Association of Monterey Bay Area Governments (AMBAG) Travel Model Peer Review Report

Appendix D Summary of Responses to AMBAG Questions

Summary of Responses to AMBAG Questions

The following questions were posed by AMBAG and other regional/local agency staff for discussion during the peer review. Most of the issues raised by the questions were covered during the discussion and recommendations phases of the peer review. Brief summaries of those discussions and recommendations, and selected additional responses are provided in this Appendix.

1. Does current AMBAG model need additional improvements to meet the FHWA/FTA as well as State standards?
   1. The current AMBAG model meets current FHWA standards, but additional improvements, as recommended by the panel, are needed to meet FTA and new State standards.
2. Are additional improvements needed to the current feedback loop procedures?
   1. The panel recommended the MPO test and report system convergence of the distribution, mode choice, assignment feedback loop. The panel also recommended the MPO consider using more than two iterations to obtain congested travel times.
3. Are there better techniques for handling trip generation and trip distribution for the size and diversity of our region?
   1. As a mid-term solution, a daily activity model could be used instead of trip generation model and a primary destination choice model could be used instead of trip distribution model to address questions related to the '5D' process and issues related to trip length frequencies. A full activity-based model is recommended as a longer term solution. In the short term, the panel suggested reinvestigating the strategic use of limited K-factors to improve trip distribution and revising trip generation rates for households rather than persons.
4. Are there checks/balances that we are not performing that should be performed? Is the current trip distribution model (gamma function) is acceptable or use the friction factors tables?
   1. It was suggested that the MPO check: 1) trip generation production and attraction values by purpose before and after adjustments, 2) distribution feedback loop convergence, 3) trip length distributions by county, 4) mode choice constants in terms of equivalent in vehicle travel time, 5) traffic assignment convergence, and 6) screenline validation in particular for Santa Clara County.
   2. The use of gamma functions for trip distribution is acceptable when applying a gravity model, but the panel recommended moving to a destination choice model due to the distribution of households and employment in the region.
5. Considering the region's diversity, how can our mode choice model be improved?
   1. Consideration should be given to supplement 2011-2012 CHTS and oversample special populations and get a good demographic description of these households. The focus of this should be to improve the future mode choice and destination choice models realizing that mode choice will also need to be supported by a transit on-board survey. The MPO should work with transit agencies to conduct a comprehensive transit on-board survey geared toward travel demand modeling. Consideration should also be given to broaden the explanatory variables in the mode choice model. Estimating a new mode choice model with socioeconomic and demographics that incorporate urban design and level of service characteristics will also provide support for '5D' analyses.
6. How can our model keep up with time-of-day requirements since the region has strong influence of SF Bay/San Jose area as well as longer commute distance/time?
   1. The panel recommended moving to five time periods: AM, midday, PM, evening and night. The model should also be calibrated by period using congested speeds and diurnal distribution data for each time period. Diurnal data could be derived from CHTS survey.
7. How can we better reflect the commute pattern between SF Bay Area and AMBAG region?
   1. Implementing an activity-based model or a hybrid activity-based/destination-choice model will help to capture longer commuting patterns, such as for the highest income group (>$75,000). The California statewide is a good resource to help identify this external trip movement. In the short term, validating the trip distribution patterns by county should also help.
8. Assessment of the reasonableness of HOV/toll traffic forecasting capabilities.
   1. According to the model documentation, the AMBAG travel demand model has HOV forecasting capabilities; however the reasonableness of the model's HOV forecasting was not discussed separately from the context of the overall mode choice model capabilities.The panel did recommend some improvements to the mode choice model, which will apply to HOV modes. Tolling capabilities do not currently exist in the AMBAG travel demand model. The panel asked if tolling was a top question for AMBAG or its partnering agencies. Tolling options are not considered a priority in the current planning, therefore it was recommended that resources be directed from developing tolling capabilities into the model now toward more pressing needs.
9. What are ways that the AMBAG's regional model can be improved to better answer questions related to Smart Growth /mixed land use development and resultant Greenhouse Gases (GHG) reduction?
   1. More detailed land use variables and model improvements such as those in questions 3 and 5 that make the travel demand model more sensitive to changes to land use will help in answering these questions. The panel's recommendation for a land use forecasting model and integration with the travel model will provide the most benefit here.
10. How can the effects of gas prices or parking cost be better implemented into the agency's model stream?
    1. This question was not specifically discussed by the peer review panel; however guidance was given on using generalized costs in trip distribution and assignment models to reflect prices in this model and to review multi-year trends to establish calibration and validation data.
11. Do peak spreading and induced travel need to be emphasized more? If so, how?
    1. This question was not discussed directly by the peer review panel, but the mid-term recommendation to develop a daily activity pattern model will produce estimates of induced travel and the long-term recommendation to develop a full activity-based model would include a time of day choice model to address peak spreading.
12. What recommendations are there for improving the current demographic allocation tool or for pursuing a new land-use forecasting model for alternative growth scenario testing as well as its integration to the future generation of the travel demand forecasting model?
    1. The peer review panel recommended developing a mid-level land use forecasting model that recognizes seasonal residents and workers and that can take advantage of parcel data. The panel offered that there is a wide spectrum of land use models, ranging from more simple models like UPLAN to more complex models like UrbanSim and PECAS. It was suggested that several mid-range land use models, such as DELTA and Cube Land, require considerably less investment to build and maintain (approximately a fifth the cost) than the more complex land use models. These mid-range models can be designed to be TAZ-based but have a parcel data option and are less data hungry than the more complex models. The panel recommended investing in the more mid-range complexity land use model, especially in light of the resource constraints of the MPO. The panel also recommended that a new land use model did not necessarily need to be fully integrated or automated with the travel model to include a transportation element.
    2. The panel recommended that AMBAG define a process for developing parcel data that could be used to support land use models through a regional data consortium. Local communities could feel a benefit from sharing data and working together by receiving information for smart growth concepts such as the '5Ds' analyses to minimize carbon footprint and VMT. This data sharing could enrich the quality of the databases and provide a cost sharing benefit.
13. What processes are we using that might not be best practices in travel demand modeling?
    1. Refer to the 'Peer Review Panel Recommendations' section of this report.
14. Are there improvements that could be implemented immediately with limited funds or implemented in a two to three year period for an overall model development?
    1. Improvements are listed in the short and mid-term recommendations of the 'Peer Review Panel Recommendations' section of this report.
15. How can we model weekend tourist/visitors activities?
    1. Answers to questions 3, 5 and 12 also apply to tourist/visitor activities. In addition, the peer review panel suggested the MPO continue to get regular updates of visitor data from tourist industry. Weekend travel is not currently modeled by other MPOs or the State and if needed, can be estimated from available data sources to pivot from current weekday models.
16. How can we model farm workers travel behavior since they are transient (field to field) and seasonal?
    1. Answers to questions 3, 5 and 12 also apply to farm worker travel behavior. Farm workers could be separated as a special generator to capture the unique aspects of this population.
17. Where should we obtain base year employment data? What do other MPOs use?
    1. This question was not discussed by the peer review panel.
18. How much and at what level field data collection is necessary to build the best practice robust Regional Model?
19. The panel recommended supporting a household travel supplement to the California Statewide survey and coordinating a regional on-board transit survey. Additional surveys for external travel or freight were not discussed. Speed data collection was useful, but the use of these data in the model were revised to include free flow speeds as an input and congested speeds as a validation measure.How can we model regional diversities (socioeconomic as well as transit services)?
    1. See answers to questions 3, 5,12, 15 and 16.
20. Should we be continuing building a regional travel demand models for all three counties and also include the neighboring region? Or also develop sub-area model?
    1. This question was not discussed by the peer review panel.
21. Considering the SB375 requirements as well as other agency needs AMBAG has designed the Model Improvement Plan (MIP). What would be staffing and funding needs to meet these requirements or does the MPO only develop a basic regional model to meet FHWA/FTA's basic requirements and enhancements are expected to built by end users?
    1. AMBAG needs to leverage partnerships realizing that all good information and analysis comes at a price. AMBAG already has budding relationships in place through the blueprint planning process and other planning efforts. The panel recommended that AMBAG continue to foster these relationships. The panel also suggested that the MPO could leverage existing grants, such as the grant to develop a bike model, to improve facets of the model. The panel emphasized continuing to explore additional grant funding sources. Even with these efforts, the panel recommended that AMBAG needs to increase funding for model improvements.
22. Any organizational changes that can facilitate our model development process more seamlessly?
    1. See answer to question 21.
23. Should we start developing a freight model? If so, how?
    1. AMBAG should replace the current truck model with goods movement model. The approach was not discussed in detail as the freight forecasting methods are changing rapidly and should be considered at the time of model development. Integration with the future statewide freight model may be beneficial.
24. Should we consider transitioning into an activity-based or tour-based model? If so, should we maintain parallel tracks of modeling? Also associated cost?
    1. The panel suggested the MPO consider moving toward an activity-based model. Given the cost and resource constraints, it was suggested that AMBAG could do a phased approach to activity-based model implantation. A hybrid activity trip based model could use daily activity patterns and primary destination choice and auto ownership models and could leverage the statewide model population synthesizer. This was a mid-term recommendation.
    2. It was recommended that the MPO implement a full activity-based model in the long-term. In developing a full activity-based model the MPO has the option of transferring an activity-based model from other locations and it was suggested that transferring from a smaller activity based model could be accomplished in about a year for about a third the cost of a full independent development. Second generation activity-based models will also come with a reduced cost as the first generation models had a steeper learning curve and have established a knowledge and user base from which to build. If AMBAG switched to an activity-based model, they would be in the second generation.
    3. Maintaining parallel tracks of modeling trip and activity-based models was not recommended due to resource constraints and because second generation activity models would not need to maintain parallel tracks.
25. Should we consider a micro simulation model and how can our current model assist with that implementation?
    1. Traffic micro simulation and dynamic traffic assignment at a regional scale was discussed by the peer review panel but placed at a lower level of priority than the other model improvements due to the needs expressed by the agencies using the model.
26. Is there any visualization techniques (sketch planning tool) available for better communications to the board and general public (education materials, flyer for "dummies") pertaining to the model methodology and results?
    1. Incorporate land use planning visualization tools to support continuing blue print planning efforts and SB 375.