Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
ADDRESSING POLICY ANALYSIS NEEDS The following discussion summarizes some specifics of how the proposed modeling approach would address some of the specific policy analysis needs described above. Pricing Analysis The traffic forecasting procedures for toll facilities and managed lanes have been a topic of considerable discus- sion recently. Various aspects of existing procedures have been criticized, including the assumed values of time for various market segments of travelers, the aggregate nature of the process (which requires fixed values of time for each segment), the difficulty in modeling time of day outside a tour- based approach, and the static nature of the traffic assignment process, which ignores the effects of the buildup and dissipation of queues. Activity- based approaches present some advantages over conventional modeling procedures in addressing some of these issues. One major advantage is that mod- eling individuals in the synthetic population provides an opportunity to use distributed values of time rather than fixed values for a relatively small number of market seg- ments. For example, say that it would take a value of time of $12/h for a certain geographic market to find using a particular toll road segment desirable. If the aver- age value of time for the market segment were $10/h, then the model would estimate that no one from that segment would use the toll road. However, if a value of time distribution were used with an average value of $10/h but with a 20% probability of having a value of time of greater than $12/h, there would be demand esti- mated for the toll road within this market segment. Another major advantage is that demand for road- ways where tolls vary by time of day can be modeled much more accurately. Time- of- day decisions for activi- ties must consider not only the time when the trip to or from the activity takes place, but also the trip in the other direction and the duration of the activity itself. For example, if someone wishes to consider shifting his departure time for a work trip to avoid a high- peak- period toll, he or she would likely also need to consider the amount of time needed to be spent at work and whether the time shift for the trip to work might shift the departure time from work to or from a peak period with a high toll. Obviously a model that treats individual trips independently cannot include such considerations. Urban Centers and Transit- Oriented Development There are several advantages to modeling travel by resi- dents, workers, and visitors in these types of develop- ments using the proposed activity- based modeling approach. First, many variables in an aggregate, trip- based model must be introduced through the use of seg- mentation, which significantly limits the number of variables that can be included in the model. Adding fur- ther segmentation to a typical cross- classification trip production model (likely with only two or three dimen- sions) to account for different trip- making characteris- tics in denser, transit- oriented areas would require the household survey data to be segmented by additional dimensions, often beyond the ability to obtain statisti- cally significant estimates of trip rates given the limita- tions of the existing sample. The activity- based modeling approach, where individual daily activity patterns are simulated, permits description of individuals using a much richer set of variables. Planning judgment and travel behavior data also sup- port the expectation that having a variety of attractions located in close proximity in the urban centers, including workplaces, other businesses, and shopping and enter- tainment opportunities, would have an effect on trip chaining, as individuals might choose to combine activi- ties that can be accomplished in the same vicinity. Obvi- ously, a tour- based approach is required to capture the effects of trip chaining. Finally, data also suggest that persons living or work- ing in higher- density transit- oriented areas should have greater opportunities to use transit and nonmotorized modes. However, properly reflecting these opportunities in the model requires a combination of capabilities: mod- eling travel in tours so that, for example, secondary tour trips, stops, and modes can be shown to be compatible with transit as the primary mode of the tour (as they will sometimes be within walking distance); destination choice models that can operate at sufficient geographic detail so as to locate some secondary stops within the transit- oriented development; and fine geographic detail on stop locations so that walk distances can be accu- rately calculated (so that walk choices in the mode choice models are accurately estimated). Transportation Project Analysis The use of disaggregate microsimulation of individuals provides some advantages to the analysis of new trans- portation projects, particularly the extensive transit investments planned for the Denver area. One of the key questions involved in the analysis of transit investments involves the identification of how specific groups of the population (for example, persons from low- income households) benefit from the investments. In conven- tional models, demographic market segmentation is not carried through beyond the mode choice step, so some model results cannot be differentiated by market seg- 179USING ACTIVITY- BASED MODELS FOR POLICY DECISION MAKING