Cities as revealed social ties in space : unraveling co-presence patterns in urban public spaces

Abstract

With unique capacity to tolerate and assimilate various cultural strands and individuals embedded in different social networks, cities involve not only physical spaces but also complex social interactions and derived social ties among urbanites. In a contemporary metropolitan context, the ubiquitous co-presence pattern of urbanites in same time and place is a form of interpersonal intimacy construction and a pre-condition of further human interactions or urban vitality. As one unique and important element of urban life, the dense, enclosed, and mobile nature of public transport space makes it complex in the spectrum of publicness and non-publicness. It reflects the lack of a theoretical framework that is applicable to public transport space. Inspired by the "mobility turn" paradigm, the paper attempts to break through the traditional theoretical boundaries of openness, accessibility, visibility, diversity, etc. in defining static and open public space, and to provide an analysis of public transport from the perspectives of "organized physical space" and "produced social space", with a zoomed focus on the co-presence phenomenon in public transport space. With the assistance of smartcard data of metro transit system in Hong Kong, this thesis conducts three empirical analyses to answer the following research questions: 1) How to identify co-present mobility pattern in public transit system and what are the spatio-temporal characteristics of such phenomenon? 2) How does the social, physical, and network attributes of the public spaces affect the co-present ridership pattern? 3) What are the typologies of co-present passengers to form their co-present pattern and how to understand it form an urban vitality view? and 4) What are the policy implications based on the analysis results? The key findings tell a comprehensive story concerning the co-presence pattern in public transit context in Hong Kong. First, approximately 6% of the entire ridership within the Hong Kong metro transit system has been identified as co-present ridership. The travel frequency and number of co-present travel fellows both present a notable long-tail distribution. By employing a cluster heatmap approach, the distinct temporal pattern of co-presence ridership at each station has been assessed. The visualizations serve as a clear tool for planners, policy-makers, and other stakeholders to better understand the collective human interactions at a spatio-temporal lens. Second, a two-step regression model is developed to examine the relationship between co-presence ridership and a set of explanatory factors. The results reveal that co-present trips are rarely like to be conducted by high income social groups, and unpaid persons in households are more likely to travel together. In addition, a directional co-present ridership flow from less-developed area to developed area across the city has been observed. Among all urban amenities, recreational facilities show the strongest explanatory power while the diversity of Points-of-Interests is not significantly important. Third, the thesis takes advantage of the one-week-long smartcard data, and performs an in-depth typology study on the in-group dynamics of co-present travel pairs. Six types of co-present typology have been identified and a series of spatial network analyses are presented to compare the internal structure differences between the social groups. Built on the typology view, two kinds of co-present riders are extracted because their trips contain a certain length of staying intervals around specific stations, which is an informative proxy to evaluate the urban vitality at station level. Therefore, a novel indicator built on the social interaction intensity to reflect urban vitality has been developed and presented. The thesis sheds light on the deep-rooted human behaviour characteristics, nature of urban life, as well as the dynamics of different social networks in cities, and develops a comprehensive and transferrable analytical framework for other kinds of social interaction assessment in complex urban scenarios.