As climate change drives an increase in extreme weather events, understanding how urban transit systems respond to these disruptions is critical. A recent study offers an illuminating exploration into the resilience of New York City’s subway ridership when faced with severe weather conditions. Conducted by researchers from NYU Tandon School of Engineering, the University of Louisville, and the University of Hong Kong, their work employs advanced statistical modeling to uncover nuanced patterns of ridership impacts across major subway stations, presenting insights that may reshape how cities prepare for and manage transit demand during extreme weather.
Rather than treating each subway station as an independent entity, this study pioneers a methodology known as vine copula modeling, a sophisticated statistical framework that captures dependencies and joint behaviors across multiple variables—in this case, ridership at different stations. This approach recognizes that a storm or weather anomaly does not affect transit points in isolation; instead, there are intricate relationships and “signatures” in how riders collectively adjust their travel decisions, depending on the local conditions and system connectivity.
By analyzing hourly ridership data collected between 2023 and 2025 from ten of New York City’s busiest stations, the researchers probed how these interdependencies manifest under two key types of extreme weather: heavy precipitation and extreme cold. The data revealed a compelling divergence in how ridership patterns shift depending on the weather stressor and the time of day, highlighting the differentiated behavioral flexibility among NYC’s transit users.
Heavy rainfall emerges as the most impactful weather event during the evening rush hour, which spans from 4 to 5 PM. Here, the median reduction in ridership varies dramatically across stations, with Columbus Circle experiencing a decline approaching 29 percent. This sharp drop points to a willingness among commuters to rearrange or forgo discretionary travel in the face of inconvenience. Nearby stations, such as Flushing–Main Street in Queens, also see pronounced reductions nearing 26 percent. These findings suggest that while essential travel largely continues, the less urgent or more flexible trips tend to be curtailed or rescheduled when heavy rain strikes.
Conversely, the effect of extreme cold is relatively muted during traditional peak commute hours. The study observed modest median declines ranging between 1 and 2.4 percent during the 8 to 9 AM window across the evaluated stations. This subtle change reflects a prioritization of routine, with many commuters maintaining regular schedules despite harsh temperatures. However, off-peak periods paint a different picture, with discretionary trips subject to cancellation or mode shifts when temperatures plunge, underscoring that weather-induced ridership changes are contextual and time-dependent.
One of the most compelling revelations of this research lies in the pronounced differences in weather sensitivity between proximate stations. For example, Columbus Circle is notably vulnerable to rider decline during heavy precipitation, while Grand Central Terminal, less than two miles away, exhibits relative resilience with only an 8 percent median decrease. This spatial disparity indicates that borough location alone does not dictate susceptibility to weather disruptions. Instead, a constellation of factors, including infrastructure robustness, station architectural design, network connectivity, and surrounding urban land use, all influence how riders respond to adverse conditions.
Joseph Chow, a lead author and Associate Professor at NYU Tandon, emphasizes the importance of perceiving the transit system as an interconnected entity exhibiting complex, structured responses to extreme weather. He likens these interactions to unique “signatures” that, once understood, can empower transit planners to forecast and manage ridership fluctuations more effectively. By moving beyond simplistic single-station analysis, the deployment of vine copula models provides a holistic view, enabling simulation of plausible ridership scenarios under varied adverse weather conditions.
This approach carries valuable implications for urban transit resilience planning. With climate change poised to intensify the frequency and severity of storms and temperature extremes, transit agencies face mounting challenges to sustain reliable service and equitable access. Identifying stations and corridors most sensitive to weather disruptions enables more targeted investments in infrastructure upgrades, such as enhanced sheltering, drainage improvements, or adaptive scheduling that accounts for anticipated rider behavior under different scenarios.
Moreover, the study touches on social equity dimensions inherent in transit disruptions. Since certain neighborhoods rely more heavily on public transportation than others, disproportionate drops in ridership linked to weather events could exacerbate inequalities in mobility access and impose heavier burdens on communities with limited alternatives. Understanding these dynamics allows policymakers to prioritize support and alternative transit options to mitigate the uneven effects of climate-induced volatility.
Omar Wani, co-author and Assistant Professor at NYU Tandon, highlights another practical advantage offered by their methodology: the capacity to generate realistic alternative ridership outcomes that are consistent with observed data patterns but extend beyond recorded events. This capability is particularly crucial because extreme weather incidents are relatively infrequent, limiting historical data reliability. The vine copula modeling framework fills this gap by enabling planners to explore and prepare for a broader spectrum of possible situations.
Despite its contributions, the research acknowledges limitations inherent in its scope. The analysis focuses on ten high-ridership stations, leaving out the myriad smaller or less frequented stations that also compose the subway network. Likewise, the rarity of extreme weather events restricts the volume of direct observations, necessitating model-based extrapolations rather than purely empirical conclusions. Accordingly, the findings should be interpreted as estimates of likely system responses rather than definitive averages or forecasts.
In conclusion, this study offers a nuanced, data-driven lens into how New York City’s subway ridership adapts to extreme weather, revealing complex inter-station dynamics and time-sensitive behavioral shifts. The innovative use of vine copula modeling to understand resilience at a network-wide level marks significant progress in urban transit research. As cities around the globe grapple with climate change, integrating such analytical tools into transit planning could prove vital for fostering sustainable, equitable, and adaptive public transportation systems.
Subject of Research: Not applicable
Article Title: Assessing subway ridership resilience under extreme weather with vine copula modeling
News Publication Date: 1-Apr-2026
Web References:
– https://www.nature.com/articles/s44333-026-00094-4
– http://dx.doi.org/10.1038/s44333-026-00094-4
References:
– Chow, J., Wani, O., Guo, Y., He, B.Y., Su, Z. (2026). Assessing subway ridership resilience under extreme weather with vine copula modeling. npj Sustainable Mobility and Transport. DOI: 10.1038/s44333-026-00094-4.
Keywords: Transportation, subway ridership, extreme weather, heavy rain, extreme cold, vine copula modeling, urban transit resilience, climate change, New York City subway, statistical analysis
