| Description |
I have opted for a three-paper dissertation, studying the relationship between travel and the built environment for three types of trips: walk and bike trips by the entire population, trips from home to school and back for students, and trips of all types by the elderly. As part of my dissertation, I have gathered the most extensive set of regional travel surveys that anyone has ever collected, specifically including 815,160 trips by 81,056 households in 23 regions. I have also linked travel records to so-called D variables for buffers of different widths around households and routes from home to school. The five D variables, widely used in travel research, are development density, land use diversity, street network design or connectivity, destination accessibility, and distance to transit. The main goal of this dissertation is to determine how we can promote walking and biking, especially for students and seniors. From the first paper, walk mode choice in the 23 regions depends primarily on land use diversity, street connectivity, and transit accessibility, while bike mode choice depends primarily on street connectivity and transit accessibility. The resulting trip chain shows that accessibility of destinations to one another may be almost as important as accessibility of residences to destinations. The second paper analyzes student travel to school in the 14 regions. I find that the most important D variables in the decision to walk or bike to school is development density and street network design or connectivity, and the least important is land use diversity. While not a D variable exactly, the need to cross major roads or commercial developments has strong negative impacts on active travel to school. In the third paper, the analysis of variance (ANOVA) tests show that seniors living in compact neighborhoods are more active than those living in sprawl neighborhoods. They generally travel more and travel more by walking and public transportation, yet travel less by automobile. The resulting models and findings in this dissertation are appropriate for post-processing outputs of conventional travel demand models, and for sketch planning applications in traffic impact analysis, climate action planning, and health policy implementations. |
