Comparing and modelling land use organization in cities.

Area of Science:

• Urban Science
• Computational Social Science
• Geoinformatics

Background:

• Geolocated information and communication technologies offer new tools for urban studies, particularly in data-scarce regions.
• Understanding urban spatial dynamics is crucial for city planning and scientific analysis.

Purpose of the Study:

• To apply a functional network approach to mobile phone records for determining urban land use patterns.
• To systematically compare land use patterns across Spanish cities of varying sizes.
• To develop a model explaining observed land use patterns.

Main Methods:

• Functional network analysis of anonymized mobile phone records.
• Detection and classification of four major land use types based on temporal activity patterns.
• Spatial analysis of land use proportions, organization, and scaling.
• Agent-based modeling inspired by Schelling's segregation model.

Main Results:

• Identification of four distinct land use types with characteristic temporal signatures.
• Strong similarities in land use proportions, spatial organization, and scaling observed across different Spanish cities.
• Land use mixing at a local scale is highly specific to individual urban areas.
• A Schelling-inspired model successfully reproduces the observed land use patterns and scaling properties.

Conclusions:

• Mobile phone data provide a versatile tool for uncovering urban land use dynamics.
• Despite overall similarities, local-scale land use heterogeneity is a key feature of urban environments.
• Simple interaction rules can explain complex emergent land use patterns in cities.