Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 27
TOUR-BASED MODELS 15
miles, with a population of approximately 20 million. the calibration process is to properly identify the source
There are approximately 4,000 traffic assignment zones, of any discrepancy. It is important to identify the prob-
six trip purposes, and 10 motorized modes in the model, lem first before rushing to make adjustments in the
as well as four types of urban development. The four model. Another suggestion is not to overadjust. A good
urban development types range from the very dense cen- approach is to begin with the largest discrepancies.
tral business district (CBD) to rural areas. · Reporting and analysis are also important ele-
· The NYMTC model has four major consecutive ments. Most staff who operate the models are not
modules. The first module is tour generation that includes involved in all aspects of the planning process. Reporting
household synthesis, automobile ownership, and journey the results represents the link between the modelers,
frequency choice models. The second module is tour planners, and policy makers. Take care to distinguish
mode and destination choice that includes premode between reporting options per se and reliability of the
choice, primary destination choice, entire tour mode com- forecasts at a fine level of typological, spatial, or tempo-
bination choice, stop-frequency choice, and stop-location ral detail. Activity-based models are more exposed than
choice. The third module is the time-of-day choice and aggregate models because more detailed reports may be
preassignment processor that includes tour time-of-day generated from these models.
choice for outbound and inbound directions, trip mode
choice, and construction of mode-specific and time-of-
day period-specific trip tables. The fourth module is traf- THE SAN FRANCISCO MODEL IN PRACTICE:
fic and transit simulation that is implemented by VALIDATION, TESTING, AND APPLICATION
time-of-day periods. The first three modules are imple-
mented as fully disaggregate microsimulation procedures Maren Outwater and Billy Charlton
working with individual records for the synthesized pop-
ulation, which includes households, persons, and tours. Maren Outwater and Billy Charlton described the valida-
The fourth module is based on standard aggregate zone- tion, testing, and application of the San Francisco County
to-zone assignments implemented in TransCAD. Chained Activity Modeling Process (SF-CHAMP), which
· A tour-based model was used because it was the was developed for the San Francisco County Transporta-
best approach to addressing complex mode and time-of- tion Authority (SFCTA). They discussed the development
day choice in a consistent manner. The model also pro- of the model, the model validation process, applications
vides the best method to understand and forecast the of the model, and comparisons of the model to the four-
highly diverse demographic and travel patterns in the step model. Volume 2 includes a paper on this topic.3 The
region. following points were covered in their presentation.
· There are numerous stakeholders in the New York
region. NYMTC staff undertook an outreach and edu- · SF-CHAMP was developed in 2000 and 2001. The
cation program to explain the need for the new model model was developed to provide detailed forecasts of
and to build support among different groups. These travel demand for various planning applications in the
efforts included technical outreach to staff at other agen- county. These applications include countywide plans,
cies, as well as meetings with policy makers. corridor and project-level evaluations, transit plans, and
· The model has been used in a number of applica- neighborhood plans. The objective was to accurately
tions, including air quality conformity analysis, major represent the complexity of the destination, temporal,
investment studies, and local planning activities. Exam- and modal options, as well as to provide detailed infor-
ples include the Tappan Zee Bridge study, the Goethals mation on travelers making discrete choices. A tour-
Bridge study, and the Manhattan area pricing study. The based model using synthesized population as the basis of
experience to date with the use of the model has been decision making rather than zonal-level aggregate data
positive. sources best met this objective.
· Work is under way to make the model accessible · The development of the model was influenced by
on the Internet. A new wave of data collection will also limited resources and time availability. These limitations
be starting. This data collection will include a major were considered in the development process. First, no
household travel survey, a work place survey, and obtain- transit onboard survey data were available to validate
ing information on visitor trips. The results of the data the mode choice elements of the model. This limitation
collection efforts, which are anticipated to take approxi- would have been an issue for a trip- or a tour-based
mately 3 years, will be used to update the model. model. New onboard survey data have recently been col-
· It is important that models are calibrated to ensure lected and are being used to update the model. The peak-
that they replicate reality. A nontrivial and often not
explored question relates to validating and calibrating a 3See Outwater, M. L., and B. Charlton. The San Francisco Model in
model chain rather than a single model. A key aspect in Practice: Validation, Testing, and Application. Volume 2, pp. 2429.
