Fidelity of Tropical Cyclone Intensity and Structure within Reanalyses
3c. Tegtmeier SPARC WDAC · 2017. 4. 18. · data, despite some issues in JRA-25 and ERA-40 •...
Transcript of 3c. Tegtmeier SPARC WDAC · 2017. 4. 18. · data, despite some issues in JRA-25 and ERA-40 •...
SPARCStratosphere-troposphereProcesses
AndtheirRoleinClimate
SRIP–SPARCReanalysisIntercomparison Project
SusannTegtmeierGEOMAR,Kiel,Germany
SPARC• Promotes research on how chemical and physical processesin the atmosphere interact with climate variability andchange.• Historically concentrated on the role of the stratosphere inclimate, but now includes foci throughout the atmosphere.
SPARCImplementationPlan•Theme1.AtmosphericDynamicsandPredictability•Theme2.ChemistryandClimate•Theme3.Long-termRecordsforClimateUnderstanding
OCTAV-UTLS– ObservedCompositionTrendsAndVariabilityintheUTLSLOTUS– OzoneTrends
AtmosphericTemperatureChangesandtheirDrivers
New/emergingactivities:
SPARCTheme:Long-termRecordsforClimateUnderstanding
• VerticallyandspatiallyresolvedtrendsinSPARC-relevantECVs?
Ø Whichspeciesandstatevariablesareneeded,andwithwhatresolution,frequency,anduncertainty?
Ø Requirementsonobservingprogramme (uncertaintiesandsamplingregimens)forreliabletrenddetection?
• Istheatmosphere(chemistryanddynamics)evolvinginawaythatisconsistentwithourunderstanding?
• Temporalevolutionofglobalandregionalforcing(naturalandanthropogenic)oftheclimatesystem?
• Docurrentobservationstestourknowledgeofatmosphericcompositionanddynamics?Whichobservationswouldprovidemorerobusttests?
WDACActionsitemsfromlastyear
RegardingthesatellitemissiongapitwasdecidedtosendalettertoCEOS/CGMSWorkingGrouponClimate
• Atmosphericcomposition:limbsounders
Regardingthein-situobservingnetworksatriskitwasdecidedthatWDACwillpreparealetterforJSCtoWMO.
• Continuationofstationswithlong-termrecords,especiallylidar andozonesondes• Homogeneityacrossthegroundnetworks,e.g.O3S-DQAforozonesonde,GRUAN.• ObservationsintropicsandSouthernmid-latitudes
FISAPS- FineScaleAtmosphericProcessesandStructures
M.Geller,H.Chun,P.Love
• Focusonatmosphericprocessesimportantforlarge-scaledynamicsandchemicalcompositionthatoccuronverticalscalesoflessthan1km
• Goals:enhanceavailabilityofhighvertical-resolutionradiosondedata(HVRRD)
Radiosondereportssortedbyverticallevels
IncreasingneedforresearchaccesstoHVRRD(currentlygrowinguseinreal-timeforecasting)
EncouragedataproducerstoworktowardshighresolutionBUFRdata
CurrentdiscussionofavailabilitythroughIGRA(verypreliminary)
SRIP- SPARCReanalysisIntercomparison Project
Masatomo Fujiwara(HokkaidoUniv.,Japan),LesleyGray(OxfordUniv.,UK),GloriaManney (NWRA,USA)
http://s-rip.ees.hokudai.ac.jp/
Objectives• Compare all (or some of the newer) reanalysis data sets for key diagnostics• Identify and understand the causes of differences amongst reanalyses• Provide guidance on the appropriate usage of various reanalysis products in
scientific studies• Establish collaborative links between reanalysis centres and the SPARC
community• Contribute to future improvements in reanalysis products
S-RIP focuses on reanalysis outputs in the stratosphere, upper troposphere andlower mesosphere
Prime Output:• S-RIP “interim” Report (2016), four ‘basic’ chapters• S-RIP “full” Report (2018), twelve chapters in total
Motivation• Multiplestudieshaveidentifiedsubstantialdifferencesamongstreanalyses
withrespecttokeyvariablesanddiagnostics.
ReanalysisCentre
Products Contacts for S-RIP
ECMWF ERA-40, ERA-Interim, ERA-20C, [ERA5] Rossana DraganiJMA (&CRIEPI) JRA-25/JCDAS, JRA-55 Yayoi Harada &
Kazutoshi OnogiNASA MERRA, MERRA-2 Krzysztof Wargan
NOAANCEP NCEP/NCAR (R-1), NCEP/DOE (R-2),NCEP-CFSR
Craig Long &Wesley Ebisuzaki
NOAA&Univ.Colorado
20CR Gilbert Compo &Jeffrey S. Whitaker
GlobalAtmosphericReanalysisDataSets
• Focusonnewerreanalysissystemsthatassimilateupper-airmeasurements(ERA-Interim,JRA-55,MERRA,MERRA-2,CFSR)
• ERA5 willnotbeincluded(sincetimeperiod1979-2009notavailableuntil2018)• Whereappropriate,includeotherreanalyses:
• Long-term,surface-inputreanalyses (e.g.NOAA-CIRES20CRandERA-20C)• Olderreanalyses (NCEP-NCARR1,NCEP-DOER2,ERA-40,andJRA-25/JCDAS)
• Theintercomparison periodcommontoallchapters:1979-2012• Somechapterswillalsoconsiderthepre-satelliteeraand/orincluderesultsforrecentyears.
Chapter Title Chapter Co-leads1 Introduction Masatomo Fujiwara, Gloria Manney,
Lesley Gray2 Description of the Reanalysis Systems Jonathon Wright, Masatomo Fujiwara,
Craig Long3 Climatology and Inter-annual Variability
of Dynamical VariablesCraig Long, Masatomo Fujiwara
4 Climatology and Inter-annual Variabilityof Ozone and Water Vapour
Michaela Hegglin, Sean Davis
5 Brewer-Dobson Circulation Thomas Birner, Beatriz Monge-Sanz6 Stratosphere-Troposphere Coupling Edwin Gerber, Patrick Martineau7 Extratropical UTLS Cameron Homeyer, Gloria Manney8 Tropical Tropopause Layer Susann Tegtmeier, Kirstin Krüger9 QBO and Tropical Variability James Anstey, Lesley Gray10 Polar Processes Michelle Santee, Alyn Lambert,
Gloria Manney11 Upper Strato. Lower Mesosphere Lynn Harvey, John Knox12 Synthesis Summary Fujiwara, Manney, Gray
Chapters1-4:Basicchapters2016(nowearly2017)Chapters5-11:Advancedchapters(end2018)
ChaptersoftheS-RIPReport
S-RIP2017InterimSPARCReport(Basicchapters)• Editors:M.Fujiwara,J.Wright,G.Manney,andL.Gray• Technicaleditors:FionaTummon andSPARCOfficecolleagues• Reportstatus:firstrevisionphase• Reviewprocessissimilartothatofjournalpublications
• minimum3-4reviewersperchapter• tworoundsofreviews
• Willbepublishedin2017
S-RIP2018FinalSPARCReport(Advancedchapters)• ZeroorderdraftsofeachchapterinOctober2016• FinaldraftsplannedforOctober2017• Beginningofreviewprocessendof2017
Chapter1:Introduction• Motivation,goals,scope,outline,communication,prospectsetc.
Chapter2:DescriptionoftheReanalysisSystems• Referenceforreanalysisusers• Understandhowreanalyses workusingsimpleterms• Comparethestructuresoftwoormorereanalyses• Sourcesofadditionalinformation,extensiveuseoftables/figures• Informationonpotentialtemporaldiscontinuities
S-RIP2017InterimSPARCReport(Basicchapters)
2.1Introduction
2.2Forecastmodels(includingmajorphysicalparametrizations andboundaryconditions)
2.3 Assimilationschemes(basicsandapplicationsinreanalysissystems
2.4Observationaldata(includingkeyupperairobservations,watervapour)
2.5Executionstreams
2.6Archiveddata
Chapter2:DescriptionoftheReanalysisSystemsJ.Wright,M.Fujiwara,C.Long
Chapter2Tables/Figures• Reanalysisdatasetwithdatesandjournalpapers• Horizontalresolution,verticallevels,modeltop• Radiativetransferschemes,cloudscheme,gravitywavescheme• SST/seaicetreatment,solarvariabilitytreatment• Ozonetreatment,aerosolschemes,CO2 +otherGHGstreatments• Assimilationschemeemployed,satelliteradianceassimilationschemeemployed• Satellitedatasetsandotherdatasets(e.g.aircraft)assimilated• Watervapour treatment
AssimilatedsatelliteradiancedatasetsSimplifiedschematicfordataassimilationstrategies
1980 2015
TOVS
ATOVS
Execution‘streams’employedtoproducethevariousreanalysisdatasets1979-2016
ERA-I
JRA-55
MERRA
MERRA-2
CFSR
• AnintroductiontoS-RIP(motivation,objectives,reportstructure)• descriptionofelevencurrentorrecentreanalyses:
• ERA-40,ERA-Interim,ERA-20C• JRA-25/JCDAS,JRA-55• MERRA,MERRA-2• NCEP-NCARR1,NCEP-DOER2,CFSR,NOAA-CIRES20CRv2
• Forecastmodelspecs(grids,dynamicalcores,parametrizations,boundaryconditions)• Dataassimilationoverview(briefaccountofmethodsemployed)• Inputobservations(conventionalandsatellite-based,changesintime,keydatasourcesandarchives,
qualitycontrolandbiascorrectionprocedures)• Ozoneandwatervapour (relevantparametrizations,detailsofdataassimilation)
AcknowledgementsandthankstothemanyS-RIPparticipantswhoprovidedinputandfeedbackforthispaper— thiswastrulyacollaborativeeffort
S-RIPoverviewpaperinACP:IntroductiontotheSPARCReanalysisIntercomparisonProject(SRIP)andOverviewoftheReanalysisSystems
Fujiwaraetal.
SpecialIssueintheAtmosphericChemistryandPhysics(ACP):
“TheSPARCReanalysisIntercomparison Project(S-RIP)”Editors:P.Haynes,G.Stiller,andW.Lahoz
• IntendedtocollectresearchwithrelevancetoS-RIP
• ParticipationinS-RIPisnotaprerequisiteforsubmission
• 10paperssofar,includinganS-RIPoverviewpaper,with8publishedand5atdiscussionstage
• Fujiwaraetal.(ACP,2017):IntroductionS-RIPandoverviewofthereanalysissystems• Hoffmannetal.,(ACPD,2017)ValidationofmeteorologicalanalysesandtrajectoriesintheAntarcticlowerstratosphere
usingConcordiasi superpressure balloonobservations• Nützel etal.,(ACP,2016)Movement,driversandbimodalityoftheSouthAsianHigh• Boothe andHomeyer (ACPD,2016)Globallarge-scalestratosphere-troposphereexchangeinmodernreanalyses• KimandChun,(ACP,2015)Momentumforcingofthequasi-biennialoscillationbyequatorialwavesinrecentreanalyses• Fujiwaraetal.(ACP,2015):Globaltemperatureresponsetothemajorvolcaniceruptionsinmultiplereanalysisdatasets• Kawatani etal.(ACP,2016):Representationofthetropicalstratosphericzonalwindinglobalatmosphericreanalyses• Miyazakietal.(ACP,2016):Inter-comparisonofstratosphericmean-meridional circulationandeddymixingamongsix
reanalysisdatasets• Friedrichetal.(ACP,2017):AcomparisonofLoonballoonobservationsandstratosphericreanalyses products• Wunderlich &Mitchell(ACP,2017):Revisitingtheobservedsurfaceclimateresponsetolargevolcaniceruptions
• Reanalysestemperaturevariabilitywithtime• Reanalysesensemblemean• Intercomparison ofthereanalyses(temperature,zonalwind)• Polarannualtemperaturecycle• Comparisonwithindependentobservations(MLS,HIRLDS,Ozonesondes,Balloons,Rocketsondes)
• Effectsofvolcaniceruptionsonreanalysestemperaturesandwind• Majorconclusions
• Reanalyes shouldNOTbeusedfortrends.• Thereisnoone'perfect'reanalyses. Theyallhaveissues.• Reanayses havetemperatureandwindbiasesinthemiddleandupperstratospherewithrespect toeachother.Witheachgenerationofreanalysesthesebiaseshavebeengettingsmaller.
• Reanalyses allhave'issues'gettingthetropicalstratosphericwindscorrect.
Chapter3:ClimatologyandInterannual VariabilityofDynamicalVariables
C.Long,M.Fujiwara
• Reanalysestemperaturevariabilitywithtime
• Reanalysesensemblemean
• Intercomparison ofthereanalyses(temperature,zonalwind)
• Polarannualtemperaturecycle
• Comparisonwithindependentobservations(MLS,HIRLDS,Ozonesondes,Balloons,Rocketsondes)
• Effectsofvolcaniceruptionsonreanalysestemperaturesandwind• Reanalysistemperaturesagreewellbelow10hPa• Largediscontinuitiesassociatedwiththe~1998transitionfromTOVS(SSU+MSU)toATOVS(AMSU)• CFSR:severaldiscontinuitiesassociatedwithproductionstreamchangesandthechangeinmodelresolution• CFSRmodelimpartsawarmbiasintheupperstratosphere• JRA-55containsayear-roundcoldbiasintheupperstratosphere• ERA-Interimisbiasedwarminthesouthernpolarvortexnear7~5hPa• MERRAcontainsaslightcoldbiasat1hPa relativetoAuraMLS• AssimilationofAuraMLS temperatureretrievalsatpressureslessthan5hPa causesajumpinMERRA-2upper
stratospherictemperatures• Disagreementsamongstreanalyses arereducedduringtheATOVSperiod(post-1998)relativetotheTOVSperiod• Significantbiasesinearlierreanalyses (JRA-25,ERA-40,NCEPR1/R2)arelarger,morepersistent,andextend
throughadeeperverticallayer• EvaluationsagainstAuraMLS,ozonesondes,long-durationballoonexperiments,andun-assimilatedradio- and
rocketsondes
Chapter3:ClimatologyandInterannual VariabilityofDynamicalVariables
C.Long,M.Fujiwara
Global Temperature Anomalies1979-2016
CFSR MERRA
MERRA2ERA-I
JRA55Global temperature anomalies reveal discontinuities in the middle and upper stratosphere due to stream and data transitions. Most notably is the transition from TOVS to ATOVS observations in late 1998. Other features observed are: El Nino warm period in troposphere in 1998, 2010 and 2016, and the lower stratospheric warming due to El Chichón and Mt Pinatubo in 1982 and 1991, respectively.
Equatorialzonalwinds(10°S–10°N) anddifferencefromensemble-mean(ERA-I+JRA-55+MERRA)
CFSR
MERRA
ERA-Interim
JRA-55
MERRA-2
Seasonal cycle Time series(1979–2015)
Jan Dec
1hPa
1979 2015
1000
Overall1.Continuedimprovementinmeteorologicalreanalyses andpastrecords2.Continuationofexistingcoremeasurements– realfundingpressure
Specific• Lackofplannedsatelliteobservations(esp.limb)ofUTScomposition• Needup-to-dateAMSUandmergedSSU-AMSUclimatedatarecords• Availabilityofhighvertical-resolutionradiosondedata• Needformorereference-qualityglobal&long-termobservations,particularlyforreanalysisintercomparisons
• Noplannedcontinuationofmesosphericradiancefortemperatures•Needforquickresponsefieldcampaignsaftervolcaniceruptions• Datasharingisachallengeinthe‘AsianMonsoonregion’
SPARCDataneedsandrequirements
Chapter4: Climatologyandinterannualvariabilityofozone
• Treatmentofozone• AllofthereanalysissystemsexceptforNCEPR1andR2includesomeformofprognosticozone
parametrizationandanalysis,butnoneoftheseincludeheterogeneouschemistry(usewithcautionforozoneholestudies)
• Todate,onlysatelliteozoneretrievalsareassimilated(noozonesondes)• Inputobservationsvarywidely,fromtotalcolumnonly(JRA-25andJRA-55)tocoarse-
resolutionprofileinformation(ERA-40,MERRA,CFSR)toabroadselectionofsatelliteproducts(ERA-Interim)
• ECMWFreanalysesuseclimatologiesratherthananalysesforradiationcalculations
• Keyconclusionsandobservationalevaluation• Climatologies,annualcyclesandinterannualvariabilitygenerallyagreewellwithobservational
data,despitesomeissuesinJRA-25andERA-40• Totalcolumnozoneismostlycapturedbyreanalyses,withsomelimitations(e.g.,noColumn
Ozone dataduringpolarnight)• Theozoneverticaldistributionisweaklyconstrainedbydataassimilation,sothatmeanbiases
inozoneproductsvarywithheight(10~50%instratosphere)• QBOsignalsarevastlyimprovedinJRA-55relativetoJRA-25• Allreanalysesfailtocapturethe“ozonevalley”associatedwiththeAsianmonsoonanticyclone
Height-latitudecolumnozoneclimatology:differencefromSBUV(DU)AlsoshownistheTOMS/OMIdifferencewithSBUV.
Ozoneverticaldistribution:%differencesofannual-av climatology(2005-2010)fromtheSPARCDataInitiativemulti-instrumentmean(MIM;i.e.Ri-MIM)/MIM*100.)
1hPa
100
S N
CFSR JRA-25 MERRA
ERA-I JRA-55 MERRA-2
OzoneQBOsignal2005-201010oS-10oN
2005 2010
1hPa
100hPa
Obs
ERA-I
JRA-55
JRA-25
CFSR
MERRA-2
MERRA
Chapter4: Climatologyandinterannualvariabilityofwatervapour
• Treatmentofwatervapour• ReanalysisWVisafunctionofdehydration,transport,andforsome
reanalysesmethaneoxidation(ERA-40,ERA-Interim,ERA-20C),icesupersaturation(ERA-InterimandERA-20C),and/orrelaxationtoclimatology(MERRAandMERRA-2)
• Dataassimilationhaslimitedornoinfluenceinthestratosphere(specificationandtreatmentoftheupperboundvaries)
• Fieldsusedforradiationcalculationsinthestratosphereoftendifferfromreanalysisproducts
• Keyconclusionsandobservationalevaluation• ModerateagreementamongstreanalysesintheUTLS• MajorissuesinJRA-55atmid–highlatitudesandpressureslessthan200
hPa• CFSRisverydry(withsomenegativemixingratios)inthestratosphere• Comparisonoftransportsimulationsandtaperecordersignalsreveals
severalimportantdiscrepanciesincirculationandtemperatures• ReanalysisstratosphericWVshouldnotbeusedinscientificstudies!
Multi-annualmeanzonalmeancross-sectionsofwatervapour (2005-2010)forSPARCDataInitiativemulti-instrumentmean(MIM;upperright)andvariousreanalysisdatasets
0.1hPa
100hPa
S N
Obs
CFSR MERRA
MERRA-2
JRA-25
JRA-55ERA-I
S-RIP2017workshop• Place:
• EuropeanCentreforMedium-RangeWeatherForecasts,ECMWF,U.K.• When (withjointDAworkshop)
• 23-27October2017
• Host:ECMWF• BeatrizMonge-Sanz,Rossana Dragani
• Facilities• LectureTheatre,meetingrooms,posterboards• Wi-Fi,cafeteria,restaurant• Parkingonsiteorpublictransportfromtown• Manyhotelsintown,offerspecialECMWFrates• Londonisreasonablyclosefor“commuting”• Twomajorairports(Heathrow,Gatwick)
• Costsestimate• Nocostformeetingroomsandequipment• Coffeebreakandlunchatextracostà registrationfee/smallSPARCcontribution
andself-fundedlunch?
Figure 1: Plot showing current isentropic level offerings from CFSR (green), ERA-I (blue), JRA-55 (purple), along with our minimum (light gray), reduced (dark gray) and full (black) recommendations up to 1500 K.
Figure 2: Plot showing sampling of our minimum (light gray), reduced (dark gray) and full (black) recommendations above 1000 K.
Motivation for Above Recommendations The full, reduced, and minimum recommendations outlined above are designed with balance in mind such that reanalysis centers could independently choose from among the three and still provide many overlapping levels throughout the atmosphere, as well as providing data for previously unrepresented regions (i.e., the USLM). As a result, two evident motivations come to mind:
1. While it would be ideal to have all datasets use the same isentropic levels, these recommendations should give the data-providers some leeway to fit their own time/space constraints.
2. There is plenty of overlap between each recommendation, which will facilitate intercomparisons, and encourage data-users to make use of multiple datasets in their analyses and studies.
We also note that there are other considerable benefits of data-providers making more isentropic levels available. One reason, in particular, is that the data-providers/reanalysis centers already have the
Figure 1: Plot showing current isentropic level offerings from CFSR (green), ERA-I (blue), JRA-55 (purple), along with our minimum (light gray), reduced (dark gray) and full (black) recommendations up to 1500 K.
Figure 2: Plot showing sampling of our minimum (light gray), reduced (dark gray) and full (black) recommendations above 1000 K.
Motivation for Above Recommendations The full, reduced, and minimum recommendations outlined above are designed with balance in mind such that reanalysis centers could independently choose from among the three and still provide many overlapping levels throughout the atmosphere, as well as providing data for previously unrepresented regions (i.e., the USLM). As a result, two evident motivations come to mind:
1. While it would be ideal to have all datasets use the same isentropic levels, these recommendations should give the data-providers some leeway to fit their own time/space constraints.
2. There is plenty of overlap between each recommendation, which will facilitate intercomparisons, and encourage data-users to make use of multiple datasets in their analyses and studies.
We also note that there are other considerable benefits of data-providers making more isentropic levels available. One reason, in particular, is that the data-providers/reanalysis centers already have the
RequestforPVonIsentropicSurfaces:conductaquicksurveytoascertainthattheserecommendationswillprovidewhatisrequiredbythecommunity?
CommentsreceivedattheSSGmeetingreportedbyLesleyGray
• TherewassomediscussionabouttherelationshipbetweenS-RIPandthetroposphericreanalysisintercomparisonthatmightfallundertheTIRAinitiative.• Therewasgreatsupportforthereanalysisinter-comparisonsgoingonwithinS-RIPanda
needtoextendthemmoretodirecttroposphericcomparisons(e.g.howwelldothereanalyses comparewithregardtospecifictroposphericfeaturessuchascirculationpatternsandfeatureslikethemonsoons)- anditwasfeltthatthiswaswhatTIRAwouldbeplanning.
• Therewassupportfortheideaofcollectinguserrequirementse.g.PVonisentropicsurfacesandfeedingthemthroughtotheReanalysisCentres.
• Therewasstrongencouragementforourideaofpullingtogetherthevarioustablesofinformationoneachreanalysisdataset,toencouragetheCentres toprovideastandardsetofinformationforeachfuturedataset.Theyhelpfullysuggestedthatthismightbeturnedintoa'livingdocument'onaweb-site(SPARC?S-RIP?)thatcouldcontinuetobeupdatedintothefuture.
• FionaTummon saidsheishappytobethehandlingeditor,sothatreviewerscanbeanonymous.