Page 1
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.1HySICSResults
RadiometricBalloon-FlightResultsfromtheHySICS
CLARREOScienceDefinitionTeamMeeting
GregKopp,P.Smith,C.Belting,Z.Castleman,G.Drake,J.Espejo,K.Heuerman
LaboratoryforAtmosphericandSpacePhysics,Univ.ofColorado,Boulder,CO
Page 2
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.2HySICSResults
HyperSpectralImagerforClimateScience
Flights: Sept. 29, 2013 and 18 Aug. 2014Observations: 8 hours at float
Principal Investigator: Greg Kopp
HySICS is an instrument intended to acquire extremely accurate radiometric images of the Earth relative to incident sunlight
Questionstobeanswered:
Climatedatabenchmarktechnique
demonstration
HySICSisaballoonpayloadthatismounted
ontheWallopsArcSecondPointer(WASP)
flyingoutofFt.Sumner,NM
Ratio ofsolarincomingtooutgoing
radiancesbenchmarksclimate
withoutrelianceonaccuracy
Page 3
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.3HySICSResults
PI: Greg Kopp / LASP
CoIs: Co-I – Peter Pilewskie / LASPBalloon Flight Manager – David Stuchlik / WFF
Build and flight test a hyperspectral imager with improved radiometric accuracies for climate science• 350-2300 nm with single FPA to reduce cost & mass• <0.2% (k=1) radiometric accuracy• <8 nm spectral resolution• 0.5 km (from LEO) IFOV and >100 km FOV• <0.13% (k=1) instrumental polarization sensitivity
Perform two high-altitude balloon flights to demonstrate solar cross-calibration approach and to acquire sample Earth and lunar radiances
Single HgCdTe FPA covers full shortwave spectral range with reduced mass, cost, volume, and complexity
Incorporate solar cross-calibration approaches demonstrated on prior IIP to provide on-orbit radiometric accuracy and stability tracking
Orthogonal configuration reduces polarization sensitivityNo-cost balloon flights from experienced team at NASA
WFF demonstrate on-orbit capabilities
Objective
Approach
HySICS to demonstrate climate science radiometric accuracies in shortwave spectral region
HyperSpectralImagerforClimateScience
LunarReconstruction
GroundReconstruction
SolarDataCube
Page 4
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.4HySICSResults
Aperture(s)
Slit
Fold
Mirror
Grating
FPA
VacuumWindow
andOrderSorting
FilterFilter(s)
HySICSInstrumentOptics
ItemsGuidingDesign:
• Highaltitudeballoonflightenvironment
• Pushbroom imagingspectrometer
• Precisionaperturestopinfrontofthe
telescope
• FPAoperatesat150K
• Lowpolarizationsensitivity
Parameter Design Requirement
SpatialResolution 2.5arcmin
FieldofView(crosstrack) 10°
IFOV 0.02°
WavelengthRange 350-2300nm
WavelengthResolution 6nm, constant,Nyquist
Aperture 0.5,10,20mmdiameter
Page 5
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.5HySICSResults
Aperture(s)
Slit
Fold
Mirror
Grating
FPA
VacuumWindow
andOrderSorting
FilterFilter(s)
RadiometricEfficiencyCalibratedOn-Orbit
Page 6
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.6HySICSResults
ImproveRadiometricAccuraciesinVisible/NIR
• Currentinstrumentshave
>2%radiometricaccuracy
– Accuracyandstabilityrelyon
groundcalibrations,on-board
lamps,cross-calibrations,solar
diffusers,orlunarobservations
Page 7
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.7HySICSResults
SolarCross-CalibrationsRequire~10-5 Attenuation
Page 8
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.8HySICSResults
AttenuationMethodsUtilizedbyHySICS
• Apertureattenuation – Reductionof
inputlight-collectingarea
– Canachieveattenuations~10-3
– Limitedbydiffraction
• Integration-timeattenuation –
Reductionoflight-collectingtime
– Canachieveattenuations~10-3
– Limitedbylinearity
• Filterattenuation – spectralfilters
calibratedwithon-orbitlunarviews
– Canachieveattenuations10-1
– LimitedbyS/N
Allattenuationmethodsarerelativemeasurements;directmeasurementsofsolarorEarthirradiancesnotrequired.
EarthViewing SolarViewing
Hyper-
spectral
Imager
FPA
Hyper-
spectral
Imager
FPA
InputAperture
Filter
time
Exposuretime
time
Exposuretime
Integration
Time
Page 9
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.9HySICSResults
ScienceandCalibrationObservations
• GroundObservation
– Acquirehyperspectraldatafromgroundscenes
• SolarIrradianceMeasurement(Cross-Slit
Scan)
– Measurespectralsolarirradiancebyintegrating
imagesaftercross-slitscanofsolardisk
• Flat-FieldCalibration(Along-SlitScan)
– Scanslitsmoothlyalongdiameterofsolardisk
– Requirespointingaccuracyof~15arcsec
• CalibrationsusingMoon
– Filters:PlaceslitacrossMoonandacquire
measurementswithandwithoutfilters
– Flat-field:along-slitscanusinglargeaperture
• Drivesyetmorestringentpointingrequirements
Observationsnotpossiblethroughvariableatmosphere,soneed>30,000maltitude
Cross-SlitScan Along-SlitScan
FilterCalibration
GroundObservation
Page 10
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.10HySICSResults
AttenuationCalibrations – Apertures
• NIST-calibratedareasprovide
uncertainties<320ppm
2.0m
Hyper-
spectral
Imager
20mmdiameter 0.5mmdiameter
Actualsystemapertures
Aperture Diameter (mm)
Aperture Area (mm2)
Area Uncertainty (ppm) (k=1)
19.9862 313.72454 180.51542 0.20865 317
Butthere’salotmoretoit...
Page 11
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.11HySICSResults
AttenuationCalibrations – Filters
• Intendedattenuationsof10-0.9
• Lunaron-orbitrelativecalibration
– Successionoffilterin/filteroutradiance
measurements
– Trackpossibledegradation
– Lowlightlevel(comparedtosolar
irradiance)limitsattenuationsto~10-1
• Absorptiveglassfiltersspanspectrum
– Bulkeffectismorestablethanthinfilm
– Lowerreflectedlight
– Simplerdependenceonangleofincidence
Hyper-
spectral
Imager
FPA
InputAperture
Filter
LunarFilterCalibration
Page 12
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.12HySICSResults
AttenuationCalibrations – IntegrationTime
• FPAsystemnon-linearity
– Non-linearitiesarecharacterized
forattenuationsof10-3
• FPEnon-linearity
– 1ppmoverattenuationsof10-3.3
FullImageLinearityCorrectionUncertainty
LinearityResidualofPixel(100,100)
FPAlinearitygreatlyexceeded
expectations
Obviatesneedforfilters
>103 range
Page 13
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.13HySICSResults
HySICSIntegratedandReadyforLaunch
Page 14
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.14HySICSResults
BalloonInflation
Page 15
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.15HySICSResults
HySICSLaunch
putshortHySICSLaunchvideohere
Page 16
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.16HySICSResults
HySICSFlight#2Summary• FLIGHTNO.: 650N
• LAUNCHDATE/TIME: August18,2014,15:36Z
• LAUNCHSITE: FortSumner,NM
• BALLOONVOLUME: 0.8MCM(29.47MCF)
• BALLOONWEIGHT: 1,675KGS(3,693LBS)
• EXPERIMENTWEIGHT: 1,925KGS(4,244LBS)
• SUSPENDEDWEIGHT: 2,722KGS(6,000LBS)
• GROSSINFLATION: 4,836KGS(10,662LBS)
• FLOATALTITUDE: 37.19KM(122KFT)
• BALLOONTHICKNESS: 20.32MICRONS(0.80MIL)
• SERIALNO.: 90/CSBFNO.1205
• DISCIPLINE: WU/SolarandHeliospheric
• TOTALFLIGHTTIME: 8hrs,54min
Page 17
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.17HySICSResults
FloatAltitude
Declared
• Launch@09:35on18Aug.2014
• Ascenttime:2hours,17minutes
• Floataltitudeof122,000feet@11:52
HySICSFlight#2Path
Page 18
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.18HySICSResults
HySICSImagesfromFlight#1
• Flight#1(EngineeringFlight)demonstratedspatial/spectral
scanningcapabilityofallthreetargetsLunarReconstruction GroundReconstruction SolarDataCube
Page 19
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.19HySICSResults
GroundScanandDataCubefromFlight#1
Groundscan(13:25:30)
R=653nm,G=535nm,B=457nm
Page 20
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.20HySICSResults
HySICSGroundScansfromFlight#2• EachFlight#2groundscanacquiredin~5min.from4200images
Page 21
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.21HySICSResults
EarthLimbScansAcquiredfromHySICS
Page 22
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.22HySICSResults
LunarDataCubefromHySICSFlight#2
Page 23
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.23HySICSResults
HySICSCross-CalibrationFormalismSpatial/spectralground-images,Smeas_obj(l)[DNs],areconvertedtophysicalspectral-irradianceunits[W m-2 nm-1]byanon-orbit-
determinedunit-conversionfactor,C(l)[W m-2 nm-1 DN-1],andthe
radiance-attenuationfactor,A(l)[unitless],whichcorrectsfortheoptical-throughputandintegration-timesusedforsolar- vs.Earth-
viewing
whereSSI(l)representstheradianceoftheobservedsceneinSI-traceable,physicalunits andC(l)istheunit-conversionfactor
withSSI(l)beingtheSSI(providedbyanindependentspace-flightinstrumentorasolarmodel)andSmeas_Sun(l)theHySICS’sin-flightmeasurementoftheSSIinDNsacquiredbyspatially-integrated
cross-slitscansofthesolardisk.
SSI(l) = Smeas_obj(l) A(l) C(l)
C(l) = SSI(l) / Smeas_Sun(l)
Page 24
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.24HySICSResults
HySICSCross-CalibrationFormalismBeingameasurementofratiosacquiredoverashortperiodoftime,
theHySICSsolarcross-calibrationapproachdoesnotrelyonintrinsic
calibrationaccuraciesorlong-termstability
Allfollowinguncertaintiesarek=1unlessotherwisenoted
SSI(l) = SSI(l) · Smeas_obj(l) / Smeas_Sun(l) · A(l)
Page 25
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.25HySICSResults
FPACorrections&Uncertainties• Bad-Pixelremoval
• ReadNoise:determinedforeachpixel;8.3DNaverage
• DarkSignal:0.29DNforlongestintegrationtimes(negligible)
• ThermalBackground:scene- andtemperature-dependent
– Canalsocontributetoshotnoise
• Linearity:~0.30%�0.12%
• Gain:determinedforeachpixel;12.01�0.12e-/DNaverage
– <0.003%ofa15%full-scalesignal
• Flat-Field:Acquiredon-orbitbysolarandlunarscans
Severaluncertaintiescanbereducedbymultiple-imageorrepeated-scanobservationsofsource(ifstatic)
ApplicabletoSunandMoonbutnotgroundobservations
Page 26
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.26HySICSResults
Instrument-LevelUncertainties• ShotNoise:signal-dependent;dominantsourceofuncertainties
forgroundscenesacrossmuchofspectrum
• Diffraction&Scatter:measuredandmodeled;affectssolar
observationsatlongwavelengths
– UsingNIST-quoted10%uncertainties
• SpectralScale:Determinedon-orbit;scaleswithspectralgradient
– Rarelyadominantcontributortonetuncertainties(exc.solarinUV)
• BrightnessOffset:Often2nd largestuncertaintyforgroundscenes
• Polarization:Causedbygratingsensitivity
– Affectsradiometryofpolarizedscenes
Severaluncertaintiescanbereducedbymultiple-imageorrepeated-scanobservationsofsource(ifstatic)
ApplicabletoSunandMoonbutnotgroundobservations
Page 27
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.27HySICSResults
MeasurementUncertainties
SSI(l) = SSI(l) · Smeas_obj(l) / Smeas_Sun(l) · A(l)
seeKoppetal.2017forfinaluncertainties
solarground aperture
Page 28
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.28HySICSResults
AttenuationUncertainties – IntegrationTimeMethodprovidesgreaterattenuation
rangewithloweruncertaintiesthan
anticipated,contributinglittleto
attenuationuncertaintiesIntegration-TimeAttenuation-MethodUncertaintiesUncertaintyParameter BrightScene
(53%FS)[%]Max.Int.
(75%FS)[%]ElectronicLinearity 0.00016 0.00016GainNon-linearity 0.050 0.120Total 0.050 0.120
>103 range
Typicalbright
groundlevel
Page 29
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.29HySICSResults
AttenuationUncertainties – FilterRatio(Sun)• Filtercalibrationuncertainties0.04%- 0.1%from1000- 2300nm
– Dominatedbyread- andshot-noise,especiallyintheUV/Visibleregion
FilterRatioUncertainty:NG4#2 FilterRatioUncertainty:NG5#2 FilterRatioUncertainty:BG25
Filter-CalibrationAttenuation-MethodUncertaintiesParameter MeasurementUncertainty(%) MeasurementUncertainty(%) MeasurementUncertainty(%)
550nm 1000nm 2000nm 550nm 1000nm 2000nm 550nm 1000nm 2000nmFilter(SolarCalibration) NG4#2 NG5#2 BG25ShotNoise 0.94 0.1 0.047 0.25 0.065 0.035 NA 0.027 0.018ReadNoise 0.97 0.097 0.048 0.25 0.06 0.035 NA 0.015 0.011WavelengthBinLocation 0.027 0.015 0.009 0.027 0.015 0.009 0.027 0.015 0.009Filter-outUncertainty 0.35 0.049 0.064 0.35 0.049 0.064 0.35 0.049 0.064BackgroundLevelCorrection 0.23 0.023 0.011 0.082 0.02 0.011 NA 0.005 0.003BlackbodyRadiationCorrection 0.0001 0 0 0 0 0 NA 0 0DarkImageReadNoise 0.069 0.007 0.003 0.018 0.004 0.002 NA 0.001 0.0008DarkImageShotNoise 0.003 0.0003 0.0001 0.0007 0.0002 0 NA 0 0Total 1.416 0.150 0.094 0.505 0.104 0.082 NA 0.060 0.068
Page 30
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.30HySICSResults
• Filtercalibrationsandflat-fieldingfromMoonoflimitedsuccess
– Limitedbythelowlunarsignalfromthenarrowlunarcrescentandpointing
sensitivity
– Lunarfilterratiouncertainty1%- 2%from600- 2300nm
LunarCalibrationsLimitedatTimeofFlight
LunarPhase>90°
29/4/16, 3:48 PMGrapeEnvy Moon Phase Calendar
Page 1 of 1https://stardate.org/cgi-bin/mooncal/mooncal.cgi
Aug 2014
Sun Mon Tue Wed Thu Fri Sat
1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30
31
Page 31
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.31HySICSResults
HySICSEfficienciesCharacterizedOn-OrbitOn-orbitSI-calibrationobviates
needtotransferground
calibrationtospace
Thisisthewholepurposeofthe
solarcross-calibrationapproach
On-orbitcalibrationscome
frommeasurementsofratios,
soaccuracydoesnotrelyon
absolutecalibrationsoron
long-termstability
Page 32
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.32HySICSResults
EndResult– RadiometricGroundImage• Applyingspectralsolar
irradiancecalibrationsto
theHySICSdataenables
radiometrically-calibrated
datacubes
1233nm
Page 33
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.33HySICSResults
HySICSReferences• G.Kopp,P.Smith,C.Belting,Z.Castleman,G.Drake,J.Espejo,K.Heuerman,J.
Lanzi,andD.Stuchlik,“RadiometricflightresultsfromtheHyperSpectralImager
forClimateScience(HySICS),”GeoscientificInstrumentation,2017,inpress• Kopp,G.,Belting,C.,Castleman,Z.,Drake,G.,Espejo,J.,Heuerman,K.,
Lamprecht,B.,Lanzi,J.,Smith,P.,Stuchlik,D.,andVermeer,B.,“Firstresults
fromtheHyperSpectralImagerforClimateScience(HySICS),”Proc.SPIE 9088,AlgorithmsandTechnologiesforMultispectral,Hyperspectral,andUltraspectral
ImageryXX,90880Q,June13,2014,doi:10.1117/12.2053426
• Kopp,G.,Pilewskie,P.,Belting,C.,Castleman,Z.,Drake,G.,Espejo,J.,
Heuerman,K.,Lamprecht,B.,Smith,P.,andVermeer,B.,“RadiometricAbsolute
AccuracyImprovementsforImagingSpectrometrywithHySICS,”IGARSS2013,Melbourne,Australia,pp.3518-3521,July2013,978-1-4799-1114-1/13.
• Espejo,J.,Belting,C.,Drake,G.,Heuerman,K.,Kopp,G.,Lieber,A.,Smith,P.,and
Vermeer,B.,“AHyperspectralImagerforHighRadiometricAccuracyEarth
ClimateStudies”,SPIEProc.,21-25Aug.2011.• P.Smith,G.Drake,J.Espejo,K.Heuerman,andG.Kopp,“ASolarIrradiance
Cross-CalibrationMethodEnablingClimateStudiesRequiring0.2%Radiometric
Accuracies,”ESTF2011,June2011.
Page 34
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.34HySICSResults
HySICSAccomplishmentsSummary• Designed,built,andtestedahyperspectralimagerforEarthviewing
• Demonstratedflightsolarcross-calibrationtechniqueandquantifiedspectrally-
dependentuncertaintiesfor3attenuationmethodsviatwohigh-altitude
balloonflights
– Aperturearearatios10-3.2,integrationtiming10-3,filters10-0.9
– Netattenuationsof10-7.1 exceedIIP’s10-4.7 goal
– Filterattenuationmethodnotneededforspaceflightinstrument,savingmassandcomplexity
• Demonstrated~2ximprovementinradiometricaccuraciesinflight– Instrumentandcalibrationchangesexpectedtoprovideanother~2ximprovement
• AppliedSI-traceableradiometricscaletomeasuredradiances
• Demonstratedsinglefocal-plane-arrayspectrometerspansdesiredwavelengths
– Reducesmass,volume,power,andcostforspaceflightinstrument
• Improvedground-basedtestfacilityforcalibrationsofsolarattenuatorsystem
andquantificationsofuncertainties
ElevatedTRLfrom3to6
Page 35
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.35HySICSResults
PI: Greg Kopp / LASP
CoIs: Co-I – Peter Pilewskie / LASPBalloon Flight Manager – David Stuchlik / WFF
Build and flight test a hyperspectral imager with improved radiometric accuracies for climate science• 350-2300 nm with single FPA to reduce cost & mass• <0.2% (k=1) radiometric accuracy• <8 nm spectral resolution• 0.5 km (from LEO) IFOV and >100 km FOV• <0.13% (k=1) instrumental polarization sensitivity
Perform two high-altitude balloon flights to demonstrate solar cross-calibration approach and to acquire sample Earth and lunar radiances
Single HgCdTe FPA covers full shortwave spectral range with reduced mass, cost, volume, and complexity
Incorporate solar cross-calibration approaches demonstrated on prior IIP to provide on-orbit radiometric accuracy and stability tracking
Orthogonal configuration reduces polarization sensitivityNo-cost balloon flights from experienced team at NASA
WFF demonstrate on-orbit capabilities
Objective
Approach
HySICS to demonstrate climate science radiometric accuracies in shortwave spectral region
HySICS
LunarReconstruction
GroundReconstruction
SolarDataCube
• HySICSdemonstratesimprovedradiometricaccuraciesbased
onsolarcross-calibrationsunderrealisticflight conditions
• HySICSacquiredrepresentative350-2300nmspatial/spectral
dataoftheSun,Moon,andEarth
• HySICSdemonstratesthefeasibilityofacquiringreflectedsolar
datawithasinglespectrometer
– Reducesmass,power,volume,cost,risk,andcomplexity
• HySICSbuildsonandimprovesneededgroundtestequipment
• HySICSdemonstratesseveralflightcapabilitiesofCLARREO-like
reflectedsolarinstrument
Page 36
CLARREOSDTMeetingHampton,VA,29Nov.2016 GregKopp- p.36HySICSResults
HySICS’ViewofInstrumentTeam
Thank
You!