Vermont Agency of Transportation (VTrans) Statewide Travel Model Peer Review Report
7.0 Panel Discussion and Recommendations
After reviewing VTrans's application for peer review, participating in the three initial peer review sessions, and providing individual comments based on previous experiences, the panelists' recommendations were aggregated and presented to VTrans and their associated staff.
The following summarizes the panel's comments and recommendations on the topics of interest to VTrans, as well as general guidance for the future of the statewide model. This summary follows the panel's final presentation to VTrans at the concluding Session #4 of the peer review held on July 31, 2013.
7.1 General Comments and Recommendations
First and foremost, the panelists emphasized the importance of defining one consistent model platform, either CUBE or TransCAD, and maintaining all data and processes in the specified format. This consistent platform would allow for simplicity in validation and calibration efforts and increase the reliability of the modeling outputs.
Next, the panelists highly recommended that VTrans internally strengthen their agency's understanding of the model, specifically with regard to its sensitivities and appropriate uses for model application at the statewide level. The panel recommended that the model developer, whether in-house or external, provide features in support of desired analysis by the agency. Furthermore, a VTrans staff person should have a strong understanding of the application side of the model in order to conduct defendable analyses. The panel also noted it critical that agency staff are able to illustrate the value of the model as a planning tool to gain financial support from agency management.
Another overarching issue discussed in the peer review sessions was the need for VTrans to minimize dependence on the statewide model by developing tools in addition to the statewide model that have the ability to meet agency needs while managing resources and effort. Particularly, the panelists noted that one model cannot provide the analytical power required for different levels of spatial acuity. Therefore, VTrans would benefit from maintaining a variety of tools to meet analytical needs. The tools should be consistent and compatible with each other and use data collected by the agency in a streamlined and automated manner.
Finally, the panel underscored the importance of identifying project types and metrics desired for project prioritization prior to the redesign of model features. Panelists lauded the ambitious nature of VTrans's model enhancement goals; however, noted that it will be imperative to first achieve basic functionality and incorporate comments from FHWA before any mid- to long-term goals that require extensive model development efforts are realized.
The following subsections partition panelist comments by potential timeframe for implementation: short-, mid-, and long-term.
7.2.1 Recommended Shorter-Term Priorities
The panel feels that VTrans should focus on the following priorities in the next year:
- Address the comments from FHWA's review of the current model:
- Undertake the list of fundamental model development considerations from FHWA provided in Section 5.1.
- Develop a statewide model users' guide and technical reference.
- Define short/mid/long term priorities based on the current model to create a detailed model development plan.
- Include new tools or model metrics for resiliency planning in the model:
- Recognize that emergency contingency planning is associated with links damaged by an emergency event not general facility design; therefore, the consideration of dynamic traffic assignment to assess traffic patterns in emergency response may be a preferable method.
- Identify metrics for emergency scenario comparison to guide model development if the agency selects the model as the tool for resiliency planning.
- Develop an at-risk location inventory in the model network via link attributes and automate their incorporation into the network if the agency selects the model as the tool for resiliency planning.
- Incorporate various model improvements to address model network and structure issues identified by the peer review panelists:
- Enlarge the external model area by including a halo over the state line.
- Ensure that the roadway network includes all interstates, major arterials, and collectors with accurate speeds, lengths, and classifications.
- Reassess centroid connectors.
- Consider seasonal trip tables.
- Differentiate between short- and long-distance trips.
- Expand to a future year beyond 2030.
- Decide on one freight model component based on either commodity flows or truck/rail vehicles.
- Review the following references for additional ideas for statewide modeling best practices:
- Special Report 288 "Metropolitan Travel Forecasting"
- TCRP Report 95 " Traveler Response to Transportation System Changes Handbook"
- NCHRP Project 836-B Task 91 "Final Report: Validation and Sensitivity Considerations for Statewide Models"
- NCHRP Report 735 "Long-Distance and Rural Travel Transferable Parameters for Statewide Travel Forecasting Models"
- NCHRP Synthesis 406 "Advanced Practices in Travel Forecasting"
- A Transportation Modeling Primer, Edward A. Beimborn Center for Urban Transportation Studies University of Wisconsin-Milwaukee, May 1995, Updated June 2006
7.2.2 Recommended Mid-Term Improvements
Over the next two to three years, the panel recommended VTrans consider the following:
- Establish a methodology for evaluating system preservation and disinvestment:
- Coordinate with pavement program staff to determine need for this type of effort.
- Identify the performance measures desired for project prioritization prior to adjusting the model.
- Consider evaluating volumes and road wear for project prioritization.
- Review Oregon's use of HERS-ST as a working example of transportation investment optimization.
- Include model components for the evaluation of performance measures to address MAP-21 and asset management:
- Identify and prioritize model design features for each performance metric desired based on agency needs.
- Apply economic assessment software to model output to assess economic impacts of transportation features.
- Develop post processing methodology to determine economic impact/GSP value of individual links.
- Consider use of a separate project-specific benefit/cost model.
- Implement the determined freight model component based on either commodity flows or truck/rail vehicles
7.2.3 Recommended Longer-Term Improvements
The panel also identified potential improvements for VTrans to consider over the longer term (beyond the next three years):
- Apply the model to incorporate the assessment of fair-share methodologies:
- Develop VMT estimates for new development by land use type and trip purpose to determine change over time and assess impact fees.
- Recognize that statewide model resolution is not adequate for a post processing methodology to determine long-range growth rates for background traffic.
- Consider a micro-simulation model, which applies future volumes and growth rates from the regional model.
- Review off-model techniques that can be used as separate/compatible tool for development impact assessment, such as the ITE Trip Generation Manual.
- Develop methodologies to assess transit and non-motorized for corridor prioritization:
- Recognize that statewide model may not be appropriate resolution for evaluating non-motorized transportation improvements.
- Develop separate/compatible tool for non-motorized transportation.
- Consider micro-simulation models for local area analysis.
- Consider survey efforts to understand current travel by mode.
- Consider a tiered approach to activity-based model development for non-motorized travel as a long-term priority if the agency envisions the statewide model as the preferred tool for non-motorized transportation assessment.
- Determine the best methodology for assessing energy use and emissions:
- Include a mode choice model component.
- Use MOVES in conjunction with model output once the model includes a mode choice component to estimate emissions.
- Identify and test sensitivities in energy/emission performance measures.
- Recognize the difficulty in addressing performance measures given the scale and resolution of the statewide model.
- Consider a separate aggregate model to apply data from both the statewide model and the MPO model to evaluate energy and emissions data.
- Consider scenario testing in the long-term.