WRFAtmosphericDataAssimila3on:lessonslearnedfromASR

NCAR/MMM:ZhiquanLiu,Hui‐ChuanLin

UCAR/COSMIC:Ying‐HwaKuo,Tae‐KwonWee

OSU/BPRC:DaveBromwich,LeshengBai,KeithHines,Sheng‐HungWang

WCRP/ICR4‐May9,2012 1

Outline

•  IntroductionofWRFDASystem

•  SeasonalvariationofWRFmodelerroratstratosphere

•  RadiancebiascorrectionstrategyforregionalDA

WCRP/ICR4‐May9,2012 2

3

WRFDA(formerlyWRF‐Var):AcommunityfacilityBarkeretal.2012,BAMS

•  DevelopedandsupportedbyNCAR/MMMwithcontributionsfromresearchcommunity

•  Includeoptionsfor3DVAR,4DVAR,andhybridVar/Ensemble–  3DVARusedforASR

•  Canassimilateconventionaldata,GPSRO(refractivity),satelliteradiances,radar,precipitation(4DVARonly)

–  Variationalbiascorrection(VarBC)forradianceDA

•  Controlvariables–  Streamfunction,unbalanced(velocitypotential,T,Ps),andpseudoRH

•  NotdirectlyanalyzeT2m/Q2m/U10/V10–  ThoseareWRFmodeldiagnosticvariables,notprognosticvariables.–  T2m/Q2m/U10/V10obsusedtoanalyzethelowestmodellevels’state.

WCRP/ICR4‐May9,2012

Observa3onsusedinASR•  Surface

–  U/V,T,Q,P:SYNOP,METAR,SHIPS,BUOY,SONDE_SFC,–  U/V:QuikSCAToverocean

•  Upperair–  SOUND(U/V,T,Q),AIREP(U/V,T),PROFILER(U/V),

PolarAMV,GeoAMV(U/V)

•  GPSRadioOccultation(refractivity)

•  Microwaveradiancesfrompolarsatellites

WCRP/ICR4‐May9,2012 4

NCEP BUFR/PrepBufr data provided by Jack Woollen

WCRP/ICR4‐May9,2012 5

synop + metar ship buoy * sound gpsref profiler airep quikscat SatWind

 More than 4000 surface stations

 Around 300 sounding stations

observa3oncoveragesnapshotat2007120100with3‐h3mewindow

WCRP/ICR4‐May9,2012 6

Seasonalvaria3onofmodelforecasterrorsatstratosphere(standardderiva3onsta3s3csfrom24hFC‐12hFCvalidatsame3me)

This is the basis for the background error covariance statistics.

100 hPa

Discon3nuitycausedbyBECchange

WCRP/ICR4‐May9,2012 7

Compare to ERA-Interim.

One of reasons to use grid-nudging near the top.

SatelliteMWradiancedataused(2000~)

WCRP/ICR4‐May9,2012 8

AMSU‐A AMSU‐B MHS

NOAA‐15 X X

NOAA‐16 X X

NOAA‐17 X

NOAA‐18 X X

NOAA‐19 X X

METOP‐2 X X

EOS‐2(Aqua) X

AMSU-A: assimilate channels 5~9. AMSU-B/MHS: assimilate channels 3~5.

Total 12 sensors from 7 satellites

DataAvailability

WCRP/ICR4‐May9,2012 9

LTAN(localtimeofascendingnode)

Launchdate availabledateforASR

NOAA‐15 16:38:14 1998/05/13 Sincethebeginning

NOAA‐16 20:00:43 2000/09/21 2001/01/24

EOS‐2(Aqua) PMorbit 2002/05/04 2004/04/13

NOAA‐17 19:45:41 2002/06/24 2002/09/11

NOAA‐18 14:31:03 2005/05/20 2005/08/23

METOP‐2 21:31:00 2006/10/19 2007/04/24

NOAA‐19 13:31:47 2009/02/06 ??

WCRP/ICR4‐May9,2012 10

Radiance Monitoring against ERA-Interim over ASR domain

WCRP/ICR4‐May9,2012 11

WCRP/ICR4‐May9,2012 12

METOP-2

13WCRP/ICR4‐May9,2012

RadianceMonitoringagainstERA‐InterimGlobalvs.Regionalsta3s3cs

spin-up 6-h cycling exps

BC statistical different over globe & sub-regions

Bias correction coefficients: “offset” evolution from global, ASR-domain, South-Pole domain.

Global statistics

Statistics over ASR-domain Statistics over south-pole domain

Biascorrec3onsensi3vityexperiments

•  Reducedresolution–  60km/57L/10hPatop,nogrid‐nudgingontop,singledomain

•  4experiments–  gts:Non‐radiancedataassimilated

expswithconventionaldataandAMSU‐Aradiancesassimilated–  varbc_cs:uses“cold‐start”VarBCcoeffsatthebeginningof

cycles,i.e.,noknowledgeofcoeffs.

–  varbc_ws:uses“warm‐start”VarBCcoeffsatthebeginningofcycles,i.e.,regionalspun‐upcoefffsfromMay~July.

–  glb_freez:turn‐offVarBC,butupdateBCcoeffseachcyclefromglobalstatistics

WCRP/ICR4‐May9,2012 15

T:analysis@12Zvs.ERA‐Interim

WCRP/ICR4‐May9,2012 16

Model Level

Time (day #)

Q:analysis@12Zvs.ERA‐Interim

WCRP/ICR4‐May9,2012 17

Model Level

Time (day #)

Z:analysis@12Zvs.ERA‐Interim

WCRP/ICR4‐May9,2012 18

Model Level

Time (day #)

T/Qbias:24hFC@12Zvs.ERA‐Interim

WCRP/ICR4‐May9,2012 19

Model Level

Time (day #)

ComparetoASCOSsoundingAug.3~Sep.7,2008

WCRP/ICR4‐May9,2012 20

T/QAnalysisvs.ASCOS

WCRP/ICR4‐May9,2012 21

Bias Bias

RMSE RMSE

T/Q24hFCvs.ASCOS

WCRP/ICR4‐May9,2012 22

Bias Bias

RMSE RMSE

Planfor10‐kmASRproduc3on•  Diagnose30kmASR‐Interimandidentify/fixtheissues

•  ImplementinWRFDAradianceblacklisttableusedbyERA‐Interim–  AlsorefertoNESDISinstrumentmonitoringandNCEPradianceblacklisttable–  http://www.star.nesdis.noaa.gov/smcd/spb/icvs/–  http://www.emc.ncep.noaa.gov/mmb/data_processing/

Satellite_Historical_Documentation.htm

•  Conservativeusageofsurface‐sensitivechannels–  e.g.,useAMSU‐Ach5,AMSU‐B/MHSch5onlyoverwater.

•  PerformradiancemonitoringrungloballypriortoASRproduction–  Pre‐computeglobalBCcoefficientsstatisticsusingERA‐Interim(ina1°x1°grid)

overthewhole11‐yearperiod.

WCRP/ICR4‐May9,2012 23