Flagstaff Metropolitan Planning Organization (FMPO) Peer Review
5.0 Peer Review Panel Recommendations
This section summarizes the recommendations of the panel.
There were several aspects of the FMPO modeling and planning process that the panel approved of; however, the panel concluded that the model required extensive changes to be brought up to the state of the practice and to be useful in the TIA process. The panel recommendations were made with a particular mind to achieving the following goals expressed by FMPO during the peer review meeting:
- Model that employs best practice approaches;
- A model that can be used to support the TIA process;
- Consideration for non-auto modes (transit, bike, pedestrian); and
- Solid forecast capabilities.
The following sections summarize the panel recommendations from the previous section by priority. Priority was determined according to the relative effort and benefit of each item. In addition to the recommendations listed below, the panel recommended that FMPO review published guidelines, such as NCHRP Report 765: Analytical Travel Forecasting Approaches for Project-Level Planning and Design, and peer models like Parkersburg, WV-OH for examples of solid, four-step models that are feasible for FMPO to implement and maintain and are appropriate to support project-level travel forecasting. The panel emphasized that regional models are not appropriate for direct project-level forecast use and that operational model should be used instead. Moreover, in order for regional model outputs to be used as reference, the model must be well validated and the outputs used used with great caution.
The high-priority model suggestions include fairly extensive changes throughout all components of the model. Given that FMPO has only one full-time modeler on staff, the panel assumed that consulting resources would be used to complete the recommendations. The estimated cost of such a model update would be about $80-120K and take about one year.
- Produce trips using cross-classified person trip generation rates based on socio-economic data.
- Implement a truck model based on the Quick Response Freight Manual.
- Replace the 24-hour model with a series of shorter traffic assignment procedures that include one-hour assignment periods during congested times.
- Implement a more traditional logit-based mode choice model using transferred/asserted parameters.
- Implement an assignment process that utilizes an intersection delay function.
- Create a transit network for skimming and assignment.
- Implement a speed-feedback loop whereby the congested travel times and distances are used to redistribute trips.
- Calibrate the model to represent the fall travel season.
- Collect counts on a monthly basis at 10-12 locations to confirm fall travel season peak assumption and gain more insight into seasonal traffic variations.
- Improve the validation process, specifically:
- Calibrate trip distribution using measures that examine the trip length frequency curve (e.g. coincidence ratio);
- Examine flows between key sub-areas of the network;
- Utilize HERE and/or INRIX travel time data; and
- Follow the example of states such as Ohio, Michigan and Florida for guidance on measures and reports.
- Develop forecast years with 10 year intervals through 2040.
The medium priority improvements suggested are those that the panel would like to see implemented, but are not as critical for FMPO's needs.
- Home-Based School Trip Purpose: Identify K-12 school locations and enrollment and add a home-based school purpose to the model, especially considering that some charter schools use the public transit system as the school bus.
- Trip Attractions: Convert the current approach from using ITE trip rates to a more traditional employment based attraction rates by employment type. Alternatively, the ITE employment based trip rates may be used, which is an improvement over the land use square footage rates.
- University Trip Special Generator: FMPO should collect or borrow university special generator data to better capture the large population of university students in the region.
- Update networks and inputs on an annual basis.
- Develop forecast years with 5-year intervals.
Low priority model improvements are included as things that FMPO should keep in mind through future model development.
- External trips should continue to be derived from the statewide model and a cordon survey taken to validate the statewide model and improve the trip purpose splits for external travel.
- Develop a summer travel season model that includes a hotel-based trip purpose.
The model recommendations above were made with the intention of making the model outputs more suited to use in a TIA process. This section summarizes the post-processing and ways to utilize the model outputs in a TIA.
- Follow the method defined in NCHRP-765 for intersection and link volume post-processing.
- Use the model to inform the radius of impact determination and identify significantly impacted intersections.
- Identify model volumes using select link analysis to calculate proportional share from new developments.
- Observe mode shares by zone to support non-auto analyses and examine travel through areas of the network with identified deficiencies.
