Analyzing the Effect of Urban Function Distribution on Overall Mobility Patterns

📌 Project Overview

This project investigates how the distribution of urban functions — such as points of interest (POIs) and workplace locations — influences human mobility within two major U.S. cities: New York and Los Angeles. By comparing these cities’ structural and transportation differences, we aim to uncover how urban form affects commute distances, accessibility, and transportation reliance.

🎯 Objectives

• Compare urban spatial structures and mobility patterns in NY and LA.
• Analyze the spatial distribution of POIs and workplaces using geospatial methods.
• Understand how urban design and transit availability affect commute behavior.
• Generate insights for urban planning and transportation policy.

🛠️ Tools & Technologies

• Python
  • GeoPandas, Shapely — for geospatial analysis
  • Matplotlib — for data visualization
• ArcGIS — for multivariate spatial clustering
• Moran’s I — index for spatial autocorrelation and clustering

Datasets Used

• OpenStreetMap POI Data — spatial layout of urban functions
• COVID19USFlows (June 2019) — human mobility patterns
• TIGER/Line Shapefiles — geographic reference for census mapping
• LEHD Origin-Destination Employment Statistics (LODES) — home-workplace mapping
• NYC MTA Turnstile Data — subway usage (New York only)

Key Findings

Spatial Clustering (Moran's I)

City	Moran’s I	P-Value
New York	0.469	0.001
Los Angeles	0.296	0.001

• New York: Dense POI clusters near downtown — transit-oriented and economically compact
• Los Angeles: More evenly distributed POIs — car-centric development and broader spatial reach

Workplace Distribution

• New York: Centralized, promoting shorter commutes (1–2 miles)
• Los Angeles: Decentralized, longer commute distances (20+ miles)

Travel Distances

• NY: Commute peaks around 1–2 miles
• LA: Commute peaks beyond 20 miles

Conclusions

• New York exhibits compact urban form and transit-based mobility.
• Los Angeles shows urban sprawl with car-reliant travel.
• Both cities display a shared short-commute trend influenced by workplace proximity.
• Urban planning should focus on:
  • Developing local hubs
  • Enhancing public transit
  • Minimizing long-distance travel needs

Future Work

• Incorporate network-based analysis of transportation infrastructure.
• Integrate machine learning models for predictive urban flow analysis.
• Expand to include more cities and longitudinal data to track changes over time.