Crossing Multiple Methods GSICS Annual Meeting 24-28th March 2014, Darmstadt, Germany Synergic Calibration Crossing Multiple Methods Bertrand Fougnie and CNES Calibration Center (CNES-DCT/MO and CNES-DCT/ME)

Crossing Multiple Methods Synergic Calibration Crossing Multiple Methods Context CNES Calibration Tool Box Outlines advantage / drawbacks or limitations Application for PARASOL end-of-life reCalibration Application in the OCR-Virtual Constellation context Application to SEVIRI

Context In the past years, several calibration methods were developed using natural targets various calibration methods, different approaches operational configuration now available In the past years, several sensors provided extensive calibration time series a large experience has been developed on the use of each method a feedback exists on the real advantage and limitations

CNES Calibration Tool Box (MUSCLE)

SADE/Muscle – The Operational Arsenal Several calibration methods are operational Desert, Rayleigh, Sunglint, Cloud-DCC, Antarctica, Moon Deserts Domes Sunglint Moon DCC Rayleigh

SADE/Muscle – The Operational Arsenal Muscle / SADE (Tools) (Database)

Outlines, strength, limitations Several calibration methods are available Desert, Rayleigh, Sunglint, Cloud-DCC, Antarctica, Moon Each target has its own behavior : Magnitude: from very dark to very bright Spectral shape : from white to very pronounced Angular signature : from nearly uniform to large BRDF Polarized properties : from non-polarized to nearly fully polarized Short-term stability : from variable to fully stable Long-term stability : from seasonal variable to fully stable So efficiency range … Indicative behavior of targets May sensitively vary with various parameters

Synergic : What does it mean ? So the observation is : Calibration methods are like “Bordeaux Wines” : every method is good but in fact, all the methods show limitations it is impossible to address all calibration/radiometric aspects (the so-called “system calibration”) with one single method Basic idea = develop the synergic use of several method in order to : take advantage of the complementarities of all method document the confidence from consistency between methods improve the “system calibration” when integrating various results assess radiometric aspects others that the absolute calibration “Indicative” cartography – range of efficiency for each method

Absolute cal. Rayleigh Scattering Interband cal. Desert sites The “Great Mixture” ! Absolute cal. Rayleigh Scattering Interband cal. Desert sites Field-of-view cal. Snow (domes) Temporal monitoring Clouds (DCC) Cross-calibration Sunglint Polarisation Moon Characterization Method

Applications to - PARASOL end-of-life reCalibration - Validation of MERIS calibration - MODIS-Aqua validation - SEVIRI evaluation

1/ PARASOL end-of-life reCalibration Recent example#1 = PARASOL : bidirectional polarimeter : ± 50° wide fov optic + 2D focal plane 1400x2000km² up to 16 viewing directions 9 channels, and 3 polarized (490, 670, and 865nm) a “3M” concept Multi-directional + Multi-spectral + Multi-polarization No on-board calibration device in-flight calibration only based on natural targets (Fougnie et al., IEEE-TGARS, 2007 + 2009) Multi-method Synergic Approach Launched in dec-2004, on A-train up to end-2009 currently drifting, End-of-life in September 2013 3MI follow-on serie of instrument is expected on Post-EPS (~2018) 2000 km along-track 1400 km cross-track Satellite PARASOL image zenith viewing angle ⇒ No on-board calibration device

1/ PARASOL end-of-life reCalibration Synergic calibration for the in field-of-view evolution Clouds suppose the reference band is stable (765nm) Desert (reference = POLDER1) : more noisy, validation of cloud results Combination of 765 from desert (small evolution) + interband from clouds Desert 865/765 670/765 565/670 490/765 765 Clouds combination

1/ PARASOL end-of-life reCalibration Synergy : calibration of the temporal monitoring Calibration versus month B565 = 0.11 B865 = 0.024 B1020 = 0.018 B765 = 0.01

1/ PARASOL end-of-life reCalibration Synergic calibration for the calibration adjustment Final adjustment = best compromise Confidence where consistency is observed : e.g. 865nm Investigation conducted when curious behavior are observed : 443nm due to a stray light problem 565nm over clouds due to ozone Remaining investigations to improve the results : e.g. versus MERIS

2/ Validation of MERIS calibration Validation of the monitoring Interband stability over DCC and desert : perfect stability, artifacts reference VIS DESERT DCC reference reference NIR

2/ Validation of MERIS calibration Validation of the interband behavior Interband over DCC and sunglint GLI (10 yrs) N=30,560 442 nm DCC (10 yrs) N=9,391 442 nm vs reflectance vs reflectance GLI (10 yrs) N=30,560 665 nm DCC (10 yrs) N=9,391 665 nm vs reflectance vs reflectance GLI (10 yrs) N=30,560 885 nm DCC (10 yrs) N=9,391 885 nm vs reflectance vs reflectance

2/ Validation of MERIS calibration Behavior within the field-of-view : Versus viewing angle DES (10 yrs) N=15,963 DES (10 yrs) N=15,963 442 nm 620 nm vs VZA vs VZA RAY (10 yrs) N=99,631 RAY (10 yrs) N=99,631 620 nm 442 nm vs # camera vs # camera DCC (10 yrs) N=9,371 DCC (10 yrs) N=9,371 620 nm 442 nm vs # camera vs # camera

2/ Validation of MERIS calibration Validation of the calibration through Synergy Agreement – validation within 1% Some light discrepancies : 1/ known limitation of the method 2/ unknown limitation of the method 3/ significant signature ? May be under the accuracy of individual method, but Synergy may help ! Compare to System Ocean Color Vicarious Calibration

3/ Aqua-MODIS validation Validation of the radiometric stability Absolute and Interband over Desert and Rayleigh Calibration stability Rayleigh Desert Interband stability Desert Rayleigh

3/ Aqua-MODIS validation Validation of the interband behavior for SWIR Interband over Sunglint and Desert GLI (Dec 2003 - N=7,861) DES (2003 - N=4,839)

3/ Aqua-MODIS validation Validation of the calibration through Synergy Agreement – validation within 1% for OC bands, some discrepancies for Land bands Some light discrepancies : 1/ known limitation of the method 2/ unknown limitation of the method 3/ significant signature ? May be under the accuracy of individual method, but Synergy may help ! Compare to System Ocean Color Vicarious Calibration Lunar Calibration

4/ Application to GEO and SEVIRI Same approach for GEO sensor with SEVIRI/MSG2 Rayleigh, Desert, Sunglint This preliminary result has to be developed Backscattering signature on Rayleigh scattering to be explained Angular dependency with sun azimuth angle on Desert to be investigated Cross-calibration with MODIS to be generated (here very preliminary results)

Final comments Cross different methods = a powerful diagnostic whatever the nominal calibration method a coverage of most of the system calibration aspects (absolute, angular, spectral, temporal, cross) A good consistency between multiple methods = a good confidence on the final calibration (nominal + validation) a “realistic” estimation of the final calibration accuracy (as if the theoretical evaluation of the error budget for the nominal cal method is excellent) no radiometric artifacts remain : such as non-linearity, straylight, offset, polarization, inside the field-of-view… Differences between results from methods = need investigation to understand why – radiometric artifact require a good knowledge of the instrumental characteristics The state of the art cannot be reached through only 1 method One calibration method has its own accuracy : e.g. 2%, 3%... we can try to go down this limitation by mixing results A good consistency between sensors AND for multiple methods = a very high level of confidence can be claimed