Spatial Econometric Models for Panel Data
Cities are constantly evolving, complex systems, and modeling them, both theoretically and empirically, is a complicated task. However, understanding the manner in which developed regions change over time and space can be important for transportation researchers and planners. In this paper, methodologies for modeling developed areas are presented, and spatial and temporal effects of the data are incorporated into the methodologies. The work emphasizes spatial relationships between various geographic, land use, and demographic variables that characterize fine zones across regions. It derives and combines land cover data for the Austin, Texas, region from a panel of satellite images and U.S. Census of Population data. Models for population, vehicle ownership, and developed, residential, and agricultural land cover are estimated; the effects of space and time on the models are shown to be statistically significant. Simulations of population and land cover for the year 2020 help to illustrate the strengths and limitations of the models.