A Large-Scale Geographically Explicit Synthetic Population with Social Networks for the United States

Area of Science:

• Geographic Information Science
• Computational Social Science
• Socio-technical Systems

Background:

• Agent-based modeling and micro-simulation require synthetic populations for realistic geo-simulation.
• Existing methods for synthetic population generation are time-consuming and often omit crucial social networks.
• Social networks are vital for understanding human interactions in simulations like disease spread or disaster response.

Purpose of the Study:

• To develop an efficient Python-based method for generating large-scale, geographically explicit synthetic populations.
• To incorporate stylized social networks (home, work, school) into synthetic population data.
• To provide a valuable resource for geo-simulation research, agent-based modeling, and urban digital twin development.

Main Methods:

• Utilized open data, including the 2020 U.S. Census, to create synthetic populations.
• Developed a Python-based approach for generating geographically explicit data.
• Integrated the creation of stylized social networks alongside demographic data.

Main Results:

• Generated a realistic, large-scale synthetic population for all 50 U.S. states and Washington D.C.
• Successfully incorporated home, work, and school social network structures.
• Created a versatile dataset applicable to various geo-simulation models.

Conclusions:

• The developed method addresses the challenges of synthetic population generation in geo-simulation.
• The inclusion of social networks enhances the realism and utility of synthetic populations.
• This work supports advanced research in human-environment interactions and urban digital twins.