WPCD GcVފB$;6o.D[{)mJ;h5hͶm\ n }yI3Ee{=i 4Uۗ#2K B`tAt;C!ZiўS]֨ cُxqU}mG \Bl A@,j D sd/Ɗ4`:5vZo/YoDcT4#ǡ/Y6@ydu4^=khBg_kϖ)FifJ՛O^u&^q#/j .Ϛa z!*5yhښx>Xt>)U+H:!9hՎg:u3F؉GǶFFNV^fnv~U ,N $,4<<DLT\dlt| $,4<<DLT\dlt|EN (08@@HPX`hpx (08@@HPX`hpx 0N ""*2:BJRZbjrz ""*2:BJRZbjrz 0wN !)19AIQYaiqy !)19AIQYaiqU:yN %-5==EMU]emu} %-5==EMU]emu} 0 N              ! !!!!!!)!1!9!A!I!Q!Y!a!i!q!y!!!!!!!!!!!!!!!!!!" """"!")"1"9"A"I"Q"Y"a"i"q"y""" 0" 0h#<$ND$F$N$V$^$f$n$v$~$$$$$$$$$$$$$$$$$$%%%%&%.%6%>%F%F%N%V%^%f%n%v%~%%%%%%%%%%%%%%%%%%&&&&&&.&6& 0C>&&N&&&&&&&&&&&&&&&&' ' '''#'+'3';'C'K'S'['c'k's'{''''''''''''''''''( ( (((#(+(3(;(C(K(S([(c(k(s({(N((N(((((((((((((((()))))')/)7)?)G)O)W)_)g)o)w)))))))))))))))))))*****'*/*7*?*G*O*W*_*g*o*w****N***************+ ++++!+)+1+9+A+I+Q+Y+a+i+q+y++++++++++++++++++, ,,,,!,),1,9,A,I,Q,Y,a,i,q,y,,,,N,,,,,,,,,,,,,,,- ----#-+-3-;-C-K-S-[-c-k-s-{------------------. ....#.+.3.;.C.K.S.[.c.k.s.{...Q/NY/[/c/k/s/{//////////////////0 000#0+030;0C0K0S0[0[0c0k0s0{0000000000000000001 111#1+131;1C1K1S12N!2#2+232;2C2K2S2[2c2k2s2{2222222222222222223 333#3#3+333;3C3K3S3[3c3k3s3{3333333333333333334 4444N44445 555#5+535;5C5K5S5[5c5k5k5s5{5555555555555555556 666#6+636;6C6K6S6[6c6k6k6s6{66666666666666 067N7777777778 888$8,848<8D8D8L8T8\8d8l8t8|8888888888888888889 999$9,949<9D9D9L9T9\9d9l9t9|99999999N99N99999999:::: :(:0:8:@:H:H:P:X:`:h:p:x::::::::::::::::::;;;; ;(;0;8;@;H;H;P;X;`;h;p;x;;;;;;;;;;;N;;;;;;;;< <<<"<*<2<:<B<J<J<R<Z<b<j<r<z<<<<<<<<<<<<<<<<<<= ==="=*=2=:=B=J=J=R=Z=b=j=r=z===========N========> >>>$>,>4><>D>L>L>T>\>d>l>t>|>>>>>>>>>>>>>>>>>>? ???$?,?4?@F@N@N@V@^@f@n@v@~@@@@@@@@@@@@@@@@@@AAAA&A.A6A>AFANANAVA^AfAnAvA~AAAAAAAAA 0ABNBBBBBBBBBBBBBC CCC$C$C,C4CKFKNKVK^KfKnKvK~KKKKKKKKKKKKKKKKKKLLLL&L.L.L6L>LFLNLVL^LfLnLvL~LLLLL 0LwMNMMMMMMMMMMMMMMMMMNN NNN!N)N1N9NANINQNYNaNiNqNyNNNNNNNNNNNNNNNNNNOO OOO!O)O1O9OAOIOQOYOaOiOqONyO{ONOOOOOOOOOOOOOOOOOPP PPP%P-P5P=PEPMPUP]PePmPuP}PPPPPPPPPPPPPPPPPPQQ QQQ%Q-Q5Q=QEQMQUQ]QeQmQuQ}Q}QNQQQQQQQQQQQQQQQQQRRRRR'R/R7R?RGRORWR_RgRoRwRRRRRRRRRRRRRRRRRRRSSSSS'S/S7S?SGSOSWS_SgSoSwSSSNSSSSSSSSSSSSSSSST T TTT!T)T1T9TATITQTYTaTiTqTyTTTTTTTTTTTTTTTTTTU U UUU!U)U1U9UAUIUQUYUaUiUqUyUUUNUUUUUUUUUUUUUUUUV V VVV#V+V3V;VCVKVSV[VcVkVsV{VVVVVVVVVVVVVVVVVVW W WWW#W+W3W;WCWKWSW[WcWkWsW{W 0WQXNYX[XcXkXsX{XXXXXXXXXXXXXXXXXXY YYY#Y+Y3Y;YCYKYSY[Y[YcYkYsY{YYYYYYYYYYYYYYYYYYZ ZZZ#Z+Z3Z;ZCZKZNSZUZUZN]Z_ZgZoZwZZZZZZZZZZZZZZZZZZZ[[[['[/[7[?[G[O[W[_[_[g[o[w[[[[[[[[[[[[[[[[[[[\\\\'\/\7\?\G\O\W\W\N_\a\i\q\y\\\\\\\\\\\\\\\\\\] ]]]!])]1]9]A]I]Q]Y]a]a]i]q]y]]]]]]]]]]]]]]]]]]^ ^^^!^)^1^9^A^I^Q^Y^Y^Na^c^k^s^{^^^^^^^^^^^^^^^^^^_ ___#_+_3_;_C_K_S_[_c_c_k_s_{__________________` ```#`+`3`;`C`K`S` 0[`&aN.a0a8a@aHaPaXa`ahapaxaaaaaaaaaaaaaaaaaabbbb b(b0b0b8b@bHbPbXb`bhbpbxbbbbbbbbbbbbbbbbbbcccc cN(c*cN2c4ceFeNeVe^efeneve~eeeeeeeeeeeeeeeeeeffff&f.f6f6f>fFfNfVf^fffnfvf~ffffffffffffffffffgggg&g.g.gN6g8g@gHgPgXg`ghgpgxgggggggggggggggggghhhh h(h0h8h8h@hHhPhXh`hhhphxhhhhhhhhhhhhhhhhhhiiii i(i0i0iN8i:iBiJiRiZibijiriziiiiiiiiiiiiiiiiiij jjj"j*j2j:j:jBjJjRjZjbjjjrjzjjjjjjjjjjjjjjjjjjk kkk"k*k 02kkNll lll$l,l4ltFtNtVt^tftntvt~ttttttttttttttttttuuuuu&u.u6u>uFuNuVu^ufunuvu~uuuuuuuuuuuuuuuuuu 0vvNvvvvvvvw www%w-w5w=wEwMwUwUw]wewmwuw}wwwwwwwwwwwwwwwwwwx xxx%x-x5x=xExMxUxUx]xexmxux}xxxxxxxxxx 0xyNyyyyyyyyyyyyyyyzzzzz'z/z7z?zGzOzWz_zgzozwzzzzzzzzzzzzzzzzzzz{{{{{'{/{7{?{G{O{W{_{g{o{w{{{N{{N{{{{{{{{{{{{{{| ||||#|+|3|;|C|K|S|[|c|k|s|{||||||||||||||||||} }}}}#}+}3};}C}K}S}[}c}k}s}{}}}}}N}}}}}}}}}}}}}}~ ~~~~%~-~5~=~E~M~U~]~e~m~u~}~~~~~~~~~~~~~~~~~~ %-5=EMU]emu}N'/7?GOW_gowǀπ׀߀'/7?GOW_gowNɁсف !!)19AIQYaiqyɂтق !!)19AIQYaiqy 0@كN #+3;CKS[ccks{Ä˄ӄۄ #+3;CKS[ccks{Å˅ӅNۅ݅N'/7?GOW_ggowdžφ׆߆'/7?GOW_ggowLJχׇ߇߇N !)19AIQYaiiqyɈшو !)19AIQYaiiqyɉщىN #+3;CKS[ckks{Êˊӊۊ #+3;CKS[ckks{ËˋӋۋN %-5=EMU]emmu}Ō͌Ռ݌ %-5=EMU]emmu}ō͍Սݍ 0NÎˎӎێ #+3;CCKS[cks{ÏÏˏӏۏ #+3;CCKS[cks{NNŐǐϐאߐ'/7?GGOW_gowǑǑϑבߑ'/7?GGOW_gowNǒɒђْ !)19AIIQYaiqyɓɓѓٓ !)19AIIQYaiqyNɔ˔Ӕ۔ #+3;CKKS[cks{Õ˕˕ӕە #+3;CKKS[cks{ÖÖN˖͖Ֆݖ %-5=EMMU]emu}ŗ͗͗՗ݗ %-5=EMMU]emu} 0ŘNșЙؙ  (08@HPX`hpxȚКؚ  (08@HPX`hpxNNě̛ԛܛ $$,4<DLT\dlt|Ĝ̜Ԝܜ $$,4<DLT\dlt|NƝΝ֝ޝ&&.6>FNV^fnv~ƞΞ֞ޞ&&.6>FNV^fnv~ 0NlnnnNnpppNprrrNrttt 0t 0?N    0 N΢ТТТ 0ТN 0NQSSS 00SN 0<Näää 0äN 0>Nååå 0våN9;;; 0`;N 0`N 0`N_aaa 0`aNççç 0`çN#%%% 08%N]___ 0_N1333 023 0eNaccc 0:cN 0YN 0_NY[[[ 0 [Neggg 0jgNѭӭӭӭ 0ӭNWYYY B*YN D/N D+N 0Ngiii D/iN 0Nhjjj 06jN 0N 0:N 0^NFHHH 0<HNNNN 0N*,,,N,...N.000N0222N2444N4666N6888N8:::N:<<<N<>>>N>@@@U6@U>vU*UB޵z  BB B Bܶ B* BBH B*p 0D$ 0lh 0DԻԻԻԻԻԻԻԻԻԻԻU,3|x? Triple+l d d d,X($u,KKIUphaser4 TX140  0    6k$Triangle0      (O$        02Ap`Helvetica #AZ*Humanst521 Cn BT                 *+ (_2623  ..                ,               aaaa'dxd dcdef`C<<CLevel 1Level 2Level 3Level 4Level 5(3$ !  (2$ !  ($$   1  z,s TABLE ATABLE A  d<:Default ParaLevel 1Level 2Level 3Level 4Level 5*G+M (_25   /%` ` hp x /23  ..  5+ ` hp x 5  *D+M (_24   ," hp x ,23  ..  5+ ` hp x 5                   *A+M (_23  ` ) hp x )23  ..  5+ ` hp x 5                  ,  AZArial                                ,@0AZArial                                                                                                                                 *>+M (_22   &hhp x &23  ..  5+ ` hp x 5                                                                  *;+M (_21   #p x #23  ..  5+ ` hp x 5                                                                  %j\AZ"Arial Narrow                                                                 *8+M (_20  h  p x 23  ..  5+ ` hp x 5                                                                  *5+M (_19   pp x 23  ..  5+ ` hp x 5  *2+M (_18    x 23  ..  5+ ` hp x 5                                                                  * (_1723  Ԁ                                                                                                                                                                                                                                                                Level 1Level 2Level 3Level 4Level 5                                                                Level 1Level 2Level 3Level 4Level 5                                                                Level 1Level 2Level 3Level 4Level 5                                                                *GM (_16   /%` ` hp x /23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  *DM (_15   ," hp x ,23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  *AM (_14  ` ) hp x )23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  *>M (_13   &hhp x &23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                  *;M (_12   #p x #23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  *8M (_11  h  p x 23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  *5M (_10   pp x 23  Ԁ  5+ ` hp x 5                                                                  (2M &_9    x 23  Ԁ  5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  ( &_823                                                                                                                                                                                                                                                                                                                                  (GM &_7   /%` ` hp x /23   5+ ` hp x 5                                                                                                                                                                                                                                                                                                                                  (DM &_6   ," hp x ,23   5+ ` hp x 5                                                                                                                                                                                                  (AM &_5  ` ) hp x )23   5+ ` hp x 5  (>M &_4   &hhp x &23   5+ ` hp x 5  (;M &_3   #p x #23   5+ ` hp x 5  (8M &_2  h  p x 23   5+ ` hp x 5  (5M &_1   pp x 23   5+ ` hp x 5  &2M $_    x 23   5+ ` hp x 5  0.Normal<:Definition T<AA:Definition L , 5+ ` hp x 5  5+ ` hp x 586Definition(''&H1      (&H2  (&H3  (&H4 XXX (&H5  (&H6  20Address8MM6Blockquote , , 5+ ` hp x 5   5+ ` hp x 5,*CITE,dl*CODEKf<6X9`(Courier NewKSg\  `&Times New RomanS42Emphasis64Hyperlink    <:FollowedHype    4go2Keyboard Kf<6X9`(Courier NewKSg\  `&Times New RomanS <:Preformatted/%  ,Kk %#/Kf<6X9`(Courier NewKSg\  `&Times New RomanS/%  ,Kk %#/<:zBottom of 7(X7  ?h%2A`Arial?  Sg\  `&Times New RomanS7(X7&  d d)1dxd'dxd<:zTop of For7 (X7  ?h%2A`Arial?  Sg\  `&Times New RomanS7(X7)2dxd0KS.SampleKf<6X9`(Courier NewKSg\  `&Times New RomanS0.Strong 8dl6TypewriterKf<6X9`(Courier NewKSg\  `&Times New RomanS42Variable: 8HTML MarkupB      2 0CommentB  :228Body Text 2         <6X9`(Courier New\  `&Times New Roman%2A`Arial\  `*Times New RomanTThttp://www.emc.nceo.noaa.gov/resarch/osse/swa/sys_errors.htm>Thin/Thick 2[ -d d>Thin/Thick 3l <d d&p d Pd>Thick Double Pd Pd&P d d f d"  ________________X,tXX,XGXX,tAa) dE1 A 22 *Corresponding_Ԁauthorsaddress:Michiko_Masutani_,  _NOAA/_NWS_/NCEP/_EMC_(_RSIS_),_Ԁ5200AuthRd.Camp ss Springs,MD20746.Email:michiko.masutani@noaa.gov.#X,tXG=# XX,t(3$ !  ((3$ !      0  (#$  0  ,@0AZArial !  _ X,tXX,X12.6    UXX,tUImpactAssessmentofaDopplerWindLidarforOSSE/NPOESS #U##X,tXU|#  Michiko_Masutani_*,JohnC.Woollen,StephenJ.Lord,JohnC.Derber 4 NOAA/_NWS_/NCEP/_EMC_,CampSprings,MD  G.David_Emmitt_,SidneyA.Wood,StevenGreco   SimpsonWeatherAssociates,Charlottesville,VA |  Ѐ b  JosephTerry,RobertAtlas H   NASA/_GSFC_,_Greenbelt_,MD .~  Ѐ d  ThomasJ._Kleespies_,_Haibing_ԀSun J  NOAA/_NESDIS_,CampSprings,MD 0  http://www.emc.noaa.gov/research/osse   i  XX,t(,(#(#(  1.INTRODUCTION   x   ThefutureNational_POES_ԀSystem(NPOESS)is & scheduledtoflyduringthe2007-2010period.Forthe  next10years,aconsiderableamountofeffortmusttake *z placetodefine,developandbuildthesuiteofinstruments $ whichwillcomprisetheNPOESS.Theforecastimpactof ~ currentinstrumentscanbeassessedbyObserving (x SystemExperiments(_OSEs_),inwhichalreadyexisting " observationsaredeniedoraddedtoobservationsfroma | standarddatabase.However,theimpactoffuture &v instrumentsmustbeassessedwithexperimentsusing   simulatedobservations.Theseexperimentsareknown z asObservingSystemSimulationExperiments(OSSEs) $t (Lordetal.1997).VV)V    ForeachOSSE,alongintegrationofan x  atmosphericgeneralcirculationmodel(GCM)isrequired "r! toprovidea"trueatmosphere"fortheexperiment.Thisis " calledthe"naturerun"(NR).Thenaturerunneedstobe v # sufficientlyrepresentativeoftheactualatmospherebut  !p$ differentfromthemodelusedforthedataassimilation. !% Theobservationaldataforexistingandfutureinstruments t"& issimulatedfromNRandimpacttestsareperformedfor #n' bothrealandsimulateddata.Thenaturerun,thedata #( assimilationsystemandforecastmodelusedinthese r$) experimentsaredescribedin_Masutani_Ԁetal(2002a). %l * wbtA1-j|T&" `@E+""+"w  AmongvariouscandidateinstrumentsDoppler %!+ wind_lidar_Ԁ(DWL,Baker1995)dataareproducedasline p&!, ofsight(LOS)windsby_SWA_Ԁusingtheir_Lidar_ԀSimulation 'j"- Model(_LSM_).Bracketingsensitivityexperimentsare '#. beingperformedforvariousDWLtechnologyneutral n(#/ conceptstoboundthepotentialimpact(_Emmitt_Ԁ1999, )h$0 _Emimtt_Ԁetal.2001b).Scanning,andvariousdata )%1  VV      -)7 samplingstrategies,arebeingtestedwiththese x7 experiments.AnalysisimpactofDWLarepresentedin "r8 Lordetal.(2002).Inthispaper,theforecastimpactis 9 presentedforselectedcases.Mainly,thefocusisonthe v: impactofscanning.  p;  2.SIMULATIONOFDWLDATA  t=     Thedetailsofprocedurestosimulate "? observationaldataaredescribedinAtlasandTerry(2002) |@ andreferencesofLord(2002)and_Masutani_Ԁetal. &vA (2002a,2002b).InthispapertheimpactofDWLis  B assessedwithexistinginstruments.However,itis zC importantthattheassessmentisalsodonewiththemore $tD advancedinstrumentsexpectedwhenDWLwouldbe E actuallylaunched.Higherdensitycloudmotionvectors xF (_CMVs_)andmoreadvancedsounders,suchas "rG AtmosphericInfraredSounder(AIRS),willbeincludedin H impacttheassessment. v I 0  02(#(# 2(#2(# 2.1SimulationofDWLdata !K     ThesimulationofDWLdataincludeseffortswith %#uM DWLperformancemodels,atmosphericcirculation #N modelsandatmosphericopticalmodels(_Emmitt_Ԁ1999, y$O _Emmitt_Ԁetal.2001b).Theinstrumentparametersare #%s P providedbytheengineeringcommunity.Scanningand %!Q samplingrequirementsareprovidedbythescience w&!R communityanddefinevariousinstrumentscenarios. !'q"S Thesescenariosaretestedinitiallybyexaminingthe '#T sensitivityofanalysestothevariousscenarios.A u(#U candidateDWLconceptisthenchosenforafullOSSE, )o$V andanimpactstudyisconductedandevaluatedbya )%W technologyneutralgroup. s*%X   ThebracketingOSSEsarebeingperformedfor +m&Y variousDWLconceptstoboundthepotentialimpact. +'Z LaterOSSEswillbeperformedformorespecific q,'[ instruments.Thefollowing technologyneutral -k(\ observationcoverageandmeasurementerror -)] characterizationswillbeexplored .  EXP1(Best): UltimateDWLthatprovidesfulltropospheric h LOSsoundings,cloudspermitting.   EXP2(_PBL_+cloud): Aninstrumentthatprovidesonly  p windobservationsfromcloudsandthe_PBL_. $ Ѐ0  ~ VV  EXP3(Upper): Aninstrumentthatprovidesmidand ( x uppertroposphericwindsonlydowntothelevelsof  ,  significantcloudcoverage.     Exp4(NonScan): Anonscanninginstrumentthat  *  providesfulltroposphericLOSsoundings,clouds    permitting,alongasinglelinethatparallelstheground 8  track. 2   TargetedResolutionVolume(_TRV_ ):200kmx200km 6  xT :    T:Thicknessofthe_TRV_Ԁ     0.25kmifz<2km,1kmifz>2km >  Ѐ0  0.25kmforcloudreturn8VV  Swathwidth: 2000kmexceptforEXP4(nonscanning) <   Nomeasurementerrorisassignedfortheinitial  test.Strategiesforsystematicerrorsarediscussedby D _Emmitt_Ԁ(2000a).Onemeasurementisanaverageof > manyshots.Dataproductsbaseduponclusteredand  distributedshotsaregeneratedforeachexperiment.The B clustereddataproductisbaseduponaveragingthe < observationsassociatedwithshotsclusteredwithinan  areathatisverysmallcomparedtothebaseareaofthe @  _TRV_.Thedistributeddataproductisbasedupon :! averagingtheobservationsofshotsdistributedthroughout " the_TRV_Ԁaswouldresultfromcontinuousconical ># scanning. 8$   Distributedshotsforthenonscanexperiment % (EXP4)arenotrealistic.However,itisusedtotestthe <& penetrationthroughcloud.Intherealatmosphere,cloud 6' hasporositywhichisnotdescribedintheNRarchive. ( CloudporosityletssomeDWLshotspassthroughthe : ) cloud.ThisnotpossiblefortheNRcloudastheclouds  4* areuniformwithinagridintheNR.Distributed !+ measurementscollectmanyshotswithinthe_TRV_Ԁand 8", thereismorechanceofpenetratingtheatmosphere.This "2- doesnotexactlymodeltheporosityofthecloudbutitis #. usedtocheckthepenetrationduetoporosity. 6$/   EXP2andEXP3aresimulatedtotestvarious $0 0 wavelengthsandinstruments.Theyaretestedbutnot % 1 presentedinthispaper. 4&!2  3.IMPACTASSESSMENTFORDWLWINDS '"4    Priortotestingfutureinstruments,dataimpact (6$6 testsofexistinginstrumentsareperformedtocalibrate )$7 OSSEs(_Masutani_Ԁetal.2002a).Theresultsshowthat :*%8 therearereasonableagreementsbetweensimulatedand *4&9 realdataimpactbuttheinterpretationneedscaution. +&:   Amongmanycandidateinstrumentsforthe 8,'; OSSE,DWLwindsaresimulatedby_SWA_.Accordingto ,2(< thestrategyforbracketingsensitivityexperiments(Lordet -(= al.2001a,Lordetal.2001b,_Masutani_Ԁetal.2001), 6.)> scanningornonscanning,variouswavelengths,numbers > ofLOSpermeasurement,arebeingtested.Sensitivityto Z? weightinthedataassimilationhasbeentested. T@   Forfirstfewdays,morethan20casesaretested A withvariouscombinationsandselectedcasesare XB completedforthewholeOSSEperiod(00zFebruary13 RC 00zMarch7,1993).Analysisimpactsforthewhole D periodareevaluatedfor13cases(Lordetal.2002).In V E thispapertheforecastimpactofeightexperimentsis  PF presented.Experimentsdiscussedinthispaperarelisted  G inTable1.Thedistributeddataforthenonscanning T H scenarioisnotrealistic.However,itisusedtotestthe  NI effectofpenetration.Becauseoftheaveragingofeach  J 200Kmsquarearea,moreDWLshotspenetratetolower R K levelsfordistributedshots.Theamountofpenetrationis  L L stillanunknownquantityandneedstobeinvestigated.  M Forclusteredshots,representativenesserror7m/sis P N assignedwhile1m/sisassignedtodistributedshots.This J O istomodelthatabout50timeshotsareinvolvedin  P distributedshotscomparedtoclusteredshots.Inthe N Q analysisimpact,theimpactwithrepresentativenesserror H R 7m/sisabout1020%lessthanthatof1m/s,butthe  S geographicaldistributionoftheimpactdoesnotchange. LT   InTable2,thecorrelationbetweenNRand72 FU hourforecastfieldsarepresented.Comparedtocontrol V experiments,anyDWLdataimprovedthewindfields JW globallyatalllevelsforallexperiments.Theforecast DX impactissimilartotheanalysisimpact.Major Y improvementsareoverthetropicsifT1Bisincludedin HZ _CTL_.Marseilleetal.(2001)showedmajorimpactin_SH_, B[ becauseintheirexperiment_CTL_ԀdoesnotincludeT1B. \ IfT1Bareincluded,themajorimprovementin_SH_Ԁhas F] alreadybeenachievedbyT1Bandthemajor @^ improvementsduetoDWLoccursinthetropicsinstead. _ ЀHoweverwithoutT1B,significantimprovementis D` achievedinthe_SH_ԀevenintheworstcaseofDWL >a (Dex4cr7).AlthoughT1BandDex4cr7showsimilar b magnitudeofimpactin_SH_ԀandminimumimpactinNH, Bc therearesignificantdifferencesbetweenexperiments <d withT1BandexperimentwithDex4cr7(Lord2002). e Therefore,bothT1BandDex4cr7togetherallowafurther @f improvementtobeachieved.InNHneithertheT1Bnor :g Dex4cr7producesignificantimpact.Significantimpact,  h whichiscomparativetoRAOBwinds,isachievedinthe >!i bestcaseofDWLwithscanningdistributeddata. !8j     4.COMMENTSANDFURTHERPLANSFORDWL #6m IMPACTTEST  2  $n   DWLisevaluatedwiththe1993datadistribution. %H!p However,DWLwindsalsoneedstobeevaluatedwith &!q boththecurrentdatadistributionandthefuturedata L'"r distributioncorrespondingtowhentheDWLdatawillbe 'F#s actuallyused. (#t   Inthispapernomeasurementerrorisincluded J)$u intheDWL.Systematicerrorsarediscussedby_Emmitt_ )D%v (2000a)andotherlargescalecorrelatederrorneedtobe *%w designedandaddedtotheassessment.Varioussampling H+&x strategiessuchastheseparationbetweenforwardand +B'y backwardscan,andadaptiveobservationsneedtobe ,'z tested. F-({   InthispaperonlyresultsfromUarepresented. -@)| Theimpactonmeridionalwind(V)issimilartothatonU.  ЀImpactintemperaturefieldsismoresensitiveand Z complicated.Impactontemperaturefromradiancedata T andRTempinvolvemanyproceduresthatalterthe  results,suchasthebiascorrection.Impacton X temperaturefromDWLwindisevenmorecomplicated R becausebalancebetweentemperatureandwindsinthe  dataassimilationsystemis_imvolved_. V    Itisfoundthatsurfacedataaretoooptimisticin  P simulationexperimentsbecauseNRsurface    characteristicsaretoosimplecomparedtothereal T   surfacedata.Therefore,impactofotherdata,including  N  T1BandDWL,areunderestimatedinthisOSSE.More    realisticerrorforsurfacedataarebeingevaluated. R     Exp2andExp3arealsobeingtestedtoevaluate  L  differenttypesofinstruments.TheOSSEdata   assimilationsystemwillbeupgradedto2002operational P  system.Withnewsystem,AIRSdataandhighdensity J  _CMV_ԀwillbeanalyzedwiththeDWLdata.Moredetailsof   thefutureplanningarediscussedin_Masutani_Ԁetal. N  (2002b).AIRSdataisbeingsimulated(Goldbergetal. H  2001)andsimulationof_CMV_Ԁisintheprocessoffinal   adjustment(_OHandley_Ԁ2001and_Atlasand_ԀTerry2002). L   ACKNOWLEDGMENT  T   WereceivedmuchassistancefromtheData  ServicesSectionandDr.Anthony_Hollingsworth_Ԁof \ ECMWFinsupplyingthenaturerun.Throughoutthis V projectNOAA/_NWS_/NCEP,NASA/DAOand  NOAA/_NESDIS_Ԁstaffsprovidedmuchtechnical Z assistanceandadvice.Especially,wewouldliketothank T  W.YangandR._Treadon_ԀofNCEP,G._Brin_,S.Bloomand ! N.WolfsonofDAO,andV.Kapoor,P.LiandW.Wolfof X" _NESDIS_.Drs.E._Kalnay_,W.Baker,J._Yoe_ԀandR.Daley R# providedexpertadvice.Weappreciatetheconstructive $ commentsfrommembersoftheOSSEReviewPanel. V% ThisprojectissponsoredbytheIntegratedProgram P& Office(IPO)forNPOESSandbytheNOAAOfficeof ' AtmosphericResearch(OAR)andtheNOAANational T( EnvironmentalSatellite,DataandInformationService N) (_NESDIS_).WethankDrs.StephenMango,Alexander  * MacDonald,JohnGaynor,Jim_Ellickson_ԀandJohnPereira R!+ fortheirsupportandassistanceinthisproject.  !L, REFERENCES  Z#. 0   Atlas,R.1997:Atmosphericobservationandexperiments $^/ toassesstheirusefulnessindataassimilation. $ 0 J.Meteor.Soc.Japan, 75 ,111130. h b% 1VV 0   Atlas,RandJ.Terry2002:ObservingSystemSimulation &f!2 ExperimentsatNASA. &"3 http://dao.gsfc.nasa.gov/DAO_people/terryj'"4VV 0   Baker,_W.E._,G.D._Emmitt_,F.Robertson,_R.M._ԀAtlas,_J.E._ (d#5 Molinari,D.A._Bowdle_,J._Paegle_,_R.M._Ԁ_Hardesty_, ($6 _R.T._Ԁ_Menzies_,_T.N._Ԁ_Krishnamurti_,R.A.Brown, h)$7 _M.J._ԀPost,_J.R._ԀAnderson,A.C._Lorenc_ԀandJ. *b%8 _McElroy_,1995:_Lidar_-measuredwindsfrom * &9 space:Anessentialcomponentforweatherand f+&: climateprediction.Bull.Amer.Meteor.Soc.,76, ,`'; 869-888., (<VV 0   Becker,B.D.,H.Roquet,andA._Stofflen_Ԁ1996:A d-(= simulatedfutureatmosphericobservation.^)>VV databaseincluding_ATOVS_,_ASCAT_,andDWL. > BAMS, 10 ,22792294.0(#(#  Z?(#(# 0   _Emmitt_,G.D.,1999:ExpandedRationaleforthe ^@ IPO/NOAABracketingOSSEs A http://www.emc.ncep.noaa.gov/research/osse/ bB swa/i DWLexp.htm# i 1BDex4dr1  "r ? Y  "r @ Y  "r A Y  "r B D 4* "r C  ?1?41 3)'"r D  ?1 ? 3N  "r E Bestinnonscan  "r F 1BDex4cr7  ` G Y  ` H Y  ` I Y  ` J C 4* ` K  @7@47 3)'` L  @7 @ 3N  ` M Worstinnonscan  ` N Dex4cr7   O N   P Y   Q Y   R C 4*  S  @7@47 3)' T  @7 @ 3N   U WorstcaseofDWLaddedto  V NTV(NoT1B) H W Table1.Experimentsdescribedinthispaper.AllotherconventionaldataincludingRAOBtemperature,ACARdata, X cloudmotionvector,etcareincludedinallexperiments. vY    `  *Pajd dFdd F dd! edd!eKdd!KXdd!Xdd!dd!M dd!M gai(#(#, dd ,dd ,dd ,|dd ,gdd +  "on]n " +o^o +NHU500 +o_o +_SH_ԀU500 +o`o +TRU200 +oao +TRU850 $bo $1B(Control) 7- c ffffffU@85.6ffffffU@785.6 XN*d ffffffU@85.6 ffffffU@ YS@77.4YS@X77.4 XN*e YS@77.4 YS@ fffff&T@80.6fffff&T@X80.6 XN*f fffff&T@80.6 fffff&T@ 9P@64.99P@X64.9 6,*g 9P@64.9 9P@ 6NTV 7- Nh yU@85.9yU@785.9 XN*Ni yU@85.9 yU@ yQ@69.9yQ@X69.9 XN*Nj yQ@69.9 yQ@ fffffS@79.6fffffS@X79.6 XN*Nk fffffS@79.6 fffffS@ 9P@64.99P@X64.9 6,*Nl 9P@64.9 9P@ 61BNWIN 7- <m T@83.5T@783.5 XN*<n T@83.5 T@ fffffS@76.1fffffS@X76.1 XN*<o fffffS@76.1 fffffS@ fffffS@78.6fffffS@X78.6 XN*<p fffffS@78.6 fffffS@ 33333sO@62.933333sO@X62.9 6,*<q 33333sO@62.9 33333sO@ 61BNTMP 7- zr 333333U@84.8333333U@784.8 XN*zs 333333U@84.8 333333U@ 9S@76.99S@X76.9 XN*zt 9S@76.9 9S@ fffffFT@81.1fffffFT@X81.1 XN*zu fffffFT@81.1 fffffFT@ P@66.4P@X66.4 6,*zv P@66.4 P@ 61BDex1dr1 7- w 33333U@86.833333U@786.8 XN*x 33333U@86.8 33333U@ ffffffT@81.6ffffffT@X81.6 XN*y ffffffT@81.6 ffffffT@ fffffU@84.1fffffU@X84.1 XN*z fffffU@84.1 fffffU@ fffffQ@70.1fffffQ@X70.1 6,*{ fffffQ@70.1 fffffQ@ 61BDex1xr7 7- F| U@86.4U@786.4 XN*F} U@86.4 U@ ffffffT@81.6ffffffT@X81.6 XN*F~ ffffffT@81.6 ffffffT@ 33333T@83.833333T@X83.8 XN*F 33333T@83.8 33333T@ P@67.9P@X67.9 6,*F P@67.9 P@ 61BDex4dr1 7- 4  yU@85.9yU@785.9 XN*4  yU@85.9 yU@ 33333S@78.833333S@X78.8 XN*4  33333S@78.8 33333S@ ffffffT@81.6ffffffT@X81.6 XN*4  ffffffT@81.6 ffffffT@ P@67.2P@X67.2 6,*4  P@67.2 P@ 61BDex4cr7 7- r! 33333sU@85.833333sU@785.8 XN*r! 33333sU@85.8 33333sU@ 33333S@78.333333S@X78.3 XN*r! 33333S@78.3 33333S@ 33333ST@81.333333ST@X81.3 XN*r! 33333ST@81.3 33333ST@ lP@65.7lP@X65.7 6,*r! lP@65.7 lP@ 6Dex4cr7 7- " fffffU@86.1fffffU@786.1 XN*" fffffU@86.1 fffffU@ @S@77.0@S@X77.0 XN*" @S@77.0 @S@ 33333ST@81.333333ST@X81.3 XN*" 33333ST@81.3 33333ST@ 33333sP@65.833333sP@X65.8:0." 33333sP@65.8 33333sP@ :Table2. 4$ Anomalycorrelationwiththenaturerunfor72hourforecastfields.ForNHvaluesareaveragedover20Nto80N.For $.  tropics20Sto20N;For_SH_Ԁ80Sto20S.Valuesareaveragedfrom00Z16February1993to12ZFebruary28,1993.For %  every12hours.Theyarepresentedaspercent. 2&!      `     h