GTFS for an Interborough Express timetable

Our lab typically participates in the NYU ARISE program each year, working with high school students from NYC for about a month in the summer. This year we were fortunate to have two outstanding students, Lauren Whang and Thahera Rahman. This article highlights one of the projects that Lauren worked on: impacts of IBX designs.

The IBX is a project announced by Gov. Hochul to provide light rail service connecting Brooklyn and Queens, as shown below:

As the project is being studied by the MTA, we wanted to ask the question of how this service could potentially impact communities, particularly underserved ones in transit desert areas. Our lab previously developed proprietary data for the city and state that allows us to model mode choice behavior shifts for any census block group to block group pair (developed by Xiyuan Ren using Replica synthetic data); if we could quantify the transit travel time changes imposed by IBX, we could quantify the impacts on mode choice. That is what Lauren set out to do this summer.

Working with her mentors Hai Yang and Farnoosh Namdarpour, she used best practices to identify hypothetical stop locations and create a timetable. The timetable was then converted to a GTFS file, which has been made publicly accessible on Zenodo. Based on this GTFS file, we could then merge it with our existing MTA GTFS data set and compare changes to all zone-to-zone travel times using a transit skim tree tool from the Swiss National Railway, which could then be used to predict a number of different demand impacts using our mode choice models.

For example, we wanted to examine the Flatbush area, not only because it tends to have lower average income than the overall NYC and is predominantly minority residents, but also since we are involved in the CTAP project with Dollaride to electrify dollar vans for NYC with Flatbush as a candidate neighborhood. Based on our models, the travel time saved is highlighted in Figure 2.

Transit ridership from Flatbush is expected to increase from a prior market share of 23% (based on 2019 numbers) to a new share of 37%. While some of these trips are substituting non-motorized modes, about 4.5 percentage points of mode share are substituting driving or taxi/ride-hail. Our synthetic population data can be further broken down into different income levels and can be used to analyze the impact to the rest of the city, and quantify the impact on consumer surplus.

The GTFS file in Zenodo may be of interest to NYC transit planners, policymakers, and researchers. If there are other policy questions or interest in exploring this data more with us, feel free to leave us a comment.