<%BANNER%>

UFIR



The influences of ENSO and the Subtropical Indian Ocean Dipole on tropical cyclone trajectories in the South Indian Ocean
CITATION PDF VIEWER
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/IR00001200/00001
 Material Information
Title: The influences of ENSO and the Subtropical Indian Ocean Dipole on tropical cyclone trajectories in the South Indian Ocean
Series Title: International Journal of Climatology
Physical Description: Journal Article
Creator: Matyas, Corene ( Researcher )
Publication Date: 2012
Copyright Date: 2012
 Notes
Abstract: El Ni˜no-Southern Oscillation (ENSO) is known to associate with variability of tropical cyclone (TC) trajectories in the southwestern Indian Ocean. However, consideration of ENSO phase alone does not account for all variability of TC tracks within this region. This study demonstrates that the subtropical Indian Ocean Dipole (SIOD) sea-surface temperature anomaly pattern is also significantly associated with variability in southwestern Indian Ocean TC tracks. Hierarchical cluster analysis is employed to group TC trajectories by their initial and final positions. Median monthly values of the Ni˜no-3.4 index and Subtropical Dipole Index corresponding to the life cycles of TCs in each group are compared using non-parametric analysis of variance. The results suggest that both ENSO and SIOD are significantly associated with different types of southwestern Indian Ocean TC trajectories. Furthermore, significant interactions of ENSO and SIOD phases are found to influence certain types of TC tracks using contingency table tests. During simultaneous warm ENSO and negative SIOD phases, TCs moving across the southwestern Indian Ocean tend to follow more southward or southeastward tracks. During neutral or cool ENSO and positive SIOD phases, TCs moving through the southwestern Indian Ocean tend toward more westward trajectories. These findings suggest that use of an SIOD index in addition to an ENSO index could improve intraseasonal to seasonal statistical prediction of southwestern Indian basin TC activity.
Acquisition: Collected for University of Florida's Institutional Repository by the UFIR Self-Submittal tool. Submitted by Corene Matyas.
Publication Status: Published
 Record Information
Source Institution: University of Florida Institutional Repository
Holding Location: University of Florida
Rights Management: All rights reserved by the submitter.
System ID: IR00001200:00001

Downloads

This item is only available as the following downloads:

AshMatyasIJOC2012 ( PDF )


Full Text

PAGE 1

INTERNATIONALJOURNALOFCLIMATOLOGY Int.J.Climatol. 32 :41–56(2012) Publishedonline5November2010inWileyOnlineLibrary (wileyonlinelibrary.com)DOI:10.1002/joc.2249 TheinuencesofENSOandthesubtropicalIndianOcean Dipoleontropicalcyclonetrajectoriesinthesouthwestern IndianOceanKevinD.Asha*andCoreneJ.MatyasbaHazardsandVulnerabilityResearchInstitute,DepartmentofGeography,UniversityofSouthCarolina,709BullStreet,Columbia,SC29208, USAbDepartmentofGeography,UniversityofFlorida,3141TurlingtonHall,Gainesville,FL32611,USA ABSTRACT: ElNi no-SouthernOscillation(ENSO)isknowntoassociatewithvariabilityoftropicalcyclone(TC) trajectoriesinthesouthwesternIndianOcean.However,considerationofENSOphasealonedoesnotaccountforall variabilityofTCtrackswithinthisregion.ThisstudydemonstratesthatthesubtropicalIndianOceanDipole(SIOD) sea-surfacetemperatureanomalypatternisalsosignicantlyassociatedwithvariabilityinsouthwesternIndianOcean TCtracks.HierarchicalclusteranalysisisemployedtogroupTCtrajectoriesbytheirinitialandnalpositions.Median monthlyvaluesoftheNi no-3.4indexandSubtropicalDipoleIndexcorrespondingtothelifecyclesofTCsineachgroup arecomparedusingnon-parametricanalysisofvariance.TheresultssuggestthatbothENSOandSIODaresignicantly associatedwithdifferenttypesofsouthwesternIndianOceanTCtrajectories.Furthermore,signicantinteractionsofENSO andSIODphasesarefoundtoinuencecertaintypesofTCtracksusingcontingencytabletests.Duringsimultaneouswarm ENSOandnegativeSIODphases,TCsmovingacrossthesouthwesternIndianOceantendtofollowmoresouthwardor southeastwardtracks.DuringneutralorcoolENSOandpositiveSIODphases,TCsmovingthroughthesouthwesternIndian Oceantendtowardmorewestwardtrajectories.ThesendingssuggestthatuseofanSIODindexinadditiontoanENSO indexcouldimproveintraseasonaltoseasonalstatisticalpredictionofsouthwesternIndianbasinTCactivity.Copyright2010RoyalMeteorologicalSocietyKEYWORDStropicalcyclones;ENSO;subtropicalIndianOceandipole;trajectories;clusteranalysis;tropicaltemperate troughs Received17June2010;Revised4October2010;Accepted7October20101.Introduction Tropicalcyclones(TCs)arearecurringphenomenonin thesouthwesternIndianOcean,mostfrequentlyduring TCseasonfromNovembertoApril,andtheinhabited regionsofthewesternrimofthebasinareparticularly pronetorepeatedTCimpacts.TCsfrequentlypassnear thesmallislandsofMauritiusandR eunionandcan causeseveredamageandsocietaldisruption,though lossofhumanlifeistypicallymitigatedbythewellorganisedandexecutedTCpreparednessandwarning systemsoftheseislands(Parker,1999;Roux etal ., 2004).MadagascarhasenduredmanydevastatingTC impacts(Jury etal .,1993;NaeraaandJury,1998; Chang-SengandJury,2010b),andthepopulation’s socialvulnerabilitytonegativeimpactsfromTCstrikes isincreasedduetowidespreadlivelihooddependence onagricultureandthelessereconomicstatusofthe country(Brown,2009).Mozambiqueis,likewise,ina precariouseconomicandclimatologicalposition,though *Correspondenceto:KevinD.Ash,HazardsandVulnerability ResearchInstitute,DepartmentofGeography,UniversityofSouthCarolina,709BullStreet,Columbia,SC29208,USA. E-mail:ashkd@email.sc.eduTClandfallsarenotasfrequentasinMadagascar (Vitart etal .,2003;ReasonandKeibel,2004;Reason, 2007;KlinmanandReason,2008;Silva etal .,2010). TCimpactsalsoextendovertheopenoceanthrough disruptionofbusyshippinglanesacrosstheentire southernIndianOcean,threateningthelivesofvessel crewsandtheirvaluablecargoesastheysteambetween maritimeeconomichubsinEurope,theAmericas,South Asia,andtheFarEast(Chang-SengandJury,2010b). Largescale,low-frequencymodesofocean-atmosphere variabilityrelatedtoTCgenesisinthesouthernIndian Oceanhavebeenstudiedpreviouslyandarerelativelywellunderstood.TheElNi no-SouthernOscillation (ENSO),theMadden-JulianOscillation,andtheIndian OceanDipole,orZonalMode,areallknowntoinuence thespatialpatternandfrequencyofTCgenesisinthe southernIndianOcean(KuleshovanddeHoedt,2003; BessaandWheeler,2006;Ho etal .,2006;Camargo etal .,2007a;Kuleshov etal .,2008;LeroyandWheeler, 2008;ChanandLiu,2009;Kuleshov etal .,2009;Vitart etal .,2010).However,giventhatTCsforminginthe centralorsoutheasternIndianOceanmaytraversethousandsofkilometerstothewesttoimpacttheaforementionedcountriesonthewesternfringeofthebasin,itisCopyright2010RoyalMeteorologicalSociety

PAGE 2

42K.D.ASHANDC.J.MATYASalsoimportanttoimproveunderstandingofTCtrajectory variabilityinrelationtoocean-atmospherevariabilityat intraseasonaltoseasonaltimescales.Lessresearchhas specicallyfocusedonthetopicofintraseasonalorseasonalsouthwesternIndianOceanTCtrackvariability, savefortheworkofVitart etal .(2003)whichsuggested anENSOlinktoTCtrackvariabilityintheregion.OutsidethebodyofsouthernIndianOceanTCresearch, ENSO,thesubtropicalIndianOceanDipole(SIOD),and tropicaltemperatetroughs(TTTs)haveallbeenshown toplayimportantrolesinair–seainteractionswithinthe southwesternIndianOceanwhichcouldinuenceTC trajectoryvariabilityonintraseasonaltoseasonaltime scales.Inthisstudy,wespecicallyconsidertheinuenceofENSOandSIODonsouthwesternIndianOcean TCtracks,andusethepreviouslyestablishedTTTframeworkforabroaderunderstandingofthemechanismby whichENSOandSIODmayalterTCtrackpatternsin thisregion. Ithasbeendemonstratedineachoftheworld’sprincipalTCregionsthatENSOcanaltertheglobalorregional atmosphericcirculationtoeitherchangethespatialpatternsofTCgenesisoraffecttheirdirectionsofmovement (GrayandSheaffer,1991;Landsea,2000;Chu,2004; Camargo etal .,2010).TheimportanceofENSOinuenceonthecoupledocean-atmospheresysteminthe southernIndianOceanaloneisverywelldocumented (vanLoonandRogers,1981;PanandOort,1983;Meehl, 1987;Karoly,1989;Nicholson,1997;Chambers etal ., 1999;Klein etal .,1999;Reason etal .,2000;Venzke etal .,2000;LarkinandHarrison,2002;HermesandReason,2005;Yoo etal .,2006;HuangandShukla,2007a). JustasinthenorthernandsouthernPacicOceans,there isnotasignicantbasin-widecorrelationinthesouthern IndianOceanbetweenENSOandTCfrequency(Jury, 1993);insteaditisthespatialpatternsofgenesisand directionsofmovementthatexhibitlinkstoENSO(RevellandGoulter,1986;Lander,1994).ENSOhasbeen demonstratedasasignicantpredictorofsouthernIndian TCactivityatweeklytomonthlytimescalesandisnow utilisedinoperationalstatisticalpredictionofTCactivityintheregion(LeroyandWheeler,2008;Vitart etal ., 2010). DuringElNi no,TCgenesisismorefrequentoverthe southwesternIndianOcean(westofabout75E–85E) thanintheeasternocean(EvansandAllan,1992; KuleshovanddeHoedt,2003;Ho etal .,2006;Kuleshov etal .,2008).Favourableconditionsfordeepconvectioninthewesternoceanaregeneratedfollowingthe occurrenceofeasterlylowertroposphericwindanomaliessouthandwestofSumatraduringearlyaustralspring (lateSeptember–November),whichexciteawestwardpropagatingoceanicRossbywavethat,coupledwith increasedpolewardEkmantransportassociatedwiththe easterlyanomaliesinthetropicaleasternpartofthe basin,induceanunusuallydeeppoolofwarmseasurfacetemperatureanomalies(SSTA)centerednear15S, 60E(Chambers etal .,1999;Jury etal .,1999;Reason etal .,2000;Xie etal .,2002;JuryandHuang, 2004).CoolerSSTAinthesoutheasternIndianOcean andanticyclonicvorticityfromthelowertropospheric easterlyanomaliesaccountforareductionofTCgenesis,mostnotablyinlateaustralspringandearlysummer(Ho etal .,2006;Camargo etal .,2007a;Kuleshov etal .,2008;Kuleshov etal .,2009).Conversely,during anENSOcoolevent,thewesternportionofthebasin ischaracterisedbyreducedTCgenesiswhiletheeastern (eastofabout75E–85E)experiencesmore,withpositiveSSTA,increasedmid-troposphericrelativehumidity, andincreasedlowertroposphericcyclonicvorticitycontributingtoanenvironmentconducivetomorefrequent TCgenesis(Wolter,1987;Ho etal .,2006;Camargo etal .,2007a;Kuleshov etal .,2008;ChanandLiu,2009; Kuleshov etal .,2009). AsforthetracksofsouthernIndianTCsinrelation toENSO,Vitart etal .(2003)notethatzonalsteering owaveragedover850–200hPaacrossthetropicaland subtropicalportionsofthebasinismorewesterly(easterly)duringElNi no(LaNi na).Consequently,MozambiqueshouldbeatgreaterriskforlandfallduringLa Ni na,whereaswesterlysteeringowsaresuggestiveof increasedincidencesofre-curvingTCsjusteastofMadagascarduringElNi no(JuryandPathack,1991;Vitart etal .,2003;ReasonandKeibel,2004;Ho etal .,2006; Camargo etal .,2007a;Kuleshov etal .,2009).ChangSengandJury(2010a)agreethatLaNi naisassociatedwithincreasedfrequencyoflonger-livedandmore intenseTCsinthesouthwesternIndianbasin.However, thoughENSOisthemostimportantsourceofclimate variabilityintheglobaltropicsandsubtropics,itdoesnot accountforallvariabilityintheglobaltropicalandsubtropicalocean–atmospheresystem,nordoesitaccountin totalityforthevariabilityoftrajectoriesamongstsouthwesternIndianTCs(Fauchereau etal .,2003;Vitart etal ., 2003;HuangandShukla,2007b;KlinmanandReason, 2008). DuringElNi noinaustralsummer,tropicaltemperate troughs(TTTs)inuenceatmosphericconditionsinthe TCregionsofthesouthwesternIndianOcean(Harangozo andHarrison,1983;Lyons,1991;vandenHeever etal ., 1997;WashingtonandTodd,1999).TheTTTsfrequently shift30–35eastwardandcollocatewiththedeep warmpooloverthetropicalwesternocean(Lindesay etal .,1986;MasonandJury,1997;Cook,2000;Tyson andPreston-Whyte,2000;Nicholson,2003;Fauchereau etal .,2009;Pohl etal .,2009;Manhique etal .,2011). Thiseastwardshiftoftheregionaltropicalconvective maximumandaccompanyingpolewardoutowisconsistentwiththenon-linearElNi noteleconnectionframeworkofHoerling etal .(1997),andplaceswithina broadersynopticcontextthepositivezonalwindanomaliesobservedinpreviousstudiesofthetropicalandsubtropicalsouthwesternIndianbasin(Reason etal .,2000; Yoo etal .,2006).Heretofore,theENSO-TTTframework hasnotbeenappliedtothesouthwesternIndianTCgenesisortrackresearch,asonlyMavume etal .(2009)briey mentionTTTsaspotentiallyimportantinrelationtoTC trackswithintheMozambiqueChannel.Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 3

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES43 TheSIODhasalsobeenidentiedasanimportant sourceofocean–atmospherevariabilityinthesouthern IndianOcean.ItisrepresentedbythesecondempiricalorthogonalfunctionofsouthernIndianOceantropicalandsubtropicalSSTA(Behera etal .,2000;Behera andYamagata,2001;Qian etal .,2002;Suzuki etal ., 2004;HuangandShukla,2007b),andSIODphases havebeensuggestedtoexhibitvariationsindependent ofthesimultaneousENSOphase(BeheraandYamagata,2001;Fauchereau etal .,2003;Washingtonand Preston,2006;HuangandShukla,2007b).Thepatternismostprominentduringtheaustralwarmseason andischaracterisedinthepositive(negative)modeby cool(warm)SSTAinthesoutheasternIndianOcean, whereasthesouthwesternIndianOceanissimultaneously warm(cool).Themechanismsforcoolingintheeastern poleduringapositivemodearestrengthenedsoutheasterlytradewindsandresultantenhancedoceansurface evaporationandmixing,whilethewesternwarmpole developsconcurrentlywithincreasedpolewardEkman transportofwarmSSTsfromtropicallatitudescombinedwithdecreasedequatorwardcoldairadvection andoceansurfaceevaporation(Reason,1999;Venzke etal .,2000;BeheraandYamagata,2001;Qian etal ., 2002;HermesandReason,2005;ChiodiandHarrison, 2007;HuangandShukla,2007b).Innegativemode,the SSTApolesaregenerallyreversedwithcoldairadvectionandequatorwardEkmanpumpingoccurringoverthe southwesternIndianOceaninconjunctionwithmorefrequentcoldfrontalpassages,whilewarmairadvectionis locatedmorefrequentlyoverthesoutheasternpartofthe basin. ThereisevidenceintheliteraturetosuggestthateastwardshiftsoftheAfricanTTTsarenotonlyassociated withENSO,butalsowiththenegativephaseoftheSIOD. Reason(2002)andFauchereau etal .(2009)notedthe strongsimilaritiesbetweenthepositiveSIODSSTand windanomalypatterns,andtheatmosphericanomalies associatedwithTTTsoversouthernAfrica.ThisisimportantbecauseitraisesthepossibilityoflocalSSTAinuenceinthevariabilityofTTTs,whichhaveoftenbeen attributedtotheinuencefromENSO.Duringanegative SIODphase,apersistenteastwardshiftofTTTsoverthe southwesternIndianOceanshouldinuenceTCstofollowmorere-curvingpolewardoreveneastwardtracks (ParkerandJury,1999;Chang-SengandJury,2010b). Therefore,ifnegativeSIODisassociatedwitheastward phaseshiftingofTTTs,andifSIODphasescanexhibit variationsindependentofthesimultaneousENSOphase (Fauchereau etal .,2003;WashingtonandPreston,2006; HuangandShukla,2007b),thenaninteractiveconsiderationofbothENSOandSIODiswarrantedtobetter accountforvariabilityinsouthwesternIndianOceanTC trajectories. InSection2,wedescribetheclusteringprocedure employedtoreduceasampleof191TCsto6types basedontheirgenesislocationswithin3subregions ofthesouthernIndianbasinmaindevelopmentregion (54E–110E)andthenfurthersubdividedbytheir directionsofmovementfromeachgenesissubregion. Wealsooutlinetheanalysisofvariance(ANOVA)tests usedtoinferdifferencesofassociationwithSIODand ENSObetweenthenalsixTCtrajectorygroups.Section 3conveystheclusteringprocedureresults,presentsthe ANOVAtestresults,andcomparesdifferentTCtrajectory groupsthatsharecommongenesisregions.Expanded discussionoftheresultsandpotentialapplicationsare giveninSection4,andSection5summarisesthe importantndingsfromthisstudyandofferspossible directionsforfutureresearch. 2.Dataandmethods 2.1.Tropicalcyclonedata TheTCtrajectoriesinthisstudyarefromtheJoint TyphoonWarningCenter(JTWC)best-trackdatasetfor theSouthernHemisphere(Chu etal .,2002).Becausethis researchfocusesonthesouthwesternIndianOcean,we consideronlythosestormswhichpassedwestof90E longitudeduringtheirrespectivelifecycles.The90th meridianischosenasthewest–eastboundaryinaccordancewiththeofcialforecastingareasofresponsibilityassignedbytheWorldMeteorologicalOrganization (WMO)toRegionalSpecializedMeteorologicalCenters (RSMC)LaR eunion(westof90E)andPerth(eastof 90E)(Caroff,2009).TheJTWCbest-trackdataarepreferredovertheRSMCLaR eunionbest-trackarchive forthisstudybecausetheLaR euniondatasetdoesnot extendasfartotheeastof90EastheJTWCdata duetotheRSMCadministrativeboundary.Byusingthe JTWCarchive,weincludemorecompletetrajectories ofTCsthatformedeastof90E,crossedthatmeridian,andcontinuedwestwardintothesouthwesternIndian basin. Additionally,weconsideredonlyTCsthatreacheda maximumlifetimeintensityofatleast30msŠ 1maximum1-minutesustainedwind.ThisprovisoaddressesTC dataqualityconcernsthroughinclusionofstrongerand betterorganisedTCs,whicharelesspronetopositionxerrorsthanweaker,poorlyorganisedtropicalsystems (Yip etal .,2006).ThesecautiousstepsaredeemednecessaryduetothelackofcompleteandpermanentgeostationarysatellitecoverageoverthesouthernIndianregion priortoMay1998(Kossin etal .,2007;Chang-Sengand Jury,2010a;Kuleshov etal .,2010).Despitethelimited satellitecoverage,KnaffandSampson(2009)notethat JTWCTCdataaresuitableforanalysisfromabout1980. Inkeepingwiththeirsuggestion,weassembledasample of191TCsfromtherecentthirty-yearperiodspanning 1979–2008.AnequallyimportantconsiderationofstudyingTCsfromthisperiodistoresearchprincipallythe ocean–atmosphereinteractionsoperatingaftertheknown Indo-Pacicregionclimateshiftof1976–1977,asspatialpatternsofocean–atmospherevariabilityassociated withENSOandSIODinteractionsareknowntodifferon multi-decadaltimescales(Trenberth,1990;Zinke etal ., 2004;TerrayandDominiak,2005).Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 4

44K.D.ASHANDC.J.MATYAS2.2.Clusteringprocedure TotestforinuencesfromENSOandSIOD,werst breakthe191trajectoriesintogroupsusingcluster analysis.Clusteranalysisiswidelyusedingeophysical researchforclassicationordataexploration,anda well-structuredclusteringsolution,whetherornotit revealsthenaturalmodalityofthedata,canaidinthe discernmentofmechanismsthatshapedifferencesand similaritiesofcomplexeventsorphenomena(Gongand Richman,1995;Wilks,2006).Clusteranalysishasbeen employedrecentlyasatooltoresearchvariabilityin TCtrajectoriesinthenorthwesternPacic(Elsnerand Liu,2003;Camargo etal .,2007b;Choi etal .,2009), thenortheasternPacic(Camargo etal .,2008),andthe NorthAtlantic(Elsner,2003;Kossin etal .,2010).The presentstudyrepresentstherstknownapplicationof clusteranalysistoanalyseTCtrackvariabilityinthe southwesternIndianOcean. Weimplementatwo-stageagglomerativehierarchical clusteringprocedureusingaEuclideandistancemeasure andthegroupaveragelinkagemethod.Agglomerative hierarchicalclusteringbeginswitheachobservationas itsowngroupandtheniterativelyjoinsthetwoclosestgroupsuntilthereisonegroupthatincludesevery observation(Lattin etal .,2003).Thecopheneticcorrelationmaythenbeusedasadiagnosticofthestrength ofaclusteringstructureintandemwithadendrogram (oratreediagram)thatillustratesthestructureoflinkagescarriedoutbytheclusteringalgorithm.Cophenetic correlation,therefore,indicateshowwellthevisualstructureillustratedbythedendrogramreectsrealclusteringstructuresinthedata.Ifthecopheneticcorrelation approachesorexceeds0.8,visualinspectionofthedendrogrammayallowfordiscernmentofanappropriate numberofclustersbycuttingthedendrogramwherethe averagedissimilaritybetweenthegroupsjumpsconsiderablybetweengroupingiterations(Romesburg,1984).At bothstagesoftheclusteringprocedureforthisstudy,the copheneticcorrelationsexceed0.7.Bythismethod,the nalnumberofclustersextractedmaynotalwaysreect thetruemodalityofthedata,butdoesfacilitateanalysis andinterpretationofthedata(Wilks,2006). Thetwo-stageclusteringsolutionemployedinour studyallowsforincorporationofknownphysicalmechanismsrelatedtoENSOthatinuencevariabilityinspatialpatternsofsouthernIndianTCgenesis,whilealso furthersubdividingtheTCsbythedirectionsoftheir movement.Ithasbeenshownpreviouslyinobservational studiesthatElNi no(LaNi na)isassociatedwithmore TCsformingwest(east)ofabout75E–85E(Ho etal ., 2006;Camargo etal .,2007a;Kuleshov etal .,2008). Thisisalsotheoreticallyjustiedinthatafavourable poolofwarmSSTAandanunusuallydeepthermocline arepresentinthesouthwesternIndianOceanduring strongENSOwarmevents(Xie etal .,2002;Juryand Huang,2004;Camargo etal .,2007a;Kuleshov etal ., 2009),whileunfavourablenegativeSSTAs,lowertroposphericeasterlyanomalies,andelevatedmeansea levelpressure(MSLP)arepresentinthesoutheastern Indianbasin(Reason etal .,2000;LarkinandHarrison,2002).ItfollowsthatastraticationofthelongitudesofTCgenesispointsshouldechotheestablished regionalphysicalcharacteristicsofENSO.Thus,therst stageisaunivariateclusteringconsideringonlytheinitiallongitudeofeachTC,fromwhichveclustersare extracted. QualitativecomparisonoftheTCgenesisclusters (Figure1)withpreviousresearchsuggeststhatstage 1oftheanalysisapproximatesthestraticationofTC genesisassociatedwithENSOinuencesasdescribed above(refertoFigures3and4fromKuleshov etal ., 2008,andFigure2(c)fromHo etal .,2006).Thethree largegroupsinthemaindevelopmentregionofthe southernIndianOceancontain89%ofthe191TCsin theinitialclusteranalysis.Thewestern(eastern)group isinthefavourablegenesisareaduringawarm(cool) phaseofENSO.WhereasHo etal .(2006)placedthe boundaryseparatingthewestandeastgenesisregionsat 75E,Kuleshov etal .(2008)suggested85E.Therefore, instage1oftheanalysis,thecentralgrouprepresents aregionoftransitionwherethedifferinginuencesof ENSOphasesonTCgenesisarenotwelldened.Asthe threelargegroupsacrossthemainbodyofthesouthern IndianbasincomprisethemajorityofTCsinthesample, Figure1.MapofveclustersolutionofsouthwesternIndianOceantropicalcyclonesclusteredbyinitiallongitude.Onlythethreemain developmentsubregions(western,central,eastern)areconsideredinsubsequentanalysis.NotethatthisstudyonlyconsidersTCswhichformed orpassedwestof90Eduringtheirlifecycles. Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 5

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES45 Figure2.SixTCtrajectoryclusters,arrangedaccordingtogroupsize,withinthemaindevelopmentregionsforthesouthernIndian Oceanbetween54Eand110E.a)C1,easterngenesis/southwest–southmovement;b)C2,central/west-southwest;c)C3,western/ west-southwest;d)C4,western/south-southeast;e)C5,eastern/west;f)C6,central/south-southeast.onlythesearecarriedforthforthesecondstageof clustering. Thesecondstagesubsequentlyappliesabivariate clusteranalysiswithineachofthethreegenesisclusters inthemainTCdevelopmentregionsboundedby54E and110E.Eachgenesisgroupisfurthersubdivided intoeastwardandwestwardtrajectoriesusingthenal latitudeandlongitudeofeachrespectiveTC.Oneof theprincipalgoalsofthisstudyistoinvestigatethe relationshipofENSOandSIODtothedirectionsofTC movements.Thesecondstageofclusteringthereforeis intendedtofurthersubdivideTCswithineachregionthat movemostlywestwardandthreatenlandfromthosethat movesouthwardoreastwardandremainatsea.Thetwostagesolution,incorporatingboththeinitialandnal geographiclocationsoftheTCs,resultsinsixgroups oftrajectorieswithinthemainbodyofthetropicaland subtropicalsouthernIndianOcean.Thesearepresented ingreaterdetailinSection3.1. 2.3.Analysisofvariance Thesixtrajectorygroupsarecomparedintermsoftheir medianmonthlyvaluesofSSTAindicesrepresenting bothENSOandSIODthroughANOVA.EachTCis assignedthemonthlyvaluesoftheindexcorrespondingtothemonth(s)spanningeachTC’slifecycle.Ifa storm’slifecyclebridgedtwomonths,thetwoindexvaluesareaveraged.ForENSO,ve-monthrunningmeans ofstandardisedNi no-3.4regionanomalies(N3.4,5N, 5S,170W,120W)areused(dataobtainedonline athttp://www.cpc.noaa.gov/data/indices/),andforSIOD theSubtropicalDipoleIndex(SDI)isused(obtainedat http://www.jamstec.go.jp/res/ress/behera/iosdindex.html). SDIiscalculatedbysubtractingtheeastpoleSSTA (18S–28S,90E–100E)fromthewestpoleSSTA (27S–37S,55E–65E)(BeheraandYamagata,2001). Anon-parametricranktest,theKruskal-Wallis(KW) testwithtiesadjustment(KruskalandWallis,1952; Higgins,2004),isusedtocomparethemediansofN3.4Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 6

46K.D.ASHANDC.J.MATYASTableI.SixmaingroupsofTCtrajectories,rankedbynumberofTCsineachgroup.Groupmedianinitialandnallongitudes areindecimaldegreeseast,andmediannallatitudesareindecimaldegreessouth. Cluster designation Number ofTCs Medianinitial longitude(E) Genesis region Mediannal latitude(S) Mediannal longitude(E) Directionof movement C14094.3Eastern23.482.4SW/S C23978.0Central25.256.1W/SW C33964.3Western26.949.8W/SW C42266.1Western26.568.0S/SE C51699.1Eastern18.758.0W C61481.2Central20.583.6S/SE andSDIforthesixprincipaltrajectorygroups.Thenull hypothesesforthetestsarethatthemedianvaluesofN3.4 andSDIarenotsignicantlydifferentacrossallsixTC clusters.Thealternativehypothesesarethatthereexist signicantdifferencesinatleastonepairofTCgroupsin theirmedianvaluesofN3.4orSDI.Toensuretheresults oftheKWtestsareappropriatelyinterpreted,ModiedLeveneEqual-VarianceTestsarealsoapplied(Brownand Forsythe,1974).Dunn’sranksumsprocedureisused toteststatisticalsignicanceinmultiplecomparisons (Dunn,1964). TocomplementtheANOVAtests,SSTAcomposites arealsoconstructedforeachclusterusingtheNational OceanicandAtmosphericAdministration(NOAA)OptimumInterpolation(OI)SSTVersion2(Reynolds etal ., 2002).AnSSTaverageisextractedbasedontheinitialandnaldatesforeachrespectiveTC’slifecycle. MonthlySSTaveragesarealsocreatedbasedonthe period1981–2008.ToobtainSSTanomalycomposites foreachTC,the26-yearmonthlycompositeissubtracted fromeachrespectiveTCeventcomposite.Forexample,TCJaya’slifecyclespannedMarch–Aprilof2007. Therefore,theSSTAcompositemapforJayaistheSST compositeforthatTC’slifecycleminusthecombined March–April26-yearSSTcomposite. Finally,totesttheinteractionofENSOandSIOD phaseswithrespecttotheTCtrajectorytypes,three contingencytablesareconstructedstratifyingTCsthat formedinthesameTCgenesissubregionsbysixcategoriesofENSOandSIODinteractions:ENSOwarm eventwithpositiveSIOD(E + S + ),ENSOwarmevent withnegativeSIOD(E + S Š ),ENSOneutralwithpositiveSIOD(E S + ),ENSOneutralwithnegativeSIOD (E S Š ),ENSOcooleventwithpositiveSIOD(E Š S + ), andENSOcooleventwithnegativeSIOD(E Š S Š ). MonthsinwhicheachTCoccurredareassignedasENSO warm,neutral,orcoolphasesbasedonacriteriaofat leastsixconsecutivemonthswith + / Š 0.4Canomalies forve-monthrunningmeansofthestandardisedN3.4 index.ThisisaverysimilarmethodologyfordeterminingENSOphaseasoutlinedbyTrenberth(1997)and implementedrecentlyinLau etal .(2008)andAntico (2009).Theprincipaldifferenceisthatourbaseperiodfor standardisationoftheN3.4indexislimitedtothestudy periodof1979–2008.Fisher’sExactTestprovidesexact p -valuesforcontingencytabletests,thusitisapplied suchthatrobustresultsmaybeobtaineddespitethesmall samplesizes(Higgins,2004). 3.Results 3.1.Clusteranalysis Thetwo-stageclusteranalysisassignsthe170main developmentregionTCsinto6groupsoftrajectories, accordingtotheirinitiallongitudesandnallatitudes andlongitudes.TheyarerankedbythenumberofTCs withineachgroup,designatedsimplyasC1throughC6 (TableI).Almost70%ofthe170TCsarecontained withinthe3largestgroups(C1-C3),andthesedisplaya netsouthwesterlycomponentofmovement.Thisistobe expectedgiventhattypicalbackgroundconditionswould allowforsuchanaveragemotionaroundthenorthwest fringesofthesouthernIndiansubtropicalanticyclone. ThelargestgroupisC1with40TCs.Theseformed between87Eand110Eandfollowedsouthwestward andsouthwardtrajectories(Figure2(a)).C2iscomprised of39TCswhichformedbetween73Eand87Eand movedwestwardorsouthwestward(Figure2(b)).Group C2threatenedlandwithrelativefrequency(16of39 within200kmofMadagascar,Mauritius,LaR eunion,or Mozambique)whileC1stormsremainedovertheopen ocean.Thethirdlargestgroup,C3(39TCs),originated between54Eand73Ewithalargeproportion(33 of39)thatmovedwestwardorsouthwestwardwithin 200kmofinhabitedareas(Figure2(c)). Theremaining30%ofTCscomprisethe3smaller groups.Theserepresentmarkeddeparturesfromthetypicalnetsouthwestwardmovement,andarethereforelikely candidatesforassociationwithnotabledeparturesfrom themoretypicalsouthwestwardsteeringowregime, suchasmightoccurwithstrongphasesofENSOor SIOD.C4(Figure2(d)),incontrasttoC3,developed between54Eand73Ewithadominantdisplacement southwardandsoutheastward.Thoughthe22TCsin groupC4formedincloseproximitytotheinhabited islandsinthesouthwesternIndianOcean,theyseldom passedwithin200kmofMadagascarortheMascarene islands.Havingformedbetween87Eand110E,group C5(Figure2(e))isdistinguishedfromC1bylongerdistancetrackswithmorewestwardtrajectories,10ofwhich passedwestof60Eandthreatenedpopulatedregions.Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 7

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES47TableII.MedianstandardisedanomaliesforNi no-3.4(N3.4) seasurfacetemperatureandSubtropicalDipoleIndex(SDI). ClusterIDNi no-3.4SDI C1 Š 0.05 Š 0.19 C2 Š 0.04 Š 0.06 C3 Š 0.220.34 C40.73 Š 0.70 C5 Š 0.370.77 C60.160.14 Thesmallestofthe6trajectoryclusters,C6,formed between73Eand87E(Figure2(f)).Directionofmovementissouthandsoutheastward,keepingthesesystems wellouttosea. 3.2.KWANOVA 3.2.1.ANOVAresults Theclusteringprocedureusinginitiallongitudeandnal latitudeandlongitudeallowsTCtrajectoryclusterswith commongeographicoriginsanddivergingpathstobe compared.ThisisaccomplishedthroughANOVA,where wecomparethemedianvaluesoftheN3.4indexandthe SDI(TableII)rstacrossallgroupsandtheninmultiple pair-wisecomparisons.TheresultsoftheKWtestsare torejectH0ofnodifferencebetweenthemedianvalues ofbothN3.4andSDI( = 0 001, p -values0.00019and 0.00012,respectively).ModiedLeveneEqual-Variance testsarealsoemployed,andH0ofhomoscedasticityis notrejectedforeitherN3.4orSDI( = 0 05, p -values 0.88and0.22,respectively).Thus,thereisstatistical evidencetosupporttheassertionthatSSTsintheN3.4 andSDIregionsaresignicantlyandcontemporaneously associatedwithdifferenttypesofsouthwesternIndianTC trajectories. Severalothervariablesweretestedusingthesame KWANOVAmethod.PreviousstudiessuggestedalinkagebetweenthetropicalIndianOceanDipoleorZonal Mode(Saji etal .,1999;Webster etal .,1999)andsouthernIndianTCactivity(LeroyandWheeler,2008;Chan andLiu,2009;Vitart etal .,2010).KlinmanandReason(2008)andChang-SengandJury(2010a)notedthat variabilityoftheSouthernHemisphereplanetarywaves couldalsoaffectsouthwesternIndianTCtrajectories. Vitart etal .(2010)includetheTrans-Ni noIndex(TrenberthandStepaniak,2001)asasignicantpredictorof southernIndianTCactivityinassociationwithENSO variability,consistentwithrecentndingsofassociation betweenENSOvariabilityandTCactivityintheNorth AtlanticandNorthPacicbasins(Kim etal .,2009;Chen andTam,2010;Kim etal .,2010).Accordingly,wecarriedoutthesameANOVAprocedureusingTNI,Dipole ModeIndex(DMI)(Saji etal .,1999),andanAntarcticOscillationIndex(AAO)(obtainedfromNOAAat http://www.cpc.noaa.gov/products/precip/CWlink/daily ao index/aao/aao index.html).WedidnotndsignicantTableIII.Multiplecomparison Z -valuetestsforNi no-3.4 regionstratiedbythesixTCtrajectoryclusters. Ni no-3.4C1C2C3C4C5C6 C10.79291.02093.2288b1.80090.4907 C20.22663.8834b1.19331.0618 C34.0759b1.02051.2265 C44.2297b2.0611aC51.8720 C6 aIndicatessignicantdifferencesofmedianNi no-3.4valuesat = 0 05.bIndicatessignicanceat = 0 01.TableIV.Multiplecomparison Z -valuetestsforSDIregion stratiedbythesixTCtrajectoryclusters. SDIC1C2C3C4C5C6 C10.60041.90282.1531a3.1125b1.0592 C21.30242.6538b2.6414b0.6178 C33.7465b1.61180.3485 C44.5597b2.6445bC51.6232 C6 aIndicatessignicantdifferencesofmedianSDIvaluesat = 0 05.bIndicatessignicanceat = 0 01.differencesbetweenthesixTCtrajectorytypesbased uponthesethreeadditionalindices(resultsnotshown). 3.2.2.WesternregionTCtypes:C3versusC4 ToascertainexactlywhichTCtypesaremoresignificantlyassociatedwithphasesofENSOand/orSIOD, Dunn’srank-sumprocedureisemployedformultiple comparisonsofthemedianvaluesofN3.4andSDIacross the6trajectoryclusters.ResultsofthemultiplecomparisonsforN3.4(TableIII)andSDI(TableIV)indicate thatTCtypeC4issignicantlymoreassociatedwithEl Ni noandthenegativeSIODmodethantheother5trajectorytypes.TheSSTAcompositesforC4corroborate theseresults.ThecompositeSSTApatternintheIndian Ocean(Figure3(d))issuggestiveofthenegativeSIOD, whichischaracterisedbywarmanomaliesfromMadagascararcingeastwardandsoutheastwardtowardsouthwesternAustralia(Reason,1999;Behera etal .,2000; BeheraandYamagata,2001;Qian etal .,2002;Suzuki etal .,2004).PositiveSSTAexceeding1Carealsoevidentbetween120Wand170WintheN3.4region (Figure4(d)). ElNi noiswellknowntoassociatewithwesterlywind anomaliesoverthesouthwesternIndianbasinduringan australwarmseason(Vitart etal .,2003;Todd etal ., 2004;Yoo etal .,2006).TTTsare,likewise,associated withwesterlyanomaliesoverthesamegeographicregion, withverticalextentfromtheboundarylayerupto 500hPaandawest–eastsignaturethatcanextendacrossCopyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 8

48K.D.ASHANDC.J.MATYAS Figure3.SeasurfacetemperatureanomalycompositesforthesouthernIndianOceancorrespondingtothesixTCtrajectoryclusterspresented inFigure2.SDIWest(27S–37S,55E–65E)andSDIEast(18S–28S,90E–100E)regionsareshownasboxesineachimage.Positive (negative)valuesaresymbolisedbydarksolid(lightbroken)lineswithunitsinC.theentiresouthernIndianOcean(ToddandWashington, 1999).Reason(2002)foundtheTTTsalsotovaryin associationwithSSTAdipolessimilarinpatterntothe SIODofBeheraandYamagata(2001),andthedeepest convectioncoincidentwithTTTshaslikewisebeen foundtoshiftfromsouthernAfricatothesouthwestern IndianOceaninassociationwithwarmENSOevents (Fauchereau etal .,2009;Pohl etal .,2009;Manhique etal .,2011).GiventhehighsignicanceofElNi no andnegativeSIODforC4relativetoallotherTC trajectorytypes,coupledwiththepropensityofTTTs toextendeastwardoverMadagascarandthewestern oceanduringbothofthesephases,thissuggeststhat C4typeTCsarehighlylikelytooccurwhenENSOis inwarmphaseandSIODissimultaneouslyinnegative mode.This30–35eastwardshiftintheregionaltropical convectivemaximumandtherelatedeastwardshiftof polewardconvectiveoutow,andthesubtropicaljet stream,providesalogicalexplanationbywhichC4typeTCsareoftensweptsouthandsoutheastwardaway fromMadagascar.ThisisconsistentwiththesynopticscaleteleconnectionsofHoerling etal .(1997)andthe S-movingTCkinematicandthermodynamicproleof Chang-SengandJury(2010b). Havingestablishedtheocean–atmosphericconnectionsforC4,itisnowappropriatetocomparewithC3, whichformsinthesamewesternregion,yetapproaches inhabitedregionswithgreaterfrequency.Thisgroupis signicantlydifferentfromC4inmedianvaluesofN3.4 andSDI(TablesIIIandIV),suggestingthatC3occurs morefrequentlyduringLaNi naorapositiveSIODphase. TheSSTAcompositemapsforC3(Figures3(c)and4(c))Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 9

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES49 Figure4.SeasurfacetemperatureanomalycompositesfortheequatorialPacicOceancorrespondingtothesixTCtrajectoryclusterspresented inFigure2.Ni no-3.4region(5N–5S,120W–170W)isshownwithaboxineachimage.Positive(negative)valuesaresymbolisedbydark solid(lightbroken)lineswithunitsinC.depictwarmanomaliessouthandsoutheastofMadagascar,andcoolanomaliesintheequatorialPacic.SSTA compositedifferencemapsforC4andC3(Figures5(a) and6(a))furthersupporttheimportanceoftheSDIand N3.4regionsindistinguishingbetweenthesetrajectory types.Enhancedeasterlyandsoutheasterlytradewinds acrossthesubtropicalandtropicalsouthernIndianbasin duringLaNi natendtocoincidewithmorewestward movingTCsthatwouldthreateninhabitedlandmasses withhigherfrequencysuchasthoseingroupC3(Reason etal .,2000;Vitart etal .,2003;Chang-SengandJury, 2010aand2010b).Furthermore,inaccordancewitha positiveSIODphase,warmSSTAinthesouthwestern IndianOceanareassociatedwithincreasedconvective precipitationoversouthernAfricaastheTTTsaremore frequentlyanchoredtotheAngolathermallow(Toddand Washington,1999;Reason,2001;Reason,2002).ForC3typeTCsthatturnpolewardof25Slateintheirlife cycle,interactionwithTTTsovertheAfricansubcontinentwouldhelpexplainlatere-curvature. FurthersupportingthedifferencebetweenTCtracks belongingtogroupsC3andC4,theresultoftheFisher’s ExactTestindicatesahighlysignicantassociation betweentheTCtypes(C3andC4)andthetypeof interactionbetweenENSOandSIOD( p -value = 0.0007, TableV).WhenENSOwasinwarmphaseandSIOD wassimultaneouslynegative,typeC4TCsoccurred frequently,andtypeC3TCsseldomoccurred.These ndingssupporttheassertionthattypeC4stormsare highlyassociatedwithantiphasingofENSO(warm)Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 10

50K.D.ASHANDC.J.MATYAS Figure5.SeasurfacetemperatureanomalycompositedifferencesinthesouthernIndianOceanfor(a)C4minusC3;(b)C1minusC5;and (c)C6minusC2.SDIWest(27S–37S,55E–65E)andSDIEast(18S–28S,90E–100E)regionsareshownasboxesineachimage. Positive(negative)valuesaresymbolisedbydarksolid(lightbroken)lineswithunitsinC.TableV.SouthernIndianOceanwesternsubregiontropical cyclonesbytypeandENSOandSIODinteractions.E + representsawarmENSOevent,E isaneutralevent,and E Š acoolevent.S + signiesapositiveSIODeventandS Š anegativeSIODevent. C3C4Total E + S + 527 E + S Š 21214 E S + 13215 E S Š 549 E Š S + 303 E Š S Š 628 Total342256 andSIOD(negative)whereinconditionsarefavourable forTC–troughinteractionoverthesouthwesternIndian Ocean.Incontrast,typeC3stormsoccursignicantly moreoftenthanC4stormswhenENSOisneutral andSIODispositive.Inthissituation,theMascarene anticycloneandassociatedtradewindsarepresentover thecentralandsouthwesternIndianbasin,andTTTs typicallyremainanchoredovertheAfricansubcontinent. Theseresultsunderscoretheinsufciencyofanopposing symmetricalElNi no–LaNi naparadigmindiscerning inuencesonsouthwesternIndianTCtrajectorieswithout considerationofENSO-independentvariability(Vitart etal .,2003;KlinmanandReason,2008). 3.2.3.EasternregionTCtypes:C1versusC5 TheTCtrajectorytypethatdevelopsintheeasternregion withatypicallywestwardmotion,C5,ismoststrongly associatedwiththepositiveSIODmoderelativetothe otherclusters.ReadilyidentiablepositiveSIODsignals areapparentbothinthehighsignicanceofC5inthe multiplecomparisons(TableIV),andintheSSTAcompositemaps(Figure3(e))whichshowspatiallycoherent warmanomaliesoffthesoutheastcoastofAfricainto thesouthwesternIndianOcean,aswellascoolanomaliesextendingeastwardfromMadagascaralong15S, 25S.ALaNi naSSTApatternisalsovisibleacrossthe equatorialcentralPacic(Figure4(e)).ThecoolSSTAs inthesouthernIndianregionarelargelyinducedby decreasedair-to-sealatentheatuxandEkmantransportassociatedwithstrongtradewindsandconcomitant equatorwardadvectionofrelativelydrymid-latitudeair massesacrossatightnorth–southSSTgradient(Behera andYamagata,2001;HermesandReason,2005;Chiodi andHarrison,2007;HuangandShukla,2007b).The presenceoftheseSSTAinfersanomalouslystrongtrade windsacrossthesouthernIndianOceannorthof25SCopyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 11

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES51 Figure6.SeasurfacetemperatureanomalycompositedifferencesintheequatorialPacicOceanfor(a)C4minusC3;(b)C1minusC5;and (c)C6minusC2.Ni no-3.4region(5N–5S,120W–170W)isshownwithaboxineachimage.Positive(negative)valuesaresymbolised bydarksolid(lightbroken)lineswithunitsinC.whichcouldaidinsteeringTCsonlongerdurationand lowerlatitudewestwardtracks,asinVitart etal .(2003). Therefore,whenapositiveSIODphaseoccursintandemwithanENSOneutralorcoolphase,theTCs oftypeC5tendtoremainatlowerlatitudesbecause theseasonallystrongsubtropicalsouthernIndiananticycloneprecludesrepeatednorthwardintrusionsbyplanetarywaves(L’HeureuxandThompson,2006)which thenfailtofostereastwardextensionofconvectiveclustersandtheirpolewardoutowsassociatedwiththe AngolathermaltroughandSouthIndianConvergence Zone(SICZ). WhileTCsingroupC5developinthesoutheastern Indianbasin,theyoccasionallythreatenlandasthey progressfarwestward.TCsinthecounterpartgroupC1 alsodevelopintheeastbutre-curveintothemiddle latitudesinthecentraloreasternocean.TheSSTAcompositemapsforC1(Figures3(a)and4(a))donotexhibit astrongENSOwarmorcoolpatterninthePacic.However,thedatadepictanorthwesttosoutheastregion ofwarmSSTAinthecentralandsoutheastsubtropical southernIndianOceanwhichresemblesthespatialSSTA patternoftheSIODnegativemode.MapsofthedifferencesbetweentheC1andC5composites(Figures5(b) and6(b))showmarkeddifferencesinSSTApatterns inboththeSIODandENSOregions.Specically,the presenceofwarmSSTAinthecentralandeasterntropicalandsubtropicalsouthernIndianbasinandcoolSSTA inthesouthwestsubtropicalIndianOceanindicatesweakenedtradewindsandincreasedfrequencyoffrontalintrusions.Thiswouldresultinstrongercoldairadvection promotingdecreasedair-to-sealatentheatuxinthe subtropicalsouthernocean(BeheraandYamagata,2001; ChiodiandHarrison,2007).Also,reducedEkmantransportfromtherelativelyweaktradewindstothenorth, inconjunctionwithincreasedair-to-seaheatuxfrom warmadvectionaheadofcoldfronts,wouldallowfor warmSSTApatternsassociatedwiththenegativeSIOD phaseorElNi no.SimilartoTCtypeC4,troughspenetratingfarthernorthintothecentralandsoutheastern IndianbasinwouldinuenceC1TCstore-curvemore abruptlyintothemid-latitudesthanC5TCs. FollowingthesamemethodologyasfortypesC3and C4above,acontingencytableofENSOandSIOD interactionsisconstructedfortypesC1andC5and Fisher’sExactTestisagainapplied(TableVI).Results indicateasignicantassociationbetweenthetwoeastern regiontrajectorytypesandENSOandSIODphase interactions,albeitwithamoreliberalalphalevel( = 0 1, p -value = 0.097).TypeC5TCtrajectories,the westwardmovinggroup,didnotoccuratallcoincident withanyENSOwarmphaseduringourstudyperiod,Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 12

52K.D.ASHANDC.J.MATYASTableVI.AsinTableV,butforeasternsubregiontropical cyclones. C1C5Total E + S + 303 E + S Š 606 E S + 8816 E S Š 12214 E Š S + 549 E Š S Š 527 Total391655 andonlytwicewhenENSOwasneutralbutSIODwas innegativemode.TypeC5stormsoccurredmostoften duringpositiveSIODphaseswithonlytwoduringLa Ni nawhenSIODwasnegative.Thedatasuggestforthe easternregionthatitisnotsufcienttoascribeTCsof longdurationthatmoveconsistentlywestwardonlyto LaNi na.ContemporaneousconsiderationofSIODphase isespeciallyrelevantwhenENSOisneutral,anditis physicallyconsistentthatapositiveSIODphasewould inuencemorewestwardmovingTCs,asseeninboth trajectorytypesC3andC5.Finally,itisalsoevidentin analysisofgroupC1thatawarmENSOeventcombined withanegativeSIODphaseresultsinahigherfrequency ofre-curvingTCtrajectoriesthanduringanENSOwarm eventcombinedwithapositiveSIODmode. 3.2.4.CentralregionTCtypes:C2versusC6 TheTCgroupwithgenesisinthecentralregionand westwardandsouthwestwardmovementisC2.C2isonly signicantlydifferentfromC4intheN3.4comparisons (TableIII),andisonlysignicantlydifferentthanC5 intheSDIcomparisons(TableIV).However,thelatter resultshouldnotbeinterpretedtomeanthatC2exhibits astrongassociationwithSIODthatisoppositeinsignto thestrongassociationbetweenC5andthepositiveSIOD phase.Thestatisticalsignicanceisinastrongpositive association(C5)comparedtoaveryweakassociation (C2),notastrongpositiveassociationcomparedtoa strongnegativeassociation.Itisclearfromboththe SDIvaluesandtheSSTAcompositemap(Figure3(b)) thatinthisstudyC2TCsarenotassociatedobviously witheitherphaseoftheSIOD.TheequatorialPacic SSTAcompositemapforC2(Figure4(b))displaysa recognisableLaNi napatterneastof180,thoughthe N3.4valuesarenotstatisticallysignicantinmultiple comparisons. Asintheprevioussections,itisusefultocompare C2tothecounterpartcentralregionTCtypeC6,which ischaracterisedbymoreeastwardtrajectoriesthanC2. ThesetwogroupsarenotstatisticallydifferentincomparisonsoftheirmedianvaluesofN3.4andSDI.Nor aretherecontiguouswarmSSTAintheequatorialPacic (Figure4(f))tosuggestanElNi noassociation.ThecompositedifferencemapsofC6minusC2(Figures5(c) and6(c))identifytheequatorialPacicasaregionofTableVII.AsinTableV,butforcentralsubregiontropical cyclones. C2C6Total E + S + 426 E + S Š 415 E S + 9413 E S Š 549 E Š S + 606 E Š S Š 10313 Total381452 noteindistinguishingthetwoTCtypes,butthereisnot asignicantstatisticalnoranobviousgeospatialdifferencethatwouldallowformoreconciseinterpretation. Finally,acontingencytableisconstructedstratifyingthe TCsbytypeandbycongurationsofENSOandSIOD (TableVII).TheresultsofFisher’stestalsosuggest nosignicantassociationtodistinguishTCtrajectories intypesC2andC6bytheirfrequenciesduringdifferentinteractionsofENSOandSIODphases( p -value = 0.559).Whenseparatedfromthewesternandeastern regions,thecentralregionTCtrajectoriesdonotdisplay obviousmodesrelatingre-curvingandeastwardtracksto ElNi noandnegativeSIOD,norwestwardtrackstoneutralENSOandpositiveSIODorLaNi na.Giventhesmall sizeofgroupC6andthathalfoftheTCsinthegroup wereearlyorlateseasonstorms(September–October, orApril–May),perhapsitisnotsurprisingthatnoclear interactivesignalbetweenENSOandSIODisfoundin thecentralregion.BothENSOandSIODareknownto exhibitstrongersignalsattheheightofaustralsummer astheyarephaselockedtotheseasonalcycle. 4.Discussion Inthisstudy,webuildontheENSO-TCtrackframework ofVitart etal .(2003)andKlinmanandReason(2008)by consideringinteractionsoftheENSOandSIODphases. WendthatsouthwardandsoutheastwardmovingTCs (typesC1andC4)aremorelikelytooccur,particularlyin thesouthwesternIndianOcean,whenENSOisinwarm phaseandSIODissimultaneouslyinnegativemode. Furthermore,westwardandsouthwestwardmovingTCs (C3andC5)arelikelyduringacoolENSOevent(relative towarmENSO),butaremostlikelytooccurwhen ENSOisneutralandSIODisinpositivemode.Akey implicationoftheseresultsisthatperiodsofeitherstrong westerlyoreasterlysteeringowinthesouthwestern IndianOceanshouldnotbeattributedsolelytoany particularENSOphase,butthesimultaneousSIODphase shouldbeconsideredaswell. TheSIOD-TCtrajectoryrelationshipproposedinthis papermaybeappliedtoTCFavioof2007toexplain itsunusualdirectionofmotionsouthofMadagascar andbacktothenorthwestoverMozambique.Klinman andReason(2008)notedthatsuchawestwardtrackCopyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 13

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES53 duringaweakElNi noyeardidnotfollowtheENSO-TC steeringmodelofVitart etal .(2003),andsuggestedthat monthlydatamightbemoreappropriatethanseasonal dataindiscerningsouthernIndianTCtrackvariability. Usingmonthlydata,theSDIwasstronglypositive(1.35) inFebruary2007,whichisconsistentwithboththe warmSSTAsouthofMadagascarandpersistenttrade windswhichallowedFaviotoremainatalowerlatitude (ratherthanre-curvingintohigherlatitudes)andprovided afavourablethermodynamicenvironmentforittoreintensifysouthwestofMadagascar.Whiletheexampleof FavioiswellsuitedtodemonstratehowSIODmodecan assistinunderstandingTCtrackbehaviourwhenENSO isneutralorweaklyinwarmorcoolphase,itmustalsobe acknowledgedthatSIODmode,likeENSOphase,isnot aperfectpredictorofsouthernIndianTCtrackdirection. Thisstudysuggeststhatinclusionofanindexfor SIODmodecouldaddpredictivepowerinstatistical modellingofsouthwesternIndianTCoccurrenceat intraseasonalorevenseasonaltemporalscales.Indirect evidenceinsupportofthisassertionexistsinthepeerreviewedliterature.LeroyandWheeler(2008)include thesecondrotatedprincipalcomponentofIndo-Pacic SSTAasasignicantpredictor(variableSST2)intheir logisticregressionmodelofSouthernHemisphereweekly TCgenesisandactivity.TheirSST2variableisvery similarinconstructionandspatialpattern(shownin theirFigure5(b),whichwasadaptedfromDrosdowsky andChambers,2001)tothesubtropicalSSTAdipole patternsidentiedinnumerouspreviousstudies(Reason, 1999;BeheraandYamagata,2001;Qian etal .,2002; Suzuki etal .,2004;HuangandShukla,2007b).Inthe updatedoperationalstatisticalmodel(Vitart etal .,2010), theSST2variableisreplacedbythetropicalDipoleMode IndexofSaji etal .(1999).Ourresultssuggestthatan indexcapturingthesubtropicaldipolephenomenon(such asSDI)couldbeatleastasimportantasthetropicalmode inmodellingofsouthwesternIndianTCoccurrenceor trackdirection.Asaproxymeasureofthevariabilityin strengthandpositionoftheTTTsandsouthernIndian subtropicalanticycloneatsubseasonalintervals,itis physicallyconsistentthatSDIshouldbesignicantly associatedwithvariabilityofsouthwesternIndianTC trajectoriesand,therefore,occurrenceatsubseasonal intervals.AsHermesandReason(2005)observe,the SIODanomaliesmaynotalwaysbeadequatelycaptured bytheSDIasconstructedbyBeheraandYamagata (2001).MoreresearchisneededtoindexSIOD-like patternswithvaryingspatialsignaturesbeforethesecan mosteffectivelybeappliedinpredictionofTCactivity ortrackdirection. ThestrongtendenciesforcertainTCtrajectoriesto occurwithcertaincongurationsofENSOandSIOD (TablesVandVI)alsoraisethepossibilitythatthese congurationsarenotonlyspecictotypesofTCtrajectories,butactuallyreectpreferredocean–atmosphere patternsduringsouthernIndianTCseason.Usingthe samemethodologyasdescribedinSection2.1forassignmentofElNi no,neutral,andLaNi namonths,SDI Figure7.PercentageofmonthsduringNovember–Marchoverthe period1979–2007whenENSOwasincoolphase(ENSO Š ),neutral phase(ENSO ),orwarmphase(ENSO + )andSDIwassimultaneously innegativeorpositivemode.wasnegative64%ofthetimecoincidentwithElNi no (Figure7)duringNovember–Marchovertheperiod 1979–2007,whereasSDIwaspositive61%ofthetime coincidentwithneutralENSOconditions.LaNi naand bothpositiveandnegativeSIODmodesoccurredwith equalfrequencyduringthestudyperiod.Thesemonthly percentagesaresimilartodailypercentagesforcoincidencesofAntarcticOscillation(AAO)andENSO phasespresentedinCarvalho etal .(2005)(seenintheir Figure6).ThereisastrongpossibilitythattheAAOor SouthernAnnularMode(SAM)isassociatedwithboth ENSOandSIODandsubsequentlywithTTTsandTC trajectoryvariabilitythroughthemodicationofplanetarywaves(HermesandReason,2005;L’Heureuxand Thompson,2006;KlinmanandReason,2008;Manhique etal .,2011;Chang-SengandJury,2010a). 5.Conclusions Thegoalofthisworkwastoinvestigatetheinuences ofboththeSIODandENSOonsouthwesternIndian OceanTCtrajectories.WegroupedTCsbytheirgenesisregionsandsubsequenttrajectoriesandthenemployed non-parametricANOVAtotestfordifferencesofmonthly valuesofN3.4andSDIcorrespondingtoeachTC’slife cycle.TheresultsindicatethatbothN3.4andSDIare associatedwithsignicantdifferencesinthedirectionsof TCtrajectoriesinthestudyregion.Specically,ourwork suggeststhatsouthernIndianbasinTCsoriginatingin theregions54E–73Eor87E–110E,andfollowing westwardandsouthwestwardtrajectories(southwardand southeastwardtrajectories)aresignicantlymoreassociatedwithacoolorneutralENSOphaseandapositive SIODmode(warmENSOandnegativeSIODmode). Wedonotndthesamestatisticallysignicantrelationshipsinthecentraloceanapproximatelyboundedby 73E–87E. Theresultspresentedinthisstudygobeyondan ENSO-onlymodelofsouthwesternIndianTCtrajectory variability.WhenconsideredtogetherwithENSO,theCopyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 14

54K.D.ASHANDC.J.MATYASSIOD(usingtheSDI)allowsaclearerunderstanding thatwhenENSOisinwarmphaseandSIODisnegativetheregionalocean–atmospherepatternsstrongly favourashiftoftropicaltemperatetroughs(TTTs)over thesouthwesternIndianOceancoincidentwiththewarm SSTApool.Inthissituation,theTTTsarepositioned tofrequentlysteerTCsawayfrominhabitedareason thewesternrimofthebasin.WhenENSOisneutral orincoolphaseandSIODispositive,TTTsdonot shiftpersistentlyeastwardoverthesouthwesternIndian Oceanandwelldevelopedtropicalsystemsarefrequently steeredalongmorewestwardandsouthwestwardtracks. Itisimportanttodiscernthedifferencesinthesecongurationsandcouldbeofgreatuseinthefutureto provideforewarningperhapsamonthormoreinadvance ofenhancedpotentialforTCstrikestotheheavilypopulatedandsociallyvulnerablenationsofMadagascarand Mozambique. Inbroadterms,muchadditionalworkisneededto understandtheasymmetricalandnon-lineartropicaland extratropicalinteractionsthatproducetheSIODSSTA patterns,andhowthesebearinuenceonsouthwestern IndianTCtrajectoriesatintraseasonalandseasonaltime scales.Inparticular,understandingoftherelationship betweenTTTsandTCactivitycouldbenetfromdeeper inquiryasthistopichasnotbeencomprehensivelyexaminedintheliterature.Futureresearchshouldalsoexplore alternativeindexingprocedurestomosteffectivelycapturethegeographiclocationandextentaswellasmagnitudeofSIODevents,andtesttheapplicationofthese alternativeSIODindexesforpotentialimprovementof operationalintraseasonaltoseasonalsouthwesternIndian basinTCpredictionschemes. Acknowledgments Theauthorsextendthankstothetwoanonymousreviewersfortheirconstructivecomments.WealsothankPeter WaylenandTimFikfortheirvaluablesuggestionsduringtheearlierstagesofthiswork.Thisresearchwas undertakenattheUniversityofFloridaaspartofthe leadauthor’sM.S.thesis.HewishestothanktheUF GeographyDepartmentforfundingandsupport. ReferencesAnticoPL.2009.Relationshipsbetweenautumnprecipitation anomaliesinsoutheasternSouthAmericaandElNi noevent classication. InternationalJournalofClimatology 29 :719–727, DOI:10.1002/joc.1734. BeheraSK,SalvekarPS,YamagataT.2000.Simulationofinterannual SSTvariabilityinthetropicalIndianOcean. JournalofClimate 13 :3487–3499,DOI:10.1175/1520-0442(2000)013 < 3487:SOISVI > 2.0.CO;2. BeheraSK,YamagataT.2001.SubtropicalSSTdipoleeventsinthe southernIndianOcean. GeophysicalResearchLetters 28 :327–330, DOI:10.1029/2000GL011451. BessaM,WheelerMC.2006.ModulationofSouthIndianOcean tropicalcyclonesbytheMaddenJulianOscillationandconvectively coupledequatorialwaves. MonthlyWeatherReview 134 :638–656, DOI:10.1175/MWR3087.1. BrownMB,ForsytheAB.1974.Robusttestsfortheequalityof variances. JournaloftheAmericanStatisticalAssociation 69 : 364–367. BrownML.2009.Madagascar’scyclonevulnerabilityandtheglobal vanillaeconomy. ThePoliticalEconomyofHazardsandDisasters JonesEC,MurphyAD.(eds).AltamiraPress:NewYork. CamargoSJ,EmanuelKA,SobelAH.2007a.Useofagenesis potentialindextodiagnoseENSOeffectsontropicalcyclonegenesis. JournalofClimate 20 :4819–4834,DOI:10.1175/JCLI4282.1. CamargoSJ,RobertsonAW,BarnstonAG,GhilM.2008.Clustering ofeasternNorthPacictropicalcyclonetracks:ENSOand MJOeffects. GeochemistryGeophysicsGeosystems 9 :Q06V05, DOI:10.1029/2007GC001861. CamargoSJ,RobertsonAW,GaffneySJ,SmythP,GhilM.2007b. Clusteranalysisoftyphoontracks.PartI:Generalproperties. Journal ofClimate 20 :3635–3653,DOI:10.1175/JCLI4188.1. CamargoSJ,SobelAH,BarnstonAG,KlotzbachPJ.2010.The inuenceofnaturalclimatevariabilityontropicalcyclones,and seasonalforecastsoftropicalcycloneactivity. Globalperspectives ontropicalcyclones:Fromsciencetomitigation ,2ndedn.ChanJCL, KepertJD(eds).WorldScientic:NewYork. CaroffP.2009.Operationalpracticesforbest-trackelaborationat RSMCLaReunion.Presented5May2009attheInternational Best-TrackArchiveforClimateStewardship(IBTrACS)Workshop, NOAANationalClimaticDataCenter:Asheville,NC. CarvalhoLM,JonesC,AmbrizziT.2005.Oppositephasesof theAntarcticOscillationandrelationshipswithintraseasonalto interannualactivityinthetropicsduringtheaustralsummer. Journal ofClimate 18 :702–718,DOI:10.1175/JCLI-3284.1. ChambersDP,TapleyBD,StewartRH.1999.Anomalouswarmingin theIndianOceancoincidentwithElNi no. JournalofGeophysical Research 104 :3035–3047,DOI:10.1029/1998JC900085. ChanJCL,LiuKS.2009.Interannualvariationsoftropicalcyclone activityintheSouthernHemisphere.Paperpresentedat9thInternationalConferenceonSouthernHemisphereMeteorologyand Oceanography,AustralianMeteorologicalSociety:Melbourne,Victoria,Australia.http://www.bom.gov.au/events/9icshmo/manuscripts /M1500 Chan.pdf Chang-SengDS,MRJury.2010a.TropicalcyclonesintheSW IndianOcean.Part1:inter-annualvariabilityandstatistical prediction. MeteorologyandAtmosphericPhysics 106 :149–162, DOI:10.1007/s00703-009-0055-2. Chang-SengDS,MRJury.2010b.TropicalcyclonesintheSWIndian Ocean.Part2:structureandimpactsattheeventscale. Meteorology andAtmosphericPhysics 106 :163–178,DOI:10.1007/s00703-0100059-y. ChenG,TamC-Y.2010.Differentimpactsoftwokindsof PacicOceanwarmingontropicalcyclonefrequencyoverthe westernNorthPacic. GeophysicalResearchLetters 37 :L01803, DOI:10.1029/2009GL041708. ChiodiAM,HarrisonDE.2007.Mechanismsofsummertimesouthern IndianOceanseasurfacetemperaturevariability:Ontheimportance ofhumidityanomaliesandthemeridionaladvectionofwatervapor. JournalofClimate 20 :4835–4852,DOI:10.1175/JCLI4271.1. ChoiK-S,KimB-J,ChoiC-Y,NamJ-C.2009.Clusteranalysisof tropicalcyclonesmakinglandfallontheKoreanpeninsula. Advances inAtmosphericSciences 26 :202–210,DOI:10.1007/s00376-0090202-1. ChuJH,SampsonCR,LevineAS,FukadaE.2002.TheJointTyphoon WarningCentertropicalcyclonebest-tracks,1945–2000. Naval ResearchLaboratoryReferenceNumberNRL/MR/7540-02-16 ChuP-S.2004.ENSOandtropicalcycloneactivity. Hurricanesand Typhoons:Past,Present,andFuture .MurnaneRJ,LiuK-B.(eds). ColumbiaUniversityPress:NewYork. CookKH.2000.TheSouthIndianConvergenceZoneandinterannual rainfallvariabilityoversouthernAfrica. JournalofClimate 13 :3789–3804,DOI:10.1175/1520-0442(2000)013 < 3789:TSICZA > 2.0.CO;2. DrosdowskyW,ChambersLE.2001.Near-Globalseasurface temperatureanomaliesaspredictorsofAustralianseasonal rainfall. JournalofClimate 14 :1677–1687,DOI:10.1175/15200442(2001)014 < 1677:NACNGS > 2.0.CO;2. DunnOJ.1964.Multiplecomparisonsusingranksums. Technometrics 6 :241–252. ElsnerJB.2003.Trackinghurricanes. BulletinoftheAmerican MeteorologicalSociety 84 :353–356,DOI:10.1175/BAMS-84-3353. ElsnerJB,LiuK-B.2003.ExaminingtheENSO-typhoonhypothesis. ClimateResearch 25 :43–54,DOI:10.3354/cr025043. EvansJL,AllanRJ.1992.ElNi no/SouthernOscillationmodication tothestructureofthemonsoonandtropicalcycloneactivityin Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 15

ENSOANDSIODINFLUENCESONSOUTHWESTERNINDIANTCTRAJECTORIES55theAustralasianregion. InternationalJournalofClimatology 12 : 611–623,DOI:10.1002/joc.3370120607. FauchereauN,PohlB,ReasonCJC,RouaultM,RichardY.2009. RecurrentdailyOLRpatternsintheSouthernAfrica/Southwest IndianOceanregion,implicationsforSouthAfricanrainfallandteleconnections. ClimateDynamics 32 :575–591,DOI:10.1007/s00382008-0426-2. FauchereauN,TrzaskaS,RichardY,RoucouP,CamberlinP.2003. Sea-surfacetemperatureco-variabilityinthesouthernAtlanticand IndianOceansanditsconnectionswiththeatmosphericcirculation intheSouthernHemisphere. InternationalJournalofClimatology 23 :663–677,DOI:10.1002/joc.905. GongX,RichmanMB.1995.Ontheapplicationofclusteranalysis togrowingseasonprecipitationdatainNorthAmericaeastof theRockies. JournalofClimate 8 :897–931,DOI:10.1175/15200442(1995)008 < 0897:OTAOCA > 2.0.CO;2. GrayWM,SheafferJD.1991.ElNi noandQBOinuenceson tropicalcycloneactivity. Teleconnectionslinkingworldwideclimate anomalies .GlantzMH,KatzRW,NichollsN.(eds).Cambridge UniversityPress:Cambridge,UK;535. HarangozoSA,HarrisonMSJ.1983.Ontheuseofsynopticdatain indicatingthepresenceofcloudbandsoversouthernAfrica. South AfricanJournalofScience 79 :413–414. HermesJC,ReasonCJC.2005.OceanmodeldiagnosisofinterannualcoevolvingSSTvariabilityintheSouthIndianand SouthAtlanticOceans. JournalofClimate 18 :2864–2882, DOI:10.1175/JCLI3422.1. HigginsJJ.2004. AnIntroductiontoModernNonparametricStatistics Brooks/Cole:PacicGrove,CA;366. HoC-H,KimJ-H,JeongJ-H,KimH-S,ChenD.2006.Variation oftropicalcycloneactivityintheSouthIndianOcean:ElNi noSouthernOscillationandMadden-JulianOscillationeffects. Journal ofGeophysicalResearch 111 :D22101,DOI:10.1029/2006JD007289. HoerlingMP,KumarA,ZhongM.1997.ElNi no,LaNi na,and thenonlinearityoftheirteleconnections. JournalofClimate 10 : 1769–1786,DOI:10.1175/1520-0442(1997)010 < 1769:ENOLNA > 2.0.CO;2. HuangB,ShuklaJ.2007a.Mechanismsfortheinterannualvariability inthetropicalIndianOcean.PartI:Theroleofremoteforcing fromthetropicalPacic. JournalofClimate 20 :2917–2936, DOI:10.1175/JCLI4151.1. HuangB,ShuklaJ.2007b.Mechanismsfortheinterannualvariability inthetropicalIndianOcean.PartII:Regionalprocesses. Journalof Climate 20 :2937–2960,DOI:10.1175/JCLI4169.1. JuryMR.1993.Apreliminarystudyofclimatologicalassociations andcharacteristicsoftropicalcyclonesintheSWIndian Ocean. MeteorologyandAtmosphericPhysics 51 :101–115, DOI:10.1007/BF01080882. JuryMR,HuangB.2004.TheRossbywaveasakeymechanismof IndianOceanclimatevariability. Deep-SeaResearchPartI:OceanographicResearchPapers 51 :2123–2136,DOI:10.1016/j.dsr.2004. 06.005. JuryMR,PathackB.1991.Astudyofclimateandweathervariability overthetropicalSouthwestIndianOcean. Meteorologyand AtmosphericPhysics 47 :37–48,DOI:10.1007/BF01025825. JuryMR,PathackB,ParkerB.1999.Climaticdeterminantsand statisticalpredictionoftropicalcyclonedaysintheSouthwestIndian Ocean. JournalofClimate 12 :1738–1746. JuryMR,PathackB,WangB,PowellM,RaholijaoN.1993.A destructivetropicalcycloneseasonintheSWIndianOcean:JanuaryFebruary1984. SouthAfricanGeographicalJournal 75 :53–59. KarolyDJ.1989.SouthernHemispherecirculationfeaturesassociated withElNi no-SouthernOscillationevents. JournalofClimate 2 :1239–1252,DOI:10.1175/1520-0442(1989)002 < 1239:SHCFAW > 2.0.CO;2. KimH-M,WebsterPJ,CurryJA.2009.Impactofshiftingpatternsof PacicOceanwarmingonNorthAtlantictropicalcyclones. Science 325 :77–80,DOI:10.1126/science.1174062. KimJ-H,HoC-H,ChuP-S.2010.Dipolarredistributionof summertimetropicalcyclonegenesisbetweenthePhilippine SeaandthenorthernSouthChinaSeaanditspossible mechanisms. JournalofGeophysicalResearch 115 :D06104, DOI:10.1029/2009JD012196. KleinSA,SodenBJ,LauN-C.1999.Remoteseasurfacetemperature variationsduringENSO:Evidenceforatropicalatmospheric bridge. JournalofClimate 12 :917–932,DOI:10.1175/15200442(1999)012 < 0917:RSSTVD > 2.0.CO;2. KlinmanMG,ReasonCJC.2008.Onthepeculiarstormtrack ofTCFavioduringthe2006–2007SouthwestIndianOcean tropicalcycloneseasonandrelationshipstoENSO. Meteorology andAtmosphericPhysics 100 :233–242,DOI:10.1007/s00703-0080306-7. KnaffJA,SampsonCR.2009.Southernhemispheretropicalcyclone intensityforecastmethodsattheJointTyphoonWarningCenter,Part I:Controlforecastsbasedonclimatologyandpersistence. Australian MeteorologicalandOceanographicJournal 58 :1–7. KossinJP,CamargoSJ,SitkowskiM.2010.Climatemodulationof NorthAtlantichurricanetracks. JournalofClimate 23 :3057–3076, DOI:10.1175/2010JCLI3497.1. KossinJP,KnappKR,VimontDJ,MurnaneRJ,HarperBA.2007. Agloballyconsistentreanalysisofhurricanevariabilityand trends. GeophysicalResearchLetters 34 :L04815,DOI:10.1029/ 2006GL028836. KruskalWH,WallisWA.1952.Useofranksinone-criterionvariance analysis. JournaloftheAmericanStatisticalAssociation 47 : 583–621. KuleshovY,deHoedtG.2003.Tropicalcycloneactivityinthe SouthernHemisphere. BulletinoftheAustralianMeteorologicaland OceanographicSociety 16 :135–137. KuleshovY,FawcettR,QiL,TrewinB,JonesD,McBrideJ, RamsayH.2010.TrendsintropicalcyclonesintheSouthIndian OceanandtheSouthPacicOcean. JournalofGeophysicalResearch 115 :D01101,DOI:10.1029/2009JD012372. KuleshovY,MingFC,QiL,ChouaibouI,HoareauC,RouxF.2009. TropicalcyclonegenesisintheSouthernHemisphereandits relationshipwiththeENSO. AnnalesGeophysicae 27 :2523–2538. KuleshovY,QiL,FawcettR,JonesD.2008.Ontropicalactivityin theSouthernHemisphere:TrendsandtheENSOconnection. GeophysicalResearchLetters 35 :L14S08,DOI:10.1029/2007GL032983. LanderMA.1994.Anexploratoryanalysisoftherelationshipbetween tropicalstormformationintheWesternNorthPacicandENSO. MonthlyWeatherReview 122 :636–651. LandseaCW.2000.ElNi no-SouthernOscillationandtheseasonalpredictabilityoftropicalcyclones. ElNi no:ImpactsofMultiscaleVariabilityonNaturalEcosystemsandSociety .DiazHF,MarkgrafV. (eds).CambridgeUniversityPress:Cambridge,UK. LarkinNK,HarrisonDE.2002.ENSOwarm(ElNi no)andcold (LaNi na)eventlifecycles:Oceansurfaceanomalypatterns,their symmetries,asymmetries,andimplications. JournalofClimate 15 : 1118–1140,DOI:10.1175/1520-0442(2002)015 < 1118:EWENOA > 2.0.CO;2. LattinJM,CarrollJD,GreenPE.2003. AnalyzingMultivariateData Brooks/Cole:PacicGrove,CA;556. LauN-C,LeetmaaA,NathMJ.2008.Interactionsbetweenthe responsesofNorthAmericanclimatetoElNi no-LaNi naandtothe secularwarmingtrendintheIndian-WesternPacicOceans. Journal ofClimate 21 :476–494,DOI:10.1175/2007JCLI1899.1. LeroyA,WheelerMC.2008.Statisticalpredictionofweeklytropical cycloneactivityintheSouthernHemisphere. MonthlyWeather Review 136 :3637–3654,DOI:10.1175/2008MWR2426.1. L’HeureuxML,ThompsonDWJ.2006.Observedrelationships betweentheElNi no-SouthernOscillationandtheextratropicalzonal-meancirculation. JournalofClimate 19 :276–287, DOI:10.1175/JCLI3617.1. LindesayJA,HarrisonMSJ,HaffnerMP.1986.TheSouthern OscillationandSouthAfricanrainfall. SouthAfricanJournalof Science 82 :196–198. LyonsSW.1991.Originsofconvectivevariabilityoverequatorial southernAfricaduringaustralsummer. JournalofClimate 4 :23–39, DOI:10.1175/1520-0442(1991)004 < 0023:OOCVOE > 2.0.CO;2. ManhiqueAJ,ReasonCJC,RydbergL,FauchereauN.2011.ENSO andIndianOceanseasurfacetemperaturesandtheirrelationships withtropicaltemperatetroughsoverMozambiqueandtheSouthwest IndianOcean. InternationalJournalofClimatology 31 :1–13, DOI:10.1002/joc.2050. MasonSJ,JuryMR.1997.ClimaticvariabilityandchangeoversouthernAfrica:areectiononunderlyingprocesses. ProgressinPhysical Geography 21 :23–50,DOI:10.1177/030913339702100103. MavumeAF,RydbergL,RouaultM,LutjeharmsJRE.2009.ClimatologyandlandfalloftropicalcyclonesintheSouth-WestIndian Ocean. WesternIndianOceanJournalofMarineScience 8 :15–36. MeehlGA.1987.Theannualcycleandinterannualvariabilityinthe tropicalPacicandIndianOceanregions. MonthlyWeatherReview 115 :27–50,DOI:10.1175/1520-0493(1987)115 < 0027:TACAIV > 2.0.CO;2. Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)

PAGE 16

56K.D.ASHANDC.J.MATYAS NaeraaM,JuryMR.1998.TropicalcyclonecompositestructureandimpactsovereasternMadagascarduringJanuaryMarch1994. MeteorologyandAtmosphericPhysics 65 :43–53, DOI:10.1007/BF01030268. NicholsonSE.1997.AnanalysisoftheENSOsignalinthe tropicalAtlanticandwesternIndianOceans. International JournalofClimatology 17 :345–375,DOI:10.1002/(SICI)10970088(19970330)17:4 < 345::AID-JOC127 > 3.0.CO;2–3. NicholsonSE.2003.Commentson“TheSouthIndianConvergenceZoneandinterannualrainfallvariabilityoversouthernAfrica”andthequestionofENSO’sinuenceonsouthernAfrica. JournalofClimate 16 :555–562,DOI:10.1175/15200442(2003)016 < 0555:COTSIC > 2.0.CO;2. PanYH,OortAH.1983.Globalclimatevariationsconnectedwithsea surfacetemperatureanomaliesintheeasternequatorialPacicOcean forthe1958-73period. MonthlyWeatherReview 111 :1244–1258, DOI:10.1175/1520-0493(1983)111 < 1244:GCVCWS > 2.0.CO;2. ParkerBA,JuryMR.1999.Synopticenvironmentofcomposite tropicalcyclonesintheSouth-WestIndianOcean. SouthAfrican JournalofMarineScience 21 :99–115. ParkerD,1999.Criteriaforevaluatingtheconditionofatropical cyclonewarningsystem. Disasters 23 :193–216,DOI:10.1111/14677717.00113. PohlB,FauchereauN,RichardY,RouaultM,ReasonCJC.2009. Interactionsbetweensynoptic,intraseasonalandinterannual convectivevariabilityoversouthernAfrica. ClimateDynamics 33 : 1033–1050,DOI:10.1007/s00382-008-0485-4. QianW,HuH,DengY,TianJ.2002.Signalsofinterannualand interdecadalvariabilityofair-seainteractioninthebasin-wideIndian Ocean. Atmosphere-Ocean 40 :1480–9214,DOI:10.3137/ao.400302. ReasonCJC.1999.Interannualwarmandcooleventsinthe subtropical/mid-latitudeSouthIndianOceanregion. Geophysical ResearchLetters 26 :215–218,DOI:10.1029/1998GL900285. ReasonCJC.2001.SubtropicalIndianOceanSSTdipoleevents andsouthernAfricanrainfall. GeophysicalResearchLetters 28 : 2225–2227,DOI:10.1029/2000GL012735. ReasonCJC.2002.SensitivityofthesouthernAfricanCirculation todipolesea-surfacetemperaturepatternsintheSouthIndian Ocean. InternationalJournalofClimatology 22 :377–393, DOI:10.1002/joc.744. ReasonCJC.2007.TropicalcycloneDera,theunusual2000/01tropical cycloneseasonintheSouthwestIndianOceanandassociatedrainfall anomaliesoversouthernAfrica. MeteorologyandAtmospheric Physics 97 :181–188,DOI:10.1007/s00703-006-0251-2. ReasonCJC,AllanRJ,LindesayJA,AnsellTJ.2000.ENSOand climaticsignalsacrosstheIndianOceanbasinintheglobal context:PartI,Interannualcompositepatterns. International JournalofClimatology 20 :1285–1327,DOI:10.1002/10970088(200009)20:11 < 1285::AID-JOC536 > 3.0.CO;2-R. ReasonCJC,KeibelA.2004.TropicalcycloneElineanditsunusual penetrationandimpactsoverthesouthernAfricanmainland. WeatherandForecasting 19 :789–805,DOI:10.1175/15200434(2004)019 < 0789:TCEAIU > 2.0.CO;2. RevellCG,GoulterSW.1986.SouthPacictropicalcyclonesandthe SouthernOscillation. MonthlyWeatherReview 114 :1138–1145. ReynoldsRW,RaynerNA,SmithTM,StokesDC,WangW.2002. AnimprovedinsituandsatelliteSSTanalysisforclimate. JournalofClimate 15 :1609–1625,DOI:10.1175/15200442(2002)015 < 1609:AIISAS > 2.0.CO;2. RomesburgHC.1984. ClusterAnalysisforResearchers .Wadsworth: NorthCarolina;334. RouxF,Chane-MingF,Lasserre-BigorryA,NuissierO.2004. StructureandevolutionofintensetropicalcycloneDina nearLaR eunionon22January2002:GB-EVTDanalysis ofsingleDopplerradarobservations. JournalofAtmospheric andOceanicTechnology 21 :1501–1518,DOI:10.1175/15200426(2004)021 < 1501:SAEOIT > 2.0.CO;2. SajiNH,GoswamiBN,VinayachandranPN,YamagataT.1999.A dipolemodeinthetropicalIndianOcean. Nature 401 :360–363. SilvaJA,EriksenS,OmbeZA.2010.DoubleexposureinMozambique’sLimpopoRiverbasin. TheGeographicalJournal 176 :6–24, DOI:10.1111/j.1475–4959.2009.00343.x. SuzukiR,BeheraSK,IizukaS,YamagataT.2004.IndianOcean subtropicaldipolesimulatedusingacoupledgeneralcirculationmodel. JournalofGeophysicalResearch 109 :C09001, DOI:10.1029/2003JC001974. TerrayP,DominiakS.2005.IndianOceanseasurfacetemperatureand ElNi no-SouthernOscillation:Anewperspective. JournalofClimate 18 :1351–1368,DOI:10.1175/JCLI3338.1. ToddM,WashingtonR.1999.Circulationanomaliesassociated withtropical-temperatetroughsinsouthernAfricaandthe SouthWestIndianOcean. ClimateDynamics 15 :937–951, DOI:10.1007/s003820050323. ToddMC,WashingtonR,PalmerPI.2004.Watervapourtransport associatedwithtropical-temperatetroughsystemsoversouthern AfricaandtheSouthwestIndianOcean. InternationalJournalof Climatology 24 :555–568,DOI:10.1002/joc.1023. TrenberthKE.1990.Recentobservedinterdecadalclimatechangesin theNorthernHemisphere. BulletinoftheAmericanMeteorological Society 71 :988–993,DOI:10.1175/1520-0477(1990)071 < 0988: ROICCI > 2.0.CO;2. TrenberthKE.1997.ThedenitionofElNi no. BulletinoftheAmerican MeteorologicalSociety 78 :2771–2777. TrenberthKE,StepaniakDP.2001.IndicesofElNi noevolution. JournalofClimate 14 :1697–1701. TysonPD,Preston-WhyteRA.2000. TheWeatherandClimateof SouthernAfrica. OxfordUniversityPress:Oxford,UK;396. vandenHeeverSC,D’AbretonPC,TysonPD.1997.Numerical simulationoftropical-temperatetroughsoversouthernAfricausing theCSURAMSmodel. SouthAfricanJournalofScience 93 : 359–365. vanLoonH,RogersJC.1981.TheSouthernOscillation.Part II:Associationswithchangesinthemiddletropospherein theNorthernwinter. MonthlyWeatherReview 109 :1163–1168, DOI:10.1175/1520-0493(1981)109 < 1163:TSOPIA > 2.0.CO;2. VenzkeS,LatifM,VillwockA.2000.TheCoupledGCMECHO-2. PartII:IndianOceanresponsetoENSO. JournalofClimate 13 : 1371–1383. VitartF,AndersonD,StockdaleT.2003.SeasonalforecastingoftropicalcyclonelandfalloverMozambique. JournalofClimate 16 : 3932–3945,DOI:10.1175/1520-0442(2003)016 < 3932:SFOTCL > 2.0.CO;2. VitartF,LeroyA,WheelerMC.2010.Acomparisonofdynamical andstatisticalpredictionsofweeklytropicalcycloneactivityinthe SouthernHemisphere. MonthlyWeatherReview 138 :3671–3682, DOI:10.1175/2010MWR3343.1. WashingtonR,PrestonA.2006.Extremewetyearsoversouthern Africa:RoleofIndianOceanseasurfacetemperatures. Journalof GeophysicalResearch 111 :D15104,DOI:10.1029/2005JD006724. WashingtonR,ToddM.1999.Tropical-TemperatelinksinsouthernAfricanandSouthwestIndianOceansatellite-deriveddaily rainfall. InternationalJournalofClimatology 19 :1601–1616, DOI:10.1002/(SICI)1097-0088(19991130)19:14 < 1601::AID-JOC 407 > 3.3.CO;2-S. WebsterPJ,MooreAM,LoschniggJP,LebenRR.1999.Coupled ocean-atmospheredynamicsintheIndianOceanduring1997-98. Nature 401 :356–360. WilksDS.2006. StatisticalMethodsintheAtmosphericSciences Elsevier:Boston;627. WolterK.1987.TheSouthernOscillationinsurfacecirculationand climateoverthetropicalAtlantic,easternPacic,andIndianOceans ascapturedbyclusteranalysis. JournalofClimateandApplied Meteorology 26 :540–558. XieS-P,AnnamalaiH,SchottFA,McCreary,JPJr.2002.StructureandmechanismsofSouthIndianOceanclimatevariability. JournalofClimate 15 :864–878,DOI:10.1175/15200442(2002)015 < 0864:SAMOSI > 2.0.CO;2. YipCL,WongKY,LiPW.2006.Datacomplexityintropical cyclonepositioningandclassication. DataComplexityinPattern Recognition .BasuM,HoTK.(eds).Springer-Verlag:Heidelberg, DOI:10.1007/978-1-84628-172-3 13. YooS-H,YangS,HoC-H.2006.VariabilityoftheIndianOcean seasurfacetemperatureanditsimpactsonAsian-Australian monsoonclimate. JournalofGeophysicalResearch 111 :D03108, DOI:10.1029/2005JD006001. ZinkeJ,DulloW-C,HeissGA,EisenhauerA.2004.ENSOand IndianOceansubtropicaldipolevariabilityisrecordedina coralrecordoffsouthwestMadagascarfortheperiod1659 to1995. EarthandPlanetaryScienceLetters 228 :177–194, DOI:10.1016/j.epsl.2004.09.028. Copyright2010RoyalMeteorologicalSociety Int.J.Climatol. 32 :41–56(2012)