GAMIT/GLOBKforGNSST.A.HerringM.A.FloydR.W.KingMassachusetts InstituteofTechnology, Cambridge,MA,USA
GPSDataProcessing andAnalysiswithGAMIT/GLOBK/TRACKUNAVCOHeadquarters, Boulder, Colorado, USA
19–23June2017
http://web.mit.edu/mfloyd/www/courses/gg/201706_UNAVCO/MaterialfromR.W.King,T.A.Herring,M.A.Floyd(MIT)andS.C.McClusky (nowatANU)
Thepromise…
• WhenGPS,Glonass,Beidou,andGalileoaredeployedandmodernized,therewillbe>100satellitesand12distinct frequenciesavailablefortracking• Obviousadvantagesforkinematicpositioningandatmosphericstudies,butalsoameanstoseparateperiodicsignalsduetoaliasing intheGPSorbits• Fulldeploymentexpectedby2020,butconstellations>20SVsbymid-2018shouldprovidehighlyusefulresults
2017/06/19 GAMIT/GLOBKforGNSS 1
GNSSavailableinGAMIT/GLOBK
• AsofGAMIT/GLOBKrelease10.61(2017-04-01), thefollowingGlobalNavigationSatelliteSystems(GNSSs)maybeprocessed✔ GPS(USA),BeiDou (China),Galileo(Europe),IRNSS(India)✘ GLONASS(Russia),QZSS(Japan)
• GNSScodeisintrialandweneedtesters• Bug-fixedandalpha-testedcodeisintheupdates/sourcedirectoryasgnss_170616.tar.gz,tobeuntarredatthegg levelafter thecurrentincremental_updates.
GNSSdatamustbeprocessedseparately foreachsysteminGAMIT,i.e.onecannotprocessGPSdataandGalileodatasimultaneously
2017/06/19 GAMIT/GLOBKforGNSS 3
Whyprocessseparately?
• Dual-frequencyobservationsarefundamentalforGNSStoremovetheionosphereandareeasilyimplementedunderthecurrentstructureofGAMIT,butprocessingdifferentsystemsacrossmorethantwofrequenciessimultaneouslyrequiresadifferentalgorithmicapproachandwilltakesometimetoimplement
• Solution(h-)filesfrommultiplesystems(aswithmultiplesub-nets)canberigorouslycombinedinGLOBKtoestimatesitecoordinatesandvelocitiesforstaticobservations
• Basedonresearchthusfar,itisnotclearthatjointprocessingwillImproveresultsforthelong-sessionsusedformm-levelmeasurements,thoughthatmaychangeasthesystemsmature(improvedorbitsandknowledgeofinter-systemsignalbiases)
2017/06/19 GAMIT/GLOBKforGNSS 4
ProcessingGNSSinGAMIT/GLOBK
2017/06/19 GAMIT/GLOBKforGNSS 12
Versionawarenessandwarnings
• AmajorchangebetweenGAMIT/GLOBK10.5and10.6wastheformatofmanytables(e.g.dcb.dat,svnav.dat,etc.)toaccommodatecodechangesforGNSS• GAMIT/GLOBK10.61nowbuildsuponthenewfilestructurestodeliverthedataprocessingcapability• Giventhesemajorchanges,manytablesusedinGAMIT/GLOBK10.6and10.61arenot backwardscompatiblewithGAMIT/GLOBK10.5andpriorreleases• Youcannot usemanytablesthatcamewithGAMIT/GLOBK10.5andpriortoprocess(GPS-onlyorGNSS)datausingGAMIT/GLOBK10.6andlater
2017/06/19 GAMIT/GLOBKforGNSS 13
Suggestionsforprocessingstrategies
• IfyouwishtocombinedatafromdifferentGNSS,processeachsysteminaseparateexperimentdirectory,e.g./2017gand/2017eforGPSandGalileo• DownloadtheRINEXfilesinadvancetocheckforavailabilityofGNSSsignals• Forsh_gamit usethe“-gnss”optiontospecifytheGNSS;andCOM1 for“-orbit”(IGSF okforGPS)• Checktheorbit-fitrms filesinthe/igs directorytoassesstheorbitquality• Combinetheresultingh-filesinGLOBKtoproduceasingleresult(timeseriesorvelocities)
2017/06/19 GAMIT/GLOBKforGNSS 14
RINEXfiles• RINEX2,whichisstillbyfarthemostcommonformatofRINEXfile,wasdesignedinanerawhenonlyGPSwasviableforobservation
• SincetheredesignofGPStobroadcastasecondcodeonL2(“L2C”)specificallyforcivilianuse,therestorationofGLONASSandtheintroductionofothernavigationsatellitesystems,RINEX2nolongersufficestotrackallavailableobservations
• Beverycarefulwithhowyoutranslateanduseotherpeople’sRINEX2fileswithL2C(seeBerglundetal.,2010;Blumeetal.,2012;andhttp://kb.unavco.org/kb/article/the-effects-of-l2c-signal-tracking-on-high-precision-carrier-phase-gps-postioning-689.html)
2017/06/19 GAMIT/GLOBKforGNSS 15
GAMIT
• Severalscriptsnowhaveanadditionaloption(“-gnss”)thatsetsthetypeofGNSS• Mostlikelytousedirectly:sh_gamit, sh_get_orbits,sh_sp3fit
• Lesslikelytousedirectly: sh_preproc, sh_bcfit,sh_rxscan, sh_get_times, sh_makexp
• Validargumentsare(onlyoneof)• G (GPS)• R (GLONASS;notyetcodedoravailabletouse)• C (BeiDou-2/COMPASS)• E (Galileo)• J (QZSS;notyetcodedoravailabletouse)• I (IRNSS)
• Thedefaultisstill“G”(GPS)2017/06/19 GAMIT/GLOBKforGNSS 16
GLOBK
• GLOBKshouldworkinthenormalmannerexceptthatyoucannotincludeorbitsintheh-file(useBASELINE inGAMITsestbl.)• SeeGLOBKlectures,includingthoseoncreatingtimeseriesusingglred andcombinationorvelocitysolutionsusingglobk
2017/06/19 GAMIT/GLOBKforGNSS 17
Resultsofinitialtests
2017/06/19 GAMIT/GLOBKforGNSS 19
Systemsprocessed:GPS(L1andL2)Galileo (E1andE5)BeiDou (C2andC7)
5 daysprocessed(2017-121 to2017-125)
SkytracksDay121atSpanishsiteVILLEachcirclecoversa4-hrwindow
2017/06/19 GAMIT/GLOBKforGNSS 20
GPS
Galileo
Beidou
PhaseRMS(for2017-121)
Fromsh_gamit_121[gec].summary (mm):
2017/06/19 GAMIT/GLOBKforGNSS 21
Site GPS Galileo BeiDou
ASCG 7.8 9.1 4.8
BOR1 5.6 5.9 6.7
BRST 8.5 9.7 7.0
BRUX 4.4 5.0 4.7
DJIG 5.9 8.0 17.1
GANP 5.9 6.6 8.4
KIRU 6.8 8.0 10.7
KOUR 10.6 12.6 5.5
Site GPS Galileo BeiDou
LAMP 4.5 6.2 6.2
MAL2 7.7 10.0 12.7
MAS1 7.1 8.1 3.1
NICO 6.1 9.1 22.1
REYK 8.4 11.1 8.3
STJ3 5.3 6.0 2.3
VILL 8.1 9.6 5.1
ALL 7.1 8.6 11.3
Ambiguityresolution
Bestfromsh_gamit_[gec].summary:
2017/06/19 GAMIT/GLOBKforGNSS 22
System Widelane(WL) Narrow lane(NL)
GPS 98% 91%
Galileo 99% 78%
BeiDou 80% 50%
Timeseriesstabilization
From“POSSTAT”linesin.org-file(s)(mm:
2017/06/19 GAMIT/GLOBKforGNSS 23
System Sites E N U
GPS 10 1-2 1–2 6–10
Galileo 9-10 2-3 2–4 11-16
BeiDou 6–9 1–7 3–9 6–18
Exampletimeseries
2017/06/19 GAMIT/GLOBKforGNSS 25
GPS Galileo BeiDou
Positiondifferences(GPSversusGalileo)
2017/06/19 GAMIT/GLOBKforGNSS 39
Initialimpressions
• Galileohasrobustphasetrackingandcould,forsmall regionalnetworkswithgoodsatellitecoverage,nowproducenearlyGPS-qualityresults• TheBeidou constelletion willneedtobefilledoutsignificantly,expectedbymid-2018• BeiDou alsoappearstobepronetosinglecycleslips,resultinginpoordetectionandcleaningoftrackedphase• Thismaybeimprovedbytuningautcln.cmd
2017/06/19 GAMIT/GLOBKforGNSS 40
Summary
• GAMIT/GLOBKisnow(asof10.61)capableofprocessingalmostallGNSSdata,except• GLONASS,whichhasvariablefrequencies and,assuch,requiresaredesignof
thesoftware’sstructureandflow• QZSS,whichisasmall,regional,high-altitudesystemoflittleusebeyondthe
narrowdesignregionoronaglobalscale(sameistrueofIRNSS,althoughthisiscoded)
• GNSSdataareavailablebutfewusersareactuallycollectingorprocessingsuchdata
• Asaresult,globalorbitsarepoorlyconstrainedbygroundstationswithaccuratecoordinatesintheterrestrialreferenceframe• SatelliteorbitalmodelsandantennadesignsarelesswellknownthanGPS• ManyGNSSsotherthanGPSareinasimilar“weak”statetowhereGPSwasin
theearly1990sbeforetheadventoftheIGS• ItisdifficulttopredictatwhattimetheothersystemswillenhanceratherthandegradeGPSresultsbutweshouldseerapidimprovementwiththelaunchofmoresatellitesandtheexpansionofthetrackingnetworkinthenext18months
2017/06/19 GAMIT/GLOBKforGNSS 44
ReferencesBerglund,H.,F.Blume,L.H.Estey,andA.A.Borsa (2010),The
EffectsofL2CSignalTrackingonHigh-PrecisionCarrierPhaseGPSPositioning,AbstractG11B-0640 presentedat2010FallMeeting,AGU,SanFrancisco,Calif.,13-17Dec.
Blume,F.,H.Berglund,andL.Estey (2012),TheEffectsofL2CSignalTrackingonHigh-PrecisionCarrierPhaseGPSPositioning:ImplicationsfortheNextGenerationofGNSSSystems,AbstractG52B-07presentedat2012FallMeeting,AGU,SanFrancisco,Calif.,3-7Dec.[http://acc.igs.org/trf/agu12_blume_l2c.pdf]
http://kb.unavco.org/kb/article/the-effects-of-l2c-signal-tracking-on-high-precision-carrier-phase-gps-postioning-689.html
Montenbruck,O.,R.Schmid,F.Mercier,P.Steigenberger,C.Noll,R.Fatkulin,S.Kogure,andA.S.Ganeshan (2015),Adv.SpaceRes.,56,1015–1029, doi:10.1016/j.asr.2015.06.019.
2017/06/19 GAMIT/GLOBKforGNSS 46