SAR Calibration Subgroup

Chairman's Report

The latest workshop of the SAR Calibration subgroup of the CEOS Working Group on Calibration and Validation (WGCV) was held at Ann Arbor, Michigan on September 28 and 29, 1994. The workshop was ably organized by Dr. Kamal Sarabandi of the University of Michigan and his staff. Their efforts are much appreciated. Attendance for this workshop was lower at 35 than for some previous meetings. This reflects a maturing in the subject matter ( a lot of the problems in SAR calibration have been solved) and the fact that our community was very well served with technical conferences this year (IGARSS and PIERS being particularly well-attended). The attendees at this latest workshop represent the 'hard-core' of people working in the field, i.e. the people who actually do the work of SAR calibration.

The workshop was organized into technical sessions and working group meetings over a two-day period. Technical sessions were held in which three or four 20-minute presentations were made in the following technical areas:

* Radarsat ScanSAR calibration

*. Spaceborne SAR calibration

* Polarimetric calibration

* SAR processor calibration

* Calibration devices

* Radiometric calibration

* Interferometric SAR

Two working group or discussion sessions were also held during which the following topics were addressed in more detail:

* Correction for ADC saturation in SAR images

The technical discussion on correction for ADC saturation in SAR images centered on a problem which has affected both ERS-1 and JERS-1 data products, namely saturation of the Analog-to-Digital converter on board the spacecraft. This can lead to a calibration error - underestimation of the radar backscatter from the data. A number of methods to correct for this problem in the image were debated. The conclusions reached by the subgroup were as follows:

1. The design of future SAR systems should incorporate features to minimize the occurrence of ADC saturation, such as Block Floating Point Quantizers, higher dynamic range ADC's, and Automatic Gain Control (AGC) systems.

2. When the problem does occur, the effects (i.e. calibration errors) should be made clear to potential users of the data and software made available to correct the problem, where possible.

* ScanSAR calibration problems

ScanSAR data, with it's very wide swaths and relatively low resolution, presents a tough technical challenge for the next year for those members of the subgroup involved in Radarsat calibration. a number of technical issues were raised during discussion, including:

3. How to use point targets to calibrate data with such low resolution (50-100m).

4. Uncertainty in beam shapes and beam pointing angles can lead to large calibration errors.

5. Over land, the altitude of the imaged site above sea level must be incorporated into the radiometric correction. Large errors (2-3dB or worse) may result if the terrain height changes significantly within an imaged scene, e.g. going from sea level on the coast to above 10,000 feet in many coastal mountain ranges. This is obviously not an issue for sea ice and ocean scenes.

6. Intercalibration of processors was regarded as a priority.

7. Variable noise levels will be present within ScanSAR images as a function of both along-track and across-track directions.

8. Geocoding of ScanSAR data has yet to be demonstrated.

9. The gamma-coefficient has been shown to be flat at both C- and X-band over a wide range of incidence angles for the Amazon rain forest.

10. During the recent SIR-C mission, data was collected in ScanSAR mode over a few sites. JPL agreed to provide copies of a sample of the raw data to members of the subgroup.

* Future directions for the subgroup

11. Following a presentation on the use of the World Wide Web to disseminate information, the subgroup decide to set up a World Wide Web server site. This will include copies of the chairman's report, plus lists of references, important papers, and e-mail addresses of sub-group members. The site will also contain a link to a Bulletin board for the subgroup members. JPL offered to set up and maintain this facility.

12. As for most technical groups, future communications between group members will be via e-mail.

13. We debated closing the subgroup down as a possible option. Given that there are still technical problems to solve and with Radarsat launch approaching, a vote decided by an overwhelming majority to continue with the subgroup's activities.

14. Activities over the next year will focus on Radarsat calibration. CCRS agreed to distribute the Radarsat ADRO to subgroup members to ensure that they gain access to the data.

15. Kamal Sarabandi is organizing a special session on SAR calibration at IGARSS '95 in Florence, Italy. All workshop attendees were invited to submit a paper.

16. The next meeting of the subgroup is planned for December 1995. The meeting will take the form of a two-day workshop held during the same week and at the same place as the December Working Group on Calibration and Validation meeting.

17. The subgroup's plan for the next five years is attached.

18. It was also recommended that the subgroup organize a special workshop for users at the first Radarsat Users Meeting sometime in early 1996.

CEOS WGCV SAR Calibration Subgroup

Five Year Plan:

1. Continue to promote technical interchange on SAR Calibration throughout the community via workshops, special journal issues and conference papers/presentations.

2. Promote joint field campaigns to deploy calibration equipment in support of SIR-C, ERS-2 and Radarsat.

3. Develop test sites for calibration of radar imaging sensors, e.g. at Edwards Air Force Base in California and Flevoland in the Netherlands.

4. Exchange data products between group members to verify calibration independently.

5. Ensure that radar image data products produced over the next five years are calibrated or stem from calibrated data.

6. Develop algorithms for removing terrain height effects on radar image calibration.

CEOS WGCV SAR Calibration Subgroup