Explaining COVID-19 Tokyo metro station-level ridership disruption and recovery: pre-pandemic demand, built environment and centrality

COVID-19 caused severe and uneven metro ridership declines. This study examines station-level annual ridership change in Tokyo from 2020 to 2023 relative to 2019, integrating pre-pandemic demand, built environment, and network centrality. Its main contribution is to identify how these factors shaped ridership resilience from acute disruption to recovery using ordinary least squares (OLS) and geographically weighted regression (GWR). Results show that high pre-pandemic demand and Production and Service intensity are associated with larger proportional losses, especially in 2020–2021. Road density is consistently linked to smaller declines and stronger recovery, indicating the importance of local access networks. Land-use mix becomes increasingly positive during recovery, suggesting that functionally diverse station areas better supported localized daily activities. Destination accessibility and centrality show more spatially uneven and temporally unstable effects. These findings suggest that resilience-oriented planning should combine targeted operations for vulnerable high-demand stations with local access improvements and mixed-use station-area development.