2012 IGARSS Conference 23-27 July 2012, Munich, Germany

2012 IGARSS Conference 23-27 July 2012, Munich, Germany

23 July 2012 1SST from NPP/VIIRS

SST from NPP/VIIRS

Sasha Ignatov, Peter Minnett, Bob Evans, Doug May,Pierre LeBorgne, Bob Arnone, Sid Jackson

P. Dash, X. Liang, J. Stroup, Y. Kihai,

B. Petrenko, J. Sapper, M. Bouali, K. Saha NOAA/NESDIS, CSU/CIRA, SSAI, Riverside

K. KilpatrickU. Miami/RSMAS

B. McKenzie, J.-F. CayulaNAVO

JPSS Program Scientist – Mitch Goldberg

NPP Project Scientist – Jim Gleason

JPSS Data, Products and Algorithms Team – Heather Kilcoyne, Janna Feeley, Bruce Gunther

VIIRS Sensor Data Records (SDR; L1b) Team – Changyong Cao, Frank DeLuccia

NESDIS/STAR JPSS Team – Ivan Csiszar, Laurie Rokke, Lihang Zhou

Acknowledgements


Outline

NPP, JPSS, VIIRS– NPP – The Suomi National Polar-orbiting Partnership – JPSS – Joint Polar Satellite System– VIIRS – Visible/Infrared Imager/Radiometer Suite

JPSS SST Team & Products– JPSS IDPS – Interface Data Processing Segment– NOAA ACSPO – Advanced Clear-Sky Processor for Oceans,

SEATEMP (NAVO), L2gen (U. Miami), OSI SAF (EUMETSAT)

ACSPO SST Imagery from VIIRS & MODIS

Monitoring and Validation of SSTs– ACSPO – IDPS Comparisons– AVHRR – MODIS – VIIRS SST Consistency

Monitoring and Validation of Radiances – Stability of VIIRS radiances– Consistency AVHRR MODIS


NPP – The Suomi National Polar-orbiting Partnership Successfully launched on 28 October 2011 Transition from NOAA/POES & NASA/EOS – to JPSS

JPSS – Joint Polar Satellite System US - European Cooperation US Contribution: VIIRS onboard NPP/JPSS in 1:30am/pm orbit.

Planned launches: JPSS-1 (2017), -2 (2022) European contribution: AVHRR onboard Metop in 9:30am/pm

orbit. Planned launches: Metop-B (Sep 2012), -C (2017?)

VIIRS – Visible/Infrared Imager/Radiometer Suite Builds on MODIS heritage: Multispectral, high spatial resolution,

high radiometric accuracy imagery Products: Raw Data Records (RDR; L1a); Sensor Data Records

(SDR; L1b); Environmental Data Records (EDR; L2) - e.g. SST

Definitions


Role / Responsibility

NOAA/POES and NASA/EOS

NPOESS JPSS

Platform, Launch Vehicle

Private Industry Private Industry Private Industry

Instruments Private Industry Private Industry Private Industry

Algorithms Government Private IndustryIn transition to Government

Data Products Government Private Industry IPO/NPOESS

Cal/Val Government Private Industry Government

Archival & Distribution

Government Government Government

JPSS Data Products:IDPS – Interface Data Processing Segment


Algorithms: IPO/NPOESS (Northrop Grumman Aerospace Systems) Operational Products: IPO/NPOESS (Raytheon) IDPS (RDRs, SDRs, EDRs)

ProductData

ProducerTeam

Member(s)Domain Heritage Format Archive

IDPS*IPO

NPOESSIgnatov

and TeamGlobal New Hdf5

NOAA CLASS

ACSPO* NOAA Ignatov GlobalAVHRR MODIS

Hdf4NOAA CLASS

L2gen U. MiamiEvans

MinnettGlobal

MO(Y)D28 AVHRR (PF)

Nc3 -

SEATEMP NAVO May Global AVHRRDEF

BUFRJPL

PO.DAAC

EUMETSAT OSI SAF Le BorgneNorth-

AtlanticAVHRR Nc3 OSI SAF

Ocean Color NRLBob

ArnoneLink to Ocean Color Cal/Val

IDPS NGASSid

JacksonLink to heritage IPO/NPOESS code

* Monitored in SQUAM www.star.nesdis.noaa.gov/sod/sst/squam/

JPSS SST Team & VIIRS Products


Current IDPS Daytime SST Algorithm


)1))(sec(()( 12113121121110 zTTaRSSTTTaTaaSST

a0, a1, a2, a3: Regression coefficientsT11: Brightness temperature at 11 mT12: Brightness temperature at 12 mRSST: Reference (First-Guess) SST (currently, NCEP GFS)z: Sensor View Zenith Angle

The algorithm is stratified by T11-T12 with a switch point at 0.8 K. For the range of 0.6 K to 1.0 K results from the two stratifications are linearly interpolated.

The regression coefficients are derived using match-ups with QCed in situ SSTs available from NOAA in situ Quality Monitor (iQuam) http://www.star.nesdis.noaa.gov/sod/sst/iquam/

Current IDPS Nighttime SST Algorithm


a0, a1, a2, a3: Regression coefficientsT3.7: Brightness temperature at 3.7 mT11: Brightness temperature at 11 mT12: Brightness temperature at 12 mRSST: Reference (First-guess) SST (currently, NCEP GFS)z: View Sensor Zenith Angle

There is no stratification for the nighttime algorithm

The regression coefficients are derived using match-ups with QCed in situ SSTs available from NOAA in situ Quality Monitor (iQuam) http://www.star.nesdis.noaa.gov/sod/sst/iquam/

)1)(sec()( 3127.321110 zaRSSTTTaTaaSST

Gulf of Mexico23 May 2012 – Night

Examples ACSPO VIIRS and MODIS SST Imagery


ACSPO_V2.10_NPP_VIIRS_2012-05-23_0730-0740_20120526.053954.hdf

NPP/VIIRS ACSPO SST 23 May 2012 0730-0740 UTC – Night – Swath Projection


ACSPO_V2.10_AQUA_MODIS_2012-05-23_0800-0805_20120527.093405.hdf

Aqua/MODIS ACSPO SST 23 May 20120800-0805 UTC – Night – Swath Projection


NPP/VIIRS ACSPO SST 23 May 2012 0730-0740 UTC – Night – Mapped onto 0.8km grid



Aqua/MODIS ACSPO SST 23 May 20120800-0805 UTC – Night – Mapped onto 0.8km grid



Great Britain and Ireland26 May 2012 – Day

Examples ACSPO VIIRS and MODIS SST Imagery


NPP/VIIRS ACSPO SST 26 May 2012 1250-1300 UTC – Day – Swath Projection



Aqua/MODIS ACSPO SST 26 May 20121300-1305 UTC – Day – Swath Projection



NPP/VIIRS ACSPO SST 26 May 2012 1250-1300 UTC – Day – Mapped onto 0.8km grid



Aqua/MODIS ACSPO SST 26 May 20121300-1305 UTC – Day – Mapped onto 0.8km grid



Wind Speed26 May 2012 – Day – 1° gridded NCEP GFS


ACSPO and IDPS SST Monitoring in SST Quality Monitor (SQUAM)

www.star.nesdis.noaa.gov/sod/sst/squam/


http://www.star.nesdis.noaa.gov/sod/sst/squam/

http://www.star.nesdis.noaa.gov/sod/sst/squam/

NIGHT: ACSPO L2 minus OSTIA L428 May 2012


• Deviation from Reference SST is flat & close to 0• Residual Cloud/Aerosol leakages seen as cold spots

NIGHT: IDPS L2 minus OSTIA L428 May 2012


• More Cloud leakages in IDPS than in ACSPO• “Limb Cooling” – due to SST equations/coefficients

NIGHT: ACSPO L2 minus OSTIA L428 May 2012


• Shape close to Gaussian • Domain & Performance Stats close to expected

NIGHT: IDPS L2 minus OSTIA L428 May 2012


• IDPS sample +29% larger compared to ACSPO • Shape less Gaussian (negative Skew / increased Kurt)• increased STDV/RSD & Larger fraction of outliers

NIGHT 28 May 2012 – Summary

NOBS (%ACSPO)

Min/ Max Mean/ STD Med/ RSD Skew/ Kurt

ACSPO 84.0M (100%) -9.4/ +13.1 +0.01/0.53 +0.01/0.43 +0.2/ +4.4

IDPS 108.5M (129%) -17.5/+25.5 +0.03/0.69 +0.05/0.48 -0.1/+15.3

ΔT = “VIIRS minus OSTIA” SST (expected ~0)

• IDPS SST domain +29% larger & Stats degraded, compared to ACSPO

• Gap between Conventional and Robust stats wider in IDPS - More outliers

Analyses are routinely performed in SQUAM.

Both IDPS and ACSPO products are being improved


NIGHT STD DEV wrt. OSTIA L4

• AVHRR & MODIS SSTs are consistent• ACSPO VIIRS is consistent with MODIS & AVHRR • IDPS VIIRS EDR shows larger STD DEV

Warm-Up Cool-Down Event

IDPS shows larger STD DEV


DAY STD DEV wrt. OSTIA L4

Warm-Up Cool-Down Event

IDPS shows much larger STD DEV


• AVHRR & MODIS SSTs are consistent• ACSPO VIIRS is consistent with MODIS & AVHRR • IDPS VIIRS EDR shows much larger STD DEV

VIIRS, MODIS, and AVHRR Radiance Monitoring in MICROS

www.star.nesdis.noaa.gov/sod/sst/micros/


Model minus Observation (“M-O”) Biases- M (Model) = Community Radiative Transfer Model (CRTM) simulated

TOA Brightness Temperatures (w/ Reynolds SST, GFS profiles as input)

- O (Observation) = Clear-Sky sensor (AVHRR, MODIS, VIIRS) BTs

Double Differences (“DD”) for Cross-Platform Consistency

- “M” used as a “Transfer Standard”

- DDs cancel out/minimize effect of systematic errors & instabilities in BTs arising from e.g.

- Errors/Instabilities in Reynolds SST & GFS

- Missing aerosol

- Possible systemic biases in CRTM

- Updates to ACSPO algorithm

M-O Biases and Double Differences (“DD”)

)]([)]([ OMREFOMSATREFSAT


Fully automated- Scripted/Croned- Back-up processing enabled

ACSPO: Identify clear-sky pixels (Fortran 95)

MICROS: Generate stats (IDL)

Post to web: Html/JS/JQuery/JQplot

4 AVHRR GACs (N-16, -18, -19, Metop-A)

1 AVHRR FRAC (Metop-A)

MODIS Terra MODIS Aqua

VIIRS/NPP

# Clear Ocean Pixels / Night ~3M × 4 ~45M ~35M × 2 ~90M

CPU Hours= ACSPO+MICROS (0.7 = 0.6+0.1) × 4 5 = 4+1 (5 = 4+1) × 2 13 = 10+3

MICROS: Set-Up, Data Volumes, Processing Time

# of clear-sky (night) ocean pixels and Processing Time for 24hrs of L1b data


Nighttime M-O Biases in MICROS-ACSPO @11µm

• Shape close to Gaussian

• M-O Bias close to zero, slightly positive (expected)

• VIIRS consistent with AVHRR; Terra consistent with Aqua

• 2 MODIS and AVHRR/VIIRS clusters shifted by 0.6K


Nighttime DD’s @3.7 µm (Ref=Metop-A GAC)

• All AVHRRs (except N16), Terra/MODIS, and VIIRS: consistent to within ±0.08K

• Aqua and Terra MODIS inconsistent by 0.3K : Calibration problem

• Aqua/MODIS out of family by 0.3K (CRTM coefficients in error)

• VIIRS Cal Change 7 Mar 2012 reset BT@M12 by +0.13K; In-family23 July 2012 32SST from NPP/VIIRS

Aqua: out of family by 0.3K

N16: unstable & out of family

VIIRS recalibration

Nighttime DD’s @11 µm (Ref=Metop-A GAC)

• All AVHRRs (except N16) and VIIRS are consistent to within ±0.05K

• Terra and Aqua/MODIS out of family by 0.6K (error in CRTM coefficients)

• Cal Change 7 Mar 2012: Reset BT@M15 by +0.14K; Now better in family



Terra & Aqua: out of family by 0.6K

VIIRS recalibration

Nighttime DD’s @12 µm (Ref=Metop-A GAC)

• All AVHRRs (except N16) and VIIRS are consistent to within ±0.08K

• Terra and Aqua/MODIS out of family by 0.3K (error in CRTM coefficients)

• VIIRS Cal Change 7 Mar 2012: BT@M15 +0.14K; Now better in family



Terra & Aqua: out of family by 0.3K

VIIRS recalibration

Nighttime DD’s SST (Ref=Metop-A GAC)

• All AVHRRs (except N16), MODISs and VIIRS are consistent to within ±0.13K

• VIIRS Cal Change 7 Mar 2012: SST +0.10K & Out of family

• New regression coefficients on 3 May 2012: SST -0.15K & back in family



VIIRS recalibration

New Reg. Coeff. used

Terra & Aqua SSTs are in family, due to use of empirical regression

Liang, Ignatov, Saha

Monitoring of IR Clear-sky Radiances over Oceans for SST from NPP VIIRS, Terra-Aqua MODIS, and NOAA-Metop AVHRR.

Poster Session “Ocean Temperature and Salinity”

Tuesday, 24 July @17:20-19:00

More on VIIRS – MODIS – AVHRR Radiance Consistency


NPP launch – 28 Oct 2011

Nadir door opens, VIS/NIR imagery acquired – 21 Nov 2011

Cryoradiator doors opens, IR imagery acquired – 18 Jan 2012

Thermal IR Cal updated – 7 Mar 2012

Sensor Data Records: Beta distributed via CLASS – June 2012

VIIRS Cloud Mask (VCM): Declared Beta – April 2012

(Performance for SST Remains suboptimal)

IDPS SST to be declared Beta VCM + 4..5 months

ACSPO SST Beta is available from NESDIS

NPP VIIRS & SST Milestones


VIIRS is a good instrument for SST Radiances: Stable & consistent with AVHRRs Imagery – comparable to or better than Aqua MODIS In ACSPO, we consider generating Level 1.5 intermediate product:

Destripe imagery & Fill in bow-tied deleted pixels

MODIS MODIS Terra-Aqua 3.7 µm inconsistent by 0.3K – working to fix All MODIS bands out of AVHRR/VIIRS family by 0.3-0.6K.

Sensor is fine, CRTM coefficients in error – working to fix

SST Products IDPS EDR – quite a bit of cloud leakage in SST. Work is underway

with the VIIRS Cloud Mask Team to fix, before declaring SST Beta ACSPO Beta is available SDRs available via CLASS – users can make their SST retrievals

Summary


2012 IGARSS Conference 23-27 July 2012, Munich, Germany

Documents

Transcript of 2012 IGARSS Conference 23-27 July 2012, Munich, Germany