Thursday 25 August 2005

G6
1330-1530 hours

396
Asian tsunami damage assessment with Radarsat-1 SAR imagery
Chang, Hsing-Chung¹, Ge, Linlin¹, Rizos, Chris¹
1 School Of Surveying and Spatial Information Systems, UNSW, Sydney, NSW, Australia
Author email: hsing-chung.chang@student.unsw.edu.au
Synthetic aperture radar (SAR) systems have advantages over optical imaging systems due to their ability of penetrating cloud and independence from solar lighting conditions. The SAR intensity image shows the characteristics of scattering mechanisms of the ground surface. For example, buildings interact with radar signal as corner reflectors and show very strong reflected signal back to radar. Interferometric SAR (InSAR) also measures the coherence which is the correlation between two SAR images acquired at different time. The signatures of backscattering intensity and coherence can be used for landcover classification and forest mapping. The impact of Asian Tsunami struck on 26 December 2004 was examined by measuring the degradation of coherence and changes in intensity in two Radarsat-1 SAR images. InSAR coherence map of Banda Aceh, Indonesia, was derived using the two images acquired on 22 July 2003 and 31 December 2004.With the perpendicular baseline of 605m and temporal separation of 528 days between the acquisitions, the coherence image revealed strong temporal decorrelation therefore the damaged areas were very difficult to be distinguished. However, the Tsunami caused serious coast line erosion, flooding, devastation to the local infrastructures and vegetations which led to the changes of the dielectric properties of the imaging pixels in the SAR images before and after the impact. The intensity difference map showed clear evidence of change in landcover, especially for the areas where the vegetation and buildings were flushed away. Visual confirmation was performed using the optical satellite images acquired by the QuickBird before and after the incident.

Return to Oral Presentations