Subsidence

1702

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 41, NO. 7, JULY 2003

JERS SAR Interferometry for Land Subsidence Monitoring Tazio Strozzi, Senior Member, IEEE, Urs Wegmüller, Senior Member, IEEE, Charles L. Werner, Member, IEEE, Andreas Wiesmann, Member, IEEE, and Volker Spreckels

Abstract—In this paper, the potential of L-band repeat-pass differential synthetic aperture radar (SAR) interferometry for land subsidence monitoring is evaluated using Japanese Earth Resources Satellite (JERS) SAR data. Bologna, Mexico City, and the Ruhrgebiet are selected as application sites representing slow to fast deformation velocities. The investigation includes feasibility aspects such as data availability, the temporal decorrelation over different landcover classes and the range of useful spatial baselines, an analysis of the achieved deformation accuracy, and considerations on the complementarity to European Remote Sensing satellite (ERS) SAR interferometry and leveling surveys. In spite of the rather limited data availability, land subsidence maps could be generated for the three selected application sites. In contrast to ERS C-band SAR data, JERS L-band interferometry permitted the retrieval of subsidence values over vegetated areas and forest when using interferograms of less than one year acquisition time interval and short baseline. In addition, the longer L-band wavelength was found to be superior in the case of large deformation gradients that lead to phase-unwrapping problems in C-band interferometry. Index Terms—Japanese Earth Resources Satellite (JERS), L-band, land subsidence, synthetic aperture radar (SAR) interferometry.

I. INTRODUCTION

L

AND SUBSIDENCE monitoring with differential synthetic aperture radar (SAR) interferometry [1] using data of the European Remote Sensing satellites 1 and 2 (ERS-1 and ERS-2) has reached operational readiness. Land subsidence maps have been generated in numerous cases for different surface deformation velocities and extents [2], [3]. In several cases, the subsidence maps were validated with leveling surveys indicating a very high accuracy of a few millimeters per year [2], [4]. This does not mean that all subsidence mapping problems are solved with SAR interferometry, but rather that the technique has a very good potential, that it has reached some robustness, and that our understanding is sufficient to evaluate its potential for new cases, i.e., to decide on the strategy to use, to assess the feasibility, to assess the expected processing effort and data costs, and to indicate an accuracy. In spite of this, it is important to keep in mind the limitations of the technique when applied to ERS SAR data. In most cases it Manuscript received March 28, 2002; revised August 22, 2002. T. Strozzi, U. Wegmüller, C. L. Werner, and A. Wiesmann are with Gamma Remote Sensing, 3074 Muri BE, Switzerland (e-mail: [email protected]). V. Spreckels is with the Departement of Engineering Surveys and Geo-Information, Deutsche Steinkohle AG (DSK), 46236 Recklinghausen-Hochlarmark, Germany. Digital Object Identifier 10.1109/TGRS.2003.813273

is not possible, for example, to generate a subsidence map with complete spatial coverage due to temporal decorrelation for certain surface types. In addition, a quantitative interpretation of the interferometric phase is often not possible in the case of high fringe rates resulting from large displacement gradients, due to phase-unwrapping problems. These limitations are closely related to the C-band frequency (5.3 GHz, 5.7-cm wavelength) of the ERS SAR. The use of a lower frequency may avoid some of these problems. L-band SAR (1.3 GHz, 23.5-cm wavelength) in particular is attractive for land subsidence mapping because of the expected lower temporal decorrelation over vegetated areas and the better measurements of large deformations. Therefore, it should achieve a more complete spatial coverage with displacement information. In this paper, the potential of L-band repeat-pass differential SAR interferometry for land subsidence monitoring is evaluated using SAR data of the Japanese Earth Resources Satellite 1 (JERS-1) [5]. The selected application sites of Mexico City, Bologna, and the Ruhrgebiet are characterized by different displacement velocities, extents of the subsiding area and landuse cover. In addition, for these sites subsidence maps derived from ERS SAR data and leveling surveys are available for comparison and validation. II. MEXICO CITY (MEXICO) Mexico City is built on highly compressible clays and because of strong groundwater extraction a total subsidence of more than 9 m has been observed over the last century [6]. Two independent ERS SAR differential interferograms in ascending and descending modes, both with an acquisition time interval of 139 days and perpendicular baselines of less than 30 m, were used to derive a subsidence map for the time period December 1995 to May 1996 in [7]. Consistent results were found with the two interferograms and their results averaged in Fig. 1(a). The observed maximum subsidence velocity was larger than 40 cm/year. The coverage with subsidence information derived from ERS SAR interferometry is limited to the urban area with missing information near the Texcoco Lake and to the south in the Chalco Plain [see Fig. 2(d)]. JERS SAR data suitable for mapping land subsidence in Mexico City were also found in the archive, but SAR processing with far-range extension [8] had to be performed in order to enlarge the coverage to the west of the city. A SAR image acquired on March 17, 1994 and five other images acquired every 44 days between April 3, 1996 and September 26, 1996

0196-2892/03$17.00 © 2003 IEEE

STROZZI et al.: JERS SAR INTERFEROMETRY FOR LAND SUBSIDENCE MONITORING

1703

(a)

(b)

(c)

(d)

Fig. 1. Mexico City. Subsidence rates in meters per year derived from (a) two ERS pairs between December 1995 and May 1996, (b) two JERS pairs between March 1994 and September 1996, (c) three JERS pairs between April 1996 and September 1996, and (d) combined ERS and JERS subsidence rates for 1996. Projection is UTM, zone 14.

were used to compute a series of interferograms with acquisition time intervals between 44–924 days and perpendicular baselines between 222–3689 m. Independently of landuse and acquisition time interval we found complete decorrelation for perpendicular baselines larger than around 2000 m (see Fig. 2), also by applying the spectral-shift filtering [1]. For acquisition time intervals larger than two years we found a visible phase signal only for urban areas near the city center but not to the south in the Chalco Plain. It has to be noticed that a coherence of 0.1 still results in a visible phase signal with JERS SAR interferometry. For acquisition time intervals of less than 176 days and baselines between around 1000–2000 m the phase signal over the urban area is visible but noisy. For baselines smaller than around 1000 m and acquisition time intervals of less than 176 days, high coherence

is observed over the urban areas and even over nonbuilt-up areas the phase signal is useful. Two interferograms formed with data acquired between March 1994 and September 1996 and three interferograms for the period between April 1996 and September 1996, all with baselines shorter than 1026 m, were selected for further analysis. The topographic phase component was estimated based on a digital elevation model (DEM) derived from an ERS-1/2 Tandem pair. For the geometric referencing between the JERS and ERS data, terrain corrected geocoding with a global DEM and a fine registration using an intensity cross-correlation method [9] were used. The terrain flattened JERS interferograms were unwrapped and stacking of the two interferograms acquired between March 1994 and September 1996 and of the three interferograms acquired between April 1996 and

1704

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 41, NO. 7, JULY 2003

(a)

(b)

(c)

(d) Fig. 2. Mexico City. JERS SAR coherence versus perpendicular baseline and acquisition time interval over selected areas. (a) Urban area near the city center. (b) Urban area in the Chalco Plain. (c) Nonbuilt-up area. (d) ERS SAR backscattering coefficient image with polygons used to compile the statistic.

September 1996 was applied to combine the individual results into two maps with reduced errors. Subsidence information derived from the interferograms with two years acquisition time interval [see Fig. 1(b)] is restricted to the urban area near the city center and is similar to that obtained with ERS SAR data in Fig. 1(a). L-band SAR interferometry for acquisition time intervals of less than 176 days [see Fig. 1(c)], on the other hand, extended subsidence information to less ur-

banized and more vegetated areas. The spatial coverage of subsidence information in Fig. 1(c) is mainly limited by the area where the ERS Tandem pair used to derive the DEM could be unwrapped. The L-band interferograms of less than 176 days acquisition time interval could be unwrapped without any particular difficulty, whereas for the ERS SAR interferograms with acquisition time intervals of 139 days and the JERS SAR interferograms with two years time interval, the fringe rate was in some areas very high and hard to be resolved. The mosaic of combined ERS and JERS results of Fig. 1(d) gives an impressive overview of the subsidence in Mexico City in 1996, with settlements of more than 50 cm/year in some areas. The location and magnitude of subsidence is also in general agreement with the results of leveling surveys and theoretical models [6]. III. BOLOGNA (ITALY) At Bologna, land subsidence is caused by ground-water exploitation for industrial, domestic and agricultural uses [10]. Maximum subsidence velocities of 6–8 cm/year were observed with precision leveling surveys during the time period 1987–1991. Using the interferogram stacking technique, we produced a subsidence map for the time period 1992 to 1993 [Fig. 3(a)] based on six ERS SAR scenes with acquisition time intervals between 140–385 days and perpendicular baselines of less than 68 m [11]. The subsidence information derived from ERS SAR interferograms of around one year acquisition time interval is restricted to built-up areas. A phase-unwrapping algorithm for sparse data [12] was used to retrieve subsidence values also for the suburbs of Bologna and the neighboring small towns. However, in comparison to leveling surveys, the spatial coverage with subsidence data outside Bologna is incomplete. Limited data availability strongly restricted the selection of JERS SAR data to map subsidence at Bologna. Only four scenes were acquired by JERS over Bologna, between July 24, 1993 and April 14, 1994, and only two interferograms with perpendicular baselines of less than 2000 m showed coherence (Fig. 4). The two interferograms have acquisition time intervals of 88 and 264 days. Considering that subsidence in Bologna does not exceed 8 cm/year, we expect large errors from the analysis of these data. However, these data were further processed in order to investigate the coherence over nonurban areas. An external DEM with a pixel size of around 200 m 200 m was used for terrain corrected geocoding and removal of the topographic phase component. The result with JERS SAR data [Fig. 3(b)] confirms the higher coherence of L-band interferometry in nonurban areas in comparison to C-band one. The subsidence signal in Fig. 3(b) is affected by large-scale errors due to baseline and atmospheric artifacts. Nevertheless, there is a general similarity in the location and magnitude of subsidence measured with JERS SAR interferometry and leveling surveys [10]. A more complete validation with leveling data was not performed, because of the unsatisfactory JERS SAR data availability. More JERS pairs with longer acquisition time intervals and short baselines would be required for an improved analysis at this site with relatively low deformation rates of a few centimeters per year.

STROZZI et al.: JERS SAR INTERFEROMETRY FOR LAND SUBSIDENCE MONITORING

1705

(a)

(b)

Fig. 3. Bologna. Subsidence rates in meters per year derived from (a) six ERS pairs between May 1992 and July 1993 and (b) two JERS pairs between July 1993 and April 1994. Projection is UTM, zone 32.

(a)

(b)

(c)

(d)

Fig. 4. Bologna. JERS coherence maps in SAR geometry, linear scale between 0.0 and 0.3. (a) 24.07.93–20.10.93, 88 days, 1008 m. (b) 24.07.93–14.04.94, 264 days, 1658 m. (c) 20.10.93–14.04.94, 176 days, 2643 m. Also shown in (d) is a JERS backscattering coefficient image.

IV. RUHRGEBIET (GERMANY) Underground coal mining in the German Ruhrgebiet causes significant surface movement. The mining company Deutsche Steinkohle AG (DSK) is legally obliged to assess the environmental impact of the excavations. Surface movement caused by mining is a very dynamic process with high spatial and temporal variability. For mining areas with high subsidence velocities, ERS interferometric pairs with acquisition time intervals

of only one or a few 35-day repeat cycles are most appropriate [13]. An example of a 70-day interferogram is shown in Fig. 5(b). For nonurbanized areas [see Fig. 5(a)], the ERS phase signal is noisy, but a clear subsidence signal can be identified in the urban area in the Southwest corner. Also shown are the active mine panels for the acquisition period, with yellow boxes indicating the mining activity up to the first acquisition date and cyan boxes

1706

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 41, NO. 7, JULY 2003

(a)

(b)

(c)

Fig. 5. Ruhrgebiet. (a) Color composite of ERS Tandem coherence (red), backscattering coefficient (green), and temporal variability of the backscattering coefficient (blue), where blue regions are generally water, green generally forest, orange generally agriculture, and yellow generally urban. (b) ERS-filtered interferogram, 07.07.96–15.09.96, 70 days, 161 m, with superimposed mining information. (c) JERS-filtered interferogram 24.06.96-20.09.96, 88 days, 502 m. Projection is Gauss–Krüger (TM), zone 2.

0

the mining works up to the second acquisition date. A major difficulty in the quantitative interpretation of the ERS interferograms was the phase unwrapping of the high fringe rates over ongoing excavation. Displacements of up to 6–8 cm in the observation period were reliably estimated. Higher deformation rates were not accurately caught resulting in a significant underestimation for settlements between 10 and 30 cm [13]. JERS SAR data selection for the Ruhrgebiet was strongly restricted by the few acquisitions found in the archive. Only seven scenes, resulting in five interferograms with baselines shorter than 2000 m and acquisition time intervals of less than 132 days, are available. Far-range extension [8] was applied to enlarge the study area and improve the coverage of active mining areas. Here, we report on interferograms with topographic and displacement phase terms not yet separated from each other. In the interferogram shown in Fig. 5(c), with a perpendicular baseline of 502 m and an acquisition time interval of 88 days, the coherence is high even for forests, resulting in a significantly increased spatial coverage of subsidence information in comparison to ERS SAR interferometry. For agricultural areas more decorrelation is observed. Three clear subsidence signals appear at positions where mining was ongoing. In all three cases the displacement causes a phase difference of only one fringe or less because of the reduced sensitivity of the long L-band wavelength. This allows us to unwrap the phase easily. The analysis of the other interferograms showed that 1) the coherence over urban areas [Fig. 6(a)] is good for all analyzed baselines up to 1350 m, 2) the coherence over forest [Fig. 6(b)] diminishes in particular with increasing baseline, but for interferograms of up to 1350 m it is still useful, and 3) the coherence over agricultural fields [Fig. 6(c)] diminishes with increasing time interval from 44–176 days. The available data were not sufficient to study seasonal effects over agricultural regions.

(a)

(b)

(c) Fig. 6. Ruhrgebiet. JERS SAR coherence versus perpendicular baseline and acquisition time interval over selected landuse classes. (a) Urban, (b) forest, and (c) agricultural.

V. DISCUSSION JERS SAR interferometry for land subsidence monitoring was applied to Mexico City, Bologna, and the Ruhrgebiet. In spite of the limited data availability, land subsidence maps for the three application sites could be generated and compared fa-

STROZZI et al.: JERS SAR INTERFEROMETRY FOR LAND SUBSIDENCE MONITORING

vorably with subsidence maps derived from ERS SAR interferometry and leveling surveys. Range extended SAR processing was applied in some of the examples to cover areas of interest of up to a few kilometers outside the normal swath. It was demonstrated that L-band interferograms of less than one year acquisition time interval and for perpendicular baselines up to 1000 m can be used to map subsidence for vegetated areas and forest. It was also found that high deformation gradients, which could not be resolved with ERS, can be resolved with L-band interferometry because of the reduced movement sensitivity at the longer wavelength. Such high gradients are, for example, often observed above active mining. Based on this we conclude that L-band differential SAR interferometry is particularly well suited for the measurement of large displacements cm/ km) and diswith strong deformation gradients (e.g., placements in forested areas. This is complementary to ERS differential SAR interferometry, which is better suited for the measurement of slow displacements in urban areas. Obvious atmospheric artifacts were not found on the JERS interferograms. This was noticed not only for the reported interferograms over Mexico City, Bologna, and the Ruhrgebiet, but also for interferograms with very short baselines and acquisition time intervals over Verona (Italy) and Maracaibo (Venezuela). We found that the main processing problem in JERS SAR interferometry is the baseline estimation because of the inaccurate orbit information of the JERS-1 satellite. When possible, the baseline was estimated using ground-control points on stable areas. Otherwise, the baseline was estimated from the interferogram fringe rate. Future work will concentrate on the quantitative validation of the displacements in the Ruhrgebiet with ground truth and on the analysis of slow displacements of few centimeters per year with the stacking of multiple interferograms. In expectation of the L-band PALSAR system onboard the Japanese ALOS satellite scheduled for launch in 2004, we conclude that L-band differential SAR interferometry has a very high potential for subsidence monitoring provided that data are regularly acquired in a single interferometric mode with small enough baselines.

1707

[5] Y. Nemoto, H. Nishino, M. Ono, H. Mizutamari, K. Nishikawa, and K. Tanaka, “Japanese Earth Resources Satellite-1 synthetic aperture radar,” Proc. IEEE, vol. 79, pp. 800–809, June 1991. [6] G. F. Vega, “Subsidence of the city of Mexico; a historical review,” in Proc. Anaheim Symp., 1976, pp. 35–38. [7] T. Strozzi and U. Wegmüller, “Land subsidence in Mexico city mapped by ERS differential SAR interferometry,” in Proc. IGARSS, Hamburg, Germany, June 28–July 2 1999. [8] C. Werner, U. Wegmüller, T. Strozzi, and A. Wiesmann, “Gamma SAR and interferometric processing software,” in Proc. ERS-ENVISAT Symp., Gothenburg, Sweden, Oct. 16–20, 2000. [9] U. Wegmüller, C. Werner, T. Strozzi, and A. Wiesmann, “Automated and precise image registration procedures,” in Proc. MultiTemp 2001 Conf., Trento, Italy, Sept. 13–14, 2001. [10] M. Barbarella, L. Pieri, and P. Russo, “Studio dell’abbassamento del suolo nel territorio bolognese mediante livallazioni ripetute: Analisi dei movimenti e considerazioni statistiche,” in INARCOS, 1990, pp. 1–19. [11] T. Strozzi, U. Wegmüller, and G. Bitelli, “Differential SAR interferometry for land subsidence mapping in Bologna,” in Land Subsidence—Vol. II—Proc. 6th Int. Symp. Land Subsidence, Ravenna, Italy, Sept. 24–29, 2000, pp. 187–192. [12] M. Costantini and P. Rosen, “A generalized phase unwrapping approach for sparse data,” in Proc. IGARSS, Hamburg, Germany, June 28–July 2 1999. [13] V. Spreckels, U. Wegmüller, T. Strozzi, J. Musiedlak, and H. C. Wichlacz, “Detection and observation of underground coal mining-induced surface deformation with differential SAR interferometry,” in Proc. Joint Workshop of ISPRS Working Groups I/2, I/5, and IV/7 “High Resolution Mapping from Space”, Hannover, Germany, Sept. 19–21, 2001.

Tazio Strozzi (M’98–SM’03) received the M.S. and Ph.D. degrees from the University of Bern, Switzerland, both in physics, in 1993 and 1996, respectively. He has been with Gamma Remote Sensing, Muri, Switzerland, since 1996, where he is responsible for the development of radar remote sensing applications and is manager of a number of research and commercial projects. From 1996 to 1998, he was a part-time Physics Teacher at the Highschool of Bellinzona, Bellinzona, Switzerland. From 1999 to 2001, he worked as a part-time Visiting Scientist at the University of Wales, Swansea, U.K. He is Principal Investigator for ERS, ENVISAT, and JERS projects on forest mapping and subsidence monitoring. His current activities include SAR and SAR interferometry for landuse applications (including forest, urban areas, and hazard mapping) and differential SAR interferometry for subsidence monitoring, glacier motion estimation, and landslide surveying.

ACKNOWLEDGMENT JERS SAR data courtesy of J-2RI-001 and ALOS-RA-94, copyright National Space Development Agency of Japan. ERS SAR data copyright European Space Agency (ESA). ERS SAR data analysis supported by the ESA Data User Programme. ERS SAR data of Mexico City courtesy of A03-178. Mining information courtesy DSK. The Italian National Geologic Survey is acknowledged for the DEM of Bologna. REFERENCES [1] R. Bamler and P. Hartl, “Synthetic aperture radar interferometry,” Inv. Prob., no. 14, pp. R1–R54, 1998. [2] T. Strozzi, U. Wegmüller, L. Tosi, G. Bitelli, and V. Spreckels, “Land subsidence monitoring with differential SAR interferometry,” Photogramm. Eng. Remote Sens., vol. 67, no. 11, pp. 1261–1270, Nov. 2001. [3] A. Ferretti, C. Prati, and F. Rocca, “Permanent scatterers in SAR interferometry,” IEEE Trans. Geosci. Remote Sensing, vol. 39, pp. 8–20, Jan. 2001. [4] M. Crosetto, C. C. Tscherning, B. Crippa, and M. Castillo, “Subsidence monitoring using SAR interferometry: Reduction of the atmospheric effects using stochastic filtering,” Geophys. Res. Lett., vol. 29, May 9, 2002.

Urs Wegmüller (M’94–SM’03) received the M.S. and Ph.D. degrees from the University of Bern, Bern, Switzerland, both in physics, in 1986 and 1990, respectively. Between 1991 and 1992, he was a Visiting Scientist at the Jet Propulsion Laboratory, California University of Technology, Pasadena, working on the retrieval of canopy parameters from microwave remote sensing data. Between 1993 and 1995, at the University of Zürich, Zürich, Switzerland, his research included interferometric data processing, theoretical modeling of scattering in forest canopies, and retrieval algorithm development for geo- and biophysical parameters using SAR interferometry. In 1995, he was a founding member of GAMMA Remote Sensing AG, Muri, Switzerland, which is active in the development of signal processing techniques and remote sensing applications. As CEO of GAMMA, he has overall responsibility for GAMMA’s activities. In addition, he is directly responsible for a number of research and commercial projects of GAMMA. His main involvement currently is in the development of applications and the definition and implementation of related services in land surface deformation mapping, hazard mapping, landuse mapping, and topographic mapping. He is Principal Investigator for ERS, ENVISAT, SRTM, and ALOS announcement of opportunity projects on SAR and SAR interferometry-related calibration issues, application development, and demonstration.

1708

IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 41, NO. 7, JULY 2003

Charles L. Werner (S’75–M’79) received the Ph.D. degree in systems engineering from the University of Pennsylvania, Philadelphia, in 1987. He has been with the Jet Propulsion Laboratory, Pasadena, CA, where he worked on digital processing and applications of airborne and spaceborne SAR, especially in the area of interferometry. This has included research in focusing algorithms, speckle reduction, scatterer classification, polarization signatures, detection of moving targets, radiometric calibration, and interferometric signal processing. He developed the system design for the Cassini Titan Radar Mapper, a multimode radar incorporating SAR, altimeter, scatterometer, and radiometer modes. While working at the University of Zürich, Zürich, Switzerland (1992–1995), he developed a high-resolution interferometric processor for both airborne and spaceborne SAR sensors included ERS-1, and the Dornier DOSAR airborne system. In 1995, he and cofounded Gamma Remote Sensing AG, Muri, Switzerland, to further the use of SAR remote sensing, particularly interferometry, for a wide range of applications including generation of digital elevation models, studies of interferometric signatures of forest and agricultural areas, and differential interferometric measurement of crustal deformation and ice motion. He is currently with Gamma Remote Sensing AG, working on many aspects of SAR interferometric signal processing and applications.

Andreas Wiesmann (M’00) received the M.S. degree in physics and the Ph.D. degree in natural sciences, both from the University of Bern, Bern, Switzerland, in 1994 and 1998, respectively. In 1998, he joined Gamma Remote Sensing AG, Muri, Switzerland, as a Senior Research Scientist. His current work includes the development of SARand InSAR-based applications for earth observation.

Volker Spreckels received the Graduate Engineer (Dipl.-Ing.) degree in geodesy from the University of Hannover, Hannover, Germany, in 1995. Between 1995 and 1996, he was a freelancer at PHOENICS, Service-Oriented Society for Digital Photogrammetry and GIS Ltd., Hannover, Germany. His main projects were the production of orthoimages, 3-D building data, and digital elevation models for the computer-based planning of cellular radio networks for private GSM network operators. From 1996 to 1997, he was employed as Photogrammetry-Engineer at Kirchner & Wolf Consult, Ltd., Hildesheim, Germany, where his task was the generation of DEM and the production of digital orthoimages. Between 1997 and 2002, he worked as a Research Assistant at the Institute for Photogrammetry and GeoInformation (IPI), University of Hannover, in four successive research and development projects set up by the German hard coal mining company Deutsche Steinkohle AG (DSK), Recklinghausen, Germany. The aim of the projects was to investigate the capability of different space- and airborne systems for the detection and monitoring of subsidence movements caused by underground hard coal mining activities. The projects dealt with DEMs derived from analytical and digital photogrammetry, three-line-scanner imagery (HRSC-A), airborne laser- (TopoSys) and radar data (AeS-1), and the use of differential SAR interferometry for the detection of relative subsidence movements. Since September 2002, he has been the Group Manager for Photogrammetry and Remote Sensing in the Department of Engineering Surveys and Geo-Information at DSK.