Digital Trace Data and Urban Mobility Patterns: Tackling Challenges in Visualisation, Analysis, and Contextualisation

Owen, Danial (2024) Digital Trace Data and Urban Mobility Patterns: Tackling Challenges in Visualisation, Analysis, and Contextualisation PhD thesis, University of Liverpool.

Text 201057453_Apr2024.pdf - Author Accepted Manuscript Download (3MB) | Preview

Abstract

Comprehending urban mobility is crucial for understanding and addressing the dynamic and complex nature of cities. Traditionally, the understanding of urban processes and the behaviour of citizens were derived from traditional sources of data such as manual counts and travel diaries. Whilst these traditional data provide valuable information, the laborious and infrequent nature of these data collection methods have limited the ability to be able to capture, predict, and understand both granular and real-time changes in mobility patterns. In recent years, this has changed due to the prolific rise of digital trace data, where information on people’s behaviours and interactions is collected by a range of digital and technical systems. Relative to traditional data, digital trace data presents higher frequency, volume, and granularity of information, providing researchers with novel opportunities to understand a wider range of urban behaviours. Yet, despite the advantages of digital trace data, significant challenges and research gaps remain. The key challenges and research gaps addressed in this thesis focus on three main areas: visualisation, analysis, and contextualisation. The first challenge, visualisation, has arisen due to the growth in size and granularity of digital trace data. Despite offering a wealth of information, traditional movement visualisation techniques cannot harness the full potential of these data because the increase in size and granularity of movement data has resulted in a rise in visualisation issues, such as visual clutter. The second challenge is analysis, where methodological challenges and gaps in the data lead to gaps in our understanding of certain urban behaviours. Despite a wealth of information now collected by digital trace data, some important aspects of urban behaviours, specifically collective pedestrian behaviour, are still less studied and understood due to a relative lack of appropriate data. The third challenge, contextualisation, concerns the imbalance between identifying patterns and understanding their context. Despite enabling researchers to quickly generate and unlock insights and patterns, digital trace data are often not accompanied by the contextual information needed to explore and understand the relationship between movement patterns and their broader social, economic, demographic, and environmental context. This thesis aims to address these deficiencies in the context of urban areas across three empirical research chapters. The first empirical chapter explores the deficiencies for movement visualisation techniques and evaluates how vector fields, a visualisation technique typically used to depict the movement of weather, fluid, and forces, can effectively visualise the movement of 10,000 flows across Milan. In this study, I find that adapting vector fields can effectively illustrate movement patterns by tackling common visualisation challenges such as visual clutter, salience bias, and spatial information loss. The second empirical chapter utilises pedestrian counting data and a clustering approach to establish the different types of responses in collective pedestrian behaviour across the centre of Melbourne during the COVID-19 pandemic. In this study, I find distinct responses in pedestrian activity 3 across the centre of Melbourne over the study period and find that differences in responses between locations, specifically retail and sustenance venues, were largely characterised by the land-use and the changing behaviours of inhabitants. The third empirical chapter builds on the second, and addresses deficiencies in the contextualisation of movement patterns. This chapter, using a multi-level modelling approach and supported by ancillary data, measures the relationship between changing movement patterns on Chicago’s rail network both during and emerging from the COVID-19 pandemic, and stations’ social, demographic, and environmental context. By integrating low-frequency contextual data with digital trace data, this study discovers that during the pandemic, more trips were made by racial and ethnic minorities and emerging from the pandemic, less trips were made at stations where people had greater access to private vehicles and home working. Collectively, these chapters have addressed and contributed to some of the key contemporary challenges and research gaps associated with digital trace data in the context of urban mobility. Many of these contributions are in the context of the COVID-19 pandemic, where the COVID-19 pandemic fundamentally changed how cities and their inhabitants behaved. This thesis helped provide valuable insights into these changes and moving beyond the pandemic, as cities and behaviours continue to evolve and access to new and emerging forms of data continue, such insights will remain valuable for ensuring effective data visualisation, analysis, and contextualisation.

Item Type:	Thesis (PhD)
Divisions:	Faculty of Science & Engineering Faculty of Science & Engineering > School of Environmental Sciences
Depositing User:	Symplectic Admin
Date Deposited:	02 Apr 2025 14:24
Last Modified:	02 Apr 2025 14:24
DOI:	10.17638/03189919
Supervisors:	Arribas-Bel, Daniel Rowe, Francisco
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3189919
Disclaimer:	The University of Liverpool is not responsible for content contained on other websites from links within repository metadata. Please contact us if you notice anything that appears incorrect or inappropriate.