Innovations in Travel Demand Modeling, Volume 1: Session Summaries (2008)

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.

documentation. There is also an issue with the lack of full integration of transit and highway networks. LION is used in New York City, while TIGER is used in other parts of the region. The lack of integration of land use models and proper feedback to NYBPM is another issue. There is a land use model, but it is not yet integrated into the NYBPM. Also, the processing time for the NYPBM is very long. Initially it was taking up to 7 days to run the model. The computer processing time alone has been reduced to approximately 4 days. A current project is examining approaches to reduce this time further. • The modeling environment also presents chal- lenges. These challenges include the diversity of the large region, software issues, and hardware issues. The region encompasses almost 10,000 square miles. There is also diversity in the size, population, and employment of each zone, as well as variation of available modes and con- nections. There is variation in the travel behavior and travel patterns of different population subgroups and there is a large group of temporary workers who are con- tinually on the move. The travel behavior of people in New Jersey is very different from that of people in New York. The term “Manhattan Syndrome” is used for unexplained travel behavior. • The software issues include changes in TransCAD and the compatibility of various software packages. The NYBPM was developed using TransCAD 4.0. The NYBPM modules are not compatible with the newer ver- sion of TransCAD. An ongoing challenge has been that users have different versions of TransCAD. Some users have experienced difficulties in running the model because of this problem. • One hardware issue relates to data storage needs. With nearly 9 million households in the base year, the journey production files are over 500 megabytes. The mode destination choice model processes over 25 million paired journeys by eight trip purposes. The output files are over 300 megabytes. There are six highway classes and four transit trip tables for each of the four time peri- ods. The combined file size is approximately 2.5 gigabytes. • There is an ongoing need to provide training to various stakeholders and users. A 1-day training course is provided for decision makers. A 3- to 5-day training course is available for individuals with some modeling background. Hands- on training, spanning several weeks, is provided for staff of member agencies who will be using the model for specific projects. These training courses focus on the purpose and use of the activity- based models and the specific use of the NYBPM. Staff from other agencies can also use the NYMTC’s com- puter facilities. The lack of trained and experienced mod- eling staff at many agencies is an ongoing challenge. Many public agencies experience high staff turnover rates. • Working with stakeholders is an ongoing priority. Involving stakeholders in defining the model needs and applications is important. It is also important to involve stakeholders in the model calibration and validation processes and in the discussion of model usage and improvements. Gaining consensus on the definition of the zonal system, the survey design, and the forecasts and calibration results was also accomplished by work- ing with stakeholders. It took time to reach an agree- ment on the socioeconomic and demographic data at the county level. • A number of model improvements are planned and a new wave of data collection is scheduled for in 2007. These efforts include a household travel survey, an air- port survey, a taxi survey, and a work place survey. Tran- sit origin–destination surveys and bridge and tunnel origin–destination surveys are also planned. Other efforts include a cordon survey, a travel time survey, and traffic counts and occupancy surveys. The results from these surveys will be used to update the NYBPM. • In addition to incorporating the results of the new surveys, other planned activities include improving the highway- transit connection, improving transit models, and integrating NYBPM with the land use model. Possi- ble web applications are being explored, including pro- viding model output analysis capabilities and model run capabilities for more of a distributed process. A plan has also been made to provide improved flowchart- based, online help and documentation. More project applica- tions in the region are anticipated. The NYBPM user’s group support will continue to meet every 2 months to help with these improvements. USING ACTIVITY- BASED MODELS FOR POLICY DECISION MAKING Erik Sabina and Thomas Rossi Erik Sabina discussed the development of a new activity- based travel forecasting model for the Denver metropol- itan area by the Denver Regional Council of Governments (DRCOG). He described the planning process in the Denver area and the proposed activity- based model. Volume 2 includes a paper on the topic.2 The following points were covered in his presentation. • A number of factors influenced consideration of a new activity- based travel forecasting model for the Den- ver area, including the Integrated Regional Model (IRM) vision phase conducted by DRCOG, the new Colorado Tolling Enterprise (CTE) within the Colorado Depart- 55THE SECRET IS IN THE SEGUE 2 See Sabina, E., and T. Rossi. Using Activity- Based Models for Policy Decision Making. Volume 2, pp. 177–180.