The Combined Use of Optical and SAR Data for Large Area Impervious Surface Mapping

Kurzfassung

One of the megatrends marking our societies today is the rapid growth of urban agglomerations which is accompanied by a continuous increase of impervious surface (IS) cover. In light of this, accurate measurement of urban IS cover as an indicator for both, urban growth and environmental quality is essential for a wide range of urban ecosystems studies. The aim of this work is to present an approach based on both optical and SAR data in order to quantify urban impervious surface as a continuous variable on regional scales. The method starts with the identification of relevant areas by a semi automated detection of settlement areas on the basis of single-polarized TerraSAR-X data. Thereby the distinct texture and the high density of dihedral corner reflectors prevailing in build-up areas are utilized to automatically delineate settlement areas by the use of an object-based image classification method. The settlement footprints then serve as reference area for the impervious surface estimation based on a Support Vector Regression (SVR) model which relates percent IS to spectral reflectance values. The training procedure is based on IS values derived from high resolution QuickBird data. The developed method is applied to SPOT HRG data from 2005 and 2009 covering almost the whole are of Can Tho Province in the Mekong Delta, Vietnam. In addition, a change detection analysis was applied in order to test the suitability of the modelled IS results for the automated detection of constructional developments within urban environments. Overall accuracies between 84 % and 91% for the derived settlement footprints and absolute mean errors below 15% for the predicted versus training percent IS values prove the suitability of the approach for an area-wide mapping of impervious surfaces thereby exclusively focusing on settlement areas on the basis of remotely sensed image data.