Quantifying changes to the urban morphology of Dublin with spatial metrics derived from medium resolution remote sensing data

Satellite images of medium resolution are cheap, widely available and are often part of extensive historic archives,which makes them ideally suited to study urban growth. Their lower resolution, on the other hand, hampers studying urban morphology and change processes at a more detailed, intra-urban level. In this paper, we develop spatial metrics for use on continuous sealed surface data produced by a sub-pixel classification of Landsat TM and ETM+ imagery. Three approaches are compared to derive the sealed surface fractions at sub-pixel level: linear regression analysis, linear spectral mixture analysis and a multi-layer perceptron (MLP). The metrics represent the shape of the cumulative frequency distribution of the estimated sub-pixel fractions within each spatial unit by fitting a
transformed logistic function with a nonlinear least-squares approach. A MLP classifier is then used to relate the metric variables to combined urban land-use classes selected from the European MOLAND topology. In combination with density information derived from the sealed surface maps, our approach allows producing maps that show changes in urban morphology.

ISBN:978-1-4244-3461-9

Publication year:2009

Keywords:urban remote sensing, multi-layer perceptrons, multi-temporal remote sensing images