intensive human activities requiring highly complex networked infrastructure for movement, public transport and myriad other spatial interactions. The planning and management of multi-modal street networks for diverse users therefore requires an understanding of urban layout beyond motorised vehicle networks as simple linear conduits of movement. In seeking to address these issues, there has been a profusion of studies of street networks in recent years, with increasing attention from network scientists such as physicists, in addition to studies from transport, geographical and urban fields. These studies take different approaches to representing street networks, each with a different focus, sophistication and level of detail. The models used are based on paradigms grounded in different traditions, often with little reflection upon which is the appropriate representation of the system for a particular application. The different approaches typically generate different results without necessarily comparing means of representation and methods of analysis for different modes and contexts. A few studies have made comparisons but none has yet been done systematically across a wider set of approaches. In this international, inter-disciplinary contribution, we identify and characterise different representations of street networks, and associated measures, and compare analytic results for a sample area to bring out the commonalities, differences and relative merits of the different approaches. This review is a first step to build a foundation for deeper and more consistent understanding of the meaning and significance of the different models, and of their utility for particular applications.