During the peer review ADOT staff and a consultant team presented an overview of the current Arizona statewide model, AZTDM2, and of the development plans and specifications for AZTDM3. (An overview of AZTDM2 taken from the model documentation provided by ADOT is found in Appendix D.) While discussion occurred around aspects of both AZTDM2 and AZTDM3, ADOT requested the panel to provide in this review insight and guidance related to specific topics of interest for AZTDM3 and AZTDM4 development. These topics helped frame the panel discussion and the panel made recommendations specific to these items. Specifically, ADOT wanted the peer panel to comment on:
One proposed application for the statewide model is to test interregional transit corridors, such as might be served by a commuter rail system. These transit corridors are between MPO areas and not fully represented by the MPO models. The MPO models do represent other transit modes, such as local bus, express bus and light rail, and which will also be represented in the statewide model.
At a statewide level, the coding of the various transit components could be a large undertaking. For example, the combined transit systems in the statewide model contain 214 unique routes (337 directional routes) in the base year. Coding this many routes, along with their supporting link structure and park-and-ride access, can be significant, especially when considering future scenarios.
To simplify the bus network coding, AZTDM3 proposes to employ a method where the local bus networks are approximated or abstracted to compute in and out of vehicle travel times and measures of accessibility by bus to access rail and intercity bus services. This method is similar to the method applied in the California statewide travel model. The abstraction method will be applied to the local bus mode only. Rail and intercity bus will still be explicitly coded in the transit network.
Local bus travel times will be based on land use and highway network variables as opposed to a local transit network. Travel times will include sensitivities to time of day and geographic accessibility to transit (catchment area) information. The abstraction model uses:
ADOT proposes to code local bus transit networks simultaneously with the local bus abstraction method in order to validate the abstraction method. The base year local bus network will be available to run on projects as needed, or the model could use the abstracted local bus network. The abstraction method of the local bus network will be the primary method used in future long range planning.
Statewide Cordon Count Data
ADOT anticipates conducting a statewide cordon survey in spring 2012. The cordon survey is designed to capture the external and through trip movements into and out of Arizona. The survey includes counts, vehicle classification, trip purpose, trip frequency and the origins and destinations of trips.
ADOT proposes to do a license plate capture survey followed by a mail-out questionnaire. The questionnaire would be used to collect the trip origin-destination, trip purpose and the frequency of the trip. ADOT is also exploring use of other technology, such as bluetooth tracking, for routing and delay data.
ADOT was interested in the panel's experience related to conducting cordon surveys, such as improving response rates, example questionnaires, and working with neighboring states. As of yet, Arizona has not pushed to do an intercept survey and it was uncertain if intercept surveys have been allowed. ADOT was also researching the use of different technologies, such as GPS or cellular.
Rural Household Travel Survey
The AZTDM2's personal travel demand model was calibrated using the 2009 National Household Travel Survey (NHTS) data as the primary source of travel behavior information. The NHTS survey data were primarily focused in the urban areas. Of the 7,157 total household surveys collected, only 89 were collected outside the MAG/PAG MPO regions.
ADOT would like to conduct a household survey for the rural areas of the state to better understand rural trip making patterns. The survey would replicate the NHTS as close as possible. ADOT is considering a sample size of around 1,000 households. ADOT requested the panel's reaction to this proposal.
The current AZTDM2 forecasts short haul and long haul truck volumes based on FHWA's Freight Analysis Framework version 3 (FAF3). ADOT would like to use a commodity-based framework using the Global Insight Transearch database that would allow ADOT to consider goods movements between different modes, such as truck and heavy rail. The model would be applied for planning and policy testing along the existing freight infrastructure as well as to test potential new freight corridors, such as the I-11 CANMEX corridor.
ADOT has purchased the Global Insight Transearch database that has commodity flows in Arizona at a TAZ level and throughout the country at a more aggregate geographic level. ADOT plans to compare the truck trips and volumes from the commodity-based freight model against the trips from the FAF3-based model before deciding on which is the better approach to use and would like the panel to provide comment on the proposed freight model method.
ADOT would like to create a hybrid version of the statewide model that could be applied to metro areas or sub-regions of the statewide model. The hybrid model would use a more refined zone structure, highway network and socioeconomic database than currently exist in the statewide model. The hybrid model would allow for more easy sub-area extraction from the statewide model which could be used for more localized or focused planning.
Central Arizona Association of Governments (CAAG) would like to see ADOT use the statewide model or hybrid statewide model in the development of their long range transportation plan. Pinal County in the CAAG area is expected to see a four-fold increase in population by 2055 and Pinal County was the only county in Arizona that exceeded its growth projection in the 2010 Census. CAAG is looking to ADOT to help model the future growth in the development, which ADOT is hoping to do in the spring of 2012.
ADOT would like to get from the panel insights on best approaches to sub-area model extractions and subsequent sub-area model usage.
ADOT is considering applying several advanced modeling techniques as part of the statewide model improvement program. These include dynamic traffic assignment, activity-based modeling, integrated transportation and land-use modeling and economic modeling. ADOT cited several examples of advanced model research projects through FHWA and Transportation Research Board (TRB). ADOT is requesting the peer review panel's guidance on the viability and best practice of implementing advanced modeling techniques at the statewide level.
