IEEE
Geoscience and Remote Sensing Society

Abbreviation: GRSS, S Code 29

SAR INTERFEROMETERY

I. Introduction

The first implementation of radar interferometry came in earth-based observations of Venus (Rogers and Ingalls, 1969). The first reported experiments to determine terrain elevation of the earth were by Graham in 1974. Ten years later, JPL began the work performing interferometric radar experiments on the airborne system Convair-990 and on the spaceborne systems Seasat and SIR-B. Since 1990 the interest in SAR interferometry has grown due to the impressive amount of data suitable for interferometry from ERS-1 and the many airborne systems available such as the JPL AIRSAR.

SAR interferometry can be implemented in two different configurations:

a) Across-track interferometry offers the capability to generate topographic maps utilizing the signals of two separated radar antennas in the across-track direction. Due to the distance between the antennas, which is also referred to as the baseline, the distance to a single point scatterer on the ground is slightly different. This slant range difference between the two antennas corresponds to a phase difference in the received signals of both antennas. The local terrain elevation above the geoid can be determined by a simple relationship between phase difference and height. In practice two different methods are applied to obtain an across-track baseline:

i) One pass interferometry: In this case both antennas are mounted rigidly to the same platform with a fixed baseline. An interferometric image is obtained after one pass of the platform.

ii) Two pass interferometry: Only one antenna is used, but the platform must pass the same terrain in the same direction along a trajectory parallel to the first one.

An extension of the two pass across-track interferometer is the Differential SAR Interferometer. Differential interferometry is used to detect very small elevation changes (on the order of 1 cm or less) while preserving the same ground range resolution and swath width as that of a conventional SAR interferometer. The basic idea is to use one antenna flown twice over the same scene. If the second flight exactly duplicates the trace of the first, the time dependent phase changes can be measured. There would be no phase changes between the images at all unless there was a physical change in the scene, such as ground swelling and buckling in fault zones, residual displacements from seismic events, etc. If the two passes are made from flight tracks that are separated, it is no longer possible to distinguish surface changes from the parallax caused by topography. However, a third image made at some other baseline may be used to remove the topography and to leave only the surface changes.

b) Along-track interferometry offers the capability to detect the motion of the illuminated targets. One application is the measurement of ocean surface currents. The SAR system must use two antennas, spaced along the radar velocity direction. The signals from each antenna are separately processed into two images, where the phase difference between the images is calculated. The phase difference of a resolution element is proportional to the radial distance moved by the resolution element in the time required for the rear antenna to move to the position formerly occupied by the forward antenna. Ocean surface currents less than 4 cm/s have been observed by Goldstein and Zebker at JPL.

II. Technology Status

a) Airborne Systems:

• AIRSAR, JPL, USA: along and across track interferometry in L and C-band. Also repeat pass interferometry in P and L band was carried out.
• Do-SAR, Dornier, Germany: along and across track interferometry in C-band.
• EMI-SAR, TUD, Danemark: repeat pass interferometry in C-band. One pass interferometry mode in C-band is under construction.
• Carabas, FOA, Sweden: repeat pass interferometry in UHF-band (10 - 90 MHz).
• SAR, DRA, England: across and along track interferometry in C-band.
• E-SAR, DLR, Germany: repeat pass interferometry in L-band. One pass interferometry mode in X-band is under construction.

b) Spaceborne Systems:

• SEASAT, JPL, USA: repeat pass interferometry in L-band.
• SIR-B, JPL, USA: repeat pass interferometry in L-band.
• ERS-1, ESA, Europe: repeat pass interferometry in C-band.
• J-ERS-1, NASDA, Japan: repeat pass interferometry in L-band.
• SIR-C/X-SAR, JPL/DARA/ASI, USA/Germany/Italy: repeat pass interferometry in L, C and X-band.
• ERS-1/ERS-2, ESA, Europe: repeat pass interferometry with two satellites in C-band (Tandem-Mission).
• Almaz, NPO, Russia: repeat pass interferometry in S-band.
• RadarSAT, Canada: repeat pass interferometry in C-band (to be launched).
• A-SAR/EnviSAT, ESA, Europe: repeat pass interferometry in C-band (under construction).
• SIR-C/X-SAR (SRL-3), JPL/DARA/ASI, USA/Germany/Italy: across track interferometry in C and X-band (planned to be launched in 1997). Only airborne systems are able to operate in the one pass interferometry configuration. The planned third launch of SIR-C/X-SAR in 1997 will be the first spaceborne mission with a one pass SAR interferometer system in C and X-band.

III. Application Areas

a) Ground Topography:

• ERS-1 measure the evaluation for digital elevation models with a grid spacing of about 50 m and a height accuracy of about 5 m. The most forested areas can not be evaluated due to the strong incoherence between the two passes. They are normally spaced by a time interval of 3, 6 or 35 days. Since May 1995 the time interval is reduced by 1 day due to the ERS-1/ERS-2 Tandem mission.
• AIRSAR and Do-SAR allow the evaluation of the digital elevation models with a grid spacing smaller than 10 m and a height accuracy around 1 m. No incoherence problems are present due to the one pass mode. The SAR derived digital elevation models (DEM) have a big impact in the field of the topography. They are replacing the stereo DEMs derived from optical systems.

b) Earth surface motion detection:

• Massonet (1993) shows the possibility of measuring the residual displacement caused by the earthquakes. The differential interferometry is clearly validated for long term survey of slow faults (typically 10 mm/year measured with ERS-1). Due to the motion errors of the aircraft it is very difficult to implement the differential interferometer in an operational way.

c) Land surface classification

• By carrying out repeat-pass interferometry, coherence maps and change detection of SAR images can be used to provide properties of land surfaces. Results using ERS-1 data show the capability of the classification of forest, open fields, urban areas and open water. Along track interferometry is mainly used for the measurements of ocean surface currents.

IV. User Requirements

a) World topography: scales between 1:25.000 and 1:1.000.000 with height accuracies between some centimeters and 100 m.

b) Earth surface motion: accuracies around 1 mm - 1 cm within a observation time up to 10 years.

c) Ocean surface current measurements: accuracies between 1 cm/s and 1 m/s.

d) Land surface classification: crop and forest classification (only possible with addition of radar polarimetry).

V. Recommendation for Further Activity

Studies on vulcanology and earth surface motion related to differential SAR interferometry will be continued. ERS-1 and ERS-2 will provide the data source for this research. Also RADARSAT and ENVISAT (A-SAR) will contribute for the two pass applications of the SAR interferometry.

A next generation of SAR satellites is also planned by some countries, which would also carry out along-track interferometry. These systems would allow the worldwide determination of ocean surface currents.

REFERENCES

Graham, L.C. (1974), "Synthetic interferometer radar for topographic mapping," Proc. IEEE, Vol. 62 (6), pp.763-768.

Goldstein, R.M. and H.A. Zebker (1987), "Interferometric radar measurement of ocean surface currents," Nature Vol. 328, pp. 707-709.

Massonnet, D. (1993), "Validation of ERS1 Interferometry at CNES," Proc. of the second ERS-1 Symposium, 11-14 Oktober 1993, Hamburg, Germany Vol.2, pp. 703-709.