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.
aggregate analysis has been resolved, with general agree- ment on the use of disaggregate analysis. Advances in dis- crete choice modeling include the use of multinomial and nested logit models, generalized extreme value models, and mixed logit models. These models better represent complex choices and substitutions among those choices. Addressing computation- related issues has been helped by the development of enhanced algorithms for many aspects of transportation analysis. The adoption of microsimulation in travel prediction and in transport systems operations has also been a major improvement. Enhanced data collection and processing has enhanced analysis techniques for control purposes. There have been significant advancements in land use modeling. We have also seen improvements in our abil- ity to represent spatial detail and realistic land market representation. There has been increased application of geographic information systems (GIS) in transport sys- tem components, trans- shipment terminals, and land use components. Dynamic assignment is one of the most central issues related to transportation service representation. How- ever, to be effective, such analysis must be informed by a good time- of- day representation of demand. Other issues include spatial detail, consistency across levels of detail, refinement of travel supply functions, and enhanced operational control. Despite the tremendous progress that has been made, we continue to need an increased understanding of behav- ioral responses. Both modelers and policy makers need to understand why we sometimes get unintended conse- quences from decisions that seem wise based on travel forecasts. We also need a better ability to focus on small areas or population groups, as well as the ability to rep- resent more complex behaviors in models. As we learn more about what we know and what we do not know, the limitations of existing models become more obvious. Other issues that need to be addressed include improved presentation of all the aspects of transportation and activity- based analysis, including supply, demand, and land use. One of the core issues is the need for better communication between modelers and decision makers. Modelers need a fuller understanding of policy issues and to improve the presentation of modeled impacts to policy makers. Modelers and decision makers also need to expand the range of decision options to be examined. STATE AND LOCAL AGENCY PERSPECTIVES ON POLICY ISSUES: ONE SIZE DOES NOT FIT ALL Michael Morris This is a very critical conference at a very critical time. My remarks focus on what state and local officials expect from the modeling community. I would like you to write down four items to remember: (a) what you do is very important; (b) things are changing very rapidly; (c) things are changing faster than our ability to respond; and (d) under certain circumstances, we need to change the foundation of the way we do our jobs. Figure 4 highlights the role of MPOs in the trans- portation decision- making process. I think Keith Lawton and others in the Portland area deserve a great deal of credit and thanks for the work they have done on testing activity- based models and other applications. I think all MPOs have an obligation to the modeling communityâ not just as users of models, but also as developers and testers of models. MPOs receive guidance from the U.S. Department of Transportation, citizens, local jurisdictions, and policy makers. Many MPOs are involved in sharing experiences with different models and applications. Within Texas, we have partnerships to help integrate models across the state. I think MPOs should be more involved in collabo- rative research and we should be encouraging more part- nerships with the private sector and with universities. We all have an obligation to increase communication, and by increasing communication we can help advance the state of the practice, the state of the art, and the decision- making process. Models are not an end to themselves. Travel models are a tool to help in the decision- making process. I think we are on the verge of a new foundation in travel mod- els. Too often in the past, the information system and the model system have been the same. We often have a prob- lem with the public and policy makers confusing the steps in the transportation planning process and the steps in the modeling process. Figure 5 illustrates the role that travel models play in the implementation process. I think there has to be a broader information system that creates a foundation for the whole transportation planning process. Travel models then become part of this broader foundation. Travel models, land use models, goods movement mod- els, and inputâoutput models are all needed. As Frank Koppelman noted, I think one shortcoming is that we do not spend enough time interpreting the model results. The goal of travel models is not to forecast the exact number of vehicles that will use a collector street in 2010. The goal is to interpret what that number means. One important issue is the different scales for fore- casting tools and the need for consistency. We need a strong national transportation system that includes national transportation data sets. The national data need to be integrated into statewide data sets. Within Texas, new intercity corridors, called the Trans- Texas Corri- dors, are being developed. Tools are needed to assess travel in these corridors that are consistent with those 4 INNOVATIONS IN TRAVEL DEMAND MODELING, VOLUME 1
