| Assessing the impact of the built environment on travel behavior can yield valuable tools for land use and transportation planning. In this study, a method is proposed by which geospatial characteristics of the built environment may be quantified and used to model spatial variability in travel behavior at an aggregate level. The GIS-based method explores many geostatistical concepts and their applicability to transportation planning. The models developed may be used to supplement four-step planning models that have little inherent sensitivity to changes in land use and other aspects of the built environment. This method is applied to a study area composed of Buffalo, NY, and its surrounding communities. Among the main conclusions of the study are that zonal mode choice is highly correlated to built environment factors, even when controlling for relevant household demographics such as vehicle ownership, and that home-based VHT and VMT are affected by the built environment to a lesser degree than by social or economic factors. Statistical concepts such as regression Cp minimization, principal component analysis, and power transformations are explored and found to be methodologically beneficial. To conclude the study, the method is applied to a hypothetical land use scenario to estimate the reduction in zonal vehicle dependency caused by high-density development in suburban areas. |
