ADOT has requested this second TMIP peer review to assist ADOT staff in identifying the best practices in statewide travel modeling for implementation in the AZTDM. The goal of the review is to enhance the methods employed in the AZTDM and to improve its utility for planning analyses at state, regional, and municipal levels.
The AZTDM is expected to have a role in a variety of transportation planning studies and projects, including:
The TMIP peer review summary report will provide useful documentation of the discussion and guidance from the peer review panel of their examination of the AZTDM. To this end, ADOT has prepared a list of specific topics for which they sought the panel's comments and recommendations. The list of topics and the peer review panel's response is presented in Section 4.0 of this report.
ADOT, along with its partner agencies, will assess the feedback from the peers when prioritizing its model development plan. While the advice of the peers is invaluable, there are many factors to work through when considering a model improvement strategy. The panel's recommendations will be regarded as recommendations only and not mandates for ADOT and its partnering agencies.
The AZTDM currently employs state-of-the-practice techniques for estimating and forecasting statewide travel demand. Over time, ADOT plans to incorporate advanced modeling techniques into the AZTDM, as shown in Figure 1.
Figure 1: AZTDM Phased Model Improvement Plan.
The first generation model (AZTDM1), delivered in 2009, was developed to provide estimates of general statewide performance of alternative system improvement strategies during the visioning activities of the Building a Quality Arizona (bqAZ) framework study. However, the basic structure of the first generation model was aggregate in nature, with a highway network containing only the highest-level roadways. AZTDM1 was determined to have limited suitability for both transportation planning applications and many of the policy questions anticipated in current statewide transportation system planning and development activities.
The second generation model (AZTDM2), initiated in March 2010 and completed in May 2011, focused on implementing personal and freight travel demand model components. The model incorporated considerable additional detail in the transportation analysis zone (TAZ) systems and the transportation networks of the metropolitan planning organizations (MPO) in the state. The personal travel demand model component was calibrated to Arizona travel behaviors by using the 2009 National Household Travel Survey (NHTS) as the primary source of travel behavior information. In addition to the base national sample collected in Arizona through the 2009 NHTS (approximately 500 samples), there were NHTS Add-Ons purchased by the Maricopa Association of Governments (MAG) (4,286 samples) and Pima Association of Governments (PAG) (2,285 samples), which formed the estimation/calibration database totaling approximately 7,000 sample households. The freight and goods movement component of the AZTDM2 was based on available data from public sources, including the FHWA Freight Analysis Framework (FAF3) and the Bureau of Transportation Statistics (BTS) databases.
The third generation model (AZTDM3), initiated in July 2011, seeks to transition the AZTDM to a 4-step model. This has involved incorporating a personal mode choice model component. The abstraction of local transit services during this enhancement provided a realistic representation of local transit while reducing the effort required to maintain and update the statewide model. However, the explicit coding of non-local services (i.e., fixed-guide-way and other line-haul transit services) was retained. The AZTDM3 has capabilities of providing estimates of transit use for major system and service improvements for multimodal planning studies. This transition to a 4-step travel demand model is intended to include feedback interactions with other AZTDM3 components (e.g., vehicle availability, trip generation, trip distribution, and network assignment) and maintain transit sensitivity to highway network congestion.
The AZTDM2 and the implementation plan for the AZTDM3 were topics of discussion during the first TMIP peer review, which was conducted on November 18, 2011. Section 3.0 of this report includes a summary of the first peer review panel recommendations and ADOT's response.
In this second peer review, ADOT requested assistance from the peer panel to examine aspects of the AZTDM3 and potential advanced modeling techniques for the fourth-generation AZTDM. ADOT's questions to the panel and the panel's responses are included in Section 4.0 of this report. Appendix D of this report contains an overview of the AZTDM3 taken from the model documentation provided by ADOT.
