Dr. Satish Srivastava Chair, CEOS WGCV SAR Subgroup Canadian Space Agency

1. Report on CEOS WGCV SAR Subgroup Activities Presented at 27th CEOS WGCV Meeting National Physical Laboratory, UK June 12 – 15, 2007 Dr. Satish Srivastava Chair, CEOS WGCV SAR Subgroup Canadian Space Agency

2. CEOS WGCV SAR Subgroup (1) • Mission: to foster high-quality synthetic aperture radar imagery from airborne and space borne SAR systems through precision calibration in radiometry, phase, and geometry, and validation of high level products. • Objectives: • Act as a forum for international technical interchange on the evolving methodologies, techniques,and equipment of SAR data processing, calibration and validation, • To determine standard definitions and calibration-validation requirements for SAR systems. • To support changes in CEOS formats and user products as appropriate, • To facilitate international cooperative programs in the calibration and validation of SAR systems, • To educate the SAR community.

3. CEOS WGCV SAR Subgroup (2) • Action Plan: • Annual Workshop/Meeting • Set up standard CAL/VAL sites – inter-sensor comparison • Calibration specification, requirements and techniques for Polarimetry, Interferrometry, POLInSAR • Recent Annual Workshop/Meeting • 2006 – Hosted by University of Edinburgh in Edinburgh, UK • 2005 – Hosted jointly by DSTO and University of Adelaide in Adelaide, Australia • 2004 - Hosted by ESA in Ulm, Germany • 2003 – ASAR 2003 Workshop, hosted jointly by CSA and CEOS WGCV SAR Subgroup in Saint-Hubert, Canada • 2002 – Hosted by BNSC in London, UK

4. Next CEOS SAR Subgroup Meeting (1) • ASAR 2007 Workshop: hosted jointly by Canadian Space Agency and CEOS WGCV SAR Subgroup • To be held in Vancouver, Canada, Sept. 11 – 13, 2007 • Organizers: Canadian Space Agency, CEOS WGCV SAR Subgroup, Defence Research and Development Canada, Canada Centre for Remote Sensing and MacDonald, Detwiller & Associates (MDA) • Website – http://www.space.gc.ca/asc/eng/events/2007/asar.asp

5. Next CEOS SAR Subgroup Meeting (2) • ASAR 2007 WorkshopTimetable • First Call: February 15, 2007 • Abstract Submission: May 25, 2007 • Notification of Acceptance: June 8, 2007 • Registration: Now Open • Workshop: September 11 – 13, 2007 • Paper Submission: October 12, 2007 • Distribution of Proceedings: December 2007

6. Next CEOS SAR Subgroup Meeting (3) • ASAR 2007 Workshop Suggested Topics • Next Generation SAR Systems • New SAR Missions • Innovative SAR Concepts • SAR Hardware and on-board Processing • Polarimetry and Interferrometry • Signal Processing Techniques • SAR User Requirements • SAR Calibration • Validation • Emerging SAR Applications • SAR Data Formats • Calibration Targets • Calibration Performance of on-going Missions • RCS Models and Scatterometers

7. International Amazon Rainforest Site Data is routinely collected and analyzed for calibration monitoring of SAR satellites including RADARSAT-1 Radiometry of the site remains stable Canadian Boreal Forest Site Radiometric characterization completed at C-band using RADARSAT-1 data Site seasonally dependent Can be used complimentary to Amazon but with a reduced radiometric accuracy Natural Calibration Sites

8. Radiometric Calibration: Amazon Rainforest • Recognized distributed target reference • Seasat • SIR-B • SIR-C • ERS-1 • ERS-2 Scatterometer • JERS-1 • RADARSAT-1 • ENVISAT • RADARSAT-2 • Well characterized site radiometrically • (late 80s early 90s) Isotropic Temporal stability Spatial uniformity

9. Extraction of In-Flight Elevation Beam Pattern from Amazon Rainforest images (Antenna pattern correction off) • PR g(qinc) G2(qel) RADARSAT-1 acquisitions to date: More than 1600 Comparison of measured pattern vs Calibrated Pattern Beam EL1

10. Boreal Forest - Hearst Site • Northwestern Ontario landmass (Hudson Bay basin) • Boreal Forest-Barrens transition • boreal spruce, balsam fir, jack pine, poplar, birch, tamarack, cedar • Seasonal Variations

11. S7 EL1 LANDSAT W1 • RADARSAT-1 • Measurement Campaign • Started in January 2003 • Covering entire range of incidence angles (5 beams) • 2 to 4 products of each beam every 24-day cycle EL1 W1W2W3S7 49° 10° S7 W3 W2 W1 EL1

12. W2 W3 S7 Deriving a BF reflectivity estimate • Linear fitof each gamma profile • Then averaged to obtain reflectivity estimate g(qinc) EL1 W1 Antenna Pattern Compensation ON Year-round Boreal Forest reflectivity • Extractions of g(qinc) from Boreal Forest images • Assuming beams remain calibrated during acquisitions g ~ centred around Rainforest value (-6.5 dB) g dependence conforms to clutter models Large variation of reflectivity levels

13. Summer period – normalized (dB) S7 EL1 W1 W3 W2 Incidence Angle (deg.) Seasonal Boreal Forest reflectivity Piece-wise linear over each beam range

14. W3 W1 Gamma Linear Fit (dB) EL1 W2 S7 Summer W1 S7 W3 Gamma Linear Fit (dB) EL1 Winter W2 Deriving estimates seasonal reflectivity • Model • g~ kqincover incidence angle range of each tested beams • Assumptions • Seasonally stable • Linear anisotropy

15. G2(qel) PR/g(qinc) Antenna Pattern Compensation OFF Elevation beam pattern extraction tests Derived a beam-wise linear reflectivity estimate g(qinc), for both reference periods In analysis software, substitute constant Rainforest reflectivity with 'summer' or 'winter' g reference Mostly experimented with EL1, W1 and S7 so far, acquired in both reference periods • Procedure (as with Amazon imagery): • Extract beam pattern • Compare with calibrated pattern in Payload Parameters File • Measure radiometric deviation: peak-peak of pattern difference (dB)

16. EL1 'Summer' example: 1.17 dB PO M0380288

17. EL1 'Winter' example: 1.26 dB PO M0398005

18. W1 'Summer' example: 0.86 dB PO M0375998

19. W1 'Winter' example: 1.33 dB PO M0404092

20. S7 'Summer' example: 0.56 dB PO M0379237

21. S7 'Winter' example: 0.88 dB PO M0396745

22. Multi-Transponder Sites in Canada In Fall 2006, ESA relocated an ENVISAT ASAR Transponder in Resolute Bay in vicinity of a RADARSAT Transponder Both Transponders are used simultaneously by ENVISAT Another ENVISAT ASAR Transponder was recently relocated in Ottawa, again in vicinity of another RADARSAT Transponder Two potential sites in Canada for inter-sensor comparisons for C-band SARs (e.g., RADARSAT-1 and 2, ENVISAT) Calibration Transponder Sites ENVISAT ASAR Image of Resolute RADARSAT Transponder ENVISAT Transponder

23. RADARSAT-1 Precision Transponders forImage Quality and Calibration Measurements Resolute Bay Prince Albert Fredericton Ottawa SAR Impulse Response

24. Recommendations from SAR Subgroup (1) • CEOS SAR Subgroup Workshop/Meeting held at Edinburgh in October 2006 concluded with a set of following recommendations to WGCV: • In order to reduce the extent of the commissioning phase of space borne SARs, effort be invested as much as possible into pre-launch calibration (e.g. antennas, TR modules, etc) • Data providers should investigate new product types that provide complex data in which the noise has been removed

25. Recommendations from SAR Subgroup (2) • In order to facilitate the use of calibrated SAR data (and to avoid the problem of continuously changing data formats), the data providers should provide: • a reader on the data CD itself • a LUT of calibration conversion factors for each pixel across a swath to covert DNs to sigma nought, beta nought and gamma • CEOS should co-ordinate a central database to log the use, location and availability of corner reflectors, in order to increase the availability of calibration targets to other agencies for appropriate SAR sensors • Future SAR Missions should be capable of full polarimetric imaging