During preparation of this report, panelists were given an opportunity to comment on an early draft. At that time, several additional comments were made by panel members. Those comments are documented in this section. Many of the comments were related to the plan for updating the model, as outlined in Section 3.2 of this report.
Data Development
As documented in Section 3.2, VDOT plans on developing the TAZ structure to be consistent with TAZ structures of the state's regional models as well as Census boundaries. One panel member emphasized that the TAZ boundaries should nest or coincide with geographic boundaries in the Census, the previous statewide model, and regional models. Section 3.2 also highlights VDOT's plan to use the HERE network for developing the highway network and Freight Analysis Framework (FAF) network for out-of-state network development. In contradiction to this plan, one of the panelists suggested that the HERE network would be suitable for the out-of-state network only, but something more granular may be needed for in-state network development. Regarding socioeconomic data needs, it was unclear to one panel member what economic variables would be used, but suggested that the socioeconomic data include income. In addition, the panel member encouraged that spatial variables be used to distinguish travel patterns differences between rural and urban areas.
Passenger Model Development
Section 3.2 details that VDOT plans on using the long-distance component of the American Travel Survey (ATS) from 1995. One panelist remarked that this data was quite old, predating the relatively quick rise in travel costs over the past two decades and changes to the nation's economy during that time. A better (more recent) data source would be desirable in that sense. Another panel member remarked on the different time-of-day periods used by Virginia's key MPO models (as shown in Table 3-2), noting that the Richmond MPO and Fredericksburg MPO both use periods, in part, defined on half-hours (e.g., the ending time of the PM peak period in Richmond is 6:30 pm). The panelist worried that VDOT would need half-hour counts in those regions or else it would be difficult to calibrate and validate the model.
Model Validation and Calibration
One panelist encouraged that VDOT review travel times and speeds from the model during the validation process, comparing to observed travel times and speeds from other data sources.
VDOT's Questions to the Panelists
Even though the panel members had a chance to respond to each of VDOT's questions during the meeting, one panelist left something out of the original response. In response to what the top challenges of statewide modeling are, one panelist commented that model results are often viewed in a vacuum, but it is important not to attempt to extract too much from model results or overstep the level of precision of the model. Uncertainty exists in the model.
