D-75724 SWOT Cal Val Plan Initial 20180129u · 29/01/2018 · A cartoon illustrating the SWOT measurement conceptis presented in Figure 1. ... download limitations, the ocean data

SWOT Project

SWOT Calibration / Validation Plan

Initial Release Document Custodian: ____________________________________ Curtis Chen Date JPL Cal/Val Lead Approved: ____________________________________ Shailen Desai Date JPL Measurement System Engineer ____________________________________ Lee-Lueng Fu Date Project Scientist ____________________________________ Tamlin Pavelsky Date JPL Hydrology Science Lead

____________________________________ Nicolas Picot Date CNES Measurement System Engineer ____________________________________ Rosemary Morrow Date CNES Oceanography Science Lead ____________________________________ Jean-Francois Cretaux Date CNES Hydrology Science Lead

Paper copies of this document may not be current and should not be relied on for official purposes. The current version is in the Product Data Management System (PDMS): https://pdms.jpl.nasa.gov/ © 2018 California Institute of Technology. Government sponsorship acknowledged. 01/29/2018 JPL D-75724

Initial Release JPL D-75724 01/29/2018 SWOT Calibration / Validation Plan

This document has been reviewed and determined not to contain export-controlled technical data. Government sponsorship acknowledged.

2

CHANGE LOG

DATE SECTIONS CHANGED REASON FOR CHANGE REVISION 1/29/2018 ALL Initial Release

Approved for export (LRR-034457) Initial Release



3

1 SWOTCalibrationandValidationScope....................................................................................................................51.1 MeasurementSystemOverview..........................................................................................................................51.2 Calibration.....................................................................................................................................................................61.3 Validation.......................................................................................................................................................................81.4 MinimumSitesandSecondTierSites...............................................................................................................91.4.1 MinimumHydrologySitesandRequirements.............................................................................101.4.2 OceanMinimumValidationSitesandRequirements...............................................................11

1.5 OverviewofCal/ValTimeline............................................................................................................................121.6 TeamRolesandResponsibilities......................................................................................................................121.7 DocumentPurposeandScope...........................................................................................................................13

2 SpecialProvisionsfortheCalibration/ValidationPhase.................................................................................152.1 PhaseCalibrationLoopandFour-ChannelRawDataDownload.......................................................152.2 FastSamplingPhase..............................................................................................................................................152.3 AirborneMeasurements......................................................................................................................................162.3.1 AirSWOT.......................................................................................................................................................172.3.2 MASSLidar..................................................................................................................................................18

3 KaRINCalibrationPlan...................................................................................................................................................223.1 CalibrationParametersDuringInstrumentCheckout............................................................................223.1.1 DifferentialRangeDelay........................................................................................................................223.1.2 CommonRangeDelay.............................................................................................................................223.1.3 FunctionalValidationoftheOceanOnboardProcessor..........................................................22

3.2 CalibrationParametersDuringtheOne-DayRepeatPhase.................................................................223.2.1 DifferentialRangeDelay........................................................................................................................223.2.2 PhaseScreen...............................................................................................................................................223.2.3 Cross-OverCalibrationforStaticPhase/RollBiases.................................................................253.2.4 StaticRangeBiases..................................................................................................................................273.2.5 RadiometricCalibration........................................................................................................................28

4 NadirAltimeterCalibrationPlan................................................................................................................................294.1 Trackingmodes........................................................................................................................................................294.1.1 TheDiode/DEMtrackingmode..........................................................................................................294.1.2 TheDiode/DEMtrackingmodevalidation....................................................................................30

4.2 InternalCalibrations..............................................................................................................................................325 WaterVaporRadiometerCalibrationPlan............................................................................................................345.1 BrightnessTemperatureCalibration..............................................................................................................345.2 Inter-beamCalibration.........................................................................................................................................34

6 SWOTValidationPlan.....................................................................................................................................................366.1 Introduction...............................................................................................................................................................366.1.1 Overview......................................................................................................................................................366.1.2 MinimumCalValengagement.............................................................................................................36

6.2 SWOTOceanErrorBudgetValidation...........................................................................................................366.2.1 Randomheighterrorvalidation.........................................................................................................366.2.2 Roll/phasedriftvalidation...................................................................................................................376.2.3 PODValidation...........................................................................................................................................386.2.4 Wet-tropodelayvalidation..................................................................................................................416.2.5 Otherpropagationdelayvalidation..................................................................................................426.2.6 EMBiasandotherwaveefffectvalidation....................................................................................436.2.7 Tidalcorrectionvalidation...................................................................................................................446.2.8 Dynamicatmosphericcorrectionvalidation................................................................................446.2.9 Rainflagvalidation..................................................................................................................................44



4

6.2.10 Iceflagvalidation......................................................................................................................................456.2.11 Landflagvalidation.................................................................................................................................46

6.3 OceanDataProductValidation.........................................................................................................................466.3.1 AbsoluteRangeandSSHvalidation..................................................................................................466.3.2 SWHValidation..........................................................................................................................................496.3.3 Oceanσ0Validation..................................................................................................................................506.3.4 WindSpeedValidation...........................................................................................................................50

6.4 ScienceValidationofOceanMeasurements................................................................................................516.5 SWOTSurfaceWaterErrorBudgetValidation...........................................................................................546.5.1 Randomheighterrorvalidation.........................................................................................................556.5.2 Absoluteinlandsurfacewaterheightvalidation........................................................................566.5.3 Inundatedsurfaceareavalidation....................................................................................................586.5.4 Rangedriftvalidation.............................................................................................................................636.5.5 Roll/phasedriftvalidation...................................................................................................................636.5.6 LandWet-TropoDelayValidation.....................................................................................................636.5.7 Otherpropagationdelayvalidation..................................................................................................646.5.8 Slopevalidation.........................................................................................................................................646.5.9 Layoverflaggingandimpactvalidation..........................................................................................656.5.10 Rainflagvalidation..................................................................................................................................656.5.11 Iceflagvalidation......................................................................................................................................656.5.12 Landflagvalidation.................................................................................................................................656.5.13 Geolocationvalidation............................................................................................................................66

6.6 SurfaceWaterDataProductValidation.........................................................................................................666.6.1 Pixelcloudproductvalidation............................................................................................................666.6.2 Rivervectorproductvalidation.........................................................................................................666.6.3 Lakevectorproductvalidation...........................................................................................................676.6.4 Rasterproductvalidation.....................................................................................................................67

6.7 DischargeCharacterization.................................................................................................................................686.7.1 Characterizationofderivedbathymetry........................................................................................686.7.2 Characterizationofderiveddischarge............................................................................................68

7 SWOTCal/ValSites..........................................................................................................................................................717.1 OceanCal/Valsites.................................................................................................................................................717.1.1 ValidationoftheabsoluteSSHbias..................................................................................................717.1.2 Validationofrelative2DSSHandcurrents...................................................................................81

7.2 HydrologyCal/Valsites........................................................................................................................................927.2.1 RiverCal/ValSites....................................................................................................................................927.2.2 LakeCal/ValSites..................................................................................................................................1177.2.3 WetlandValidationSites....................................................................................................................1287.2.4 Tidal/EstuarineValidationSites.....................................................................................................1357.2.5 GlobalPlanforTier2Cal/ValSites(U.S./FranceJointProject)........................................1417.2.6 CornerReflector/TransponderCalibrationSites....................................................................144

8 References.........................................................................................................................................................................148AppendixA. Specific Resources Required and Timing of Activities for the Inland HydrologyCal/ValProgram.......................................................................................................................................................................150



5

1 SWOTCALIBRATIONANDVALIDATIONSCOPEThissectionprovidesanintroductiontoSWOTcalibrationandvalidationactivitiestobeconductedbytheCal/Valteam.ThisintroductionprovidesthescopeandobjectivesofCal/Valwork,amid-leveldescriptionofCal/Valactivities,andtheorganizationalcontextofhowtheworkwillbeundertaken.AdditionaldetailsontheCal/Valplanitselfaregiveninsubsequentsections.

1.1 MeasurementSystemOverviewAdescriptionoftheSWOTmeasurementcharacteristicsandrequirementsispresentedintheSWOTMission Science Document (Fu et al., 2012) and in the SWOT Science Requirements Document(Rodríguez et al., 2016). An additional description of the SWOT science goals and expectedperformance is given by Durand et al (2014). In this section,we present a brief summary of themeasurementsystemkeycharacteristics.TheSWOTmission iscomposedofseveral instruments:adual-frequency(KuandC-band)nadiraltimeter;KaRIn,aKa-bandradarinterferometer;adual-beamwatervaporradiometer(AdvancedMicrowaveRadiometer,AMR);and,aDopplerOrbitographyandRadiopositioningIntegratedbySatellite(DORIS)beacon,aGlobalPositioningSystem(GPS)receiver,a Laser Retroreflector Array (LRA), star trackers, and gyros, for precision orbit and attitudedetermination.AcartoonillustratingtheSWOTmeasurementconceptispresentedinFigure1.

Figure1MeasurementconceptforSWOT.



6

Belowwesummarizetheexpectedcapabilitiesofeachofthesecomponents:

1. NadirAltimeter:This system isacloneof the Jason-classaltimeters. Itprovidesvalidationandlong-wavelengthmeasurementsofseasurfaceheight(SSH).

2. KaRIn: This is the main instrument for measuring high-resolution elevations for SSH andsurfacewatermeasurements. ItconsistsofadualbeamKa-bandradar interferometer,eachbeamprovidingabsoluteelevationmeasurementsoveranominal50kmswaththatextendsfrom 10 km to 60 km on either side of the altimeter nadir track. In order to meet datadownloadlimitations,theoceandataareprocessedonboardtoapostingof250m.Dataoverland is downlinked at a higher data rate, enabling estimation of elevations with a spatialresolution,aftertakingazimuthlooks,ontheorderof25mintheazimuthdirectionand70m-10mintherangedirection,dependingonthecross-trackdistance.

3. AMR:Thisinstrumentisanevolutionoftheadvancedwatervaporradiometer(AMR)intheJason-3 mission and provides estimates of wet tropospheric delays over the ocean with aresolutionforitslowestfrequencyofabout40km.ThemaindifferencebetweentheAMRonSWOTandtheoneon Jason-3 is thepresenceof twobeams,centeredontheKaRInswaths,ratherthanasinglebeampointedalongthenadirdirection.

4. OrbitDetermination:TheSWOTmissionwillcarryanorbitdeterminationinstrumentsuitevery similar to the one that has been used by the Jason altimeter series (DORIS, GPS, andLRA).

5. AttitudeDetermination: SWOT includesa star-trackernear the instrumentandadditionalgyrosintheinstrumentsuiteforimproveddeterminationoftheinterferometricbaseline.

1.2 CalibrationThe scope of the SWOT calibration activities will be to conduct appropriate independentmeasurements to determine SWOT static system parameters used in ground processing. Theseparametersareexpectedtobeconstantintime.Long-termCal/Valactivitieswillmonitorfordriftsinthe calibration parameters and will allow for calibration parameter updates as a contingencyscenario,butthebaselineCal/Valplanassumesthatthecalibrationparameterscanbesetonceandwillremainfixedforthedurationofthemission.Parametersusedinon-boardprocessingwillbedeterminedduringthecheckoutandcommissioningphaseandarenotstrictlypartofCal/Valactivities(SeeSect.1.5).Notethatinsomecases,thesameparametermay takeondifferent values for on-boardprocessing than for groundprocessing. Thismay typically occur when a coarse estimate of the parameter is needed for on-board processingsimply to avoid interferometric decorrelation, while a finer estimate of the parameter is used forgroundprocessinginordertoprovidethefinestpossibleabsoluteaccuracyofthemeasurement.Dynamic variations for some calibration-related parameters are addressed by specific operationalscienceprocessingalgorithms;thesetreatedseparatelyfromthestaticparameterstobeestimatedinthecalibrationphase.Thefulllistofparameterswillbegiveninsubsequentsections,butkeyparametersarelistedhere:



7

CommonRangeDelay:ThisistheaverageoftherangebiasbetweenthetwoKaRInchannels,andiscausedbydelaysintheinstrumentthatcouldnotbecalibratedpriortolaunch.Itistheequivalentofthenadirrangebias.DifferentialRangeDelay:ThisisthedifferencebetweenthetwoKaRInchannels.Itsmaineffectwillbetocausethetwochannelstobemiss-registered,leadingtoalossofcorrelationandaphasebias.StaticDifferentialPhase:ThisisanyresidualphasebetweenthetwoKaRInchannelsthatisstaticor varying very slowly (on the scale of months or years). It should not be confused with theinstantaneous channel-to-channel phase that can vary due to the changes in temperature ormechanicaldilationsbetweenchannels;suchdynamiceffectsareaddressedviaoperationalsciencealgorithmprocessingusingdownlinkeddynamiccalibrationdataaswellascrossovercorrections.StaticRollAngle:Thisisanyerrorinknowledgeofthebaselineandantennaorientationafterspacedeployment. As with the previous parameter, we only calibrate the static part and recognize thatadditionalrollerrorswillbepresentduetouncertaintiesintheIMUrollestimation.Inpractice,onecannot differentiate between static roll and static phase biases, and an effective roll (or effectivephase)willbetheonlyparameterestimated(perswath).BaselineLength:Thisparameterisexpectedtobeknownwithhighaccuracypriortodeployment,but itwillberefinedduringcalibration(perswath).Notethatthebaseline lengthandanglecanbeequivalentlyexpressedintermsoftheleverarmstotheantennaphasecentersintheKaRInreferenceframe. Dynamic variations are addressed in operational science algorithmprocessing through theuseofcrossovercorrections.ReferencePointLocation:Theeffectoflocationerrorsforthereferencepointinthenadirdirectionis(nearly)identicaltoacommonrangedelay,andwillbeincorporatedintothatparameter.Errorsinlocation in the orthogonal planewill lead to geolocation errors. It is assumed that the cross-planelocationof the referencepointwill beknown to sufficient accuracy (<10 cm), so that the effectonabsolutegeolocationcanbeneglected.PhaseScreen:Experiencehasshownthat it is impossible inpractice tomatchexactly the far fieldphaseofbothantennas.Differencesinthephasefar-fieldpattern,whichmaybecausedbyinteractionwiththebaselineandspacecraftstructures,willresultinphasedifferencesbetweenthechannelsthatvariesasafunctionof lookangle,or,equivalently,absolutephase.Unliketheotherparameters,thephasescreenisnotasinglevalue,butacontinuousfunctionthatmustbeestimatedacrosstheentireswath.Ithasneverbeencalibratedpriortodeployment.AbsoluteSystemGain:Althoughnotstrictlynecessary for interferometry, theabsolutegainof thesystemisdesirableifonewantstorelatethewindmodelfunctionderivedbySWOTtothatderivedbyothersystems.However,ifthemodelfunctionisderivedfromSWOTdataalone,thereisnoneedforabsolutegaincalibration.Swath-to-Swath Gain Calibration: Although each of the swaths has a different polarization, therestricted set of near-nadir incidence angles implies that a single wind model function will besufficient forbothswaths. Inorderforthistobethecase, thegain inbothchannelswillhavetoberelativelycalibrated.



8

AntennaPatternRelativeGain:Again,thisisnotrequiredforinterferometry,butisrequiredfortheestimationofmeansquaredslopefromthedecayofthecrosssectionasafunctionofangle.Itwillbeassumed that both antennas arematched sufficientlyprior to launch so that this calibration is notnecessary.Thenadiraltimeter,ontheotherhand, isasimplerinstrumentandforthepurposesofSWOT,onlythe altimeter range bias (and drift) needs to be calibrated against a reference constellation ofaltimetersandagainsttheKaRINinterferometerrangebias.

1.3 ValidationTheSWOTvalidationactivitieswillbedividedbetweentheprojects, theSWOTscience team,otheragenciesandforeignpartners.ThescopeoftheprojectresponsibilitiesaregovernedbythefollowingrequirementsfromtheSWOTScienceRequirementDocument(revB):

2.7.6 [Requirement] The SWOT ocean performance shall be verified by payloadindependent measurements or analysis during a post-launch calibration/validationperiod.2.8.12 [Requirement] The SWOT surfacewater elevation shall be verified by a payloadindependentmeasurementoranalysisduringapost-launchvalidationperiodaswellasduringthemissionlifetime.2.8.14[Requirement]TheSWOTdischargeperformanceshallbequantifiedbyapayloadindependentmeasurementoranalysisduringapost-launchvalidationperiodaswellasduringthemissionlifetime.2.8.15[Requirement]SWOTelevationandinundationextentperformanceinvegetatedwetlandsshallbequantifiedbyapayloadindependentmeasurementoranalysisduringapost-launchvalidationperiodaswellasduringthemissionlifetime.2.6.3.a[Requirement]ALevel-2pixelclouddataproductshallbeproducedforthesurfacewaterdata.Thepixelclouddataproductincludes:

● […]● As noted below, SWOT required performance will be evaluated using non-

vegetated water bodies meeting the minimum size criteria set in the sciencerequirements, i.e., water bodies with area greater than (250m)2 and rivers ofwidthgreaterthan100m.However,theSWOTperformancewillbecharacterizedfornon-vegetatedwaterbodiesmeetingtheminimumsizecriteriainthesciencegoals; i.e., water bodies with area greater than (100 m)2 and rivers of widthgreater than 50 m. Only non-vegetated water bodies in regions of moderatetopographicrelief(i.e.,wherelayovercontaminationisnegligible)aretobeusedtoassessSWOTperformance.

● […]Although not explicitly stated in the Science Requirements Document, all mission product typesdescribedthere,withtheinlandwaterbodiesexceptionsnotedabove,willbevalidated.Thatis,forthe cases of significant layover, wetlands, water bodies below minimum size requirements, etc.,SWOTperformancewillbequantifiedandevaluated,buttheresultswillnotbecountedagainstthe



9

performance requirements; the science requirements are not applicable in such cases due to theexclusionsexplicitlydefinedbytherequirements.Theproject activitieswill consistof validationof the systemparameters listedabove, validationoftheSWOTerrorbudget,andvalidationofthedataproductsreleasedbytheproject.Inthecontextofthis document, “validation of the error budget” means confirming that the SWOT measurementperformance, includingerrorcontributions thatmaybeseparatelyobservableonly in intermediatedata, matches expectations based on the team’s best understanding of the system (includinginstrument,spacecraft,andgroundprocessing).Ontheotherhand,“validationofthedataproducts”means confirming that the SWOT measurement performance, as achieved in the data productsavailable tousers, is consistentwith the SWOTscience requirements. Validationwill occurover arangeofconditionssufficienttocapturerepresentativeglobalperformance.Theocean science requirements impose an elevation error accuracy that is defined in the spectraldomain.This shouldbe contrasted to the traditional altimeter requirements,where the total errorintegratedoverallscalesisspecified.Inpractice,thisdifferencewillmeanthatthevalidationoftheSWOTmeasurementsmustbedoneover anextended test site.This is in contrast to the altimeter,wherepointtestsites(PointConcepcion,Corsica,Bassstrait)weresufficienttoprovideacompletevalidationofthemeasurementerrorbudget.Anotherdifferencewith traditionalaltimetry is thatawaterbodyextentrequirementmustalsobevalidatedforfreshwaterbodies.Inordertoperformthisvalidation,independentandsimultaneousdeterminationofwaterextentmustbeperformedduringperformancevalidation.Additionalactivities,beyondthosecoveredbytherequirementsabovemaybeproposedandselectedbypeerreviewunderNASAROSESorCNESTOSCAfunding.Theseactivitieswillnotbecoveredbythisdocumentastheyprovideadditionalcalibrationand/orvalidationbeyondthatneededtomeettheSWOTrequirements, thesubjectofthisdocument. Ingeneral,coordinationbetweentheprojectfunded activities described here and science team calibration/validation activitieswill be pursuedactivelytoreduceoverlapsandutilizepotentialsynergybetweendifferentprojects.

1.4 MinimumSitesandSecondTierSitesThe SWOTproject has selected a set of set of sites and instrumentation thatwill be theminimumneeded tomeet the validation requirements, and this sectionpresents anoverviewof the selectedminimumsites.Inadditiontotheseminimumsites,anadditionalsetofsecondtiersites,orsitesofopportunity, are presented which may become available through leveraging suitable foreign oragencypartners,asdescribedlaterinthedocument.TheCalValactivitiesforthemissionwillbejointlyfinancedbytheSWOTpartners,withthefundingfortheminimalsitescomingmostlyfromtheNASAandCNESProjects,withadditionalcontributionsfrom Canada or other national agencies for selected sites. At this point the workshare for theseactivities is under finalization. NASA has included a minimal set of sites sufficient for minimumvalidationunderitsbudget.FortheSWOTCalValsitesdescribedinthefollowingsections,therewillbeaminimumengagementbytheCNESprojectfor2mainoceanCalValsites,and2mainhydrologysites.Anumberofsecondarysites are also included in this document, which cover ocean and hydrology regionswith differentdynamics and phenomenology. The CalVal activities at these secondary in-situ CalVal siteswill be



10

accomplished with a best effort contribution from the SWOT project, combined with additionalnationalandEuropeanfinanceforcertaininstrumentationandanalyses.

1.4.1 MinimumHydrologySitesandRequirementsIn this section, we describe the minimum requirements for a comprehensive validation of SWOThydrologyobjectives.Theplansandsiteslistedhereareaminimumsubsetofthosefoundlaterinthereport. More detailed plans for validating SWOT hydrology measurements and data products aredescribedinSection6.5,andeachoftheCal/ValfieldsitesaredescribedindetailinSection7.2.Theminimumvalidation requirements for rivers require that SWOTobservationsofwater-surfaceheight, slope, and inundation extent as well as discharge characterizationmust be validated for arangeofriversizes,climatezones,andphysiographiccharacteristics. Toaccomplishthisobjective,wehavefocusedonasmallnumberofso-calledTier1Cal/Valsites:theWillametteRiver(small,mid-latitude temperate, single-to-multichannel), the Tanana River (large, sub-Arctic, braided), theConnecticut River (medium-sized, mid-latitude temperate, single-channel), the lower MississippiRiver (large,mid-latitude sub-tropical, single-channel), the St. Lawrence River (large,mid-latitude,single-channel),andatleastonelarge,tropicalriverinSouthAmerica,tobeconductedincooperationwithcolleaguesinBrazil.InadditiontotheTier1siteslistedabove,aminimumofonehundredso-called Tier 2 Cal/Val river sites will be utilized, which rely heavily on existing river gages andinstrumentation.InadditiontotheUSsites,therewillbeaTier1FrenchriversitethatwillcomplementtheAmericanriversdescribedabovetoaddadditionalrivertypestothevalidationdataset.Theminimum validation requirements for lakes require that SWOT observations ofwater surfaceelevationandinundationextentmustbevalidatedoverarangeofdifferentlakesizes,climatezones,and physiographic characteristics. The set of lake Cal/Val Tier 1 sites include: Lake Tahoe (large,mid-latitude,moderate elevation), a groupof lakes in the SierraNevada (small,mountainous, highelevation), a group of lakes in the Yukon Flats region (small- to medium-sized, sub-Arctic, low-topography) and a group of lakes in the Prairie Potholes region of the U.S. great plains (small- tomoderate-sized,low-topography,lowelevation).Inadditiontothesiteslistedabove,aminimumsetoffiftyTier2Cal/Vallakesiteswillbeutilized.TheminimumvalidationrequirementsforinundatedwetlandsrequirethatSWOTcharacterizationofwater-surfaceelevationandinundationextentmustbevalidatedoverarangeofwetlandvegetationtypes, climate zones and physiographic characteristics. The minimum validation requirements forwetlandswillbemetusingthefollowingTier1Cal/Valwetlandsites:theYukonFlats(boreal,sub-Arctic wetlands with sparse low-lying vegetation), the Atchafalaya Delta (subtropical, sparsely- tomoderately-wooded wetlands) and the Everglades (subtropical grassland wetlands with slopedwater-surfaceelevations).The minimum validation requirements for tidal sites require that SWOT observations of watersurface elevation, slope, and inundation extentmust be validated over a range of tidal conditions.TheminimumvalidationrequirementsfortidalareaswillbemetsolelyusingtheConnecticutRiverTier1Cal/Valtidalsite(mid-latitude,slightly-tomoderately-tidal).



11

Finally,minimumvalidationofSWOTice, rain, layover,andother flagswill takeplaceat theTier1sites listed above and as many of the Tier 2 sites as possible that have adequate nearbyinstrumentation.

1.4.2 OceanMinimumValidationSitesandRequirementsForscalessmallerthan150km,themainactivityforvalidationofSWOTovertheoceanconcentrateson validating the SWOT error spectrum contained in the science requirements document. Thevalidation of the error spectrum requires synoptic coincident measurements of SSH over a site.AirbornelidaristheprimarycandidatemethodforcollectingindependenttruthdataofabsoluteSSHfor SWOT ocean validation at ocean wavelengths as short as 15km. Lidar data will be collectedmainlyovertheprincipalUSCal/Valsite,althoughlidarflightsoverotherregionswillbeconsideredaswell. Thisapproach isdescribedfurther inSect.2. Sciencevalidationofdynamicheightwith insitumeasurementsisdescribedinSect.6.4.TheselectionofminimalsitesisguidedbytherequirementtobenearaSWOTcross-overduringtheSWOT1-dayrepeatphase inorder tomaximize thesamplingof thesite, theability tocollect truthdataatthesite,andthesizeofthesignalthatwillbepresentatthesite.Giventheserequirements,three candidate sites aremost promising: theGulf Stream (7.1.2.1) site, theCaliforniaCurrent site(7.1.2.2), andasite in theMediterranean(7.1.2.3).TheCaliforniaCurrentsitehasbeenselectedasprimary because the strong currents at the Gulf Stream site may pose a problem for in situinstrumentation. Moreover, thedynamicnature of theGulf Stream sitemaymake SWOTanomalyresolution more difficult if unanticipated issues arise. However, the Gulf Stream site will beconsideredasaback-up, incaseproblemsare identifiedwiththeCaliforniaCurrentsiteduringthepre-launch characterization phase. CNES will contribute a Mediterranean site (7.1.2.3), the sitespecificsandinstrumentationremaintobeconfirmed,afirstcampaignwilloccurinMay2018totestvariousmeasurementsmeansoverthissite.



12

1.5 OverviewofCal/ValTimelineTheprojectmissiontimeline,includingcalibrationandvalidationactivities,ispresentedinFigure2.

(The timeline is provided here for information only; the timeline is governed by other projectdocuments.)Thekeyitemsrelatedtothistimelineare:

1. Checkout and commissioning (85days),where the first rough setof instrument calibrationparameterswillbederived. On-boardparameterswillbeupdatedduring thecheckoutandcommissioning phase; science data collected from this phase may not meet SWOTperformance requirements even after ground reprocessing. Validation of the On BoardProcessor (OBP), comparing it to a reference algorithm on the ground, also occurs duringKaRIncommissioning.

2. Thecalibrationphase(90days),wherethecalibrationparametersarerefinedandvalidationover selected sites takes place in the fast-sampling orbit described in Section 2.2. Thecalibrationphasebeginsafteron-boardparametershavestabilizedandbeenvalidatedsuchthat science data collected after this point can be reprocessed on the ground to achievenominalSWOTscienceperformance.

3. Primary validation activities to continue until the calibration and validationmeeting, withlow-levelextended-validationactivitiesoccurringfortheremainderofthemission.

4. The first calibration and validation meeting occurs approximately 1.5 years after launch.Nominalinstrumentandprocessingparametersaredefined.

5. Ongoingcalibrationandvalidationactivities tomonitor forsystemdriftuntil theendof themission.

1.6 TeamRolesandResponsibilitiesDuring the checkoutandcommissioningphase, activities related toeach instrumentare ledby therespective instrument system engineering (SE) or instrument science teams. The Cal/Val team

Figure2SWOTnominalmissiontimeline,includingcalibrationandvalidationperiod.



13

supportsthesecommissioningactivities(forexample,bydeployingcornerreflectorsthattheKaRInSE team will use); the Cal/Val team participates in a background or “shadow” capacity. GroundalgorithmtestingandvalidationoccursduringbothcommissioningandCal/Valperiods(seebelow).Once the calibration phase begins, the Cal/Val team takes on overall leadership of calibration andvalidationactivitiesgoingforward.Atthispoint,theKaRInSEteamisdissolved,withkeypersonnelfrom that team transition to the Cal/Val team. The Cal/Val team has responsibility for gatheringexternal truth data for comparison to SWOTmeasurements, whether these data are collected viaSWOTCal/Valactivitiesortheyareproducedbyotherexistingorganizationsorassets.TheCal/Valteam will pull these data sets together and perform the comparisons of SWOT measurements toexternaldata.Calibrationandvalidationarethenbasedonthesecomparisons.TheCal/Valteamwilllead anomaly-resolution activities. Science representatives on the Cal/Val team will provide theinterfacetotheScienceTeamforvalidationactivities.TheserelationshipsareillustratedinFigure3.

Figure3.TeamorganizationduringtheCal/Valphase.

The Algorithm Development Team (ADT) will be responsible for validating the science algorithmsoftwaretothefullestpracticalextentusingSWOTdataonly,withouttheuseofexternaltruthdata.That is, the ADT is responsible for ensuring that ADT software gives self-consistent SWOTmeasurementsthatarefreeofgross,obviouserrors.TheADTwillalsoberesponsibleforfixinganyADT software problems, whether discovered using external truth data or not. The ADT willincorporate calibration parameters derived by the Cal/Val team, along with any other neededparameter or software updates, into new software releases that are delivered to the Science DataSystem(SDS)foroperationalprocessing.The Cal/Val team will interact with the relevant operations team(s) for any needed on-boardparameterupdatesifneededtoresolveanomalies,thoughnoneareplanned.

1.7 DocumentPurposeandScopeThisdocumentservesto(1)articulatethehigh-levelscopeofplannedCal/Valactivities inordertoensurecoordinationacrossprojectelementsandexternalorganizations;(2)definespecificplansforCal/Val activities with sufficient detail that these activities can be reconciledwith project budget,

Science Team

Cal/ValAlgorithm

Development Team (ADT)

Science Data

System (SDS)

Instr Ops

Instr SE

Science team representatives participate in Cal/Val site characterization, data analysis, and assessment

Instrument SE personnel roll into Cal/Val element after commissioning

Ops requests if needed, (e.g., high-res mask, onboard parameter updates, etc.)

Cal params

Alg update requestsProcessing software for evaluation

Software for release

Cal/Valteamincludesrepresentationfrom:JPLandCNESprojects

USandEuropeanoceanandhydroscienceExternalgovernmentagencies

Science team representatives participate algorithm development data product defintion



14

schedule,andworkforceplans,allowingprojectresourcestobeallocatedappropriately;(3)capturedetailed information regarding planned or potential Cal/Val activities in order to facilitate thetransfer of knowledge across the Cal/Val team. Given the important implications of the first twoobjectivesabove,thisdocumentwillbeconfigurationcontrolledasaprojectdocument.Aswithanyplan,however, it is anticipated that thisdocumentwill become less and lessuseful asactivitiesmove from theplanningphase through theexecutionphase. That is, onceaspectsof theplan are carriedout, recordsof these activities as they are executed supersede theplan andmakethose parts of the plan obsolete. Therefore, only major changes to the plan that have significantbearing on objectives (1) and (2)will trigger formal revisions to and re-release of this document.Minorchangestodetailsmaybehandledinformally.



15

2 SPECIALPROVISIONSFORTHECALIBRATION/VALIDATIONPHASEInthissection,weprovideabriefoverviewofspecialprovisionsthathavebeenmadetoensuretheappropriatecalibrationandvalidationofSWOT.

2.1 PhaseCalibrationLoopandFour-ChannelRawDataDownloadByimplementingacalibrationloopthatmeasuresthephaseforasignificantfractionofthetransmitandreceivepathsforeachchannel,thebulkofchannelphaseimbalances,includingalltheactiveRFcomponents, can be accounted for operationally. This calibration loopwill significantly reduce theneedfortheindependentcalibrationofchannel-to-channelphaseanddelayvariations.However,thephasecalibrationloopcannotincludethefeeds,antennas,orsomeofthepassiveelementsthatmayintroducephaseimbalance.However,sincetheuncalibratedpathsarenotactive,calibrationofstaticphase and delay differences between the channels should be sufficient to calibrate the paths notincludedinthephaseloop.Inaddition,thereareprovisionsfordownloadingtherawreturnsfromallchannelsforsmallsubsetsof the data. For nominal operations, only the returns from two channels (those that transmit andreceivefromthesameantenna)arerequiredfortheestimationofelevation.

2.2 FastSamplingPhaseThe calibration of static parameters in the presence of noise and varying parameterswill requireaveragingovermultipleobservations.Forthenominalmissionorbit,anygivencalibrationsitewillbevisitedonaverageonceevery11days.Acquiringasufficientnumberofsamplesforcalibrationwillrequireadelayofthenominalmissiondataflow,sincedataprocessingforscienceproductsrequiresthatthecalibrationvariablesbeavailable.(Noticethatthedatacollectedduringthecalibrationphasewill,inalllikelihood,stillbevalidformakingsciencedataproductsafterthecalibrationconstantsaredetermined).In order to expedite the calibration and error budget validation phases, the project has chosen tostart themissionwith a fast sampling phase thatwill significantly speed up the acquisition of thecalibrationoftheinstrumentanditsperformancevalidation.Thisfastsamplingphaseusesanorbitalaltitude that is only slightly different than the nominal altitude such that Cal/Val results andconclusions from the fast-samplingorbitwill generally carryover into thenominalorbit. Figure4shows sample coverage for the fast sampling orbit currently baselined by the project. The 1 dayrepeat timeof thisorbitallowsmuch fastercalibration than the21dayrepeat timeof thenominalorbit.



16

Figure4.FastsamplingphaseorbitcoverageduringtheSWOT1-dayrepeatphase.

Thefastsamplingphasealsoallowstheinvestigationofphenomenaoccurringattimescalessmallerthan the nominal mission sampling, which will benefit the determination of ocean submesoscaledecorrelationtimes,andthesynopticstudyoffloodwavepropagation,amongothers.Unfortunately,as is evident from Figure 4, the fast sampling can only be achieved at the cost of poor spatialsampling,andminimizingthecalibrationtimetotransitiontothenominalmissionphasewillbeanimportantconsideration.

2.3 AirborneMeasurementsA major challenge for the calibration and validation of SWOT is to obtain independent synopticmeasurementsbefore thephenomenabeingobservedchangesignificantlycomparedto thedesiredprecision. One way to obtain fast synoptic measurements over large scales is to use airborneplatforms. For SWOT calibration and validation, two airborne instruments are currently planned:AirSWOT,aKa-bandradarinterferometerdevelopedunderNASAtechnologyfundingandcurrently



17

funded and operated by theUS SWOTproject; andMASS, a lidar systemdeveloped byK.Melville,ScrippsInstitutionofOceanography(SIO),aSWOTscienceteammember(Melvilleetal.,2016).NotethatAirSWOThadbeenoriginallyplannedforbothoceanandhydrologyCal/Val.However,theoceanperformanceofAirSWOTwasfoundtobeinadequateforoceanCal/Valduetowave-bunchingeffects that had not been anticipated. AirSWOT is therefore no longer planned for ocean Cal/Val,althoughitisstillplannedforhydrologyCal/Valandotherpotentialpre-launchphenomenologyandriskmitigationpurposes. NotethattheeffectsofwavebunchingonSWOTaresmallandhavebeenincorporatedintotheSWOTmissionperformanceerrorbudget(D-79084revA).Materialsfromanindependent review of the AirSWOT ocean issues can be found in SWOT Docushare collection222168.

2.3.1 AirSWOTAt the core of the AirSWOT payload is a Ka-band interferometric radar (KaSPAR, Ka-band SWOTPhenomenologyAirborneRadar),whosemain characteristics are given inTable 1. AirSWOT radardataarecollectedoveran inner swath that coversSWOT-like incidenceanglesandanouter swaththat can map up to approximately 20-30˚ incidence angles, depending on surface reflectivity andaircraft attitude. AirSWOT radar data generally have finer intrinsic spatial resolution than SWOT.AirSWOTheightaccuraciesdependonanumberoffactorsbutaregenerallycomparabletoexpectedSWOTperformance(atanequivalentposting)forhydrologytargets.AirSWOTfliesonaNASAB-200SuperKingAiraircraftoperatedbyArmstrongFlightResearchCenter.

Table1AirSWOTInstrumentCharacteristics

Parameter Value CommentsNumberofantennas 5 Nominally used for 2 cross-track and 2 along-

trackinterferometryswathsPolarization V-pol ToenhanceSNRatfarrangeRangebandwidths 80MHz/400MHz 80 MHz for wide swath, 400 MHz for narrow

swathSwaths 4km/500m Typicalazimuthresolution 3m-5m Includes>30azimuthlooksTransmitPower 100W Platformaltitude 8000m Toprovidelongerwavelengthcorrectionsandpositioning,theAirSWOTpackageincludesastateoftheartInertialMotionUnit(IMU),includingahighprecisiongyroscopecoupledtoaGPSreceiver.The final component of the AirSWOTpayload is a color-infrared camerawith pixel resolution andgeolocationaccuracy<10mandaswathonthesameorderastheAirSWOTswath.ThiscameracanbeusedtovalidatetheSWOTwaterbodydelineationmeasurements.As of January2018, AirSWOT outer-swath height measurements have demonstrated performancethat is suitable for hydrology Cal/Val. Inner-swath performance validation has been hampered byinstabilitiesintheAirSWOTantennahardware,butplanstoexploitAirSWOTinner-swathdatatothe



18

extentpossibleremaininplace.Itshouldbenoted,however,thatCal/Valdoesexplicitlyrelyonanyparticularlevelofinner-swathperformance.In the pre-launch timeframe, theAirSWOT teamwill continue tomake improvements toAirSWOTgroundprocessingsoftwareinordertoimprovetheefficiencywithwhichdataareprocessed,makingthe processing lessmanually intensive andmore robust. Additional pre-launchAirSWOTanalyseswillalsosupportphenomenology investigationsthataresharedbetweenCal/Val,ADT,andScienceTeamefforts.

2.3.2 MASSLidarAdescriptionofthefullModularAerialSensingSystem(MASS)isgivenbyMelvilleetal.(2016):“Thecoreofthesystemforoceanwaveandseasurfaceheight(SSH)measurementsisaQ680iwaveformscanning lidar (Riegl,Austria)whichhasamaximumpulse repetition rateof400kHz, amaximum±30orasterscanrateof200Hz,andhasbeenusedataltitudesofupto1500mwithgoodreturnsforsurface-wavemeasurements.Thetheoreticalswathwidthoverwateristypicallyproportionaltothealtitudeoftheaircraft,anditseffectivewidthisalsodependentonthewindspeedandseastate.”Thesystemalsoincludesvisible,infrared,andhyperspectralcamerasaswellasaGPS/IMUsystem.AsofJanuary2018,theMASSsystemhasbeenmainlyflownonaPartenaviaP68aircraft.MASS lidarSSHmeasurementswillprovidedirectmeasurementsof theabsoluteSSH,which is thefundamentalphysicalquantitySWOTwillmeasure.TheMASSSSHmeasurementsthereforeprovideameansofdirectvalidationofSWOTperformance in termsofbotherrorbudgetanddataproductvalidation.Additionally,MASSflightpatternsthatcrosstheSWOTtrackcanvalidatethe2-Dnatureof the SWOT measurements, which is a new aspect of SWOT compared with traditional nadiraltimetry.ThesedatacanbeusedfortroubleshootingandanomalyresolutionaswellasSWOTphasescreen calibration. MASS measurements of directional wave spectra will allow for validation ofSWOTsignificantwaveheight(SWH)estimatesandrelatedphenomenology.Furthermore,MASSseasurface temperature measurements will provide additional insights into the phenomenology ofSWOTobservations.The MASS SSHmeasurements have been validated against the Jason-1 altimeter and have showngood agreement (see Figure 5) for the large (>150km) wavelengths that Jason-1 can resolve. Inadditiontothelidar,theMASSsystemalsohasaninfraredcamera,ahyperspectralcamera,avideocamera,andahigh-precisioncoupledIMU/GPSsystem.



19

Figure5. (Upperpanel)Heightabove theellipsoid measuredbyMASSand Jason-1. (Lowerpanel)Sameasabove,butaftersubtractingthemeanseasurface.

ValidationofthenoisefloorandtheheighterrorspectrumoftheMASSsystematshortwavelengthshasbeendemonstratedbya2016flightovertheAlgodonesDunes(seeFigure6).Inthisexperiment,theaircraftwasflownbackandforthovermanyrepeating,reciprocalpassesoveradesertdunefield.Variousfractionsofthelidarreturnswerediscardedtosimulatethedifferenceinalbedobetweenthedune field and an ocean surface. The dune field was chosen over a real ocean target for thisexperimentinordertoensurethatthesurfacedidnotchangesignificantlyoverthetimeoftheflight.The high-spatial-frequency noise floor of the MASS measurements meets SWOT Cal/Val needs,especially considering that a significant portion of the height error in Figure 6 is attributable tohorizontalerrorsover thesteepdunes; sucherrorswouldnotoccurover theoceanbecauseoceanwaveshavemuchshallowerslopes.

37

665

FIG. 10. Sea Surface Height (SSH), Sea Surface Height Anomaly (SSHA) vs latitude (N) measured 666

from Jason-I satellite altimeter and the MASS lidar. The dashed rectangle shows the location of 667

the subset of data plotted in Figure 11. 668

669

670

671

672

673

674



20

Figure6. MASS lidarheighterror spectra (black, red,gray-bluecurves)obtainedover theAlgodonesdunes,with theSWOToceanrequirementsandtypicalnadiraltimetrysignallevelsoverlaid.

DespitetheencouragingresultsfromtheMASSsystemdescribedabove,however,thesystemcannotbeconsideredfullyvalidatedandreadyforSWOToceanCal/Valpurposesbecausethesystemhasnotyetdemonstratedsuitableheighterrorspectralperformanceintherangeofoceanwavelengthsfrom15-150km,whichiskeyforSWOT.SpectraoverthisrangehavebeengeneratedfrompreviousMASSexperiments over the ocean, but they have generally been inconclusive, usually due to lack ofindependent truth data with appropriate spatial resolution for comparison. The range of the P68aircraftisalsotoolimitedtopracticallysupporttheselectedoceanCal/Valsites.ThislackofindependentdataforcomparisonposesthemaindifficultyinvalidatingtheMASSheighterror spectral performance to the shortest length scales. Ocean validation of theMASS system istherefore perhaps best accomplished through an examination of self-consistency of the MASSsystem’sownSSHmeasurementsacquiredonrepeating,reciprocalpasses.ThisapproachisdifficultwiththecurrentP68aircraft,however,becausetherelativelyshortrangeandenduranceoftheP68limitthenumberofpassesthatcanbecollectedatsuitablelengthscales. ThelowspeedoftheP68alsoimplieslongertimesbetweenrepeatingpassesandhencegreatertemporaldecorrelationoftheoceansurfacebetweenobservations.Thelimitedrangeoftheaircraftalsodoesnotallowittogetfarfromshore,where tidalvariationsarebettermodeledandareconsequentlymorereadilyremovedfromthedata.ForCal/Val,accesstothecrossoverdiamondsalsorequiresgreateraircraftrange.Note that temporal changes in SSH over the time of an aircraft flight (typically around 5 hr, butpossiblylongerorshorterdependingonaircrafttype)areproblematicnotonlyforMASSvalidation,but also for SWOT Cal/Val aswell. If the ocean changes appreciably over the time of the aircraftflight, only the airborne data collected close enough in time to the nearly instantaneous SWOT



21

overflight will be useful. (It would be neither logistically nor programmatically feasible to fly theaircraftmany,manytimestocollectonlyasingleusefulflightlineoneachsortie.)Becauseof the limitationsposedbyP68aircraft, thecurrentplan is tomove theMASSsystemtoafaster, longer-range aircraft. A NASA P-3 Orion aircraft operated by Wallops Flight Facility iscurrentlybaselined, thoughotherP-3optionsareavailable. Anextended-rangeTwinOtteraircraftwithlowerspeedthantheP-3isalsopossibleasadeeperfallbackoption.TheMASSsystemwillbeintegrated onto the new aircraft and validated during an ocean experiment at the primary(California) oceanCal/Val site in thepre-launch timeframe. This campaignwill validate theMASSsystem as intended for SWOT Cal/Val and demonstrate readiness for Cal/Val operations byaddressingthefollowingobjectives:

1. Demonstrate suitable height spectral error from theMASS system over the ocean at oceanwavelengths from 15-150km. This demonstrationwill encompass the performance of theMASS system itself, logistical and operational constraints (including weather, aircraftavailability,airspaceavailability,etc.),andthespatio-temporalvariabilityoftheoceansurfaceitself.

2. Characterize thespatialand temporalvariabilityof theoceanat theprimaryCal/Val site inordertohelpinthedesignofflightpatternsandsamplingapproachesbestsuitedforSWOTCal/Val.

3. Help establish the linkages between lidar-based SSH measurements and coincidenthydrographic in situ measurements of dynamic height (see Sect. 6.4) for the purposes ofsciencevalidation.

Assuming the successful validation of theMASS systemduring the pre-launch period as describedabove,thesystemwillbeflownattheprimaryoceanCal/Valsiteoveranintensivethree-weekperiodatthebeginningofthecalibrationphaseofthemissionandagainoveratwo-weekperiodneartheendofthecalibrationphase.ThefirstflightcampaignwillprovideinitialdataforSWOTcalibration,validation, and troubleshooting. The second campaignwill enable the validation of any on-boardupdatesmadeinresponsetoinsightsgainedfromthefirstcampaignandwillalsoallowforupdatedlidardataacquisitionstrategiesbasedonexperiencefromthefirstcampaign.While theMASS lidarwill be themain SSH validation tool, used over the principalUS Cal/Val siteduring the fast sampling phase, tests are to be conducted with a separate French airborne lidarsystem forpossibleuse at a secondocean siteduring the fast samplingphase (e.g.,MediterraneanSite).Ifconclusive,thiswouldenableSSHcalibrationandvalidationovermorediversedynamicsandsurfaceroughness(wind,wave)conditions.



22

3 KARINCALIBRATIONPLAN

3.1 CalibrationParametersDuringInstrumentCheckoutDuring the instrument checkout phase, the JPL Cal/Val and KaRIn System Engineering teamswillcollaborate in determining instrument parameters required before transitioning into the one-dayrepeatorbitsciencephase.Bytheendoftheinstrumentcheckoutphase,allparametersresident intheflighthardwareandusedforon-boardprocessingwillbecalibratedwithsufficientaccuracythatdata fromtheCal/Valandsciencephasescanbereprocessedonthegroundata laterdatetomeetperformance requirementswith updated, ground-derived calibration values. The parameters to becalibratedandthecalibrationmethodologiesarelistedbelow:

3.1.1 DifferentialRangeDelayChanneltochannelimagecorrelation.Seesection3.2.1

3.1.2 CommonRangeDelayCornerreflectorsites,comparisonwithnadiraltimeter,andradiometer.Seesection3.2.4

3.1.3 FunctionalValidationoftheOceanOnboardProcessorDuring the fast sampling phase, high resolution raw data will be transmitted to the ground forselectedoceanregions,togetherwithdataprocessedbytheonboardprocessorforthesameregions.Thesedatawillbeprocessedwithhardwareand“goldenmodel”softwaresimulatorsoftheonboardprocessor to validate its performance. The data will also be compared against averaged dataproducedbyanindependenthigh-resolutioninterferometricprocessorasanindependentfunctionalvalidationoftheonboardprocessingapproach.

3.2 CalibrationParametersDuringtheOne-DayRepeatPhase

3.2.1 DifferentialRangeDelayThisisthesimplestparametertocalibrate,asitcanbedonewithouttheneedforexternaldata.Theprocesstoestimatethedifferentialrangedelayistoperformrangecross-correlationmeasurementsbetween images of the ocean and to vary the relative range delay in the images until the crosscorrelation ismaximized.Theaccuracy for thisprocessdependson thenumberof scenesused forcross-calibration, and the expected accuracy easily exceeds 1/100 of a range pixel. This techniquewasdemonstratedinthecalibrationofSRTM(Farretal.,2007).

3.2.2 PhaseScreenThe phase screen can be estimated by comparison to SWOT-independent data or by SWOT self-consistencyapproaches.Considering SWOT-independent sources of truthdata sufficient forphase screen calibration,MASSunder-flightsare theprimaryapproach forestimatingorvalidating thephasescreen.A100kmby140 km swath is built at the time of a SWOT overflight, and the resulting topography mosaic isinterpolated to the SWOT along-track/cross-track swath coordinate system. This area may besubsampledbytheaircraftifneededduetocoveragelimitations,butalargeareaisrequiredinordertobeatdownSWOTnoise.Theadditionalerrorduetosubsamplingthetruthmustbeincorporatedinto the error analysis and flight pattern design. The resulting topography is subtracted from the



23

SWOTtopography,and, foreachalong-trackposition,aconstantbiasandlineartrendareremoved(since these are considered part of the static phase/roll biases of Sect. 3.2.3 by convention). Theestimatedbiasesandtrendsareused,togetherwithotherdata,asexplainedbelow,toestimatethestatic range and roll biases. The residual topographic variations for each swath can be related tophasevariationsbymeansoftheequationwhereδφ istheresidualphase,δh istheresidualheight,k istheelectromagneticwavenumber,B isthebaselinelength,andxisthecross-trackdistance.Noticethatthisequationdoesnotdependonthealong-trackpositionandone canaverage in thealong-trackdirection to reduce the randomheightnoise frombothKaRInandMASS.Thesubsequentprofile isaggregatedovermultipleunder-flightsand the final average is fitwith aChebyshevpolynomial of relatively loworder, toprotect againstover-fitting.Theresultingtwofunctions(oneperswath)constitutetheinterferometricphasescreencorrection,whichissubsequentlyappliedtobothlow-resolutionoceandataandhigh-resolutionlanddata.Theminimumnumberofflightsrequiredtomeetthephasescreenrequirementisdominatedbythenoise in theKaRInmeasurement itself. An estimate has been derived (Rodríguez and Chen, 2015)giventheneedtomeetthephasescreenhydrologyrequirement,whichleadstotherequirementforphase screen residuals to be less than 1.2cm, and the KaRIn performance estimated by systemengineering(D.Esteban-Fernandez,personalcommunication).TheresultsarepresentedinFigure7,which shows that a relatively small number of sorties will be required in order to provideappropriatecalibration.

Figure7.Minimumnumberofindependentlinesrequiredtoestimateorvalidateindependentlythephasescreencorrection.



24

Analternateapproachtoestimatingthephasescreenandstaticphase/rollbiasesistouseKaRInandaltimetercross-overs thatoccurwithinperiodsof less thanoneday,whichwillalwaysbe thecaseduringthefastsamplingphase. Inthatcase,thevariability inthecross-overisprimarilyduetotheuncalibratedphasescreen,witha low-wavelengthrandomcontributiondue to troposphericdelaysandpotentiallylow-wavelengthEMbiasesorinternaltideseffects.RodríguezandChen(2015)haveshownthatitispossibletoestimatethephasescreentomuchgreateraccuracythanthatrequiredbythehydrologyphasescreenrequirementbynon-parametricinversionofphasescreenfromtheone-daycross-overdifferences.Theaccuracyof the inversionwillvarywith thesignificantwaveheightand the spatialvariabilityof thephase screen itself.Asaworst-caseexample,Figure8. shows theestimatedphasescreenerrorsforasignificantwaveheightof6mandhighfrequencyphasescreenvariationsbasedonretrievingthephasescreenusingasinglecross-over.Itisclearthatthismethodwillbemuchmoreaccuratethantheairborneinstrumentphasescreenestimates,althoughitisnotindependent of theKaRIn data itself. The nominal planwill be to use a number of cross-overs forphasescreencalibrationandvalidationwillbedoneusingamixtureofairborneflights(independentvalidation)orcross-overestimatesnotusedtoderivethecalibrationparameters.The phase screen can also be estimated from SWOT data only along with oceanmodel estimates(Dibarboure,2016). ThedisadvantageofSWOT-onlyestimationapproachesforthephasescreenisthatmanytypesoferrorscanmasqueradeasphasescreeneffectsbutwouldresultinunstablephasescreenestimates. Thisisbecauseofthemanydegreesoffreedominthephasescreen,whichallowforoverfittingofothererrorcontributionsthatmayvaryintimeorwithdependenciesdifferentthanthephasescreenitself. TheuseofindependentdatahashistoricallybeenhelpfulindiagnosingandresolvingsuchissueswhencalibratinginterferometricSARsystems.At theendof thephasescreencalibrationexperiments,onewillalsohaveanestimateof thestaticphase/rollanglebiasforeachswath.Thisestimatewillbeaveragedtothatobtainedfromthecross-overcalibrationdescribedbelow.



25

3.2.3 Cross-OverCalibrationforStaticPhase/RollBiasesWhile the MASS under-flights will produce estimates of the static phase/roll bias along with thephase screen, the accuracyof the estimateswill be limitedbyhaving to averageoverdynamic rollerrors that will be present simultaneously. Additional information regarding this bias can beobtained and incorporated by assuming that, after correcting for tides and high frequencyphenomenon (pressure and winds) but also sea state bias, the ocean surface does not movesignificantly between ascending and descending passes at orbit cross-overs so that theinterferometricphase/rollerrorsatthecross-overdiamondscanbeestimatedfromtheKaRINheightdifferences(FuandRodriguez,2004). (Notethattoperformthephase/rollbiasestimates, it isnotnecessarytousethenadiraltimeterdata.)SincetheKaRINphasebiaswillbecommonbetweenascendinganddescendingpasses,itwillcancelwhen calculating the cross-over differences. However, therewill some differences in the elevationmeasurements due to changes in the wet troposphere correction and the sea surface height. Thetropospheric errors are partially mitigated by application of the radiometer correction, althoughsomerandomcross-trackvariationsmayremain.Thecontaminationduetotheevolutionoftheseasurfaceheightispartiallymitigatedbyusingcross-overs during the 1-day fast sampling phase, when the time difference between ascending anddescendingpasseswillbelessthanoneday.Toassesstheimpactofoceanmotion,weusethehighresolution ECCO2 oceanmodel (seeMenemenlis et al., 2008 for a review) of theNorthAtlantic tosimulatetheaccuracyofphase/rollerrorretrieval(seeFigure9).

Figure8.ResidualestimatedbiasesandstandarddeviationsforphasescreenretrievalbasedonaMonteCarlosimulationofphasescreenretrievalfor6mSWHassumingonly1cross-overisused.



26

Toobtainboundsfortheproblemoftheeffectsofoceanmotion,theretrievalaccuracyisestimatedincluding instrument errors, but no surfacemotion, and assuming that the nominal SWOT 21-dayorbit determines the cross-over revisit time. The results of this simulation for the residual heighterror after phase/roll error corrections are presented in Figure 10. Clearly, the biases can beretrieved with more than sufficient accuracy in the absence of ocean motion. Given the typicaltemporalcorrelationtimeoftheoceansurfacemesoscalecirculation,whichisontheorderof20days(see,e.g.,LeTraonetal.,1998),weexpecttheresultsfromthefastsamplingphasetocloselymatchtheseresults,especiallyifonestaysawayfromareasofsignificantmesoscaleactivity.

Figure9.MesoscalevariabilityfromAVISOmaps.



27

Figure 10. Probability density function (upper panels) and cumulative distribution function (lower panels) of height errorsaftercorrectingforunknownphase/rollerrorsinthecasewherethereisoceanmotion(leftcolumn)andinthecasewhenthenominalorbitcross-oversetisused(rightcolumn).Inthecasewhennomotionispresent,almostallerrorsarebelow0.03cm.Whensignificantoceanmotionispresent,68%oftheerrorsare<1.5cm,80%oftheerrorsare<2.0cm,and90%oftheerrorsare<5.0cm.

Thecrossoverestimates for thenominalmissionwillbeoptimallymergedwith thedata from theonboard gyro, using the known correlations and variances for gyro, phase, and sea surface height.Theseestimateswillbeusedduringthenominalmissiontofurtherreducethedynamicvariabilityoftheroll/phasebiases.Asanexampleofthebenefitsofusingthiscross-overinformation,ordynamicocean calibration, for continuous dynamic calibration, Figure 11 presents the results of optimalmergingoceanandgyro information to improve theroll correctionover landsignificantlyover theresult that could be obtained using the gyro information alone. Significantly better resultswill beobtainedovertheoceanduetothedensityofcross-overpointsandtheshortalong-tracktemporalseparationbetweenthem(althoughmeetingthemissionerrorbudgetdoesnotrelyonthisprocess).

Figure11.Expectedresidualphase/rollerrorsoverlandincentimetersforthescienceorbit.

3.2.4 StaticRangeBiasesThe problem of absolute range calibration is one of the most challenging for both conventionalaltimetersandradar interferometers. In fact, forallaltimetersystemstodate,adhoc constantbiascorrectionshavebeenrequiredtoensureconsistencyforthemeasuredseasurface.Therefore,forSWOT,ratherthanrequiringthattheabsoluterangebemeasuredprecisely,werequirethatthemeanseasurfaceproducedbeconsistentwiththatproducedbythehistoricalTopex-Jason1-Jason 3 climate data set, which has been cross-calibrated with data collections obtainedsimultaneously.Both thenadir altimeter andSWOTneed tobe calibrated to eachother and to thereference altimeterdata set, including thememberof the Jason series operating at the timeof theSWOTlaunch.(Notethatifnosuchsatelliteexists,theprocessoutlinedbelowwillstillinsurethattheSWOTaltimeterandKaRINareatleastconsistentwitheachother).As a first step, we cross-calibrate KaRIN and the nadir altimeter on SWOT. Rather than using theascending-descending cross-overs, which are contaminated with residual errors due to temporaldifferences inwet troposphere and sea surface dynamics,weuse along track data collected at thesame time to obtain this calibration. The KaRIN sea surface height (SSH) field is optimallyinterpolatedacrossthe20kmnadirgapandtheresultingelevationsarecomparedtothealtimeterheight estimates. Since the ocean SSH spectrum decays quickly with decreasing spatial scale, one



28

expects thecontributiondue to interpolationerrors tobezero-meanandsignificantlysmaller than1cm, especially over regions of lowmesoscale activity. Thus, by averaging the interpolated heightdifference over the entire fast-sampling phase, one obtains an average range difference that issignificantly smaller than that required by the SWOT error budget. This process can be furtheroptimizediftheoptionalKaRINnadirchannelisimplemented(byaddinganadirlookingreceivertotheradiometerantenna),asthiswouldallowdirectcomparisonsofrangedelayatnadir,withouttheneedforspatialinterpolation.Obtainingconsistencywiththereferencealtimeterconstellationmustrelyoncross-overdata,whichis not collected simultaneously and will be contaminated by wet troposphere and SSH dynamics.However, assuming slow drifts for the range bias for both systems, an appropriate accuracy canobtainedbyaveragingtheobserveddifferencesoverasuitabletime(months).

3.2.5 RadiometricCalibrationRadiometric calibrationof SWOTwill beobtainedby comparing radar cross section for SWOTandAirSWOTwhen coincidentmeasurements are available at the same set of incidence angles. SWOTestimates of reflectivity will also be compared to model estimates based on wind speeds fromweather models and buoy data. The radiometric calibration will also be informed by examiningcorner reflector data, although discrete-target effects may add some uncertainty to the cornerreflector reflectivity estimates. Transponders can be considered as an alternative to cornerreflectors.



29

4 NADIRALTIMETERCALIBRATIONPLANRequirementsontheperformanceoftheSWOTnadiraltimeterareverydemanding,notablybecausethedriftoflong-wavelengthsischallengingtovalidatewitha3-yearmission(asopposedto10+forJason-class). Many instrumental features will be checked (functionally) and characterized notablywith respect to the ground acceptance test measurements. It concerns mainly: the trackingcapabilities,theshapesofpointtargetresponse(PTR)andthelowpassfilter(LPF),thevaluesofCNGattenuators.Theimpactsofthevariousconfigurationsofthealtimeterperformancewillbeevaluatedindetail.Thehighlevelrequirementsforthein-flightassessmentofthealtimeterwillbeprovidedbytheCNESteaminchargeoftheinstrument.Inthefollowingsections,themainSWOTnadiraltimeterfunctionalitiesthatwillbecarefullycheckedarerecalled.

4.1 TrackingmodesThetrackingmodesareinheritedfromthePoseidon-3altimeterofJason-3:● Closeloop:theacquisitionmodeismedianorDiode,andtheonlytrackingmodeismedian.● Open loop: the acquisition and trackingmode is called Diode/DEM (Digital ElevationModel).

Note that itwillbepossible toevaluate theDiode/DEMmodeevenwhen it isnotoperatingonboard.Thetelemetryactuallycontainssimultaneously,themediantrackerdataandtrackerdatacomputedwiththismode.Thispossibilitywillallowacompletecrosscomparisonofthemodescharacteristics.

Tracking modes will be validated during the Cal/Val phases. The objectives are to check theoperability and the performances of each trackermode. Typicalmetrics are data availability, datacoverage andglobal altimetricperformances in a classicalCalValmeasurementsquality sense.Thetrackingmodeswillbeassessedoveroceanbutalsooversea/landtransitionsandoverhydrological,sea-iceandlandiceareas.Before the end of the assessment phase, the project team will have the necessary metrics andcomparison studies between these modes to be able to select the “nominal tracker mode”. ThevalidationplanfortheDiode/DEMmodeisdetailedinthefollowingsection.

4.1.1 TheDiode/DEMtrackingmodeThis tracking mode has been already implemented on the Poseidon-3 altimeter (on-board Jason-2&3), on SARAL and on the Sentinel3missions.Moreover, Poseidon-3Cwill have the capability toswitchfromtheautonomoustrackingmodetotheDiode/DEMcoupledtrackingmodeautomaticallydepending on the actual position on the orbit. This automatic transition can be activated ordeactivatedbygroundtelecommand.

Thetarget’spredicteddistance iscalculateddirectlyby thealtimeter,combiningaltitudedata fromDORIS/Diode with the altitude from a pseudo Digital Elevation Model (DEM) recorded in thealtimeter'sonboardmemory.Dependingonthequalityof thisDEM,thepositioningaccuracyof thereturnechointhealtimeter'sreceivingwindowisoftheorderofafewmeters.ThecombineduseofDiode data and the altitude from the DEM enables the position of the reception window to becontrolleddirectly,whichinturnenablesanytargettobetrackedindependentlyofthetypeofreturnecho.Thismodeisthereforeveryusefulwhentrackingoverareasofspecialinterest,suchasriversandlakesandcoastalareas.TheonboardDEMisaseriesofwateraltitudeswithrespecttotheDiodegeoid,sampledataconstantangular step along the satellite path (0.01°) and corrected from the mean atmospheric and



30

ionospheric delays. Since the projection of the radar spot on the ground covers a circular area ofabout 8km of radius, the radar signalmay be composed of bothwater and land contributions incoastalorlake/riverareas(Figure12).Asthemissionobjectiveisthemeasurementofwatersurfacealtitudes, suchDEMpoints are assigned towater points and the correspondingDEM altitudes arerecordedasthealtitudeofthenearestwaterpoint(Figure13).Thisenablesthealtimetertofollowthewaterareasbeforeandaftertheyhavebeeneffectivelyoverflownbythesatellite.

Figure12.DEMsamplingalongtheorbitpath.

Figure13.DEMsamplingalongtheorbitpath.Extensionofareasofinterest(water).

TwocodingalgorithmsareusedfortheonboardDEMstorage:

● Absolutecoding:Successivepointsofaltitudevariationswithinagiventhreshold(±2metersoverwatersurfaces)aregatheredintosegmentsofsamealtitude(2bytesforthealtitudeofthefirstpoint+2bytesforthenumberofpoints inthesegment).Theoceanandtheinlandwaters(flatsurfaces)arecodedthisway.

● Incrementalcoding:Successivepointswithhigheraltitudevariationsarecodedasfollows:Altitudeofthefirstpoint(2bytes),thenaltitudeincrements(1byteeach).

Astheonboardmemoryislimited,thewholeDEMcannotbeuploaded.Thestrategyistocodeonlythewell-knownhydrologicalsurfaces,intermsoflocationandelevation.ItwillbepossibleduringthemissionlifetimetoupdatetheDEMbyaddingotherhydrologicalsurfacesofinterest.Theoceanhasalsobeencoded.

4.1.2 TheDiode/DEMtrackingmodevalidationSince theDEMaltitudes recorded in theonboardmemoryaredependenton the satelliteorbit, theprerequisiteofthisplanisthecomplianceoftheactualsatelliteorbitwiththenominalorbit.

Ocean

Land

Altimeter



31

Performanceassessmentindices:TheDEM trackingmode performancewill be assessedwith two indices giving information on thedataavailabilityandthedataaccuracy:

Data availabilitywill be assessed by comparing the number ofmeasurements acquired using theDEM tracking mode with the number of measurements acquired over the same area using theautonomoustrackingmode.A“dataavailabilityindex”combiningthesetwonumberswillbedefinedinordertogiveanimmediateassessmentoftheacquisitionperformanceoftheDEMtrackingmode.Data accuracy assessmentwill consist in studying the location of thewaveforms in the receptionwindow. Especially, one will check that the waveforms are statistically centered in the receptionwindowandbounded in the authorizedextension (±10metersoverwater surfaces).As geographicinformationwill be simultaneously availablewith the altitudemeasurements, itwill bepossible toidentifyareaswherereceptionsignaldonotcomplywiththeserequirements.PerformanceassessmenttargetsAsthealtimetercannotrunsimultaneouslytwotrackingmodes,datawillbecomparedfordifferentacquisitioncycles(e.g.acycleinclassicaltrackingloopandacycleinDEMtrackingmode).However,it has to be noticed that when the altimeter operates in closed-loop tracking mode, the trackingcommandthatwouldbeappliedinDiode/DEMmodeisavailableintheproduct.Asaconsequence,thetwotrackingcommandscanbecomparedatthesametimeandthecenteringofthewaveformcanbeassessed.

Furthermore,inordertomakevalidationresultsmorerepresentativeandaccurate,itisproposedtoperformthevalidationstudybyconsideringfourdifferentregiontypes:

● Overdeep-seaareas, it shouldbeproved that theDEMtrackingmodeprovidesat leastasmanymeasurement as the closed-loop trackingmode and that thesemeasurements are atleastasaccurateasthemeasurementsacquiredwiththeautonomousopen-trackingloop.

● Since the onboard DEM does not take into account temporal effects such as tides andvariations inpath-delay corrections, an important validationpoint is the assessment of thedecrease of the tracking performance in coastal areas. But one skill of the DEM trackingmode is that the reception window is always close to the measured surface whereas theautonomous tracking mode is usually unable to follow sea surfaces at distances less thanabout 5-10km offshore. Over coastal areas the same “data availability” indices as thosedefinedoverdeep-seasurfaceswillbeconsidered.Twocaseswillhavetobeaddressed:

o Areasoflowtidaleffects:Theseareaswillbeofinteresttoassesstheincreaseofdataacquiredwith theDEMtrackingmodewithrespect to thenumberofmeasurementsacquiredinclosed-looptrackingmode.Theseregionswillalsobeusedtocheckthattheslantmeasurementoftheseasurfaces,asdescribedin“§TheDiode/DEMtrackingmode”,isefficientduringsea-landandland-seatransitions.

o Areasofhightidaleffects:Overtheseareas,theremightbelessdatathanoverareaswhere lowtidaleffectsoccur.However,since theDEMtrackingmode isdesignedtocenterthebackscatteredsignalinthereceptionwindow,somemeasurementswillbeexploitable if they are not too far from the tide-free altitudes recorded in theDEM.Therefore, it shouldbepossible to compute the ‘data availability’ index and then toassesstheperformanceoftheDEMtrackingmode.



32

● TheDEMisintendedtoenablealtimetrymeasurementsoverriversandlakes.Thevalidationprocess will be conducted in two steps and by considering each lake and each riverindividually:

o Dataavailability:TheindexdefinedforthevalidationoveroceanareaswillbeusedtoassesstheimprovementofdataacquisitionintheDEMtrackingmode.Furthermore,the DEM tracking mode for slant measurements will have to be assessed inland/watertransitions.

o Data accuracy: coherence ofmeasurements acquired in the two trackingmodes bycomputingthebiasandthescatteringofthedifferencesofaltitudeswillbechecked.Incaseofevidenceofanexistingbiasinthemeasurementforoneorseverallakes/rivers(e.g.Wave formnot centered in the receptionwindow),onewillpropose to correctthecorrespondingaltitudesrecordedintheDEMandtoproceedtoanupload.

4.2 InternalCalibrationsThealtimeterprovidesmeasurementsbetweentheoverflownsurfaceandthephysicalpointwherethederampprocessisappliedinsidetheinstrument.Thealtimeterrangeisthencomputedbetweenthe antenna and the ocean surface. Themeasured range has consequently to be corrected for theinternal groupdelay that is computedpreciselyongroundbefore launch.Ofparticular importanceare the group delays introduced by the duplexer and the antenna. The measured delays will betreatedascorrectionsinthegroundprocessing.However,apartofthegroupdelaycanbecomputed(andmonitored)thankstothepointtargetresponse(PTR)measurement.Itwillthusbepossibletocontinuouslyupdatethiscontributionafterlaunch.Two internal calibrationmodes are implemented in the Poseidon-3C instrument (as for Poseidon-3B):

● The first mode (CAL1) gives themeasurement of the instrument point target response byfeedingthesignalfromtheemissionchannelbacktothereceiverchannel.

● Thesecondmode(CAL2)givesthetransferfunctionofthealtimeterreceivingchain.These calibration modes (with various parameters configurations) are fully characterized beforelaunchduring thegroundacceptance test.During theassessmentphase,acompletesetofscenariomust be played in order to functionally validate a large set of possible configurations of thecalibration parameters. The objective is to guarantee the good operation of the instrument and tocharacterize the various features of the calibration responses. Then, and for the whole life of themission,thetwomodesofcalibration(inthesameconfigurations)willbeactivatedseveraltimesaday by telecommands (3 times for OSTM/Jason-2, depending on the calibration stability) with adoubleobjective:✓ ThefirstobjectiveistocontinuouslycharacterizetheshapesandpositionsofthePointTarget

Response and Low Pass Filters in order to daily introduce updated corrections in thealtimeter processing chains that generate the level 2 products. Themain corrections usingcalibrationresultsarethefollowing:

o correctionofthewaveformsbythelowpassfilterbeforebeingretracked,o computationofthetotalpowerofthePTRinordertocorrectthesigma0estimation,o computationof thedifferenceof internalpathsbetween theemissionand reference

channelsinordertocorrecttherangeestimation.✓ The second objective is to ensure an unceasing monitoring of the instrument, in order to

monitortheelectronicsagingandtocheckthegoodhealthoftheequipment.



33

Acomparisonbetweenmeasurementsbeforeandafterlaunchwillbeperformedtoquantifypossibleevolutionsduetothelaunch.CNG(NumericalGainCommand)Inthealtimeterreceivingchain,twonumericalgaincommands(CNG)arepresentinordertoadjustthe amplitude of the echo return to a nominal value. In the ground processing, the sigma naughtcoefficient isdeterminedthanksto theestimatedpowercomputedbytheretrackingalgorithmandthenumerical gain commands thatwereusedonboard. It is thennecessary topreciselyknow therealattenuationvaluethathasbeenapplied.TheCNGvalueshavebeenmeasuredduringthegroundacceptancetest.However,theyareopentodriftwithtimeduetotheagingofthecomponents. It isthusnecessarytoregularlycalibratetheCNGthankstoamethodbasedontheanalysisofasetofPTRmeasurements. This calibration will have to be done during the assessment phase and repeatedregularly (every3or6months,dependingon thecalibrationstability)during thewhole lifeof themission.SimulatorofperformancesDeveloped first in the frame of the SSALT Poseidon-1 altimeter onboard TOPEX, a simulator ofperformances has been consolidated and updated with characteristics and new functionalities ofPoseidon-2 and Poseidon-3 altimeters (new trackingmodes, updated instrumental characteristics,point target response and low pass filter, CNG tables, and various hardware characteristics). Thissimulator (validated notably by the good agreement between the ground tests performance ofPoseidon-2instrumentandsimulationresults)willplayagainanimportantroleduringthePoseidon-3C assessment phase. Performances computed with the pre-launch Poseidon-3C acceptance testmeasurementswillserveasreferencesforthein-flightassessmentphase.ThissimulatorofperformanceswillalsobeusedtogeneratetheLookUpTables(LUT)corrections(forrange,significantwave-heightandsigmanaughtcoefficient),beforeandafterlaunch.ThoseLUTarerequiredtoaccountforthepotentialgroundprocessingapproximations(forexampletheMLE4canbeimplementedwithanapproximationofthePTRbasedonauniqueGaussianfunction).AltimetermeasurementsMeasuredaltimeterparameterswillbeevaluatedafterlaunch.Firstofall,thescienceparameterswillbestudied:e.g.,range,SWH,backscattercoefficient,mispointingangle,andwaveforms.Thesestudieswillincludenoise-levelestimatesusingFourierTransformanalysisaswellascomputationofalong-track statistics (mean and standard deviation) over the ocean and other surfaces. Histograms anddispersion diagramswill also be computed for these parameters. The resultswill be compared toequivalentresultsfrompreviousPoseidon-3altimeters.



34

5 WATERVAPORRADIOMETERCALIBRATIONPLANThe SWOTmicrowave radiometer provides an estimate of thewet tropospheric path delay in thecenter of each KaRin swath. The radiometer system consists of two independent three-frequencyradiometersfeedingashared1-mreflector. TheradiometerdesignisbasedontheJasonAdvancedMicrowaveRadiometer.The radiometer gain and offset are calibrated at a plane internal to the radiometer using a noisesourceandDicke switch toa reference load. Thegainandoffset are referenced to the instrumentinput by correcting for the loss and self-emission in the RF front-end components outside of thecalibration loop. The internalreferencesandthe front-endpath lossarecalibratedpre-launchandtunedon-orbit.

5.1 BrightnessTemperatureCalibrationThe AMR brightness temperatures (TBs) will be calibrated to on-Earth brightness temperaturereferences. Theon-Earthreferencesareaso-calledvicariouscoldreference(Ruf,2000),whichisastatistical lower bound on ocean surface brightness temperature andpseudo-blackbody regions intheAmazonrainforest(BrownandRuf,2005).ThesereferenceshavebeenusedforthecalibrationofallpreviousNASAaltimeterradiometers.DuringtheinitialCal/Valperiod,dependenciesofthecalibrationoninstrumenttemperaturewillberemoved by sampling the TB references as a function of the AMR thermistor measurements andreducing the slope to zero. After that, anAdvancedRadiometerCalibrationSystem(ARCS)willbeused on a continuous basis during the mission to facilitate the long term calibration. ARCS wasoriginallydevelopedforJason-2andJason-3,andwillbedevelopedforSentinel-6andSWOT.ARCSuses the comparisons to the on-Earth references to both monitor and correct the long termcalibration.The on-Earth referenceswill bemainly used to set the radiometer absolute TB calibration and toidentifyandremoveanyresidualinstrumenttemperaturedependenterrors.Toidentifyandcorrectforotherpotentialsystematicerrorsthatarespatiallyortemporallycorrelated,comparisonsoftheradiometertomodeledTBsandmeasurementsfromotherradiometersensorswillbeused.ModelTBsaregeneratedusingnumericalweatherpredictionmodel fieldsanda radiative transfermodeltosimulatewhatthesensorshouldbeobserving.Inter-sensorcomparisonsareperformedbyfinding co-incident match-ups and deriving AMR equivalent TBs from the other sensors TBobservations. Both the model TB and inter-sensor TBs have about a 2-3K uncertainty for anindividualmatch-up. But, thisuncertainty isGaussiandistributedandaveragesdownwith a largeenoughsampleset.Typically,comparisonsover1monthreducetheuncertaintyinthecomparisontothe 0.1K level. These match-ups are averaged spatially and temporally to identify errors in theantennapatterncorrectionalgorithm,whichappearasspatiallycorrelatederrors.

5.2 Inter-beamCalibrationBecausetwoindependentradiometersareusedtoderivethepathdelayslopeacrosstheswath,itisessential that they arewell inter-calibrated. Statically, thedifferencebetween the two radiometermeasurements,whichareseparatedbyabout80kmacross-swath, canberepresentedasa randomprocesswithazero-meanGaussiandistribution. Thismeans thatgivena largeenoughsampleset,any deviations from zero represent real calibration errors between the sensors. Inter-beamdifferenceswill be compared in a numberofways to assess the inter-calibrationquality. Monthly



35

mapsof thedifferenceswillbeused toassessgeographically correlatederrors. Scatterplotsof theTBswillbeusedtoassessrelativedifferencesintheslopeandoffsetofthecalibration. Finally,thedailyaveragedglobalmean inter-sensordifferencewillbeused tocalibraterelativedriftsbetweenthetworadiometers.



36

6 SWOTVALIDATIONPLAN

6.1 Introduction

6.1.1 OverviewIn this chapter,we describe the overall performance validation plan for ocean (Sects. 6.2-6.4) andsurfacewater(Sects.6.5-6.7)products.In Section 6.2 we describe how each contributor to the ocean error budget will be validatedindependently (e.g. to confirm that themeasurement system, including instrument, spacecraft, andprocessing,behavesasexpected)andsection6.3tacklestheoverallvalidationoftheoceanproduct,notably in terms of wavelength decomposition (as per the science requirements). Section 6.4addressessciencevalidationofoceandynamicheights.InSection6.5,wedescribehoweachcontributortothesurfacewatererrorbudgetwillbevalidatedindependently (e.g. to confirm that themeasurement system, including instrument, spacecraft, andprocessing,behavesasexpected)andsection6.6 tackles theoverallvalidationof thesurfacewaterproduct.Lastlysection6.7outlinestheplantocharacterizetheriverdischargeparameters,asperthesciencerequirements.

6.1.2 MinimumCalValengagementTheCalVal activities for themissionwill be jointly financedby theNASAandCNESProjects.CNESengagestoactivelyparticipateintheSWOTCalValactivitiesdetailedinthisdocument,includingtheglobalstatisticalCalValandparticipationincertainin-situCalValsites,asithasdonewithallofthepastNASA/CNESaltimetricmissions–Topex/Poseidon,Jason-1,-2,-3series,etc.

6.2 SWOTOceanErrorBudgetValidationIn this section, we describe how each individual component of the SWOT error budget will bevalidated.ThevalidationofNadiraltimeterdatawillinheritfromthemethodsusedforconventionalaltimetermissionssuchasJason-3.TovalidateSWOTdataglobally,threemethodswillbeused:

● Statisticalanalysisofasingledataset(Nadiraltimeter,radiometerorKaRIN)● DifferentialanalysiswithoverlappingmeasurementsfromNadir,andKaRIN● Comparison to external assets: remote sensing products, in-situ data, airborne data, or

models

6.2.1 RandomheighterrorvalidationThepurposeofthissectionistovalidatetherandomheighterrorthatisdominatingtheerrorbudgetat the highest frequencies. This is notably the limiting factor for Sea Surface Height (SSH)observabilityforwavelengthsrangingfrom1to30km.

6.2.1.1 Nadiraltimeterrandomheighterrorvalidation

As for the Jason-classmissions, thenadiraltimeter randomnoisewillbe inferred from theplateauobserved on power spectral densities (PSD) and/or from the variance of high-pass filtered SSHmeasurements. Geographical variations and themodulation of randomheight errors by significantwave height (SWH) will be analyzed with global maps and statistics (e.g. PDF, relationship with



37

SWH). Lastly the temporal variations (seasonal, inter-annual, drifts) of the random error will bemonitoredthroughoutthemission’slifetime.

6.2.1.2 KaRInrandomheighterrorvalidationTovalidatetheKaRINrandomheighterrors,thesamemetricswillbeusedasforthenadiraltimeter.TheglobalmeanandregionalPSDwillbeanalyzedtoinfertheenergyoftherandomnoiseplateau,andtoderivetherandomheighterror,aswellasitsgeographicalandtemporalvariations.Furthermore, specific cross-track classifications will be carried out to infer the cross-trackdependencyoftherandomnoise(perFigure14),aswellasitmodulationbyseastate(e.g.surf-boarderror).AdditionalexternalanalyseswillbeperformedusingWAMandWaveWatch3modeloutputstoquantifytheinfluenceofswellamplitudeandorientationwithrespecttothesatellitetrackontherandomheighterror.By launch, the mean sea surface (MSS) and geoid models are unlikely to resolve 1-km features.Consequently, residual geoid error will be present in SSH anomalies from SWOT. The MSS/geoiderroristhesameofallSWOTpixelsinagivenlocationthereforedifferentialanalyses(cross-over,orcycle-to-cycle during the1-dayphase)will be used to gauge the SSHvariance cancelled out in thedifference and its relationshipwithMSS/geoid signatures (orMSS/geoid formal errormaps). Thisapproach will help separate the systematic high frequency error (geoid) from the random KaRINerror.

Figure14.Swathandnoisevariation.Thefullwidthofthesatelliteswath,from-61to+61km,ismeasured,withanominalgapof ±3 km around nadir. However, the data quality is degraded outside of the nominal 10-60-km range. KaRIn randominstrumentnoisevariesacross theswath,as illustrated in the figurebelow,withaswath-averaged(10kmto60km)heighterrorof2.4cmforSWH=2mfora1km²pixel. (from JPL D-79084)

Inadditiontothissystematicvalidationcarriedoutglobally,more localanalyseswillbeperformedontheso-called9-beamexpertproductinordertovalidatethecross-trackvariationsoftherandomheighterroroneachbeam.TheseadditionalanalyseswillhelpdeterminehoweachbeamcontributestotherandomerroroftheLevel-2combiningSSHcontentfromthe9beams.

6.2.2 Roll/phasedriftvalidation



38

The error budget of KaRIN is primarily controlled for wavelengths shorter than 1000 km, asdiscussed inmissionperformanceanderrorbudget (SWOT_D-79084). In contrast, long termdriftsand biases are beyond the ocean science requirements. The longer wavelengths and drifts arehowever relevant for hydrology. To that extent, the LR processing chain includes an empiricalcalibrationoftheKaRINrangedriftthatisprovidedasacorrectioninLow-Resolution(LR)andHigh-ResolutionHRproducts.Thiscrossover-basedcorrectionwillbeadirectmeasurementofthelongerwavelengths(>1000km)ofroll/phase/baselineerrors.Itisthereforeimportanttomonitorthiscorrectioninthelongrun(e.g.daily or cycle averages, cyclic maps, regional trends throughout the mission lifetime), to try anddetect contaminationwith other phenomena (e.g. verify relationshipwith in-orbit parameters, seastateoratmosphericconditions)andtoanalyzeextremeandsuspiciousevents.To gauge the error bar of this correction, it is possible to compare the product corrections withindependentdriftestimatesfromotheralgorithms(e.g.directandsub-cycle,crossoveroverlapswiththealtimeterconstellation,asperDibarboureandUbelmann2013).Similarly,external in-situsites(likebiglakes)willbeusedtocheckthecross-tracktopographyprofiles,theseprofilescanbeusedtoinferthelongtermdriftoftheroll/phase/baselinesystematicerrors(seesection6.5.5).

6.2.3 PODValidation

6.2.3.1 OverviewThe precise orbit determination (POD) verification activity will rely on a cooperative

investigation among project POD teams (at CNES and JPL) and PIs investigators (GSFC, ESOC…)workinginthisarea.CNEShastheresponsibilityforproducingthepreciseorbitestimatesthatwillbeincludedinthesciencedataproducts.TheCNESPODverificationeffortwilltakeadvantageofallavailable trackingdata toproduce, on a routinebasis, an estimateof theorbit error, aswell as anevaluationoftheperformanceofthetrackinginstruments.

Themethodsdevelopedtoverify theaccuracyofTOPEX/Poseidon, Jason-1&2&3orbitswillbe extensively used for SWOT. The achievement of the radial accuracy has been confirmed for allnadir missions (refer for example to the OSTST report chapter 6.4 Precise Orbit Determination :http://www.aviso.altimetry.fr/fileadmin/documents/OSTST/2014/OSTST_2014_Meeting_Report.pdf) and the mainfocushasnowmovedtotheassessmentofthelongtermcoherenceoftheorbitsandontheimpactofgeographicallycorrelatederrorsonboththeglobalandregionalMeanSeaLevelestimates:thesenewobjectives arebeyond the scopeof the SWOTscienceobjectivesbut they are relevant in the sensethattheycontributetothelongwavelengthSSHerroronoceanandontheabsoluterangeerrorforhydrology.

6.2.3.2 TwotypesofPODvalidationThemostcritical issuesconcern thestabilityof thereference frameused toprocessDORIS,

GPSandSLRtrackingmeasurements,theaccuracyandfidelityoftheforcemodelsthatunderpinthePODcomputationsandtheoverallqualityoftheavailabletrackingdata.

The verification activities will be conducted both during the orbit production process(operationalverification)andafterwards(expertverification).Thegoaloftheoperationalverificationis to ensure that the orbits meet mission accuracy requirements. The operations team performsoperational verification during the production of the orbits, and results are summarized in the



39

verificationreport,whichisprovidedalongwiththeorbit.TheprojectPODteamanalyzestheresultsoftheverificationandauthorizesthedeliveryoftheorbit.Theexpertverificationfocusesonamoredetailedunderstandingofthenatureoftheorbiterror,andof its impact on the end users. It includes long termmonitoring of the orbit quality, especially toenabletheearlydetectionofpotentialdrifts.ThisverificationisperformedbothbytheprojectPODteam and by members of the POD Working Group of the Ocean Topography Science Team. Thisverificationisconductedyearround,andwithoutaformaltimeconstraintbetweentheproductionofanorbitanditsexpertverification.TheprojectPODteamexpertverificationstartsduringtheorbitproductionprocess.ThemembersofthePODWorkingTeamconducttheirverificationeffortsoncetheorbitsareofficiallyavailable.

6.2.3.3 TypicalPODvalidationmetricsThe tools of orbit verification are traditionally divided among internal and external tests. Internaltestsdonotneedanydataother than thoseused fororbitproduction.Theirkey feature is the factthattheycanbeperformedduringtheorbitproductionprocessitself.Ontheotherhand,theyusuallylacktheabilitytoidentifysystematicerrors.Externaltestsarebasedontheuseofdatanotincludedintheorbitdeterminationoronorbitsproducedbydifferentgroupsusingdifferentsoftwareand/orconfigurations.Thesetestsarethereforedependentontheavailabilityofthesedata.However, theyare very powerful at detecting systematic errors and long-term trends. In addition, external testsperformed using altimeter data evaluate the orbit quality in terms, which are relevant to theoceanographicusers.ThelistofexistingtestsisgiveninTable2.Manyancillaryparametersareestimatedintheorbitdeterminationprocess.Someofthoserepresentmeaningful physical quantities for which valid ranges are known. Others can be correlated withexternalinformation.Whencollectedtogether,theseverificationsgiveadifferentvisionoftheinnerworkingsoftheorbitdeterminationprocess.Asanexample,observingtheamplitudeoftheadjustedempiricalforcesgivesagoodindicationoferrorsinthemodelingofthesatellitesurfaceforces.Theexperience gainedwith othermissionswill eventually allow identification and correction of theseproblemsearlierandmoreefficiently.TheparametersthatshouldbemonitoredaregiveninTable3.



40

Table2.PreciseOrbitDeterminationVerificationTests

Test Description NotesDataresidualsanalysis

Analysisofthestatisticaldistributionoftheresiduals

Dataresidualsinterpretation

Decomposition of the residuals into time and rangebiases and analysis of the fluctuations and trends inthesebiases

The meaning of this test islimited because a cut-offcriteria is applied to thesebiasesduringdataediting

SLRResiduals SLRResidualscumulatedrmsvaluesformeasurementsperformedabovecurrentelevation

Allowstoobserve therelativecontributions of transverseorbiterrors

HighelevationSLRresiduals

Selectedhighelevation laser trackingpassesprovideanaccuratemeasureofthespacecraftrangewhenitisclosetothezenithandthusisagoodestimateofthespacecraftaltitude

Single dataorbit cross-comparison

DORISandGPSareusedindependentlytoproduceJasonorbits, which are then compared together to evaluatesystematicerrors.SLR residuals are computed for both of these orbits toevaluatetheconsistencyofthe3datatypes.

Systematic biases betweendata types due to incoherentreference systems mightoverwhelmthesetests

Overlaps Orbitscomputedforthesametimeperiodusingdifferentdatasetsarecompared.Thistestcanbeusedindifferentways- overlapbetweensuccessiveorbits(comparisonover

thefewhoursincommon)- overlap between a 7-day arc and a shorter arc (in

this case all thedata of the short arc is common tobothorbits)

- - etc.

These tests provide a goodevaluationoftheorbitqualityOverlaps with reduceddynamicsorbitswhichcontaindata in common do notprovide any informationbecause theorbitvery closelyfollowsthedata

Altimeterdatacross-overresiduals

Residuals of the altimeter measurements at cross-overpointsarecomputed

Theresidualsignalduetotidemodel errors and oceanvariability is so high that thistest does not provide a goodestimate of orbit error.However, it is useful toevaluatetherelativequalityofdifferentorbits.

Comparisonbetweenorbits

Orbits computed by different groups using differentconfiguration and/or different software are compared;when longseriesareavailable, themain focus isputongeographically correlated radial differences and on theNorth/South shift between different solutions; specialcare is taken in observing the stability of thesecharacteristicsignaturesovertime

The usual contributors to thePOD expert verificationactivities are NASA GSFC, JPLandCNES



41

Table3.PreciseOrbitDeterminationAncillaryParametersandAssociatedTests

Parameter Function TestDynamicalparametersDragcoefficient Correct errors in the atmosphere

densitymodelShould correlate with solar activityvariations

Solar radiation pressurecoefficient

Correct global error in the surfaceforcemodel

Shouldbenearlyconstant

Amplitudeof1/revterms Absorb errors in the surface forcemodelattheorbitalperiod

Variation with solar angle indicative ofproblemswith solar radiation pressuremodel

Amplitude of thestochasticempiricalforce

Absorbs residual dynamical modelerrors

Level should remain at the 10-9 m/s2level

DORISparametersFrequencybiasperpass Absorbsfrequencyoffsetofbeacons Long term evolution should be

compatiblewithUSOqualityclockTroposphere bias perpass

Empirical value of the zenith wettropospheredelay

On-boardUSOfrequency Measures frequency of the on-boardoscillator

Longtermevolutionshouldberelativelysmooth

SLRparametersRange bias per passTimebiasperpass

Absorbsstationcalibrationerrors Shouldberelativelyconstantperstationand should correlate well with dataobtainedwithothersatellites

GPSparameters Clockoffset Offset of the station and satellite

clocksShould behave in a reasonable clockfashion. Should correlate well with theIGSvalues

6.2.4 Wet-tropodelayvalidationTheoceanwettropospherecorrectionisgeneratedusingthetwo-beamradiometerandinterpolatingthe measured path delay throughout the entire swath. The validation of the wet tropospherecorrection uses a two-step approach: 1.) validation of the wet troposphere path delay from eachbeamand2.)validationofthewet-troposphereintheKaRINswath.TheformerusesmetricsinheritedfromtheglobalvalidationofJason-class(e.g.AdvancedMicrowaveRadiometer or AMR) radiometers:maps and time series of SSH and crossover variance reductionwithrespect toamodel-basedcorrection, longtermmonitoringofdriftsandoffsets.Thesemetricswillbeusedseparatelyoneachradiometer.Thecomparisonbetweenbothbeamswillbeperformedtoruleoutanomalies,jumpsanddriftsthatwouldbespecifictoagivenbeam.ThesecondstepistovalidatetheinterpolationmechanismthatyieldsawettropospherecorrectioninallpixelsoftheKaRINswath.Thevalidationwillusesimilarmetrics(e.g.differencebetweentheradiometer-basedcorrectionandglobalmodelsuchasECMWForNCEP)butappliedinbinsofcross-trackdistance: the influenceof the interpolationwillbemeasuredasbiasorvaryingrandomerror(varianceincrease)asfunctionofthecross-trackdistance.



42

Additionaloptionalvalidationswillbeconsideredduringthe1-dayphase:co-locationofthe2Dwet-troposphere correction used on KaRIN product with AMSU-like radiometer imagery. The goal isprimarily tovalidate thecross-track interpolationmechanismand theresidualerrors illustrated inFigure15(e.g.nonlineareffectsthatareactuallymeasuredbyradiometerimagery).

Figure15.Principleoflinearregressionbetweenleftandrighthandradiometer

6.2.5 Otherpropagationdelayvalidation

6.2.5.1 IonosphereValidationNadirionospherecorrectionThe validation of the altimeter correction (based on the Ku/C dual frequency measurement) isheritedfromJason-classvalidationwithmapsandtimeseriesofSSHandcrossoverbiasandvariancereduction.SWOT’s orbit is not sun-synchronous and the ionosphere path delay will bemodulated by the socalledbeta-primeangleoftheorbitplane.Tothatextentadditionalstatisticswillbecomputedasafunctionofsolartimeinordertoquantifyandvalidatethatthecorrectionbehavesconsistentlywithclimatology from the TOPEX/Jason series as well as external measurement from Sentinel-3 andJason-CSaltimeters.Off-nadirionospherecorrectionThe off-nadir correction is based on JPLGIMmaps, a very smoothGPS-basedmodel. Although thecross-track variations are generally weak over SWOT’s 120-km swath, one can anticipate rareoccurrencesofionospherescintillationsevents,especiallywhenSWOT’ssolartimegoesthroughlatehoursoftheday(10PM).

NRadiom

Radiom 1 1

5

5



43

AcomparisonwithlocalGPSreceivermeasurementswillbeconsideredtoquantifytheir impactonSWOTproducts, although this comparison is likelydifficult toperformdue to thevery local in-situnatureofthemeasurementandthepossiblecorruptionofcoastalorinlandmeasurementsfromthealtimeter.

6.2.5.2 DrytroposphereValidationThe validation of the altimeter correction (based on a global model such as NCEP or ECMWF) isinherited from Jason-class validationwith cross-model comparisonswithmaps and time series ofSSH and crossover bias and variance reduction. Additional validations and comparison will betentativelycarriedoutbytheatmosphericmodelcommunityaspartofthescienceteam.

6.2.6 EMBiasandotherwaveefffectvalidation

6.2.6.1 NadirEMBiasValidationThe nadir altimeter SSB algorithmwill be inherited from Jason-class algorithms, and the Ka-bandcorrectionwillbederivedfromAltiKa(Ka-bandnadiraltimeter)sinceMilletetal(2005)haveshownthattheoff-nadiranglemeasurementshouldbenegligible.Tothatextent, theirvalidationisderivedfrompastaltimetermissionwithmapsandtimeseriesofSSHandcrossoverbiasandvariancereduction,aswellascomparisonswithotherSSBmodels(e.g.impact of using a non parametric solution or a 4-parameter solution, usingWW3 parameters, …).AlthoughdifficulttocompareduetotheinfluenceofSWHonKa-bandsigma0,theSSBcorrectioninKu/C-band(altimeter)willbeanalyzedsystematically.

6.2.6.2 Off-NadirEMBiasValidationDefining the validationof off-nadirEMbias inKaRIN images is challenging considering the lackofmaturityof simulations andSSBalgorithmdefinition.Thevalidationof the geometricbiaswill usemetricsderivedfromnadiraltimetry(seeabove)aswellasspectralanalysestoinfertheinfluenceoftheEMbiascorrectiononallscales.Forglobalocean,thevalidationwillbebasedonthecomparisonbetween3EMBSolutionsbasedon:

- ananalyticalmodel,availablefromthebeginningofthemissionandfedwithSWOTderivedproducts(SWH,Wind...)andpotentiallywithothersea-stateenvironmentalconditionsfromauxiliarydatasets,

- aparametric tablecomputedfromtheoneday fastsamplingphaseresiduals(seebelow),6monthsfromlaunch,

- andaparametrictablecomputedfromthecrossoveranalysis,after1yearofmission.Comparingthesesolutionswillconsistinevidencingpotentialdependenciestoadditionalparametersinordertorefineourunderstandingoftheobservedbias,e.g.,stemmingfromsurfacewavevelocityandorientation.Itwillprobablydeservetheingestionofancillaryand/orauxiliarydatasetssuchasWave Watch 3 global model (wave period, spectral components,…), the OBP wave-mitigationproductsdeliveredbytheon-boardprocessor...)(Tranetal.2010).Formore localvalidation,additionalmethodstoevidencedependenciesareenvisaged, for instancewith parameters derived from systematic co-locationwith SAR imagery (e.g. operationalmissionssuch as Sentinel-1 : http://www.esa.int/Our_Activities/Observing_the_Earth/Copernicus/Sentinel-1).ThiswillinferseastateconditionsintheKaRINimage.



44

OneoftheadvantagesofSWOTovertraditionalaltimetermissionsforthevalidationoftheEMbiasistheone-day fast samplingphase.During thatphase,notwithstanding internal tidesandotherhigh-frequency phenomena, one of the main differences at the repeat passes or cross-overs (aftercalibrationofthephasescreen)willbetheseastatedependentEMbias,whichwillchangefromdaytoday,andverylittlealiasingisexpectedfromthemesoscaleSSHsignaltemporalvariation.Thus,thefastsamplingphasewillpresentauniqueopportunitytoderiveestimatesoftheangulardependenceof the EM bias using repeat pass differences, coupledwith estimates of sea state. The cross-oversdifferencesduringthefastsamplingphasecanthenbeusedtovalidatethattheEMbiascorrectionsreducethecross-overvariability,ashasbeendonetraditionally.

6.2.7 TidalcorrectionvalidationThetidalcorrectionismeasuredinSWOT’sSSHanditisnotapartoftheerrorbudget.Tothatextentits validation is beyond the scope of critical project activities. However, the validation of globalbarotropictidemodelsistraditionallycarriedoutincollaborationwiththescienceteam(e.g.recentreview of global models by Stammer et al, 2014) using SSH and crossover variance reduction,temporalharmonicanalysesorspace/time2Dspectra,andcomparisonwithin-situtidegauges.The case of internal tides is somewhat specific since it is an active research topic. Placeholderalgorithmsandcorrectionaredefinedfortheoceanproductandmayevolveinthenearfuture.Theinternaltidecorrectionwillbevalidatedusingtheabovemetricswithafocusonwavelengthsrangingfrom 50 to 500 km, and accounting for the regional and directional nature of the variability ofinternaltides.ProjectvalidationwillbecarriedoutinclosecooperationwiththeScienceteam.Validation of SWOT during the 1-day phase will also leverage the daily revisit and the very slowvariations of the local measurement time over 60 to 90 days in order to infer the fraction of theinternaltidesvariancethatisphaselockedandpredictablewithstaticmodels,andthefractionoftheinternaltidescontinuumthatmustbecorrectedwithdynamicalgorithms.

6.2.8 DynamicatmosphericcorrectionvalidationThe dynamic atmospheric correction (DAC) ismeasured in SWOT’s SSH and it is not a part of theerrorbudget.Tothatextentitsvalidationisbeyondthescopeofcriticalprojectactivities.However,the validation of global barotropicwind and pressure-forcedmodels is traditionally carried out incollaboration with the science team using SSH and crossover variance reduction, and comparisonwith simple Inverse Barometer solutions based on global atmosphericmodels such as ECMWF orNCEP.

6.2.9 RainflagvalidationConcerningflagvalidation,theSRDstipulatesthat:“SWOTshallprovideflaggingofheightpostingsaffectedbyrain/seaicewith68%accuracyoftherain/seaice(Morethan68%ofcontaminateddatamustbecorrectlyflagged).”And,eventhoughtheflagsalgorithmsarestillbeingdevelopedandnotfullydefinedwecanalreadytakeadvantageofnadiraltimetryexperience(AltiKa,Jason-3,…)toaddressthefactthat:

• flagsarewellpositionedwithin68%=lessthan30%offalsealarms/nondetection• andresultingSSHisrelevantaftertheirapplication

Themetricsofquantificationwillrelyon:



45

- Geometricalmethodsbasedonmaskscomparisons(Differenceofsurfacestatistics,differencesofRMSbetweencontours…)

- Statisticalmethodsbasedonsingledimensionflagcomparisons(histograms,dispersionsdiagrams…

Foreach,products,flagcanbe:- Intrinsictothemeasure:mostlybasedonsurfacecharacteristicssensitivityofthe

instruments,itisdependantontheinstrumentanditsresolution- Basedonexternaldatasets:colocalised(temporallyandgeographically)inbothRadiometer,

NadirandKaRinproducts

6.2.9.1 NadirrainflagvalidationThevalidationofthenadirrainflagwillbeprimarilybasedonthealtimeter(sigma0drops)andthetworadiometerbeams(tensofkilometersawayfromthenadirposition).UsingavariantofAltiKa-basedmatching pursuit algorithms fromTournadre (2009), a systematic comparison of the pointsand segments flaggedby eachdatasetwill be performed to validate the behavior of each flag. ThevalidationofSWOTproductswillalsousethesystematiccomparisonwithradiometer imagersthatarecurrentlyusedforSARAL/AltiKa.

6.2.9.2 Off-nadirrainflagvalidationTheKaRINoff-nadirrainflagisbasedonrapiddropsinradarcross-section(energylostduetorain)thatwillbevisibleinthe250-msigma0meanandvariance.Statistics,mapsandtemporalserieswillbe used to validate the geographic distributionof rain and the typical size of rain cells. Additionalcomparisonwithrainradarsandradiometerwillbeusedtovalidateper-pixelflags,albeitonlyonco-locatedimages.The positive influence of the rain flag will be gauged using statistics on the SSHwhen the flag isappliedornotinvariousregionsandseasons.The 1-day phase will also be leveraged to compare subsequent 1-day images in terms of SSH,radiometer wet-tropo and rain flag. One can anticipate that the atmospheric conditions are morelikely tochangethantheSSH.Theseanalyseswillhelpvalidate that therain flag isconsistentwithspuriousSSHpixelsorspuriouswet-tropomeasurements.

6.2.10 Iceflagvalidation

6.2.10.1 NadiraltimetericeflagvalidationThe validation of the nadir ice flag will be primarily based on a combination of the Ku-C-bandaltimeterandthetworadiometerbeams(inheritedfromJasonmethods).Moreover,waveformsfromthenadiraltimeterwillbeanalyzedusingsupervisedorunsupervisedclassificationtoidentifysea-iceechoes (approach inherited from AltiKa algorithms). The positive influence of the ice flag will begaugedusingstatisticsontheSSHwhentheflagisappliedornotinvariousregionsandseasons.Thevalidationofthenadiriceflagwillalsousethesystematiccomparisonwithseaiceconcentrationandcontours(e.g.Eumetsat’sOSI-SAF)andSARimagerstogetthecontextnearthenadirtracks.

6.2.10.2 KaRINiceflagvalidationTheKaRINoff-nadir ice flag isbasedonrapid increases/drops inradarcross-section(e.g. theGPMKa-bandmissionobservesonaveragea+/-10dBduringoceantoicetransitions)thatwillbevisible



46

inthe250-msigma0meanandvariance.Statistics,mapsandtemporalserieswillbeusedtovalidatethe geographic distribution of sea ice. The positive influence of the ice flag will be gauged usingstatisticsontheSSHwhentheflagisappliedornotinvariousregionsandseasons.Additional comparison with radar cross-section from SAR imagers and sea ice concentration andcontours(e.g.Eumetsat’sOSI-SAF)willbeusedtovalidateper-pixel flags,albeitonlyonco-locatedimagesinsomededicatedareas.

6.2.11 Landflagvalidation

6.2.11.1 Altimeterlandflagvalidation Alongtrackpollutionofdatabylandwillbeflaggedbothinaltimeterandradiometerproducts.Theirvalidationwillconsistin:comparingthemintermsofcoverageandqualityofSLAwhentheflagsareapplied. These results will be mainly analyzed in the fringe near the coasts given by static field(https://www.soest.hawaii.edu/pwessel/gshhg or Globecover, SRTM, ASTER…) or dynamic aftercloudpreprocessing(LandSat,S3…).

6.2.11.2 KaRInlandflagvalidation A land ocean flag limit will be deduced from KaRIN information (sigma0, phase, coherence andinformationfromtheaveragestep).Tovalidateit,acomparisontoahighresolution(km)landmaskisenvisaged.On the one hand, to validate occurrences of non detection of land pollution in ocean images, wepropose to compare the SSH statistics, (jointlywith Sigma0 and interferometric information)withbothmasks (external landmaskorKarin flag) in a limited fringe (below20km fromcoast and forlatitudesbelow50°inordertoavoidmixedinformationfromicepollution).On the other hand, to avoid false alarms,we propose to focus on a set of regions and tomap theremoveddata,superimposedtotheexternallandmask.Theseareaswillbechosenforvariousshoretypes(rockysteepcoasts,largeflatbeaches…)andthiswithdifferenttidalbehaviors:fromnegligibletolarge.Collocationwithopticalimagescanalsobeenvisagedforspecificareas(LandSat,S3…).

6.3 OceanDataProductValidationIn addition to the validation of each component of SWOT’s error budget (section6.2), the mainparametersoftheLevel-2oceanproductswillbevalidated.Seasurfaceheight(SSH)isaddressedinsection6.3.1usingthewavelengthdecompositionofthesciencerequirements,mostlywithusingSSH(or SSH anomalies), i.e. a composite of the instrument range and POD, after all path delay andgeophysicalcorrectionsareapplied.Tothatextent, thissectioncomplementsthevalidationofeachcomponentdescribedinsection.Thevalidationofsignificantwaveheight(SWH)isthendiscussedinsection6.3.2.Thevalidationofsigma0measurementsisalsopresentedinsection6.3.3,andthevalidationofwindspeedin6.3.4.

6.3.1 AbsoluteRangeandSSHvalidation

6.3.1.1 ValidationoftheOceanPerformancefrom15kmto150km



47

Validationof theSWOToceanabsoluteheight error spectrumatwavelengths shorter than150kmwillbeachievedthroughthemethodsdescribedinSect.2.3.2(airbornelidar). SciencevalidationofdynamicheightisdiscussedinSect.6.4.Additionally,we canvalidate someof the smallermesoscale structures (50-150km)using1DSARnadir altimeterSSHobservations collocated in spaceand timeacross theSWOTswaths (limited tomeasurementswithina fewdays). The Jason-CSandSentinel classaltimeterswillbe flyingwithaSAR mode, allowing 1D spectral performance down to 30-50 km depending on the geographicalregionandsea-stateconditions.However,sincetheSARaltimetertrackswillnotbeexactlyalignedwith the SWOT track, thismethodwill have some limitations for validating the SWOT along-trackspectrumperformance.

6.3.1.2 ValidationoftheOceanPerformancefrom150kmto1000kmForscaleslargerthantheoceanswath,thealong-trackspectrafromtheSWOTaltimeterandKaRINmustcoincide(withinthenoisefloorcapabilitiesofthealtimeter).Therefore,theKaRINalong-trackspectrum for these scales will be validated by direct comparison against the simultaneouslymeasuredaltimeter spectrum.Spectral estimateswillbeperformed forall cross-trackpixels in theSWOT swath, to validate the consistency of the KaRIN data and the effect of variations in thewettroposphere,EMbias, and ionospheric corrections that aremadebasedon thenadir altimeterandradiometer measurements. Other altimeter missions flying during this period can also provideadditionalvalidationdata,fortrackswhicharecollocatedinspaceandwithinafewdaysoftheSWOTobservations,aswellas2Dgriddedmapswhichprovideasynopticviewofthemediummesoscale.

6.3.1.3 NadirabsoluterangeandvalidationofLong-WavelengthHeightErrorsForwavelengths longer than1000km, thenadiraltimeterprovidesa Jason-classSSHreference forKaRIN.ThemethodsusedtovalidatethelongwavelengthsofthealtimeterareinheritedfromJason-class Cal/Val activities, and in particular the comparisonwith Jason-3, and Jason-CS (alternativelySentinel-3A and 3B operational altimeters operating at the time of the SWOT launch :http://www.esa.int/Our_Activities/Observing_the_Earth/Copernicus/Sentinel-3 ). The Harvest,Corsica,andBassStraitsiteswillbeconsideredaswell.Inadditiontotheabsoluteglobalmeanbias,thevalidation includes theassessmentof regionaldiscrepancies (mapsof regionalbias), aswellasinvestigations on correlationswith in-orbit conditions or geophysical parameters (e.g. SSB). Thesemetricswillbecompletedbycomparisonswithglobalin-situnetworkssuchastidegaugesandARGO(localcalibrationisaddressedinsection7.1).Thedifferences between the nadir altimeter datasetwill provide additionalmetrics to be used forvalidating the off-nadir KaRIN SSH, as well as some insights on the Ku-band versus Ka-banddiscrepancies(e.g.ionospherebiasesandtrends).MostmetricswillbeusedsystematicallytoderivetemporalvariationsoftheLong-Wavelengtherrors(e.g.seasonalbiases,inter-annualvariability,orrelationshipwithbetaprimeangle…).Ocean measurements will also be used to infer the altimeter regional inland bias using sphericalharmonicand/orEOFanalysis.

6.3.1.4 Off-nadirabsoluterangeandvalidationofLong-WavelengthHeightErrorsanddriftOverview



48

On ocean, the range error budget of KaRIN is defined for wavelengths shorter than 1000 km, asdiscussed in the mission performance and error budget (SWOT_D-79084). In contrast, long termdrifts andbiasesarebeyond theocean science requirements foroceanographyeven if they canbemeasuredonocean.Longwavelengthrangeerrorsanddriftsarehoweverrelevantforthehydrologyerror budget. To that extent, the KaRIN processing chain includes an empirical calibration of theKaRINrangedriftthatisprovidedasacorrectioninLRandHRproducts.Oneproposedalgorithmistocomparethealtimeter-basednadirprofiletoanequivalent1Dprofilefrom KaRIN (e.g. cross-track average in both swaths) and to interpret the difference as a directmeasurement of the KaRIN range bias, thus reducing the KaRIN range drift to small values for acorrectedtopography.Itisthereforeimportanttovalidatethiscorrectionandtheresidualbias(iftheoperationalKaRINrangecorrectionisperfecttheresidualbiaswillbezero).ValidationoftheoperationalrangecalibrationInadditionto theabsoluteglobalmeanbiasbetweenKaRINandthenadiraltimeter, thevalidationincludestheassessmentofregionaldiscrepancies(mapsofregionalbias),aswellasinvestigationsonsystematicerrorslinkedwithin-orbitconditions(e.g.separationofascendinganddescendingpasses,relationshipthermalconditions)orgeophysicalparameters(e.g.relationshipwithSWHorseastateconditions).Thesemetricswillbecompletedbycomparisonswithglobal in-situnetworks (e.g. tidegaugesandARGO)andpreciselocalcalibrationsites(discussedinsection7.1).TheywillbeusedsystematicallytoderivetemporalvariationsoftheLong-Wavelengtherrors(e.g.dailyorcycleaverages,cyclicmaps,regional trends throughout the mission lifetime), to try and detect contamination by otherphenomenaandtoanalyzeextremeandsuspiciousevents.AnalysisofresidualrangediscrepanciesbetweenKaRINandthenadiraltimeterThestudyfromDibarboureetUbelmann(2014)showsthat3othermethodscanbeusedtoquantifythe longwavelength errors of KaRIN, namely the sub-cycle, collinear and crossovermethods. Thecrossovermethodisinessencewhatisdonewithaltimeterrangecalibrationwhennadircrossoversareusedtodeterminealong-termdriftinthealtimeterrange.TheothertwoarevariantsspecifictotheKaRINgeometry.ThecombinedusedofKaRINandthealtimeterwillmakeitpossibletoseparatetheplatformheightbiasfromPODresidualerrorsthatiscommontothealtimeterandKaRIN,andtoisolatetherangedriftthatisspecifictoKaRIN.ThestrengthofthesemethodsarenotaffectedbythespatialvariabilityassociatedwithKaRIN/nadircomparisonsandlesssensitivetodirectionalKaRINerrorssincetheyuseaperfectspatialco-locationbetweentwoKaRINimagesorbetweenaKaRINimageandanadirprofile.Thedownsideisthatthetimedifferencebetweenco-locateddatasetswill introducetemporalvariability.Tothatextent, it islikelythattherangedriftwillrelyoncrossoverwithshorttimedifferences.Residualhigh-frequencyvariability will be very small with respect to the ocean topography signal: measurements fromoverlapsbetweenJason-1GM/Jason-2showthattheSSHvariabilityforperiodsshorterthan4daysrepresentslessthan10%ofthetotalSSHvarianceforwavelengthslongerthan300km(Dibarboure,2015).



49

Duringthefirstmonthsofmission,thecrossoverandsub-cyclecoveragewillbesparse,andthemainreferencemethodshouldbethecollinearmethod(differencebetweensubsequent1-daycycles).

6.3.2 SWHValidation

6.3.2.1 NadiraltimeterSWHValidationThenadiraltimeterSWHwillbevalidatedusingheritage from Jason-classmissionswith statisticalanalyses (e.g. PDF, spectra),maps of bias and variance, analyses of crossover differencewith veryshort time differences, correlation with in-orbit and geophysical parameters. The SWOT nadiraltimeterwillalsobecomparedwithin-situbuoydataandglobalmodels,aswellaswithoperationalaltimeters that are concurrentwith SWOT (e.g. Jason-CS or Sentinel-3). The temporal evolution ofthesemetricswill also bemonitored throughout themission’s lifetime to infer possible drifts andjumpsassociatedwithonboardeventsorprocessorchanges.

6.3.2.2 Off-nadirKaRINSWHValidationTheoff-nadirSWHderivedfromtheKaRINcoherenceislikelysimilartothealtimeterSWHprofiles,althoughtheyarebasedonafitthatextendsthroughtheentireswath.Thevalidationofthisproductwill use the same statistical analyses and external comparison as for the nadir altimeter. Cross-comparisons between the nadir and off-nadir SWH estimates will be performed to rule out thepresenceofsystematicbias,andtoquantifytheprecisionoftheoff-nadirKaRINSWH.Moreover,toinferhowthenaturalSWHvariabilitythroughoutthe120-kmswathmightaffecttheoff-nadir SWHproduct, systematic co-locationswith SAR imagesordownstreamproducts (e.g. virtualbuoys from Collard et al 2009 in Figure 16) will be used, in particular during the one-day phasewhere high-latitude crossovers have an extremely short time difference, enhancing the value ofcomparing twoKaRINmeasurementswith limited external assets. Lastly, theCFOsatmission fromCNES and China (https://cfosat.cnes.fr/en/CFOSAT/index.htm ) will provide an additional globalreferencewith crossoverwave spectra thatwill be comparedwith SWOTmeasurements (e.g. as afunctionofwavelengthortheKaRINazimuth/rangeangles).



50

Figure16.TrajectoriesofalltheswellobservationsgivenbytheSARwavemodeandassociatedtothestormof11April2008,South-EastofNew-Zealand.Thegenerationregionissymbolizedbyareddiskandthecoloralongthetrajectoriesindicatesthedaysoftravelsincegeneration.Bluedisksareplacedatobservationlocations.Theirsizeindicatesthesignificantswellheightatthismoment.(Hussonetal.2012)

6.3.3 Oceanσ0Validation

6.3.3.1 Nadirσ0validationThe nadir altimeter σ0 will be validated using heritage from Jason-class missions with statisticalanalyses (e.g. PDF, spectra),maps of bias and variance, correlationswith in-orbit and geophysicalparameters, analyses of crossover differences with very short time differences,. The SWOT nadiraltimeterwillalsobecomparedtooperationalaltimetersthatareconcurrentwithSWOT(e.g.Jason-CS or Sentinel-3). The temporal evolution of thesemetrics will also bemonitored throughout themission’s lifetime to infer possible drifts and jumps associated with onboard events or processorchanges.

6.3.3.2 KaRInσ0validationAlthoughthereisnosciencerequirementontheperformanceoftheoff-nadirKaRInσ0performance,thevalidationofthisproductwillusethesamestatisticalanalysesandexternalcomparisonasforthenadiraltimeter.Tovalidatethe2Dcomponentoftheσ0 images,andtovalidatetheprecisionandaccuracyinKaRInmeasurements, co-location with SAR images and Ka-band radars (e.g. GPM/TRMM) will be used.Moreover, internal comparisons betweenKaRInproductswill be performedon a systematic basis.Lastly,the9-beamexpertproductwillbeusedinofflinestudiestovalidatethehigh-resolutiondatausedand combined todeliver theLevel-2σ0map, and to explain the sourceof residual variability.Thesestudieswillbecarriedinclosecooperationbetweentheprojectandthescienceteam.

6.3.4 WindSpeedValidation

6.3.4.1 NadirwindspeedvalidationThevalidationthealtimeterwindspeedproductislinkedwiththevalidationoftheσ0(discussedinsection6.3.3)andthe transfer functionusedtoderiveanabsolutewindspeedmodulus.Therefore,thenadiraltimeterσ0willbevalidatedusingheritagefromJason-classmissionsanddiscussedabove.Specific comparisons with external data such as global model and scatterometers (e.g. CFOsatoperates concurrently a wind scatterometer and a wave scatterometer, thus yielding a very richreferenceproductforcross-comparisonswithSWOT).

6.3.4.2 Off-nadirwindspeedvalidationAlthough there is no science requirement on the performance of the off-nadir KaRIN windperformance,thevalidationofthisproductwillrelyonthesameapproachasfornadiraltimetry,andin particular on comparison with scatterometry products and high resolution model outputs tovalidate the2Dwindproducts fromKaRIN, even thoughSWOTprovidesonly thewind speed (notdirection).TheKaRINσ0inversionalgorithmusedtoderivewindspeedmayrequiremodelfieldsasaninput,andalternativealgorithmsandinputfieldswillbegeneratedandcomparedtothereferenceproductinordertoinferthevariabilityandprecisionofthewindspeedinversionprocess.



51

An alternate for of validation will be the development of a wind model function from the GlobalPrecipitationMission(GPM)Ka-bandradar,whichsamples theSWOT incidenceangles.ThismodelfunctioncanthenbeusedtocompareagainsttheonederivedfromSWOTdata.

6.4 ScienceValidationofOceanMeasurementsAprimaryobjectiveofsatellitealtimetryhasbeentomaptheoceandynamicheightforthestudyofocean circulation. The thrust of SWOT in oceanography is to extend ocean dynamic height towavelengthsshorterthanthe2-dimensionalresolutionofconventionalaltimetry.Hydrographicinsitumeasurementshavebeenamainstayofoceanographysincelongbeforetheadventofspacebornealtimetry.Thesemeasurementsprovidedirectinsightintophysicalquantitiesofoceanographicinterest,particularlyoceancirculation.Whilegeoidandhigh-frequencyeffectsmustberemovedfromtheSWOTmeasurementsofabsoluteSSHbeforecirculationcanbeestimated,hydrographicinsituapproachesmeasuredynamicheight,whichisthequantityofmorefundamentalscienceinterestthatunderpinstheSWOToceanographicscienceobjectives.TherelationshipbetweenabsoluteSSHanddynamicheighthavebeendemonstratedatwavelengthslongerthan150km,buttheirrelationshipdowntowavelengthsasshortas15kmhavenotyetbeenfullyvalidated.Hydrographicinsituapproachesrelyonconductivity-temperature-depth(CTD)measurementsoververticalprofiles(ie,asafunctionofdepth)atagivenhorizontallocationinordertointegrateverticallythedynamicheight.AspatiallydistributedarrayofinsitumeasurementswouldbeneededinordertovalidatethespectralperformanceofSWOTusinginsituapproaches.A1-DarrayoffixedmooringswithCTDinstrumentswouldbetheidealmeasurementapproach.Thearrayspacingwouldbe7.5kminordertoNyquistsamplethe15kmminimumwavelengthrequiredoftheSWOTmeasurement.Thearraywouldextendfor150km(20moorings)inordertocapturewavelengthsuptotheregimethatnadiraltimetryalonewouldoffersufficientvalidation.EachmooringwouldideallyincludeCTDinstrumentsspanningthefulldepthoftheocean.Eachmooringwouldideallyalsoincludesufficientnear-real-timecommunicationthatthehealthandfunctionofthearraycouldbeconfirmed(orfaultyelementscouldbereplacedinatimelymannersoasnottojeopardizethetightSWOTCal/Valtimeline),anddatacouldbeexaminedwithsufficienttimetoreacttoanysurprisesduringtheCal/Valperiod.Unfortunately,thisidealmeasurementconceptusingmooringsisnotfeasiblelogisticallyorprogrammatically.However,anarrayofstationkeepingautonomousunderwatervehicles(AUVs)maybeabletoreasonablyapproximatethismeasurement.UnderwaterglidersareAUVsthattypicallypropelthemselvesbychangingtheirbuoyancy,usingwingstoconvertverticalmotionintohorizontalmotion.Somemayalsobeequippedwithfoldable,propeller-driventhrusters.Gliderstypicallyincludesatellite-basedcommunicationslinkstooperators.WithgliderscarryingCTDinstrumentsanddivingupanddownwhilemaintainingapproximatelyfixedlocationshorizontally,dataintegratedoverthegliderprofilescangiveestimatesofdynamicheightforvalidationofSWOTscienceandspectralrequirements.ItisnotpossibletosamplethefulldepthoftheoceanwithglidersatthecandidateCal/Valsites,butsimulationssuggestthatsamplingtheupper500moftheoceancapturesmostoftheoceandynamicssuchthatthemeasurementsoffercomparablespectralperformancetotheSWOTbaselinerequirements,assumingthattheSWOTmeasurementsarecorrectedforthedifferencesbetweenabsoluteSSHanddynamicheight(theSWOTrequirementsonSSHerrorspectrahavebeenchosentobeconsiderablysmallerthantheexpectedspectrallevelsofSSHsignals).



52

Figure17. Simulation results consideringanarrayof 20 stationkeepinggliders spacedat 7.5kmapart sampling theupper500moftheocean(seeWangetal.,2017).

Inordertomitigaterisksassociatedwiththegliderconcept,anexperimentwasconductedinsummer2017,comparingrealgliderperformancetoamooringinMontereyBay.Theexperimentresultsshowgoodagreementbetweenthemooringandglidermeasurementsofdynamicheight.Theresultsalsoshowedthatthegliderstestedwereabletoadequatelymaintaintheirhorizontalpositionsandthatsamplingonlytheupperoceanratherthanthefulldepthcapturesvariationsindynamicheightovertime.However,theoceancurrentsanddynamicsatthelocationofthisexperimentarenotnecessarilyrepresentativeoftheprimaryCal/Valsite.



53

Figure18.ComparisonofastationkeepingglidertoafixedmooringinMontereyBayfromanexperimentinJune2017.

ThefollowinguncertaintiesthereforeremainintheabilityofagliderarraytomeetSWOTCal/Valneeds:

1. Thereisariskthattheglidersmaynotbeabletostayonstationinthepresenceofthecurrentsattheprimary(California)SWOTCal/Valsite.Notethatanalysissuggeststhatthegliderswouldnotbeabletomaintainstationatthebackup(GulfStream)Cal/Valsite,atwhichthecurrentsareevenstronger.AnalysissuggeststhatthestationkeepingabilityattheCaliforniasiteismarginal.Hybridgliders,whichcanusetheirthrustersattheexpenseofbatterypower(andhenceoperatingduration),shouldhavesufficientcontrolauthorityforstationkeeping(viaanalysis).However,assumingtheuseofthrustersforSWOTCal/Valimpliesthat(1)onlyhybridgliderscanbebaselined,whichcouldincreasecostsandlogisticdifficultyinsecuringtheuseofsuchgliders;and(2)additionalcost,especiallyinshiptime,willbeinvolvedtoreplacethegliderbatteriesduringthecourseofthecalibrationphase.Constantuseofthrustersshortenstheoperationaldurationoftheglidersfromanestimatedfourmonthstotwomonths,socoveringthe90daycalibrationphaseofthemissionmayrequireanadditionalshipcampaigntoreplacethebatteriesofthegliderarrayifthethrustersareneeded.

2. Thereisariskthatsamplingonlytheupper500misnotsufficientlyrepresentativeoffull-depthsamplingattheprimary(California)SWOTCal/Valsite.NotethatthedepthattheCal/Valsiteisapproximately4000m,whilethedepthatthesiteoftheMontereyexperimentisapproximately1000m.Increasingthedivedepthofthegliderprofileswouldallowdeepersampling(uptothemaximumdepthoftheglider,whichvariesbyglidermodel),butsinceeachdivewouldrequiremoretime,thetemporalresolutionofthemeasurementswoulddegrade,therebyintroducingadditionalerror.Temporalresolutioncouldberegainedbyoperatingpairsofglidersateacharraylocation,witheachgliderexecutingadiveprofilethatisphased180degreesapartfromtheother.Theuseoftwoglidersperarraylocationwould



54

doublethetotalnumberofgliders,however,therebyincreasingthecostandoperationalcomplexityoftheexperiment.

Inordertoaddresstherisksabove,apre-launchexperimentattheCaliforniaCal/Valsiteisbeingplannedthatwouldinvolvethedeploymentofonefixedmooringandtheconcurrentoperationoftwogliders.ThesegliderswouldbeSlocumhybridgliderswiththrusters.Thedesiredoutcomeoftheexperimentisthat(1)theglidersareabletomaintainstationwithnoorminimaluseofthrusters;(2)thedynamicheightmeasurementsoftheglidersandthemooringagree,demonstratingthatsamplingonlytheupper500moftheoceanissufficienttoachievealevelofaccuracyappropriateforthesciencevalidationobjectives.Assumingthatthepre-launchexperimentachievesthedesiredoutcomeabove,thereisaproposalforthepost-launchsciencevalidationtocompriseanarrayof20stationkeepingglidersspaced7.5kmapartcoveringa150kmlineattheCaliforniacrossoversite.Eachgliderwouldsampletheupper500moftheoceanforthe90daydurationoftheSWOTcalibrationphase.Additionally,onefixedmooringsamplingthefulloceandepthmaybedeployedtohelpcross-calibratethegliders.GPSinstrumentsarenotpartofthebaselineinsituproposalbutcouldbeincludedasacontributionand/orincludedifwarrantedbasedonotherGPSinvestigationsthatwillbeoccurringinthepre-launchtimeframe.NotethatunderwaterCTD(UCTD)measurementsonamovingshiphavebeenevaluatedforSWOTCal/Val,butduetotheslowspeedoftheshipcomparedtotheveryfastoverflighttimeofthespacecraft,thisapproachcannotadequatelycapturethetemporalvariabilityoftheocean.

6.5 SWOTSurfaceWaterErrorBudgetValidationInthissectionwedescribehowtheoverallsurfacewaterperformanceofSWOTwillbevalidated.Wediscusshoweachcontributortotheerrorbudgetwillbevalidatedindependently.ThesurfacewaterperformanceofSWOTwillbevalidatedwithacombinationofmeasurementsfrominsituinstrumentsandairborneinstruments,includingAirSWOT,duringandaftertheendofthefastsamplingphase.Thevalidationsiteswillbedistributedtocharacterizetheeffectsofuncalibratedphase/rolldriftinthe interior of continents, aswell as a variety of lake, river, andwetland characteristics, includingsize,topography,andvegetationtype.Validationsiteswillbedividedintotwotypes:Tier1sitesthatwill involvedirect fieldmeasurementsbySWOTvalidation teammembersandTier2site thatwillleverageexistingmeasurementassetswithminimaladditionalfieldmeasurements(e.g.USGSstreamgauges) andwill be used to estimate the spatial and temporal variability in SWOTmeasurements.Therewill be a total of about 15Tier 1 sites (see Sections7.2.1–7.2.4) and~100Tier 2 sites (seeSection7.2.5).InsituobservationsoflakeandriverlevelandslopewillbeobtainedatTier1andTier2siteswithGPSobservationsincombinationwithtemporalvariationsmeasuredbyexistingrivergauges,and/ortemporarilyinstalledpressuretransducersanddischargegauges.Thelakeandriversurfaceareawillbemeasuredusing theAirSWOTnear-infraredcamera (ora similar systemonadifferentairborneplatform) at the same time as a SWOT pass. In situ information will also be collected regardingvegetationdistribution,height,andcanopycharacteristics(LeafAreaIndex(LAI),canopyclosure),aswell as a high accuracy digital elevationmodel of the surrounding topography for layover studies.



55

Thesemeasurements will enable the validation of SWOT elevation and surface water extent on acontinentalbasis.Thevalidationperiodusedtoassessmissionsuccesswilltakeplaceduringthefirstsix months to a year of the start of the nominal mission phase, but validation will continuethroughoutthelifetimeofthemission.InthissectionwedescribehoweachindividualcomponentofthesurfacewaterSWOTerrorbudgetwillbevalidated.

6.5.1 RandomheighterrorvalidationThe randomheight error for hydrology targets can be assessed by simply examining the standarddeviationoftheSWOTheightestimatesoverareasthataresufficientlylarge(forexample,largelakesthatarefreefromlayover).ComparisonofthenoisestatisticsbetweenLRandHRdata(allowingfordifferencesinrandomerrorperformanceduetopresumming)canalsovalidatetheHRrandomerrorperformance.Datafromsmallerwaterbodieswillalsobeaggregated,withmodelsusedtoaggregatethedatastatistically,toassessrandom-errorperformance.

Figure19.Randomerrorasafunctionofthecross-trackposition(HRproduct).FromJPLD-79084.

Figure20.Randomerrorasafunctionofthecross-trackposition(LRproduct).FromJPLD-79084.



56

6.5.2 AbsoluteinlandsurfacewaterheightvalidationThe validation of the absolute heights requires that the absolute height of the surface of theconsideredwaterbodyismeasuredindependentlyofSWOT.SeveralprotocolswillbeimplementedandwillvarybetweenTier1andTier2sites.Thefollowingsections(6.5.2.1to6.5.2.4)describethevariousmethodsbywhichabsoluteheightwillbevalidated.

6.5.2.1 AbsolutewaterheightvalidationbypressuretransducersDensenetworksofGNSS-leveledpressuretransducersorGPSbuoyswillbeinstalledalongriversandinlakeswithintheCal/Valsitesdetailedinsection7.2.Forrivers,transducerswillbeinstalledevery6-20 river widths; for small lakes, one transducer will be sufficient; for large lakes, five to tentransducers (located at different shoreline locations and in the lake center) will be used. Thesenetworks will allow validation of absolute SWOT height variations in space and time over shortlengthscales.ComparisonbetweenSWOTheightsandwaterabsolutelevelswillbeperformedoveralongperiodof time, startingduring the1dayorbitandcontinuing forat leastoneyearduring thenominalorbit.Akeyissuetoresolveforuseofthismethodwillbewatersurfacecurvature,forbothriversandlakes.For rivers, if we can assume that any curvature of the water surface long profile is below thedetection limit of SWOT data then a single point measurement in the middle of a reach willcharacterize the average water surface height over the reach that SWOT will measure. In otherwords,ifwecanassumealinearwatersurfaceslopethenourmeasurementtaskisgreatlysimplified.Pre-launchassessmentswillneedtobemadeateachoftheTier1riversitestodeterminethattheydonotshowsignificantlongprofileslopechangeswithvariationsinstagethatmayprecludetheuseofthisoption.

Forsmall lakes,curvatureeffectsare likely tobesmall (i.e. smaller than theSWOTdetection limit)andcanbeminimizedbypositioningpressure transducersorGPSbuoysat lakecenters. For largelakes,therequirementisforpressuretransducersorGPSbuoystobepositionedatleast1kmawayfromtheshoresuchthatanywaterheightvariationoverthe~1km2SWOTaveragingwindowhasalinearslope.Thiswillensurethatthepointwaterheightmeasurementisequivalenttoanaverageofground-based water height measurements over the SWOT averaging window. Pre-launchassessmentwillneedtobemadeatlargelakestocollectdataonpossiblewaterheightvariationandlengthandheightscalesofwaterslopecurvature.Thiswillbeachievedviaaninstallationoffivetotenpressuretransducersaroundtheperimeterofthelake.

For rivers, the existence of significant cross channel elevation changes (for example due tohydrodynamicsuper-elevationofthewatersurface)attheTier1sitesalsoneedtobediscountedasthiswouldotherwisesuggestthatpointelevationmeasurementsofwaterheightwouldhavebias.Apre-launch field campaign is required at the river sites to measure the scale of such effects anddeterminewhetherornotthesewillbesmallerthantheSWOTdetectionlimit.

PastAirSWOTdatacanalsobeusedtoexamineforthepresenceofalltheaboveeffects,andthiswillbeanimmediatetaskforpre-launchactivities.



57

6.5.2.2 Direct measurement of the free surface height at the exact timing of SWOT overpassing

DirectmeasurementofwaterheightatthetimeofSWOToverpassingwillbeperformedbymeansofGNSS systems, mostly based on the US GPS system. Several devices with floating GPS antennascurrentlyarebeingdevelopedandtestedbytheSWOTscienceteam.Themostadvancedconsistsofafloating sheet of ~10m² bearing a GPS antenna for which height and attitude is continuouslymonitored(CalNaGeo,https://swot.jpl.nasa.gov/docs/jun17_stm_101_seine.pdf). Lightandeasy tosetup,thisequipmentcanbeusedtoperformthemappingofseveralkm²withinafewhours.Othersystems include GPS floats mounted on Sontek hydroboards coincident with Acoustic DopplerCurrent Profiler (ADCP) measurements. Prior to launch, comparisons will be made between theaccuracyofGPSmeasurementslikelytobeobtainedfromallGPSmeasurementplatforms.ThesitestobemeasuredthiswaywillbeselectedamongtheofficialCal/Valsitesoftheproject,preferentiallysitesthatwillbeoverflownduringthe1dayorbit

6.5.2.3 AbsolutewaterheightvalidationfromHydroweb/HysopeExternalvalidationofthenadiraltimeterandKaRINproductswillbedeliveredinnearrealtimebythe Hydroweb/Hysope network over large global lakes with an accuracy of water level at sub-decimeter level reported continuously during the mission lifetime (Hydroweb:http://www.legos.obs-mip.fr/en/soa/hydrologie/hydroweb/; Hysope: http://hydroweb.theia-land.fr/?lang=en&). The Hydroweb and Hysope sites already are used for external validation byaltimetermissions.ThequalitycheckofthenadiraltimetryproductsisdoneusingasetofinsitulakelevelcollectedthroughnationalhydrologicalservicesinUSA,inRussia,inChile,andinArgentina.

Figure21.TheCalNaGeoinstrumentusedtoproducethehighestprecisionwatersurfaceelevationsfromGNSSmeasurements.

Therequiredaccuracyfortheabsolutesurfacewaterheightvalidationisanabsoluteverticalaccuracy to ±5cm at 1σ (minimum) or ±2cm at 1σ (target). Cal/ValsiteswhereabsolutesurfacewaterheightmeasurementswillbeperformedincludealloftheTier1sites.TheseincludesitesoverRivers,Lakes,WetlandsandEstuaries.



58

6.5.2.4 AbsolutewaterheightvalidationusingTier2Cal/ValsitesInaddition to theabovemethods,wewilldevelopanetworkofTier2Cal/Val siteswherewewillleverageexistinggaugemeasurementsof stage anddischarge, andwill addaGNSS-levellingof thegage datum. Once the reference point of the gauge is leveled to GNSS accuracy, the water levelsrecorded by the gauge (eithermanually or automatically) can be converted into absolute heights,directly comparable with the SWOT measurements in close vicinity. We plan to acquire a globaldatabase of a few hundred leveled gauges by launch. The countrieswhere thiswork is already ingood progress are the USA, France and the countries sharing the Amazon basin. Extension of thisdatabase to SouthAsia countries (India, Bangladesh, etc…), African countries (Niger, Congo, RDC)andEuropeancountrieswillbeperformedaccordingtoopportunity.The Tier 2 Cal/Val network optimallywill consist of~200-300 siteswith good global coverage ofdifferenthydroclimaticandecosystemzones.Eachgagewillbelevelledtohavearequiredminimumverticalaccuracy to ±5cm at 1σ (minimum) or ±2cm at 1σ (target).

Figure 22. The Sontek hydroboard system thatwill also be used to produce precision GPSmeasurements of water surfaceelevation.

6.5.3 Inundatedsurfaceareavalidation

6.5.3.1 RiverinundatedsurfaceareavalidationTo validate that SWOT can measure inundated area in rivers with sufficient accuracy to meetrequirements presented in the SWOT Science Requirements Document, we will validate riverinundated surface area using one or more of the following three methods, all of which will beevaluatedindetailduringprelaunchactivities:1. Todirectlyvalidateinundationextentinrivers,wewillacquireonceateachTier1fieldsitea

high-resolution (~1 m resolution) airborne dataset of near-infrared or mid-infraredphotography (equivalent to Landsat TM band 4 or, ideally, band 5). This type of datacurrently is being collected using the AirSWOTColor Infrared (CIR) Camera,whichwill beevaluatedforsuitabilitybasedonflightsconductedin2015and2017.TheairborneimagerymustbeacquiredsimultaneoustoaSWOToverpass(<3hrsdifferent,butdependingonthewater dynamic on each site) during clear-sky conditions in order to provide a direct



59

comparison. There is a long heritage of measuring inundation extent using this type ofimagery inrivers. SeeFigure23foranexampleofsuch imagery,acquiredovertheTananaRiverinsummer2015.

2. Toindirectlyvalidateinundationextentinrivers,wewillusetheintersectionofahigh-qualitytopographic DEM and fieldmeasurements of surface-water elevation collected from eitherinstalledpressuretransducersorfromboat-measurements.Whilethisvalidationmethodhasthe advantage that it does not require directmeasurements of inundation coincident withAirSWOT,itdoesnotprovidedirectmeasuresofinundationextent.

3. During field campaigns to be conducted during both the fast sampling and nominal orbits,field Cal/Val teamswill walk selected sections of water/land boundaries (shoreline) usingGPSwith<2mhorizontalprecisioninordertoprovidevalidation. Thissecondstepwillbeparticularlycriticaltoperforminareaswithlarge,wetsandbarsadjacenttosediment-ladenrivers,asthesefeaturescanlookverysimilarinnear-IRphotography.

Figure23.ExampleofcolorinfraredphotographyacquiredovertheTananaRiver,Alaskaduringsummer,2015.ThistypeofimagerywillbeusedtovalidateSWOTmeasurementsofinundationextentinrivers.

The Cal/Val Tier 1 river sites where inundated surface area will be validated include:WillametteRiver, Tanana River,Mississippi River, Connecticut River, Garonne River, and a tropical river. Therequired imageaccuracy for inundatedextent is7.5% (minimum)and1.5% (target) over a 1 km2 area. For GPS surveys of inundation extent, we require horizontal measurement accuracy equal to ½ of the SWOT pixel size in the range direction (~12.5 m in the middle of the swath, minimum) and 1/10 of the SWOT pixel size, or ~2.5 m (ideal).

6.5.3.2 SmallLakeinundatedsurfaceareavalidation

TheabilityofSWOTtomeetthesciencerequirementsforinundationextentinlakeswillbevalidatedusingoneormoreofthesamethreemethodsdescribedaboveforrivers:1. Todirectlyvalidateinundationextentinsmalllakes,wewillacquireonceateachTier1field

site a high-resolution (~1m resolution) airborne dataset of near-infrared or mid-infrared



60

photography (equivalent to Landsat TMband4 or, ideally, band5). The airborne imagerymustbeacquiredsimultaneoustoaSWOToverpass(sameday)duringclear-skyconditionsinordertoprovideadirectcomparison.Thereisalongheritageofmeasuringinundationextentusingthistypeofimageryinsmalllakes.

2. We will leverage relationships between lake height and inundation extent to predictinundation extent, which will then be directly compared to SWOT measurements. Therelationshipbetweenheightandsurfaceareainmostsmalllakesdoesnotvarysubstantiallyaslongastheshorelinedoesnotchange.Inlakesthatdonothavehighlyvariableshorelinecharacteristics, we will validate the inundation extent using a combination of a high-resolutiondigital elevationmodel (e.g. lidar)andmeasurementsofwater surfaceelevation.Whilethisvalidationmethodhastheadvantagethatitdoesnotrequiredirectmeasurementsof inundation coincident with SWOT, it does not provide direct measures of inundationextent.

3. During field campaigns to be conducted during both the fast sampling and nominal orbits,fieldCal/Val teamswillwalk selected sectionsofwater/landboundaries (shorelines)usingGPSwith<2mhorizontalprecisioninordertoprovidevalidation.

Small lake inundated surface area will be validated at all Tier 1 small lake sites. The images ofinundationextentshouldbeaccurateto7.5%(minimum)and1.5%(target)overa1km2area. ForGPS surveys of inundation extent,we require horizontalmeasurement accuracy equal to½ of theSWOTpixelsizeintherangedirection(~12.5minthemiddleoftheswath,minimum)and1/10oftheSWOTpixelsize,or2.5m(ideal).

6.5.3.3 LargelakeinundatedsurfaceareavalidationMeasurement of inundation extent for large lakeswill havedifferent types of complexity thanwillvalidationof small lake inundationextentmeasurements. Most small lakeswillbemeasuredviaasingleSWOToverpass,whilemanylargelakeswillonlybepartiallyobservedinanygivenoverpass.Itwillnotbepractical toacquireairborne imageryover largeareasdirectlycoincidentwithSWOToverpasses.However,thelargelakeschosentobeprimaryvalidationsitesdonotvarysubstantiallyinsurfaceareaovershorttimeperiods.Moreover,itisexpectedthatitwillbemucheasiertomeettheSWOTinundatedareaaccuracyrequirementforlargelakesthanforsmalllakessimplybecauseamuchlarger fractionof theirarea isdistant fromlandcontamination. Assuch, thefollowingthreemeasurement strategieswill be used, neither ofwhich requires acquisition of newdatasets by theSWOTmission:1. High-tomoderate-resolutionsatelliteimagery(e.g.Sentinel2,Landsat)willbeacquiredclose

in time to SWOToverpasses, and inundation extent derivedusing existingmethods (e.g. LiandSheng,2012)willbedirectlycomparedtoSWOT-derivedinundationextent.

2. Bathymetry of shallow lakes acquired using existing altimeters or in situmeasurements ofheightandhigh-resolutionimagesofinundationextentwillbeusedtodeveloppreciseratingcurves between inundation extent and elevation. These rating curves will allow preciseestimationofinundationextentgivenknowledgeofwatersurfaceelevationduringtheSWOTCal/Valphase.PleaseseethecasestudyonLakePoopodescribedbelow.

3. Finally, inundated areas derived from rating curves developed between inundation extentand water surface elevation derived from existing altimetry resources will be comparedagainst SWOTmeasurements for a variety of large lakes globally (e.g. Figure 24). A set ofabout100lakesamongthemhalf locatedontheTibetanPlateaualreadyexist(LEGOSworkfor Hydroweb database) and will be completed before the launch. This will serve as an



61

external sourceofvalidation forwaterextentvalidationalthoughnotstrictlyof land/waterclassification. If water height is validated by other means, then for each water heightmeasured, a water extent can be calculated using polygon coefficients of the hypsometrycurveandcomparedtothesurfaceextentdirectlymeasuredbySWOT.

TheTier1Cal/Val largelakesiteswheretheinundatedsurfaceareawillbevalidatedinclude:LakeIssykkul, LakeTahoe, and the large global lakesdataset (Hydroweb).The required accuracyof theinundationextentshouldbeaccurateto7.5%(minimum)and1.5%(target) over a 1 km2 area.LakePoopoCaseStudy: In2014,aDEMoftheLakePoopo,whichis locatedovertheAltiplanoinSouthAmerica,wasdevelopedusingacombinationofsatellite imagery(setof landsat images)andlaseraltimetryonIcesat.LakePoopoisveryshallow,withhighseasonalandinter-annualarealextent(andheight)variability.Everyyearinwinter,itisinundatedandduringtherestoftheyearitshrinksduetoveryhighevaporation.Atinter-annualtimescales,thiscycleofinundationanddroughtalsoisvery unstable, with some very wet years contrasting with very dry ones (see Figure 25). Inconsequence, the derived DEM of Lake Poopo is valid from a minimum surface close to the fulldrought to amaximumwhen the lake is almost entirely inundated (red lines on Figure 25). Theprecision of this DEM has been established at better than 10 cm. It therefore can be used forvalidationoflakesurfaceextent.ForeachwaterextentmeasuredbySWOT,wecansimplyprojectthecorrespondingwatermasktotheDEManddeterminetheclosesttheoreticalmaskdeducedfromtheDEMalone.RepeatingthisprocedurepassafterpasswillgivequantitativevalidationofwatermaskinferredfromSWOTmeasurements.

Figure24.HypsometrycurvefortheLakeNganga-Ringco(TibetanPlateau).



62

Figure25.LakePooposurfaceextentmeasuredbyModisfrom2000to2012every8days

Figure26.ImageofthebathymetryofthelakePoopo.

6.5.3.4 WetlandinundatedsurfaceareavalidationSWOTmeasurementofwetlandsurfaceareawillbevalidatedusingoneormoreofthefollowingfourmethods:1. There is substantial precedent for measuring inundation extent under even very dense

vegetationusingL-bandSARsensors. InundationextentwillbemeasuredathighspatialbytheUAVSAR system, a JPL facility, concurrentwith SWOToverpasses of at least twoof theLower Mississippi, Everglades, and Yukon Flats Tier 1 field sites. The resulting UAVSAR



63

inundation heights will then be compared directly against SWOT returns to assess SWOTperformanceinwetlandsoveravarietyofvegetationtypes.

2. Varioussatellitesensorscapableofmeasuringinundationextent invegetatedenvironmentsat high resolution may be available concurrent with SWOT during the Cal/Val phase,includingSentinel1,NISAR,andfutureRADARSATandALOSmissions.AlthoughthismethodwillnotbeourprimarymeansofvalidatingSWOT,itwillallowgreatergeographicdiversityinthetypesofwetlandsvalidated.

3. Wewill use high-resolution lidar DEMs of wetland topography, where available, to assessinundationextentbasedonvariationsinwatersurfaceelevationmeasuredfromSWOT.ThismethodhastheadvantagethatitdoesnotdependonsimultaneousimageacquisitionwithaSWOToverflight,thoughitdoesrequireaprioricollectionofalidarDEM.

4. To understand the influence of vegetation on SWOT inundation extent and water surfaceelevationreturns,wewillcollectairborne lidarmeasurements includingbothwatersurfaceelevationandvegetationheightsimultaneouslywithUAVSARandSWOTmeasurementsoverat least two of the Lower Mississippi, Everglades, and Yukon Flats field sites. Thesemeasurements will allow us to assess vegetation height and canopy closure in a way notpossible using ground-based measurements and will allow full understanding of SWOTcapabilitiesinwetlands.

TheTier1wetlandCal/Valsiteswhere inundatedsurfaceareawillbevalidated include theLowerMississippi,YukonFlats,andEverglades.Frenchsiteswillbeselectedlater.Therequiredaccuracyoftheexternalvalidationmethodissuchthatatleastonemethodmusthaveaccuracyofatleast7.5%over1km2,withanidealaccuracyof1.5%over1km2.

6.5.4 RangedriftvalidationTheabsolute rangedriftofKaRInwillbevalidatedbycomparison tonadiraltimeterdataover theocean,asdescribedinprevioussectionsonlong-wavelengthoceanvalidation.

6.5.5 Roll/phasedriftvalidationOverland,therollwillbevalidatedoverlakes.Adatabaseoflargelakeswhosesurfaceisnotsubjecttorapidtilts(establishedfromongoingaltimetrymissions)isbeingbuilt.Thesesurfaceswillbeusedas a reference to infer cross track tilts due to rolling. Thanks to the repeat orbit of the SWOTaltimeter,wemaycalculateameanverticalprofilealongeachofthetracksforeachchosenlake.Inthis process we benefit from the long-term time series of altimetry data since the launch ofTOPEX/Poseidon.Foreachlake,althoughabsoluteheightischangingduetohydrology,therelativeheightbetweeneachtrackshouldn’tchangefromonecycletoanother,exceptifseichesareobserved.Thesameapproachcouldalsobeusedovermajorriverswherethemeanslopeisknownwithagoodaccuracy.

6.5.6 LandWet-TropoDelayValidationThe wet troposphere estimates over land are based on models. The models can be validated bycomparisontoradiometer,radiosonde,andGPSmeasurements.Forlargelakes,themodelscanalsobecomparedtotheSWOTradiometerdata.



64

6.5.7 OtherpropagationdelayvalidationIonosphere model corrections can be validated by comparison to local GPS estimates. Drytroposphere estimates are based on ECMWF models and have been reasonably well validatedalready.

6.5.8 Slopevalidation

Validation of the river slope requirement requires that the absolute height of the surface of theconsideredwaterbodyismeasuredindependentlyofSWOT.Severalprotocolswillbeimplemented,andthesewillvarybetweenTier1andTier2sites.AttheCal/Valsitesthefollowingmeasurementswillbemade:1- Dense networks of leveled pressure transducers or GPS buoys will be installed along

rivers within the project calibration/validation sites detailed in section 7. Rivermeasurementsarerequiredapproximatelyevery10riverwidthsoveralengthofatleast50km.U.S.locationswillincludetheWillametteRiver,TananaRiver,ConnecticutRiver,andMississippiRiver. ThesenetworkswillallowvalidationofSWOTslopevariationsinspaceandtimeovershort lengthscales. ComparisonbetweenSWOTheightsandwaterabsolute levelswill be performed over a long period of time, starting during the 1 dayorbitandcontinuingforatleastoneyearduringthenominalorbit.

For large rivers (principally the Mississippi) the existence of significant cross channelelevation changes (for example due to hydrodynamic super-elevation of the watersurface)attheTier1sitesalsoneedtobediscountedasthiswouldotherwisemeanthatslopemeasurementswouldvarysubstantiallydependingonwhichsideoftherivertheywereobtained from. Apre-launch field campaign is required at the large river sites tomeasure the scale of such effects and determinewhether or not there are going to besmallerthantheSWOTdetectionlimit.

2- DirectmeasurementofthefreesurfaceheightattheexacttimingofSWOToverpassing.

SuchmeasurementswillbeperformedbymeansofGNSSsystems,mostlybasedontheUSGPS system. Several types of GPS floats are currently in use by the group (see section6.5.2.2).InsitumeasurementsofslopewillbemadeatleasttwiceatdifferentdischargesforallTier1sitesusingthismethodcoincidentwithSWOToverpassesduringthecal/valphaseofthemission.

3- Whilethetwomethodsabovearecapableofmeasuringslopesoverrelativelyshortriver

reaches, they are not capable of synopticallymeasuring slopes over long river reaches(e.g. >100 km). The only tool currently capable of validating such slopes derived fromSWOT is AirSWOT (Icesat2 could provide additional inputs but this mission is not yetlaunched). As such, AirSWOT will serve as a crucial tool for validating SWOT slopemeasurements.AirSWOTwillbeflownatleasttwiceovertheWillamette,Mississippi,andConnecticut Rivers at different discharges during the cal/val phase in order to validateSWOTslopevaluesoverlongreaches.

4- AtTier2sites,pairsofaccuratelyleveledgaugeswillprovideestimatesofwatersurface

slopebetweenthem,thoughnotwiththedegreeoffidelityprovidedbythepreviousthreemethods. Nonetheless, thesesiteswillbeusedtovalidateSWOTslope inenvironmentswhere SWOT-dedicated measurements are not feasible due to cost or logisticaldifficulties.



65

TheCal/ValriversiteswhereslopewillbevalidatedincludeTier2riversitesandthefollowingTier1sites:Willamette,Connecticut,Tanana,Mississippi,andGaronneRivers.Therequiredaccuracyoftheindependent slope measurements (from AirSWOT, pressure transducer arrays, and GPS drifters)shouldbeatleast8.5µrad(minimum)andpreferablyasgoodas1.7µrad(ideal).

6.5.9 LayoverflaggingandimpactvalidationThe impact of layover on height and slope estimateswill be characterized by comparison to truthmeasurementsasdescribedinprevioussections.Thecontributionsduetolayoverwillbeseparatedfromothercontributorsbycomparisontopredictionsusinghigh-fidelityDEMsandmodelsofsigma0,whichcanbeinformedbyairborneKa-bandreflectivityestimates.Ageometricflaggingassessmentonlyrequirestheuseofahigh-fidelityDEM.

6.5.10 RainflagvalidationThe rain flag is designed to alert users to the presence of rain during measurements, which cancompromise SWOT performance. There are several ways in which the presence of rain will bevalidated. First, for Tier I field sites at least, localmet stationswill be erected, and these stationsshould include tipping buckets or other means of precipitation measurement. This directmeasurement at the site will give the time and intensity of rain events for validation. Second,commercial Doppler radar is effective at detecting rain events, andmany civilian and governmentwebsitesallowuserstoviewanddownloadDopplerimages.Inthecaseofarainflagwithoutaninsitumetstation,theseDopplermapswillbeusedtoverifythepresenceandintensityofrain.Finally,there are several satellite products (e.g. theGlobal PrecipitationMission) that identify rain events,althoughtheseproductsarefarlessreliablethaninsitumeasurementsorlocalDopplerradar.Theseproducts will be used to validate the rain flag in cases where the primary two validationmeasurementsareunavailable.

6.5.11 IceflagvalidationThe SWOT ice flag is designed to indicate where and when SWOT-observable water bodies arecoveredwith snow and ice. Ice flaggingwill be nominally based upon the use of optical satelliteimagery,whichcanrobustlydifferentiateiceandwateratarangeofspatialresolutions.Wewillusedailymoderate-resolutionimagesfromMODISand/orVIIRStodetecticebreakuptiminginthepan-Arcticregion(wherethemajorityofriverandlakeiceoccurs).Wewillusehigher-resolutionimagery(e.g. Landsat, Sentinel 2, ASTER) to assess the detailed patterns of SWOT ice flag accuracy duringbreakupandvalidatetheiceflags.However,becauseSWOTandtheseopticalimagerswillnotbeinsynchronousorbits,itwillbepossibletovalidatetheiceflagusingspace-basedassetsaloneonlyatthe reach scale. In order to validate pixel-scale flags,wewill obtain airborne optical imagery of aportionoftheTananaRiver,AlaskacoincidentwithSWOToverflightsduringandbeforeicebreakup.IcecoverwillbemappedusingbothSWOTandtheopticalimagery,andtheresultswillbecompared.

6.5.12 LandflagvalidationTheSWOTlandflagisdesignedforobjectsthatgiveareasonablybrightradarreturnbutthatarenotwater, and are thus commission classification errors. In essence the land flag is the inverse of thewatermask,andwillbevalidatedinthesameway.However,validationoflakeandriverinundatedarea will leverage other radars that may have similar commission issues as SWOT. Therefore,validationoflandflaggedproductswillrelyprimarilyonaerialandsatelliteimageryandinsituGPSmapsofwaterextentwhereavailable.Theonlyadditionalresourcesrequiredforthisvalidationwill



66

bestafftimetocheckeachlandflagobject,astheinsitudataandimageryrequiredforvalidationwillbecollectedforotherpurposes.

6.5.13 GeolocationvalidationThereisatradeoffbetweenthegeolocationaccuracyandspatialprecisionofSWOTdata.ToensurethatSWOTproductsmeetthegeolocationsciencerequirement(2.6.3a),comparisonoftheirSWOT-derivedpositionwillbemadeagainstpositiondataofknownprecision.SinceSWOTdatawillhaveirregularspacing,geolocationvalidationmustbeperformedonapixelbypixelbasis,andvalidationwillbeperformed fordistinct targets thatareeasily identified. Ideally, thisvalidationwillbemadeusing precise GPS coordinates of the corner reflectors already deployed at Tier I field sites. Othertargetsforgeolocationvalidationwillbedeterminedbasedonthesiteconditionsateachcal/valsite,andappropriateobjectsthatcanreliablybedetectedinSWOTdatashouldbeidentifiedandtheirGPSpositionsrecorded. In thecaseswhere theseobjectscannotbe found in the field,aerialorsatelliteimagery shouldbe obtained to cross reference SWOTdata and geolocation error determined fromtheseproducts.

6.6 SurfaceWaterDataProductValidationInadditiontothevalidationofeachcomponentofSWOT’ssurfacewatererrorbudget(section6.5),themainparameters of theLevel-2hydrologyproductswill be validated. Inmany cases, thedataproductvalidationwilltakeplaceconcurrentlywitherrorbudgetvalidation.Whenthisisthecase,itwillbenotedbelow.

6.6.1 PixelcloudproductvalidationTheSWOTpixelcloudproductisalevel2productintendedtoprovideaccesstoheight,water/landclassification, and relevant quality flags in their rawest form. It will include both geolocated andslant/range coordinates and will be the basis for development of raster and vector productsdescribed below. There is no independent requirement on pixel cloud height accuracy that isdifferent fromtheoverallheightrequirementsdescribed insection6.4.2. Assuch,wewillvalidateheights in the pixel cloud by comparing spatial averages of pixel heights within a water body atrelevantscalesof(250m)2and1km2againstfieldmeasurementsofheightcollectedasdescribedinsection6.5.2.Similarly,wewillvalidateclassificationaccuracyagainstmeasurementsfromairborneinfraredimageryavailableatsubstantiallyhigherresolutionthanSWOT.WewillfocusonvalidatinginundationextentclassificationaccuracyatSWOT-relevantscalesdescribedaboveforheight.Otherquantities, including ice, rain, and layover flags, will be validated on a pixel-by-pixel basis asdescribedinsection6.5.

6.6.2 RivervectorproductvalidationPass-basedvectordataproductforriversPass-based river vector productswill include point, line, and/or polygon features that are derivedfromthepixelcloudofjustoneSWOToverpass.Theywillbetheprimaryrepositoryforreach-scaleheight, slope, width/inundation extent, and discharge data on rivers. Unlike the raw pixel cloudproduct, the vector productwill have already aggregated SWOTheight and classification data intodefined reaches. Values forheight, slope, and inundation extent in these reacheswill be validatedusingmethods described in Sections 6.5.2, 6.5.3, and 6.5.8. However, the successful translation ofSWOT data from pixel cloud to reachwill also be evaluated. Wewill compare the flow length ofreaches derived from SWOT datawith similar reaches derived from high-resolution airborne (e.g.AirSWOT) or satellite (e.g. SPOT, WorldView) imagery over Tier 1 validation sites including the



67

Willamette,Garonne,Connecticut, andTananaRivers. This comparisonwillbecritically importantforunderstandingtheerrorcharacteristicsofSWOT-derivedslope,whichdependsontheaccuracyofbothSWOT-derivedheightsandthelengthoftheriverreach.Wewilldirectlycompareslopesinthepass-basedvectorproductsagainst slopesmeasured in situ, asdescribed insection6.5.8. Wewillalso compare reach-averaged heights and inundation extents against manually aggregated valuesfromthepixelcloudproductandfield-measuredvaluesatTier1sitesinordertoensureconsistency.Cycle-basedvectordataproductforriversIn addition to producing pass-based vector products, vector products will also be created thatincorporatedatafromanentireSWOTorbitcycle(21days).Unlikepass-basedproducts,thesecycle-basedproductsoftencannotbeeffectivelyevaluatedagainstinstantaneousmeasurementsofheight,slope,inundatedarea,andotherquantities. Inthecaseofheightandslope,wewillrelyonexistingstreamgaugesandtheinstallationofnetworksoftemporarygaugesthatwillmeasurewatersurfaceelevation every 15minutes (or less), as described in sections 6.5.2 and6.5.8. Validation of cycle-based inundationextentwillbemorecomplex,as there isno feasiblemethodofdirectlyobservingvariationsovera21-daytimeframe.Instead,forTier1siteswithhigh-qualitybathymetricDEMswewilluseinundationextent-stageratingcurvesasdescribedinsection6.5.3.

6.6.3 LakevectorproductvalidationPass-basedvectordataproductforlakesPass-based lake vector products will consist of polygons derived from the pixel cloud productrepresenting lakeboundaries.Theywill be theprimary repository forwhole-lake valuesof height,inundationextent,andrelevantqualityflags.Unliketherawpixelcloudproduct,thevectorproductwillhavealreadyaggregatedSWOTheightandclassificationdata. Valuesforheightandinundationextent forwhole lakeswillbevalidatedusingmethodsdescribed inSections6.5.2,6.5.3, and6.5.8.However, the successful translation of SWOT data from pixel cloud to whole lake will also beevaluated. Wewillcompare the inundatedareasandboundariesof lakesderived fromSWOTdatawith similar values derived from high-resolution airborne (e.g. AirSWOT) or satellite (e.g. SPOT,WorldView) imagery over Tier 1 validation sites including Lake Tahoe, the Prairie Potholes, theYukonFlats,mountainlakesinCalifornia,andothertargetsasdescribedinSection6.5.3. Wewillalsocomparewhole-lakeheightsand inundatedareasagainstmanuallyaggregatedvalues fromthepixelcloudproductandfield-measuredvaluesatTier1sitesinordertoensureconsistency.Cycle-basedvectordataproductforlakesInaddition toproducingpass-basedvectorproducts, lakevectorproductswillalsobecreated thatincorporatedatafromanentireSWOTorbitcycle(21days).Unlikepass-basedproducts,thesecycle-basedproductsoftencannotbeeffectivelyevaluatedagainstinstantaneousmeasurementsofheightandinundatedarea. Inthecaseofheight,wewillrelyontheinstallationofnetworksoftemporarygaugesthatwillmeasurewatersurfaceelevationevery15minutes(orless),asdescribedinsection6.5.2. Validation of cycle-based inundation extent will be more complex, as there is no feasiblemethodofdirectlyobservingvariationsovera21-daytimeframe.Instead,forTier1siteswithhigh-quality bathymetric DEMs we will used inundation extent-stage rating curves as described insection6.5.3.

6.6.4 RasterproductvalidationAmethodwillbeprovidedtogenerateapass-basedrasterproductfromthepixelcloudproductatarangeof spatial resolutions. This rasterwill include (at least) informationon location, land/water



68

classification,height,andbrightness. Wewillvalidatetherasterdataproductbycomparingittoinsituandairbornedataonheightandinundationextentasdiscussedinsections6.5.2and6.5.3.

6.7 DischargeCharacterization

6.7.1 CharacterizationofderivedbathymetryAclassofmodelscurrentlyavailable toderivedischarge fromasetofheight,widthandslopealsoneedsabathymetryoftheriverreach.Whennotavailable,thisbathymetryispredictedtogetherwiththedischarges.InthecasewhenmodelsofthiskindisretainedbytheDischargeWorkinggroupforthe estimate of SWOT discharge products, the bathymetry predicted by the algorithms will becharacterizedinthetwofollowingways:

1. The bathymetry will be characterized by comparison with actual cross sections, mostlycollectedduringADCPmeasurements (seesection6.7.2).Adatabaseof suchcrosssections,preferablyleveled,willbeconstituted.USGSpossesseshundredsofthousandsofsuchcross-sections that could bemade available to the project, contingent uponUSGS involvement. Itwouldbepreferableifthesecrosssectionswereleveled.Also,suchadatasetofhundredsofcrosssectionsexistfortheAmazonbasin,thelowerpartoftheGBM(Gange-Brahmaputra,-Meghna) river system, the major Brazilian rivers, French and Italian rivers. It is alreadyagreedthatthesecrosssectionswillbemadeavailableandcanbeusedbytheproject.Weareaware that such cross sections exist for some rivers running in other South American andEuropeancountriesbuttheirintegrationintothedatabasewillbemadeonlyona“besteffort”basis,dependingonthegoodwillof theagenciespossessingthesedatatoprovidethemforfree.

2. The SWOT bathymetry will be characterized by comparison with river bed elevationsestimatedfromotherindependentsources,inparticularfromratingcurveswhichusewaterdepth toderivedischarge(insteadof thewaterelevations).Adatabaseofsuchvirtualriverbed elevations (VRBE, in opposition to bed elevations actually obtained by directmeasurement)willbeconstituted.ScientificprojectshavealreadyproducedsuchadatabaseintheAmazonbasin.ThatfortheCongobasinhasbeenconstitutedin2017,withanongoingworkonNigerriver

6.7.2 Characterizationofderiveddischarge SWOT-deriveddischargeisacriticalhydrologyproductexpectedtobeofgreatinteresttotheinternationalhydrologycommunity.AprimarypurposeoftheSWOTdischargeproductistoproduceestimates in ungauged basins and in regions where current discharge knowledge is spatiallydiscontinuous.While isbydefinition impossibletocharacterizeaSWOTdischargeproduct inthesesituations,dischargecal/valactivities forriversofknowndischargearecritically important for themissionasawhole. Characterization of discharge is straightforward, and is performed by directly comparingSWOT derived-discharge to some known discharge. Objective characterizationwill be achieved bycalculatingasuiteofmetricsfirstproposedbyBjerklieetal.,2005.ThesemetricsincludetheRMSE,RRMSE, model selection criteria (MSC), and the mean and standard deviation of each of the raw,relative, and log residuals between SWOT-derived and measured discharge. These metrics allowassessmentofdischargebias,stability,andtotalerror.Thecharacterizationofdischargewillbeperformedonthebasisof:

1. theSTprojectsselectedbytheROSES/TOSCAcall.



69

2. The discharge algorithm(s) finally selected by the ST to be implemented in theproductionchain.

TheprimaryCal/ValactivityforcharacterizingdischargeistoproduceareferencedischargeagainstwhichSWOTmaybecompared.SWOT-deriveddischargewillbecomparedwithdischargederivedinsitubytheCal/Valteamusingthefollowingmethodologies:

6.7.2.1 DirectdischargemeasurementPerhapsthemostaccurateandstraightforwardwaytoproduceareferencedischargeisbymeasuringdischargedirectlyinthefieldatthetimeofaSWOToverpass.Today,themostup-to-dateinstrumentto measure discharge is an ADCP with GNSS positioning. In the case that new technologies areavailable at the time of the launch, these will be utilized. ADCP instruments are expensive, andmaking spatially distributed measurements with them is time consuming (although orders ofmagnitudemoreefficientthanprevioustechnology).Therefore, thistechniquewillbe implementedataverylimitednumberofsitesdependingonthefundingcapabilities.ThelocationswillbeselectedwithinthesitesincludedinthelistofofficialTier1sites.Duringthe1-dayphase,themeasurementswillbeperformeddaily,ascloselyaspossibletothetimeofoverpassingbySWOT(thesecannotbeexactlysimultaneoussinceADCPmeasurementscantakeatleastonehourforlargerrivers).Duringthenominalphase,insitumeasurementswillbemadeatthedayofpassing(e.g.twicepercycle)andshouldattempttoincludeatimewindowcoveringthelargestandlowestflows(halfofahydrologicalcycle)Resources required: The ROSES/TOSCA call will determine the personnel that will perform thesemeasurements. Foreachdischargemeasurementinsitu,fundsareneededforpersonneltravelandlodging, and it is expected that teamswill have access to or have requested funding for an ADCP.Additionalfundsareneededforwatercraftandtransportationofwatercrafttothefieldsites.

6.7.2.2 IndirectdischargemeasurementsConsideringtheimpracticalityofdirectlymeasuringdischargeatnumerousworldrivers,theCal/Valteamwill leverage stream gauges and rating curves to produce reference discharge for numeroussites.MostofthedischargevaluespublishedintheWorld’sbasinsarederivedfromrivergaugesanda rating curve (RC, stage/discharge relationships), and these gauges can provide continuousestimates of discharge at a station. SWOT-derived discharge will be compared to these rateddischargesata listofsitesestablishedby theScienceTeamprior to the launch,andgaugeswillbeusedtocharacterizethedischargeatdifferenttimescales(instantaneousdischarge,seasonalmean,annual mean). Such RC are already available for thousands of USGS gauges, and the USGS shouldmake them available to the project through a proposal at the ROSES call. Scientific projects arecurrentlyestablishing suchRCovera largevarietyof rivers (from the typical scalesof100m3/s to100,000m3/s)inthelargebasins(Amazon,Congo,GBM).ThisdatabasewillbemadeavailabletotheprojectforthecharacterizationoftheSWOTproductatalargescale.Resourcesrequired:Itisexpectedthatthesegaugedatawillbeacquiredviathepublicdomain(intheUSandFrance),bytheUSGSpendingaROSESproposal,orviaexistingandongoingscientificwork.However, for Tier 1 Cal/Val sites, gauges and rating curves should be established prior to launchwithinthetargetSWOTreaches.Theinstrumentationrequiredtoestablishthesegaugesisidenticaltothoseneededforslopevalidation,sofundingandschedulefortheseactivitiesisidenticaltothoselistedin6.5.8.Additionalfundingwillbeneededtomakeinsitumeasurementsofdischargeatthese



70

gaugespriortolaunchtoestablishtheratingcurveateach,andextrafundsshouldbeallocatedtotheslopevalidationinstallationssopersonnelcantakethetimetodeployandADCPwithinthereach.

6.7.2.3 ModeloutputWhile field measurements and gauge estimation of discharge are highly respected and accuratemeansofproducingreferencedischarge,hydraulicmodelsarealsoabletoproduce dischargewithgoodaccuracy inmanycases.Atsomeof theSTsites(see forexample theGaronnesite inFrance),high accuracy hydraulic models have been developed and produce accurate discharge estimatesgiven top-of-reach in situ inputs. These estimateswill beused to characterize the SWOTproducts.The list of sites/models to be used thiswaywill be established in agreementwith the ST and thedischarge Working Group. All models will be furnished by members of the ST, and any modeldevelopmentwilloccurintheframeofROSES/TOSCAproposals.

6.7.2.4 Statistical/morphologicalestimatesofdischargeFinally,theabovemethodologiesareabletoproducereferencedischargeforsinglechannelswithafairdegreeoffieldlabororpreviousinfrastructuredevelopment.SinceSWOTestimatesofdischargeareperhapsofmostinterestinungaugedbasins,classicmorphologicalestimatesofdischargewillbeused to broadly characterize discharge in these regions. It has long been established that meandischarges rely on themorphologyof the river reaches (or, equivalently, that themorphologyof areachisforcedbytheamountofwaterthathastoflowthroughit),inparticularitsmeandepth,widthand slope.At a global scale, the SWOTdischargewill be characterizedwith respect to the existingrules of thumb. These standard practices include development of regional power laws betweendrainagearea,width,depth,andslopeofriverchannels.Inthesecases,thesuiteofmetricsproposedby Bjkerlie et al. will not be used to characterize discharge, and characterization will be morequalitativeasbefitsthenatureofthereferencedischarge.



71

7 SWOTCAL/VALSITESThis section describes the sites that will be instrumented by the SWOT project, ST members, orcollaboratingorganizations, andhow theseCal/Val siteswill beused to accomplish the calibrationandvalidationneedsdescribedabove.Italsooutlinesthepre-launchandpost-launchactivitiesthatwillbeperformedforeachsite.TheresponsibilityformaintenanceoftheCal/ValsiteswillvarydependingonthepurposeofthesiteandtheneedforittoderiveorvalidateparametersrequiredbytheSWOTproject.InadditiontoUSor French project supported sites, it is envisioned that community sites will be developed by thescienceteamorbecontributedbyotheragencyorforeignpartners.Below,weindicatetheprimaryresponsibilityforeachsite,althoughsitescanservemultiplepurposes.WealsoproposetoleverageexistingJasonCal/Valsites,supplementingtheircapabilitiesasnecessary.

7.1 OceanCal/ValsitesOcean Cal/Val sites are divided in two complementary categories: absolute range/SSH bias andrelative 2D SSH and its derivatives (e.g. geostrophic currents). The former extensively leveragesexistingCal/ValsitesusedovertwodecadesfornadiraltimetersandwillmonitorSWOTbiasesandtrends in different regions during the 3-year nominal phase. The latter combines existinginfrastructures andnewmeasurement techniques during the fast-sampling phase and the nominalphase.

7.1.1 ValidationoftheabsoluteSSHbiasOver more than two decades of nadir altimetry Cal/Val, the ability of precise in situ Cal/Val tomeasure the local bias has been largely demonstrated. These techniques are complementarywithglobal metrics (e.g. statistical or global in-situ networks). Having at least two or three absolutevalidation sitesmakes it possible for local/global comparisons to infer the influence of errors thatdependson in-orbit andgeophysics conditions (e.g. geoid, tides, corruptionby coastal layover).Tothatextent, theoceancalibrationsitesdescribedinthissectionfeatureprovideawiderangeofseastateandgeographicconditions.NotethatneithertheUS(Harvest)norFrench(Corsica)long-termcalibrationsitesforabsoluterangebiaseswillbeobservedduringthefastsamplingphase,buttheywillformacriticalcomponentofthelong-termCal/ValmonitoringofSWOTperformanceduringthesciencephase.TheKaRInswathwillobservetheBassStrait(Australia)long-termcalibrationsiteduringthefastsamplingphase.

7.1.1.1 HarvestCal/ValSite

7.1.1.1.1 SiteDescriptionTheHarvestOilPlatform(Figure27)is locatedabout10kmoffthecoastofcentralCalifornia,nearVandenbergU.S.AirForceBase,thesitefortheupcomingJason-3launch.Theplatformisfixedtotheseafloorandsitsinabout200mofwaternearthewesternentrancetotheSantaBarbaraChannel.ConditionsatHarvestaretypicaloftheopenocean:windwavesandswellaverageabout2m,thoughwaves up to 10 m have been experienced during powerful winter storms. Built in 1985 andoperationalsince1991,Harvestcontinuestoserveasproductionplatform,drawingoilandgasfromthe Arguello reservoir. Harvest has also served as the NASA prime calibration site for theTOPEX/POSEIDON(1992–2005),Jason-1(2001–2013)andOSTM/Jason-2(2008–)missions,andassuchisanimportantinternationalresourceforthestudyofsealevelfromspace.TheJason-3(2016–)



72

andJason-CS(2020–)missionswillfollowthesamegroundtrack,implyingthatHarvestwillcontinuetoserveacrucialroleinvalidatingdatafromprecisespaceborneradaraltimetersystems.

Figure27. MapofHarvestPlatformvicinity showingground tracks for Jason referencemissions (left).Theplatform (right)hosts one of the oldest tide gauge/GPS collocations in the world, and has provided for continuous monitoring and of theTOPEX/PoseidonandJasonseriesofreferencemissionssince1992.

Harvest offers a number of advantages as a spaceborne altimeter calibration site. The platform islocated sufficiently far offshore so that the area illuminated by the traditional (pulse-limited)altimeter iscoveredentirelybyoceanwhenthesatellite isdirectlyoverhead.Atthesametime,theplatformitselfissmallenoughsothatitdoesnothaveameaningfulinfluenceonthereflectedradarsignal. Equally important, the open-ocean environment implies that the spacecraft measurementsystemsaremonitoredintheconditionsunderwhichtheyaredesignedtobestoperate.

7.1.1.1.2 SiteGoalsTheprincipalgoaloftheHarvestexperimentismaintainthisvitallong-termcalibrationrecord,andto improve the skill with which the experiment can detect systematic errors in the altimetermeasurementsystems.

7.1.1.1.3 SiteInstrumentationThe Harvest experiment features carefully designed collocations of space-geodetic and tide-gaugesystems to support theabsolutecalibrationof thealtimetric sea-surfaceheight (SSH).Theprimarytidegaugesystemisadual-redundantNitrogenBubbler/pressuretransducerfromNOAA,whichhasprovided continuous data (with a few short exceptions) since 1992. A lidar is operated by theUniversityofColorado,and twonewradargaugesareslated for installationbefore Jason-3 launch.Thesecompetingtechnologiesformeasuringthewaterlevelwillofferanunprecedentedopportunitytocharacterizethesystematicerrorsexperienced indynamicsea-stateenvironments.TheplatformGPSstationisoneoftheoldestcontinuouslyoperatingsitesintheInternationalGNSSService(IGS)network. A newGPS station (and antenna)was installed at a different location on the platform inearly 2015 to provide competing measurements of the platform subsidence and zenith wettroposphere path delay under different multipath conditions. The wealth of information from theHarvest experiment underscores the unique contributions of a dedicated, well-instrumented andcontinuously maintained calibration site. The platform sensors can be complemented by buoycampaignsforparticularapplications.



73

7.1.1.1.4 Pre-launchSiteCharacterizationPre-launchactivitieswill focusonadaptingtheHarvestexperimentalstrategytosupporttheSWOTmissionrepeatgroundtrackconfiguration.AsshowninFigure28,Harvestisnotintheswathforthe1-d fast repeat phase of the SWOTmission. Regional calibration techniques (using, e.g., mean seasurfaceprofiles),supplementaltidegaugeandprecisionmooredbuoysareallcandidatesforbridgingthe gap from the open-ocean SWOT ground track to Harvest. We will in particular leverage theresults from a current JPL/NOAA initiative to develop a prototype precision GPS buoy for long-durationmonitoringofwaterlevelandatmosphericproperties.ThenominalSWOTorbithasagroundtrackthatpassesveryclosetoHarvest(Figure29).Thisisveryfavorable approach from theopenocean (similar to thatof the Jason referencemissions), andwillenablearobustdeterminationofthebiasoftheSWOTnadiraltimetersystemagainstthebackdropofthe TOPEX/Poseidon and Jason climate-scale calibration record from the platform. For this 21-drepeat, Harvest also lies in the swath of a descending pass (307), a geometric configuration thatpromisestolendnewinsightsonthelinkfromthenadirmeasurementtotheswath.

7.1.1.1.5 Post-launchCal/ValActivitiesPost-launch activities will focus on the careful addition of SWOT data to the Harvest calibrationrecord.Wewillalsostudytechniques(e.g.,relyingondistributedbuoys)toprovideajointcalibrationoftheswathandnadirgroundtrackasthecalibrationpointisoverflown.

Figure28. LocationofSWOT1-d fast repeat swath tracks in relation toHarvest.The twobandsoutlinedby thewhite linesdepicttheswathtracks.



74

Figure29. LocationofSWOT21-drepeatswathtracks inrelationtoHarvest.Theblue linepassingclosest toHarvest is thegroundtracktracedbythenadirpointofascendingpass294.Harvestisalsointheswathofthedescendingpass307.



75

7.1.1.2 CorsicaCal/ValSite

7.1.1.2.1 SiteDescription

Figure30.ConfigurationoftheCorsicacalibrationsite.Thecolorcontourmapscorrespondtothe“localgeoids”

TheprimaryfacilitiesfortheCorsicacalibrationsitehavebeenoperationalatSenetosasince1998tomonitor the TOPEX/Poseidon and Jasonmissions (Figure 30). The site was expanded to facilitatecomparison at two additional comparison points, Ajaccio and Capraia, also adding the ability tomonitor the Envisat and SARAL/AltiKa satellites. This provides a unique opportunity to crosscalibrateall thesedisparatemissionswithcommonprocessesandstandards.Forexample, thewettropospheric path delays for comparison to the satellite radiometer estimates can be determinedfrom GPS and ground-basedmeteorology stations located at both Ajaccio and Senetosa. The closeproximityofeachsitealsoprovideseconomicandlogisticaladvantages,suchastheabilitytousetothesameGPS-basedsealevelmeasurementsystemstoregularlyperformindependentcalibrationsatthe various comparison points. An evolution of the “overhead” calibration methods to a regionalapproachhasbeenalsodevelopedbasedontheextendedCorsicasitecapabilities(Cancetetal.2013;Janetal.2004).

7.1.1.2.2 SiteGoalsThe traditional “overhead” concept of in situ altimeter calibration involves the direct satelliteoverflightofathoroughlyinstrumentedexperimentsite.Itisessentialthatsuchacalibrationsitehassome means of observing sea level in situ (using for example a conventional tide gauge, oceanmooring or GPS-based sea level measurement systems) and subsequently tying the sea levelestimatestoaterrestrialreferenceframecomparabletothesatellitealtimeter.Inanidealsituation,



76

theexperimentsiteislocatedonarepeatinggroundtrack(orbetterstillacross-overofanascendinganddescendingaltimeterpass), sufficientlyout in theopen-ocean toavoidcontaminationofeitherthealtimeterorradiometerfootprintsbytheland.Onthecontrary,acoastalsitesuchasCorsicacanbeusedasavaluabletooltoestimatetheerrorsinaltimetrywhenapproachingthecoast,eitherforthe range itself or the associated corrections (Wet troposphere) as well as for Significant WaveHeight.Two distinct methodologies exist for the measurement of the in situ Sea Surface Height at acomparisonpointthatissubsequentlyusedforcomparisonagainstthealtimeterSSH.Thetechniquesandunderlyingalgorithmsarequitedisparatedependingontheparticularapplicationandwillnotbedeveloped here (Bonnefond et al. 2011).However, itmust generally consider— either directly orindirectly—geophysical,oceanographicandatmosphericphenomenathatcausethevariationinsealevelovertimeandspace.

Directmethod:Inthiscase,SSHisphysicallyobservedataoffshorecomparisonpointusing,forexample,GPS-basedsystemsoroffshoreinstrumentation,asitisthecasefortheHarvestplatform(Hainesetal.2003).

Indirectmethod:Inthiscase,theSSHmeasurementinvolvestheobservationofsealevelaway

from the comparison point, typically using a tide gauge at nearby (typically coastal) location. TheoffshorealtimetricSSHisthen“transferred”or“extrapolated”atthelocationoftheinsituinstrumentthrough theuse of precise regional geoidmodels, and inmany cases, numerical tidemodels. Tidalmodels are not used in Corsica (especially at Senetosa) because the estimated impact, even usinghigh-resolutionmodels, is at the level of a fewmillimeters over the considered area (Cancet et al.2013).InthecaseoftheCorsicaexperimentreportedhere,bothdirectandindirectmethodsareused.Thus,thelocalgeoidsbuiltundertheT/PandJasongroundtrack#085attheSenetosaCape(Bonnefondetal., 2003b), and under the Envisat and SARAL/AltiKa ground track #130 near Ajaccio are keycomponents to imposing the datum for our absolute calibration process when using the indirectmethod. Details about the SSH bias processing (SSHaltimetry – SSHin situ) and the general parametersusedarenotrecalledherebutcanbefoundinBonnefondetal.(2003a,2011and2015).InCorsicatwo independent instruments(tidegaugeandGPS-basedsea levelmeasurementsystems)areusedwithdifferencesintermsofprocessingtocomputetheSSHbias.As alreadyplanned, the 1-day orbit ground track is too far fromCorsica to use any of the currentinstrumentations,sowewillfocusourCal/Valactivitiestothenominalphase.IfCorsicaisselectedtobeoverflown,Figure25 illustrates thebest scenario,usingadescending track thatpassover the2existinggeoidsrespectivelyatAjaccioandSenetosasites.ThisshouldallowperformingtheCal/Valforbothnadirandswath.Thisconfigurationwillpermit tovalidateswathmeasurement invarioussituations: therightswathwillbealways inopen-oceanconditions,while leftswathwillencountercoastal conditions and then permit to study the impact of land contamination (Bonnefond et al.,2015).Moreover, “Boussole” a buoydesigned for Cal/Val of ocean color sensors (MERIS, SeaWiFS,and MODIS) is located in the right part of the swath (Antoine et al. 2008; http://www.obs-vlfr.fr/Boussole/html/home/home.php).Thisbuoy(Figure31left)ismaintainedintheframeworkofSentinel-3missionbyCNESandESAandmaybeusedtoalsotoinstalldedicatedinstrumentsintheframeworkofSWOTmission.



77

Figure31.Possible locationofSWOTgroundtrackduringthenominalphase.Left:overviewwithnadir lightblueandswathlimits in green. Right: zoom over Ajaccio and Senetosa sites with othermission ground tracks (red = Sentinel-3A, yellow =SARAL/AltiKa,Envisat,…,purple=T/P&Jasons)

7.1.1.2.3 SiteInstrumentationTheSenetosasiteisequippedwith4tidegauges,locatedoneachsideofT/PandJasongroundtrackdedicated to the altimeter calibration process. A permanentGPS station is operational since 2003.Surveysofthegeodeticmarkersandtidegaugelocationshavebeenundertakenregularlysince1998and the repeatability of the GPS solutions and the optical leveling are below 1 cm and 5 mmrespectively.AtAjaccio,apermanentGPSstation (IGN)andanautomatic radar tidegauge (SHOM)havebeeninstalledsince1999.Since 2000, a GPS buoy has also been used in the calibration process at Corsica: the GPS buoy isdeployedfor~1hrsurroundingoverflights(~10kmoffshore)wheneversea-stateconditionsarenottooharshtoensuresafenavigation.Bonnefondetal. (2015)provideanexampleofwhereboththedirectandindirectmethodologiesarecombined.Amajorchangehasbeenimplementedsince2012:the traditional waverider buoy was replaced by a zodiac (termed GPS-zodiac hereon) for bothSenetosaandAjacciocalibrationsites.Themainreasonisthatawaveriderbuoycan’tbetowedbyaboat.Asaconsequence,handling(fromseatoboatandviceversa)leadstolossesoftheGPSsignalthataffecttheambiguityresolutioninthedataprocessing.TheuseoftheGPS-zodiacinsteadoftheprevious buoy avoids these problems and thus allowed us to record SSH continuously at 1Hz(Bonnefondetal.2015).AtSenetosa,aweatherstationhasbeeninstalledatthelighthousesince2000,neartheGPSreferencepoint. The main goal of this station is to provide atmospheric pressure to correct the tide gaugemeasurements and to derive the dry component of the tropospheric correction. AtmosphericpressuresfromAjaccio(~40kmNorth)andFigari(40kmEast)providedbyMétéo-Francearealsousedasbackupincaseoflocalstationoutages.

7.1.1.2.4 Pre-launchSiteCharacterizationSynopsisofthepre-launchactivities:



78

● geoidextensionusingtheCalNaGeosystem(Figure21)atdedicatedlocationundertheSWOTswath(and/ornadir).ThealreadyplannedgeoidextensionintheframeofSentinel-3missionneedsveryfewchangestobeadaptedtoSWOTifthenominalorbitisphasedasillustratedinFigure25(right)

● feasibilitystudytoinstallnewsensors(tidegauge,GPS,…)onthe“Boussole”buoy(Figure25left)

● regular deployment of the CalNaGeo at the previous mapped locations to derive thedifferences in termofoceanic signal compared to coastal tidegaugesmeasurements.Thesemeasurementswillbeusedasconstraintsforthedevelopmentofaspecificoceandynamicsmodelwithhighspaceandtimeresolutions.

● adaptationoftheregionalcalibrationmethodtoSWOT● useofthesimulatortogenerateSWOTmeasurementandderivesimulatedSSHbiases:

- fromdirectandindirectmethods- fromtheregionalcalibrationmethod

7.1.1.2.5 Post-launchCal/ValActivitiesSynopsisofthepost-launchactivities:

● regulardeployementoftheCalNaGeoatthepreviousmappedlocationsatthetimeofSWOToverflights

● useofthe“Boussole”measurements(sealevel)inthecalibrationprocess(directmethod)● deriveSSHbiasestimeseriesusing:

- directandindirectmethods- regionalcalibrationmethod

● compareradiometerwettroposphericpathdelaystothosederivedfromGPSmeasurements.

7.1.1.3 BassStraitCal/ValSite

7.1.1.3.1 SiteDescriptionThe Bass Strait calibration and validation site has been used in the derivation of absolute biasestimatesfortheJason-classsatellitealtimeterssincethelaunchoftheTOPEX/Poseidonmissionin1992. The site is one of three primary validation facilities contributing to the Ocean SurfaceTopography Science Team, and the sole site located in the Southern hemisphere. The historicalcomparison point is located off the northwest coast of Tasmania, Australia (40° 39’S, 145° 36’ E,Figure26),andisnowpermanentlyinstrumentedwithasuiteofmooredoceanographicinstruments.Themooredoceansensorsenabletheproductionofaprecisetimeseriesofseasurfaceheight,withanabsolutedatumimposedthroughepisodicdeploymentsofGPSequippedbuoys.Additionaldataisobtained from the land based GPS and tide gauge located in Burnie (Figure 32). This historicalcomparisonsite,togetherwithcomparisonpoints(CPs)forSentinel3Aand3B(seelater)areshowninFigure32–alsoshownisthenominal21daySWOTorbit.



79

Figure32.TheBassStraitCal/ValfacilitywithJason-series“historical”comparisonpoint(CP),Sentinel-3Aand3BCPs.SWOT21d orbit is shown in cyan, with inner (10 km) and outer swath in green. SWOT orbit fromSWOT_Science_Option4_ScienceP59_over_CalValP3_Swath_10_60.

In 2015, the Bass Strait facility will be augmented with a second comparison point to facilitatevalidationof theSentinel-3Aaltimeter (S3ACP,40°33’S,145°34.5’E, Figure32).PendingongoingsupportfromAustralianfundingagencies,weplantodeployathirdcomparisonpointtotheWestofthe primary comparison point to enable validation of the Sentinel-3Bmission (S3BCP, 40° 33.5’S,145°06’E,Figure32).ThislocationispositionedclosetotheHuntergroupofIslandsinBassStrait,providingavalidation target that is closer to the coast (~12km)and ina regionofmore complexocean dynamics compared with the primary Jason-class comparison point. Together, these siteslocatedinthesoutheastcornerofBassStrait,augmentedwithhighresolutionmodelingwillprovidethebasisfortheAustraliancontributiontoabsolutevalidationfortheSWOTmission.The21dayorbit forSWOThas the followingcharacteristicswithrespect to thehistorical,S3AandS3BCPs(referFigure32):

● SWOT nadir crossover is ~50 km to the north of historical CP (difficult to instrumentgivenseafloorsedimentatthislocation).

● The historical and S3A CPs just within inner swath of Desc Pass 65 (9 km from nadirtrack).GPSbuoyscouldbeusedtoobserveSSHslope,andhighresolutionmodelsusedfordifferencesintide.TheseCPswouldalsobesuitableforKaRINofPass328(Asc).

● TheS3BCP is justoutside innerswathofPass328(Asc)butsuitable forKaRINofPass328(Asc)andPass65(Desc).

Thenominal1dayorbit forSWOT(Figure33)isnadir~35kmtotheeastofthehistoricalCP.SitesuitableforKaRINonly.



80

Figure33.AsperFigure26butshowingtheSWOT1dayorbitfromSWOT_Science_Option4_ScienceP59_over_CalValP3_Swath_10_60.

7.1.1.3.2 SiteGoalsThe primary goal of the Bass Strait validation site is to provide cycle-by-cycle absolute biasestimationfortheSWOTmission.Ourcontributionswillbecenteredonthelocationsdesignatedbythehistorical (40°39’S,145°36’E)andS3B(40°33.5’S,145°06’E)comparisonpoints (seeFigure32).Highspatialandtemporalresolutionoceanmodelingwillaugmentinsitudatathatwillincludevarious point based absolute sea level and surface topographymeasurements, and network basedestimatesofintegratedtotalzenithdelayfromthetroposphere.WeawaitdecisiononthefinalSWOTgroundtracktofullydetermineourexperimentdesign.

7.1.1.3.3 SiteInstrumentationThe primary instrumentation will include moored oceanographic sensors at the two comparisonpoints.Themooredarraysprovidepreciseseasurfaceheightona5-minutetimebasederivedfrombottom pressure and integrated temperature and salinity observations through thewater column.Wewill investigate the acquisition ofwave data from upward lookingmoored sensors, or surfacebasedwavebuoys.Currentswillalsobeobservedthroughthewatercolumn.TheabsolutedatumofthesealeveldatawillbederivedusingepisodicGPSbuoydeployments,withprocessingagainstlandbasedGPSreferencestations.TheS3Bcomparisonpointat40°33.5’S,145°06’E(Figure32)isclosertothecoastandneartheHunterIslandgroupinBassStrait.ThiswillenablethecollectionofadensenetworkofintegratedwatervapourmeasurementsfromGPSstationstobedeployedinthearea.Tofacilitatehighresolutionoceanmodeling,campaignbasedacquisitionofseasurfaceheightdatafromautonomousvehicleswillbeinvestigated(pendingAustralianresources),inadditiontoacquisitionofsupplementarybathymetrydatatoaidmodeldevelopment.

7.1.1.3.4 Pre-launchSiteCharacterizationThe various comparison pointswill bewell characterized prior to the launch of SWOT given theirintendeduse for Jason-3, Sentinel-3A and Senitnel-3B.Aspects of this characterizationwill include



81

modeldevelopmentandvalidationthatwillenableimprovedabilitytoresolveprocesseswithintheSWOTswath.

7.1.1.3.5 Post-launchCal/ValActivitiesPost-launchactivitieswill includeongoingmooringandbuoydatacollection.Resourcingwill likelydictate data will be downloaded from the moorings every six months, with this to be optimizedaroundthetimeoflaunchofSWOT.HighresolutionoceanmodelingwillbeundertakenonaregulartimestepusingadomainspanningBassStrait,nestedinalargermodel.GPSdatawillbedownloadedremotelyenablingregularproductionofzenithwetdelays.Dedicatedcampaignsforprofilingsurfacetopographywillbeundertakenatsetepochs,dependentonavailableresources.AsecondaryactivitywillbecomparisonofSWOTdataagainsttheglobaltidegaugenetworkinordertoassesstheabilityofSWOTtoassessaccuratechangesinregionalmeansealevel.

7.1.2 Validationofrelative2DSSHandcurrentsTheCal/Val sitesdescribed in this sectionaimatvalidating the relative2DSSHand itsderivatives(e.g. geostrophic currents),oftencombiningexisting infrastructures (e.g.HFradars, shipandADCPcurrent data and glider data) and new measurement techniques such as lidar and/or in situinstrumentation.These localvalidationsitesarecomplementarywithglobalmetrics (e.g. statisticalor global in-situ networks). Having at least two or three validation sites makes it possible forlocal/globalcomparisonstoinfertheinfluenceofanyerrorthatdependsonin-orbitandgeophysicsconditions(e.g.geoid,tides,corruptionbycoastallayover).Tothatextent,theoceancalibrationsitesdescribedinthisdocumentprovidearangeofoceandynamics,tides,bathymetry,windandseastateconditions.Inadditiontotheirparticulardynamics,allofthevalidationsiteswill:

● evaluatetheevolutionofthedynamics(feature&frontdetection,2Dspectra)overthefast-repeatmissionphaseusingSWOTandmulti-satelliteanalysis,andavailableairborneandin-situsurfaceobservations

● evaluatetheassociatedverticalstructureoftheSWOTSSH,incomparisonwithavailablein-situdataandHRmodels

● validate 2D SSH reconstruction techniques for gridded fields based on SWOT & availablenadiraltimeterdataduringthefastsamplingphaseandthenominalphase

7.1.2.1 GulfStreamValidationSite(BackupUSProjectSite)

7.1.2.1.1 SiteDescriptionTheGulfStreamCal/ValsiteislocatedoffCapeHatterasinNorthCarolinaofftheUSeastcoast(seeFigure34).Itiscenteredonthecrossoverpointofthe1-dayrepeatorbit.



82

Figure34.Tracksofthe1-dayrepeatphaseofSWOT,withacrossoverpointoffCapeHatterasinNorthCarolinaoftheUSeastcoast.

TheGulfStreamsiteisselectedasthebackupUSprojectsitebecauseofthefollowingconsiderations:

1) The Gulf Stream is the strongest western boundary current. It is typically 100kilometreswideandcanbetraceddowntoatleast1000m.Thecurrentvelocityisfastest near the surface, with the maximum speed over 2 m/s. Beginning in theCaribbeanandendinginthenorthernNorthAtlantic,theGulfStreamSystemplaysanimportantroleinthepolewardtransferofheatandsaltandservestowarmtheEuropeansubcontinent. TheGulfStreamsystemisamong, ifnot, themoststudiedoceanographic feature. In addition to the strong current, the Gulf Stream isassociatedwithhighdegreeofmesoscaleandsubmesoscaleactivities. Asa result,theGulf Stream system and itsmesoscale and submesoscale variability have beenthe focus for numerous observational, modelling and theoretical studies (e.g., therecent LATMIX seasonal submesoscale experiment). In addition to major fieldprograms, theGulf Streamwas also seen as the first oceanographic featureby theearliestaltimetrysatellites(e.g.,Seasat,Geosat).

2) In thedeepoceanregionof theSWOTcrossover, theGulfStreamcurrent isstrongand so this site is notwell suited tomaintaining station-keeping gliders or repeatgliderlines,orstablemooringlines.Assuch,ithasnotbeenproposedastheprimaryUSprojectsite



83

3) AnotherconsiderationfortheGulfStreamsiteasaUSprojectsiteissimplybecauseof its proximity to the coast so as to support both airborne flights anddeployment/recoveryofinsitusensors/platformsincludingshipsandautonomousvehicles. However, the strong currentsmaymake operation of some such assetsinfeasible.

4) AlthoughtheFast-samplingPhasecrossoverpositionisinthehighlyenergeticdeepbasin, which is challenging for maintaining an extensive array of in-situobservations,the1-dayswathextendstowardsthecoastoverthewideshelfoftheMid-Atlantic Bight. This shelf region is alreadywell-sampledwith theMARACOOScoastalobservingsystem,allowinggoodexternalin-situvalidationoftheSWOTSSHandgeostrophiccurrents.ThisinshoresitewillprovideasecondbackupsitefortheUSCalValactivities.

7.1.2.1.2 SiteGoalsStartingwiththeSWOTFast-samplingphase,thegoalsforthissiteare:

● TovalidatetheSWOTSSHobservationsandswathaveragedspectra from15to150kmforthismid-latitudewestern boundary site with highmesoscale and sub-mesoscale dynamicsandmoderatetidesandinthewell-sampledMid-AtlanticBightregionusingwhateverdataareavailablegiventhatthisisnotaprimarysite.

● evaluate the surface wave conditions for this particular site with moderate swell, and theimpactontheseastatebiasestimation

● evaluate the SWOT signal to noise for the particular dynamics at this site (eg, strongsubmesoscalesignalinwinter,higherSWHinwinter,theoppositeinsummerconditions)

7.1.2.1.3 SiteInstrumentationThe Gulf Stream system is routinely monitored by the Mid-Atlantic Regional Association CoastalOcean Observing System (MARACOOS), a regional association of the national Integrated OceanObservingSystem(http://maracoos.org/). MARACOOSmaintainsanumberofobservationalassetsincluding:coastalweathernetwork,primaryandback-upsatellitedataacquisitioncenters,atriple-nestedmultistaticHFRadarnetwork,anacceleratingautonomousunderwaterglidercapability,andmission-specificstatisticalanddynamicaloceanforecastmodels.InthecomingyearsleadingtothelaunchoftheSWOT,MARCOOSwillsurelymaintaintheseexistingobserving systems. More importantly, MARACOOS will introduce new technology as they becomeavailable in the future. MARACOOS is currently focusingon thedevelopmentanddeploymentof afleet of gliders to continuously patrol the coastal oceans. In order to achieve this goal, we areemploying some of the same “smart” technologies thatNASA has used in deploying earth-orbitingsatelliteconstellations.Thistechnologyallowsthegliderstoadjusttheircurrentcoursebasedonthepreviouslycollectedphysicalandopticaldata.Whenrealized,thiswillallowfor24-hour-a-daydatacollectionwithoutconstantsupervisionbyahumanscientist.Theendresultwillbeagliderfleetthatwillbeabletodetectandtrackoceanicfeatures(i.e.:upwellingevents,red-tides,andcoastaleddies)fromtheirformationtodissipation,improvingourcurrentunderstandingofthedynamicalnatureofcoastalecosystemsandprovidingearlierdetectionofoceanicfeaturesthatdevelopoffshoreandareadvectedintocoastalwaters.



84

7.1.2.1.4 Pre-launchSiteCharacterizationThe Gulf Stream system has already been studied extensively. No additional SWOT-dedicatedcharacterizationisneededatthissitebeforelaunch.

7.1.2.1.5 Post-launchCal/ValActivitiesIftheprimaryCal/ValsiteoftheUSwestcoastisfoundnottobeviableforsomereason,someorallof the planned Cal/Val activities for the primary site can bemoved to the backup site. This is acontingencyscenariothatwillbedefinedwhen/ifitbecomesnecessary.

7.1.2.2 CaliforniaCurrent/CoastCal/ValSite(PrimaryUSProjectSite)

7.1.2.2.1 SiteDescriptionTheCaliforniaCurrentCal/Valsite is locatedoff thecoastofcentralCalifornia intheeasternNorthPacific Ocean (see Figure 35). It is centered on the crossover point of the 1-day repeat phase ofSWOT. There is also extensive instrumentation over the portion of the 1-day swath that extendstowardtheCaliforniacoast,whichcanbeconsideredtobepartoftheCal/Valsite.



85

Figure35.Tracksofthe1-dayrepeatphasewithacrossoverpointoffthecentralCaliforniacoast

TheCaliforniaCurrent is anEasternboundary current and is alsopartof theNorthPacificGyre, alargeswirlingcurrentthatoccupiesthenorthernbasinofthePacific.TheCaliforniaCurrentSystemis associated with major upwelling zones to support a very productive ecosystem and thereforefishery.ItispartofaPacificOceancurrentthatmovessouthwardalongthewesterncoastofNorthAmerica, beginning off southern British Columbia and ending off southern Baja California. ThemovementofnorthernwaterssouthwardmakesthecoastalwaterscoolerthanthecoastalareasofcomparablelatitudeontheeastcoastoftheUnitedStates(e.g.,GulfStream).Additionally,extensiveupwellingofcoldersub-surfacewatersoccurs,causedbytheprevailingnorthwesterlywindsactingthrough the Ekman Effect. The winds drive surface water to the right of the wind flow, that isoffshore,whichdrawswaterupfrombelowtoreplaceit.TheupwellingfurthercoolsthealreadycoolCaliforniaCurrent.ThisisthemechanismthatproducesCalifornia'scharacteristiccoastalfog.TheCaliforniaCurrentregionwasthelocationofAirSWOTandMASSoceanflightsaswellasUCTDdatacollectionduringApril2015. Thiscampaignencounteredmanyweather-relatedand logisticalchallenges.Sincethattime,theSWOTfast-repeatorbithasbeenchangedsothatthecrossoverpoint



86

off theUSwest coast has been shifted further to thewest, out of USNavywarning zones, but theadditionaldistancefromshorepresentsfurtherchallengesforshipsandaircraft.ThetypicalcurrentattheCaliforniasiteisbenignenoughforgliderstomaintainstation,however,sotheCaliforniasitehasbeenselectedastheprimaryUSprojectoceanCal/Valsite.

7.1.2.2.2 SiteGoalsThegoal fortheCaliforniasite is tocarryouttheSWOToceancal/valatshortwavelengths. TherewillalsobeopportunitiesformanylocalfieldcampaignsthatcanbecoordinatedinordertolearntheSWOTcalibrationparameters.StartingwiththeSWOTFast-samplingphase,themaingoalsforthissiteare:

● TovalidatetheSWOTSSHobservationsandswathaveragedspectra from15to150kmforthiseasternboundarysitewithmoderatetides,andmoderatemesoscaleandsub-mesoscaledynamics

● evaluatethesurfacewaveconditionsforthisparticularsitewithlargeswelldevelopedovertheentirePacificOcean,andtheimpactontheseastatebiasestimation

● evaluatetheSWOTsignaltonoisefortheparticulardynamicsatthissite

7.1.2.2.3 SiteInstrumentationTheCaliforniaCal/ValsiteisthebaselinelocationforthegliderarraydescribedinSect.6.4.Currently, a number of data sets are being collected in real-time by the Central and NorthernCalifornia Ocean Observing System (CeNCOOS) and Southern California Coastal Ocean ObservingSystem (SCCOOS) fundedbyNOAA IntegratedOceanObservingSystem (IOOS). TheMontereyBayAquarium Research Institute (MBARI) also has a more than 30 years history to measure theMontereyBayanditsnearbycentralCaliforniacoastalocean.Specially, the surface current is measured by a network of high-frequency (HF) radars along theCaliforniacoastathourlyintervalandaspatialresolutionof1km.Amooring(knownasM1)atthecenterof theMontereyBay is collectinga continuous timeseriesof temperatureandsalinityatalldepths. VerticalprofilesoftemperatureandsalinityalongthreeSpraygliderscontinuouslyintime.MBARIfrequentlydeploystheirautonomousunderwatervehicles(AUVs)tomeasureverticalprofilesoftemperatureandsalinityathigherspatialresolution.AdedicatedeffortofdevelopinganestedsystemofoceangeneralcirculationmodelshasbeenputinplacetosupporttheSWOToceansciencevalidationattheCaliforniasite.ThemodelconfigurationisshowninFigure36.Themodelwillassimilatedatafromthein-situoceanobservingsystem(gliderarrayasthebaseline)plusotherroutinelyavailablein-situandsatelliteobservationstoproducenowcastandforecasttosupporttheSWOToceanvalidationactivities.ThiseffortisaimedtoproduceoptimalestimatesofthestateoftheoceanforcomparisontoSWOTobservationsforvalidationandunderstanding.ThedatafromtheMASSlidarsystemwouldalsopossiblybeincorporatedbythesystemforsupportingCalVal.



87

Figure36.ConfigurationofassimilativemodelattheCaliforniaCal/Valsite.

7.1.2.2.4 Pre-launchSiteCharacterizationTheCaliforniaCal/Valcrossoversiteisthelocationforthepre-launchgliderexperimentdescribedinSect.6.4.The pre-launch characterization of the California Current site began in April 2015 with in situmeasurements coincident to AirSWOT and MASS overflights. During the 2015 data collection,problemswerefoundwiththesiteaccessibilityduetoNavyairspacerestrictions,sothesitehasbeenshiftedslightlytothewesttoavoidtheserestrictionsduringthepost-launchcalibrationphase.Duringthe2015campaign,threetypesofinstrumentsweredeployedoveraperiodoftwoweeks.AnunderwayCTDsensorwasdeployedfromashipgoingatabout10kt.TheUCTDmeasuredverticalprofilesoftemperatureandsalinitydowntoabout200metersalongtheshiproute.ThreeEM-APEXfloats were deployed to measure not only vertical profiles of temperature and salinity as theconventionalArgofloatsbutverticalprofilesofvelocityaswell. Multiplesurfacedrifterswerealsodeployed to measure the surface current that was used to validate the HF radar derived surfacecurrentandquantify theeffectofsurfacecurrentonAirSWOTmeasurements. Areal-time3Ddataassimilativeoceangeneral circulationmodelwasalsoused to facilitate theplanningof in situ fieldcampaignsaswellasdatainterpretationandvalidation.

7.1.2.2.5 Post-launchCal/ValActivitiesDuring thepost-launchcalibrationphase,Cal/Val activitiesat theprimaryCaliforniaoceanCal/Valsitewill include theMASS and/or in situ glider campaigns. See Sects. 2.3.2 and6.4 for details ontheseactivities.

Cal/Val&ROMS&Configura1on&&

Online&nes1ng&&Upda%ng((the(boundary(condi%on(every(%me(step(for(the(%dal(and(high8frequency(energy(to(propagate(from(the(large(to(small(domain.(The(%me(step(for(the(1/3(km(resolu%on(is(20s.(

Climatological&WOA13&+&monthly&anomaly&T/S&and&geostrophic&veloci1es&from&gridded&Argo,&and&AVISO&SSH.&&

9Ikm&resolu1on& 3Ikm&resolu1on&

Tidal&forcing&(10&cons1tuents)&&&

1&km&resolu1on&

Star&at&(125.5W,&35.7&N)&

1/3&km&resolu1on&

100&Ver1cal&&Levels&&



88

7.1.2.3 MediterraneanValidationSite(FrenchProjectSite)

7.1.2.3.1 SiteDescriptionThewesternMediterraneansiteischosenasamid-latitudesitewithmoderateeddyenergylevels& weak tides. The Fast Sampling Phase of SWOT will allow validation of the 15-150 km oceanprocessesattwositeshere(Figure37).ComparedtoregionsintheAtlanticorPacificbasinssituatedatthesamelatitudes,theMediterraneanSeadynamicsareenergeticatshortscales(theRossbyradius is about 10-15 km; feature scales of 30 km). These small dynamics are partly captured byalongtrackaltimetrybutnotwellby today’smappeddata.Thesesmalldynamicsaredominatedbygeostrophicmotions (> 90% in the energetic currents). Being an enclosed sea, theMediterraneanbasin has a small horizontal extent but relatively strongwinds, generating short rapid sea stateconditions rather than long regular swell, different from other ocean validation sites in theAtlanticorthePacific.Twoexperimentalsitesareplanned–oneatthe1-dayrepeatcrossoverpointsbetweentheBalearesIslandsandAlgeria,theotheroffshorefromToulonontheFrenchcoast.

Figure37.Left:Tracksofthe1-dayrepeatphaseofSWOT,withacrossoverbetweentheBalearicIslandsandAlgeria.ThereddashedcircleshowstheproposedvalidationregionneartheSWOT1-daycrossover,andthepinkdashedcurvethevalidationregionintheLigura-ProvencalCurrent.Right:Meaneddykineticenergy(cm2/s2)basedon1kmSymphoniemodelresultsintheNWMediterraneanSea.Duetothedominanceofsmall-scalestructuresmostoftheenergyisatshortscalesof20-30km

7.1.2.3.2 SiteGoalsDuringtheSWOTFast-samplingphase,thegoalsareto:

● validatetheSWOTSSHobservationsforthisparticularsitewithweaktides,moderateeddyenergyanddominantsmall-scaledynamics

● evaluate the surfacewave conditions for this particular sitewith low swell, and dominantshortsea-state,anditstheimpactontheSWOTSSHestimation

● evaluatetheSWOTsignaltonoisefortheparticulardynamicsatthissite

7.1.2.3.3 SiteInstrumentationTheaimistodeployaselfcontained,rapidlydeployedandlowcostexperimentalsetupaimedatquantifying the two-dimensionalSSHstructuresatscales froma fewto tensofkm,aswellas theirassociatedphytoplanktoncommunitystructures.



89

Theplannedinstrumentationincludespairsorabatchofthreeglidersthatwillperformaformationflightwithin the SWOT swath. Gliders should be available from the Spanish IMEDEA/SOCIB group(southernsite)andFrenchMOOSEobservations(northernsite).ThegliderswillbefittedwithCTDandasuiteofopticalsensors(fluoresenceandbackscatterinseveralwavelengths).Thegliderswillmakeshallowdives(to100mdepth), remainingata fixeddistanceone fromanother(about1kmapart). Eachof themwillmonitor the surface currentduring the time spentdrifting at the surfacebetween two successive dives. We will have access to the surface velocity gradient tensor atkilometric scale, along the gliders path. This in situ datasetwill allow assessing quantitatively thecapabilityofSWOTtocapturethesurfacecurrent(dominatedbygeostrophy),butalsokeydynamicalquantitiesderivedfromit(vorticity,divergenceandstrainrate).The plan is also tomap the physical and biological parameters including the phytoplanktoncommunitywith an underway instrument and a towed vehicule, operating around the glider fleet.Thesmallship-basedinstrumentationshouldincludeaMovingVesselProfiler200(MVP),abenchflow cytometer CytoSense connected to the ship surface pumping system, and an echosounder.TheMVP is an automatic winch system that is used to deploy a freefall “fish”which contains thevarious instruments (CTD, LOPC and fluorometer). The “fish” is towed behind the moving shipperforming high-frequency free-falling profiles from the surface to a given depth. TheMVP allowsrepeatedsynopticsectionsoffast-evolvingsubmesoscalestructures.Theobservationscanbeusedtoreconstruct the vertical distribution of both physical and biological parameters within the watercolumn.Onlyhalf thecrossover isovertheocean,anda largepartof this iswithintheAlgerianEEZwaterswhich can be difficult to access for airborne or in-situ deployments. The priority is therefore toinstrumentthe1-dayswaththatextendsbacktowardstheBalearicIslands,inaregionofmoderateeddy energy. This site is easily reached by airborne instruments, including the possibility to use aFrenchairborneLidarsystemwithmutli-spectralcameratobettersituatethepositionofthesurfacefronts.Thetwoproposedsiteswillbeusedforadetailedcross-validationofSSHfromSWOTandavailablenadiraltimeters.ThedeploymentofGPSbuoys ispossiblefortheFastsamplingphaseatonesite.HRrealisticmodels(1kmresolution)areavailableinNRTforSSHandwaves.Waveriderbuoys,adaptedtomeasurethepredominantshortwaves,willbetested.HFRadarisavailablenearToulonprovidingsurfacecurrentsinthe150kmcoastalbandforthenorthernsite.Thevicinityoftheproposedexperimentalareastoourlogisticalsites(FrancemainlandandBalearicIslands) makes this region a good benchmark for our project. Aside from the logistics issues, thediplomaticaspects (EEZclearancesandrelated issues)havealsobeenproven tobe tractable,withcurrentlyongoingglidertransectsoffAlgiersaspartoftheSOMBAprojectledbyLOCEAN.

7.1.2.3.4 Pre-launchSiteCharacterizationPre-launchcampaignshavestartedandothersareplannedatbothsitesinthenextyears,inordertopractice the experimental deployment and fine-tuning submesoscale resolving adaptive samplingtechniques.Figure38displays the legsof theOSCAHRcampaign(GulfofLion,October2015,PI:A.Doglioli) where quasi-synoptic, high resolution CTD casts have been recorded by a towed vehiclealongoneJason-2andtwoSARAL/Altikatrackswiththeaimofcomparingaltimetry-derivedandinsitu estimations of sea level anomalies, together with the deployment of surface drifters, glideroperations,andHFradarobservations.HRrealisticmodels (1kmresolution)areavailable forSSHandwaveanalysispre-launch.Waveriderbuoys,adaptedtomeasurethepredominantshortwaves,



90

willalsobetestedprelaunch.HFRadarisavailablenearToulonprovidingsurfacecurrentsinthe50kmcoastalbandforthenorthernsite.JointanalysisofcollocatedSARnadiraltimetry,gliders,andHFradarareunderwayandwillcontinueatbothsitestocharacterizetheobserveddynamics.

Figure38.LegsoftheOSCAHRcampaign(blackcrosses)andtracksofJason-2andSARAL/Altika(resp.redandmagentalines).The background color represents GHRSST L4 Sea Surface Temperature. The campaign (PI: A. Doglioli) took place inOctober/November2015.Amongotherobjectives,thecampaignaimedattestingSLAreconstructionfromquasi-synopticCTDcastsmadewithatowedvehicleandsupportedbycurrentandhydrographicobservationsfromsurfacedrifters,HFradar,andoneglider.

AHRgeoidisavailablefromtheFrenchMarineservice(SHOM)intheNWMediterranean.AHRmeanseasurface (MSS)canbedeveloped ineachregion,using theHRgeoid,newsatellitealtimetryandgravitymissiondata,andin-situobservationsincludingtowedGPSobservations.

7.1.2.3.5 Post-launchCal/ValActivitiesPost-launch validation will concentrate on the self contained, rapidly deployed and low costexperimental setup at both the northern and southern sites during the fast sampling phase, asdescribedinsection8.1.2.2.3and8.1.2.2.4.

7.1.2.4 LoyaltySiteThis sitehasbeenproposedasanopportunity site to theongoingROSES/TOSCA incase the1-dayorbit isconfirmedtopassovertheLoyaltyBasin. It isassociatedtotheproposalentitled“SWOTinthe Tropics, A Case Study In South West Pacific” lead by L. Gourdeau. In the case that the fastsamplingorbitfinallyselectedbytheSTdoesnotpassanymoreintheLoyaltybasin,thissitewouldnotbeincludedanymoreinthelistofpotentialCal/Valsites.



91

7.1.2.4.1 SiteDescriptionTheLoyaltybasin issituated inSouthWestPacific. It isenclosedbytheLoyaltyarchipelagoontheEastandtheGrandterreisland,allformingtheFrenchTerritoryofNewCaledonia(Figure39).Thisbasinhasbeenstudiedbyoceanographersfortensofyearsandwithin-situdatacollectedduringadozenofcruisesandconsequently,thedynamicsofthecurrentsisparticularlywellknown.ItisasiteselectedintheAltiKaScientificTeam.Sincethe90’s,ithasbeenfullymappedwithmulti-beamecho-sounders, itsgeologicalsettinghasbeenthesubjectofseveralpublications,allmakingpossible thecomputationofahighresolutiongeoidsurface.

Figure39.Theyellowlinesstandsfortheswathlimitsandtheredlinestandsforthenadirtrack.Theyellowpinsgivelocationoftheexistingtidegageswhenthewhitepinsshowthelocationoftheplannedones.

7.1.2.4.2 SiteGoalsThissitewillbededicatedtocollectingseasurfaceheightanddynamicheightmeasurementsinthealong-track and cross-track directions, to provide surfacewave data andmean sea surface heightundulations.Thissiteat20°SisalsoalowerlatitudeCalValsite,impactedbytropicalandsubtropicaldynamics.

7.1.2.4.3 SiteInstrumentationThearchipelagoisequippedwithadenseGPSnetwork,computeddaily,andanetworkoftidegauges,geodetically linked to the GPS network. These in-situ permanent instruments give access to theinstantaneousabsolutesea-levelatafewplaceswithintheSWOTswath.Thearchipelagoisequippedwitharesearchvessel,N/OAlis,dedicatedtoworksatseainthevicinityoftheislands.Together,theTerritoryownsavesselthatcanbeusedforscientificworknotexceedingacoupleofdays.The archipelago is equippedwith a network ofmeteo stations (at least one on each island, at theairport)Awave radar is planned to be installed on one of the islands for studies of thewind induced seasurfaceroughness



92

7.1.2.4.4 Pre-launchSiteCharacterizationTheoperationsthatwillbeconductedbeforethelaunchare:

- Officialsollicitationsofbothvesselsforthefastsamplingphase- Sollicitationsofthevesselsfortestcruisesbeforethedateoflaunch,inparticulartotestthe

useofCalNaGeo,theGPSfloatingsheet(Figure13).ThetestcruisescouldbetheopportunitytocollectprofilesofsealevelanomaliesalongthenadirtrackofSWOT.

- Computation of a high resolution geoid model by combination of surface altimetry data,shipbornegravitydataandgeopotentialinferencefromthesubmarinegeologicalstructure

- Installadditionalsensors(tidegauge,cornerreflectors,submarinepressuregaugesonreefs,etc…)accordingtoSTrecommendationsandfundings.

7.1.2.5 Post-launchCal/ValActivitiesThepostlaunchactivitieshavetobeconsideredforthe1-day-thefastsamplingorbit. During the fast sampling orbit, the vessels will be used to collect profiles of the absoluteinstantaneousseasurfaceheight.ThecruiseplanwillbeelaboratedfollowingtherecommendationsoftheST(longalong-trackprofilesvscross-trackfromoneouterrimoftheswathtotheother…).

7.2 HydrologyCal/Valsites

7.2.1 RiverCal/ValSites

7.2.1.1 WillametteRiverCal/ValSite(USProjectSite)

7.2.1.1.1 SiteDescriptionTheWillametteRiverCal/ValsiteislocatedinwesternOregon,U.S.A.,betweenthetownsofCorvallisandEugene(Figure40).Theriverinthis75-kmstudyreachisprimarilygravel-beddedwithasingle-thread channel with occasional sections of multiple-thread channel, and relatively stable bed andbanks. Though there are moderate-size flow regulation reservoirs in the upper watershed, thehydrographoftheriverinthestudyreachistypicaloftemperaterain-andsnow-fallriversystems,withdistinctandlargeregularpeaksandrecessionlimbsfromOctobertoJuneandalowflowperiodinthesummerfromJulytoSeptember.ThisstudyreachhasbeenthefocusofpreviousSWOT-relatedstudies,particularlytheAirSWOTcampaigninspring2015.TheWillametteRiverCal/ValstudysitewillbeusedvalidateSWOT’sabilitytodetectwater-surfaceelevation,slope, inundationextent,anddischarge.



93

Figure40.TheWillametteRiverFieldSite

7.2.1.1.2 SiteGoalsTherearetwoprimaryvalidationgoalfortheWillametteRiverCal/Valsite:

1. ValidateSWOT’sabilitytomeasureandcharacterizeriverswithsingle-andmultiple-threadchannels,typicaloftemperateclimaterivers.

2. CharacterizeSWOT’sabilitytomeasurewater-surfaceelevation,slope,inundationextent,anddischargeinriversnearthe100mbaselinewidthrequirement.

7.2.1.1.3 SiteInstrumentationTherearethreeactivedischargegagesonthemainstemWillametteRiverinthestudysite,andonegageon eachof the threemajor tributaries,making the study site relativelywell-characterized fordischarge.Theoverbanktopographyofthesitewasmeasuredwithaeriallidarin2008-2009,thoughin some locations the riverhas changed substantially since that time.Thebathymetryof the studyreachhashistoricalcoarsebathymetryfromwidely-spaced(2+kmspacing)crosssectionscollectedin2002.Morerecently,newcrosssectionswithapproximately2-kmspacingandseverallongprofileshavebeenmeasuredaspartoftheAirSWOTcampaigninspring2015.Besidesthegagesalreadyinoperationontheriver,therearenootherpermanentmeasurementsbeingmade.



94

One of the primary needs for theWillamette River Cal/Val site is a calibrated 1D or 2D hydraulicmodeltobeabletotestSWOTdataproductsinboththepre-andpost-launchperiods.The1DmodelmaybedevelopedinconjunctionwiththeUSGSfollowingtheAirSWOT2015campaign,however,itisnota fundedprojectatpresent.The temporarydeploymentofpressure transducersalong the longprofile as well as boat-conducted surveys of water-surface elevation, such as utilized during theAirSWOTcampaignin2015,aresuggestedforinstrumentationinthepost-launchperiod.

7.2.1.1.4 Pre-launchSiteCharacterizationThepre-launchcharacterizationoftheWillametteRiversitebeganinspring2015withbathymetric,water-surface elevation and discharge measurements coincident to AirSWOT overflights. FieldmeasurementswerecollectedbyUSGSOregonWaterScienceCenterandUniversityofOregon(MarkFonstad). Approximately six water-surface elevation long profiles and 20 cross sections weremeasured over a range of flows, though no data were collected during substantial overbankdischarges.Dischargewasmeasuredatapproximately15locationsalongthelongprofileduringtheflightsandoverarangeof flows,aswellasseveralsmaller tributaries. Inadditionto the fielddatacollection, the Willamette was flown six times by AirSWOT, including KaSPAR and near-infraredimagery, and thesedatawill provide extensive additional information to help characterize the siteand potential SWOT performance for water-surface elevations, slope, inundation extent, anddischarge.ThedataprocessingfortheAirSWOTdataisongoingandisexpectedtobecompletedbyfall2015. Additional suggestedpre-launchcharacterizationactivities for theWillametteRiverCal/Valsitewillbethedevelopmentofa1Dhydraulicmodelandpotentially2Dhydraulicmodelsinshortersubreaches.Inaddition,the1-dayfastrepeatcyclemayincludethelowerWillametteRiver(CorvallistoPortland)forwhichasparsedatacollectioneffortmightbewarrantedtocharacterizethispartoftheriver.

7.2.1.1.5 Post-launchCal/ValActivitiesTheWillametteRiverCal/Valsiteisoneoftheonlyriversitesthatwillbeincludedinthe1-dayfastrepeatcycleimmediatelyfollowinglaunch.Assuch,duringthe1-dayfastrepeatcycletheWillametteRiverCal/ValsitewillbethefocusofintensivemeasurementstohelpvalidateSWOTmeasurementsof water-surface elevation, slope, inundation extent and discharge. Field measurements will besimilar to the techniques utilized in the spring 2015 AirSWOT campaign, including boat-basedmeasurements ofwater-surface elevation, slope, and discharge, aswell as deployment of pressuretransducers tomeasure temporal changes inwater-surfaceelevationandslope. Inaddition to fieldmeasurements, theWillametteRiverCal/Valsite isa likelycandidate forcoincidentunderflightsofAirSWOT, with measurements by KaSPAR and near-infrared camera to constrain water-surfaceelevation, slope, and inundation extent.Once developed, the 1D and/or 2Dmodelswill be used todiagnoseand trouble-shootany issueswith theSWOTestimates forwater-surfaceelevation, slope,inundationextentanddischarge.

7.2.1.2 GaronneRiverValidationSite(FrenchProjectSite)

7.2.1.2.1 SiteDescriptionTheGaronneRiver(SouthWestofFrance)isthe4th longestriverinthecountrywitha55930km2drainagearea.TwospecificGaronnereachesareproposedasvalidationsites (Figure41).Foreachreach,therearethreeoperationalwaterlevelgages(withratingcurve),multiplebathymetrycross-sectionsand1D/2Dhydraulicmodelsalreadyimplemented.Thetworeacheswillbefullycoveredbythe1dayorbit(Figure42).TheupstreamreachbetweenBlagnacandMalauseis80kmlongwithameanriverwidth~150m(lowerduringthelowflowseasonwithsandbars)andriverbathymetry



95

slope around0.9m/1000m.The downstream reach betweenTonneins and LaRéole is 50 km longwithameanriverwidth~180m(alsowithsandbarsduringthelowflowperiod)andariverslopearound0.3m/1000m.

Figure41.GaronneRiverwatershedandvalidationsites(inlightblue)

Figure42.Cal/ValGaronnereachesandSWOT1dayorbitswathscoverage(delimitedbyyellowlines,redlinecorrespondstothesatellitenadir)



96

Onthesetworeaches,theriverchanneliswelldefined(theriverisnotbraided),nomaintributaryjoins theriverwithin thereaches(theTarnriver join theGaronne fewkilometersdownstreamtheBlagnac-Malause reachand theLot river a fewkilometersupstream theTonneins-LaRéole reach).There isriparianforestontheriverbanks,butthesurroundingfloodplain ismainlycropscovered.Alongtheriverbanks,manyleveeswerebuilttoavoidflooding.TheGaronneRivervalleyislessthanadozenofkilometerswide(varyingbetweenreaches),steep-sidedbyhills(uptoabout100mhigherthanthevalley)thatgenerateerrorsduetolayovereffects.

7.2.1.2.2 SiteGoalsThese sites will be used to validate: water surface elevation locally and at multiple points, watersurfaceslopealongtheriver,dischargealgorithmsefficiency.

7.2.1.2.3 SiteInstrumentationOntheupstreamreach(Figure38),thereisoneoperationalgagelocatedinthemiddleofthereach(at Verdun-sur-Garonne), another one situated just upstream of the reach (at Portet-sur-Garonne)andalastoneaboutonedozensofkilometersdownstreamofthereach(atLamagistère,butbeyondtheconfluencewiththeTarnRiver).Water levelsaremeasuredevery15minutesandtheellipsoidheight of the gage is available for Verdun-sur-Garonne and Lamagistère stations (not available atPortet-sur-Garonne,but itcanbedoneeasily).Dischargesfromratingcurvearealsoavailablefromoperationalagencies.Similarly,onthedownstreamreachthreegagemeasurementsareavailable(atTonneins,MarmandeandLaRéole,seeFigure43),withthesamemeasurementcharacteristicsasforupstreamgages.These six gages aremaintained by operational governmental agencies in order to alert in case offloodingorextremelylowflows.ThereisthereforeaveryhighprobabilitythatthesegageswillstillbeoperationalwhenSWOTwillbeflyingandthedatawillbeavailableforCal/Valactivities.

Figure 43. Longitudinal profile along the river axis for the upstream reach showing available river bed elevation (frombathymetric cross-sections).Reddots showcurrent operational gages.Greenarrows showwater surface slopebreak,whereadditionalGPSmeasurementscouldbedone.Credit:IMFT



97

Figure44.Longitudinalprofilealongtheriveraxisforthedownstreamreachshowingavailableriverbedelevationfromcross-sections,water surface elevation for 01March 2010 fromMascaretmodel. Red dots show current operational gages. Greenarrowsshowwatersurfaceslopebreak,whereadditionalGPSmeasurementscouldbedone.Credit:LNHE/CERFACS

Ontheupstreamreach,203bathymetrycross-sectionsareavailableandhavebeenusedtosetupa1D and a 2D hydrodynamic modelings of the reach. Some of these cross-sections are quite old(~1995) and they should be updated.Moreover, theywere not originally georeferenced, so itwasnecessarytodothismoreorlesspreciselybyfittingwithlidardata.AlidarDEMofthefloodplainwasprovidedbyIGN(butunfortunatelynotscannedduringlowwaterlevel)andhasbeenincludedinthe2DmodelingbyIMFT.On thedownstreamreach,83bathymetrycross-sectionsareavailable,butherealso theyarequiteoldandneedtobeupdated. 1Dand2Dhydrodynamicmodelingof thereach isalsoavailable.ThefloodplainlidarDEMisalsoavailableonthisreachbutneedstobeincludedinthe2Dhydrodynamicmodeling.

7.2.1.2.4 Pre-launchSiteCharacterizationA few years up to severalmonths before launch, older existing cross-sections for the two reachesshould be updated and georeferenced, and new ones should be made using ADCP (providingbathymetry and also local discharge, the value of which could be compared to rating curvesestimates).TheellipsoidheightatPortet-sur-Garonnegagewillbemeasured.Inaddition,SWOTdatasimulationwill be computedwithin ST activities and could be used to characterize zones that areexpectedtohavehighesterrors(layover,watermaskdetectionerror,etc.).Thesezonescouldthenbesubjecttomoreintensivemonitoringduringthepost-launchCal/Valphase.ApreciseDEMforthesesimulationcouldbebroughttoIGNifneedbe.

7.2.1.2.5 Post-launchCal/ValActivitiesOnFigure43andFigure44,greenarrowsshowlocationswithbathymetryandwatersurfaceslopebreaks.TheselocationsaregoodcandidatesforadditionalGPSmeasurementofwaterlevelforsomeshortperiodoftimeduringthe1-dayorbitphase.Water-surfaceslopemeasurementsalongtherivercouldalsobedonewithasimilarsystemtotheoneshownonFigure21(ifavailable).



98

Currently available resources from the involved teams (CERFACS, IMFT, IMT, INSA Strasbourg,LEGOS, LNHE,Meteo-France): one boat and oneADCP could be used by IMFT (providing they areavailable during theCal/Val phase), a fewGPS at LEGOS and somewind speed gages fromMeteo-France(atsomecost).Amongallthelaboratories,thereareseveralwelltrainedpeopleforeachkindof equipment (ADCP, GPS) and additional people (a dozen)with no field experience that could beavailable.One operational agency (DREAL/SPC Garonne) could update their rating curves for our Cal/Valactivitiesandcouldinstallfewephemeralgages(~3,000euros/gage).

7.2.1.3 LowerMississippiRiver(U.S.ProjectSite)

7.2.1.3.1 SiteDescriptionThelowerMississippiisaverywide(~1km)andlow-slope(~7cm/km)river.Itwasselectedasatier1validationsitebecauseitiseasilyaccessibleandrelativelywell-instrumented,alongwithbeingthelargestriverbydischargeinNorthAmerica.ThestudyreachselectedisbetweenVicksburgandNatchez,MS,alengthof~115km(Figure45).Theriverisprimarilycomposedof1-2channels,withonlyafewmeanderbendsandnoactivecontrolstructureswithinthereach.Theriverexhibitsandannualcycleinflow,withmaximumflowusuallyoccurringinspringorearlysummerandminimumflow in the fall and winter (Figure 46). The site is sufficiently far downstream that it does notexperienceiceformationinthewinter,makingitagoodtargetforyear-roundvalidationactivities.

Figure45.LocationoftheLowerMississippiFieldSite

7.2.1.3.2 SiteGoalsThe primary goal of this field site is to validate SWOT slope, height, and inundation extentmeasurementsonalarge,alluvial,low-sloperiver.TheMississippiistheonlyriverintheContinentalU.S.wherethisvalidationisfeasible,giventhesizeoftheriverandlackofcontrolstructures.



99

7.2.1.3.3 SiteInstrumentationTherearecurrentlygaugesateitherendofthestudyreach(VicksburgandNatchez)operatedbytheArmyCorpsofEngineerstomeasureriverstage.Dischargemeasurementsarenotavailableatthesesitesbutareavailable~200kmdownstreamatBatonRouge,LA.

Figure46.StagehydrographoftheMississippiatVicksburgfrom2012-2015.

7.2.1.3.4 Pre-launchSiteCharacterizationAsanumberofotherriversiteswillbecharacterizedprelaunch,itisnotenvisionedthatasignificantprelaunchcampaignwillbeconductedforthissite.

7.2.1.3.5 Post-launchCal/ValActivitiesTheLowerMississippi isnotcoveredbythe1-dayfastsamplingorbit. Assuch,primaryvalidationactivitieswill takeplaceafter itsconclusion. Themost importantSWOTvariabletobevalidatedatthissitewillbeslope,sincemeasurementerrorsforinundatedareaandheightfromSWOTarelikelyto be small given the size of the river. Wewill conduct four long-profile surveys ofwater surfaceelevationfromamotorizedboatasdescribedinsection6.4.8.Thesesurveyswillbeconductedfromamotorizedboat. Inaddition,twoAirSWOToverflightswillbeconductedwithlong-profilesurveysinorder toassess thesimultaneousaccuracyofSWOT-derivedheight, inundationextent,andslopeover the entire reach length. The USGS has a field office in the region and has the capabilities toconduct the kind of survey described here. It may be prudent to partner with them in order tocompletethesesurveysinthemostefficientway.

7.2.1.4 ConnecticutRiver(U.S.ProjectSite)

7.2.1.4.1 SiteDescriptionThe Connecticut River Cal/Val site extends from the USGS gaging station at Thompsonville,Connecticut,USA,upstreamtotheUSGSgagingstationatMontagueCity,Massachusetts,USA(Figure47).Theriverinthis80-kmstudyreachisprimarilygravel-beddedwithasingle-threadchannelandrelatively stable bed and banks, and one run-of-river dam located near Holyoake, Massachusetts.



100

There is extensive existing gage infrastructure, including three real-timemainstemand fourmajortributariesgagingsitesoperatedbytheUSGS.TheConnecticutRiverCal/ValsiteiswithintheSWOT1-dayFastRepeatOrbit,andisdirectlyupstreamoftheConnecticutRiverTidalCal/Valsite.Thoughthere are small flow regulation reservoirs in the upper watershed, the hydrograph of the river istypical of temperate rain- and snow-fall river systems, with distinct and large regular peaks andrecessionlimbsfromOctobertoJuneandalowflowperiodinthesummerfromJulytoSeptember.Inaddition, this reach of the Connecticut River freezes during thewinter,with breakup occurring inMarchorApril.

Figure47.MapoftheConnecticutRiverCal/Valsite,ConnecticutandMassachusetts,USA.ThetwoyellowlinesshowtheextentoftheKaRINfootprintduringthe1-dayFastRepeatOrbit,theredareaisthestudyreach,andthewhite-encircledsymbolsareUSGSgages.

7.2.1.4.2 SiteGoalsThe Connecticut River Tidal Cal/Val study sitewill be used validate SWOT’s ability tomeasure orcharacterize water-surface elevation, slope, inundation extent, and discharge as well as validateSWOT’slayover-,ice-andrain-flags.

7.2.1.4.3 SiteInstrumentationTherearethreereal-timemainstemdischargegagesonthemainstemConnecticutRiverinthestudyreach,andfourdischargegagesonthemajortributaries,alloperatedbyUSGS.Allofthesegagesrely



101

on rating curves for computing discharges, though discharge ismeasured at each site from ten totwenty times per year. During periods of ice accumulation, the records at the gaging stations areflaggedintheUSGSdatarecord(seeFigure48).Severalaeriallidardatacollectionshavebeenflownoverthestudyreach,includingtwoFEMAflightsin2004,apost-HurricaneSandyUSGSflightin2014,andaUSGS3DEPflight in2015.Thehigh-resolutionDEMproductsfromthe2014and2015flightsareexpectedtobecompletedinearly2016.

Figure 48. Gage data from the USGS gage on the Connecticut River at Holyoke, MA. The top panel shows river discharge,includingiceflags,andstageforthecorrespondingperiodisshowninthebottompanel.

7.2.1.4.4 Pre-launchSiteCharacterizationThepre-launchsitecharacterizationoftheConnecticutRiverCal/ValsiteincludesashortAirSWOTcampaign consisting of two-days with multiple AirSWOT passes during various river stagesscheduledfor2018.Concurrently,agroundcampaignwillhaveinstalledandlevelledapproximately30-50 pressure transducers, aswell as collecting day-of-flight longitudinalwater-surface elevationand discharge measurements using coupled GNSS and ADCP instruments. If not performedpreviously,GNSS-levelingoftheexistingUSGSgageswillberequiredpre-launch.SWOTflaggingwillbeevaluatedusingexistinghigh-resolutionlidarDEMsforlayoverflags,satellite-observationsoficeandsnowforiceflags,andlocalradarandweatherstationsforrainflags.Inthemonthsimmediatelypreceeding launch,approximately30-50pressure transducerswillbe installed for theone-day fastrepeatorbit.

7.2.1.4.5 Post-launchCal/ValActivitiesThe post-launch Cal/Val activities for the Connecticut River site include monitoring the installedpressure transducersduring the one-day fast repeat orbit, aswell as twoone-week campaigns forboat-basedlongitudinalwater-surfaceelevationanddischargemeasurements.

7.2.1.5 TananaRiverValidationSite(USProjectSite)

7.2.1.5.1 SiteDescriptionIn Alaska, SWOT validation efforts will focus on two regions that will have received prior SWOT-related study: the Tanana River and the Yukon Flats (See section 8.2.2.X). The Tanana is a large



102

braided river draining the northernportions of theAlaskaRange (Figure 49). Because it receivessubstantial inflow from glaciers it also transports large sediment loads both in suspension and asbedload.Assuch,thechannelplanformchangesrapidlyintimeandspace.TheTananais>1000kmin total length, but validation efforts will focus principally on a ~150 km reach in the vicinity ofFairbanks. This reach beginswith a highly braided planformbut transitions into a single-channelplanform due to the influence of topography to the north of the river. As such,within this singlereachitispossibletovalidateSWOT’sabilitytodetectwatersurfaceelevation,slope,andinundationextentundermanydifferentconditionscharacteristicofnorthernrivers.Thesitewascharacterizedin detail during a summer 2015 field campaign, includingmultipleAirSWOToverflights, field datacollection of temporally continuous water surface elevation measurements at 23 locations,measurementofwatersurfaceelevationprofilesdowntheentireregionshowninredinFigure49,installationofcornerreflectors,andmeasurementofdatarelatedtoKa-bandradarphenomenology.In addition, a 2-D hydrodynamic model based on LisFLOOD-FP has been developed usinginterpolated data from an in situ bathymetric survey conducted in 2013 by SWOT cal/val teammembers.

Figure49.TheTananaRiverValidationSite



103

Figure50.TheTananaRiverValidationSite,ascharacterizedduringSummer2015

7.2.1.5.2 SiteGoalsTherearethreeprimaryvalidationgoalsfortheAlaskasites:

1. ValidateSWOT’sabilitytomeasureorcharacterizeriverswithcomplex,braidedplanforms,such as the Tanana River. Characteristics to be studied include inundation extent, watersurfaceelevation,andslope.

3. TestSWOTriverdischargemeasurementsinmultithreadedandbraidedriverenvironments.4. UnderstandSWOTcapabilitiestoaccuratelyflagriverice.

7.2.1.5.3 SiteInstrumentationOntheTananaRiver,therearecurrentlytwooperationaldischargegauges,oneatFairbanksandoneat Nenana, approximately 90 river km downstream. These gauges will allow for validation ofdischarge algorithms on the Tanana. Beyond these two gauges, however, there is no permanentinstrumentationon the study reach. However, an airborne IfSARDEMat5m spatial resolution isavailableovertheentirestudyreach.Inaddition,datacollectedduringthe2015fieldcampaigncanbeusedtocharacterizetheoverallpatternsofslope,height,andwidththatSWOTwilllikelyobserveontheTanana,absentmajorplanformchangesbetweennowandtheCal/Valperiod.

7.2.1.5.4 Pre-launchSiteCharacterizationFieldworktocharacterizetheTananaRiverstudyareabegan in2013withabathymetricsurveyofthe~90kmbetweenFairbanksandNenana.Thisallowedconstructionofa2-Dhydrodynamicmodelfor thereach,whichcanhelpwithvalidation. Substantialadditionaldatawascollected insummer2015, including water level variations at 23 locations via pressure transducer (Figure 50) and 8AirSWOToverflightsoftheentirestudyarea,includingbothKaSPARdataanddigitalinfraredcameraimagery. Vegetationconditionsandsandbarcharacteristics, includinggrainsizeandsoilmoisture,



104

werealsobecollected,aswere longprofilesofwatersurfaceelevationmeasuredusingaprecisionGPSmountedonaSontekHydroboard.

7.2.1.5.5 Post-launchCal/ValActivitiesCal/Val activities after launch will focus on the first summer post-launch. Because the currentlyplanned 1-day repeat orbit does not intersect the Tanana Tier 1 Site, no effort will be made toconductvalidationactivitiesduringthis3monthsamplingperiod. FortheTananaRiver,validationactivities will focus on the capabilities of SWOT to measure inundation extent, water surfaceelevation and slope and to estimate discharge. Aerial near-infrared imagery will be collectedcoincidentwith SWOT overflights in order to provide validation of inundation extent. A series ofpressure transducer water level loggers will be installed along the Tanana in order to provideestimates of variations inwater surface elevation and slope. Discharge canprimarily be validatedusingthetwostreamgaugesatFairbanksandNenana,butadditionalmeasurementswillbecollectedbyADCPasnecessary.In addition to these basic measurements, two additional sets of measurements would providedesirablevalidationcapabilities.AirSWOTmeasurementswouldprovidefull,2-DvalidationofSWOTwatersurfaceelevationsinboththeTananaRiverandYukonFlatslakes.AirborneL-bandSAR(e.g.UAVSAR)wouldproviderobustmeasurementsofwatersurfaceextentundervegetationintheYukonFlats,whichwouldofferasubstantialimprovementoverground-basedsurveys.Finally,becauseofitsproximitytoFairbanks,theTananawillbeusedtovalidatetheSWOTiceflag.Airborne visible and near infrared imagerywill be collected simultaneouswith a SWOT overflightduring ice breakup,whichwill providedirect validationof SWOT’s ability to differentiate ice fromopenwater.

7.2.1.6 CanadianValidationSites(CanadianCal/ValSites)

7.2.1.6.1 SiteDescriptionIntandemwiththeUSGSandSTmembers,EnvironmentCanadawillproposethefollowingsites1.ThePeace-AthabascaDelta(PAD)andLakeAthabasca2.TheYukonPorcupineRiverfromWhitehorse/YukontoStevensVillage/Alaska3.TheSt.LawrenceRivernearTroisRivieres4.TheNorthSaskatchewanRiveratPrinceALbert5.TheMackenzieRiveratInuvik6.TheSlaveRiveratGreatSlaveLakeECwill lead these effortswith USGS support alongwith other international partners including STmembers. Of the above sites, 3will be Tier I sites: the PAD, the SlaveRiver and the St. Lawrence,describedbrieflybelow.TherestofthesiteswillfocusonvalidationofselectSWOTobservations,asdescribed in the following section. These sites cover a broad range of physiographic and climaticregimes, from permafrost-free temperate (St. Lawrence) to continuous permafrost Arctic(Mackenzie)anddiscontinuouspermafrostsubArctic(SlaveRiver).



105

Figure51.MapofthePAD.

The PAD is located in northeast Alberta. It is the largest freshwater inland river delta in NorthAmerica.Thedeltaencompassesaround5800squarekilometersandisformedwherethePeaceandAthabasca rivers converge at thewest-end of LakeAthabasca and drain north via the SlaveRiver.Twomain Northern Cal/Val sites are located in this complex region. The PADwas designated awetlandofinternationalimportanceundertheRamsarConventionin1982.About80%oftheareaisprotected within the Wood Buffalo National Park (established 1922), which was designated aUNESCO World Heritage site in 1983. Given the international importance, the PAD has been thesubject of many studies by Canadian agencies - mostly Parks Canada, Environment and ClimateChangeCanada(EC)andNaturalResourcesofCanada(NRCAN)–anduniversities. ThePADdrainsnorthwardintheSlaveRiverwhichhasastableflowratingcurveandalonghistoryofflowrecords.



106

Therivergaugeisupstreamofanaturalcontrolandhasfewtributaryinputattheproposedcal/valsection.

Figure52.ThePeaceAthabascaDelta(PAD)delimitedbyLakeClairetotheWest, thewesternendofLakeAthabascatotheEast,thePeaceandSlaveriverstotheNorthandtheAthabascaRivertotheSouth.

TheSt.LawrenceRiver is thethird largestriver inNorthAmerica,withacatchmentareaof~1.6×106km2,andanaveragefreshwaterdischargeof12200m3s-1atQuebecCity.Itisthedownstream



107

waterbodyof theGreatLakesandOttawaRiver systemsand it is subject to largevariabilityat allscales. While decadal time scale variability is caused by climate dynamics, management of itsupstreamsourcescansignificantlyimpactwaterlevelsattheseasonaltimescale,inparticularintheMontrealarchipelago.Between Montreal and Trois-Rivières, more precisely downstream of Sorel, the river widenssignificantly.Thearea,knownasLacSaint-Pierre,isaUNESCObiospherereserve.Thisecosystemishost to a vast number ofmigratory birds, and sustains an important recreational and commercialfishingindustry.DownstreamofTrois-Rivières,significanttidalinfluencesareobserved,withwaterleveldifferencesbetweenlowandhightideofover5matQuebecCity.Strongcurrentreversalsareobserved,combinedtorapidchanges inwettedareasovershallowtopography,asaresultof thesevariations.

Figure53.TopographyoftheSt.LawrenceRiver,includingtheMontrealarchipelago,LacSaint-PierreandthefluvialestuarydowntoÎle-aux-Coudres.



108

7.2.1.6.2 SiteGoalsAt all sites, gaugedatawill beused to characterizewater elevation, local slopes and inmanycasesSWOT-deriveddischarge.Each sitewill alsovalidateSWOTheight, slope, and inundationarea.TheTier Isiteswillvalidate thesequantitiesovergreaterareasandwithgreaterprecision,and includemeteorologicalstationsanddeploymentstothefieldconcurrentwithaSWOToverpass.Collectively,the sites will also validate SWOT’s performance vis-a-vis permafrost controls, which introducecomplexbankmorphologyandpotentiallyverywetsoilwhenthawed.Anothergoalacross thesesites is todevelopandcontribute toastandardmethodology formakingcal/valmeasurements,andECandtheSTaresettobeginmakingpre-launchmeasurementsasearlyas2016 to test theprotocolsoutlined in thisdocument. Wherepossible these sitesare co-locatedwhereECalreadyhasexistinghydrodynamicmodelsestablishedorinoperations.

7.2.1.6.3 SiteInstrumentationECmaintainsavastnetworkofrivergauges,andgaugesarelocatedeitherwithin,adjacentupstreamordownstream,oradjacentbothupstreamanddownstreamtoallproposedsites.ECalsomaintainsseveralhydrodynamicmodelsofvariousrivers.Specifically, for thePAD,TheWaterSurveyofCanadaestablishedhydrometricstations in theearly1970sandseasonalpressure transducers for researchpurposesenhance the long-termmonitoringnetwork.Historicalandpresentwaterextentsbasedonremotesensingtechnologyarealsoavailable.Elevation benchmarks for the hydrometric have been converted to the new geoid-based verticaldatumCGVD2013withhighprecisionGNSSsurveys.Usinganyglobalgeoidmodelisthusalreadyapossibilityonthissite.LiDARdataoverthePADisavailableonaportionofthePAD;surveyswereflownin2000,2012and2013, with the rest of the area covered by lower resolution Space Shuttle Topography Mission(SRTM)data.Vegetationcharacterizationhasalsobeendoneoveraportionofthesiteviamorethan35vegetationtransectsmaintainedbyWoodBuffaloNationalParksincethe1990s.



109

Figure54.Availableseasonalpressuretransducers(whiteandyellow)andpermanenthydrometricstation(red)overthePADregion.



110

Figure55.LiDARdatapotentiallyavailableoverthePADregion.Surveysweredoneinearly2000(red),2012(green)and2013(blue).

Thereisagreatamountofsatellite imageryavailableoverthePADfrom2008uptothisday.LidardataisalsoavailableoverpartsofthePADandtheSt-Lawrenceseaway.Radarsat2imageryisalmostguaranteed tobeavailableover thePADonaperiodicbasis,givingaccess tohi-resolution imagerydataforwaterextentmappingregardlessoftheconditions(unlikeopticalimagery).Thereisalsothepossibilityofrunninga2DhydrodynamicmodeloverthewholePADregioninthecomingyearsthatmight also be available before the launch. Wind forecast models - which are being verified forinclusion in the 2D hydrodynamic model over Lake Champlain in non-stationary mode - canpotentiallyalsobeverifiedoverthePADtoassesstheeffectofwindoverthesites.LakeAthabascaheightswill be available in CGVD2013 (Geoid CGG2013) .Wind forecastmodels -which are being



111

verifiedforinclusioninthe2DhydrodynamicmodeloverLakeChamplaininnon-stationarymode-canpotentiallyalsobeverifiedoverthechosenlakestoassesstheeffectofwindoverthesites.LakeAthabascaheightswillbeavailableinCGVD2013(GeoidCGG2013)InthecaseoftheSt.Lawrence,adetailedhydrodynamicmodelabletoprovidewaterlevel,currentsand temperature information is available from Montréal to Quebec City, with an average spatialresolution of 190m and refinements down to a fewmeters. BetweenMontreal and Trois-Rivières,accuracyofwaterlevelpredictionsisontheorderofafewcentimetres.Downstream,errorsinwaterlevelsarebelow5%of the local tidalranges.Ahydrologicmodelcalculating the total inflowto theriverfromtheGreatLakestoQuebecCityisalsoavailablefromtheCentreMétéorologiqueCanadien(CMC).Thesitehaspermanenthydrometricstationscoveringtheentiredomainthatprovidereal-timewaterlevelanddischarge(fewsitesonly)data(Figure56).Mapsof the floodplain,emerging,aquaticandshore vegetation, substrate, roughness as well as derived manning coefficients are also available.LiDARandbathymetricsurveyswereconductedduringthe2000stogetahighresolutionDEMofthewholearea.As of this day themodel runs in stationarymode betweenMontréal and Trois-Rivières on a dailybasis on the CMC computers. The entire system, including the tidal part of the river, is pendingcompletion.Predictionsofwater levelsarealsoavailableonthewholedomainfromanoperationalone-dimensionalmodel.A2Dhydrodynamicmodelofoneoftherivertributaries,theRichelieuRiver,isalreadyavailable

Figure56.PermanentandseasonalhydrometricstationsontheSt.LawrenceRiver.



112

7.2.1.6.4 Pre-launchSiteCharacterizationAt each site, EC and the STwill install networks of pressure transducers to validate SWOT heightmeasurements along river channels. These transducers can also be calibrated with in the fielddischargemeasurementstomakeanetworkofgauges,andthiswillbeperformedpriortolaunchatleast at the Tier I sites, and this information will be used to develop well calibrated dischargealgorithmsforthesites.AttheTierIsites,thesewillbepermanentinstallations,surveyedintoplaceusingGNSSGPS,andleft inplaceuntilat leasttheendofthefast-samplingphase. InearlySeptemberofeachyeartheinstallationswillberevisitedfordatadownload.Thiseffortwill be led byEC, usingEC boats, equipment, and field technicians. Also at the Tier I sites,meteorologicalstationswillbeinstalledpriortothefastsamplingphasetounderstandbothenergybalanceforamorecompletehydrologicunderstandingofthebasinsandtovalidateSWOTsrainflags.Inadditionto theabove,specificmeasureswillbe lookedat for thePAD. Highprecision fieldworkwas conductedduring the1990sand resumed in2010withnewgroundLiDARsurveys, enhancedwater level monitoring, characterization of surface water connectivity, and should intensify inupcomingyearsfortheSWOTprogram.WorkfrombothNRCANandECwillallowthecreationofahigh resolution DEM extending further out of the floodplain and more towards the limits of thedomain,allowingthecreationofa2DhydrodynamicmodelontheregionwhichisbeingworkedonbyECandtheuniversityofSherbrooke.Waterlevelmonitoringonapermanentandseasonalbasisisdoneeveryyearonmorethan25sitesandisplannedtocontinueandexpandtomoresitesaspartofthe Joint Alberta-Canada Oil SandsMonitoring Program. A UAV overflight for validation of a SARsurfacewaterandfloodedvegetationproductwasconducted in2015byNRCAN.Cornerreflectors,meteorologicaldata,bathymetricandsinglelaserLiDARwereusedduringthissurveyandcouldhelpinthehighresolutionDEMcreation.Radarsat-2derivedwaterextentdetectionovertheregionwillbe continued. It is planned to useADCP and shallowwatermultibeam echosounder on the lowerAthabasca River , along with high precision GNSS to characterize sensible areas on the rivermainstemin2016andmoreisplannedinthecomingyearsforkeyconnectingchannels.Astudyofsome of the lakes phenomenology (wind seiche, emergent vegetation, vegetation on lakemargins,mudandfloatingvegetation,wavesandlowslopemudshorelines)andtheireffectsoncorrectwaterlevel and extentmeasurements and instrumentation placement limitationswill also be conducted.Windforecastsmodel,alongwithmeasurementsonsitewillalsocontinuetobeworkedontostudywindseicheondifferentlakesinCanada,includingLakeAthabasca.PlansforthestretchesalongtheSt.LawrenceincludepermanentinstallationofanacousticDopplervelocity meter (ADVM) in Quebec City, scheduled for installation prior to the SWOT launch.Temporaryinstallationofthedevicehasbeentestedandhasshownsatisfyingresults,providingreal-timedischargevaluesfortheSaint-Lawrence.Theuseofabetterhydrological(runoffs)estimationofthetributariesandtheungaugedbasinsasaninputtothehydrodynamicmodelisundertestingandshould improve the overall precision of the model. Pressure transducers will be placed in thetributariestobetterestimatethehydrologicaldelaybetweenthehydrometricstationsandtheriver.Thehydrodynamicmodelofthesiteisscheduledtoberuninnon-stationarymodefromMontrealtodownstreamofQuebecCitybytheendofthe2016-17fiscalyear.Predictionsinnon-stationarymodeshould be made available in the coming years. The International Great Lakes Datum of 1985(IGLD85) is scheduled to be updated in 2018 andwill be fixed to a geoid. Thiswillmake verticalinformationofthesiteeasiertoconverttothechosenverticaldatumforthemission. Furtherfieldwork to update and improve the current information about the site will be conducted whennecessary.



113

7.2.1.6.5 Post-launchCal/ValActivitiesDepending on the time of year of the fast sampling phase, the Canadian / Northern sites are alsoopportunetovalidateSWOT’siceflags,andmayprovideacomplexicebreakupenvironmentforthisvalidation. During SWOT’s mission, remotely sensed datasets will be used to validate inundationarea, and transducer installations from the pre-launch permanent installations will be used tovalidateSWOTheightandcharacterizeriverdischargealgorithms.SpecificallyinthePAD,Thefast-samplingorbitcoversthewesternendofLakeAthabascaandtotheNorthontheSlaveRiver,thelatterbeinganotherCanadianvalidationsite.Detectionofwindseichewillbepossiblealongwithmorelimitedwaterinundationextentdetectionduringthisphaseofthemission.For the PAD specifically, during the science and fast sampling phase of the mission, Radarsat-2derived imagery representing thewater extentwill be collected close to or coincidentwith SWOToverflights to validate water detection. Unlike optical imagery, this will provide water extentinformation regardless of the conditions. Pressure transducers will be installed on sensibleinundationareasandsmalllakesaswellasonthePeace,SlaveandAthabascamainstemstoenhancetheexistingWSChydrometricnetworktomonitorwaterlevelvariations. ADCPcanbeinstalledfordischargemeasurements where no permanent real-time hydrometric station is available. Ground-based surveys are also planned to better define inundated vegetation “shoreline” areas of waterbodiesandwaterelevationmeasurementsandmodelling.In theST.Lawrence,Radarsat-2derived imageryof the iceandwaterextentclose toorcoincidentwithaSWOToverflightwillbemadeavailableregardlessofthemeteorologicalconditionasopposedtoopticalimagery.ACODARHFradarsystemcouldalsobeputonsitetogetinformationaboutthesurfacecurrentunderthefastsamplingorbitorelsewhereonthesite.Thefast-samplingorbitcoversthetidalpartoftheriverdownstreamofQuebecCity.Waterlevelanddischargedata,aswellashigh-resolution (~200m average) 2D hydrodynamic simulations and predictionswill bemade availablealongwithSWOTsimulations.Groundbasedsurveyswillalsobeconductedtomaintainthequalityofthedataprovidedduringthewholemissionandonanad-hocbasisshouldtheneedforadditionalinformation arise. The final 2D hydrodynamic model will also be available to simulate SWOToverflight to provide other validation strategies prior to an actual SWOTmeasurement should anAirSWOToverflightbeunavailable.

7.2.1.7 South American Validation Sites: South American Rivers (France & Foreign PartnerSites)

7.2.1.7.1 SitedescriptionWorkinginSouthAmericanriversimplieshavingresearchprojectsendorsedbylocalinstitutions.Inordertosetupprojectsandagreementswiththeselocalinstitutions,agroupofSWOTearlyadopterswasconstituted,withakick-offmeetingheldinMay,12-14th,2015inRiodeJaneiro,Brazil.Asecondmeetingwilloccur inMarch2018.The final listofsites,andtheirrankingasTier-1site,additionalsite,etcwillultimatelydependontheprojectsandagreementsthatcouldbesetupbylaunch,andbythefundingavailablefortheseprojects.ThreepassesofthefastsamplingorbitcrosscuttheSouthAmericacontinent,includingtheAmazonbasin,thelargestwatershedintheworldwiththousandsofcontributorriversofallsizes.Thisbasinhaslongbeenusedasavalidationsiteforaltimetrymissionsandmostofthetechnologiesproposed



114

in thisdocument for thevalidationovercontinentalwatershavebeendevelopedandtested in thisbasin.ThisincludesinstallationoflevelledgaugesandprofilesoffreesurfacesbyGPSonboardboats.Today,morethan20,000kmofGPSprofileshavebeencollectedatdifferentseasonsoverthemajortributariesoftheAmazonbasin.ItisestablishedbyMoreiraetal.(inprep)thattheaccuracyoftheseprofilesisatthe2cmlevelatthe2kmhorizontalscale.Theseprofilesrevealthechangesinheightandslopealongtheprofileswiththeriseorfallinthehydrologicalcycle,inparticularhowfarfromtherivermouthdothedampingoftheslopepropagatebecauseofthebackwatereffect.Today,theseGPS profiles constitute the best way to derive continuous profiles of the free surface slope.Consequently, this technique of GPS systems embarked on large boats will be privileged in theCal/ValoperationsdealingwiththevalidationofheightsandslopesintheSouthAmericanrivers.

Figure57.CoverageofFastSamplingOrbitoversouthAmerica.ThecourseofthemajorcontributorsoftheAmazonareshowninlightblue.



115

Figure58.LevelingofagaugebydirectobservationwithaGPSsystemontopofagauge(foreground),andsurveyoftheriverheightandslopewithaGPSsysteminstalledontheroofoftheboat(background).

Figure59.ExampleofGPSalongtheNegroriveratdifferentphasesofthehydrologicalcycle.Notetheflatteningoftheprofileattherivermouth,duetobackwatereffectcontrolledbythelevelintheAmazonriveritself.



116

Figure 60. Longitudinal derivative of a GPS profile of the free surface of the Madeira river (averaged every 10 km), usingdifferentgeoidmodels.Notethatthemajorundulationsexhibitedbytheslopeprofilearenotsignificantlyalteredbythechoiceofthegeoidmodel.

7.2.1.7.2 SiteGoalsTheSouthAmericansiteswillbeusedto:-1validatetheSWOTheights-2ValidatetheSWOTslopes- 3 check for long wavelength error. The Fast Sampling tracks of SWOT over South America willcrosscutriverreachesallalongthepass.Jointanalysisoftheheighterrorsallalongthepasseswillbeconductedtoevidencepossiblelongwavelengtherrors.-4checkforrollerror-5Characterizedischarge

7.2.1.7.3 SiteInstrumentationBrazilmaintainsanetworkofhundredsofhydrometricstationsthroughoutthecountry.Thesedataare distributed freely and consequently theywill bemade available to the project. Specifically forSWOT,a listofprioritystations thatmustbedelivered inNearReal timewillbeestablished in theframeofaSWOTEarlyAdoptersproject.TheAndean countries collectwater levels but today, theydonot distribute thedatapublically.AnagreementwiththerelevantorganisationswillbeestablishedbetweentheFrenchandUSprojectsononesideandthelocaloperatorsontheotherside,inorderthatthedataaremadeavailableatleastduringtheFastSamplingorbit.TheseagreementsconstituteoneoftheobjectivesoftheSWOTearlyAdoptersgroup.

7.2.1.7.4 Pre-LaunchsitecharacterizationAllthesitesthatwillbeselectedwillhaveatleastapairofgaugesinorderthattherolleffectcanbeevaluated.Ifnotexisting,thesegaugeswillbeinstalled.



117

Allthegaugeswillbelevelled.ifnotalreadydonebylocalagencies,GPSlevellingofthegaugeswillbeperformedbeforethelaunchofSWOT.Allthesiteswillhaveat leastonemeasurementofthereachcrosssection.Ifnotavailablefromthelocalagencies,atleastoneADCPmeasurementwillbeperformedforthesitesselected.All thesiteswillhavea stage -dischargerelationship (ratingcurve). Ifnotavailable fromthe localauthorities,aratingcurvewillbeestablishedona“besteffort”basisusingsatellitealtimetryandarain-dischargemodel.Asmuchaspossible,thesiteswillbeselectedatcrossingswithotheraltimetricmissions,inparticulartheSentinel-3missions.Thewaterlevelseriesgainedfromthesemissionswillbeusedtoestablishtheratingcurves.Acapabilitytosendthein-situinformationwithinafewwillbeestablishedwhenevernecessary.In the frameof theSouthAmericanEarlyAdoptergroup, aneffortwill be carriedout at the inter-agencyleveltolevelasmuchaspossiblethenetworksofexistinggauges,withprioritygiventothegaugeslocatedintotheswathsoftheFastSamplingorbit

7.2.1.7.5 Post-LaunchactivitiesAllthesiteswillbeselectedastobeoverflownduringthefastsamplingorbit.Duringthisphase,thein-situ stages will be collected daily at the time of overpassing. As much as possible, an ADCPmeasurementwillbeperformeddaily.Itwillprovidethevalueofthedischargeandreachwidth.On a longer term, the SWOT heightswill be compared to the series of vertically referencedwaterlevelscollectedatthegaugeslevelledpriorthethelaunch.Formanystations,thedataarecollectbymeansofatechnicianhiringaboattogofromonestationtotheother.Asmuchaspossible,thesecruiseswillbeconductedatthetimeofoverpassingbySWOTsincetheywillbeusedtocollectprofilesofthefreesurfacebymeansofGNSSstationinstalledontopoftheboat(ortrailedbehindtheboat).TheseprofileswillbecomputedusingCNES’softwareGINS-PCandusedtoassesstheSWOTslopeproducts.

7.2.2 LakeCal/ValSites

7.2.2.1 LakeIssykkulValidationSite(FrenchProjectSite)

7.2.2.1.1 SiteDescriptionLakeIssykkulislocatedinCentralAsia,inKyrgyzstan,andserveofficiallysince2008asaCal/ValsiteforsatellitealtimetryonJason-2,Envisat,SARAL/AltiKaandfuturenadiraltimeters.Ithasalengthof180kmandawidthofabout60/70km.Westpartofthelake’sshorelineisveryshallow,whileeast,northandsouthpartiscoveredwithhighmountains.Itistheregionoftheworldthemostfarfromanyocean.WelllocatedinthecenteroftheEurasiaitwillbeaperfectlocationforadditionalsiteforcalibrationofcrosstrackerrorinparticulartorollandphaseerror.Theseichesarenotfrequent,nottoo high (generally smaller than 10 cm) and they could be monitored as they are preferentiallyorientedintheEast/Westdirection:aninsitugaugewithdatatimesamplingof5minutesisalreadyinstalledintheEastsideofthelakewheretheeffectisthehighest.Theaccessisquiteeasythanksto10yearsofcollaborationbetweenLegosandKyrgyzinstituteofhydrology.Vesselfornavigationoverthe entire lake is possible all the year at a reasonableprice for suchboat. Basedonmore than10yearsofcollaborationwithlocalauthoritieswhichhasbeenconfirmedforthefollowingyears,aMeanlakesurfaceathighspatialresolutionisunderconstruction.



118

ItislargeenoughtobecrossedbySWOTseveraltimesovereachcycleof21days.LakeIssykkul,inthehalfeastpartwillbefullycoveredbySWOTduringthe1-Dorbitfastsampling.LakeIssykkulhasbeeninstrumentedwithpermanentGPSreceivers,2waterheightgauges,andweatherstations.

Figure61.PreliminarymeanlakesurfaceofIssykkulfromacombinationofCryosat-2,IcesatandGPSfieldworksfrom2004to2010.Worktobecontinuedbeforethelaunch.



119

Figure62.LakeIssykkul(fromGoogleEarth)with1Dfastsamplingorbitrepresentation ‘red:nadir,andyellowlines,swath’slimits).

7.2.2.1.2 SiteGoalsDuringthefastsampling1Dorbit,theLakeIssykkulwillbeusedto-validateheightdetermination-validateslope-calibratetherollandphaseerrorsusingameanlakesurfacewhichwillbeinferredfromabout15yearsoffieldmeasurementswithmovingGPSovera1kmresolutionontheentiresurfaceofthelake.

7.2.2.1.3 SiteInstrumentation-Weatherstationatthreelocations- 2 in situ level gauges, one with measurement twice daily, one with one measurement every 5minutes-2permanentGPSreceivers-Vesselavailableforfieldworkonthelake

7.2.2.1.4 Pre-launchSiteCharacterizationThemain task forpre launchwillbe tocomplete thecalculationof themean lakesurfacebysomeadditionalfieldworkintheframeofOSTSTprogramforCal/Valofnadiraltimeters

7.2.2.1.5 Post-launchCal/ValActivitiesDuringthe1Dfastsamplingphase:-dailymeasurementofwaterlevelofthelakeattheexactdateofpassofthesatellite-installationofaGPSlocalnetworkalongtheshorelineofthepartofthelakecoveredbySWOTfordeterminationoftroposphericdelay-monitoringofseicheeffectusinginsitudataandwindfieldmeasurements(networkofanenometerwillbetemporaryinstalled)- GPS kinematic profile on the cross track direction right over the Swath to calibrate for roll andphaseerrorandcomparewithmeanlakesurfaceDuringthenominalphase:-comparepassperpassthewaterlevelmeasuredwithSWOTwithinsitumeasurement- compare the mean lake surface measure by SWOT with those obtained from the historical GPSfieldwork.-usetheseverticalprofilesbetweenthenadirandtheswathtochecktheconsistencybetweenKaRINandthenadiraltimeteravoidinganyerrorsduetointerpolationwithinthenadirgap.-Using insituwaterheightathightemporalresolution(5minutes)andthemean lakesurfacewillalsoallowcalculatingthestaticrangebiasesofbothKaRInandthenadiraltimeter.

7.2.2.2 LakeTahoeCal/ValSite(U.S.ProjectSite)

7.2.2.2.1 SiteDescriptionTheLakeTahoeCal/Val Site is locatedatLakeTahoeon theborderofCaliforniaandNevada,USA(Figure63).LakeTahoe is locatedatrelativelyhighaltitude,1,900m,and isrelatively largewithasurfaceareaofapproximately490squarekilometers.LakeTahoeisagoodlocationforaCal/Valsiteduetothelargeamountofcompileddataandactiveresearchonthelake.Activeresearchonthelakethat isofparticular interest toSWOTis thatdonebytheU.C.DavisTahoeEnvironmentalResearchCenter,TERC(http://terc.ucdavis.edu/),which includesanetworkofbuoysandboatdeployments,



120

thatareusedincombinationwithJPLforcalibrationofotherNASAsatellites,suchasTerra,Landsat,Aqua,andEnvisat.TheLakeTahoeCal/ValSiteisnotwithintheSWOT1-dayFastRepeatOrbit,andassuch,thepost-launchactivitiesprimarilywilltakeplaceduringtheScienceOrbit.

Figure63.LakeTahoe.

7.2.2.2.2 SiteGoalsThehydrologygoalsoftheLakeTahoeCal/ValSitearetovalidateSWOTmeasurementsofabsolutewater surface height, and inundated surface area for a large, high-altitude lake. In addition, othercomponentsof the SWOTerrorbudgetmaybevalidatedheredue to theuniquepropertiesof thissite,suchasrandomheighterror,rangedrift,androll/phasedrift.

7.2.2.2.3 SiteInstrumentationLakeTahoehasavarietyofinsituinstruments,includingseveralstagemeasurementdevices(someoperatedbyUSGS,othersbyTERC/JPL,andUSCoastGuard),severalmetrologystations(operatedbyNWSandTERC),inflow/outflowrivergagesatallthemajortributaries(operatedbyUSGS),andfourstationary buoys (operated by TERC and JPL; http://laketahoe.jpl.nasa.gov/get_met_weather) thatrecordatmospheric,radiation,andwaterqualitydatabutnotwaterstagedata.Inaddition,anumberof research vessels are available for making day-of-flight measurements. Aerial lidar for the fullwatershedwascollectedin2010and2012andhigh-resolutionDEMsoftheshorelineareavailable.



121

Figure64.TERC/JPLstationarybuoyinLakeTahoe.

7.2.2.2.4 Pre-launchSiteCharacterizationThepre-launchsitecharacterizationof theLakeTahoeCal/ValSitewasstarted in2013during theAirSWOT flights of that year. Those experiments included boat-towed buoys with GNSS stagemeasurements. Additional pre-launch characterization of the site will include supplementing theexisting instrumentationwithstagerecorders (suchasaddingstagemeasurementcapability to theexisting stationary buoy instrumentation), as well as deploying approximately fifteen pressuretransducersaroundthelaketostudywind-drivenchangesinsurfacewaterelevationsacrossthelakeif these data have not been previously collected. Because Lake Tahoe is not under the 1-day FastRepeat Orbit, no immediate pre-launch setup will need to take place. Instead, the pressuretransducerswillbedeployedpost-launch.

7.2.2.2.5 Post-launchCal/ValActivitiesPost-launchCal/ValactivitiesattheLakeTahoeSitewillincludeinstallinganarrayofapproximately20 pressure transducers spaced around the lake shoreline (15 in lake, 5 atmospheric). ThesetransducersmaycomefromotherCal/Valsitesthathadbeenstudiedduringthe1-dayFastRepeatOrbit. Deployment of the transducers will help to supplement existing in situ stage recorders todetermine water surface variations across the surface of the lake. If water surface slopes aredeterminedtobesignificant,two-daysofboat-measuredtransectsrecordingwatersurfaceelevationsacrossthelakewillbeperformedtocorrespondwithoverflightsbySWOT.

7.2.2.3 PrairiePotholesSmallLakesCal/ValSite(U.S.ProjectSite)

7.2.2.3.1 SiteDescriptionThePairiePotholesSmallLakesCal/ValSite is locatednear Jamestown,NorthDakota,USA(Figure65).ThesiteistypicalofthePrairiePotholesRegion,apost-glaciallandscapestretchingfromIowatoAlberta, Canada, that is composed of numerous small waterbodies, ranging from several tens ofsquare meters up to several kilometers in surface area. Some of the smaller waterbodies areephemeralandgodryduringtheearlyfall,butmostofthelargerwaterbodiesareperennialandhavedynamic(1+m)changes inwater-surfaceelevations. Inaddition, thePrairiePotholeCal/ValSite is



122

affectedby ice,snow,andrain,making itagood locationforalsotestingtheperformanceofSWOTflagged data products. The Prairie Potholes Validation Site is the location of amulti-decadal studyfundedbyUSEPAandUSGSinvestigatingwaterdynamicsinthisregion.

Figure65.MapofthePrairiePotholesValidationSite,nearJamestown,NorthDakota,USA.Secondpanelshowsaclose-upofthenumeroussmallwaterbodiestypicaloftheregion.

7.2.2.3.2 SiteGoalsThegoalsofthePrairiePotholesValidationSitearetovalidateSWOTmeasurementsofsurface-waterheightandinundatedareaforsmallwaterbodies,someofwhichareephemeral.



123

7.2.2.3.3 SiteInstrumentationAerial lidar was flown in 2010-2012 and high-resolution DEMs are available for the site. It isunknowniftheaeriallidarwasflownwhenthepotholeshadlowwaterlevels.Thereisoneexistingstream gage in the upperwatershed and one in the lowerwatershed, both are operated byUSGS.Therearehistoricaldatasetsofsurfacewaterchangesinseveralofthelargerwaterbodiesandeachyear approximately five to ten pressure transducers are deployed in waterbodies to record theirdynamics.Inaddition,thewatershedoftheCal/Valsitehasseveralcalibratedcoupledsurface-andground-water numerical models at various resolutions from medium to fine, with all the modelsavailablefromtheUSGSandUSEPA.

7.2.2.3.4 Pre-launchSiteCharacterizationTherearethreeprimarypre-launchactivitiesthatareproposedforthePrairiePotholesregion:1.)Astudy to understand wind setup on small lakes (proposed for 2017); 2.) A study to determineaccuracyofderivedinundationextentfromtheintersectionoflidardemswithwaterlevel(proposedfor2017);3.)Astudytoevaluateaccuracyofinundationextentvalidationmethodsinlow-elevation,non-wetland environments using AirSWOT (proposed for 2017, will require 2-days of AirSWOTflights).Forthesepre-launchactivitiesproposedfor2017,approximately40-60pressuretransducerswillbeinstalledtomeasurewater-surfaceelevations(~50insmalllakes,~10foratmosphericcorrections).Windsetupwillbeevaluatedusing thepressure transducernetwork,aswellasdeploymentof tenweatherstationsthataretobepurchasedforCal/Valuses.LidarDEMsalreadyexistforthesite,butapproximately twoweeksof fieldworkwillbeneededtovalidate inundationextentusing the fieldmethods proposed earlier (Section 6.4.3.2), andwill be synchronizedwith two-days of concurrentAirSWOToverflights.Ashort2-3dayfieldvisitwillberequiredtoremovethepressuretransducersattheendofthedeploymentperiod.Therewillneedtobeapproximatelythreemonthsofstafftimesetasidetoevaluateandcompareallthevariousdatacomponents.

7.2.2.3.5 Post-launchCal/ValActivitiesThePrairiePotholesCal/ValsiteisnotlocatedundertheSWOT1-dayfastrepeatcycle,assuch,post-launch Cal/Val activities primarily will occur during the science orbit portion of themission. Thepost-launchCal/ValactivitiesatthePrairiePotholesCal/Valsitewillrelyprimarilyontheinstallationof approximately 30 pressure transducers (~25 in small lakes, ~5 for atmospheric correction).DependingonthetimeofyearofthelaunchofSWOTandthelikelihoodoficecover,theinstallationofpressuretransducersatthissitewilloccureitherpre-launchorduringthe1-dayfastorbit.Ashort1-2dayfieldvisitwillberequiredtoremovethepressuretransducersandapproximatelytwoweeksofstafftimewillberequiredtoworkupthedataandcompareittoSWOTdataproducts.

7.2.2.4 YukonFlatsLake&WetlandValidationSite(USProjectSite)Seesection8.2.3.2forfulldescription

7.2.2.5 SierraNevadaAlpineSmallLakesValidationSite(USProjectSite)

7.2.2.5.1 SiteDescriptionThe Sierra Nevada Alpine Small Lakes Validation Site (SNASL), is located in the Sierra NevadaMountainsofCalifornia,approximately50kmsouthofLakeTahoe(Figure66).Thissitewaschosentohelpisolatethewettropospherecomponentoftheerrorbudgetduetotherelativelylowamountof wet troposphere compared to more lowland sites. In addition, high-elevation lakes provide an



124

important source of water for many human populations and understanding the errors associatedwiththeseenvironmentswillcontributetoSWOTsutility.

Figure66.MapoftheSierraNevadasmalllakesCal/Valsite,withLakeTahoetothenorth,andMonoLaketotheEast.



125

Figure67.CloseupofSierraNevadasmalllakesCal/Valsite.Thereareapproximately80smalllakesintheenclosedarea,manywithrelativelygoodroadaccess,andthesiteisprimarilyinYosemiteNationalPark,whichisinterestedinstudyingtheselakes.

7.2.2.5.2 SiteGoalsThegoalsforthisCal/ValsitearetounderstandSWOTdynamicsforsmallhigh-altitudemountainouslakes.Theselakesareofprimaryinterestbecausetheyhaveaverylowwettropospherecomponentandareimportantcontributorstothewatersupplyforhumanpopulations.

7.2.2.5.3 SiteInstrumentationAtpresent,thereisnoexistinginstrumentationatthesesites.

7.2.2.5.4 Pre-launchSiteCharacterizationTheSierraNevadasmalllakesCal/ValsiteisoutsidetheSWOT1-dayfastrepeatorbit,assuch,pre-launch site characterization will be minimal. Permissions to install pressure transducers and siteidentificationwillrequireapproximatelytwoweeksofstafftime.

7.2.2.5.5 Post-launchCal/ValActivitiesPost-launch Cal/Val activities include installing pressure transducers with GNSS-survey levelaccuracy,most likely intheperiodimmediatelyafterthe1-dayfastorbit. Sitecharacterizationwillconsist of installing approximately 30 pressure transducers (~25 small lake sites,~5 atmospheric



126

sensors),primarilyintheregioncoveredbyaeriallidar,justsouthwestofMonoLake(FigureXX).Ateamoftwofieldmemberswillneedapproximatelyoneweektoinstallthepressuretransducers,andapproximatelyoneweektoremovethem.Approximatelyone-twoweeksstafftimewillbeneededfordataprocessingoftheGNSSdataasthestaticcollectionsaremorecomplexthanRTKcollectionsatotherCal/Valsites.

7.2.2.6 SouthAmericanValidationSites:AndeanLakes(France&ForeignPartnerSites)

7.2.2.6.1 SiteDescriptionIn southChile there is a set of few lakes forwhichdaily gageheights are available fromDirecciónGeneraldeAguas(DGA)throughasystemofpublicrequest(www.dga.cl).TheLakesinsitudatahavetobeleveledusingGPSpositioningandwillthenbeavailableforrealtimevalidationofwaterheightfromSWOT.These data have already been used to validate the SARAL/Altika data and compare with Envisatproducts.

Figure 68. Examples of In Situ water level for Lake Ranco and Llanquihue compared to Envisat and SARAL/AltiKameasurements.

Moreover,followingafirstcampaigndonein2005over3ofthelakesoftheregionsnamed“loslagos”near the city of Puerto Mont, some additional leveling of the lake surface using kinematic GPSmeasurements will be performed in cooperation with the university of Concepcion: under theframeworkoftheSouthAmericanGroupofSWOTearlyadopters.



127

Figure69.MapoftheLosLagosandsomepreliminaryverticalprofilesalongthelakesRancoandTodosdelosSantos.

7.2.2.6.2 SiteGoals-InsituvalidationofwaterheightmeasurementsfromSWOTforlakesinmountainarea.-ValidationofslopeaccuracyusingverticalGPSprofileoverthelakeThe topography of the surrounding of the lakes of theLos Lagos regionwill allow testing layovereffect (mountainsandvolcanoesarepresentaswellas forestsalong theshore)but these lakesarealso large enough to present some non-negligible geoid variations (see Figure 69). Some smallislandsandverycomplexshorelinedelineation(especiallyfortheTodosdelosSantos)maybeagoodtargetforclassificationvalidation.

7.2.2.6.3 SiteInstrumentationTheselakesareequippedwithinsitugaugesdeliveringdailydatawithoutanyrestrictions.Theregioniseasilyaccessibleandalllakesareconnectedbyroads

7.2.2.6.4 Pre-launchSiteCharacterization-Delineatingthelake’sshorelinesfromGPSsurveycouldbedonepriortothelaunch.-FieldworkwithkinematicGPSinordertodeterminegeoidslopeoverthelakesRanco,LlanquihueandTodosdeLosSantos.Thereisaneedtorepeatthislevelingatdifferentperiodsinordertocheckthepotentialseicheeffects.-checkavailabilityofmeteostation(anemometerinparticular)fordeterminationofpotentialseichesovertheselectedlakes-GPSlevelingoftheinsitugauges(incoordinationbetweenuniversityofConcepcionandDGA

7.2.2.6.5 Post-launchCal/ValActivities-regularGPSlevelingofthegauges-usetheNearRealTimeinsitudataforvalidationofwaterlevelmeasuredbySWOT-usetheverticalprofileforvalidationofslopeandcalibrationofroll/phaseerror.



128

7.2.3 WetlandValidationSites

7.2.3.1 LowerMississippiRiverWetlandCal/ValSite(U.S.ProjectSite)

7.2.3.1.1 SiteDescriptionTheLowerMississippiRiverWetlandCal/Valsite is located inSouthernLouisiana(Figure70).Thesitehas a rangeofdifferentwaterbodieswithvaryingdegreesof inundationandvegetation cover,whichwillbeusedtotestSWOTperformanceformeasuringwater-surfaceelevationandinundationextentinwaterbodiesunderavarietyofvegetationtypes.AkeycomponentoftheLowerMississippiRiverWetlandCal/Valsiteistheextensiveexistinggagenetwork,theCoastwideRegionalMonitoringSystem(CRMS).TheCRMSnetworkconsistsofnearlyfourhundredstationsatwhichmeasurementsof water-level elevation, vegetation classes, percent vegetation cover and other parameters aremeasured (http://pubs.usgs.gov/fs/2010/3018/pdf/FS2010-3018.pdf).Data canbeaccessedeasilyonlineattheCRMSdatawebsite(http://lacoast.gov/crms2/home.aspx).

Figure70.Mapof theLowerMississippiWetlandCal/Val site.TheCoastwideRegionalMonitoringSystem(CRMS)gagesareshowningreen.

7.2.3.1.2 SiteGoalsThegoaloftheLowerMississippiWetlandCal/Valsiteisto:

1. ValidatetheabilityofSWOTmeasurementstoaccuratelycharacterizeandmeasurewaterpresence under a range of vegetation types, from high-canopy cover to emergentvegetationtomixedopenwater.

2. Validate the ability of SWOT measurements to accurately characterize and measurewater-surfaceelevationandinundationextentoverarangeofvegetationtypesinwaterbodies,includingephemeralandfluctuatingwaterlevels.

7.2.3.1.3 SiteInstrumentationTheLowerMississippiWetlandCal/ValsitehasalargenumberofexistinggagesthatshouldsufficeforSWOTandAirSWOTcomparisonsbutifanyadditionalstagemeasurementsarerequired,project



129

pressure transducers will be deployed. In addition, the survey quality of the water-levels of theexistinggageswillbeverified.Oneormoreweatherstations,potentiallyincludingstationsinstalledwithinthevegetationwouldbeusefultohelpcontrolforvegetationmotionfromwindandwavesontheunderlyingwatersurface.FlightsofAirSWOTwithKaSPARandanear-infraredcamerawouldberecommended.

7.2.3.1.4 Pre-launchSiteCharacterizationThewetlandsitewillneedtohaveaeriallidardataavailabletocharacterizethetopographyandthevegetationheights,withthedatacollectionpreferablycollectedasclosetoSWOTlaunchaspossible.There are several existing lidar datasets, with significant overlap between years and with somerelativelyrecent(2013,2015),suggestingthatthissiteisofsufficientinteresttorequirerepeatlidarflights. Ifarecent lidarflight isnotavailablewithinayearpre-SWOTlaunch,a lidardatacollectionflightwouldberecommended.Morerecentaeriallidardatawillbepreferredtositeswitholderdatabecausethevegetationmayhavechangedsignificantlysincethelidarwasflown.

Aerial lidardatawillbeused tosegment thestudysite intoclassesofvegetation (basedonheight,canopy structure, and density) as well as inundation depth determined from fieldmeasurements.FieldmeasurementsthatwillbehelpfulcharacterizetheCal/Valwetlandsiteincludeinstallationofpressuretransducersinwaterbodieswithinarangeofdifferentvegetationtypes,aswellasweatherstations tohelpparameterizeeffectsofvegetationmotionandwavesonSWOTmeasurementsanddata products. AirSWOT flights over the chosen wetland site in the pre-launch period have beenperformedbutadditional flights, ifneeded,wouldberecommendedtohelpdevelopandtestSWOTproductsforwater-surfaceelevationandinundationextentundervegetationcanopies.

7.2.3.1.5 Post-launchCal/ValActivitiesAllofthechosenpotentialwetlandCal/ValsitesarewithintheSWOT1-dayfastorbitpath.Assuch,initialSWOTproducts,suchasinundationextentandwater-surfaceelevation,willbevalidatedfromthe wetland Cal/Val site. The post-launch characterization will consist of the installation of 20 ormore temporary pressure transducers in water bodies underlying a range of vegetation types.Installationofweather stationsmaybenecessary if they are found tobehelpful in thepre-launchtesting. AirSWOT underflights, with KaSPAR and near-infrared camera measurements, coincidentwithSWOTpasseswillbeuseful forvalidatingtheSWOTmeasurementsofwater-surfaceelevationandinundationextent.

7.2.3.2 YukonFlatsLake&WetlandValidationSite(USProjectSite)

7.2.3.2.1 SiteDescription TheYukonFlatsisalarge,tectonically-controlledbasininCentralAlaska,approximately150km north of Fairbanks. The Yukon River flows through the heart of the basin, andwetlands andthousands of small lakes dominate the surrounding areas. These lakes and wetlands arecharacteristicofsimilarfeaturesfoundatnorthernhighlatitudes,includingWestSiberianLowlands,portions of Eastern Siberia, and lowland areas inNorthernCanada. As such, validationof SWOT’sability todetect lakewater surface elevation, inundation extent and storage change, aswell as theability todetectheight and inundationextent inboreal vegetatedareas. TheYukonFlats receivedextensivestudyviaground-basedandAirSWOTmeasurementsduringsummer2015. Assuch, it isalreadycomparativelywell-characterizedrelativetosomeotherSWOTvalidationsites.



130

Figure71.TheYukonFlatsFieldSite

7.2.3.2.2 SiteGoalsTheprimarygoalsoftheYukonFlatsvalidationsiteareto:

5. Validatemeasurements ofwater surface elevation, inundation extent, andwater storage insmall wetland lakes. These lakes occur frequently across boreal regions and will be keytargetsforSWOT,andtheYukonFlatsisahighlyrepresentativeexample.

6. UnderstandtheabilityofSWOTtomeasurewetlandinundationunderborealvegetation.7. Althoughitisnotaprimarygoal,thissitewillalsoincludeatleastonelargeArcticRiver(the

Yukon), as well as multiple smaller tributaries. These could be used to validate SWOT-derivedheightandslopeifappropriatefielddatawerecollected.

7.2.3.2.3 SiteInstrumentationTheYukonFlatsstudyareaisnotwell-instrumented.Theonlypermanentstreamgaugeintheregionis on theYukonRiver at theDaltonHighwaybridge (listed as theYukonRiver at StevensVillage),substantially downstream of the Yukon Flats. Much of the Flats is locatedwithin the Yukon FlatsNationalWildlifeRefuge,andtherefugestaffhaveinstalledtemporarygaugesinanumberoflakesinthe past. Similarly, a research team from theUSGS led byMichelleWalvoord andRob Striegl hascollectedextensivehydrologicandbiogeochemicalmeasurements in theYukonFlats in thepast. Aportionoftheflatswasflownwithlidarin2010,buttheremainderoftheareaisnotyetcoveredbyahigh-resolutionDEM. Inaddition,a team ledbyTamlinPavelskyconductedextensive fieldwork intheFlatsduringsummers2015and2017.Waterleveltimeseriesweresuccessfullymeasuredin13lakes, data characterizing vegetation, wind speed, inundation extent were collected, and multipleAirSWOTdatacollectionsoccurred.One,onJune15th,occurredduringclear-skyconditionsandthusproducedsimultaneousradarandopticalimagesoftheregion.

7.2.3.2.4 Pre-launchSiteCharacterizationA significant degree of site characterization has already occurred during the summer 2015 fieldseason, including collection ofmultitemporalAirSWOT radar data, AirSWOToptical data, and fieldmeasurements of water surface elevation, wind speed, water/land boundaries, and vegetationcharacteristics.However,thelackofasuitablehigh-resolutionDEMpointstotheneedforadditionalworkprelaunch. Assuch,werecommendcollectionofahigh-resolutionLiDARDEMoverat leasta



131

portion of the study area prior to launch. Unlike other types of DEM, this data will provideinformationonbothvegetationheightandbareearthelevation. Thiscombinationwillproveusefulin characterizing the ability of SWOT to observe water under vegetation and will also help tocharacterizevegetation-inducedlayoverinnorthernwetlandenvironments.

7.2.3.2.5 Post-launchCal/ValActivitiesCal/Val activities after launch will focus on the first summer post-launch. In the Yukon Flats,pressuretransducerswillbeinstalledinlakestomeasurevariationsinwatersurfaceelevationandinundation extent. Lakeswill be selected to cover a range of sizes spanning the lower bounds ofSWOT detectability, from ~1 ha to >5 km2. Near-infrared aerial photography will be collectedcoincidentwithSWOToverflightstovalidatedetectionofinundationextent. Inaddition,theYukonFlats will serve as the primary validation site for SWOT detection of inundation under borealvegetation. Point measurements of water surface elevation, vegetation characteristics, andinundationextentwillbemadeusingground-basedsurveys. Inadditiontothesebasicmeasurements,twoadditionalsetsofmeasurementswouldprovidedesirablevalidationcapabilities.AirSWOTmeasurementswouldprovidefull,2-DvalidationofSWOTwater surface elevations in Yukon Flats lakes andwetlands andwould helpwith interpretation ofSWOTwaterclassification,asAirSWOTwillprovidesimilardataatmuchhigherspatialresolution.Inaddition,airborneL-bandSAR(e.g.UAVSAR)wouldproviderobustmeasurementsofwatersurfaceextent under vegetation in the Yukon Flats, which would offer a substantial improvement overground-basedsurveys.

AportionoftheYukonFlatsisincludedinthe1-dayrepeatorbit(Figure72).Ifthisorbitisavailable during the open-water seasons, it would be highly valuable to conduct a field campaignduringduring this timeperiod, as thehydrologyofnorthernbasins like theYukon tends toevolverapidlyduringandimmediatelyafterthespringbreakupofriverice.

Figure72.TheYukonFlatsFieldSite&theonedayrepeatorbit.



132

7.2.3.3 EvergladesWetlandValidationSite(USProjectSite)

7.2.3.3.1 SiteDescriptionThe Everglades Wetland Validation site is located in Southern Florida, USA, 75 km west of FortLauderdale(Figure73).Commonlyreferredtoasthe“RiverofGrass”,theEvergladesareaseriesofdynamicwetlandsandwaterbodiesthathavesignificantvariationsinwater-surfaceelevations,bothintra-and inter-annually,withcomplexwetlandwatersurfacesdue tovariations invegetationandtopography. One of the primary benefits of using the Everglades as a validation site is the largeamountofavailabledata,models,andsciencereports(over800atlastcountin2014).Muchofthisscience work has been funded through the multi-decadal, mutli-billion dollar ComprehensiveEvergladesRestorationPlan (CERP), supportedby theU.S.ArmyCorpsof Engineers and theUSGSGreater Everglades Priority Ecosystems Science Program. Funding for CERP and science in theEverglades is expected to continue for decades to come. In addition, the Everglades containsEvergladesNationalPark,BigCypressNationalPreserve,and theEvergladesLong-TermEcologicalReference (LTER) site (http://fcelter.fiu.edu/), all of which have extensive ongoing research anddata-collectionprogramsaswellasnumerousresearchersworkingthroughouttheregion.The hydrology of the Everglades is driven by rainfall, with an average rainfall of one and a halfmeters.Atypicalyearconsistsofarelativelydrylatefall,winterandearlyspring,followedbyawetsummer,when themajorityof rainoccurs.Typical intra-annual stagevariationsareapproximatelyone meter (Figure 74). During the high-water season, water spreads out over large areas, with adense multitude of water bodies present, ranging from open water of one square kilometer, toexpansesofwetlandfromseveralsquaremetersinareauptoseveraltensofkilometers.Allofthesewaterbodieschangedynamicallythroughouttheyearandsomegodryinthedryseason.Vegetationconsists primarily of varying densities of one- to three-meter high grass and some overstory treecanopy.The selected location of the ~600 square-kilometer Everglades Validation site is outside theEvergladesNationalParkboundaries toavoidpossible issueswith instrument installationanddatacollection,andfarenoughwesttoavoidpotentialair-trafficconflictswiththebusyMiami-DadeandFortLauderdaleairports.ThesiteisoutsidetheSWOTone-dayfastrepeatorbit,andassuch,willbeusedasavalidationsiteoncetheSWOTscienceorbitisachieved.



133

Figure73.LocationmapoftheEvergladesWetlandValidationsite,westofFortLauderdale,Florida,USA.TheredpolygonisthelocationoftheEvergladesWetlandValidationsite,andthewhitesquaresarethelocationofEDENstagegages.



134

Figure74.Plotofwater stage leveland rainfall fromanEDENreal-time stage recorder,L28S1, forWaterYears2013-2015(availableat:http://sofia.usgs.gov/eden/eve/).This site is typicalofmostEvergladeswaterbodies,withapproximatelyone-meterofstagechangethroughouttheyear.Thisparticularsite isoneofaboutadozen locatedwithintheSWOTEvergladesWetlandValidationsite.

7.2.3.3.2 SiteGoalsThegoals of theEvergladesWetlandValidation site are to validate SWOT rain-flagging, andSWOTmeasurementsofwater-surfaceelevationsandinundatedarea,particularlyunderamixedmoderate-heightvegetativecanopytypicaloflowlandwetlands.

7.2.3.3.3 SiteInstrumentationTheEvergladeshaveanextensiveandongoingdatacollectionnetworkforwater-surfaceelevationsandmeteorological conditions, aswell as high-resolution topographicDEMs, near-real timewater-surfacegeneration,andawealthofadditionaldatasetsthathavebeencompiled.MuchofthisdataisavailablethroughtheSouthFloridaInformationAccesssystem(SOFIA;http://sofia.usgs.gov/).Akeycomponent of the CERP and of particular use for SWOTCal/Val purposes is the EvergladesDepthEstimationNetwork(EDEN;http://sofia.usgs.gov/eden/),aseriesofapproximately300stagegageslocated throughout the Everglades, with the majority of the stage gages surveyed to high-qualityGNSSaccuracyandreportinginreal-time.



135

For rainfall and meteorologic conditions in the Everglades, there are approximately two-hundredreal-timemeteorologicstationsavailablefromtheSouthFloridaWaterManagementDistrictand81non-realtimerainfallgages,aswellas15-minuteNEXRADcoveragefromtheU.S.NationalWeatherService. A 2 km x 2 km gridded, 15-min rainfall data product is created by locally-correcting theNEXRAD data with the dense meteorologic station network (available throughhttp://sofia.usgs.gov/eden/eve/).

7.2.3.3.4 Pre-launchSiteCharacterizationPre-launchactivitiesfortheEvergladesWetlandValidationsiteincludeestablishingconnectionswithUSGSandotherresearchersworkingintheareaaswellasdevelopinganunderstandingthepotentialerrorsorbiases in theexistingdatacollectionnetwork(e.g.howaccurateNEXRAD-derivedrainfallestimatesmightbe,andtheaccuracyofwaterstageelevationsfromtheEDENnetwork).BecausethisValidationsiteisnotwithintheSWOTone-dayfastrepeatorbit,mostofthefieldactivitieswilloccurpostlaunch,thoughmostofthedatausedtovalidateSWOTwillrelyontheexistingdatacollectionnetwork.

7.2.3.3.5 Post-launchCal/ValActivitiesPost-launchactivitiesfortheEvergladesWetlandValidationsiteincludeapotentialaeriallidarflighttohelpcharacterizevegetationandcreateahigh-qualityDEM, if thesedataarenotavailableorthesite has not been flownwith lidar previously through theUSGS 3DEP program. CoordinationwithUSGSresearcherswillincludethecoordinationofasatelliteoraerialphotographydatacollectionforinundatedareaextent (e.g.near IR imagery), coincidentwithSWOToverflights.TheexistingEDENgagenetworkshouldbesufficientforcomparisontoSWOTwater-surfaceelevationandtheNEXRAD-derived rainfall estimates should be sufficient for SWOT rain-flagging, though the uncertainties ofthesenetworksshouldbewellunderstoodpre-launch.

7.2.4 Tidal/EstuarineValidationSites

7.2.4.1 SevernEstuaryandRiverValidationSite(UKProjectSite)

7.2.4.1.1 SiteDescriptionThe Severn River Cal/Val site is located in the South-West of the UK., downstream of the town ofGloucester (Figure 75) which is the upstream tidal limit. The river in this 90-km study reach isestuarinewithahugetidalrange(~14m,the2ndor3rdhighestintheworld)andextensivetidalflats.It is a single-threaded low energy meandering channel carrying a moderate sediment load andexperiencessomeevolutionofthebedandbanks.Thereachvariesinwidthbetween~100mattheupper end (approximately 5km below Gloucester) to 24km at the lower limit of the SWOT fastsamplingorbitswath(seefiguresbelow)



136

Figure75.RiverSevern study site (bluepolygon) covering thearea from the seaward limitof theSWOT fast samplingorbitswathupstreamtowherethechannelwidthdecreasesbelow~100m.



137

Figure76.ZoomofRiverSevernstudysite(bluepolygon).

Thereachexperiencessignificantriverflowinputs(lowflow~30-50m3s-1,floodflowsupto~1000m3s-1),whichareonlymarginally regulated. The tidalprism ismassive, and is sufficient to formaboreupto~2mhighduringspringtides.Thereareextensivemudflats,whichareinundatedathightide. River flood flowscanoccuratany timeof theyear,butaremorecommon in thewinter,andsignificantstormsurges(upto2mofskewsurge)canalsooccuraslowpressuresystemstrackoverthe area. The estuary has also been the proposed site of a large tidal barrage designed to harvestrenewableenergywithamaximumpotentialoutputofapproximately7%oftheUK’senergyneeds.TheRiverSevernhasnotbeen the focusofSWOT-relatedstudies todate, largelyasa resultof theUK’srelativelylateentrytotheMission.However,itistheonlymajoreastuarythatwillbesampledduringtheSWOTfastsamplingphaseanditisthereforeanessentialCal/Valandsciencesiteforthemission. The River Severn Cal/Val study sitewill therefore be used to validate SWOT’s ability todetect water-surface elevation, slope, inundation extent, and discharge in a highly dynamic tidalestuary.

7.2.4.1.2 SiteGoalsThereisoneprimaryvalidationgoalfortheRiverSevernCal/Valsite:

8. ValidateSWOT’sabilitytomeasureandcharacterizedynamictidalestuaries.Characteristicstobestudiedincludewater-surfaceelevation,slope,inundationextent,anddischarge.

7.2.4.1.3 SiteInstrumentationAlidarsurveyofsomepartsoftheshorelineexist,butthiswillneedtobeupdatedwithanewandspatiallycompletesurveytakenatlowwaterintheperiodpriortotheSWOTlaunch.Thiswillneed



138

tobesupplementedwithahighwaterbathymetricsurveyusingsidescansonarandthetwosurveysmergedintoaseamlessDEMproduct.Tide gaugesmaintained by theUKNational Tide and Sea Level Facility exist at Avonmouth in themiddleof thestudydomainandatHinckleyPoint justbeyondtheseaward limit. Thenormal tidallimit is atGloucesterwhereaweirprevents the furtherupstreampropagationof the tide. Severalriver gauging stations exist at Gloucester and above. The storm surge climate is relatively wellcharacterizedbythetidegauges,but thequalityof therivergaugeswillneedtobeassessedand ifnecessaryratingcurveswillneedtoberevised(ordetailed2Dhydrodynamicmodelsconstructedforthegaugesitestoextrapolatetheratingcurves).Oneof theprimarypre-launchneeds for theRiver SevernCal/Val site is a calibrated2Dhydraulicmodel (e.g. TELEMAC) to be able to test SWOT data products in both the pre- and post-launchperiods.s

7.2.4.1.4 Pre-launchSiteCharacterizationThefollowingtaskswillneedtobeundertakenpre-launch:

● Installation of a dense network of geodetically leveled pressure transducers every 1-2kmalong the reach, including on both banks where the channel becomes wide enough forsignificantcross-channelelevationdifferencestoform.

● GPS boat surveys of water surface elevation to determine long profile water slopes (andparticularlyslopecurvature)atdifferenttidalstates.

● Quality assessment and, if necessary, revision of discharge rating curves for river gaugingsites.

● ADCPmeasurementsofdischargeandvelocityatapproximately~10crosssectionsalongthereachprofile.

● Lowwaterlidarandhighwatersidescansonarsurveys.● Airphotosurvey.● Development,calibrationandvalidationofa2Dmodelofthesite.

Thesearenecessary tobothgain insights intoprocessesat thesite,butalso to testandrefinesafesamplingproceduresinsuchahighlydynamic,andthereforedangerous,environment.

7.2.4.1.5 Post-launchCal/ValActivitiesDuring the 1-day fast repeat cycle the River Severn Cal/Val site will be the focus of intensivemeasurements to help validate SWOTmeasurements of water-surface elevation, slope, inundationextentanddischarge.Fieldmeasurementswillbesimilartothetechniquesutilizedinthepre-launchcharacterization phase, including boat-based measurements of water-surface elevation, slope, anddischarge, as well as deployment of pressure transducers tomeasure temporal changes inwater-surfaceelevationandslope. AnumberofsetsofairphotoswillalsobecapturedtovalidateSWOTmeasurements of inundation extent. Once developed, the 2Dmodelwill be used to diagnose andtrouble-shoot any issues with the SWOT estimates for water-surface elevation, slope, inundationextentanddischarge.



139

7.2.4.2 ConnecticutRiverTidalCal/ValSite(USProjectSite)

7.2.4.3 SiteDescriptionTheConnecticutRiverTidalCal/ValsiteextendsfromtheUSGSgagingstationatMiddleHaddam,CT,upstreamtotheUSGSgagingstationatThompsonville,CT,USA(Figure77).Theriverinthis65-kmstudyreachistidally-influenced,withastrongtidalinfluenceonlow-andmoderate-flowsextendingthrough themiddle of the reachnearHartford, CT. There is extensive existing gage infrastructure,includingthreemainstem(includingoneindex-velocitygageatthedownstreamend)andfourmajortributary gaging sites operated by theUSGS. All of the Cal/Val site iswithin the SWOT1-day FastRepeatOrbit thoughtheriverdownstreamof theCal/Valsite isunder the1-day fastrepeatorbit’snadirlocation(e.g.notcoveredbyKaRINbutcoveredbythealtimeter).Thoughtherearesmallflowregulationreservoirsintheupperwatershed,thehydrographoftheriveristypicaloftemperaterain-andsnow-fallriversystems,withdistinctandlargeregularpeaksandrecessionlimbsfromOctoberto June and a low flow period in the summer from July to September. In addition, parts of theConnecticutfreezeduringthewinter,withbreakupoccurringinMarchorApril.AsecondaryreachoftheConnecticutRiverupstreamofHaddam,CT,servesastheConnecticutRiverCal/Valsite(notidalinfluence)–seeSect.7.2.1.4.

Figure77.Mapof theConnecticutRiverTidalCal/Val site,with thereachshown inred, theSWOTone-day fastrepeatorbitswathextentsofKaRINshownbytheyellowlines,thenadiraltimetershownwiththeredline,andUSGSgagesshownbythewhiteencircledmarkers.

7.2.4.4 SiteGoalsThe goals of the Connecticut River Tidal Cal/Val are to validate SWOT’s ability to measure orcharacterize water-surface elevation, slope, inundation extent, and discharge as well as validateSWOT’s layover-, ice- and rain-flags. In addition to the Severn River, U.K., the tidal reach of theConnecticut River is the only tidally-influenced Cal/Val site and has amoremoderate tidal action



140

thandoestheSevern.Assuch, itwillplayanimportantroleinvalidatingSWOTmeasurementsanddataproductsfornear-shorecoastalareas.

7.2.4.5 SiteInstrumentationTheConnecticutRiverTidal Cal/Val site has threemainstemand fourmajor tributary gaging sitesoperatedbytheUSGS(Figure77).Duetothestrongtidalinfluenceondischargeandstage,thelowestsite,ConnecticutRiveratMiddleHaddam, isan IndexVelocitygagestationwithapermanentside-lookingADVM(Figure78).AtHartford,theriverisstillstronglytidal,andassuch,theHartfordgagehasawaterheightrecorderonly.TheuppermainstemgageatThompsonville,CT,usesagageheightrecorderwith rating curve to compute discharge and is entirely riverine dominatedwith no tidalinfluence.There are twogagesdownstreamandoutsideof theCal/Val site but these gages recordgage height only and arewithin the nadir gap of the SWOT footprint during SWOT’s one-day fastrepeatorbit.



141

Figure78.DifferencesinriverineandtidaldischargemeasurementsfortheConnecticutRiverTidalCal/ValsiteforOctober14-21, 2015. TheConnecticutRiver atMiddleHaddamUSGSgage is at the lowest boundary of theCal/Val site and is stronglyinfluencedbytidaleffects(toppanel);theConnecticutRiveratThompsonvilleUSGSgagehasnotidalinfluenceandislocatedattheupperboundaryoftheCal/Valsite(bottompanel).

TherearetworecentaeriallidardatasetscoveringtheConnecticutRiver:oneflownbyNRCSin2010(available), and another flown in 2014byUSGS followingHurricane Sandy (in process). The 2014aeriallidarcurrentlyisbeingprocessedandisnotyetpubliclyavailablebutitisexpectedtobecomeavailablein2016.Thereareseveral1Dand2DmodelsavailablefortheConnecticutRiver,thoughthe2Dmodelsdonotextendintothetidally-influencedreach.

7.2.4.6 Pre-launchSiteCharacterizationThe pre-launch site characterization of the Connecticut River Tidal Cal/Val site include a shortAirSWOTcampaignconsistingoftwo-dayswithmultipleAirSWOTpassesduringvarioustidalstagesscheduledfor2018.Concurrently,agroundcampaignwillhaveinstalledandlevelledapproximately30pressuretransducers,aswellascollectingday-of-flight longitudinalwater-surfaceelevationanddischargemeasurements.Ifnotperformedpreviously,GNSS-levelingoftheexistingUSGSgageswillberequiredpre-launch.SWOTflaggingwillbeevaluatedusingexistinghigh-resolutionlidarDEMsforlayoverflags,satellite-observationsoficeandsnowforiceflags,andlocalradarandweatherstationsforrain flags. Inthemonths immediatelypreceding launch,approximately30pressuretransducerswillbeinstalledfortheone-dayfastrepeatorbit.

7.2.4.7 Post-launchCal/ValActivitiesThepost-launchCal/ValactivitiesfortheConnecticutRiverTidalsiteincludetheinstalledpressuretransducermeasurementsduringtheone-dayfastrepeatorbit,aswellastwoone-weekcampaignsforboat-basedlongitudinalwater-surfaceelevationanddischargemeasurements.

7.2.5 GlobalPlanforTier2Cal/ValSites(U.S./FranceJointProject)TheglobalnetworkofnumerousTier2Cal/Valsiteswillbuilduponexistinggagingstationnetworksin member countries by converting existing stage recording data into high-accuracy real-worldsurfacewaterelevations.Forexample,theUSGSstreamandlakegagingstationnetworkintheUSAisfreelyavailableandeasilyaccessible,andrepresentsapproximately9,900stationsreportingstagesand/or discharges at 15- to 60-minute intervals. Themajority of these stations report stageswitheitherlowabsoluteelevationaccuracies(e.g.+/-5m)thoughtherelativevariationsinheightwillbemuchmoreaccurate,orthereportedstageis inanarbitraryelevationframe,nottiedtoreal-worldcoordinates.TobeabletousethesesitesforSWOTCal/Valpurposes,itwouldbemostusefultohavethesestationelevationstiedintoreal-worldcoordinates,andthatistheprimarytechniqueproposedhere.A typical Tier 2 Cal/Val sitewill consist of an existing stage and/or discharge gaging station,withhourlyormorefrequentdatarecording,combinedwithonehigh-accuracyGNSSmeasurementofthestageatagiven time toconvert thestagedata intoreal-worldwatersurfaceelevations.TheTier2Cal/Val sites have been broken into River and Lake sites in the sections below, though theinfrastructureofthetwoareverysimilar.TheadvantageoftheTier2sitesishavingalargenumberofsitescoveringthefullSWOTfootprintswathandlargeareacoverageacrossthecontinentstohelpwith Cal/Val of SWOT measurements and data products. The disadvantage of these types ofminimally-instrumentedTier2sitesisthatdiagnosingerrorsfromSWOTbecomemoredifficultthan



142

in Tier 1 sites. In addition, achieving a large fraction of global coveragewill depend uponwillinginternationalcooperation.

Figure79.MapofpotentialglobalTier2RiverandLakeCal/Valsites.WhitelinesshowtheSWOTone-dayFastRepeatOrbit.CountriesshowninredareeitherSWOTprojectpartnersorunofficially-confirmedparticipants,countriesshowninyellowareWorldMeteorological Organizationmembers and are potential partners for Tier 2 sites. Already identified potential Tier 2Cal/Valsitesareshownbyreddots(approximately10,500sites),withthosesitesundertheSWOTone-dayFastRepeatOrbitshowninyellow(approximately1,200sites).

7.2.5.1 Tier2RiverSites(U.S./FranceJointProject)

7.2.5.1.1 SiteDescriptionEachTier2RiverCal/Valsitewillincludeastagerecorder,whichmayormaynothaveanassociatedrating curve to computedischarge (though this ispreferred).Eachof these stage recorderswillbesurveyed to high-level GNSS accuracy, with survey-grade being possible in most cases, by eitherSWOT Team Members or by the host country representatives with the corresponding GNSSmeasurementbeingrelatedtothestagerecordermeasurementspatiallyandtemporally.

Establishing several hundred of these Tier 2 River Cal/Val siteswith near global coveragewouldbeanidealsituation.Realistically,however,afewhundredsitesismorelikely,thougheffortshavebeenstartedtoinitiatediscussionswithglobalmeteorologicalorganizations,suchastheWorldMeteorological Organization, to help gain participation of countries interested in SWOT data. Inaddition,theUSGSInternationalProgramsOffice,whichcommonlyinterfacesoninternationalwaterissueswith theUSDepartmentofState, the InternationalMonetaryFund,and theWorldBank,hasofferedassistanceingainingaccesstothewaterrepresentativesfromhostcountries.

7.2.5.1.2 SiteGoalsThegoalsof theTier2RiverCal/Valsitesare tovalidate theabsoluteheight,randomheighterror,layover flagging, range drift, and roll/phase drift. Validation of some of these errors will beparticularlyimportanttoevaluatewhenthemeasurementsarefarfromtheocean,wherecalibrationusingthealtimeterisnotpossible.Inaddition,forsomeoftheTier2siteslocatednearmeteorologicstations, itmaybepossibletoevaluatewet-tropodelayerrorsandpotentiallytherain-andice-flagdataproducts.



143

7.2.5.1.3 SiteInstrumentationAt each of the Tier 2 Cal/Val sites, it will be determined that the stage recorder is of sufficientaccuracy for measuring water-surface elevations, such that the recorder meets or exceeds USGSgaging standards, and that the data are recorded at hourly or more frequent time intervals. Inaddition,siteswithweb-orreal-timestreamingdataavailabilitywillbegivenpreference,thoughinsomecountries,thesetypesofgagesarenotusedanddataisrecordedofflinetobepublishedlater.There are some sites that will be given preference due to their location under SWOT cross-overpointsduringeithertheone-dayFastRepeatOrbitortheScienceOrbit.

7.2.5.1.4 Pre-launchSiteCharacterizationThetasksforpre-launchcharacterizationoftheCal/Valsiteswillinclude:

- Interfacingwithcountryrepresentativestodeterminewillingnesstoparticipate.- Selectionofappropriategages,giventhecriteriaabove.- High-accuracyGNSSmeasurementsatthelocationofthestagerecordinggageandtemporally

tiedtothestagerecord.

7.2.5.1.5 Post-launchCal/ValActivitiesPost-launchCal/Val activities are primarily related to data processing and synthesis,with perhapssomeminimalfieldworktocheckanydiscrepanciesinthegagingdata.

7.2.5.2 Tier2LakeSites(U.S./FranceJointProject)

7.2.5.2.1 SiteDescriptionEachTier2LakeCal/Valsitewillincludeastagerecorder,whichismostlikelylocatednearthelakeshore.Eachofthesestagerecorderswillbesurveyedtohigh-levelGNSSaccuracy,withsurvey-gradebeingpossibleinmostcases,byeitherSWOTTeamMembersorbythehostcountryrepresentatives.ThecorrespondingGNSSmeasurementwillberelated to thestagerecordermeasurementspatiallyand temporally. For the Lake Cal/Val sites, some associated fraction of lake area will need to beidentified as consistent with the stage recorder because the lake point measurement will not beuniversallyapplicabletotheentirelake.

7.2.5.2.2 SiteGoalsThegoalsof theTier2LakeCal/Val sitesare tovalidate theabsoluteheight, randomheighterror,layover flagging, range drift, and roll/phase drift. Validation of some of these errors will beparticularlyimportanttoevaluatewhenthemeasurementsarefarfromtheocean,wherecalibrationusingthealtimeterisnotpossible.Inaddition,forsomeoftheTier2siteslocatednearmeteorologicstations, itmaybepossibletoevaluatewet-tropodelayerrorsandpotentiallytherain-andice-flagdataproducts.

7.2.5.2.3 SiteInstrumentationAt each of the Tier 2 Cal/Val sites, it will be determined that the stage recorder is of sufficientaccuracy for measuring water-surface elevations, such that the recorder meets or exceeds USGSgaging standards, and that the data are recorded at hourly or more frequent time intervals. Inaddition, sites with web- or real-time data availability will be given preference, though in somecountries,thesetypesofgagesarenotusedanddataisrecordedofflinetobepublishedlater.

7.2.5.2.4 Pre-launchSiteCharacterizationThetasksforpre-launchcharacterizationoftheCal/Valsiteswillinclude:

- Interfacingwithcountryrepresentativestodeterminewillingnesstoparticipate.



144

- Selectionofappropriategages,giventhecriteriaabove.- High-accuracyGNSSmeasurementsatthelocationofthestagerecordinggageandtemporally

tiedtothestagerecord.

7.2.5.2.5 Post-launchCal/ValActivitiesPost-launchCal/Val activities are primarily related to data processing and synthesis,with perhapssomeminimalfieldworktocheckanydiscrepanciesinthegagingdata.

7.2.6 CornerReflector/TransponderCalibrationSites

7.2.6.1 Oklahoma/KansasSites(USProjectSite)

7.2.6.1.1 SiteDescriptionThissiteiscomprisedofanarrayofcornerreflectorslocatedatthefast-samplingcrossoverdiamondoverOklahoma,Texas,andKansas.Thecornerreflectorswillnominallybearrangedinaneast-westline to span the swaths of the ascending anddescending orbits. Specific locations for each cornerreflectorwillbechosenbasedonthefollowingcriteria:

● Accesstothesiteandlogisticaleaseofinstallation,maintenance,andremoval● Stabilityof the reflector, includingavoidingdisturbances in targetpositionorattitude from

humans,livestock,wildlife,vegetationgrowth,weather,etc.● Flatnessof the surrounding terrain inorder to simplify the interpretationof thedata. The

surroundingsmust at least be flat enough to provide a clear view of the sky for good GPStrackingwhensurveyingthepositionsofthereflectors.

● Avoidanceoffeaturesthatmaybebrightenoughtocontaminatethecornerreflectorechoes● Avoidance of overhead vegetation thatmay attenuate theRF signal (vegetationwould also

makeaccessandmaintenancemoredifficult)● Evenspacingofreflectorsinthecross-trackdirection● Minimalbutslightstaggeringofthereflectorsinthealong-trackdirectioninordertoisolate

thetargetimpulseresponseswhileensuringthatthetargetsareimagedoverashortenoughperiodoftimethatvariationsininstrumentstateareminimized.

● Abilitytoimagereflectorsinbothascendinganddescendingpassesinordertomaximizethebenefitofeachreflector

Sevencornerreflectorswillbearrangedacrosseachof theKaRInswaths(14total),givingacross-track spacing of approximately 7 km, commensurate with the sampling required for the scienceobjectiveofresolving15kmwavelengths.



145

Figure80.UScornerreflectorsiteattheOklahomacrossoverofthe1-dayorbit.

7.2.6.1.2 SiteGoalsArrays of corner reflectors are commonly used for the calibration and low-level validation of SARsystems.Theywillbeusedtoprovidecoarse,initialestimatesfortheabsoluteanddifferentialrangedelays, to validate point target responses and geolocation accuracy, and secondary phase screenvalidation.

7.2.6.1.3 SiteInstrumentationThelocationsofthecornerreflectorswillbesurveyedwithGPSinstrumentstocentimetricaccuracypriortotheinstrumentcheckoutphaseofthemission.Ifthereisevidencethatthecornerreflectorshavebeendisturbed(inconsistenciesinheights,horizontalposition,orreflectivityinSWOTdata,orknowledgeofextremeweather,etc.),maintenanceoftheaffectedreflectorswillbedoneasnecessaryand the reflectors resurveyed. The reflectors will be surveyed once again before they are takendown.



146

7.2.6.1.4 Pre-launchSiteCharacterizationPre-launchactivityconsistsofthefollowing:

● Identifyingthespecific locationsofallcornerreflectors,coordinatingwith local landownersororganizationstosecureaccesspermissions,etc.

● Designing and building the corner reflectors, or identifying and securing existing cornerreflectors. Thisincludestheelectricaldesign(size,triangulartrihedralvs.othershape,etc.)aswellasmechanicalmountingandsurveyingprovisions.

● Development of the detailed plan for who will deploy and survey the reflectors, whendeployment will occur, how as-needed maintenance will be pursued, and how/when thereflectorswillbetakendown.

● Deploymentof the corner reflectors, including surveying. Thismayalsooccurafter launchbutpriortoKaRIncheckout.

● ImagingthecornerreflectorswithAirSWOTwouldprovideausefulvalidationoftheirsetupforriskreductionpurposesbutisnotstrictlyrequired.

7.2.6.1.5 Post-launchCal/ValActivitiesPost-launchactivityconsistsofthefollowing:

● Deployment of the corner reflectors if this was not done before launch (deploymentmustoccurpriortoKaRIncheckout).Waitinguntilaslateaspossiblemayminimizethelikelihoodofthereflectorsbeingdisturbed.

● MaintenanceofthecornerreflectorsintheeventofextremeweatherorinconsistenciesintheKaRIndata.

● Removalofthecornerreflectors,includingasurveyofthereflectorpositions.

7.2.6.2 AustraliaSites(USProjectSite)

7.2.6.2.1 SiteDescriptionTheAustraliancornerreflectorsite isanalogous to theUScornerreflectorsitedescribedabove. Asecond site is needed in order todetect anddiagnose effects that have a latitudedependence (theSRTM processor had a software bug that was diagnosed through the use of a secondary cornerreflectorsiteinAustralia)andtoprovidecorroboratingabackupsitetotheprimaryUSsite.Thissitewill be located at the fast-sampling crossover diamondnear theBass Strait inAustralia, providingsynergisticvalidationwithanyBassStraitdataorexperiments.Thenumberofreflectorsforthissitemaybereducedasadescopeoption,thoughdoingsoreducestherobustnessoftheCal/Valprogramtounexpectedproblemsencounteredafterlaunch.



147

Figure81.Australiancornerreflectorsiteatthe1-daycrossoverinsoutheasternAustralia.

7.2.6.2.2 SiteGoalsSameasfortheUScornerreflectorsite,butatadifferentlatitude.

7.2.6.2.3 SiteInstrumentationSameasfortheUScornerreflectorsite.Simplificationmaybeadescopeoption.

7.2.6.2.4 Pre-launchSiteCharacterizationSameasfortheUScornerreflectorsite.Simplificationmaybeadescopeoption.

7.2.6.2.5 Post-launchCal/ValActivitiesSameasfortheUScornerreflectorsite.Simplificationmaybeadescopeoption.



148

8 REFERENCESAntoine,D.,P.Guevel,J.-F.Desté,G.Bécu,F.Louis,A.J.ScottandP.Bardey,2008.The«BOUSSOLE»

buoy –Anew transparent-to-swell tautmooringdedicated tomarine optics : design, tests andperformanceatsea,JournalofAtmosphericandOceanicTechnology,25,968-989.

Bonnefond,P.,O.Laurain,P.Exertier,A.Guillot,P.Picot,M.Cancet,F.Lyard.SARAL/AltiKaabsolutecalibrationfromthemulti-missionCorsicafacilities,SARAL/AltiKaspecialissue,MarineGeodesy,2015.Inpress

Bonnefond,P.,P.Exertier,O.Laurain,P.ThibautandF.Mercier,GPS-basedsealevelmeasurementstohelp the characterization of land contamination in coastal areas. 2013. Advances in SpaceResearch,Volume51,Issue8,Pages1383-1399,ISSN0273-1177,10.1016/j.asr.2012.07.007.

Bonnefond,P.,B.HainesandC.Watson.2011.InSituCalibrationandValidation:ALinkfromCoastaltoOpen-oceanaltimetry,chapter11inCoastalAltimetry,pp259-296,editedbyS.Vignudelli,A.Kostianoy,P.Cipollini,J.Benveniste,Springer,ISBN:978-3-642-12795-3.

Bonnefond,P., P.Exertier,O. Laurain,Y.Menard,A.Orsoni,G. Jan, andE. Jeansou.2003a.AbsoluteCalibration of Jason-1 and TOPEX/Poseidon Altimeters in Corsica. Special Issue on Jason-1Calibration/Validation,Part1.MarGeod.26(3-4):261-284. doi:10.1080/714044521.

Bonnefond,P.,P.Exertier,O.Laurain,Y.Menard,A.Orsoni,E.Jeansou,B.Haines,D.Kubitschek,andG.Born. 2003b. Leveling Sea Surface using a GPS catamaran. Special Issue on Jason-1Calibration/Validation,Part1.MarGeod.26(3-4):319-334.doi:10.1080/714044524.

Cancet,M.,S.Bijac,J.Chimot,P.Bonnefond,E.Jeansou,O.Laurain,F.Lyard,E.Bronner,P.Féménias.2013.Regionalinsituvalidationofsatellitealtimeters:calibrationandcross-calibrationresultsatthe Corsican sites. Adv. Space Res. 51(8): 1400-1417. ISSN 0273-1177. doi:10.1016/j.asr.2012.06.017.

M. Durand, L. L. Fu, D. P. Lettenmaier, D. E. Alsdorf, E. Rodriguez, and D. Esteban-Fernandez. Thesurface water and ocean topographymission: Observing terrestrial surface water and oceanicsubme-soscaleeddies.PROCEEDINGSOFTHEIEEE,98(5):766–779,May2010.

L.Fu,D.Alsdorf,R.Morrow,E.Rodrıguez,andN.Mognard,editors.SWOT:TheSurfaceWaterandOcean TopographyMission:Wide-Swath AltimetricMeasurement ofWater Elevation on Earth.NumberJPL-Publication12-05.JetPropulsionLaboratory,Pasadena,California,2012.

Fu,L.andRodriguez,E.(2004).High-resolutionmeasurementofoceansurfacetopographybyradarinterferometry for oceanographic and geophysical applications. In The State of the Planet:FrontiersandChallengesinGeophysics,volume19ofIUGGGeophysicalMonograph,pages209–224.InternationalUnionofGeodesyandGeophysicsandtheAmericanGeophysicalUnion.

JanG,MénardY,FaillotM,LyardF,JeansouE,BonnefondP(2004)OffshoreAbsoluteCalibrationofSpace Borne Radar Altimeters. In: Special Issue on Jason-1 Calibration/Validation, Part 3. MarGeod27(No3-4):615-629

Haines, B J, D. Dong, G H. Born, S K. Gill. 2003. The Harvest experiment: Monitoring Jason-1 andTOPEX/POSEIDON from a California offshore platform. In: Special Issue on Jason-1Calibration/Validation,Part1.MarGeod.26(3-4):239-259.doi:10.1080/714044520.

W.K.Melville,L.Lenain,D.R.Cayan,M.Kahru,J.P.Kleissl,P.Linden,andN.M.Statom.Themodularaerialsensingsystem.JournalofAtmosphericandOceanicTechnology,(2016),2016.

Millet F., Warnick K and Arnold D. (2005) Electromagnetic bias at off nadir incidence angles,JGeophys.Res.110,C09017,doi:10.1029/2004JC002704E.Rodríguez.SurfaceWaterandOceanTopographyMission (SWOT)ScienceRequirementsDocument.Technical report, JetPropulsionLaboratory,4800OakGroveDr.,Pasadena,CA91109,March2015.

N.Tran,D.Vandemark,S.Labroue,H.Feng,B.Chapron,H.L.Tolman,J.Lambin,N.Picot(2010)Seastate bias in altimeter sea level estimates determined by combining wavemodel and satellitedata,JGeophys.Res.115,C03020,doi:10.1029/2009JC005534



149

J.Wang,L.-L.Fu,B.Qiu,D.Menemenlis, J.T.Farrar,Y.Chao,A.Thompson,andM.Flexas(2018)Anobservingsystemsimulationexperiment for thecalibrationandvalidationof theSurfaceWaterandOceanTopographyseasurfaceheightmeasurementusingin-situplatforms.J.Atmos.OceanicTechnol.,doi:10.1175/JTECH-D-17-0076.1.



150

AppendixA. SPECIFICRESOURCESREQUIREDANDTIMINGOFACTIVITIESFORTHEINLANDHYDROLOGYCAL/VALPROGRAM

ThefollowingappendixdescribesthetimingofactivitiesandthespecificresourcesrequiredforthecomponentsoftheInlandHydrologyCal/ValprogramofSWOT. B.1Absoluteinlandsurfacewaterheightvalidation-timingofactivitiesandresourcesrequired: Pre-launch:

● 2016. Staff time to analyze past AirSWOT data fromWillamette, Sacramento, Tanana andYukon Flats to understand cross-channel height variability, curvature of river slopes. Stafftime, travel andequipment to conductpressure transducer testsofwindsetupon lakes10kmacrossandsmaller.

● 2016 or 2017. Staff time, travel and equipment to conduct GPS campaign on LowerMississippitounderstandcross-channelheights&curvatureeffects.

● Spring2019. Staff time,AirSWOToverflights, travelandequipment toperformdryrun forriverCal/ValontheWillamette, including installationofpressuretransducers. Lidar flightsovertheWillamettealsorequired. Agreebythistimeonacceptableequipment,techniques,measurementmethods,etc.

● 2018-launch: Staff time, travel and equipment to obtain all permissions for installation ofCal/Valequipment&conductmeasurementsatallCal/Valsites.WorkwithUSGStosurveyinallgaugestobeusedinSWOTheight&slopevalidation.

Post-launch:

● Stafftime,travelandequipmenttodeployfieldteamstotheTier1sitestocheckdatacaptureandcollectGPSdrifterdataofwatersurfaceelevation.

Tier2Cal/Valsites-timingandresourcesrequired Pre-launch:

● Stafftimetosetupinternationalagreementsandagreetodatatransferformats.Stafftimetocreatethedatabase.Paymentstobuydataifnecessary.

Post-launch: ● Stafftimetomanagedatasharingagreementsandcollatedatabase.

B.2Inundatedsurfaceareavalidation-timingofactivitiesandresourcesrequired: Pre-launch:

● 2016 to 2017. Staff time to analyze past AirSWOTdata fromWillamette, Sacramento, andTananatounderstandabilityofAirSWOTtomeetrequirementsforvalidation.StafftimetotestmethodsforextractinginundationextentsfromlidarDEMsandwatersurfaceelevations.

● Spring2019. Staff time,AirSWOToverflights, travelandequipment toperformdryrun forrivercal/valontheWillamette.LidarflightsovertheWillamettealsorequired.Agreebythistimeonacceptableequipment,techniques,measurementmethods,etc.

● 2018-launch:Stafftimeandbudgettocollectsampleaerialimagesoveraportionofallfieldsitesusingtheproviderslikelytobeusedduringpost-launchcampaigns.

● 2019-launch:StafftimetoassembleandanalyzealluseablelidarDEMSovertier1andtier2sitesthatcanbeusedtoestimateinundationextent.

Post-launch:



151

● AirSWOT flightsor flight time forotherplatform tocollectnecessary imagery. Funding forfield teams to deploy on-the-ground measurements. Staff time to analyze lidar, field, andairbornedata.

B.3Smalllakeinundatedsurfacearea-timingofactivitiesandresourcesrequired: Pre-launch:

● 2016to2017.StafftimetoanalyzepastAirSWOTdatafromYukonFlatstounderstandabilityofAirSWOT tomeet requirements forvalidation. Staff time to testmethods for extractinginundationextentsfromlidarDEMsandwatersurfaceelevations.

● 2018-launch:Stafftimeandbudgettocollectsampleaerialimagesoveraportionofallfieldsitesusingtheproviderslikelytobeusedduringpost-launchcampaigns.

● 2019-launch:StafftimetoassembleandanalyzealluseablelidarDEMSovertier1andtier2sitesthatcanbeusedtoestimateinundationextent.

Post-launch:

● AirSWOTflightsorflighttimeforotherplatformstocollectnecessaryimagery. Fundingforfield teams to deploy on-the-ground measurements. Staff time to analyze lidar, field, andairbornedata.

B.4Lakelakeinundatedsurfaceareavalidation-timingofactivitesandresourcesrequired: Pre-launch

● 2016-2019: Staff time to develop height/inundation extent rating curves for many lakes,globally(tobeledbyFrench).

Post-launch:

● Staff time to download and process satellite imagery coincidentwith SWOT overflights (towithin +/-3 days formost lakes) and to compare SWOT inundation extents against valuesderivedfromratingcurves.

B.5Wetlandinundatedsurfaceareavalidation-timingofactivitiesandresourcesrequired:Pre-launch:

● 2016-2017: Staff time to analyze data collected in 2015 overMississippi Delta and YukonFlats to determine SWOT ability tomeasure inundation extent under vegetation of variousdensitiesandelevations.

● 2017-2018: AirSWOT campaign to either Mississippi Delta or Everglades to characterizeSWOT ability to measure inundation extent under vegetation, contingent on outcome ofanalysis of 2015 campaigns. This is likely tobenecessary inpart becauseAirSWOTswathcoveringSWOTincidenceanglesindetailwasnotcollectedoverMississippiin2016.

Post-launch:

● UAVSARandlidarflightsoverfieldsites.Stafftimetoprocessandanalyzeresultingdataandtodownloadandprocessrelevantsatelliteimageryoverwetlandsites.

B.6SlopeValidation-timingofactivitiesandresourcesrequired:Pre-launch:

● 2016or 2017. Staff time, travel and equipment to conductGPS campaignonMississippi tounderstandcross-channelheights&curvatureeffects.



152

● 2016-2017: Staff timetoanalyzeandcompareslopes fromallmethodsproposedaboveontheSacramento,Willamette,Tanana, andMississippiRivers. Goal is toverify that accuracyrequirementscanbemet.

● Spring2019. Staff time,AirSWOToverflights, travelandequipment toperformdryrun forriver cal/val on theWillamette, including installation of pressure transducer array. Lidarflights over the Willamette also required. Agree by this time on acceptable equipment,techniques,measurementmethods,etc.

● 2018-launch: Staff time, travel and equipment to obtain all permissions for installation ofcal/valequipment&conductmeasurementsatallcal/valsites.WorkwithUSGStosurveyinallgaugestobeusedinSWOTslopevalidation.

● 2019-launch:InstallpressuretransducerarraysinallTier1riversites.Post-launch:

● Stafftime,travelandequipmenttodeployfieldteamstotheTier1sitestocheckdatacaptureandcollectGPSdrifterdataofwatersurfaceslope.AirSWOTflightsoveratleastthreerivers(Willamette,Connecticut,Mississippi)tovalidateslopesoverreachlengths>50km.

D-75724 SWOT Cal Val Plan Initial 20180129u · 29/01/2018 · A cartoon illustrating the SWOT measurement conceptis presented in Figure 1. ... download limitations, the ocean data

Documents