VIIRS (Visible/Infrared Imaging Radiometer Suite) on Suomi ...€¦ · National Polar-orbiting Partnership (NPP) satellite was launched on October 28, 2011; In January 2012, NPP was

1/32

VIIRS (Visible/Infrared Imaging Radiometer Suite) on Suomi NPP (SNPP) satellite:

New capabilities for cloud and surface mapping in the ArcticAlexander P. Trishchenko

Canada Centre for Remote Sensing, NRCan, OttawaPh: 613- 995 578 E-mail: [email protected]

Environment Canada, November 23, 2012

2/32

Presentation outline

NPP/SUOMI satellite

Visible/Infrared Imaging Radiometer Suite – VIIRS• Spectral features• Spatial resolution• Examples of images and applications

VIIRS and CrIS (imaging and sounding) synergy

How to achieve high refresh rate over Polar Regions Highly Elliptical Orbit (HEO) vs Low Earth Orbiters (LEO)

Conclusions

3/32

NPP/SUOMI satelliteNational Polar-orbiting Partnership (NPP) satellite was launched on October 28, 2011;

In January 2012, NPP was re-named SUOMI after Professor Verner E. Suomi, a renowned American Scientist who is considered the father of satellite meteorology. NPP SNPP;

SNPP orbit: sun synch. (1:30pm), H=824km, VIIRS swath – 3040km

Suomi NPP carries 5 five key instruments• Advanced Technology Microwave Sounder (ATMS)

• Cross-track Infrared Sounder (CrIS)

• Ozone Mapping and Profiler Suite (OMPS)

• Clouds and the Earth's Radiant Energy System (CERES)

• Visible Infrared Imaging Radiometer Suite (VIIRS) – replaced AVHRR

On 02.21.2012 SNPP Direct Broadcast transmission was turned ON

SNPP is the first satellite of future US/NOAA Joint Polar Satellite System (JPSS) that will include JPSS-1 & 2 satellites.

4/32

VIIRS Spectral Bands

VIS

/NIR

SWIR

IR

22 bands• 14 solar• 7 thermal• 1 Day/Night Band• (DNB)

Band grouping:I-bands: 5 (375m)M-bands: 16 (750m)DNB: 1 (750m)

5/32

VIIRS bands

0.030.072340300Single1.60.7511.538 - 12.488SSTM16

Moderate resolution bands

Imaging bands

0.750.750.5 – 0.9Day/ Night BandDND

0.030.07343300Single1.60.7510.263 - 11.263SSTM15

0.060.091336270Single1.60.758.400 - 8.700Cloud TopM14

0.0420.107/0.423343/634300/380H/L1.60.753.973 - 4.128SST/Fire detectionM13

0.130.396353270Single1.60.753.660 - 3.840SSTM12

221031.80.12Single1.60.752.225 - 2.275Clouds/AerosolM11

55034271.27.3Single1.60.751.580 - 1.640Snow fractionM10

2278377.16Single1.60.751.371 - 1.386Ci cloud detectionM9

221741655.4Single1.60.751.230 - 1.250Cloud particle/ snow grain size

M8

435/631215/34029/3496.4/33.4H/L1.60.750.846 - 0.885Ocn color/AerosolM7

355199419.6Single1.60.750.739 - 0.754Atrm correctionM6

321/631242/36059/65110/68H/L1.60.750.662 - 0.682Ocn color/AerosolM5




578/974352/316135/61544.9/155H/L1.60.750.402 - 0.422Ocn color/AerosolM1

0.41.5340210Single0.80.37510.50 - 12.40Imagery cloudsI5

0.42.5353270Single0.80.3753.550 - 3.930Imagery cloudsI4

149672.57.3Single0.80.3751.580 - 1.640Snow/iceI3

25115034925Single0.80.3750.846 - 0.885Land Imagery/NDVII2

21411971822Single0.80.3750.600 - 0.680Vis Imagery/NDVII1

ObservedSNR/ NT

SpecsSNR/ NT

Max value

[W/m2/sr/m or K]

Typical value

[W/m2/sr/m or K]

GainEdgeNadir

Wave length (m)Primary parameterBand

Spatial resolution [km]

refle

ctiv

eem

issi

vere

flect

ive

emis

sive

Source: NOAA on May 1, 201212 bit quantization

6/32

VIIRS products and formatsAll VIIRS files are in HDF5 format (http://www.hdfgroup.org/HDF5/). Generally, no reprocessing after processing system updates, except possibly CDR – climate data recordsData types:

• RDR – Raw Data Records• SDR – Sensor Data Records - L1B analog• EDR – Environmental Data Records – products or enhanced imagery (level 2)

Data availability:• NOAA CLASS archive for VIIRS

http://www.class.ncdc.noaa.gov/saa/products/search?sub_id=0&datatype_family=VIIRS&submit.x=32&submit.y=8

• List of publicly released products for SNPP sensorshttp://www.class.ncdc.noaa.gov/release/data_available/npp/npp_public_data_table.htm

• SNPP ftp-rolling archive (90-days) at NOAA ftp://ftp-npp.class.ngdc.noaa.gov/• Near-real time ftp-access from U of Wisconsin/SSEC (Madison, USA)

ftp://ftp.ssec.wisc.edu/pub/eosdb/npp/

VIIRS SDR products:• GMODO, GIMGO – geolocation files (ellipsoid) –M-bands MO, I-bands IM• GMTCO, GITCO - geolocation files (Terrain COrrected) M- and I bands• SVM[01-16], SVI[01-05] – imagery for corresponding M- and I –bands

VIIRS EDR products• Currently limited to enhanced imagery, intermediate cloud mask and fire products

7/32

VIIRS data granularity

VIIRSSmall granule: 86 seconds of data6400 pixels x 1536 lines: I-bands3200 pixels x 768 lines: M-bands4064 pixels x 768 lines: DNB

Big granule: 4 small granulesshown on the left

344 sec (<6 min) or ~2000km6400 pixels x 6144 lines: I-bands3200 pixels x 3072 lines: M-bands 4064 pixels x 3072 lines: DNB

= = = = = = = = = = = = = = = = = = =MODIS L1B and L2 data are split into 5 min granules

Nov 10, 2012

8/32

VIIRS Native Swath Structure3-detector aggregation 2-detector aggregation 1-detector aggregation

Scan is obtained by 32 detectors in imaging bands (I-bands)16 detectors in moderate resolution bands (M-bands)

Scanlines are overlapped off-nadir due to bow-tie effect. Some edge pixels in each scanline are removed in 2-detector and 1-detector aggregation zones to save the data volume for satellite downlink

9/32

Spatial resolution

Source: M. Goldberg, NOAA and Northrup Grumman & Raytheon

Bands 1&2MODIS Bands 3-7

3 detectors aggregation

2 detector aggregation

1 detectoraggregation

VIIRS Swath -3040kmOrbit height - 824kmLocal solar time 1:30pm(ascending node)

10/32

Spatial resolution along the swath

VIIRS image quality (spatial resolution) is preserved much better at the swath edge than AVHRR and MODIS

May 10, 2012 18:04UT

EdgeCenterCenter

11/32

Ice over Hudson Bay from VIIRS imagery

Red - I3Green- I2Blue - I1

Imaging bands@ 375m

12/32

VIIRS DNB – Day/Night Band @ 750m constant spatial resolution

South East USA, Oct 8, 2012 7:50-7:56 Z

Houston

Dallas

Atlanta

Florida

13/32

Calibration trends of VIIRS VIS/NIR bands

Source: NOAA/NESDIS/STAR

I2, M7 are of significant concern due to rapid degradation, although SNR is still within specsReason is the tungsten oxide contamination of the VIIRS mirror.

14/32

Calibration trends of VIIRS SW bands


15/32

VIIRS BlackBody (BB) temperature

Source: NOAA/NESDIS/STARVIIRS BB ( sometimes called also as Internal Calibration Target - ICT) is stable to within 0.07K @ 292.7K

16/32

Noise (Digital Numbers – DN) in VIIRS Thermal Bands M14, 15 and 16


17/32

VIIRS Geometric Performance:Band-to-Band Registration Error

Source: R.Wolfe et al. , 2012 (NASA)

• VIIRS was designed to collocate 2x2 I-band pixels into one M-band pixel

• Preliminary evaluation conducted by NASA (R.Wolfe et al. 2012) concluded that all VIIRS spectral bands are collocated within 0.1 M-band sampling interval (i.e. 0.2 I-band sampling interval).

• Early on orbit performance is generally consistent with pre-launch TVac testing

2x2 I-band 1 M-band

18/32

VIIRS Geometric Performance:Absolute Geolocation Error

Source: R.Wolfe et al. , 2012 (NASA)

• VIIRS geolocation results are very good

• NASA VIIRS geolocation team is working to further improve accuracy by implementing more precise parameterization

19/32

VIIRS Cloud Mask Intermediate Product -IICMO

Input bands; I-bands 1,2,4,5; M-bands: 1,4,5,7,9,10,11,12,13,14,15,1648-bit (6-byte) cloud mask is produced as result of series of test

• Main output is a confidence level of cloud/clear-sky detection• 00 = Confident Clear• 01 = Probably Clear• 10 = Probably Cloudy• 11 = Confident Cloudy

Cloud shadow is also determined through geometric approachCloud mask contains also flags for snow/ice, dust/smoke/volcanic ash, fire, land/water, Ci, sun glint and others.Details of algorithm and outputs are described in Joint Polar Satellite System (JPSS) VIIRS Cloud Mask (VCM) Algorithm Theoretical Basis Document (ATBD) # 474-00033

20/32

Clouds over low density vegetationSouthern US. Nov 10, 2012

Bright surface decreases the quality of cloud detection in the VIIRS Cloud Mask Intermediate Product IICMO_npp

21/32

Clouds and snow detection: Eternal questionNorthern Canada, Nov 10, 2012

VIIRS IICMO_npp product is not completely off, there is a room for improvement

22/32

Clouds and snow detection: Another example

VIIRS IICMO_npp product is not completely off, there is a room for improvement

23/32

VIIRS potential for snow retrievals

Several VIIRS bands have good sensitivity to snow structure, especially SWIR bands M8,10,11

Overlaid curves show model snow albedo for various grain sizes

24/32

Example of VIIRS 1-day composite produced at CCRS

Completed (initial version)• VIIRS data ingestor from

HDF5 format (SDR)

• Image re-projection routine using gradient search

• Simplified cloud ID mask• Compositing routine

On-going activities

• Improved cloud mask (or ingestion of VCM)

• BRDF retrievals/correction• Atmospheric correction

Focus is on surface products:• Spectral reflectance• FPAR• Snow/Ice• LAI• Fires/Burnt area• Land cover disturbances

25/32

CrIS – CrossTrack Infrared Sounder

717

437

163

# of bins

0.625648.75-1096.25LW IR

1.251207.5-1752.5MW IR

2.52150-2555SW IR

Spacing [cm-1]Spectral range [cm-1]Band

2200km+/-500, 30 FOR3 FOV x 3FOV1.10 x1.10

Swath widthScan structureField of RegardFOR

Field of View FOV

1.10

1.10

14kmdiameter

FOV2200km = 30 FORs in 1 scan

26/32

Examples of CrIS Spectrum. Nov 10, 2012

During 2nd NOAA satellite sceince week, the NOAA/NESDIS/STAR stated that NET for CrIS is better that IASI and AIRS

Synergetic use of VIIRS and CrIS can enhance cloud property and atmospheric profile retrievals and improve the accuracy of atmospheric correction for land applications.

27/32

Can a LEO system achieve the final goal: Observe weather at any time over the Polar Regions ?The key atmospheric applications include a number dynamic phenomena

• Atmospheric Motion Vector (AMV) retrievals• Ice/snow monitoring• Volcanic ash, smoke and aerosol dynamics• Atmospheric stability indices and aircraft icing threat• Radiative fluxes and surface albedo• Forest fire monitoring

Typical GEO refresh rate for next generation systems:• 1-min for local rapid scan• 5-min to regional rapid scan• 15-min for entire Earth disc

Motion vectors are normally retrieved from image triplets=> 10-15 min refresh rate is optimalTherefore, continuous observations of a GEO type are required over the Polar Regions!

28/32

GEO LEO (4-sat)

HEO

Trishchenko and Garand, 2012

29/32

Monthly mean zonal coverage

Example for May, 2011 with daily statistics averaged over 1 month period

Single images Image pairs Image triplets

Significant reduction of statistics for image pairs relative to single images (30% - 50% or more), and image triplets relative image pairs for LEO constellations ( >50%). HEO statistics is not affected above 60degN, and very slightly reduced below 60degN. (Trishchenko and Garand, CJRS, 2012)

30/32

Temporal coverage for combined GEO and HEO systems

Trishchenko and Garand, 2012

31/32

Earth view from HEO (Three Apogee -TAP) orbit

MODIS blue-marble imagery generated by NASA was used

0h

2h4h

6h

14.50

15.3017.90

26.20

32/32

ConclusionsVIIRS (Visible/Infrared Imaging Radiometer Suite) on NPP/Suomi satellite opens a new era in operational meteorological satellite imaging

Data is becoming operational right now

VIIRS has great potential for cloud and surface mapping in the Arctic

VIIRS has wide swath (3040km) and high-quality imagery with less degradation of spatial resolution at large Viewing Zenith Angle than MODIS and AVHRR

To observe weather at any point and any time over the Polar Region one need to have 2 satellites on HEO orbit.

GEO+HEO = Continuous Weather Observations from Space

We would be happy collaborate with EC/CMC to share our tools, methods and experience with VIIRS

AcknowledgementsVIIRS data were acquired from the NOAA CLASS and the Space Science and Engineering Center (SSEC) University of Wisconsin-Madison

VIIRS calibration results were acquired from Satellite Integrated Calibration / Validation System (ICVS) developed by the NOAA/NESDIS/STAR

VIIRS geolocation accuracy assessment was provided by Robert Wolfe, NASA

VIIRS (Visible/Infrared Imaging Radiometer Suite) on Suomi ...€¦ · National Polar-orbiting Partnership (NPP) satellite was launched on October 28, 2011; In January 2012, NPP was

Documents