MONITORING SURFACE DEFORMATION IN ACTIVE MARGIN SETTINGS WITH THE SCANSAR TECHNIQUE USING WIDE SWATH ENVISAT DATA J. Anderssohn (1) and H. Kaufmann (2) (1) GeoForschungsZentrum Potsdam, Department of Geodesy and Remote Sensing, Telegrafenberg, D-14473 Potsdam (Germany), Email: calypso@gfz-potsdam.de (2) GeoForschungsZentrum Potsdam, Department of Geodesy and Remote Sensing, Telegrafenberg, D-14473 Potsdam (Germany), Email: charly@gfz-potsdam.de ABSTRACT Active continental margins are the Earth's principal regions of active deformation and significant earthquake hazard. We apply the enhanced method called Scan Synthetic Aperture Radar (ScanSAR) interferometry to quantify active surface deformation in the coastal region of the South American active margin in central Chile between 22 S and 27 S. Here, as in other active margin settings, crustal deformation may reach far inland and affects hundreds of square kilometer wide areas. The normal ENVISAT ASAR Strip-Mode covers a 100km wide stripe and will therefore not be sufficient for such deformation study. We use the ENVISAT ASAR Wide Swath (WS) data that cover in generally an area of 400x400km². Because of this we investigate the potential of WS data to monitor such wide-stretched surface displacement. Therewith a high spatiotemporal surface deformation monitoring in active margin settings can be given. 1. OBJECTIVES Active margin settings, here in the meaning of active subduction zone, are very complex systems of interaction between the overriding and subducting plate. The high peaks and volcanoes of the Andes and the great earthquakes along the South American coast are dramatic manifestations of the oblique subduction of the Nazca plate beneath the continental South America plate [1]. Convergence occurs over a 500- to 1000 km wide boundary zone and plays a major role of South American continent s evolution. Here in this study we aim to evaluate and present the potential of ScanSAR interferometry in terms of detecting large-scale or wide spread surface deformation. Studies of such surface deformation can significantly improve the understanding of the complex subduction zone processes. The more the subduction zone processes will be understood, the merrier hazard potential can be assessed and the risk of human tragedies caused by earthquake and volcano eruption reduced. 2. SURFACE DEFORMATIONS IN ACTIVE MARGIN SETTINGS (AMS) The relative plate motion between the Nazca and South America plate is assumed to be partitioned into several components. On the interface s locked portion at the trench some motion accumulates and releases abruptly when the interface ruptures in large thrust earthquakes (co-seismic). The most intense ever recorded co-seismic event on Earth, the 1960 Valdivia earthquake at a rate of M w 9.5 [2] took place here on the West coast of South America. Other prominent earthquakes, as the 1985 M w 7.9 Valparaiso and the 1995 M w 8.0 Antofagasta earthquakes [3], are just two examples of the strength of the active subduction zones. The accumulated motion on the locked portion of the interface at the trench between co-seismic events causes transient elastic deformation and is denoted as interseismic deformation (few centimetres horizontal displacement per year). Both, interseismic and co-seismic processes are well monitored and documented in the past by GPS measurements and processing [3][4][5][6]. Another consequence of an active subduction zone is the movement of magma (volcanic evolution) or hydrothermal fluids at depth [7]. The co- and interseismic deformation as well the volcanic evolution can be measured by certain techniques such as GPS, levelling and gravimetry. 3. INTERFEROMETRIC SYNTHETIC APERTURE RADAR (INSAR) IN AMS The lack of the mentioned techniques is the sparse ground resolution. Space geodetic techniques like interferometric synthetic aperture RADAR (InSAR) have the potential to overcome this lack. SAR Strip- Mode (SM) products acquired by the European satellites ERS and ENVISAT have spatial extension of about 100x100km² with a ground resolution of about 20m. This allows covering and monitoring large areas with a high continuously area-wide ground resolution. With absence of significant vegetation and a satellite recurrence of 35 days both satellites provides a unique high spatiotemporal technique for surface deformation Proc. Envisat Symposium 2007, Montreux, Switzerland 23 27 April 2007 (ESA SP-636, July 2007)
studies. References [7] and [8] demonstrated the potential for surface deformation studies applying InSAR in AMS. Through these mentioned studies, a number of active volcanoes are known and InSAR results for interseismic signal monitoring have been demonstrated. But the interseismic-induced surface displacement may reach far inland affecting hundreds of square kilometers wide areas and can easily exceed the SM ground coverage. This makes it quite difficult to estimate the interseismic-induced signal in the SM products and distinguish from orbit error influences. Furthermore the SM products cannot be used to calibrate the relative measurements in terms of absolute values due to the inherent relative InSAR measurements. Geodetic, geologic and seismic studies demonstrated decreasing amplitude of interseismic surface deformation up to insignificant deformation with increasing distance, in the meaning of hundreds of kilometers, to the West coast of South America. The width of SM products is insufficient to cover the area between the coast and the supposed inland stable part. slave data acquisitions would be required with 16x a small baseline and 16x sufficient coherence. But because of the varying data acquisition time for all 16 tiles and different potential error influences on each interferogram, no useful merge or mosaic is feasible. To conclude, we can summarize ScanSAR interferometry is currently the best remote sensing technique for largescale or wide-spread deformation monitoring. 4. SCANSAR APPLICATION IN STUDY AREA To overcome this problem we apply the ScanSAR interferometry using Wide Swath (WS) data acquired by ENVISAT ASAR instrument. Fig. 1 depicts a principal scheme of WS data acquisition by ENVISAT ASAR. The system transmits pulses to and receives echoes from a sub-swath for a period long enough to synthesize a radar image of the area within the beam footprint at the required resolution. It then switches beams to illuminate a different sub-swath and continues in this manner until the full-wide swath is covered at which point it returns to the original sub-swath and the scanning cycle is repeated [9]. Such WS data have a coverage of about 405km width. Processing the so called level-0 data (raw data) enables the user to process a long stripe of 405km width and hundreds up to thousands kilometer length, only limited by data acquisition s azimuth length. For all large-scale deformation studies on Earth this is a unique technique to monitor such wide areas with a ground resolution of about 100 meters. There is no other geodetic survey technique with comparable potential. 4.1. InSAR strip-mode vs. ScanSAR WS mode Beside the mentioned big advantage to cover an area of four times larger width, some more advantages of ScanSAR are note worthy, in terms of large-scale or wide-spread deformation studies, with respect to the SM products. A total of 16 SM level-1 ENISAT ASAR products, with a fixed physical size of 100x100km², would be required to cover the same area as using one WS (400x400km²) product. This means 16 master and Figure 1. Principal scheme of WS data acquisition [9] 4.2. Wide Swath Observations in study area Our study area is a subset of the West coast of South America ranging 22 S 27 S and extends about 400km inland. In the Southwest of the study site the city Antofagasta is located, here one of the most intense recent earthquakes occurred in 1995 with a magnitude of M w 7.9. We choose one ENVISAT track to fit our study area and selected seven WS data acquisition. These selected data sets have the potential to generate 19 interferograms. One important condition for successful ScanSAR interferometry processing is the presence of the Burst Synchronization (BS). Technical details can be found in [10][11]. Here we aim to focus this paper more on the application than on theoretical details. Five potential interferograms could not be generated due to the absence of BS. Seven other interferograms were refused because of low coherence caused by large baselines, so that seven interferograms
were finally available for our study. These data sets can be separated in three division: three interferograms are spanning a time range of 2-3, two interferograms have a temporal baseline of about six month and two monthly interferograms. considering GPS derived interseismic-induce large-scale surface deformation. Residual fringes as given in these interferograms are mostly caused by inaccurate satellite orbit data. The subject matter of orbit residual fringes is already known, based on our experiences of SM data processing. 4.2.1. Satellite Orbit correction Figure 2. Exemplary ScanSAR results Fig. 2 shows exemplary the three few years spanning interferograms and the two monthly interferograms. In almost all interferograms a certain number of fringes is clear recognizable. At least for the monthly pairs such fringes are not expectable assuming no deformation took place. For the three 2-3 years spanning interferograms the amount of fringes is not reasonable We used the SARscape software papckage (www.sarmap.ch), where the so-called baseline fit procedure provides us the possibility to correct the satellite orbit. Using self-specified ground control points (gcp) and an interferogram corresponding SRTM DEM a least square adjustment of the observed (by InSAR) gcp elevations and the given SRTM elevations can be performed. As result of baseline fit procedure five satellite correction parameters are estimated and considered in a re-processing of the data. Fig. 3 depicts the original and the baseline fit corrected interferograms. The subtraction of the satellite orbit induced fringes is clearly visible in the corrected interferogram that uncovers the actual signal in the interferogram. Figure 3. Original interferogram (left) and satellite orbit corrected interferogram (right) 4.2.2. Recognized Signals Two signals are prominent in the corrected interferogram (Fig. 3). In the Southwestern part a bubble shaped signal is visible and above this a clear signal of about two fringes can be recognized. These highlight the Lazufre volcanic complex evolution, first monitored and published by [7]. This publication showed that no uplift deformation was detected 10 years ago. But afterwards InSAR could evidence a significant deformation for this area. This prominent example
demonstrates the potential of ScanSAR, because the wide coverage makes recognition of unknown deformation in active margin settings easier and faster. Based on this WS result an ENVISAT SM data study was initialised for detailed investigations in the time span between 2003 and 2006, in which we could deduce a significant increase of uplift deformation up to ~3cm/a compared to [7]. The deformation signal of Lazufre volcanic complex is affecting an area of about 1100km², that is comparable in size to super volcanoes such Yellowstone or Long Valley. The bubble shaped signal is directly located above the Salar de Atacam (salt lake) and is assumed as atmospheric/tropospheric contribution. Investigation on it considering predictable water vapour content and microwave path delay using MODIS data was accomplished in [11]. This result and the existence of the signal in all our interferograms, even in the monthly interferograms, support the assumption of atmospheric/tropospheric induced signal. Continuative investigation using MODIS data has to be done for final assessment of this signal. Further ongoing studies of all inteferograms will be used to recognize other (small) signals combined with detailed investigations using SM data. 5. MOTIVATION FOR ENHANCEMENT OF WS DATA ACQUISITION As mentioned in section 2. another interesting and important aspect of surface deformation in active subduction zones are the co- and interseismic related surface displacements. Here the interseismic deformation and its large-scale deformation pattern are up to now key questions and less understood. Unfortunately the current data amount of WS data in our study area is insufficiently to deduce any reliable results. Based on the existing dense SAGA GPS network and the pointwise known deformation [5] and the stable coherence over few years it is worthwhile to continue our study. Using potential future WS data acquisitions and the circumstance of existing master scene from 2003 enable us to generate long-time (3-6 years) interferograms. For this purpose it is highly recommended to enable future WS data acquisitions by ENVISAT, in the meaning of data itself, comparable geometry, small baselines and existing burst synchronisation related to the master scene taken on 2003 September 22. 6. ACKNOWLEDGEMENT We thank P. Pasquali and sarmap s.a. for supporting us regarding technical information about ScanSAR and data processing. The Research has been funded by the German Ministry of Education and Research (BMBF) and the German Research Foundation (DFG), grant 03G0594 (to J.A.). This is publication no. GEOTECH- 268 of the R&D-Program GEOTECHNOLOGIEN. 7. REFERENCES 1. Norabuena, E., Leffler-Griffin, L., Mao, A., Dixon, T., Stein, S., Sacks, I.S., Ocola, L. & Ellis, M. (1998). Space Geodetic Observations of Nazca-South America Convergence Across the Central Andes. Science. 279. 358-362 2. Barazangi, M., Isacks, B.L. (1976). Spatial distribution of earthquakes and subduction of the Nazca plate beneath South America. Geology. 4(11). 686 692 3. Pritchard, M., Simons, M., Rosen, P., Hensley, S. & Webb, F.H. (2002). Co-seismic slip from the 1995 July 30 Mw =8.1 Antofagasta, Chile, earthquake as constrained by InSAR and GPS observations. Geophysical Journal International. 150. 362-376 4. Klotz, J., Angermann, D., Michel, G.W., Porth, R., Reigber, C., Reinking, J., Viramonte, J., Perdomo, R., Rios, V.H., Barrientos, S., Barriga, R. & Cifuentes, O. (1999). GPSderived Deformation of the Central Andes Including the 1995 Antofagasta Mw = 8.0 Earthquake. Pure and Applied Geophysics. 154. 709-730 5. Klotz, J., Khazaradze, G., Angermann, D., Reigber, C., Perdomo, R. & Cifuentes, O. (2001). Earthquake cycle dominates contemporary crustal deformation in Central and Southern Andes. Earth and Planetary Science Letters. 193. 437-446 6. Khazaradze, G. & Klotz, J. (2003). Short- and longterm effects of GPS measured crustal deformation rates along the south central Andes. Journal of Geophysical Research. 108. 2289 7. Pritchard, M.E. & Simons, M. (2002). A satellite geodetic survey of large-scale deformation of volcanic centres in the central Andes. Nature. 418. 167-171 8. Chlieh, M., de Chabalier, J.B., Ruegg, J.C., Armijo, R.. Dmowska, R., Campos, J. & Feigl, K.L. (2004). Crustal deformation and fault slip during the seismic cycle in the North Chile subduction zone, from GPS and InSAR
observations. Geophysical Journal International. 158. 695-711 9. http://envisat.esa.int/dataproducts/asar/cntr1-1- 3.htm 10. A. Monti Guarnieri & Prati. ScanSAR focusing and interferometry. (1996). Geoscience and Remote Sensing, IEEE Transactions on. 34. 1029-1038 11. Holzner, J. (2004). Signal theory and processing for burst-mode and ScanSAR interferometry. Deutsches Zentrum fuer Luft- und Raumfahrt. Forschungsberichte. 2004-06. 224. 12. Pavez, A., Remy, D., Bonvalot, S., Diament, M., Gabalda, G., Froger, J., Julien, P., Legrand, D. & Moisset, D. (2006). Insight into ground deformations at Lascar volcano (Chile) from SAR interferometry, photogrammetry and GPS data: Implications on volcano dynamics and future space monitoring. Remote Sensing of Environment. 100. 307-320