NPP VIIRS Pre-Launch Performance
and SDR Validation
Frank De Luccia, The Aerospace Corporation, El Segundo, CA
Bruce Guenther, NOAA - Joint Polar Satellite System, Goddard Space Flight
Center, Greenbelt, MD
Chris Moeller, University of Wisconsin, Madison, WI,
Xiaoxiong Xiong and Robert Wolfe , NASA’s Goddard Space Flight Center,
Greenbelt, MD
2011 IEEE International Geoscience
and Remote Sensing Symposium (IGARSS)
24-29 July, Vancouver, Canada
Topics
• Instrument Background
• Pre-Launch Performance
– Comparisons to MODIS Aqua as reference
•
•
•
•
Spectral characteristics
Spatial characteristics
Radiometric sensitivity - SNR and NEdT
Polarization sensitivity
– Performance summary and issues
• SDR Validation
– Calibration and Validation (Cal/Val) team
– Cal/Val plans and task structure
– Principal activities per Cal/Val phase
• Summary
2
Instrument Background
Instrument Background
• NPP VIIRS manufactured by
Raytheon under
subcontract to Northrop
Grumman for NPOESS
– NPP instruments now under
JPSS program management
• VIIRS design draws on
MODIS heritage
– MODIS also built by
Raytheon (SBRS)
– On-board calibrators and
calibration strategy very
similar
VIIRS
MODIS
Orbit
824 km, sun-synchronous,
near-polar
705 km, sun-synchronous,
near-polar
Scan Rate
33.6 rpm, cross-track
20.3 rpm, cross-track
Swath
Dimensions
3000 km across track by
13 km along track at nadir
2330 km across track by 10
km along track at nadir
Size
1.3 m x 1.4 m x 0.9 m
1.0 m x 1.6 m x 1.0 m
Weight
263 kg
250 kg
Power
154 W
225 W
Data Rate
10.4 Mbps (peak)
11 Mbps (peak)
Quantization
12 bits
12 bits
Design Life
7 years
5 years
4
Instrument Background – cont’d
Separately Mounted Electronics Module
MODIS
Solar Diffuser
MODIS
Blackbody
MODIS derived
Solar Diffuser
Stability Monitor
(SDSM)
3-Mirror Anastigmat
(TMA) All reflective
Rotating telescope
4-Mirror Anastigmat
(FMA) All Reflective
Aft Optics Imager
Flat-panel Cryoradiator
Half-angle Mirror
Cold FPA
Dewar Assembly
5
Environmental Data Products (EDRs) Derived
from VIIRS Sensor Data Records (SDRs)
Product Group
EDRs/Other Products
Imagery
Imagery*
Surface
Temperatures
Sea Surface Temperature*
Ice Surface Temperature
Land Surface Temperature
Clouds
Cloud Base Height
Cloud Cover/Layers
Cloud Particle Size
Cloud Optical Thickness
Cloud Top Height
Cloud Top Pressure
Cloud Top Temperature
Cloud Mask**
Aerosols
Aerosol Optical Thickness
Aerosol Particle Size
Ocean Biosphere
Ocean Color/Chlorophyll
Land Biosphere
Vegetation Index
Vegetation Type
Albedo
Snow & Sea Ice
Snow Cover
Sea Ice Characterization
Fire
Fire Detection
Other
Soil Moisture
Net Heat Flux
Suspended Matter
Geolocation**
* Key Performance Parameter (KPP)
** Intermediate Product (not an EDR)
6
Pre-Launch Performance
VIIRS Spectral Characteristics
•
VIIRS has 22 bands from 0.4 m to 12.5 m
(8 dual gain)
–
–
–
7 moderate resolution, 2 imaging resolution
and 1 broadband day/night band in 0.4 to
1.0 m range
6 moderate resolution and 2 imaging
resolution bands in 1 to 5 m range
3 moderate resolution and 1 imaging
resolution bands in 5 to 15 m range
VIIRS and MODIS Spectral Coverage: 1 m to 5 m
MODIS - 1000 m
MODIS - 500 m
MODIS - 250 m
VIIRS - 750 m
VIIRS - 375 m
H2O
H2O
H2O
H2O, CO2
N2O
H 2O
H 2O
CO2, N2O
Atmospheric
Transmission
(Mid-latitude
Summer)
Wavelength (microns)
VIIRS and MODIS Spectral Coverage: 0.4 m to 1.0 m
VIIRS and MODIS Spectral Coverage: 5 m to 15 m
MODIS - 1000 m
MODIS - 1000 m
MODIS - 500 m
MODIS - 500 m
MODIS - 250 m
MODIS - 250 m
VIIRS - 750 m
VIIRS - 750 m
VIIRS - 375 m
VIIRS - 375 m
H2O
H2O
H2O
O3
H2O
CO2
Atmospheric
Transmission
O2
(Mid-latitude
Summer)
H2O
H2O
O2
Wavelength (microns)
Wavelength (microns)
8
VIIRS Spatial Characteristics
•
16 moderate resolution bands with 750
m footprint at nadir
5 imaging resolution bands with 375 m
footprint at nadir
1 DNB with 750 m footprint constant
across scan.
Pixel aggregation used to control
footprint growth across scan.
•
•
•
VIIRS and MODIS Along-Scan Spatial Sampling
Agg = 3 : 1
|SA| < 32 deg
742 m
1104 m
1600 m
262 m
630 m
786 m
1260 m
1600 m
SA = Scan Angle
Agg = Aggregation
VIIRS and MODIS Along-Track Spatial Sampling
2.5
7
VIIRS 375 M
Along-Track Ground Sampling Interval (km)
VIIRS 375 M
Along-Scan Ground Sampling Interval (km)
Agg = 1 : 1
45 deg < |SA| < 56 deg
Agg = 2 : 1
32 deg < |SA| < 45 deg
6
VIIRS 750 M
MODIS 250 M
5
MODIS 500 M
4
MODIS 1000 M
3
2
1
0
VIIRS 750 M
2
MODIS 250 M
MODIS 500 M
1.5
MODIS 1000 M
1
0.5
0
0
10
20
30
Scan Angle (deg)
40
50
60
0
10
20
30
40
50
60
Scan Angle (deg)
9
VIIRS Radiometric Sensitivity
Comparison of VIIRS and MODIS Aqua Reflected Solar Band SNR
2200
2000
1800
better
VIIRS SNR and NEdT are comparable or
superior to MODIS SNR and NEdT when
normalized to the same spatial scale and
the same
radiance level MODIS Substitute
VIIRS
1600
1400
1200
VIIRS Band Spectral Range (um) Nadir HSR (m)
MODIS Band(s)
Range
HSR
SNR
1000
DNB
0.500 - 0.900
M1
M2
0.402 - 0.422
0.436 - 0.454
750
750
M3
0.478 - 0.498
750
M4
0.545 - 0.565
750
4 or 12
I1
0.600 - 0.680
375
1
M5
0.662 - 0.682
750
13 or 14
8
9
3
10
15
2
0.405 - 0.420
0.438 - 0.448
0.459 - 0.479
0.483 - 0.493
0.545 - 0.565
0.546 - 0.556
0.620 - 0.670
0.662 - 0.672
0.673 - 0.683
0.743 - 0.753
0.841 - 0.876
0.862 - 0.877
0.841 - 0.876
SAME
1.360 - 1.390
1.628 - 1.652
1.628 - 1.652
2.105 - 2.155
3.660 - 3.840
SAME
1000
1000
500
1000
500
1000
250
1000
1000
1000
250
1000
250
500
1000
500
500
500
1000
1000
M6
I2
0.739 - 0.754
0.846 - 0.885
750
375
M7
M8
M9
I3
M10
M11
I4
M12
0.846 - 0.885
1.230 - 1.250
1.371 - 1.386
1.580 - 1.640
1.580 - 1.640
2.225 - 2.275
3.550 - 3.930
3.660 - 3.840
750
750
750
375
750
750
375
750
M13
3.973 - 4.128
750
21 or 22
3.929 - 3.989
3.929 - 3.989
1000
1000
M14
8.400 - 8.700
750
29
SAME
1000
M15
10.263 - 11.263
750
31
10.780 - 11.280
1000
I5
10.500 - 12.400
375
31 or 32
10.780 - 11.280
11.770 - 12.270
1000
1000
M16
11.538 - 12.488
750
32
11.770 - 12.270
1000
16 or 2
5
26
6
6
7
20
20
800
600
400
200
0
M1/8 M2/9 M3/3 M3/10 M4/4 M4/12 I1/1 M5/13 M5/14 M6/15 I2/2
MODIS Aqua PL
MODIS Aqua On-Orbit
M7/2 M7/16 M8/5 M9/26 I3/6 M10/6 M11/7
VIIRS PL
VIIRS PL Scaled
Comparison of VIIRS and MODIS Aqua Thermal Emissive Band NEdT
I4/20
M12/20
M13/21
M13/22
M14/29
M15/31
I5/31
I5/32
M16/32
1
better
•
0.1
NEdT (K)
0.01
0.001
Dual-gain Band
MODIS Aqua PL
MODIS Aqua On-Orbit
VIIRS PL
VIIRS PL Scaled
10
VIIRS Polarization Sensitivity
VIIRS VisNIR Polarization Sensitivity
4.5%
Polarization Sensitivity (%)
•
5.0%
3.5%
3.0%
VIIRS HAM A, - 8 deg SA
2.5%
VIIRS HAM B, -8 deg SA
2.0%
MODIS MS 1, 0 deg SA
1.5%
MODIS MS 2, 0 deg SA
1.0%
0.0%
0
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17
Detector Index
MODIS Aqua VisNIR Polarization Sensitivity
6%
6%
5%
4%
-55.5 deg
-45 deg
3%
-22 deg
-8 deg
+20 deg
2%
+45 deg
+55.5 deg
1%
0%
300
4.0%
0.5%
400
500
600
700
Band Center Wavelength (nm)
800
900
Max. Polarization Sensitivity (%)
•
Polarization Sensitivity vs. Detector Index
VIIRS polarization performance is
generally better than that of MODIS.
Unlike MODIS, NPP VIIRS polarization
sensitivity varies strongly with field angle
resulting in detector dependence.
VIIRS pre-launch characterization provides
high-quality, per-pixel characterization of
polarization sensitivity and phase for use
in data product processing.
Max. Polarization Sensitivity (%)
•
5%
4%
-45 deg
3%
-22.5 deg
0 deg
+22.5 deg
2%
+45 deg
1%
0%
300
400
500
600
700
800
900
1000
Band Center Wavelength (nm)
11
Performance Summary and Issues
• VIIRS performance is comparable or superior to that of MODIS Aqua
in corresponding bands in all key performance areas except
crosstalk and out-of-band spectral response.
– NPP VIIRS has optical crosstalk in the VisNIR (0.8% average) due to high
angle scattering in the spectral filters
• May impact Ocean Color/Chlorophyll and Aerosol data products.
– NPP VIIRS out-of-band (OOB) spectral response somewhat higher than
MODIS Aqua due to same scattering mechanism in filters
– Second and subsequent VIIRS flight units will have reduced optical
crosstalk and OOB response due to improved spectral filters.
• VIIRS polarization sensitivity is lower than that of MODIS Aqua but
is detector dependent
– Requires per-pixel correction in Ocean Color processing
– Banding effects in SDRs displayed as images
• In other key performance areas not addressed here, such as bandto-band registration, near-field response and stray light response,
VIIRS performance compares favorably with that of MODIS Aqua.
12
SDR Validation
VIIRS SDR Cal/Val Team
Organization
Key Personnel
NOAA/NESDIS Center for
Satellite Applications and
Research (STAR)
Changyong Cao - VIIRS SDR Cal/Val Management Lead
Fuzhong Weng – JPSS SDR Cal/Val Chair
Mark Liu, Tim Chang
The Aerospace Corporation
Frank De Luccia - VIIRS SDR Cal/Val Technical Lead
David Moyer, Scott Houchin, Kameron Rausch, Christopher
Florio, Jason Cardema, Evan Haas, Patrick Yuen, Allen
Raines, Jeff Lipeles, Aaron Myrick, Zaven Petrosyan
NASA GSFC
• NPP Instrument Characterization
Support Team (NICST)
• NPP Instrument Calibration and
Support Element (NICSE)
Jack Xiong and Kurt Thome - Instrument Scientists
Hassan Oudrari, Robert Wolfe, Robert Barnes, Gene Eplee,
Fred Patt
NICST: Jeff McIntire (Lead), Ning Lei, Thomas Schwarting,
Junqiang Sun, Alin Tolea, Shihyan Lee, Aisheng Wu
NICSE: Vincent Chiang (Lead), Mash Nishihama, Gary Lin
University of Wisconsin
Chris Moeller
MIT Lincoln Laboratory
Juliette Costa, Ed Bicknell
Northrop Grumman Airborne
Systems
Lushalan Liao
14
Evolution of VIIRS Calibration
and Validation Plans
• On-orbit cal/val tasks were defined in an intensive governmentcontractor team collaboration
– Cal/val strategies draw heavily on MODIS experience
– Inputs on specific activities distilled into 54 tasks
• VIIRS SDR Calibration Plan produced and peer reviewed in 2009
–
–
–
–
Comprehensive description of VIIRS calibration program
Describes team, team interactions, tasks, tools and ground truth resources
Maps cal/val activities into cal/val phases
Provides metrics for evaluating SDR product maturity
• VIIRS SDR Calibration/Validation Operations Concept (OPSCON)
Document developed in 2011
– Infrastructure support definition
– Data flow and process definition to support cal/val rehearsals
• Managerial leadership has transitioned to Changyong Cao of NOAA
STAR
– Roles and responsibilities have evolved to reflect increased participation
of NOAA STAR
15
Six Broad Categories
of On-Orbit Cal/Val Tasks
•
Functional Performance and Format Evaluation (FPF 1-7)
– FPF tasks involve evaluating instrument functions and verifying the correctness of data
formats. Performed early in the mission, and will not be repeated unless the instrument
suffers a catastrophic event.
•
Calibration System Evaluation (CSE 1-6)
– CSE tasks evaluate the performance of the onboard calibration system and update the
calibration algorithm databases accordingly.
•
Image Quality Evaluation (IMG 1-4)
– IMG tasks evaluate the quantitative and qualitative spatial performance characteristics
of the instrument.
•
Radiometric Evaluation (RAD 1-25)
– RAD tasks evaluate the radiometric performance of the data product algorithm.
Radiometric evaluation will include evaluation of spectral characteristics since changes
in these characteristics relative to the pre-launch baseline will mainly manifest
themselves as in-band radiometric errors.
•
Geolocation Evaluation (GEO 1-7)
– GEO tasks evaluate the geolocation accuracy of the data product.
•
Performance and Telemetry Trending (PTT 1-5)
– PTT tasks evaluate long-term changes in the performance of both the instrument and
the data product.
16
Pre-Launch Activities
• Continued analysis and refinement of performance baseline
based on test program
• Verification of at-launch SDR algorithm look-up tables (LUTs)
– Improvements where warranted
• Operational code review and error checking
• Cal/val tool development
– RDR and SDR readers, LUT readers/writers
– Custom tools for cal/val analysis tasks
• Practice and training in operation of VIIRS SDR operational
code
– Algorithm Development Library (ADL) version of code allows cal/val
team to modify and test LUT updates and potential algorithm
improvements
• Task rehearsals to demonstrate readiness for on-orbit cal/val
17
Early Orbit Checkout (EOC)
• Verify instrument operability and functionality
– Signal
– Noise
– Scan rate
• Compare signal and noise characteristics to pre-launch
baseline
• Verify RDR, SDR and intermediate product formats and
validity of content
• Early use of Solar Diffuser Stability Monitor
– Critical for beginning time series to trend change in Solar Diffuser
reflectance
18
Intensive Calibration & Validation (ICV)
• Radiance match-ups
–
–
–
–
Ground sites
Inter-satellite comparisons
Comparisons with Cross-Track Infrared Sounder (CrIS)
Aircraft underflights
• Geolocation match-ups
– Ground Control Points (GCPs) from Landsat
•
•
•
•
•
Feedback on SDR quality from EDR validation efforts
Maneuvers
Exploitation of lunar data
SDR algorithm tuning and LUT updates
Continued performance and telemetry trending
Objective is achievement of stable, “validated/calibrated” SDRs by end of ICV.
19
Long-Term Monitoring (LTM)
• Routine performance of reflective band off-line calibration
• Periodic measurement and trending of key instrument
characteristics
–
–
–
–
Operability
Noise/SNR
Gains
Critical temperatures, voltages, currents in telemetry
• Periodic repetition of ICV tasks to maintain calibration and
SDR quality
– Radiometry
– Geolocation
20
VIIRS Cal/Val Activities by Phase
Sensor
Characterization
Functional
Checkout
Performance &
Telemetry Trending
Baseline
Radiance
Match-Ups
Data
Inventory
Geolocation
SDR Algorithm
Initialization &
Update Capability
Performance &
Telemetry
Trending
SDR Algorithm
Tuning
SDR
Parameter
& LUT Updates
SDR
Parameter
& LUT Updates
Launch
Cal/Val Tool
Development
Pre-Launch
Phase
Geolocation
Performance &
Telemetry
Trending
RDR/SDR
Verification
SDR Cal/Val
Plan Development
Radiance
Match-Ups
Early Orbit Checkout
(EOC) Phase
Intensive Calibration
& Validation (ICV) Phase
L + 50 days
Long Term Monitoring
(LTM) Phase
L + 180 days
21
SDR Product Maturity Levels
• Beta
– Early release product, initial calibration applied, minimally validated and may still contain
significant errors
– Available to allow users to gain familiarity with data formats and parameters
– Product is not appropriate as the basis for quantitative scientific publications studies and
applications
• Provisional
– Product quality may not be optimal
– Incremental product improvements are still occurring as calibration parameters are
adjusted with sensor on-orbit characterization
– General research community is encouraged to participate in the QA and validation of the
product, but need to be aware that product validation and QA are ongoing
– Users are urged to contact NPP Cal/Val Team representatives prior to use of the data in
publications
• Validated/Calibrated
– On-orbit sensor performance characterized and calibration parameters adjusted
accordingly
– Ready for use by the Centrals, and in scientific publications
– There may be later improved versions
22
Inter-Satellite Comparisons
• Analysis of Simultaneous Nadir
Overpass (SNO) data from VIIRS and
other satellites will reveal relative
biases
• Off-nadir inter-satellite comparisons
will be exploited also, particularly to
evaluate Response vs Scan Angle
effects
• Leverages validation performed for
other satellite systems
• VIIRS/MODIS comparisons expected
to be particularly useful
– MODIS instruments well validated
– Many similar bands with similar
radiometric performance
MODIS Terra/AVHRR 11 m Band Comparison
(Xiong et al., CEOS-IVOS 2004)
23
CrIS-VIIRS Cal/Val
(heritage AIRS-MODIS)
Abundant matchups
on every Aqua orbit.
14 orbits per day
(>10E6 matchups/day)
Radiometric performance as
function of scene temperature
AIRS footprints overlain
on MODIS image. Filter
the sample to retain
spatially uniform scenes.
Radiometric performance as
function of scan mirror angle
24
Use of Maneuvers for VIIRS SDR Validation (1)
• S/C Maneuvers for VIIRS Calibration
– Scientific benefits and implementation strategies are based on
lessons and experience from heritage missions and sensors, such as
MODIS and SeaWiFS
– All 3 types of VIIRS maneuvers have been approved for the NPP
mission
• Roll Maneuvers
– Near monthly event with lunar phase angles at 55 degree and roll
angles less than -14 degrees
– Provide an accurate and long-term monitoring of VIIRS RSB
calibration stability and independent monitoring of SD degradation
• Yaw Maneuvers
– Once during initial S/C checkout; could repeat every 3 years
depending on the SD degradation rate
– Validate SD solar attenuation screen (SAS) and SDSM sun view screen
transmission as a function of solar illumination angles and assure
VIIRS SDR quality for the RSB
25
Use of Maneuvers for VIIRS SDR Validation (2)
• Pitch Maneuvers
– Once during initial S/C checkout; could repeat depending on scanangle dependent changes in TEB responses
– Validate sensor response versus scan angle (RVS) and assure VIIRS
SDR quality for the TEB
• Operation and Data Analysis Support
– Maneuver implementation working group established to coordinate
among different groups (and instruments)
– VIIRS maneuver data analysis tools developed primarily from MODIS
experience
26
Use of Lunar Data for VIIRS SDR Validation (1)
• Track Changes in Sensor Responses for VIIRS Reflective
Solar Bands
– Both MODIS and SeaWiFS have used lunar time series to track
changes in sensor responses; lunar viewing geometry corrections are
made using ROLO lunar model
• Provide Independent Monitoring of VIIRS Solar Diffuser
Degradation
– The angle of incidence (AOI) of VIIRS space view, through which the
lunar observations are made, is identical to the AOI of SD
observations
• Enable Calibration Inter-comparison with Other Sensors
(e.g. MODIS)
– This is similar to the effort made to inter-compare Terra and Aqua
MODIS, MODIS and SeaWiFS
27
Use of Lunar Data for VIIRS SDR Validation (2)
SeaWiFS Normalized Lunar Radiance
Terra and Aqua MODIS B1 Lunar Irradiance
28
VIIRS Geolocation Matchups
•
•
•
Example of first 90 days of control point
matchups from MODIS/Terra
Good (> 60% normalized cross correlation)
control point matchups per day: 260
Used to perform initial refinement of
instrument to spacecraft alignment and to
refine rotating telescope and half angle
mirror geometric parameters
Number of Control Point Matchups
Track Residuals (nadir adjusted)
Scan Residuals (nadir adjusted)
29
Aircraft Campaigns
• Aircraft underflights can provide high quality radiometric validation
across the VIIRS spectrum
– Direct observations of the integrated upwelling earth scene radiance in
nearly same time, space, and geometry as the on-orbit sensor
– NIST traceable uncertainties for some airborne instruments
• Approach based upon heritage validation of AIRS, MODIS, IASI.
• Preliminary plans for aircraft campaigns have been developed but
are currently “on hold” due to funding constraints
• Platforms: ER-2 and/or WB-57
• Instruments desired for VIIRS SDR validation
– Enhanced MODIS Airborne Simulator (EMAS)
• 50 channel Vis/IR spectrometer, 50 m resolution, 36 km swath
• Expected to include reflective solar band (RSB) Hyperspectral Imager (HSI)
– Scanning High-resolution Interferometer Sounder (SHIS)
• Scanning M/LWIR 0.5 cm-1 interferometer, 2 km resolution, 32 km swath
• NIST traceable calibration
– NPOESS Airborne Sounder Testbed (NAST-I)
• Scanning MWIR/LWIR interferometer, 2km resolution, 40 km swath
• NIST traceable calibration
– Cloud Physics Lidar (CPL)
• Micropulse dual polarization lidar, 15 m resolution, nadir only
30
Approach based upon
heritage efforts for AIRS
and MODIS
Histogram of radiometric matchups
11.01 um
Fly instrumented ER-2 along satellite track
Matching geometry of earth scene observations
MODIS on
Terra/Aqua
SHIS, etc.
on ER-2
q
705 km
20 km
31
Summary
• VIIRS calibration and validation plans draw heavily on MODIS
strategies and experience
– Similar data products
– Similar on-board calibrators
– Similar radiometric and geolocation retrieval approaches
• VIIRS pre-launch performance baseline very well established
– Extensive instrument characterization provided in test program
– Independent data analysis by multiple contractor and government
teams has provided highly reliable parameter values for SDR algorithm
LUTs
• Experienced, multi-organization calibration/validation team is
well prepared to execute VIIRS on-orbit validation
32