A flexible framework for measuring accessibility with destination bundling

Abstract

A transit system's usefulness is governed by the freedom it provides to those who use it. This freedom, typically quantified as accessibility, is proportional to the amount and variety of destinations available to a potential transit user. Often, transit systems are designed with the commuter in mind; employment is a typical stand-in measure for all destinations when measuring accessibility in a city. This paper proposes a framework to “bundle” destination types into a more comprehensive profile of accessibility. The framework is flexible enough to adapt to local conditions and data availability, and provides a potential planner with the ability to tell a more nuanced story of transit accessibility in a city. Using population, employment, and crowd-sourced destination data in Calgary, Canada, we perform a comparison of destination bundling approach to find that the relative level of access to destinations varies greatly with the bundle of destinations used. We also analyze correlations between quality of access to destinations, suggesting that certain destinations can act as substitutes for others, and that using destinations with low correlations in their quality of access increases the results' sensitivity to the transit network. As this approach uses open data sources available in most jurisdictions, it can be easily applied to different urban areas, destination sets, and accessibility measures to tell a more comprehensive story of transit accessibility in cities.