Determining the effect of lower public transit frequencies in COVID-19 timetables on perceived door-to-door travel times using Pareto optimal range queries
Using modern transit routing algorithms it became feasible to compute all fastest travel options between all traffic zones in larger cities, allowing for an in-depth analysis how frequencies affect journeys between different origins, destinations and departure times. We use the rooftop method to calculate a realistic model of how a public transit user may perceive travel time, taking into account waiting time and/or adaption time to fit appointments in someone's schedule. The higher the frequencies, the lower those waiting times will be and vice versa. The rooftop method calculates a travel impedance for any given moment in the travel time. Furthermore, as few journeys will start and end at a transit stop, some walk component is often also involved. We sample 6693 addresses for 799 zones to compute travel times door-to-door in Amsterdam and surrounding area, explicitly including walk access and egress time to and from transit. In this study we focus on the transit timetables before, during, and in the current phase of the COVID-19 pandemic in order to investigate the effect of changed schedules on accessibility and mobility by public transit. This is particularly relevant for services that have been reduced and may remain reduced for the near future moving ahead. We expect this application of methods outlined in this paper to be of interest to public authorities and transit providers in making difficult decisions during COVID-19.
|Transit accessibility, Transit routing, COVID-19, Pareto optimal range queries, GTFS, Rooftop method|
|Annual Meeting of the Transportation Research Board|
Koch, T, & Dugundji, E.R. (2021). Determining the effect of lower public transit frequencies in COVID-19 timetables on perceived door-to-door travel times using Pareto optimal range queries. In Proceedings of the Transportation Research Board 100th Annual Meeting.