Echelon analysis of the relationship between population and land cover pattern based on remote sensing data

With continuing proliferation of human influences on landscapes, there is mounting incentive to undertake quantification of relationships between spatial patterns of human populations and vegetation. In considering such quantification, it is apparent that investigations must be conducted at different scales and in a comparative manner across regions. At the broader scales it becomes necessary to utilize remote sensing of vegetation for comparative studies against map referenced census data. This paper explores such an approach for the urbanized area in the Tokyo vicinity. Vegetation is represented by the normalized difference vegetation index (NDVI) as determined from data acquired by the thematic mapper (TM) sensor of the Landsat satellite. Sparseness of vegetation is analyzed in relation to density of human residence, first by regression analysis involving stratified distance zones and then by the recent echelon approach for characterization of surfaces. Echelons reveal structural organization of surfaces in an objective and explicit manner. The virtual surface determined by census data collected on a grid is shown to have structural correspondence with the surface representing vegetation greenness as reflected in magnitude of NDVI values computed from red and infrared bands of image data.