Descriptive and Comparative Studies of 1990 Urban Extent Data for the New York Metropolitan Region

Regional planners and decision makers are teaming up with data providers and urban growth modelers in the ongoing effort to map historic extents of urbanization, forecast future extents, and predict impacts on the local environment. Spatial data layers that are nationally consistent across space and time are valuable input variables for some predictive models, but can be of limited use because of varying definitions of urban land and methods of data capture. This article reports on a descriptive and comparative study of two national data sets for mapping the extent of 1990 urbanization for the New York-New Jersey-Connecticut metropolitan region: 1) urban land classes based on Multi Resolution Land Characteristics data, and 2) block group urbanized area data from the United States Census Bureau. As multiple data sets from the earth, natural and social sciences proliferate and are injected into models, it is imperative that researchers and practitioners conduct and share findings of such comparative studies as part of the emergent metadata culture. Urban Growth Modeling for the NYNJ-CT Metropolitan Region In general, urban dynamics research intends to evaluate, utilize and enhance predictive models that at a minimum would allow urban planners, local government officials and the general public to visualize future urbanization growth patterns and potential impacts on the local environment. One such model, the SLEUTH model, also called the Clarke Cellular Automata Urban Growth Model and the Clarke Urban Growth Model, has six data inputs: slope, land use, exclusion, urban, transportation, and hill shading (Clarke et al. 1997, U.S. Environmental Protection Agency 2000). An underlying assumption is that historic growth trends will continue and that the future can be projected based on these trends. In the SLEUTH model, urban land is defined as residential, commercial, mixed use, and industrial land uses (U.S. Environmental Protection Agency 2000) and for the 1930s to 1990s was generally compiled using aggregated, multiple source scanned graphics and digital vector and image data from historic maps, satellite imagery, aerial photos, Census statistics, and commerce records (Acevedo et al. 1996). At least four urban time periods are required, with one defined as the “seed” or start year from which to predict urban expansion. In the case of the study region, the seed year is 1990 and the main sources of data have been the U.S. Geological Survey, the Regional Planning Association, the New York Public Library, and the U.S. Census Bureau. Table 1 Counties and boroughs in study region Connecticut New York New Jersey Fairfield, Hartford, Litchfield, Middlesex, New Haven Boroughs: Borough of Queens (Queens County), Borough of Manhattan (New York County), Borough of Brooklyn (Kings County), Borough of the Bronx (Bronx County), Borough of Staten Island (Richmond County).Counties: Dutchess, Nassau, Orange, Putnam, Rockland, Suffolk, Sullivan, Ulster, Westchester Bergen, Essex, Hudson, Hunterdon, Mercer, Middlesex, Monmouth, Morris, Ocean, Passaic, Somerset, Sussex, Union, Warren