Monitoring Crustal Dynamics in the Mediterranean Using ERS Differential Radar Interferometry
Contact Persons
Thomas Egli |
Keywords
2 pass interferometry, synthetic interferogram, differential interferogram.
Quick Reference
see: Differential SAR-Interferometry
Some Results
An Application of Differential SAR-Interferometry: The Earthquakes from August 1996 in the Region of Konitsa (Northern Greece)
In August 1996 two medium-strength earthquakes occured in the area of Konitsa (near the greek-albanian border). The first event occurred on 5th of August with a magnitude of 5.7 at Richter-scale, the second one on the 20th of the same month with a magnitude of 5.3. In order to detect the topography displacements the 2-pass method of differential SAR interferometry was applied.
For the execution of the 2-pass method a digital elevation model (DEM) and at least two SAR scenes are required. The two SAR scenes are used to generate an interferogram (real interferogram). The phase of this interferogram correlates with topography and changes in topography. To analyze the topographical displacements the topography-dependant part of the phase needs to be eliminated, requiring a DEM. The height values need to be recorded and converted to synthetic phase-values (synthetic interferogram). In a next step the phase-values of the real and synthetic interferogram have to be subtracted from each other. In this way residual phase-values are obtained, resulting in a differential interferogram, which is uniquely correlated to variations in the topography. However, before computing the amount of displacement, the influence of errors arising from the DEM or baseline-induced phasetrends have to be determined.
For the phase investigation six suitable SAR scenes were found. From these six scenes 3 interferometric pairs were produced. But only two of them could be used for the interpretation. To generate the synthetic interferograms a DEM calculated from SPOT data was used.
A first interferometric pair was acquired during the ERS tandem-mission (mid-May 1996). As a result, the time difference between the two acquisitions was just one day, yielding a reference differential interferogram not influenced by any earthquakes. It was used to judge the DEM-accuracy, which was proven to be high in the range of the Konitsa plaine.
Reference differential interferogram not influenced by earthquakes.
A first interferometric pair was acquired during the ERS tandem-mission (mid-May 1996). As a result, the time difference between the two acquisitions was just one day, yielding a reference differential interferogram not influenced by any earthquakes. It was used to judge the DEM-accuracy, which was proven to be high in the range of the Konitsa plaine.
For the most important interferometric pair of our investigation, data from July and August were acquired. The time difference between these two acquisitions is 35 days. It is during this time period that the two earthquakes mentioned above took place. Despite of the long time period between the two acquisitons, the coherence-values are unexpectedly high (about 0.5). This is probably due to the lack of major variations in the vegation during this time. In contrast to the differential interferogram acquired in May 1996, there are very obvious phase trends existing; two fringes are clearly visible. The two fringes mainly cover the Konitsa plaine and the hilly area further north of the plain. Comparing this July-August interferogram with the differential interferogram of May, we can exclude DEM-error as a primary phase influence source. With a height of ambiguity of 135m for the July-August pair a DEM-error would have been about 270m in magnitude. This is impossible, because otherwise the May pair would have displayed a similar of phase-pattern.
A baseline-induced phase trend as the cause for the two fringes can also be excluded, as filtering gave no evidence to support this. It can be shown that the differential phase from the fringe-margin to the fringe-centre represents a subside movement. Because of the soft subsoil of the Konitsa plaine, this is the expected displacement direction.
In spite of a lack of geological maps, which would have aided in the determination of exact correlations between the differential fringes and the bedrock, two points should be noted: At the southeastern edge of the Konitsa plane, the fringes stop abruptly. In this area the mountains reach elevations of over 2000m a.s.l. This phase pattern can be explained by an abrupt transition between the soft subsoil of the Konitsa plaine and the step mountains. At the northern edge of this plaine the situation looks different. The fringes also cover some of the hilly region further north. Comparison with the topographical map indicates that this hilly area seems to be subject to much erosion, which implies soft subsoil and hence a reason for the spreading out of the fringes in this area.
All the points mentioned above support a tectonic cause for these differential fringes. However, it is difficult to quantify the crustal movements exactly, because of too few differential interferograms with which to compare the data. Because one fringe corresponds to a slant-range difference of a half-wavelength (wavelength of ERS-1/2=5.6cm), this also provides us with an estimate of the order of magnitude of the crustal displacement. That explains why a maximum subsidence of 2*1/2-wavelength (5.6 cm) was seen.
Publications
- EGLI, T. [1998]:
Eine Anwendung der differentiellen SAR-Interferometrie: Die Erdbeben vom August 1996 in der Region Konitsa (Nord-Griechenland), Diplomarbeit Geographisches Institut, Universität Zürich - BIEGGER, S. [1998]:
Untersuchungen zur Generierung eines DHM's von Kefalonia (Griechenland) mit Multi-Baseline SAR-Interferometrie, Diplomarbeit Geographisches Institut, Universität Zürich
Partnerships
This work is done under the auspices of the Inter-University Collaboration for Earth Observation and Geoinformatics by the
- Remote Sensing Laboratories, Department of Geography, University of Zürich
- Institute of Geodesy and Photogrammetry, ETH Zürich
- Institute of Geophysics, ETH Zürich