Considering context and dynamics: A classification of transit-orientated development for New York City
Yunzhe Liu; Alex Singleton; Daniel Arribas-Bel (2020). Journal of Transport Geography, 85, 102711. DOI: 10.1016/j.jtrangeo.2020.102711
Abstract
Transit-Oriented Development (TOD) is a widely recognised planning strategy for encouraging the use of mass and active transport over other less sustainable modes. Typological approaches to TOD areas can be utilised to either retrospectively or prospectively assist urban planners with evidence-based information on the delivery or monitoring of TOD. However, existing studies aiming to create TOD typologies overwhelmingly concentrate input measures around three dimensions of: density, diversity and design; which might be argued as not effectively capturing a fuller picture of context. Moreover, such emphasis on static attributes overlooks the importance of human mobility patterns that are signatures of the dynamics of cities. This study proposes a framework to address this research gap by enhancing a conventional TOD typology through the addition of measures detailing the spatiotemporal dynamics of activity at transit stations; implemented for the selected case study area, New York City.
Extended Summary
This research develops an enhanced framework for classifying Transit-Oriented Development (TOD) areas that combines traditional contextual measures with dynamic patterns of human mobility to create more comprehensive urban planning typologies. Traditional TOD classification systems focus primarily on the ‘three Ds’ of density, diversity and design, which may not capture the full complexity of urban transit environments. The study addresses this limitation by incorporating spatiotemporal dynamics of activity at transit stations alongside conventional measures. The methodology employs a Self-Organising Map (SOM) combined with hierarchical k-means clustering to analyse 64 variables across four domains: Land Use and Built Environment, Transit-related, Location and Accessibility, and Socioeconomic and Demographic factors. Data sources include the American Community Survey, National Walkability Index, Smart Location Database, NYC Open Data, NYC Planning, and Metropolitan Transportation Authority turnstile data covering 472 subway stations in New York City within 800-metre walking catchments. The analysis produces four contextual TOD typologies: Commercial Core (characterised by highly educated populations in management occupations living in high-density apartment areas with excellent transit access), Blue-Collar Domicile (residents in service and production occupations living in family-oriented rented apartments), Young Family Residential (families with school-age children in detached properties with higher car dependency), and Older Family Residential (college-educated middle-aged residents in owned detached properties on the urban periphery with high vehicle ownership). Additionally, five temporal clusters emerge from turnstile data analysis: Typical Work-Oriented, Home-Work Mixed, Entertainment and Work, Off-Peak Average, and Typical Home-Oriented stations. The integration of contextual and temporal classifications reveals meaningful patterns, with Commercial Core areas predominantly corresponding to work-oriented travel patterns, whilst residential areas align with home-oriented mobility patterns. This enhanced framework provides urban planners with more nuanced evidence for TOD policy development and monitoring, capturing both the static built environment characteristics and the dynamic human activity patterns that define successful transit-oriented development.
Key Findings
- Four distinct TOD typologies identified in New York City ranging from commercial cores to residential areas with varying transit dependency levels
- Five temporal mobility clusters revealed through subway turnstile data analysis showing different patterns of work, home, and mixed-use travel behaviours
- Strong correspondence between contextual TOD classifications and temporal travel patterns validates the integrated analytical framework approach
- Traditional ‘three Ds’ approach enhanced by incorporating spatiotemporal dynamics provides more comprehensive understanding of transit-oriented development
- Self-Organising Map methodology successfully handles high-dimensional urban data without requiring normal distribution assumptions of conventional statistical methods
Citation
@article{liu2020considering,
author = {Yunzhe Liu; Alex Singleton; Daniel Arribas-Bel},
title = {Considering context and dynamics: A classification of transit-orientated development for New York City},
journal = {Journal of Transport Geography},
year = {2020},
volume = {85},
pages = {102711},
doi = {10.1016/j.jtrangeo.2020.102711}
}