[Back]


Diploma and Master Theses (authored and supervised):

A. Dostálová:
"On the use of the Doppler centroid anomaly for the detection of water bodies under the severe wind conditions";
Supervisor: W. Wagner; Department für Geodäsie und Geoinformation, 2015; final examination: 2015-01-19.



English abstract:
Over the last years, methods for automated retrieval of the water bodies and flooded areas from Synthetic Aperture Radar (SAR) satellite data are an important research topic. However, most of these methods have restrictions over rough water surfaces caused by high wind conditions. Therefore there is the need for method that could work also over roughened water surface. One possibility could be the use of other information that is available from the SAR data apart from the traditionally used normalized radar cross-section (NRCS) images.
One of the ancillary datasets delivered along the Envisat Advanced Synthetic Aperture Radar (ASAR) Wide Swath (WS) mode data is a measured Doppler centroid value. The Doppler centroid value is dependent on a relative speed of the satellite and Earth surface towards or from each other. With the knowledge of satellite orbit and attitudes and Earth rotation speed, the theoretical, modelled, Doppler centroid values may be computed. For a stable Earth surface, the difference between these two values (Doppler centroid anomaly) should be close to 0, whereas for the moving surface, such as water surface influenced by surface currents and winds, the anomaly should be dependent on the relative speed of the water surface in the perpendicular direction to the satellite orbit track.
The radial surface velocities based on the Doppler centroid anomalies have already been successfully used for retrieving wind speed and direction over the oceans as well as a usefulness for the studies of ocean currents has been investigated. In this Thesis, the use of the radial surface velocity for detection of inland water surfaces is studied - whether the inland water surfaces are detectable in the Doppler centroid anomaly data and what the limitations of this method are.
Although the relationship between wind speed in the radar line of sight direction and radial surface velocity data was confirmed also for inland water pixels (Pearson correlation coefficient of 0.79), the currently available ASAR WS radial surface velocity data were found unsuitable for the inland water mapping. The main limitations are very coarse spatial resolution and low retrieval accuracy, especially in densely urbanized areas. However, both the spatial resolution and retrieval accuracy are expected to improve in case of the radial surface velocity data derived from the recently launched Sentinel-1 satellite.


Electronic version of the publication:
http://repositum.tuwien.ac.at/urn:nbn:at:at-ubtuw:1-72365


Created from the Publication Database of the Vienna University of Technology.