1

Collection6.1ChangeDocumentTERRAMOD07AtmosphericProfileProducts

EVAE.BORBAS,W.PAULMENZEL,CHRISMOLLERCIMSS,UNIV.WISCONSIN;1225WESTDAYTONSTREET;MADISON,WI

Background ObservationsintheTerraMODISPhotovoltaic(PVLWIR)bands27–30areknowntobeinfluencedbyelectroniccrosstalk.ThemagnitudeofthiscrosstalkaffectingL1Bradianceshasbeensteadilyincreasingthroughoutthemissionlifetime,causing,forexample,earthsurfacefeaturestobecomeprominentinatmosphericband27,increaseddetectorstriping,andtrendsinradiometricbiases.

Inaddition,onFebruary18,2016theTerraMODISinstrumententeredasafemodestatuswhentheTerraspacecraftexceededspecifiedlimitsonpointingduringaninclinationadjustmentmaneuver.MODISwasinsafemodefrom18to24Februaryanditsfocalplanetemperatureswereuncontrolled.Afterthesafemode,theelectroniccrosstalkbehaviorshowedasignificantincreaseforallPVLWIRbandsinamplitude(Wilsonetal.,2017).

Inrecentyears,thecrosstalkcontaminationhasbeenrecognizedascompromisingtheclimatequalitystatusofseveralMODISL2andL3scienceproductsthatdependonthePVLWIRbands.Inresponse,theMODISCharacterizationSupportTeam(MCST)hasgeneratedacrosstalkcorrectionalgorithmintheoperationalL1Bradiancealgorithm.ThecrosstalkcorrectedL1Bradianceshavebeentested(Moelleretal.,2014,2017)inseveralTerraMODISL2scienceproductalgorithms,includingMOD35(CloudMask),MOD06(CloudFraction,CloudParticlePhase,CloudTopProperties),andMOD07(AtmosphereProfiles).Thisdocumentdescribestheseso-calledCollection6.1(061)changestoMODISMOD07MoistureandTotalOzoneAtmosphericProfileProducts.

TheMODISMOD07L2operationalalgorithmforretrievingtemperatureandmoistureprofiles,totalcolumnozone,totalandlayerintegratedwatervaporfrominfrared(IR)radiancesobservedbytheNASATerraMODISinstrumentisaclearskysyntheticregressionretrievalalgorithmcalledMOD07.MOD07inputsaretheclear-skyradiancesoverlandandoceanforbothdayandnightfromelevenMODISinfraredchannels(25,27-36)thatincludethecrosstalkaffectedPVLWIRbands.Theoperationalalgorithmconsistsofseveralproceduresthatincludecloudfiltering(MOD35),averagingclearradiancesover5by5field-of-view(FOV)areas,forwardmodelcalculations,andaregressionbasedretrieval.TheradiativetransfercalculationoftheMODISspectralbandradiancesisperformedusingtheJCSDACommunityRadiativeTransferModel(CRTM)transmittancemodel(Hanetal.,2005)includingspectralshiftstosomebands(27-28,30,34-36)tobringtheirradiancesinlinewiththoseoftheAIRS(Tobinetal.,2006)andIASIsensors(Quinnetal.,2010).

2

TERRA/MOD07WaterVaporProducts Trends in theC6MOD07upper (UTWV,above440hPa),middle (MTWV,between440and640hPa)andlower(LTWV,below640hPa)tropospherewatervaporshowanoticeablenegativetrendintheMTWVandamodestpositivetrendintheUTWV,especiallyoverthetropics(Figure1).Thesetrendsarenotfoundincorresponding AquaMODIS C6 products and point towards a biasing of the TerraMODIS trends due tocrosstalkincreasinginthePVLWIRbands.

Figure 1. Time series of C6 monthly mean Terra MODIS upper, middle and lower troposphere integrated water vapor (UTWV, MTWV, LTWV, respectively) for the mid-latitude North (30N - 60N, red), mid-latitude South (30S - 60S, green) and tropical latitude (30N - 30S, blue) zones at nighttime. Toinvestigatethis,theNASADAACprovidedTerraMODIScrosstalkcorrectedL1BplusMOD07andMOD08(L3)productsforthemonthofAprilineveryyearfrom2000through2016,plusJuly2016.Figure2showsthree(TerraC6,AquaC6,andTerracrosstalkcorrected)fieldsofMTWVforApril2015.ThebottompanelsshowtheMTWVdifferencesbetweenC6TerraandC6Aqua(left)productsandbetweenthecrosstalkcorrectedTerraandC6Aqua(right).TheC6TerraMTWV(topleftpanel)isartificiallyreducedbycrosstalkerrorsinband27andband28,especiallyinthetropicsoverocean.However,thecrosstalkcorrectedMTWV(toprightpanel)muchmorecloselymatchesthatofAquaMYD07(topmiddlepanel)withdifferencesfallingtolessthan2mmovertheentireglobe(bottomrightpanel).

3

Figure 2: The MOD08 integrated water vapor of the middle troposphere (MTWV) for April 2015 from Terra MODIS C6 (top left), Aqua MODIS C6 (top middle) and crosstalk corrected Terra MODIS (top right). Differences between Terra C6 and Aqua C6 (bottom left) and Terra crosstalk corrected and Aqua C6 (bottom right) are also shown. ThedistributionofthelatitudinalmeandifferencesofMTWVforApril2015isplottedinFigure3.TheleftpanelrepresentsthemeanandstandarddeviationofthedifferencesbetweenC6TerraandAquaandtherightpanelshowsthesameusingtheTerracrosstalkcorrecteddata.TheTerraC6differencesshowsastronglatitudinaldependenceinthedistributionwiththemaximummonthlymeanoccurringintheEquatorialzone(5mm;~50%oftheabsolutevalue);afterthecrosstalkcorrection,themeandifferencesarereducedatalllatitudesandthelatitudinaldependenceislargelyremovedwithequatorialzonedifferenceslessthan0.8mm(<10%ofabsolutemeanvalue).Standarddeviationsarealsosignificantlylowered,suggestingthattheuncertaintyoftheMOD08productisreduced.

Figure 3: The latitudinal profile of the April 2015 MOD08 MTWV differences between Aqua MODIS C6 and Terra MODIS C6 (left), and between Aqua MODIS C6 and Terra MODIS crosstalk corrected (right).

4

Figures4and5aresimilartoFigures2and3,exceptshowingUTWVresults.TheeffectofthecrosstalkerrorontheUTWVistoonaverageincreasethewatervaporamount(oppositeofMTWV).Again,aftercrosstalkcorrection,theTerraMODISUTWVproductmuchmorecloselymatchesthatofAquaMODIS.Figure5illustratesthatthemaximumdifferencesoccurringintheequatorialzonearealsolargelyeliminatedandstandarddeviationsarereducedbythecrosstalkcorrection(rightpanelofFig5).

Figure 4. Same as Figure 2, except for upper troposphere (UTWV).

Figure 5. Same as Figure 3, except for upper troposphere (UTWV).

5

Figure6,atimeseriesoftheL3total,LTWV,MTWVandUTWVintegratedwatervaporinthetropics,indicatesthatthecrosstalkerrorinC6radiancescausedamoisturedeficitinthelowerandmiddlelayerofthetroposphereandexcessivemoistureintheuppertroposphere.Thetimeseriesshowthisbecomingmorepronouncedinabout2010(seegreencrosses)forLTWVandMTWVandevidentalreadyin2003forUTWV(notethatthescalingisdifferentforeachplotandmaybemaskingbehaviorofLTWVandMTWVearlierinthemission).Thetrendsfrom2001tothepresentforcrosstalkcorrectedTerraTPW,UTWV,MTWVandLTWV(redcrosses)showthatthecrosstalkcorrectionismovingtheTerramoistureproductsclosertotheAquadeterminations(especiallyforUTWVandMTWV).ThecrosstalkcorrectedTerratrendsmaybefurtheralteredbyadjustmentstothespectralshiftsofbands27-28and30.Theseshiftswillbere-evaluatedusingcrosstalkcorrectedL1Bradiances.

Figure 6: Terra (MOD) and Aqua (MYD) tropical integrated water vapor 16 year trends for total (TPW), lower, middle and upper layers for C6 and for Terra crosstalk corrected (“Test”). The crosstalk correction moves the Terra moisture trends into closer agreement with the Aqua trends in all layers.

6

TERRA/MOD07TotalOzoneProduct DeterminationoftotalozoneintheMOD07Algorithmisperformedbyintegratingozoneprofilesthroughtheatmosphericcolumn.Thetimeseries(seeFigure6)oftheL3totalozoneseparatelyovertheNorthernandSouthernHemisphere(between30and60-degreelatitude)andTropicsindicatesthatthecrosstalkerrorinC6radiancescausedexcessiveozoneforallthreelatitudinalbands.Thetimeseriesshowthisbecomingmorepronouncedinabout2010(seegreencrosses),andafterthesafehold(Feb2016)theincreasebecamemuchstronger(>100%).Thetrendsfrom2001tothepresentforcrosstalkcorrectedTerraTOZshowthatthecrosstalkcorrectionismovingtheTerraTOZproductsclosertotheAquadeterminations,althoughtherearestillsomeincrease(~20-30DU)afterthesafeholdmode(Feb2016)perhapsduetosomeresidualelectroniccrosstalkinband30afterthecrosstalkcorrection.

Figure 7: Terra (MOD) and Aqua (MYD Total Ozone 16 year trends over Northern Hemisphere (top), Tropic (middle) and Southern Hemisphere (bottom) for C6 and for Terra crosstalk corrected (“Test”). The crosstalk correction moves the Terra total ozone trends into closer agreement with the Aqua trends in latitudinal region.

7

Toinvestigatethespatialdistributionoftheeffectofthecrosstalkcorrection,theglobalmapandlatitudinalzonalmeanisplotted.Figure8showsthethree(TerraC6,AquaC6,andTerracrosstalkcorrected)fieldsofTOZforApril2016.ThebottompanelsshowtheTOZdifferencesbetweenC6TerraandC6Aqua(left)productsandbetweenthecrosstalkcorrectedTerraandC6Aqua(right).TheC6TerraTOZ(topleftpanel)isartificiallyincreased(maxdiffis650DU)bycrosstalkerrorsinband30,especiallyinabovethe30degreeLatitudes.However,thecrosstalkcorrectedTOZ(toprightpanel)muchmorecloselymatchesthatofAquaMYD07(topmiddlepanel)withdifferencesfallingtolessthan100DUovertheentireglobe(bottomrightpanel).

Figure 8: The MOD08 Total Ozone for April 2016 from Terra MODIS C6 (top left), Aqua MODIS C6 (top middle) and crosstalk corrected Terra MODIS (top right). Differences between Terra C6 and Aqua C6 (bottom left) and Terra crosstalk corrected and Aqua C6 (bottom right) are also shown. Note that the scaling is different for the bottom right panel. ThedistributionofthelatitudinalmeandifferencesforApril2015(leftpanels),April2016(middlepanels)andJuly2016(rightpanels)isplottedinFigure9.ThetoppanelsrepresentthemeanandstandarddeviationofthedifferencesbetweenC6TerraandAquaandthebottompanelsshowthesameusingtheTerracrosstalkcorrecteddata.TheTerraC6differencesshowsastronglatitudinaldependenceinthedistributionwiththemaximummonthlymeanoccurringinbothSouthernandNorthern60-70degreelatitudes(600DU;~100%oftheabsolutevalue);afterthecrosstalkcorrection,themeandifferencesarereducedatalllatitudesapartfromthehighpolarlatitudes(80-90degrees),andthelatitudinaldependenceislargelyremovedwithdifferenceswellbelow100DU(<15%ofabsolutemeanvalue).Standarddeviationsarealsosignificantlylowered,suggestingthattheuncertaintyoftheMOD08TOZproductisreduced.

8

Figure 9: The latitudinal profile of the April 2015 (left column), April 2016 (middle column) and July 2016 (right column) MOD08 Total Ozone differences between Aqua MODIS C6 and Terra MODIS C6 (upper panels), and between Aqua MODIS C6 and Terra MODIS crosstalk corrected (bottom panels).

SUMMARY

AcrosstalkcorrectiondevelopedbyMCSTforTerraMODISPVLWIRbands27-30hasbeentestedintheL3MOD07moistureandtotalozonescienceproductstoaddresspoorperformancefoundinTerraMODISCollection6productsthatusethePVLWIRbands.TheeffectofthecrosstalkerrorontheMOD07moistureproductshasastronglatitudinaldependenceinthedistributionpeakingattheTropicscausingahugemoisturedeficit(50%)inthelowerandmiddlelayerofthetroposphereandamoistureincrease(50%)intheuppertroposphere.ThecorrectionhasbroughttheTerraMODISmoistureintomuchcloseragreementwiththatofAquaMODIS,reducingthemoistureincreaseinthelowandmiddlelayeranddeficitintheupperlayertolessthan10%.Alatitudinaldependencecausedbythecrosstalkerrorisalsolargelyeliminated.

ThecrosstalkerrorinC6radiancescausedexcessiveozoneforalloverwithastronglatitudinaldependenceinthedistributionwiththemaximummonthlymeanoccurringinbothSouthernandNorthern60-70degreelatitudes.Thiseffectbecamemorepronouncedinabout2010,andafterthesafehold(Feb2016)theincreasebecamemuchstronger(>100%~600DU).Wedemonstratedthatthecrosstalkcorrectionreducedthemeandifferencesatalllatitudesandthelatitudinaldependenceislargelyremovedalthoughtherearestillsomeincrease(max100DUatthe60-70degreelatitudes)afterthesafeholdmode(Feb2016).

9

Afterthecrosstalkcorrection,somedeparturesforboththemoistureandozoneproductsyetexistwhichmaysignalthatthespectralshiftsappliedtobands27,28and30forC6MOD07processingarenotaccurateafterthecrosstalkcorrection.Theseshiftswillbereviewedandupdated.ThisworksuggeststhattheseTerraMODISproductscanberestoredtoclimatequalitystatusbythecrosstalkcorrection.

ACKNOWLEDGEMENTS

TestdatasetsweregeneratedbythestaffatMODAPS.ThisworkhasbeenfundedunderNASAGrantNNX14AH48G.

REFERENCES Han,Y.,P.Delst,Q.Liu,F.weng,B.Yan,andJ.Derber,2005:User’sGuidetotheJCSDACommunity

RadiativeTransferModel(BetaVersion),http://www.star.nesdis.noaa.gov/smcd/spb/CRTM/crtm-code/CRTM_UserGuide-beta.pdf.

Moeller,C.,W.P.Menzel,andG.Quinn,2014:ReviewofTerraMODISthermalemissivebandL1BRadiometricPerformance.Proc.SPIE9218,EarthObservingSystemsXIX,92180T(September26,2014);doi:10.1117/12.2062138.

Moeller,C.,RichardFrey,EvaBorbas,W.PaulMenzel,TrumanWilson,AishengWu,XuGeng,"ImprovementstoTerraMODISL1B,L2,andL3scienceproductsthroughusingcrosstalkcorrectedL1Bradiances,"Proc.SPIE10402,EarthObservingSystemsXXII,104020O(5September2017);doi:10.1117/12.2274340

Quinnetal,2010,http://www.ssec.wisc.edu/~gregq/iasi-modis/plots.html

Tobin,D.C.;Revercomb,HenryE.;Moeller,ChristopherC.andPagano,ThomasS.2006:UseofAtmosphericInfraredSounderhigh-spectralresolutionspectratoassessthecalibrationtoModerateresolutionImagingSpectroradiometeronEOSAqua.JournalofGeophysicalResearch,Volume111,2006,Doi:10.1029/2005JD006095,2006.CallNumber:Reprint#5053.http://www.agu.org/pubs/crossref/2006/2005JD006095.shtml

Wilson, T., Wu, A., Shrestha, A., Geng, X., Wang, Z., Moeller, C., Frey, R. and Xiong, X., “Development and implementation of an electronic crosstalk correction for bands 27-30 in Terra MODIS collection 6,” Remote Sens. 9(6), 569 (2017).