Walking to the station: The effects of street connectivity on walkability and access to transit

This thesis analyzes an on-board transit survey conducted by the Atlanta Regional Commission in order to determine how far urban density, mixed land-uses, and street network connectivity are related to different walking behaviors, namely transit walk-mode shares and walking distances to/from stations. The data are drawn from all the stations of Atlanta's rapid transit network (MARTA). Allowing for quite a bit of noise in the data, some of the findings confirm for the case of Atlanta what a review of existing literature would lead one to expect: mixed land-use and denser street networks are associated with higher proportion of riders traveling to/from the station "walking" (noise in the data does not allow to fully distinguish with certainty walking as the sole mode of access to/from the station from walking combined with the use of bus services). The thesis also explores questions that have not been previously covered systematically in the literature. First, does urban form (including street configuration and connectivity as well as land-use patterns) affect the distance transit patrons are willing to walk? Findings suggest that street networks with denser intersections and more linear alignments of road segments support greater walking distance thresholds. Second, does the location of the station relative to the street hierarchy of the surrounding area affect the proportion of patrons walking or the distance walked? The thesis answers this question negatively. If the surrounding area, at a one mile radius, is a transit friendly urban form, the location of the station within the street hierarchy does not have a major impact on walk mode share and the distance walked. In light with the various conclusions presented in this thesis, the finding regarding the association between street connectivity and distances walked appears to be the most critical.;The research findings have several implications. They confirm that urban form (including density, land-use and street network configuration) affects the proportions of patrons walking to/from the station. Thus, they also confirm that transit oriented policies are better supported by urban development policies and zoning and subdivision regulations that encourage transit-friendly urban forms. More specifically, they suggest that the scale at which urban form has an impact on pedestrian travel is of the order of a mile radius, rather than a few blocks around the station. Findings also suggest that transit oriented policies are compatible with policies aimed at the enhancement of health and the reduction of obesity through daily physical activity (walking to/from the station can contribute a significant part of the daily activity recommended by Healthy Living Guidelines). Finally findings augment the knowledge-base that supports transit oriented development by emphasizing the contribution of the spatial structure of the street network, over and above the impact of side-walk provision and design and pedestrian safety.