A preliminary analysis of socio-economic and accessibility attributes and landscape patterns in a coastal urban area

In this paper, we report the result of our preliminary research that aims to examine how spatio-temporal landscape patterns can be related to socio-economic and accessibility attributes with a coastal urban area as the case. Our research methodology emphasizes the use of remotely sensed data, in combination with census data and other geographically referenced data. We classify two satellite images to derive land use/cover information. Census data are linearly interpolated and redistributed by using the dasymetric mapping technique. Accessibility conditions considered include terrain elevation and slope and separate Euclidian distance measurements of each pixel to selected geographic features. We analyze the empirical relationship between the spatio-temporal landscape patterns and socio-economic and accessibility attributes by using step-wise multivariate regression. Our initial results suggest that socio-economic and accessibility metrics can explain much of the variability of the landscape patterns. Our study demonstrates that remote sensing and GIS-based spatial analysis can be quite useful to examine the possible drivers leading to land use/cover changes; such an understanding can help develop more realistic models for future change prediction, which is critical for environmental planning and management in such a rapidly growing coastal urban area.