WPC^ pMA„89@DZZ#t0 h^QO*3 t d"~+p+X0\ MmswoIYWI j2ϪhVOC[g;jsYP{ Ѯ96JT'otTfF?hȱ iA ɣJk#udX}yW̸ S9<3.*>nVOkYrA'm̔Ps`E:Y$ vhkmj Iߴ ]F?Ӯy6w gWnގ(qB; Uw{UItiDiV} 'İD%uUo񂜁aKueWPRmvNRͲ\fAbm[5WдnxYg]ߐՃFq/iKc:W4y.Ujwe@Mဟ{[^-nU0 %D 0(Jwr4v m#dNlnpxz| 0F 0NHJRT\^`hjrz|U2zLT\dlt| $,4<DLLT\dlt| D3 B AQe 0D+ 0Do 0Df^ aUBf[a]fqsf9a;ONWYaiqy !)19AIQYYaiqy !)19AIU,Q}N'/7?GOW_gow'/7?GOW_gowEN  #+3;CKS[cks{  #+3;CKS[cks{{N  %-5=EMU]emu}  %-5=EMU]emu}V N^ ` h p x                       ( 0 8 @ H P X ` ` h p x                       ( 0 8 @ H P P X N` b j r z                      " * 2 : B J R Z b b j r z                      " * 2 : B J R R Z Nb d l t |                   $,4<DLT\ddlt| $,4<DLTTT\Ndfnv~&.6>FNV^ffnv~&.6>FNVV^Nfhpx (08@HPX`hhpx (08@HPXU,`N&.6>FNV^fnv~&.6>FNV^fnv~N (08@HPX`hpx (08@HPX`hpxaN&..6>FNV^fnv~&..6>FNV^fnv~lNtv~&.6>FNV^fnvv~&.6>FNV^fn4N<>FNV^fnv~&.6>>FNV^fnv~&.6N     & . 6 > F N V ^ f n v ~                  !!!!!&!.!6!>!F!N!V!^!f!n!v!~!!!!!!!!!!!!!!!!!!!!N"""" "("0"8"@"H"P"X"`"h"p"x""""""""""""""""""##### #(#0#8#@#H#P#X#`#h#p#x###################$N$ $$$"$*$2$:$B$J$R$Z$b$j$r$z$$$$$$$$$$$$$$$$$$% % %%%"%*%2%:%B%J%R%Z%b%j%r%z%%%%%%%%%%%%%%%%%%%&N & &&&$&,&4&<&D&L&T&\&d&l&t&|&&&&&&&&&&&&&&&&&&' ' '''$','4'<'D'L'T'\'d'l't'|'''''''''''''''''''(N ((((&(.(6(>(F(N(V(^(f(n(v(~(((((((((((((((((()))))&).)6)>)F)N)V)^)f)n)v)~)))))))))))))))))))*N*** *(*0*8*@*H*P*X*`*h*p*x******************+++++ +(+0+8+@+H+P+X+`+h+p+x++++++++++++++++++,,,N,,,",*,2,:,B,J,R,Z,b,j,r,z,,,,,,,,,,,,,,,,,,- ----"-*-2-:-B-J-R-Z-b-j-r-z------------------.. .N...$.,.4.<.D.L.T.\.d.l.t.|................../ ////$/,/4/0F0N0V0^0f0n0v0~00000000000000000011111&1.161>1F1N1V1^1f1n1v1~111111111111111111222N22 2(20282@2H2P2X2`2h2p2x22222222222222222233333 3(30383@3H3P3X3`3h3p3x3333333333333333334444N44"4*424:4B4J4R4Z4b4j4r4z4444444444444444445 5555"5*525:5B5J5R5Z5b5j5r5z5555555555555555556 6 66N66$6,646<6D6L6T6\6d6l6t6|6666666666666666667 7777$7,747<7D7L7T7\7d7l7t7|7777777777777777778 8 88N88&8.868>8F8N8V8^8f8n8v8~88888888888888888899999&9.969>9F9N9V9^9f9n9v9~999999999999999999::::N: :(:0:8:@:H:P:X:`:h:p:x::::::::::::::::::;;;; ; ;(;0;8;@;H;P;X;`;h;p;x;;;;;;;;;;;;;;;;;;<<<<<N <"<*<2<:<B<J<R<Z<b<j<r<z<<<<<<<<<<<<<<<<<<= ==="="=*=2=:=B=J=R=Z=b=j=r=z==================> >>>>N">$>,>4><>D>L>T>\>d>l>t>|>>>>>>>>>>>>>>>>>>? ???$?$?,?4?@F@N@V@^@f@n@v@~@@@@@@@@@@@@@@@@@@AAAA&A&A.A6A>AFANAVA^AfAnAvA~AAAAAAAAAAAAAAAAAABBBBBBN&B(B0B8B@BHBPBXB`BhBpBxBBBBBBBBBBBBBBBBBBCCCC C(C(C0C8C@CHCPCXC`ChCpCxCCCCCCCCCCCCCCCCCCDDDDD DN(D*D2D:DBDJDRDZDbDjDrDzDDDDDDDDDDDDDDDDDDE EEE"E*E*E2E:EBEJEREZEbEjErEzEEEEEEEEEEEEEEEEEEF FFFF"FN*F,F4FHFHNHVH^HfHnHvH~HHHHHHHHHHHHHHHHHHIIII&I.I.I6I>IFINIVI^IfInIvI~IIIIIIIIIIIIIIIIIIJJJJJ&JN.J0J8J@JHJPJXJ`JhJpJxJJJJJJJJJJJJJJJJJJKKKK K(K0K0K8K@KHKPKXK`KhKpKxKKKKKKKKKKKKKKKKKKLLLL L L(LN0L2L:LBLJLRLZLbLjLrLzLLLLLLLLLLLLLLLLLLM MMM"M*M2M2M:MBMJMRMZMbMjMrMzMMMMMMMMMMMMMMMMMMN NNN"N"N*NN2N4NPFPNPVP^PfPnPvP~PPPPPPPPPPPPPPPPPPQQQQ&Q.Q6Q6Q>QFQNQVQ^QfQnQvQ~QQQQQQQQQQQQQQQQQQRRRR&R&R.RN6R8R@RHRPRXR`RhRpRxRRRRRRRRRRRRRRRRRRSSSS S(S0S8S8S@SHSPSXS`ShSpSxSSSSSSSSSSSSSSSSSSTTTT T(T(T0TN8T:TBTJTRTZTbTjTrTzTTTTTTTTTTTTTTTTTTU UUU"U*U2U:U:UBUJURUZUbUjUrUzUUUUUUUUUUUUUUUUUUV VVV"V*V*V2VN:VXFXNXVX^XfXnXvX~XXXXXXXXXXXXXXXXXXYYYY&Y.Y6Y>Y>YFYNYVY^YfYnYvY~YYYYYYYYYYYYYYYYYYZZZZ&Z.Z.Z6ZN>Z@ZHZPZXZ`ZhZpZxZZZZZZZZZZZZZZZZZZ[[[[ [([0[8[@[@[H[P[X[`[h[p[x[[[[[[[[[[[[[[[[[[\\\\ \(\0\0\8\N@\B\J\R\Z\b\j\r\z\\\\\\\\\\\\\\\\\\] ]]]"]*]2]:]B]B]J]R]Z]b]j]r]z]]]]]]]]]]]]]]]]]]^ ^^^"^*^2^2^:^NB^D^L^T^\^d^l^t^|^^^^^^^^^^^^^^^^^^_ ___$_,_4_<_D_D_L_T_\_d_l_t_|__________________` ```$`,`4`4`<`ND`F`N`V`^`f`n`v`~``````````````````aaaa&a.a6a>aFaFaNaVa^afanava~aaaaaaaaaaaaaaaaaabbbb&b.b6b6b>bNFbHbPbXb`bhbpbxbbbbbbbbbbbbbbbbbbcccc c(c0c8c@cHcHcPcXc`chcpcxccccccccccccccccccdddd d(d0d8d8d@dNHdJdRdZdbdjdrdzdddddddddddddddddde eee"e*e2e:eBeJeJeReZebejerezeeeeeeeeeeeeeeeeeef fff"f*f2f:f:fBfNJfLfTf\fdflftf|ffffffffffffffffffg ggg$g,g4giFiNiNiVi^ifinivi~iiiiiiiiiiiiiiiiiijjjj&j.j6j>j>jFjNNjPjXj`jhjpjxjjjjjjjjjjjjjjjjjjkkkk k(k0k8k@kHkPkPkXk`khkpkxkkkkkkkkkkkkkkkkkkllll l(l0l8l@l@lHlNPlRlZlbljlrlzllllllllllllllllllm mmm"m*m2m:mBmJmRmRmZmbmjmrmzmmmmmmmmmmmmmmmmmmn nnn"n*n2n:nBnBnJnNRnTn\ndnlntn|nnnnnnnnnnnnnnnnnno ooo$o,o4oqFqNqVqVq^qfqnqvq~qqqqqqqqqqqqqqqqqqrrrr&r.r6r>rFrFrNrNVrXr`rhrprxrrrrrrrrrrrrrrrrrrssss s(s0s8s@sHsPsXsXs`shspsxsssssssssssssssssstttt t(t0t8t@tHtHtPtNXtZtbtjtrtzttttttttttttttttttu uuu"u*u2u:uBuJuRuZuZubujuruzuuuuuuuuuuuuuuuuuuv vvv"v*v2v:vBvJvJvRvNZv\vdvlvtv|vvvvvvvvvvvvvvvvvvw www$w,w4wyFyNyVy^y^yfynyvy~yyyyyyyyyyyyyyyyyyzzzz&z.z6z>zFzNzNzVzN^z`zhzpzxzzzzzzzzzzzzzzzzzz{{{{ {({0{8{@{H{P{X{`{`{h{p{x{{{{{{{{{{{{{{{{{{|||| |(|0|8|@|H|P|P|X|N`|b|j|r|z||||||||||||||||||} }}}"}*}2}:}B}J}R}Z}b}b}j}r}z}}}}}}}}}}}}}}}}}}~ ~~~"~*~2~:~B~J~R~R~Z~Nb~d~l~t~|~~~~~~~~~~~~~~~~~~ $,4<DLT\ddlt| $,4<DLTT\Ndfnv~ƀ΀րހ&.6>FNV^ffnv~Ɓ΁ցށ&.6>FNV3|x, nAp`Helve-WP,X($PW+KKIUHP DeskJet 1600C (Copy 2),,44,,0    6k$Triangle0      (O$        02Ap`Helvetica http://www.emc.nceo.noaa.gov/resarch/osse/swa/sys_errors.htm                                               aaaa'dxd dcdef`C<<CLevel 1Level 2Level 3Level 4Level 5(2$ !  (2$ !  ($$   1  ,dZ+s TABLE A#AZ*Humanst521 Cn BT TABLE B?Level 1Level 2Level 3Level 4Level 5TABLE C                                                                ,@0AZArial                                                                                                                                                                                                   d"  _Aa)` dE` A X   XXX,XC XXCorrespondingauthorsaddress:MichikoMasutani,  NOAA/NWS/NCEP/EMC,5200AuthRd.CampSprings,MD  20746.Email:michiko.masutani@noaa.gov.#XX C#W` XX                                                                                                                                                                                                                                                                                                                                ,@0AZArial                                                                                                                                 TABLE A                                                                Level 1Level 2Level 3Level 4Level 5                                                                Level 1Level 2Level 3Level 4Level 5                                                                Level 1Level 2Level 3Level 4Level 5                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 !  _ XXX,X3.3   fCXX Impact_Assesment_ԀofaDopplerWind_Lidar_Ԁfor_NPOESS_/_OSSE_#XXfCx#   &%~%XXStephenJ.Lord,Michiko_Masutani_,JohnC.Woollen,JohnC._Derber_  _NOAA_/_NWS_/_NCEP_/_EMC_,CampSprings,MD | #XX%&%~#,,XX&%~%,,G.David_Emmitt_,SidneyA.Wood,StevenGreco   SimpsonWeatherAssociates,Charlottesville,VA#,,%&%~A#   Ѐ x  &%~%,,RobertAtlas,JosephTerry  <  NASA/_GSFC_,_Greenbelt_,MD#,,%&%~D#    Ѐ    &%~%,,ThomasJ._Kleespies_ `  _NOAA_/_NESDIS_,CampSprings,MD#,,%&%~;# 8  ',,http://www.emc.noaa.gov/research/osse#,,'8# 4  h   ',,((#(#(1.INTRODUCTION   h    ThefutureNational_POES_ԀSystem @ (_NPOESS_)isscheduledtoflyduringthe2007-2010  period.Forthenext10years,aconsiderable p amountofeffortmusttakeplacetodefine,develop 0 andbuildthesuiteofinstrumentswhichwill @ comprisethe_NPOESS_.Theforecastimpactof  currentinstrumentscanbeassessedbyObserving p SystemExperiments(_OSEs_),inwhichalready 0 existingobservationsaredeniedoraddedto @ observationsfromastandarddatabase.However,  theimpactoffutureinstrumentsmustbeassessed p withexperimentsusingsimulatedobservations. 0 TheseexperimentsareknownasObservingSystem @ SimulationExperiments(_OSSEs_).(Atlas,1997)VV)V     Thisprojectisacollaborationamongthe p! NationalCentersforEnvironmentalPrediction 0" (_NCEP_),NASA/DataAssimilationOffice(DAO), @# SimpsonWeatherAssociates(_SWA_),andthe $ NationalEnvironmentalSatellite,Dataand p % InformationService(_NESDIS_).Throughthis 0!& collaboration,thedataassimilationandmodeling !@' communitiescanbeinvolvedininstrumentdesign "( andcanprovideinformationabouttheexpected p#) impactofnewinstruments.Furthermore,through 0$* the_OSSEs_,operationaldataassimilationsystems $@ + willbereadytohandlenewdataintimeforthe %!, launchofnewsatellites.Thisprocessinvolves p&!- preparationforfuturedatavolumesinoperations, 0'". thedevelopmentofthedatabaseanddata '@#/ processing(includingformatting)andaquality ($0 controlsystem.Allofthisdevelopmentwill p)$1 acceleratetheoperationaluseofdatafromthe 0*%2 futureinstruments(Lordetal.1997). *@&3 wbA1-j|&" `@Ep,""p,"w p,'5  p Foreach_OSSE_,alongintegrationofan h 5 atmosphericgeneralcirculationmodel(GCM)is (x 6 requiredtoprovidea"trueatmosphere"forthe 87 experiment.Thisiscalledthe"naturerun"(NR). 8 Thenaturerunneedstobesufficiently h9 representativeoftheactualatmospherebut (x: differentfromthemodelusedforthedata 8; assimilation.Theobservationaldataforexisting < andfutureinstrumentsissimulatedfromNRand h= impacttestsareperformedforbothrealand (x> simulateddata.Thenaturerunusedinthis 8? project,thedataassimilationsystemandforecast @ modeusedintheseexperimentsisdescribedin hA _Masutani_Ԁetal(2002). (xB  p Amongvariouscandidateinstruments 8C Dopplerwind_lidar_Ԁ(_DWL_)winddataareproduced D aslineofsight(LOS)windsby_SWA_Ԁusingtheir hE _Lidar_ԀSimulationModel(_LSM_).Bracketing (xF sensitivityexperimentsarebeingperformedfor 8G various_DWL_Ԁtechnologyneutralconceptsto H boundthepotentialimpact(Lordetal.2001a). h I Scanning,andvariousdatasamplingstrategies, (!xJ arebeingtestedwiththeseexperiments. !8K  3.SIMULATIONOFOBSERVEDDATA  h#M  p  p Detailsofprocedurestosimulate $@ O observationaldataaredescribedin_Masutani_Ԁet %!P al.(1999b)andLordetal.(2001a,2001b)and p&!Q thesepapersareavailableatthe_OSSE_Ԁwebsite. 0'"R ЀTheinitialsimulationusesrealobservationaldata '@#S distributionsavailableinFebruary1993._ACARS_ ($T andsatellitederivedwindsaresimulatedwith p)$U distributioninFebruary1993._TOVS_Ԁlevel1B 0*%V radiance(T1B)dataissimulatedby *@&W _NOAA_/_NESDIS_ԀanddetailsaredescribedinLord +'X etal.(2001a)and_Kleespies_Ԁ(2001).NASA/DAO p,'Y istakingtheleadinthesimulationofrealistic 0-(Z conventionalobservations,includingcloudmotion -@)[ vectors(_CMV_,_Velden_Ԁetal.1997)and_ACARS_.  Morerealistic_CMVs_ԀbasedonNRcloudandthe p morerecentdistributionof_CMVs_Ԁaresimulatedwith 0 collaborationwith_SWA_. @   Inthispapertheimpactof_DWL_Ԁis  assessedwithexistinginstruments.However,itis p importantthatassessmentisalsodonewithmore 0  advancedinstrumentsexpectedwhen_DWL_Ԁwould  @ beactuallylaunched.Higherdensity_CMVs_Ԁand   moreadvancedsoundersneedtobeincludedin p   impactassessment.TheatmosphericInfrared 0   Sounder(AIRS,Goldbergetal2001)isscheduled  @  tobeincludedinthe_NPOESS_ԀPreparatoryProject   (_NPP_)instrumentsuite.Asoneoftheadvanced p  sounders,AIRSisplannedtobeincludedinthe 0  calibrationandimpactassessment.Outlineofthe @  simulationofAIRSdataisgivenin_Masutani_Ԁetal.   (2001).TheAIRSsimulationpackagewasoriginally p  developedbyEvan_Fishbein_Ԁof_JPL_.Thesimulation 0  (i.e.,forwardcalculation)isbasedonradiative @ transfercodedevelopedby_Larrabee_Ԁ_Strow_  (_UMBC_).AIRSwillbesimulatedby_NOAA_/_NESDIS_ p usingthesimulationpackagefrom_JPL_.A 0 _scatterometer_Ԁ(e.g.,_ASCAT_)isalsoconsideredto @ beaddedandmethodsimulationisbeing  investigated. ` 0 0 V VpVV 3.1Simulationof_DWL_Ԁdata 0     Thesimulationof_DWL_Ԁdataincludes  effortswith_DWL_Ԁperformancemodels,atmospheric t circulationmodelsandatmosphericopticalmodels 4 (Atlasand_Emmitt_,1995;_Emmitt_,1995a;_Emmitt_ D andWood,1996;Woodetal.,1993;Woodetal.,   1995;Woodetal.,2001).Thestepsbetweena t! notionalconceptfora_DWL_Ԁandtheblueprintsfor 4" instrumentconstructionincludeaconsiderable D# amountofperformancemodelingand,for $ space-basedsystems,anintensiveseriesof t % _OSSEs_.DuringandsubsequenttotheLaser 4!& AtmosphericWindSounder(LAWS)study(Bakeret !D' al.,1995),amethodforassessingthepotential "( impactofanew_DWL_Ԁobservingsystemwas t#) established.Theinstrumentparametersare 4$* providedbytheengineeringcommunity.Scanning $D + andsamplingrequirementsareprovidedbythe %!, sciencecommunityanddefinevariousinstrument t&!- scenarios.Thesescenariosaretestedinitiallyby 4'". examiningthesensitivityofanalysestothevarious 'D#/ scenarios.Acandidate_DWL_Ԁconceptisthen ($0 chosenforafull_OSSE_,andanimpactstudyisthen t)$1 conductedandevaluatedbyatechnologyneutral 4*%2 group. *D&3   Thebracketing_OSSEs_Ԁarebeingperformed +'4 forvarious_DWL_Ԁconceptstoboundthepotential t,'5 impact.Later_OSSEs_Ԁwillbeperformedformore 4-(6 specificinstruments(_Emmitt_,1999).Thefollowing -D)7  technologyneutralobservationcoverageand 7 measurementerrorcharacterizationswillbe p8 explored . 09 EXP1(Best): Ultimate_DWL_Ԁthatprovidesfull ; troposphericLOSsoundings,cloudspermitting. <  EXP2(_PBL_+cloud): Aninstrumentthatprovides  X> onlywindobservationsfromcloudsandthe_PBL_.  ? Ѐ0 p  @p(#p(#  EXP3(Upper): Aninstrumentthatprovidesmid P A anduppertroposphericwindsonlydowntothe  hB levelsofsignificantcloudcoverage.  ( C  Exp4(NonScan): Anonscanninginstrument X E thatprovidesfulltroposphericLOSsoundings,  p F cloudspermitting,alongasinglelinethatparallels 0 G thegroundtrack.  H  _TRV_ :200kmx200kmxT  pJ  p T:Thicknessofthe_TRV_Ԁ 8K  p 0.25kmifz<2km,1kmifz>2km L Ѐ0 p 0.25kmforcloudreturnhMp(#p(#  Swathwidth: 2000kmexceptforEXP4(non 8O scanning) P  p Nomeasurementerrorisassignedforthe 0R initialtest.Strategiesforsystematicerrorsare @S discussedby_Emmitt_Ԁ(2000).Onemeasurement T isanaverageofmanyshots.Dataproducts pU baseduponclusteredanddistributedshotsare 0V generatedforeachexperiment.Theclustered @W dataproductisbaseduponaveragingthe X observationsassociatedwithshotsclustered pY withinanareathatisverysmallcomparedtothe 0Z baseareaofthe_TRV_.Thedistributeddata @[ productisbaseduponaveragingtheobservations  \ ofshotsdistributedthroughoutthe_TRV_Ԁaswould p!] resultfromcontinuousconicalscanning. 0"^  p Distributedshotsforthenonscan "@_ experiment(EXP4)arenotrealistic.However,it #` isusedtotestthepenetrationthroughcloud.In p$a therealatmosphere,cloudhasporositywhichis 0% b notdescribedintheNRarchive.Cloudporositylet %@!c some_DWL_Ԁshottogothroughthecloud.This &"d notpossiblefortheNRcloudasthecloudsare p'"e uniformwithinagridintheNR.Distributed 0(#f measurementscollectmanyshotswithinthe_TRV_ (@$g andthereismorechancetopenetratethroughthe )%h atmosphere.Thisdoesnotexactlymodelthe p*%i porosityofthecloudbutitisusedtocheckthe 0+&j penetrationduetoporosity. +@'k  p EXP2andEXP3aresimulatedtotest ,(l variouswavelengthandmeasurementmethods. p-(m Theyaretestedbutnotpresentedinthispaper.  Morerecentdevelopmentsarediscussedby p _Emmitt_Ԁ(2001). 0  4.INITIALRESULTSFORIMPACT  ASSESSMENTFOR_DWL_Ԁ x    Priortotestingfutureinstruments,data  P impacttestofexistinginstrumentsareperformedto   calibrate_OSSEs_Ԁ(_Masutani_Ԁetal.2002).Theresults    showedtherearereasonableagreementsbetween @   simulatedandrealdataimpactbuttheinterpretation  P  needstobeconductedwithgreatcaution.      InTable1partofexperimentsusedfor   calibrationand_DWL_Ԁimpactassessmentsarelisted. @  InTable2timeandareaaveragedrootmean P  squareerror(_RMSE_)betweencontrolexperiment   andtheNRarelisted.Theperiodusedforthe   averagingisfromFebruary14toMarch6.The @  valuesarezonallyaveragedandaveragedinthe P latitudebandindicatedinthetableforthesouthern  hemisphere(_SH_),thetropics(TROP),andthe  northernhemisphere(NH).Impactonzonalwind(U) @ for200_hPa_,500_hPa_Ԁand850_hPa_Ԁarepresented. P Changeinthe_RMSE_Ԁfromthecontrolexperiments  arepresentedasimpactinTable3.    ThefirstthreeexperimentsinTable3are @ denialtestsforexistinginstrument.These P experimentsarealsousedforcalibrationandthe  impactsarecomparedwithrealanalysis.Forthe  calibrationtheanalysiswascomparedwith1B @ experiment.Now,forthesimulateddatathe P analysiscanbecomparedwithtruth.Theyshow   theRWindhasthelargestimpactinNHandT1Bin ! _SH_.Thisresultsagreeswiththeimpacttestin @" forecastskillscore(_Masutani_Ԁetal.2001).The P# impactofRTemponwindfieldsissmallanditis $ notcleariftheyhavepositiveimpact.  % 4.1Results "P'   Amongmanycandidateinstrumentsforthe #) _OSSE_,_DWL_Ԁwindsaresimulatedby_SWA_. D$* Accordingtothestrategyforbracketingsensitivity %T + experiments(Lordetal.2001a,Lordetal.2001b, %!, _Masutani_Ԁetal.2001),scanningornonscanning, &!- variouswavelengths,numbersofLOSper D'". measurement,arebeingtested.Sensitivityto (T#/ weightinthedataassimilationisalsobeingtested. ($0   Forfirstfewdays,morethan20casesare D*%2 testedwithvariouscombinationsandselectedcases +T&3 arecompletedforthewhole_OSSE_Ԁperiod(00z +'4 February1300zMarch7,1993).Experiments ,'5 discussedinthispaperarelistedinTable1.The D-(6 distributeddataforthenonscanningscenarioisnot .T)7 realistic.However,itisusedtotesttheeffectof 7 penetration.Becauseoftheaveragingofeach p8 200Kmsquarearea,more_DWL_Ԁshotspenetrate 09 tolowerlevelsfordistributedshots.Theamount @: ofpenetrationisstillanunknownquantityand ; needstobeinvestigated. p<  p Theimpactsaremeasuredasthechange 0 = in_RMSE_ԀfromtheNR.Table2showszonallyand  @> timeaveragedvaluesforthreelatitudebandsas  ? differencesfromthe_RMSE_Ԁbetweenthecontrol p @ experiment(_CTL_)andNR.Apositivevalue 0 A indicatestheexperimenthaspositiveimpact  @B comparedto_CTL_.  C  p Rows47inTable3showanadvantage p D ofascanninginstrument(rows4,5)overanon 0 E scanninginstrument(rows6,7).Thedifferences @ F arelargestintheuppertroposphereandare  G reducedinthelowertroposphere.Distributed p H shotsarebetterthanclusteredshotsinmostof 0 I thecases.Comparisonbetweenrows4,6and @J 5,7showsthatpenetrationindistributeddatais K importantinthelowertroposphere. pL  p Representativenesserrorsof1m/sand 0M 7m/saretestedforthefirstweekandtheresults @N arepresentedinLordetal(2001c).Theimpact O withrepresentativenesserror7m/sisabout10 pP 20%lessthat_that_Ԁof1m/s,butthegeographical 0Q distributionoftheimpactdoesnotchange. @R  p  p WithT1Bin_CTL_,_DWL_Ԁdataimprovedthe pT windfieldsgloballyatalllevelsforallexperiments 0U (row47inTable3).Majorimprovementsare @V overthetropicsifT1Bisincludedin_CTL_. W Marseilleetal.(2001)showedmajorimpactin pX _SH_,becauseintheirexperiment_CTL_Ԁdoesnot 0Y includeT1B.IfT1Bareincluded,themajor @Z improvementin_SH_ԀhasalreadyachievedbyT1B [ andofthemajorimprovementsdueto_DWL_Ԁmove p \ totropics.HoweverwithoutT1Bin_CTL_,using 0!] NTVas_CTL_Ԁmoreimprovementisachievedin !@^ _SH_Ԁevenbyworstcaseof_DWL_Ԁ(Dex4cr7,row9), "_ comparedtoT1B(row8).AlthoughT1Band p#` Dex4cr7showsimilarmagnitudeofimpactin_SH_ 0$a andminimumimpactinNH,row1inTable4 $@ b showstherearesignificantdifferencesbetween %!c experiment1BandexperimentDex4cr7. p&!d Therefore,bothT1BandDex4cr7togetherallow 0'"e afurtherimprovementtobeachieved(row10in '@#f table3).InNHneithertheT1Bnorworstcaseof ($g _DWL_Ԁ(Dex4cr7)producesignificantimpact. p)$h Significantimpactisachievedbythebestcaseof 0*%i _DWL_Ԁ(Dex1dr1inrow11) *@&j  p Dex4cr7isrunwithoutRWind +'k (_NTVNWIN_)as_CTL_Ԁandtheimpactiscompared p,'l withRWind(Row12,14,table3).Theresults 0-(m showworstcaseof_DWL_Ԁ(Dex4cr7)dotproduce -@)n asmuchasimpactoverNH_compareed_ԀtoRwind.  However,theimpactofthebestcaseof_DWL_ p (Dex1dr1)istwiceasmuchasRwind(row15, 0 Table3)inNH.AddingT1Bonlycouldcause @ negativeimpactoverNH(row13,table3).The  distancebetweenNTVandF'Dex4cr7NWIN#'F:`#Ԁintable4 p indicatesthattheimpactofRwindandtheimpact 0  of_DWL_Ԁarequitedifferent.  @ 4.2CommentsontheResults  p       _DWL_Ԁisevaluatedwith1993data  D  distribution.However,_DWL_Ԁwindsalsoneedtobe    evaluatedwithboththecurrentdatadistribution t  andtheanticipatedfuturedatadistribution 4  correspondingtowhenthe_DWL_Ԁdatawillbeused. D    Inthispaperno_measuserment_Ԁerroris t  includedin_DWL_.Systematicerrorsarediscussed 4  by_Emmitt_Ԁ(2000)andotherlargescalecorrelated D errorneedtobedesignedandaddedto  assessment.Varioussamplingstrategiessuchas t separationbetweenforwardandbackwardscan, 4 adaptiveobservationneedtobetested. D   InthispaperonlyresultsfromUare  presented.Theimpactonmeridionalwind(V)is t similartothatonU.Impactintemperaturefieldsis 4 moresensitiveandcomplicated.Impacton D temperaturefromradiancedataandRTemp  involvemanyprocedurestoaltertheresults,such t asbiascorrection.Impactontemperaturefrom 4 _DWL_Ԁwindisevenmorecomplicated.Itis D interestingtonotethatwhendistributed,data   usuallygivepositiveimpactinwindfields t! comparedtoclustereddata,althoughsometimes 4" thetemperaturefieldsisbetterwiththeclustered D# data.Analysisandsimulationproceduresneedtobe $ evaluatedanddevelopedformorereliableresults. t %  5.FUTUREPLANS   !D'   Thecalibrationwillbecontinuedtogain |#) furtherconfidenceinthe_OSSE_Ԁ_system._ԀVarious <$* techniquesforaddingsystematicerrorswillbe $L + tested.ThesimulationprocedureofT1Brequires % !, furtherevaluation,includingtheformulationof |&!- observationalerrors. <'".   Inadditiontoa_DWL_ԀandAIRS,theCross 'L#/ TrackInfraredSounder(_CrIS_),Conicallyscanning ( $0 Microwave_Imager_/Sounder(CMIS),andthe |)$1 AdvancedTechnologyMicrowaveSounder(_ATMS_) <*%2 havebeenproposedascandidateinstrumentstobe *L&3 testedby_OSSEs_.Weareproceedingtodevelop + '4 appropriateforwardmodelsfortheseinstruments. |,'5   Inordertomakereliablerecommendations, <-(6 thetechniquesforcreatingsimulatedobservations -L)7 needtoberefined.Additionoflargescale 7 spatiallycorrelatederrorandsystematicerrorin p8 simulateddatamayaltertheresults. 09  p _OSSEs_Ԁalsoneedtobetestedwith @: upgradedtechniquesfordatahandlinganddata ; assimilationsystem.Sincetheamountofdata p< involvedinthefutureinstrumentsincreases 0 = drastically,effectivesuperobservationsto  @> reducethesizesofdatasetsneedtobestudied  ? (Purseretal.2001).Includinganadaptive p @ correctionforthebiasinthedataassimilationwill 0 A alsobetested(Purserand_Derber_,2001).  @B  p Futureinstrumentsneedtobetestedwith  C 2001andfuturedatadistributionssincethe1993 p D datadistributionisoutdated.Alternative_NRs_Ԁfor 0 E thesameperiodandsummertimehavealsobeen @ F generatedbyNASA/DAOandcanbeusedto  G investigateadditionalatmosphericregimes._NRs_ p H totestnorthernsummertimeresponseare 0 I important,especiallytostudytheimpacton @J tropicalstormprediction. K  p Theevaluationmetricswillbeexpanded pL toincludediagnosticsofstrengthandpositionof 0M cyclonesandjetsandastudyofextremeevents, @N aswellasstandardforecastskillscores.Cost O benefitandflightplanningwillalsobestudied. pP  ACKNOWLEDGMENT  HR  p Wereceivedmuchassistancefromthe T DataServicesSectionandDr.Anthony @U _Hollingsworth_Ԁof_ECMWF_Ԁinsupplyingthenature PV run.Throughoutthisproject_NOAA_/_NWS_/_NCEP_, W NASA/DAOand_NOAA_/_NESDIS_Ԁstaffsprovided X muchtechnicalassistanceandadvice. @Y Especially,wewouldliketothankW.Yang,R. PZ _Treadon_,Y._Zhu_,W.S.Wu,M._Iredell_,D. [ _Keyser_,W.Collins,andR._Kistler_Ԁof_NCEP_,G.  \ _Brin_,S.BloomandN.WolfsonofDAO,andV. @!] Kapoor,P.LiandW.Wolfof_NESDIS_.Wewould "P^ liketothankEvan_Fishbein_Ԁof_JPL_Ԁforproviding "_ simulationcodeforAIRSdata.Drs.E._Kalnay_, #` W.Baker,J._Yoe_ԀandR.Daleyprovidedexpert @$a advice.Weappreciatetheconstructive %P b commentsfrommembersofthe_OSSE_ԀReview %!c Panel.ThisprojectissponsoredbytheIntegrated &!d ProgramOffice(IPO)for_NPOESS_Ԁandbythe @'"e _NOAA_ԀOfficeofAtmosphericResearch(OAR) (P#f andthe_NOAA_ԀNationalEnvironmentalSatellite, ($g DataandInformationService(_NESDIS_).We )$h thankDrs.StephenMango,Alexander @*%i MacDonald,JohnGaynor,JimEllicksonand +P&j JohnPereirafortheirsupportandassistancein +'k thisproject. ,'l   .P)n REFERENCES   W` '0   j W`#j.jI#jj.Atlas,R.andG.D.Emmitt,1995:Simulationstudiesoftheimpactofspace-basedwind (x profilesonglobalclimatestudies.Proc.AMSSixthSymp.onGlobal ( ChangeStudies,Dallas,TX,January1995.VV 0   Atlas,R.1997:Atmosphericobservationandexperimentstoassesstheirusefulnessindata 8 assimilation.J.Meteor.Soc.Japan, 75 ,111130. h 8VV 0   Baker,W.E.,G.D.Emmitt,F.Robertson,R.M.Atlas,J.E.Molinari,D.A.Bowdle,J.Paegle,   R.M.Hardesty,R.T.Menzies,T.N.Krishnamurti,R.A.Brown,M.J.Post,J.R. P  Anderson,A.C.LorencandJ.McElroy,1995:Lidar-measuredwindsfrom  P  space:Anessentialcomponentforweatherandclimateprediction.Bull.    Amer.Meteor.Soc.,76,869-888.`  VV 0   Becker,B.D.,H.Roquet,andA.Stofflen1996:Asimulatedfutureatmospheric  `  observationdatabaseincludingATOVS,ASCAT,andDWL.BAMS, 10 ,2279    2294.x VV 0   Derber,J.C.andW.-S.Wu,1998:TheuseofTOVScloud-clearedradiancesintheNCEP (x  SSIanalysissystem.Mon.Wea.Rev., 126, 2287-2299.( VV 0   Emmitt,G.D.,1995a:OSSE'sinsupportofasmall-satellitemission.Paperpresentedat   theNOAAWorkingGrouponSpace-basedLidarWinds,Clearwater,FL, @  January31-February2.@ VV 0   Emmitt,G.D.andS.A.Wood1996:LidarMappingofCloudTopsandCloudTopWinds,   1996.PL-TR-96-2129,F19628-93-C-0196,1996.PVV 0   Emmitt,G.D.,2000:SystematicerrorsinsimulatedDopplerwindlidarobservations. P 4.dO  5  http://www.emc.nceo.noaa.gov/resarch/osse/swa/sys_errors.htm6Owice  7is AVV 0   EmmittG.D.,2001:Globalwindobservationalrequirementsandthehybridobserving ` systemappproach.AMSpreprintvolumefortheFifthSymposiumon ` IntegratedobservingSystems,1419January2001,Albuquerque,NM.176  178.pVV 0   Goldberg,M.D.,L.McMillin,W.Wolf,L.Zhou,Y.Qu,andM.Divakarla,2001:  p OperationalradianceproductsfromAIRS,AMS11thConferenceon   SatelliteMeteorologyandOceanography15-18October2001,Madison,  Wisconsin.555558.0VV 0   Kleespies,T.J.andD.Crosby2001:Correlatednoisemodelingforsatelliteradiance 0  simulation.AMSpreprintvolumeforthe11thConferenceonSatellite ! MeteorologyandOceanography,October2001,MadisonWisconsin.604 @" 605.@#VV 0   Lord,S.J.,E.Kalnay,R.Daley,G.D.Emmitt,andR.Atlas1997:UsingOSSEsi.nthe $ designofthefuturegenerationofintegratedobservingsystems.Preprint P% volume,1stSymposiumonIntegratedObservationSystems,LongBeach, P& CA,27February1997.'VV 0   Lord,S.J.,M.Masutani,J.S.Woollen,J.C.Derber,R.Atlas,J.Terry,G.D.Emmitt,S.A. ` ( Wood,S.Greco,T.J.Kleespies,2001a:ObservingSystemSimulation !`) ExperimentsforNPOESS,AMSPreprintvolumetheFifthSymposiumon !* IntegratedObservingSystems.1419January2001,Albuquerque,NM168 p"+ 173. #p,VV 0   Lord,S.J.,M.Masutani,J.S.Woollen,J.C.Derber,R.Atlas,J.Terry,G.D.Emmitt,S.A. # - Wood,S.Greco,T.J.Kleespies,2001b:ObservingSystemSimulation $. ExperimentsforNPOESS,AMSPreprintvolumethe14thConferenceon 0% / NumericalWeatherPrediction,30July2August2001,FortLauderdale, %0!0 Florida.167171.&!1VV 0   Lord,S.J.,M.Masutani,J.S.Woollen,J.C.Derber,R.Atlas,J.Terry,G.D.Emmitt,S.A. @'"2 Wood,S.Greco,T.J.Kleespies,2001c:ObservingSystemSimulation '@#3 ExperimentsforNPOESS,Presentationatthe14thConferenceonNumerical (#4 WeatherPrediction,30July2August2001,FortLauderdale,Florida.P)$5VV   http://www.emc.ncep.noaa.gov/research/osse. *P%6 0   Marseille,G.J.,A.Stoffelen,F.Bouttier,C.Cardinali,S.deHaanandD.Vasiljevic,2001: *&7 ImpactassessmentofaDopplerWindLidarinspaceonatmospheric `+&8 atalysisandnumericalweatherprediction.KNMI,Contract ,`'9 No.13018/98/NL/GD.,(:VV Ѐ0   0VV  Masutani,M.,J.C.Woollen,J.C.Derber,S.J.Lord,J.Terry,R.Atlas,S.A.Wood,S.Greco, p-(; G.D.Emmitt,T.J.Kleespies1999a:ObservingSystemSimulation .p)<VV ExperimentsforNPOESS,AMSPreprintvolumeforthe13thConference < onNumericalWeatherPrediction.September1999,DenverColorado,1 `= 6.`>p(#p(# 0 p  Masutani,M.K.Campana,S.Lord,andS.K.Yang1999b:NoteonCloudCoverofthe ? ECMWFnaturerunusedforOSSE/NPOESSproject.NCEPOfficeNote p@ No.427 pAp(#p(# 0 p  Masutani,M.,J.S.Woollen,J.Terry,S.J.Lord,T.J.Kleespies,G.D.Emmitt,S.A.  B Wood,S.Greco,J.C.Derber,R.Atlas,M.Goldberg2001:Calibrationand  C InitialResultsfromtheOSSEsforNPOESS,AMSpreprintvolumefor 0 D the11thConferenceonSatelliteMeteorologyandOceanography,October  0E 2001,MadisonWisconsin.696699. Fp(#p(# 0 p  MasutaniM.,J.C.Woollen,S.J.Lord,J.Terry,J.C.Derber2002:Calibrationand @ G ErrorSensitivitytestsforNPOESS/OSSE,AMSPreprintvolumetheSixth  @H SymposiumonIntegratedObservingSystems.January2002,Orlando,  I Florida.P Jp(#p(# 0 p  McMillin,L.M.,L.CroneandT.J.Kleespies,1995:Atmospherictransmittanceofan P K absorbinggas.5:ImprovementstotheOPTRANapproach.Appl.Opt.  L  34 󀀀(36)1995.` Mp(#p(# 0 p  McNally,A.P.,J.C.Derber,W.-S.WuandB.B.Katz,2000:TheuseofTOVSlevel-1 h N radiancesintheNCEPSSIanalysissystem.Q.J.Roy.Metorol.Soc., 129 ,  O 689724. Pp(#p(# 0 p  Parrish,D.F.andJ.C.Derber,1992:TheNationalMeteorologicalCenter'sspectral 0Q statisticalinterpolationanalysissystem.Mon.Wea.Rev., 120 ,1747- 0R 1763.Sp(#p(# 0 p  Purser,R.J.andJ.C.Derber,2001:UnifiedTreatmentofmeasurementbiasand HT correlationinvariationalanalysiswithconsiderationofthe HU preconditioningproblem.AMSPreprintvolumeforthe14thConference V onNumericalWeatherprediction.July2001,FortLauderdale,Florida. XW 467470.XXp(#p(# 0 p  Purser,R.J.andD.F.Parrish,2000:ABayesiantechniqueforestimating Y continuouslyvaryingstatisticalparametersofavariational hZ assimilation.NCEPOfficeNote429.(AlsosubmittedtoMeteor.Appl. h[ Phys.)\p(#p(# 0 p  Purser,R.J.,D.F.ParrishandM.Masutani2001:Meteorologicalobservationaldata x] compression;analternativetoconventional SuperObbing.NCEP (x^ OfficeNote430.(_p(#p(# 0 p  Velden,C.S.,C.M.Hyden,S.J.Nieman,W.P.Menzel,S.Wangzong,andJ.S.Goerss ` 1997:Uppertroposphericwindsderivedfromgeostationarysatellite 8a watervaporobservations.Bull.Amer.Meteor.Soc., 78 ,173195.8bp(#p(# 0 p  _Velden_,C.S.,T.L.OlanderandS.Wanzong,1998:TheimpactofmultispectralGOES c 8windinformationonAtlantictropicalcyclonetrackforecastsin1995. P d Part1:Datasetmethodology,descriptionandcaseanalysis.Mon.Wea. !Pe Rev., 126, 12021218.(#(#K(#p(#p(# !f 0 p  Wood,S.A.,G.D.Emmitt,M.Morris,L.WoodandD.Bai.Space-basedDopplerlidar h"g samplingstrategies--algorithmdevelopmentandsimulated #hh observationexperiments.FinalRept.NASAContractNAS8-38559, #i MarshallSpaceFlightCenter,266pp.,1993.x$jp(#p(# 0 p  0pp(#p(#  Wood,S.A.,G.D.Emmitt,D.Bai,L.S.Wood,andS.Greco,1995:Acoherentlidar (%x k simulationmodelforsimulatingspace-basedandaircraft-basedlidar %(!l winds.PaperpresentedattheOpticalSocietyofAmerica'sCoherent &!m LaserRadarTopicalMeeting,Keystone,CO,July23-27.8'"np(#p(# 0 p  Wood,S.A.,G.D.Emmitt.,andS.Greco2001:Thechallengesofassessingthefuture '8#o impactofspacebasedDopplerWindLidarswhileusingtodaysglobal (#p andregionalatmosphericmodels. AMSpreprintvolumefortheFifth H)$q SymposiumonIntegratedobservingSystems,1419January2001, )H%r Albuquerque,NM.95101 #j.j##'j.# *%sp(#p(# (VV(  *gaiddd Xdd Xdd X(#(#_,Fdd , dd!,edd!,Kdd!,Xdd!,dd!,dd!,M dd!+  pp Experiment $ Name   j.'T1B  $ RAOB $ WIND   DWL  $ DWL $ SHOT    DWL $  Rep_error    DWL $  ЀSCAN    #'j.#  $ 1B  8 Y  8 Y  8 N  8   8   8   8 AllexistingdataincludingT1B  8 NTV  <  N  <  Y  <  N  <    <    <    <  denyT1Bfrom1B  <  1BNWIN    Y     N   ! N   "    #    $    % denyRWindfrom1B   & 1BNTMP   4' Y   4( Y   4) N   4*    4+    4,    4- denyRTempfrom1B   4. NTVNWIN  8 / N  8 0 N  8 1 N  8 2   8 3   8 4   8 5 sfcdataandRTemp  8 6 1BDex1dr1   7 Y   8 Y   9 Y   : D 4*  ;  ?1?41 3)' <  ?1 ? 3Y   = Bestinscan   > 1BDex1cr7  0 ? Y  0 @ Y  0 A Y  0 B C 4* 0 C  @7@47 3)'0 D  @7 @ 3Y  0 E Worstwithinscan  0 F 1BDex4dr1  4 G Y  4 H Y  4 I Y  4 J D 4* 4 K  ?1?41 3)'4 L  ?1 ? 3N  4 M bestinnonscan  4 N 1BDex4cr7   O Y   P Y   Q Y   R C 4*  S  @7@47 3)' T  @7 @ 3N   U worstinnonscan   V Dex4cr7  ,W N  ,X Y  ,Y Y  ,Z C 4* ,[  @7@47 3)',\  @7 @ 3N  ,] worstcaseofDWLaddedto ,^ NTV  _ Dex1dr1  @` N  @a Y  @b Y  @c D 4* @d  ?1?41 3)'@e  ?1 ? 3Y  @f BestDWLaddedtoNTV  @g Dex4cr7NWIN  Dh N  Di N  Dj Y  Dk C 4* Dl  @7@47 3)'Dm  @7 @ 3N  Dn WorstcaseofDWLaddedto Do NTVNWIN  Tp Dex1dr1NWIN  Xq N  Xr N  Xs Y  Xt D 4* Xu  ?1?41 3)'Xv  ?1 ? 3Y  Xw BestcaseofDWLaddedto Xx ЀNTVNWIN hy Table1.Experimentsdescribedinthispaper.AllotherconventionaldataincludingRAOBtemperature, z ACARdata,CMV,etcareincludedinallexperiments. { *kalddFdd F dd! edd!eKdd!KXdd!Xdd!dd!M dd!M gai(#(#,Sdd",dd",dd",dd"+   _CTL_  < _SH_Ԁ(80S20S)  < TROP(20S20N)  < NH(20N80N)  < 1B  @! 3.4 @! 2.8 "P 1.9  " 3.6 @! 3.1 "P 2.0  " Ѐ2.3 @! Ѐ2.5 "P Ѐ2.1  " NTV  $d 4.1 $d 3.3 $$  2.2  %  3.9 $d 3.3 $$  2.1  %  2.7 $d 2.6 $$  2.1  %  NTVNWIN  &8" 5.1 &8" 3.8 '" 2.5  h(# 4.4 &8" 3.4 '" 2.4  h(# 2.7 &8" 3.0 '" 2.2 h(#  Table2RMSE(CTLNR)forzonalwind(U).Withineachcelltop:200_hpa_,middle500_hPa_Ԁbottom850hPa. *h% RMSEareaveragedbetweenFebruary14andMarch6,1993. *(&    `  +& *mah4ddSdd"Sdd"dd"dd"kal(#(#,d",d ,Td ,d ,d ,d +  00    CTL   Exp   SH  (80S T 20S)  d TROP   (20S T  20N)  d  NH   (20N T  80N) 4* d  ?1?41 3)'h   ?1 ? 31B  h  NTV h  (Deny1B)  ( x -0.67 h  -0.56 ( x -0.30   8 Є0.24 h  Є0.22 ( x Ѐ0.062   8 -0.017 h  -0.085 ( x -0.0013 4*  8  @2@42 3)'<   @2 @ 31B  <  1BNWIN <  (DenyRWind)   L -0.23 <   -0.17  L! -0.12   " -0.43 < # -0.60  L$ -0.34   % -0.45 < & -0.40  L' -0.22 4*  (  @3@43 3)'` )  @3 @ 31B  ` * 1BNTMP ` + (DenyRTemp)   , 0.018 ` - 0.038  . Ѐ-0.0091   / 0.056 ` 0 0.28  1 0.0027   2 0.0029 ` 3 -0.042  5 -0.015 4* P 6  @4@44 3)' 7  @4 @ 31B   8 1BDex1dr1  9 (add_DWL_)  d: 0.88  ; 0.69 d< 0.45  $t= 1.3  > 1.3 d? 0.69  $t@ 0.24  A 0.30 dB 0.33 4* $tC  @5@45 3)'xD  @5 @ 31B aaaa xE 1BDex1cr7 xF (add_DWL_) $8Gaaaa $0.91 xH 0.48 8I 0.21  HJ 1.0 xK 0.72 8L 0.24  HM 0.31 xN 0.23 8O 0.095 4* HP  @6@46 3)'LQ  @6 @ 31B  LR 1BDex4dr1 LS (add_DWL_)   \T 0.35 LU 0.29  \V 0.16  W 0.54 LX 0.49  \Y 0.19  Z 0.12 L[ 0.12  \\ 0.089 4* ]  @7@47 3)' p^  @7 @ 31B   p_ 1BDex4cr7  p` (add_DWL_)  0a 0.25  pb 0.12 0c 0.045  d 0.34  pe 0.13 0f 0.036  g 0.086  ph 0.029 0i 0.011 4* j   @8 @48 3)'l   @8  @ 3NTV  Dm 1B Dn (AddT1B)  o 0.67 Dp 0.56 q 0.30  tr 0.24 Ds 0.22 t 0.062  tu 0.017 Dv 0.085 w Ѐ0.00_13 4* tx  "@9"@49_ 3)' y  "@9 "@ 3NTV   z Dex4cr7  { (addworst_DWL_)  !| 0.89  } 0.47 !~ 0.18  H" 0.56   0.22 ! 0.050  H" 0.10   0.059 ! 0.014 5+ H"  $@10$@510 4*(#  $@10 $@ 4NTV  # 1BDex4cr7 # (addT1Band \$ worst_DWL_)  %l  0.92 # .68 \$ 0.34  %l  0.58 # 0.35 \$ 0.098  %l  0.10 # 0.11 \$ 0.013 5+ %l   &@11&@511 4*(p&!  &@11 &@ 4NTV  p&! Dex1dr1 p&! (Addbest_DWL_)  0'" 1.8 p&! 1.4 0'" 0.93  '@# 1.6 p&! 1.2 0'" 0.71  '@# 0.22 p&! 0.42 0'" 0.31F'5+ '@#  (@12(@512 4*(d  (@12 (@ 4_NTVNWIN_  d NTV d (AddRWind)  $ #'F#1.0 d 0.49 $ 0.29   0.71 d -0.35 $  0.055    0.30 d  0.36 $  0.031 5+    *@13*@513 4*(8  *@13 *@ 4_NTVNWIN_  8 1BNWIN 8 (AddT1B)   1.5 8 0.88  0.47  h  0.36 8 -0.33  0.046  h  -0.13 8 0.042  -0.19 h   5+ ( x  ,@14,@514 4*(|   ,@14 ,@ 4_NTVNWIN_  |  Dex4cr7NWIN |  (Addworst_DWL_)  <  1.7 |   0.71 < ! 0.35   L" 0.71 | # -0.35 < $ 0.055   L% 0.065 | & 0.033 < ' -0.16 5+  L(  .@15.@515 4*(P )  .@15 .@ 4_NTVNWIN_  P * Dex1dr1NWIN P + (Addbest_DWL_)  ` , 2.9 P - 1.9 ` . 1.1   / 2.2 P 0 1.7 ` 1 1.0   2 0.49 P 3 0.55 ` 4 0.23  5  Table3.Dataimpact_measuered_Ԁby_RMSE_(_CTL_ԄNR)_RMSE_(EXPNR)ofzonalwind.Ifthevalueispositive  6 EXPisclosertotheNRthan_CTL_ԀanddataaddedtoEXPhavepositiveimpactordatasubtractedfromEXP @ 7 havenegativeimpact.Ifthevaluesarenegative,thedatasubtractedfromEXPhavepositiveimpact. P8 PeriodusedisfromFeb14toMarch6.1993.F'ԀExceptforrow11isforfeb14toFeb19.#'F#Withineachcelltop: 9 200_hpa_,middle500_hPa_Ԁbottom850hPa.F'     p  : *oapd4dd"d Td Td d d mah(#(#,Xdd ,Xdd ,dd ,dd ,dd ,dd +  P<P   d= EXP1  d> EXP2  d? _SH_  d@ TROP  dA NH 4* dB  ?1?41 3)'C  ?1 ? 31B  D Dex4cr7  E #'F#2.8 F 2.4 xG 1.3  8H 2.5 I 2.5 xJ 1.2  8K 1.3 L 1.3 xM 0.67 4* 8N  @2@42 3)'O  @2 @ 3NTV  P Dex4cr7NWIN  Q 3.6 R 2.9 LS 1.5   \T 2.6 U 2.6 LV 1.2   \W 1.3 X 1.4 LY 0.67 4*  \Z  @3@43 3)'`[  @3 @ 3NTV  `\ Dex1dr1NWIN  `] 3.7 `^ 3.3  p_ 2.0  0` 3.4 `a 3.2  pb 1.8  0c 1.9 `d 2.4  pe 1.5 4* 0f  @4@44 3)'4 g  @4 @ 31B  4 h NTV  4 i 3.6 4 j 2.9  Dk 1.5  !l 2.6 4 m 2.6  Dn 1.2  !o 1.3 4 p 1.4  Dq 0.67 !r     `  Table4.Timeandareaaverage_RMSE_ԀofzonalwindbetweenEXP1andEXP2.PeriodusedisfromFeb14 d#s toMarch6.1993.Withineachcelltop:200_hpa_,middle500_hPa_Ԁbottom850hPa.F' x   $$tt   #'F>#