Unraveling the impact of COVID‑19 on Beijing’s subway system using a causal machine learning analysis

Authors

• Linmu Zou
• Yanhua Chen
• Rui Guo
• Peicheng Wang
• Yanrong He
• Shiyu Chen
• Zijia Wang
• Jiming Zhu

Document Type

Journal Article

Publication Date

2025

Subject Area

place - asia, place - urban, mode - subway/metro, planning - methods, planning - personal safety/crime, ridership - behaviour

Keywords

Public health, Urban rail transit, Machine learning, Causal inference, Spatiotemporal heterogeneity

Abstract

The COVID-19 pandemic has significantly disrupted urban mobility patterns, particularly in public transportation systems. Capturing the immediate behavioral responses of populations to such disruptive events is crucial for understanding their true impact and informing rapid public health interventions. This study investigates the pandemic’s immediate impact on subway ridership in Beijing, utilizing a three-stage analytical framework that integrates machine learning, causal inference methods, and model interpretability. By conducting a multi-year comparative analysis of multiple indicators across the same period in years before and during the COVID-19 wave, we examined how various demographic, land use, network, and weather-related factors influenced subway ridership. The average treatment effect revealed a significant reduction of approximately 10,000 daily passengers per station after the COVID-19 outbreak in Xinfadi, Beijing, in June 2020. Spatiotemporal variations in the effects of key factors, such as betweenness centrality, housing prices, and the presence of restaurants and enterprises, were observed before and during the pandemic. Placebo tests were designed to confirm the robustness of our estimates. Our findings highlight the need for adaptive urban planning and evidence-based public health strategies to enhance urban resilience against future pandemics. In the face of the ongoing threat of emerging infectious diseases, this framework serves as a ready-to-deploy tool for rapid pandemic response in urban system management, offering data-driven guidance on safe and flexible transit operations.

Rights

Permission to publish the abstract has been given by Elsevier, copyright remains with them.

Comments

Sustainable Cities and Society Home Page:

http://www.sciencedirect.com/science/journal/22106707

Recommended Citation

Zou, L., Chen, Y., Guo, R., Wang, P., He, Y., Chen, S., Wang, Z., & Zhu, J. (2025). Unraveling the Impact of COVID‑19 on Beijing’s Subway System Using a Causal Machine Learning Analysis. Sustainable Cities and Society, 106709.