DEM Generation Using SAR Interferometry

Contact Persons

David Small	Adrian Schubert
Christophe Magnard	Erich Meier

Keywords

SAR, InSAR, Interferometry, Topography, DEM, DSM, ERS-1, ERS-2, SRTM

Quick Reference

Multiple synthetic aperature radar (SAR) images acquired from displaced vantage points provide the basis for calculating a digital elevation model (DEM) an imaged area. The images can be obtained either simultaneously with separate antennae, or a single antenna can image a landscape two or more times successively. In the latter case, if the landscape changes between image acquisitions, decorrelation noise is introduced, and the quality of the resulting elevation model deteriorates (loss of coherence). The difference between the local phase measurements for each pixel in an image is transformed first into a difference in distance between between the two acquisitions. Together with range and Doppler information, this allows triangulation into a georeferenced map geometry.

The European Space Agency's ERS-1 and ERS-2 satellites have provided a wealth of data that can be used for DEM generation. The tandem mission in 1995-96 was the first to produce systematic coverage over most of the Earth's land mass. SAR interferometry has also been applied using data from other satellites (Seasat, SIR-B, JERS-1, SIR-C, Radarsat-1, ENVISAT ASAR, ALOS PALSAR). The SRTM mission provides improved quality thanks to its single-pass dual antenna configuration. Future missions (TerraSAR-X, TanDEM-X, Radarsat-2) will further improve upon the data store available.

Aircraft with two radar antennae can systematically acquire an area. Since the two image acquisitions are simultaneous, loss of coherence is no longer a significant problem. However, motion compensation algorithms gain importance, as does mosaicing, as the swath width is smaller than that of satellite systems.

In general, SAR Interferometry processing requires an image registration step, interferogram formation, removal of systematic fringe trends, calculation of the interferometric coherence, adaptive filtering, phase unwrapping, generation of height model, and geocoding. Height models have been generated for test sites in Germany, Switzerland, and validated against available reference information.

Geographic area of interest

Test sites in Switzerland and Germany with high resolution reference DEM's have been used for validation purposes.

Some Results

Synthetic Aperture Radar Interferometry (InSAR) enables the production of digital elevation models over large areas.

Radar ranging and Doppler information, together with knowledge of the phase difference between (at least) two data acquisitions allow for a 3D reconstruction of the area imaged. For example, data from the ERS radar satellites has been used to make height maps for an area near Bonn, Germany. Detailed discussion of the height models and their validation is available.

Methods

The geometry of the spaceborne case is shown in the following figure.

The InSAR processing chain consists of image coregistration, calculation of the interferogram, systematic phase flattening, adaptive filtering, phase unwrapping, refinement of the image geometry, height model generation and geocoding, followed by optional validation based on a reference DEM or tiepoints.

Partnerships

Research has been conducted in cooperation with the following organizations:

Publications

SMALL D., MEIER E., NÜESCH D. [2002]:
Phase Noise Countermeasures for Synthetic Interferogram Generation,
Proc. of IGARSS 2002, Toronto, Canada, June 24-28, 2002, pp. 2546-2548.
SCHUBERT A., SMALL D., MEIER E., NÜESCH D. [2002]:
Robustness of Wavelet-Based Stereo Matching for Variable Acquisition Geometries Using Simulated SAR Images,
Proc. of IGARSS 2002, Toronto, Canada, June 24-28, 2002, pp. 2759-2761.
SMALL D., PASQUALI P., HOLECZ F., MEIER E., NÜESCH D. [1998]:
Experiences with Multiresolution and Multifrequency InSAR Height Model Generation,
Proc. of IEEE-IGARSS'98, Seattle, USA, July 6-10, 1998, pp. 2671-2673.
PASQUALI P., HOLECZ F., SMALL D., MICHEL T. [1997]:
Calibration and Classification of SIR-C Polarimetric and Interferometric SAR Data in Areas with Slope Variations,
Proc. of IEEE-IGARSS'97, Singapore, August 3-8, 1997, pp. 448-450.
SMALL D., NÜESCH D. [1996]:
Validation of Height Models from ERS Interferometry,
Proc. of ESA-FRINGE'96 Workshop, Zürich, Switzerland, Sept. 30 - Oct. 2, 1996.
SMALL D., PASQUALI P., FÜGLISTALER S. [1996]:
A Comparison of Phase to Height Conversion Methods for SAR Interferometry,
Proc. of IEEE-IGARSS'96, Lincoln, Nebraska, USA, pp. 342-344.
SMALL D., WERNER C., NÜESCH D. [1995]:
Geocoding and Validation of ERS-1 InSAR-derived Digital Elevation Models,
EARSeL Advances in Remote Sensing, Volume 4, Number 2, Oct. 1995, pp. 26-39, pp. I-II.
SMALL D., HOLECZ F., NÜESCH D. [1995]:
Combination of Ascending / Descending ERS-1 InSAR Data for Calibration and Validation,
Proc. of IEEE-IGARSS'95, Florence, Italy, pp. 553-555.
SMALL D., WEGMÜLLER U., MEIER E., NÜESCH D. [1994]:
Geocoded ERS-1 InSAR-derived Digital Terrain Information,
Proc. of CEOS SAR Calibration Workshop, Ann Arbor, Michigan, USA, pp. 184-190.
SMALL D., WERNER C., NÜESCH D. [1993]:
Baseline Modelling for ERS-1 SAR Interferometry,
Proc. of IEEE-IGARSS'93, Tokyo, Japan, pp. 1204-1206.
SMALL D., WERNER C., NÜESCH D. [1993]:
Registration of ERS-1 SLC Products for SAR Interferometry,
Proc. of Fourth GEOSAR Workshop, Loipersdorf, Austria, pp. 63-66.

Department of Geography

Research Projects