Mobility census for monitoring rapid urban development

Area of Science:

• Urban Studies
• Data Science
• Geospatial Analysis

Background:

• Traditional censuses provide valuable urban data but may lack the temporal resolution needed for rapidly evolving cities.
• Big data sources like human mobility data offer potential but present challenges due to noise and unstructured formats.

Purpose of the Study:

• To develop and validate a method for extracting meaningful variables and classifications from human mobility big data.
• To demonstrate the utility of this method for monitoring urban structure and development at high spatio-temporal resolution.

Main Methods:

• Utilized human movement data derived from mobile communication in Beijing.
• Developed a novel data analysis technique to extract explanatory features and classifications from noisy, unstructured mobility data.
• Achieved analysis at 500 m spatial resolution and near real-time frequency with high computational efficiency.

Main Results:

• Successfully extracted meaningful features from mobile communication data, revealing urban dynamics.
• Demonstrated the emergence and absorption of sub-centres within the urban structure.
• Showcased the method's suitability for tracing event-driven mobility changes and their impact.

Conclusions:

• Human mobility big data, when analyzed with appropriate methods, can effectively complement traditional data for urban monitoring.
• The proposed method enables high-resolution, near real-time analysis of urban dynamics, crucial for understanding contemporary urban development.
• This approach offers a computationally efficient way to track the impact of mobility changes on urban structures.