Topics

• The Big Picture
• Model Input
• Model Development
• Model Application
  • Transportation Corridor Projects
  • Transportation Operational Studies
  • Transit Projects
    • Modes and Terminology
    • Network Coding
    • Forecasting New Modes
    • New Starts
    • Small Starts
    • Federal Railroad Administration Guidelines
    • Validation
      • Transit Service Representation Validation
      • Transit Path-Building Validation
      • Transit Assignment Validation
    • Reasonability and Sensitivity Testing
  • Metropolitan Transportation Plans
  • Statewide Transportation Plans
  • Interregional Travel
  • Large Transportation Investment Studies
  • Pricing Studies
  • Public Private Partnership Projects
  • Congestion Management
  • Multi-Modal Projects
  • Non-Motorized Projects
  • Project Prioritization
  • Emergency Evacuation
  • Energy and Fuel Consumption
  • Multiple Goal Optimization
  • Infrastructure Sizing
  • Freight
  • Land Use and Development Analysis
  • Transportation and Air Quality Conformity
  • Climate Change and Green House Gas
  • Other Policy Analysis

Transit Path-Building Validation

 Validation of transit path-building two different types of tests.  The first is a manual review of modeled best transit paths for reasonability.  This can be a long and tedious process and requires someone with a knowledge of the transit system.

The second commonly used approach is to assign an observed transit trip table from a recent on-board survey and compare the modeled numbers of boardings for the region, by mode, and by line or line group to observed boardings.

A recent innovative approach builds from the current state-of-the-practice of assigning an observed trip table and comparing total boardings by line.  The innovative practice takes advantage of information regarding the numbers of boardings reported by individual travelers responding to the on-board survey.  Modeled boardings for specific interchanges reported by the individual travelers are then compared to the reported boardings by those travelers and prediction-success tables are constructed.  An application of this process has reported that modeled numbers of boardings for individual trips matched the reported boardings for approximately two-thirds of the travelers included in the survey data (see proceedings from 11^th TRB National Transportation Planning Applications Conference).

Since transit assignment is based on minimum impedance transit paths, there is normally full consistency between the parameters used for transit path-building (required for mode choice model application) and parameters used for transit assignment.  Consistency should also be maintained between the transit path-building parameters and estimated mode choice model parameters.

In order to properly validate a transit model, adequate independently collected data on transit travel should be available.  In other words, the data used for travel model validation should not be the same data used to estimate the travel model.  Examples of  independently collected data include transit line boarding and alighting counts and on-board survey data.  Validation data should be “current.”  A good suggestion is that the data be less than four years old.

Other Resources

1. Travel Model Improvement Program, Travel Model Validation and Reasonableness Checking Manual – Second Edition
2. Travel Forecasting for New Starts Proposals – September 2007 – The Federal Transit Administration’s Office of Planning and Environment sponsored a two-day workshop September 19 & 20, 2007, at the Adam’s Mark Hotel in St. Louis, Missouri.
3. Travel Model Improvement Program, Shining a Light Inside the Black Box webinar series, especially Sessions 2 and 3: Session 1 slides, Session 2 slides, Session 3 slides, Session 4 slides.
4. Travel Forecasting for New Starts Proposal – March 2009 – The Federal Transit Administration’s Office of Planning and Environment sponsored two three-day workshops: March 3-5, 2009, at the Radisson Hotel Phoenix City Center in Phoenix, Arizona, and March 23-25, 2009, at the Embassy Suites Hotel in Tampa, Florida
5. Transit Path-Building: “To Multipath or Not to Multipath” – Presentation at 11^th TRB National Transportation Planning Applications Conference, Daytona Beach, Florida, May 6-10, 2007.