| This dissertation examines transportation-land use interactions in the Halifax-Dartmouth region. First, it investigates the changing urban form of the region. Then, it examines both directions of the transportation-land use relationship by quantifying the impact of residential land development on road development and the impact of road development on residential land development.; The dissertation begins by computing kernel estimates to investigate the decentralization and deconcentration of residential and commercial land parcels, and to examine the segregation of land uses. The results suggest that changes in urban form were a combined result of infill, contiguous, and leapfrog development. Next, univariate and bivariate K functions are estimated to measure spatial dependence within and between the classes of residential and commercial land parcels. The results suggest that residential land parcels cluster together, commercial land parcels cluster together, and over time residential and commercial land parcels have become more clustered.; With a better understanding of the changes in urban form, the dissertation then examines both directions of the transportation-land use relationship. An ordered probit model is specified where residential land development is a function of either the change in accessibility or the distance to a high-speed road and other explanatory variables. A spatial lag model is estimated where the change in accessibility is a function of residential land development and other explanatory variables. The results of the first model suggest that road development does have an impact on land development. However, the results of the second model indicate that land development does not drive road development. Collectively, the results of both models provide important insight into the transportation-land use relationship. Understanding the direction and strength of this relationship is imperative for making informed policy decisions. |
