Abstract
Mega-city regions have emerged from urban growth and improved inter-city connectivity. Due to the complexity of its spatial arrangement, quantitatively describing and predicting the functional spatial structure within mega-city regions has become a new challenge. To address this challenge, human mobility has become the hot spot of research, as it enables the exchange of ideas, goods, services, and cultural interactions that shape the dynamics of urban spaces. Human mobility is integral to the functioning of society. It could explain the relationship between micro-level individual behaviour and macro-level urban phenomenon. Therefore, this research proposes to develop an analytical framework for modelling the functional spatial structure in mega-city regions through the lens of human mobility, predicting the dynamic shift in the urban spatial structure. To achieve the main research aim of predicting the urban spatial structure, this research set a series of research objectives as follows: (1) To identify urban functional zones within mega-city regions by examining travel behaviour and differentiating intra-city from inter-city trips; (2) To develop a novel spatial interaction model that enhances travel flow predictions by incorporating residents' socio-economic characteristics; and (3) To predict the impact of urban interventions and policies on travel patterns and the mega-city region's functional spatial structure through localised changes.This study proposes several novel algorithms based on spatial-interaction models to achieve its research objectives and then tests them with case studies. The study first designed a regionalisation algorithm for delineating urban functional zones, utilising cell phone signalling data in the Great Bay Area in China. Secondly, this research proposes a novel variation of the spatial interaction model, combined with the k-means clustering algorithm, to predict the travel flows of residents using census data in the Greater London Area in the United Kingdom. Furthermore, this research integrated the tools to simulate how urban interventions and policies affect the functional spatial structure in the Great Bay Area in China from the perspective of human mobility patterns.
The primary research outcome of this study suggests that the distance decay in the spatial interaction model exhibits significant spatial heterogeneity, and this parameter could be used to represent the functional urban spatial structure. This distance decay parameter could also be associated with various factors, including spatial arrangement and non-spatial factors, such as socio-economic factors. By predicting the local variation of the distance decay with socio-economic characteristics and travel flows, we can forecast the dynamics of the urban spatial structure in the mega-city region for future scenarios. This simulation model could help governments and urban planners make informed decisions by forecasting the impacts of urban interventions on the spatial structure of mega-city regions.
Furthermore, this thesis advances the discussion on long-standing issues in spatial interaction models using human mobility big data research, such as localisation, calibration methods, and spatial heterogeneity, which contribute to solving these long-standing issues through novel approaches.
| Date of Award | 1 Sept 2024 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Zara Shabrina (Supervisor), Zhong Chen (Supervisor) & James Millington (Supervisor) |