The existing travel demand model that DVRPC maintains is an aggregate, trip-based model. It relies on aggregate traffic analysis zones that divide the region into small areas that form the basis for the model. Overall, the model includes 3,400 traffic analysis zones.
Like many trip-based models, DVRPC segments trips into a number of different categories. Total trips are split into person, vehicle, and external trips, and person trips are further segmented by travel purpose. Table 3‑1 shows the trip purposes used by the existing DVRPC model.
As is the case in many trip-based models, the majority of the demand model components are dedicated to modeling person trips. Households are further stratified by certain household characteristics (including household size, number of workers, and household income) for trip generation, but only the income stratification is carried forward through distribution and mode-split components of the model. Figure 3‑1 shows the stratification of trips through model components.
Figure 3‑1: Household Stratification
(Source: DVRPC internal memorandum documenting current travel demand model,
Chapter 7, "Demand Model." Provided to Peer Review Panel prior to meeting.)
Figure 3-1 depicts the household stratification used by the existing DVRPC trip-based model across each of the model's components and trip purposes. For trip generation model, home-based work, home-based shopping, and home-based other trip purposes are segmented by household income and number of household members or workers. For other model components (non-motorized split, trip distribution, and mode choice), these trip purposes are segmented only by high/medium and low income groups. Other trip purposes (home-based school, home-based university, non-home-based work, and non-home-based other) are not segmented across household attributes in any way. For the trip assignment model component, all trips are combined.
The trip distribution models are gravity-type models. Prior to trip distribution, nonmotorized trips are removed; thus, the trip distribution model is relevant only to motorized trips. The mode choice model splits motorized trips into auto, walk-transit, and drive-transit trips. Temporal factors are used to disaggregate trips by time of day. Four time-of-day periods are considered: morning (6 a.m. to 10 a.m.), midday (10 a.m. to 3 p.m.), evening (3 p.m. to 7 p.m.), and night (7 p.m. to 6 a.m.).
Due to some of the limitations of the existing model, DVRPC currently is developing an activity-based model to replace their existing trip-based model. The activity-based model development already is underway. Part of the peer review process was intended to help direct the development of the activity-based model with consideration to the existing model platform and processes already in place at DVRPC.
DVRPC hosted a peer review of their modeling practices in September 2009 (report available at http://www.fhwa.dot.gov/planning/tmip/resources/peer_review_program/dvrpc/). The primary purpose of this peer review was twofold. First, the peer review was charged with giving opinions on prioritizing short- and long-term improvement plans. Second, the peer review panel provided an evaluation of the modeling system used by DVRPC and compared it to prevailing industry standards.
At the conclusion of the peer review panel meeting, the panel offered a number of formal recommendations to DVRPC as outlined below.
DVRPC should:
DVRPC should:
Prior to meeting, DVRPC identified several areas for which they wanted the peer review panel to comment and make recommendations. These items were presented as a series of questions and were a focus during the meeting. The questions are detailed below:
