Search Content

Major network disruptions have significant impacts on local travelers. A good understanding of behavioral reactions to such incidents is crucial for traffic mitigation, management, and planning. Existing research on such topics is limited. The collapse of the I-35W Mississippi River Bridge (August 1, 2007) abruptly disrupted habitual routes of about 14,000 daily trips and forced even more travelers to adapt their travel pattern to evolving network conditions. The opening of the replacement bridge on November 18, 2008 generated another disturbance (this time predictable) on the network. Such "natural" experiments provide unique opportunities for behavioral studies. This study focuses on the traffic and behavioral reactions to both bridge collapse and bridge reopening and contributes to general knowledge by identifying unique patterns following different events. Three types of data collection efforts have been conducted during the appropriate frame of reference (i.e. before vs. after bridge reconstruction): 1) GPS tracking data and associated user surveys, 2) paper and internet-based survey data gauging travel behavior in the post-bridge reconstruction phase, and 3) aggregate data relating to freeway and arterial traffic flows, traffic control, and transit ridership. Differences in reactions to planned versus unplanned events were revealed. Changes in travel cost were evaluated and their temporal and spatial patterns were analyzed. This report concludes with thorough discussions of findings from this study and policy implications.

Report #7 In The Series: Access to Destinations Study. This study presents an effort to track and model land use change in the Twin Cities Metropolitan Region. To that end, we make use of a unique, high-resolution, cell-level set of land use data for the Twin Cities. The data represent 75 meter by 75 meter land use cells, observed at several points in time during the period from 1958 to 2005. These data are used to validate three different types of land use models, which then are used to forecast land use several decades into the future. The models applied in this study include Markov Chain models, Markov Chain- Cellular Automata (MC-CA) models, and an empirical model based on a logistic regression specification. The models are intended to have a simple, transparent structure that allows the user to identify sources of forecast error. Forecasts of land use are made both for the entire study area and also for a specific corridor along State Highway 610 in the northwestern suburbs of the Twin Cities. The study concludes with a brief discussion of the limitations of the models, and how they might meaningfully be expanded and applied.

A well-developed methodology is crucial for both data collection and analysis, and thus the soundness of behavioral models, especially in such a limited time window. Therefore, this paper reviews both theoretical and empirical studies on traffic and behavioral impacts of network disruptions.