Urban vegetation and social change: An analysis using remote sensing and census data.

The use of satellite data has great potential for research on urban environmental change. Prior research utilizing remote sensing data in urban areas has shown a direct relationship between amount of vegetation and the socio-economic characteristics of neighborhoods at one point in time. However, the significance and utility of remote sensing data in the analysis of environmental and social change in urban areas has not yet been established. By processing two Landsat data sets from May 1975 and May 1992, to produce a binary change image for vegetation in Wayne County, Michigan, and merging the binary change image with historic census data in a Geographical Information System this study associates change in the amount of vegetation in the City of Detroit with the social, economic, demographic and land use characteristics and changes which took place during the same time period. The overall change in amount of vegetation over time is shown to be most closely associated with the residential data from the earlier time period. A predictive model of vegetation change is tested. Through techniques of spatial data analysis vegetation change in each observed area is shown to be highly dependent on conditions and changes in surrounding neighborhoods. A final model incorporates spatial autocorrelation effects and a two-stage least squares feedback system.PhDEarth SciencesGeographyRemote sensingSocial SciencesUrban planningUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/131320/2/9840642.pd