TerraSAR InSAR Investigation of Active Crustal Deformation

Ling Lei and Roland Bürgmann


This project aims to utilize advanced analysis of Terra-SAR data to investigate the dynamics and interactions of solid Earth deformation processes, such as earthquakes and fault creep, and Earth surface processes, such as land subsidence and groundwater movements, in a densely populated, urban region, the San Francisco Bay Area. Results from TerraSAR data will be carefully compared and integrated with InSAR data from other spacecraft, including the ERS-1/2, Envisat, RADARSAT-1, and ALOS satellites, and will be integrated in a rigorous analysis and monitoring effort of active surface deformation in the region. Ongoing deformation imaging reveals a number of natural hazards, including elastic strain accumulation about seismologic faults, active landsliding, land subsidence and rebound, and settling of unconsolidated sediments that are highly susceptible to liquefaction. Up to now, we have ordered and received 20 TerraSAR-X Spotlight Single Look Complex (SLC) images and 4 Stripmap SLC images delivered by the German Aerospace Centre (DLR) and obtained a few primary preliminary results.

Data analysis and processing

The TerraSAR-X images were acquired over the San Francisco Bay Area, particularly around an area of active landsliding, coastal subsidence, and shallow Hayward fault creep near the city of Berkeley. Berkeley is situated between 37.45$^{\circ}$ and 38.00$^{\circ}$ latitude, and 237.30$^{\circ}$ and 238.00$^{\circ}$ longitude. The data acquisition interval is from November, 2008 to July, 2009. Four types of Spotlight images and one type of Stripmap image in time sequence were ordered and acquired: spot_012, spot_038, spot_049, spot_075, and strip_003, having different look angles and pass directions. The data acquired with the standard Spotlight mode are with 1m pixel resolution, 10 km wide swath, and HH polarization. A TerraSAR-X Spotlight scene covers approximately 10*5km$^{2}$. Actually, there are 122 spotlight beams called spot_001 to spot_122 and 8 different beams of Spotlight data over the Berkeley area. But the recommended performance beams of Spotlight images are spot_010 to spot_079. So we chose four recommended beams of all the available beams for the Bay Area. Details of the data set used are mentioned in Table 2.1. Access to the SAR data is via FTP about 10 days after acquisition date.

Table 2.1: Information on the data acquired
Date Time Polarization Polarization Incidence Pass Beam
    Mode Channels Angles Direction  
2008-11-26 14:23:42UTC Single HH 21.78 D spot_012
2008-12-18 14:23:40UTC Single HH 21.78 D spot_012
2009-01-09 14:23:38UTC Single HH 21.78 D spot_012
2009-02-22 14:23:38UTC Single HH 21.78 D spot_012
2009-04-07 14:23:38UTC Single HH 21.78 D spot_012
2008-11-18 2:00:28UTC Single HH 34.14 A spot_038
2009-01-12 2:00:24UTC Single HH 34.14 A spot_038
2000-02-25 2:00:24UTC Single HH 34.14 A spot_038
2009-04-10 2:00:20UTC Single HH 34.14 A spot_038
2009-07-18 2:00:29UTC Single HH 34.14 A spot_038
2008-11-21 14:15:09UTC Single HH 38.94 D spot_049
2008-12-24 14:15:08UTC Single HH 38.94 D spot_049
2009-02-28 14:15:07UTC Single HH 38.94 D spot_049
2009-04-13 14:15:02UTC Single HH 38.94 D spot_049
2009-07-21 14:15:11UTC Single HH 38.94 D spot_049
2008-11-23 2:09:01UTC Single HH 47.82 A spot_075
2008-12-26 2:08:59UTC Single HH 47.82 A spot_075
2009-03-02 2:08:56UTC Single HH 47.82 A spot_075
2009-04-15 2:08:53UTC Single HH 47.82 A spot_075
2009-07-23 2:09:02UTC Single HH 47.82 A spot_075
2009-05-10 14:23:39UTC Single HH 21.45 D strip_003
2009-05-21 14:23:43UTC Single HH 21.45 D strip_003
2009-07-26 14:23:39UTC Single HH 21.45 D strip_003
2009-08-06 14:23:43UTC Single HH 21.45 D strip_003

The data is supplied in TerraSAR-X standard SLC COSAR (COmplex SAR) format with orbital information in an Extensible Markup Language (XML) header. The file contains integer real-complex components with double sampling and calibration constants for values. The COSAR file contains all focused complex SAR data of one beam in a burst-by-burst order, together with sample validity and position annotation. Stripmap and Spotlight images consist of one burst in that sense. The huge variety of level 1b product types for TerraSAR-X (complex, detected, geocoded) requires product annotation in an extensible and dynamic format. I am using ROI_PAC to do the interferograms. But ROI_PAC was designed to process raw data rather than SLC images. So there are two main problems. One is that ROI_PAC does the azimuth spectrum filtering (cutting out the part of the azimuth spectrum that does not overlap) in the ``roi" program. So ``resamp_roi" has to be changed to do the azimuth spectrum filtering in the interferogram formations. The other is that there is a significant difference between the way that ROI_PAC and other SAR processors use the Doppler centroid in the SLC formation. ROI_PAC is set up to process the SLC in the original geometry, but the TerraSAR processor produces the SLC images in a deskewed geometry. The deskewing moves the data to adjust for the squint or angle between the radar line of sight (LOS) and projects it perpendicular to the orbit track. Another problem is the processing with the Spotlight data. In order to form coherent Spotlight interferograms from two SAR observations, the following two conditions must be met: the ground projected-range spectra of both observations must overlap, and the received Doppler frequency spectra must overlap as well. The nonstationary squint angle during spotlight imaging also causes a negative drift of the Doppler centroid frequency in the SAR data. So the linear Doppler drift in azimuth has to be removed and Doppler centroid set to zero before making interferograms. We are now still working on those problems and will hopefully get better results in the near future. Here are some preliminary results: a Stripmap interferogram (Figure 2.6) and a Spotlight interferogram (Figure 2.7). I also plotted the landslides areas around Berkeley Hills. There are four large, slow moving, deep-seated landslides. All the landslides extend through residential areas and move on the order of cm/year, each covering an area of roughly 0.25-1.00 km$^2$

Figure 2.6: Interferogram for Stripmap pair of $10^{th}$ May and $21^{st}$ May, 2009
\epsfig{file=lei09_1_1.eps, width=8cm}

Figure 2.7: Interferogram for Spotlight pair of $22^{nd}$ Feb. and $7^{th}$ April, 2009
\epsfig{file=lei09_1_2.eps, width=8cm}


We thank the German Aerospace Centre (DLR) for providing TerraSAR-X data for this project. We thank Paul Lundgren, Eric J. Fielding, and Paul Rosen for providing TerraSAR reading codes and beneficial discussions.


TerraSAR-X Ground Segment Basic Product Specification Document, TX-GS-DD-3302. TerraSAR-X Ground Segment Level 1b Product Format Specification, TX-GS-DD-3307. Bürgmann, R., E. Fielding, and J. Sukhatme, Slip along the Hayward fault, California, estimated from space-based synthetic aperture radar interferometry, Geology, 26, 559-562, 1998. Eineder, M., N. Adam, and R. Bamler, Spaceborne Spotlight SAR Interferometry With TerraSAR-X, IEEE Trans. Geosci. Remote Sens., 5, 1524-1535, 2009.

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail:
© 2007, The Regents of the University of California