Research Request - Realizable Transit Accessibility

[KatrinaMKaiser]

Research Question

Single sentence description: What is the difference between estimated (scheduled) and realizable access to jobs and destinations on transit statewide?

Detailed description:
Actual transit service provided (in terms of volumes, service frequencies, and speeds) may differ from the amount of service scheduled to be provided. This leads to systemic overestimation of access to jobs and destinations when performing accessibility analyses for transit. Liu et al. identify methodologies for retrospective accessibility and realizable accessibility for comparison. They further create an aggregate measure for accessibility unreliability . We would like to reproduce this analysis statewide in California.

How will this research be used?

What will be achieved or informed? What is the intended impact?

We would like to develop synthetic GTFS feeds that reflect retrospective accessibility (based on real-time arrivals) and realizable accessibility (based on the trips users plan to take based on the schedule, with actual travel times)? These could be put into Conveyal as network bundles so that the differences between scheduled and realizable accessibility may be measured and used to prioritize transit investment.

We would also like to compare unreliability across agencies to explore its use as a performance metric, and publish the results on the Analysis Portfolio with illustrative maps.

Stakeholders & End-Users

Who is asking the question, and who is the audience?

Metrics

• difference between scheduled and retrospective (+ realizable)
  • accessible stops
  • accessible jobs
  • accessible destinations
• accessibility unreliability at the following levels
  • route level
  • agency level
  • regional network level

Data sources

• Cal-ITP data sources: GTFS schedule vs rt stop times

• External data sources: jobs/destinations available in Conveyal

• Remaining data source questions: Analyst will need Conveyal access eventually, but not at first

Deliverables

Notebook, dashboard, narrative, presentation details.

• exploratory notebook developing proof of concept with a single agency
• exploratory notebook further developing proof of concept in a region with transfers between agencies
• synthetic GTFS network bundle for realizable accessibility (retrospective accessibility is first draft)
• report (format TBD) measuring difference between scheduled and realizable accessibility
• portfolio site illustrating unreliability metrics for all agencies

Timeline of deliverables

Estimated completion date TBD
Starting March 2025