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.
research project on time- of- day models that included a case study of a new time- of- day model (including peak spreading) for SF- CHAMP. Plans call for this new time- of- day model to be incorporated into the model. ⢠The approach to trip assignment included a tradi- tional aggregate assignment because there were too few resources in the project to implement a microsimulation assignment methodology. This approach has been used in all other tour- based model applications in the United States to date (except Transims). Nonetheless, it intro- duces aggregation bias and fails to take advantage of the disaggregate information on each traveler during route choice. ⢠SF- CHAMP combines trip tables from the MTC regional trip- based model with trips generated from the San Francisco tour- based model. As a result, only San Francisco residents are represented by the tour- based model and its advantages. These limitations were known at the outset and accepted as lesser priorities than the core objective of building a tour- based model. In some cases, these limita- tions are already undergoing change in the update of the SF- CHAMP model. There was one additional innovative aspect of the mode choice model: the inclusion of reliability and crowding as explicit variables in the transit utility func- tions; this aspect was tested and then not included in the final models. These variables were included in a stated- preference telephone survey of 407 transit users in San Francisco. Logit analysis was used to estimate trade-offs between in- vehicle time, frequency of service, reliability (defined as the percentage of days that the vehicle arrives five or more minutes late), and crowding (low plenty of seats available; medium few seats available, but plenty of room to stand; high no seats available and standing room is crowded). It was estimated that improving the percentage of vehicles arriving on time by 10% (e.g., once every 2 weeks) is equivalent to reducing the typical wait time (half the headway) by 4 min for commuters or 3 min for noncommuters. It was also esti- mated that improving the level of crowding from high to low is equivalent to reducing the typical wait time by 5 min for commuters and 9 min for noncommuters. Thus, relative to commuters, noncommuters are, on average, less sensitive to delay but more sensitive to crowding. In application, the reliability and crowding was coded in the transit network by means of observed system data collected by SFCTA. The trade-offs esti- mated between these variables and wait time were applied in performing transit assignment and found not to be coincident with the observed boardings. As a result, these variables were not used in model application. MODEL VALIDATION Travel behavior was validated by comparing travel data in a household travel survey to related travel data in the travel demand forecasting model. For the validation of the current 1998 SFCTA regional travel demand fore- casting model, the trip data in the 1990 Census and the 1990 MTC household survey data were compared with the same data in the model (2). The model components were calibrated individually by using various observed data sources. This effort involved calibrating each model separately and then reviewing highway and transit assignment results for each of the five periods to make additional adjustments in the model components. The adjustments were all made to constants within the models; there were no adjustments to model coefficients. Highlights of results of the calibration are summarized below for each model component. ⢠Vehicle Availability: The vehicle availability model was calibrated primarily on two key variablesâ number of workers per household and superdistrictâ by using the 1990 Census as the primary source of observed data. A second validation test was used to evaluate the total number of vehicles estimated by the vehicle availability model compared with Department of Motor Vehicles estimates of auto registrations. These data were different by 5%. Unfortunately, the 1990 MTC survey, which was used to estimate the model, contained different results for vehicle availability than the 1990 Census. Because the 1990 Census has a much larger sample size, these data were used to calibrate the vehicle availability model. The results, therefore, have indirect effects on the market segmentation of automobiles and workers that was car- ried out in the mode split model. ⢠Full- Day Pattern Tour Models: The full- day pattern tour models were calibrated by converting tours to trips and comparing these to the 1996 MTC survey expanded to match the 1998 population. The 1996 MTC survey was used because the number of trips within San Fran- cisco County was very low in the 1990 MTC survey due to underreporting of trips. The underreporting of trips is not consistent across time periods or across trip purposes, conditions which may have influenced model estimation that was based on the 1990 MTC survey. The differences between trips by period were confirmed with initial assignments by periods with the uncalibrated San Fran- cisco model revealing that the off- peak time periods were significantly underestimated compared with traffic counts. The vast majority of underreporting of trips in the 1990 MTC survey was in other tours. ⢠Destination (Primary- and- Intermediate Stop) Choice Models: The destination choice models were cal- 26 INNOVATIONS IN TRAVEL DEMAND MODELING, VOLUME 2
