Agence spatialeCanadian Space CanadienneAgency CEOS Working Group on Calibration & Validation (WGCV) 27th Meeting, NPL, UK Page 1 Dr. Satish Srivastava

Agence spatiale Canadian SpaceCanadienne Agency

CEOS Working Group on Calibration & Validation (WGCV)

27th Meeting, NPL, UKPage 1

Dr. Satish Srivastava

Chair, CEOS WGCV SAR Subgroup

Canadian Space Agency

Report on CEOS WGCV SAR Subgroup Activities

Presented at

27th CEOS WGCV MeetingNational Physical Laboratory, UK

June 12 – 15, 2007




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.




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




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




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




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




Natural Calibration Sites

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




Recognized distributed target referenceSeasatSIR-BSIR-CERS-1ERS-2 ScatterometerJERS-1RADARSAT-1ENVISATRADARSAT-2

Well characterized site radiometrically

(late 80s early 90s)

IsotropicTemporal stabilitySpatial uniformity

Radiometric Calibration: Amazon Rainforest




Extraction of In-Flight Elevation Beam Pattern from Amazon Rainforest images (Antenna pattern correction off)

PR (inc) G2(el)

Beam EL1

RADARSAT-1 acquisitions to date: More than 1600

Comparison of measured pattern vs Calibrated Pattern

-16

-14

-12

-10

-8

-6

-4

-2

0

7 8 9 10 11 12 13 14 15 16 17 18 19 20

Elevation Angle (deg.)

Tw

o-W

ay

Ga

in (

dB

)




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




LANDSATW1

EL1

RADARSAT-1Measurement Campaign

• Started in January 2003

• Covering entire range of incidence angles (5 beams)

• 2 to 4 products of each beam every 24-day cycle

S7

-16

-14

-12

-10

-8

-6

-4

-2

0

2

4

6

5 10 15 20 25 30 35 40 45 50

S7

EL1 W1W2 W3

EL1 W1 W2 W3 S7

10° 49°




Deriving a BF reflectivity estimate

Linear fit of each gamma profile

Then averaged to obtain reflectivity estimate (inc) EL1

W1

W2W3 S7

Year-round Boreal Forest reflectivity

~ centred around Rainforest value (-6.5 dB)

dependence conforms to clutter models

Large variation of reflectivity levels

Extractions of (inc) from Boreal Forest images Assuming beams remain calibrated during acquisitions

Antenna Pattern Compensation ON




Piece-wise linear over each beam range

-4

-3

-2

-1

0

1

2

3

4

8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50

Incidence Angle (deg.)

EL1 W1 W3 S7

Summer period – normalized (dB)

W2

Seasonal Boreal Forest reflectivity




Deriving estimates seasonal reflectivity

Model ~ kinc over incidence

angle range of each tested beams

Assumptions• Seasonally stable• Linear anisotropy

Gam

ma

Lin

ear

Fit

(dB

)G

amm

a L

inea

r F

it (d

B)

Summer

Winter

EL1

W1

W2

W3

S7

EL1

W1

W2

W3 S7

Standard Deviation of

dependance

(dB/deg.)

'Summer' 'Winter'

EL1 .0943 .1789

W1 .0915 .1784

W2 .0919 .1978

W3 .0933 .1929

S7 .0915 .1953




Antenna Pattern Compensation

OFF

Elevation beam pattern extraction tests

Average peak-peak pattern difference (dB)

100% swath

'Summer' 'Winter' Rainforest

EL1 1.35 (1.26) 1.43 (1.24) 0.83

W1 0.99 1.21 0.68

S7 0.81 1.26 0.71

G2(el) PR / (inc)

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)

In analysis software, substitute constant Rainforest reflectivity with 'summer' or 'winter' reference

Mostly experimented with EL1, W1 and S7 so far, acquired in both reference periods

Derived a beam-wise linear reflectivity estimate inc for both reference periods




EL1 'Summer' example: 1.17 dB PO M0380288

-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

7 8 9 10 11 12 13 14 15 16 17 18 19 20

Elevation Angle (deg)

Tw

o-W

ay

Re

lati

ve

Ga

in (

dB

)




EL1 'Winter' example: 1.26 dB PO M0398005

-20

-18

-16

-14

-12

-10

-8

-6

-4

-2

0

7 8 9 10 11 12 13 14 15 16 17 18 19 20


Tw

o-W

ay

Re

lati

ve G

ain

(d

B)




W1 'Summer' example: 0.86 dB PO M0375998

-9

-8

-7

-6

-5

-4

-3

-2

-1

16 17 18 19 20 21 22 23 24 25 26 27 28 29


Tw

o-W

ay

Re

lati

ve G

ain

(d

B)




W1 'Winter' example: 1.33 dB PO M0404092

-9

-8

-7

-6

-5

-4

-3

-2

-1

16 17 18 19 20 21 22 23 24 25 26 27 28 29


Tw

o-w

ay

Re

lati

ve G

ain

(d

B)




S7 'Summer' example: 0.56 dB PO M0379237

0

1

2

3

4

5

6

7

38 38.5 39 39.5 40 40.5 41 41.5 42 42.5 43


Tw

o-W

ay

Re

lati

ve G

ain

(d

B)




S7 'Winter' example: 0.88 dB PO M0396745

0

1

2

3

4

5

6

7

38 38.5 39 39.5 40 40.5 41 41.5 42 42.5 43


Tw

o-W

ay

Re

lati

ve G

ain

(d

B)




Calibration Transponder Sites

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)

RADARSATTransponder

ENVISATTransponder

ENVISAT ASAR Image of Resolute




RADARSAT-1 Precision Transponders forImage Quality and Calibration Measurements

Resolute Bay

Ottawa

Fredericton

Prince Albert

SAR Impulse Response




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




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

Documents

Agence spatialeCanadian Space CanadienneAgency CEOS Working Group on Calibration & Validation (WGCV) 27th Meeting, NPL, UK Page 1 Dr. Satish Srivastava