[Back]

S. Schlaffer, M. Hollaus, W. Wagner, P. Matgen:
"Flood delineation from synthetic aperture radar data with the help of a priori knowledge from historical acquisitions and digital elevation models in support of near-real-time flood mapping";
Talk: Earth Resources and Environmental Remote Sensing/GIS Applications III, Edinburgh, UK; 2012-09-24 - 2012-09-27; in: "Proceedings of SPIE", edited by Daniel L. Civco, Manfred Ehlers, Shahid Habib, Antonino Maltese, David Messinger, Ulrich Michel, Konstantinos G. Nikolakopoulos, Karsten Schulz,, 8538 (2012), 853813-1 - 853813-9.

The monitoring of flood events with synthetic aperture radar (SAR) sensors has attracted a considerable amount
of attention during the last decade, owing to the growing interest in using spaceborne data in near-real time flood
management. Most existing methods for classifying flood extent from SAR data rely on pure image processing
techniques. In this paper, we propose a method involving a priori knowledge about an area taken from a multitemporal
time series and a digital elevation model. A time series consisting of ENVISAT ASAR acquisitions
was geocoded and coregistered. Then, a harmonic model was fitted to each pixel time series. The standardised
residuals of the model were classified as flooded when exceeding a certain threshold value. Additionally, the
classified flood extent was limited to flood-prone areas which were derived from a freely available DEM using the
height above nearest drainage (HAND) index. Comparison with two different reference datasets for two different
flood events showed that the approach yielded realistic results but underestimated the inundation extent. Among
the possible reasons for this are the rather coarse resolution of 150 m and the sparse data coverage for a substantial
part of the time series. Nevertheless, the study shows the potential for production of rapid overviews in near-real
time in support of early response to flood crises.

Flood mapping, Synthetic aperture radar, ENVISAT, Sentinel-1, Remote sensing, HYDROSHEDS, Digital elevation model

http://dx.doi.org/10.1117/12.974503