A Global Intra-city Commuting Origin-Destination Flow Dataset for Urban Sustainable Development

Area of Science:

• Urban Science
• Data Science
• Transportation Engineering
• Human Mobility Research

Background:

• Commuting Origin-Destination (OD) flows are crucial for understanding urban dynamics and sustainable policy-making.
• Traditional methods for collecting OD flow data are expensive and time-consuming.
• A need exists for comprehensive, globally representative commuting flow data.

Purpose of the Study:

• To introduce a novel, globally comprehensive commuting OD flow dataset.
• To develop a deep generative model for estimating human mobility patterns.
• To provide a valuable resource for urban science and sustainable development research.

Main Methods:

• Collected fine-grained demographic data, satellite imagery, and points of interest (POIs) for 2,358 cities worldwide.
• Employed a deep generative model to capture complex relationships between urban geospatial features and human mobility.
• Generated commuting OD flows between urban regions based on these relationships.

Main Results:

• A large-scale commuting OD flow dataset with unprecedented global coverage was created.
• The generated OD flows demonstrated strong alignment with real-world spatial distributions upon validation.
• The model successfully captured intricate human mobility dynamics across diverse urban environments.

Conclusions:

• The developed dataset and methodology offer a cost-effective and efficient alternative to traditional data collection.
• This resource significantly advances research in urban science, data science, and transportation engineering.
• The findings support the development of data-driven sustainable urban development strategies.