KeithS.Noll
SpaceTelescopeScienceInstitute,3700SanMartinDr.,Baltimore,MD21218
noll@stsci.edu
arXiv:astro-ph/0605606v1 23 May 2006HaroldF.Levison
Dept.ofSpaceStudies,SouthwestResearchInstitute,Ste.400,1050WalnutSt.,Boulder,CO
80302
hal@boulder.swri.edu
WillM.Grundy
LowellObservatory,1400W.MarsHillRd.,Flagstaff,AZ86001
grundy@lowell.eduDeniseC.Stephens
JohnsHopkinsUniversity,Dept.PhysicsandAstronomy,Baltimore,MD21218
stephens@pha.jhu.edu
ABSTRACT
Wehaveidentifiedabinarycompanionto(42355)2002CR46inourongoingdeepsurveyusingtheHubbleSpaceTelescope’sHighResolutionCamera.Itisthefirstcom-paniontobefoundaroundanobjectinanon-resonantorbitthatcrossestheorbitsofgiantplanets.Objectsinorbitsofthiskind,theCentaurs,haveexperiencedrepeatedstrongscatteringwithoneormoregiantplanetsandthereforethesurvivalofbinariesinthistransientpopulationhasbeeninquestion.MonteCarlosimulationssuggest,however,thatbinariesin(42355)2002CR46-likeheliocentricorbitshaveahighproba-bilityofsurvivalforreasonableestimatesofthebinary’sstill-unknownsystemmassandseparation.BecauseCentaursarethoughttobeprecursorstoshortperiodcomets,thequestionoftheexistenceofbinarycometsnaturallyarises;nonehasyetbeendefini-tivelyidentified.ThediscoveryofonebinaryinasampleofeightobservedbyHSTsuggeststhatbinariesinthispopulationmaynotbeuncommon.Subjectheadings:Centaurs;KuiperBeltObjects;Satellites,General
–2–
1.
Introduction
In1977aslow-moving,apparentlyasteroidalobject,1977UB,wasdiscoveredinthecourseofasystematicsurveyoftheeclipticplane(Kowal1977).Withtheidentificationofseveralprediscoveryimagesgoingbackasfaras1895(Lilleretal.1979a,b;Kowaletal.1979)itquicklybecameapparentthatthisunusualobjectinhabitedachaoticorbitwithcloseapproachestobothSaturnandUranus(OikawaandEverhart1979,Scholl1979).Thisobjecteventuallyreceivedtheminorplanetdesignation2060andwasnamedChironatthesuggestionofKowal.HefurthersuggestedthatsimilarobjectsalsobenamedformythologicalCentaurs.ThediscoveryoftheKuiperBelt,startingwith1992QB1(JewittandLuu1993),hasledtotherealizationthatobjectslikeChiron,inshort-livedchaoticorbits,musthavebeenrecentlyperturbedfromthemorepopulousandmorestabletransneptunianpopulation.Inaddition,asmanymoreobjectshavebeenfound,thedistinctionbetweentheCentaursandtransneptunianobjects(TNOs),particularlytheScatteredDiskobjects(SDOs),hasbecomeblurred.
AfunctionaldefinitionforCentaursisthattheyareobjectsintheoutersolarsystemonnon-resonant,giant-planet-crossingorbits.Asaresultoffrequentencounterswiththeplanetswhoseorbitstheycross,theyhavedynamicallifetimesthatareshort,ontheorderoftensofmillionsofyears.Computation-intensivecalculationsarerequiredtodeterminetheorbitalstabilityofanyparticularobject(Horneretal.2004a,b).However,itispossibletoidentifyclassesofobjectsthatarelikelyhaveunstableorbits.TheoriginaldefinitionofCentaursincludedonlythoseobjectswithsemimajoraxes,a,lyingbetweentheorbitsofJupiterandNeptune.Holman(1997)hasshownthat,exceptforasmallregionofstabilityforloweccentricityorbitsbetween24and27AUandforobjectstrappedin1:1resonanceneartheL4orL5LagrangepointsofJupiterandNeptune,objectsinthisregionhavedynamicallifetimesof108yearsorlesswiththecurrentgiantplanetconfiguration.Holmanalsonotesthatplanetarymigrationwouldbelikelytoremoveeventheonesmallregionofmarginalorbitstability.Thus,thereisaveryhighlikelihoodthatanyobjectwithasemimajoraxisbetweenJupiterandNeptunewillhaveanunstableorbit.Mostoftheknownnon-TrojanobjectswithheliocentricsemimajoraxesbetweenJupiter’sperihelionandNeptune’saphelion,qJupiter –3– aswell:“nonresonantobjectswhoseosculatingperiheliaarelessthantheosculatingsemimajoraxisofNeptuneatanytimeduringtheintegration”(Elliotetal.2005).WenotethatamoreconsistentdefinitionwouldalsoincludeobjectswithperiheliathatcomewithinNeptune’sHillradiusofitsosculatingaphelion,althoughthisisaminorquibble.Othertaxonomicalschemesarepossible(e.g.Levison1996),however,inthispaperweadopttheDESdefinitionofCentaurs.IntheDESscheme,thesubjectofthispaper,(42355)2002CR46,isaCentauronanunstableorbitthatcrossestheorbitsofbothNeptuneandUranuswitha=38.1AUandq=17.5AU. Thefirsttransneptunianbinary(TNB)wasidentifiedin1978withthediscoveryofCharon(ChristyandHarrington1978),coincidentallycontemporaneouswiththediscoveryofthenear-homonymously-namedChiron.ThesubsequentdiscoveryofalargepopulationofTNOsandthemorerecentdiscoveryofnumerousboundsystemsprofoundlychangesthecontextinwhichweviewbinaries(seereviewbyNoll(2006)forarecentsummary).Binariesandmorecomplexboundsystemscannolongerberegardedaslowprobabilityoutcomesofstatisticallyunlikelyevents,butmustinsteadbeexplainedasanaturalconsequenceofsmallbodyevolutionintheoutersolarsystem.Sufficientdataonbinaryfrequencyhavenowaccumulatedthatdifferencesinthebinaryfrequencybetweendifferentdynamicalpopulationsarebeginningtobeidentifiable.StephensandNoll(2006)haveidentifiedanenhancedoccurrenceofbinariesinthecoldclassicalpopulationcomparedtootherdynamicalgroupsofTNOs.Thisdifferencehashelpedreinforcespeculationastotheinstabilityofbinariesinpopulationsthathavehadstronginteractionswithgiantplanets.Centaurshaveexperiencedrepeatedstrongscatteringeventswithoneormoregiantplanets.Itisanopenquestionwhethertypicalscatteringeventsarecapableofdisruptingasignificantfractionofweaklyboundsystems,aquestionperhapsbestaddressedbybothobservationandtheory.Inordertoobtaintheleastbiasedstatisticsontheoccurrenceofbinariesindifferentsub-populationsofsmallbodiesintheoutersolarsystemwehaveundertakenasurveyusingtheHighResolutionCamera(HRC),acomponentoftheAdvancedCameraforSurveys(ACS)ontheHubbleSpaceTelescope(HST).Ourprogramwillobserveasubsetof250possibletargetsdistributedoverallknowndynamicalclassesinasnapshotprogramthatisactivefromJuly2005throughJune2007.ObservationswiththeHRCaremuchmoresensitivetobinariesthanprevioussearchesbothintermsofthelimitingmagnitudeandintheangularresolutionoftheinstrument.TheprogramincludesanemphasisoncurrentlyundersampledpopulationsincludingtheCentaurs.Inthispaperwedescribethediscoveryofabinarycompanionto(42355)2002CR46,thefirstknownbinaryCentaur. 2.ObservationandAnalysis Observationsof(42355)2002CR46weremadewithACS/HRCon20January2006,09:50-10:14UT.TheHRCwasusedinanon-standardconfigurationwiththeClearfiltersinbothfilterwheelsrotatedintotheopticalpath.TheuseoftwoClearfiltersresultsinafactoroftwoincreaseinthroughputcomparedtothewidestavailablestandardfilteratthecostofaslightdegradation –4– inimagequality(GillilandandHartig2003).Thebandpassisdeterminedbythedetectorresponseandisequivalenttoafiltercenteredat610nmwithaFWHMof520nm.Four300secexposuresof(42355)2002CR46wereobtained.ThetelescopewasditheredbetweeneachexposureusingthestandardACS/HRCboxditherpattern.Theuseofdithersensuresthatfixed-patterndefectssuchashotpixelscanberemovedwhentheexposuresarecoadded.Byobtainingfourexposuresweareabletoeliminateessentiallyallcosmicrayswhentheindividualexposuresarecombined. Thetelescopetrackedthemotionof(42355)2002CR46duringtheexposuresandcorrectedforparallaxintroducedbyHST’sorbitalmotion.Observationsforallobjectsinourprogramhavebeenscheduledwitharestrictionthatthetargetandbackgroundsourcesmusthaverelativemotionsataminimumrateof5×10−5arcsec/sectoensurethatwecandiscriminatebetweenthetargetandbackgroundsources.Allofthetargetsobservedtodatehavebeensuccessfullyacquiredinthe29×26arcsecfieldofviewoftheHRC’s1024×1024CCDarray. DatawerereducedusingthestandardCALACSpipelinecalibration.Formovingtargets,westartouranalysiswiththedark-subtracted,flat-fieldcorrected∗flt.fitsimagefiles;the∗drz.fitscombinedimagefilefromthestandardpipelineisnotuseableformovingtargets.Thebinarycompanionto(42355)2002CR46isapparentineachoftheindividualflat-fieldedimageswithoutrequiringanyfurtherprocessing.However,inordertoproduceanimagethatisfreeofcosmicrays,hotpixels,andthathasthebestsignal-to-noiseratio(S/N),weremovegeometricdistortionandcombinetheimagesusingthemultidrizzletask(version2.6.7)intheSTSDASpackagerunninginPyRAF. Tocombinemovingtargetobservationswemustaddanonstandardsteptotheusualmultidrizzleworkflow.Becausethetargetistracked,thecoordinatekeywordsineachexposurearedifferent,reflectingthedifferentstartingpositionofeachintegration.Inordertoremovethetarget-trackingmotionofthetelescopethatwouldbemisinterpretedbymultidrizzle,weforcetheworldcoor-dinateheaderkeywordsCRVAL1andCRVAL2foreachofthefourexposurestobeequaltothevalueatthestartofthefirstexposure.Eachimageiscentroidedandshiftsarecomputedusingthetweakshiftscommand.Thecomputedshiftsareusedmakeafinalcombinationofthefourexposures.TheresultofthisprocessisshowninFigure1.ThesignalinthepeakpixelhasS/N=2700andthestandarddeviationinthebackgroundis0.0078counts/sec.The3sigma,peak-pixeldetectionlimitcorrespondstoasourceofV=27.7mag.Notethatpeak-pixelS/Ncanbeimprovedbyafactorof∼3whenthedataarefittoapoint-spread-function(PSF).ThelevelofperformanceobservedisinexcellentagreementwithpredictionsbasedontheACSexposuretimecalculatorsandconfirmsthesuccessofdataprocessing. WemeasuredtheseparationbyiterativelyfittingaPSFtoeachofthetwocomponentsof(42355)2002CR46ineachofthefourexposuresandinthecombinedexposure.TheiterativePSFfittingisdescribedindetailinStephensandNoll(2006).Wefoundaseparationof4.36±0.08pixels,equaltoanangularseparationof0.109±0.002arcsec.Thesecondarywasorientedatanangleof226.8±0.8degreesEastofNorth.PSFfittingalsoyieldedscalingfactorsthatweusedtoderive –5– relativephotometryofthetwocomponents.Wefoundthesecondarytobefainterthantheprimaryby1.47±0.04magnitudes.Themeasuredmagnitudedifferencecorrespondstoadiameterratioofd2/d1≈0.5,assumingbothcomponentshavethesamealbedo.ComputingstandardmagnitudesismoredifficultbecauseoftheextremelywideeffectivebandpassoftheClearfilterandtheuncertaincolorsoftheobjects;wedonotattemptithere. 3.Discussion (42355)2002CR46wasdiscoveredbytheNEATteamatPalomarandtheheliocentricorbitwasrefinedsoonthereafter(McNaughtetal.2002;StossandMarsden2002).AtthetimeofourHSTobservationthedistancefromEarthto(42355)2002CR46was16.675AU;aplanviewoftheorbitisshowninFigure2.Theprojectedseparationofthetwocomponentsontheplaneoftheskywas1330±130km.Theabsolutemagnitudeof(42355)2002CR46isHV=7.65±0.01(Tegleretal.2003).OpticalcolorsmeasuredforthisobjectareB−V=0.74±0.02andV−R=0.52±0.01(Tegleretal.2003);colorsthataretypicaloftherelatively“neutral”groupofCentaursandTNOs.ColorsmeasuredbyPeixinhoetal.(2004)differslightly,butaregenerallywithinthequoteduncertainties.Ortizetal.(2003)placedanupperlimitof0.15magsonanypossiblelightcurvefromtheunresolvedbinary.ThealbedooftheunresolvedbinarywasmeasuredbyStansberryetal.(2005)usingtheSpitzerSpaceTelescope;theyreportedp=0.08−0.12.Forassumedalbedosofp=0.1foreachcomponentandHV=7.65forthepair,wecalculateatotalequivalentdiameterof125km.Fora2/1diameterratiowecalculatecomponentdiametersof112kmand56kmfortheprimaryandsecondaryrespectively. Aninterestingquestiontoaskforanybinaryishowtightlyboundthesystemis.Therearetwowaystoanswerthisquestion,intermsofthetotalbindingenergyofthesystem,andintermsofthestabilityofthesystemrelativetoperturbationsfromotherbodies.Thebindingenergydependsontheseparationofthesecondaryfromtheprimary,aB,andonthesystemmass.By 3/a,assumingthatdensitiesofTNOsaresimilar(anadmittedlyshakyassumption),thequantityrpB whererpistheradiusoftheprimary,canbeusedtoinfertherelativebindingenergiesofboundsystems.Morefrequently,however,therelatedquantityaB/rpistabulated,aconventionwefollowhere.Basedonasingleobservationwecannotconstrainthesemimajoraxisofthebinary,aB,beyondthestatementaB>(1330/2)kmwhichwouldbetrueforthelimitingcaseofanorbitwitheccentricitye=1seenfaceon.Abetterguessforthesemimajoraxisistwicetheobservedseparation,2×1330≈2700km.Usingthis,weestimateaB/rp=48.ThisiscomparabletothevalueofaB/rp=56foundfor(66652)1999RZ253andsignificantlylessthanforanyotherTNBwithameasuredorbitexceptforthePluto/Charonsystem(Noll2006)indicatingthattheCR46systemismoretightlyboundthanmostotherknownbinaries.Toestimatethesystemstabilityrelativetothirdbodyperturbationsweassumeadensitynear1gcm−3givingaHillradius,rH,atthecurrentheliocentricsemimajoraxisof38.1AUontheorderof300,000km(notethattheinstantaneousHillradiusscaleswithdistancefromtheSun).Againusingtwicethecurrently –6– observedprojectedseparation,s,inplaceofthestill-to-be-determinedsemimajoraxis,aB,wefind2s/rH≈0.01.SeparationsontheorderoffewpercentoftheHillradiusarefairlytypicalinbinariesinthetransneptunian,MainBelt,andNearEarthpopulations(Noll2006)and,inthissense,(42355)2002CR46doesnotappeartobeunusual. 3.1.FrequencyofBinaryCentaurs Animportantquestioninthestudyofbinarysystemsinthetransneptunianandrelatedpopulationsiswhetherandhowthefractionofbinariesindifferentdynamicalpopulationsvary.Existingdatashowanincreasedfractionofbinariesinthecoldclassicaldiskcomparedtoanaggregateoftheremainingpopulations(StephensandNoll2006).Aplausible,butunverifiedpostulateisthatpopulationsthathaveexperienceddisruptivescatteringeventswillhaveasmallerfractionofbinariesthanpopulationsthathavehadquiescentdynamicalhistories.Centaursareveryshort-livedcomparedtotheageofthesolarsystemandduringtheirlivesundergomanyencounterswithgiantplanets.TheCentaurs,followingthislineofreasoning,mightbeexpectedtohavethelowestfractionofsurvivingbinariesasaresultofrepeatedscatteringencounterswithgiantplanets.However,whiletherehasbeeninformalspeculationonthistopic,theobservationsneededtoaddressitquantitativelyand/orthenumericalmodelingrequiredforasoundtheoreticalpredictionhavenotbeencompleted. TheDEScurrentlylists65objectsasCentaurs(Elliotetal.2005)usingtheirbroaddefinition.Noneofthesehasbeenfoundtobebinaryfromexistinggroundbasedobservations.Togaugewhetherthislackofbinarydiscoveryissurprising,itismostusefultomakeacomparisonwithbinariesfoundinthegeneralTNOpopulation.Oftheapproximately1000TNOsdiscoveredfromground-basedobservations,7havebeenidentifiedasbinariesfromnon-AOground-basedimages.Ifthesamefractionofground-detectablebinariesexistedintheCentaurpopulation,wewouldexpect1binary,atmost,tohavebeendetectedinthatsample.WenotethatthegroundbasedsurveyofSchallerandBrown(2003)failedtoidentifyanybinariesinasampleof150TNOs,againconsistentwiththeoverallinfrequencyofwidebinaries.ThesmallerheliocentricdistanceofCentaursfavorsthedetectionofsimilarlywidebinariescomparedtoTNBsbyafactorofordertwo;butthisisinsufficienttochangeourconclusionthatthecurrentlackofground-detectedCentaurbinariesdoesnotconstrainCentaurbinariestobelessfrequentthanTNBs. HighresolutionobservationswithHSTandwithadaptiveopticssystemsfromthegroundhavefoundthemajorityofknownTNBs(Noll2006),largelyduetothefactthatcomponentseparationsaretypicallymuchlessthan1arcsec.EightCentaurshavebeenobservedbyHSTusingfourdifferentimaginginstruments,theWideFieldPlanetaryCamera2(WFPC2),theSpaceTelescopeImagingSpectrograph(STIS),theNearInfraredCameraandMulti-ObjectSpectrometer(NICMOS),andtheACS/HRC(Table1).Whilethisisfarfrombeingahomogeneoussampleintermsofresolutionordepth,itisneverthelessfarlessheterogeneousthanground-baseddataandisareasonablestartingpointforestimatingthebinaryfrequencyinCentaurs. –7– Inordertoperformthemostsensitivepossiblesearchforfaintcompanions,wecombinedallavailabledataforeachobjectwithmultidrizzleusingthesamestepsasdescribedfor(42355)2002CR46.Thisanalysisstepnecessarilyinvolvedsomecompromises.Thedatawereobtainedinmultiplefil-ters,theF555W,F675WandF814WforWFPC2andtheF110WandF160WforNICMOS.Becausewearecombiningdatawithdifferentintrinsicpoint-spread-functions(PSF)welosetheabilitytoextractunresolvedbinariesfromPSF-fitting(e.g.StephensandNoll2006)butincreaseourabilitytodetectfaintresolvedcompanions.TheangulardetectionlimitfordataprocessedinthiswayisthengivenbythestandardNyquistcriterion.ForWFPC2thelimitis0.2arcsec,forNICMOSitis0.15arcsec.Wenote,asacaveat,thattheresolutionofthecombinedimagealsode-pendsontheaccuracyofthecentroidingoftheindividualframesandthecalculatedshiftsusedforalignment.ForlowS/Ndata((54520)2000PJ30)orforundersampleddata((60608)2000EE173and(87269)2000OO67)additionalsystematicerrorsmayaccrue. STISdataweretakeninfixedtargetmode;theCentaursmovebyseveralpixelsduringeachoftheintegrations.ThismotionisduemostlytoHSTparallaxandisthereforenonlinearbothindirectionandmagnitudeinthecourseofanorbit.Wecombinedasubsetofeightframesforeachoftheobjects,choosingtheintegrationswiththesmallestdegreeofblurring.Theangularresolutionofthecombinedimageiscomplexandisafunctionofpositionangle.However,weestimatethatacompanionatadistanceof0.2arcsecorgreaterwouldberesolved. Foralloftheobjectswehaveestimatedhowmuchfainteranobjectcouldhavebeendetectedinthecombinedimage.Inallcasesbutone,thecombinedimagehasasmoothbackgroundwithfewartifactsexceptforresidualsfromextendedsourcesthatareincompletelyremoved.Theapparentmotionof(87269)2000OO67relativetostarsinthefieldwassmallenoughthatthestarswerenotremovedbythemultidrizzlecosmicraystep(essentiallyamedianfilter).However,thebackgroundoutsidethesesourcesisflatandclean.Thestandarddeviationofthebackgroundwasdeterminedbyusingimstatoncleansubsectionsoftheimages.Weestimatedthedetectabilitylimitforafaintsourcetobethreetimesthisstandarddeviationinthepeakpixel.TheratioofthepeakpixelintheobservedCentaurtothisdetectabilitylimit,expressedinastronomicalmagnitudes,islistedinTable1.ThisdependsonthebrightnessoftheCentauraswellasthedepthoftheintegrationandvarieswidely.Itisworthnotingthatwiththeexceptionofthetwonewly-detectedcompanionsobservedaroundPluto(Weaveretal.2006),allotherknownbinarieshavemagnitudedifferencesof∼4.5magnitudesorless(Noll2006,Brownetal.2006);allbuttwoofourdatasetsprobeto4.5magnitudesordeeper.MostTNBs,17/25,havers/rp>0.5andwouldhavebeendetectableinallofourdatasets. ThedetectionofonebinaryintheHSTsampleofeightCentaurscorrespondstoabinaryfrequencyof13±195%atseparationsofgreaterthan0.1-0.2arcsecforcompanionswithinafactorof3ofthebrightnessoftheprimary.UpperlimitsforsignificantlyfaintercompanionsattheseseparationscanalsobederivedfromthissamplebasedonthedatainTable1.WecancomparetootherdynamicalclassesmeasuredfromNICMOSdata;StephensandNoll(2006)foundbinary 9 ratesof22±10%forobjectsinthecoldclassicalbeltand11.5±54%forSDOs.Withinthelarge –8– uncertainties,thefractionofbinaryCentaurscannotbedistinguishedfromeithertheSDOsortheColdClassicals.Itappearsunlikely,however,thatCentaurbinariesaresignificantlylessfrequentthanbinariesintheScatteredDisk. 3.2.Survivalofthe(42355)2002CR46Binary Itispossibletoestimatewhetherabinarysystemlike(42355)2002CR46couldsurviveasitevolvedfromanorbitoriginallyintheScatteredDiskintotheorbitinwhichitiscurrentlyfound.Suchanestimaterequirestwodistinctcalculations.First,wemustdetermine,atleaststatistically,thetrajectorythattheobjectfollowedafteritlefttheScatteredDisk,payingparticularattentiontocloseencounterswiththeplanets.Second,wemustdeterminewhatkindofplanetaryencounterswillleadtothetidalstrippingofthebinary. Fortunately,wecanaccomplishthefirsttaskbyemployingasetofnumericalorbitintegrationspresentedinLevisonetal.(2006a,hereafterL06).L06followedthedynamicalevolutionof2200objectsinitiallyintheScatteredDiskunderthegravitationaleffectsoftheSunandalltheplanets(exceptMercury,PlutoandotherlargeTNOs).Eachtrajectorywasfolloweduntiltheobjectwaseitherejectedfromthesolarsystem,impactedtheSunoraplanet,orreachedasemimajoraxisof1000AU,whereitwasassumeditwouldentertheOortcloud.Duringthesimulation,L06kepttrackofthecloseencounterswiththeplanets. Forourpurposes,wewillconsideranyobjectonanorbitwithsemimajoraxis37 –9– q=14.5AUandQ=30.5AU(seeLevisonandDuncan1997foradescriptionofthisbehavior).Afterthescatteringeventat12Myr,encounterswithUranusdrivetheevolutionofthisobject,andat17MyritevolvesontoaCR46-likeorbit. Intheir2003paper,TiscarenoandMalhotra(serendipitously)computedtheorbitforatestobjectwithinitialconditionsidenticalto(42355)2002CR46(TiscarenoandMalhotra2003,Fig.4).Theirtestobjectorbitevolvedwitharoughlyconstantperiheliondistancebutincreasingsemimajoraxisuntil,at34Myr,theobjectwaslosttotheinnersolarsystem.BecauseofthechaoticnatureofCentaurorbits,thisresultisnotunique,butisagaininstructiveasanexampleofthekindofbehaviorthatispossible. ThebehaviorsdescribedaboveillustratetheimportancethatcloseencounterswiththegiantplanetshaveincreatingobjectsonCR46-likeorbits.Herewedefine‘close’asanencounterinwhichtheobjectspasseswithintheHillsphereofaplanet.TheobjectshowninFigure3suffered147closeencounterswithNeptuneand41closeencounterswithUranusoverthe18Myrtime-frameshown.Notsurprisingly,forourensembleoffictitiousCR46’s,closeencounterswithNeptunearethemostprevalent.Theseobjectssufferedbetween39and1657(median288)sucheventsontheirwaytoaCR46-likeorbit.EncounterswithUranusnumberedbetween0and363,withamedianof57(only1ofour126fictitiousCR46’sdidnotsufferanencounterwithUranusatall).Interestingly,9%ofourobjectssufferedatleastoneencounterwithSaturnand3%withJupiterbeforeevolvingontoCR46-likeorbits. Whether(42355)2002CR46survivesasabinarythroughthisevolutiondepends,toalargeextent,onthestrengthoftheencounters,which,inturn,dependsontheclosestapproachdistances.InFigure4wepresentthecumulativeprobabilitythatoneofourobjectssufferedanencounterwithinagivendistanceofoneofthegiantplanetsonitswaytoaCR46-likeorbit.HalfofourfictitiousCR46’ssufferanencounternocloserthan∼0.1AUand90%ofthemdonotgetcloserthan0.03AU. Thenextstepinourinvestigationistodeterminewhethertheseencounterswilldisruptthebinary.Theexactoutcomeofanencounterdependsonmanyparametersincludingthecloseapproachdistance(D),thetotalmassofthebinarypair(mtot),themassoftheplanet(Mp),andthedetailsofthebinary’smutualorbit.However,AgnorandHamilton(2005)foundthatfornearlyparabolicencountersdisruptionbecomescommonfor D –10– (42355)2002CR46donotcontradictthisguess(Grundyetal.,inpreparation).Weestimatedabovethat(42355)2002CR46hasacombinedequivalentdiameterof125km,which,foranassumeddensityof1gcm−3,impliesasystemmassof1021g.Undertheseassumptions,wefindthatrtdrangesfrom0.0067AU(forNeptune)to0.018AU(forJupiter).TheupwardarrowsinFigure4showthevaluesofrtdforthefourgiantplanets. ComparingthecurvesinFigure4tothevaluesofrtdforthevariousplanets,weestimatethatthereisa95%chancethattheCR46binarywouldhavesurviveditstransitionfromtheScatteredDisktoitcurrentorbitifaB∼2700km.HalfofthethreattothebinarycomesfromUranusandhalffromJupiter.WecanturnthecalculationaroundanddeterminewhatvaluesofaBarevulnerabletodisruption.UsingthedatainFigure4andassumingmtot=1021g,wefindthatanobjectwithaB∼23,000kmhasonlya50%chanceofsurvival.Ifweassumethatthealbedoof(42355)2002CR46is1ratherthan0.1,i.e.theobjectsaretheminimumpossiblediametersandthereforelowerinmass,the50%survivalsemimajoraxisfallsto7300km. Finally,itisinterestingtoquestionwhether(42355)2002CR46mightbeexpectedtohaveasmallsemimajoraxisrelativetotheprimarydiameterbecauseanybinarythatwaslesstightlyboundwouldnothavesurvived.ThisquerycanbeaddressedbycalculatingrtdfortheknownTNBsandcomparingthesevaluestothecurvesinFigure4.Forthe8TNBs,excludingPluto,withmeasuredorbits(assummarizedinTable6ofNoll2006),wefindthatrtdwithrespecttoUranusrangesfrom0.007to0.09AU.Eveninthemostvulnerablecase,(88611)2001QT297,thereisa68%chancethatthebinarywouldhavesurvived.Thus,wecanconcludethat(42355)2002CR46’ssemimajoraxisisnotstronglyboundedbyadynamicalselectioneffect. 3.3.WhereAretheBinaryComets? Jupiterfamilycomets(JFCs)havebeenshowntoderivefromasourcepopulationonprograde,lowinclinationorbitsatlargerheliocentricdistancesandwere,indeed,thefirstevidencepointingtotheexistenceofatransneptunianpopulation,particularlytheSDOs(Duncanetal.2004andreferencestherein).Centaursareanaturalfeatureofthismodelassomewilltransitionfromlargetosmallheliocentricorbitsundertheinfluenceofthegiantplanets.IfbinaryCentaursexistatanyappreciablefraction,thenextnaturalquestioniswhethertherearebinaryJFCs.ThissamequestioncouldbeaskedforHalley-typecometssincetheirsourcemaybeSDOs(Levisonetal.2006b)andScatteredDiskbinarieshavebeenidentified(Noll2006;Nolletal.2006).Indeed,inviewofthefactthatbinariesseemtobeacommonphenomenoninvarioussmallbodypopulations,thequestioncouldaswellbeappliedtoOort-cloudcomets. Excludingobjectsthathavebeenobservedtodisintegrateintooneofmorepieces,therehasbeen,todate,onlyonecometofanykindwithevententativeclaimsforbinarity.Marchisetal.(1999)andSekanina(1999)bothclaimedevidenceofapossiblebinarynucleusforthelargeOort-cloudcometHale-Bopp.Sekaninaidentifiedacompanionwithadiameter0.4timesthatof –11– theprimaryinHSTWFPC2imagesprocessedtoremovelightfromthecoma.Marchisetal.(1999)pointedoutboththedoublepeakedcomatheyobtainafterimagerestorationandmorphologicalfeaturesofthejetsaspossibleevidenceofabinary.However,WeaverandLamy’s(1999)analysisofHale-BoppemployingsomeofthesameHSTdatausedbySekaninafoundapossiblysmallerdiameterforthenucleus(30-70kmvs.70km)andreportednoevidenceofasecondcomponent.Intheabsenceofadirectdetectionofabinarynucleus,wecanaskwhatotherlinesofindirectevidencemightpointtotheexistenceofbinarycomets.Cometfragmentationisawell-knownphenomenonthathasbeenobservedinmanycomets(Boehnhardt2004).Historically,theseeventshavebeeninterpretedasphysicalbreakupofasinglenucleus.Weposeaspeculativequestion:couldsomeoftheobservedsplitcometsbeduetoapreexistingbinary?Giventhepredominanceofsimilar-massbinariesinthetransneptunianpopulation(Noll2006),anexpectedoutcomeofchaos-assistedcapture(Astakhovetal.2005),themostlikelycandidatesinthisspeculativescenariowouldbethosecometsthatappeartosplitintonearlyequalsizedcomponentswithouttheobviousinfluencesolarheatingasadriverofthebreakup.AnexampleofasplitcometthatmightbeamenabletointerpretationasapreexistingbinaryisthecometpairC/2002A1andC/2002A2(LINEAR).ThesetwoobjectsareonverysimilarCentaur-likeorbitswithsemimajoraxesof18AU,periheliaat4.7AUandapheliaat31AUandnearlyidenticaleccentricities,inclinationsandmagnitudes(Spahretal.2002).Sekaninaetal.(2003)usedtheircometfragmentationcodeandrepeatedastrometricobservationstoinferthatthepairsplitnon-tidallyatadistanceof∼22AUfromtheSunintherecentpast.However,theydidnotconsiderthepossibilityofapreexistingbinaryintheirmodelanditisunclearwhatobservabledifferences,ifany,onemightexpectbetweenasingle,internallyfracturedprogenitorandatightbinary. Animportantancillaryissuemustbementioned.ThenucleiofJFCsaretypicallymuchsmallerthanknownCentaursorTNOs(Lamyetal.2004).Thedependenceofbinaryfractionondiameteriscurrentlyunknownandwillrequiresignificantinvestmentsofobservingtimetoanswerforanyofthesmall-bodypopulationsinquestion.Itisconceivable,however,thattherecouldbeadependencythatwouldweakenanycomparisonsbetweentens-of-km-scaleCentaursandkm-scaleJFCs.Nonetheless,giventheapparentexistenceofbinaryCentaurs,moreobservationalandtheoreticalattentiontothequestionofbinarycometsiscertainlyinorder. 4.Conclusions (42355)2002CR46isabinarysystem,thefirstbinaryCentaurtobeidentified.(42355)2002CR46’sorbitcrossestheorbitsofbothUranusandNeptuneanditis,therefore,inadynamicallyunstableorbit.Numericalsimulationsshowthat,onaverage,anobjectonitswaytoaCR46-likeorbitwillhavehadhundredsofcloseencounterswithgiantplanetswitha50%chanceofcomingascloseasD=0.1AUtoone.Thissystemis,then,anexcellentempiricaltestofthepropositionthatcloseplanetaryencounterscandisruptbinaries.Theexistenceofanobjectlike(42355)2002CR46rulesouttheextremeformulationofthepropositionthatallbinariesongiant-planet-crossingorbitswill –12– bedisrupted.Theseparationofthecomponentsofthe(42355)2002CR46systemcomparedtothesizeoftheprimaryissmallerthanformostknownTNBs;anindicationthatthissystemistightlybound.Inordertotesttheweakerformulationofthebinary-disruptionpostulate,i.e.thatsomefractionofplanet-crossingbinarieswillbedisruptedleavingonlythemosttightlyboundsystems,manymoreCentaurswillneedtobeobservedathighangularresolution. 5.Postscript Sincethismanuscriptwaspreparedandreviewed,fouradditionalCentaurshavebeenobservedinourongoingHSTobservingprogram.Threearesingleobjects((120061)2003CO1,(55576)Amycus,and(83982)Crantor).OnemoreCentaur,(65489)2003FX128,isbinary.Thedetectionofanotherbinaryinanunstablegiant-planet-crossingorbitstrengthenstheconclusionsofthispaper.WiththeadditionofthesefourCentaurs,thestatisticsforCentaurbinariesstandsat2outof12.Thedetailsoftheserecentobservationswillbecoveredinfuturepublications. ThisworkisbasedonobservationsmadewiththeNASA/ESAHubbleSpaceTelescope.Theseobservationsareassociatedwithprogram#10514.Supportforprogram#10514wasprovidedbyNASAthroughagrantfromtheSpaceTelescopeScienceInstitute,whichisoperatedbytheAssociationofUniversitiesforResearchinAstronomy,Inc.,underNASAcontractNAS5-26555.HFLisgratefulthePG&GandOriginsforcontinuingsupport. –13–REFERENCES Agnor,C.,andHamilton,D.P.2005.SatelliteCaptureviaBinaryExchangeReactions:Application toTriton.AGUFallMeetingAbstracts,P11B-0120.Astakhov,S.A.,Lee,E.A.,&Farrelly,D.2005.FormationofKuiper-beltbinariesthroughmultiple chaoticscatteringencounterswithlow-massintruders.MNRAS,360,401-415.Boehnhardt,H.2004.SpiltComets.In:Festou,M.C.,Keller,H.U.,andWeaver,H.A.(Eds.), CometsII.UniversityofArizonaPress,Tucson,pp.301-316.Brown,M.E.,and14colleagues2006.SatellitesoftheLargestKuiperBeltObjects.ApJ,639, L43-46.Christy,J.W.,andHarrington,R.S.1978.TheSatelliteofPluto.AJ,83,1005-1008. Duncan,M.,Levison,H.,andDones,L.2004.DynamicalEvolutionofEclipticComets.In:Festou, M.C.,Keller,H.U.,andWeaver,H.A.(Eds.),CometsII,UniversityofArizonaPress,Tucson,pp.193-204.Elliot,J.L.,and10colleagues2005.TheDeepEclipticSurvey:ASearchforKuiperBeltObjects andCentaurs.II.DynamicalClassification,theKuiperBeltPlane,andtheCorePopulation.AJ,129,1117-1162.Gilliland,R.andHartig,G.2003.StabilityandAccuracyoftheHRCandWFCShutters.Instru-mentScienceReportACS2003-03,STScI,Baltimore.Holman,M.J.1997.APossibleLong-LivedBeltofObjectsBetweenUranusandNeptune.Nature, 387,785-788.Horner,J.,Evans,N.W.,&Bailey,M.E.2004a.SimulationsofthepopulationofCentaurs-I. Thebulkstatistics.MNRAS,354,798-810.Horner,J.,Evans,N.W.,&Bailey,M.E.2004b.SimulationsofthepopulationofCentaurs-II. Individualobjects.MNRAS,355,32-329.Jewitt,D.,andLuu,J.1993.DiscoveryofthecandidateKuiperbeltobject1992QB1.Nature, 362,730-732.Kowal,C.T.1977.Slow-MovingObjectKowal.IAUCirc.,3129. Kowal,C.T.,Liller,W.,andMarsden,B.G.1979.Thediscoveryandorbitof2060Chiron.IAU Symp.81:DynamicsoftheSolarSystem,81,pp.245-250.Lamy,P.L.,Toth,I.,Fernandez,Y.R.,&Weaver,H.A.2004.Thesizes,shapes,albedos,and colorsofcometarynuclei.In:M.C.Festou,H.U.Keller,andH.A.Weaver(Eds.)CometsII.UniversityofArizonaPress,Tucson,pp.223-264. –14– Levison,H.F.1996.CometTaxonomy.In:Rettig,T.W.andHahn,J.M.(Eds.)Completingthe InventoryoftheSolarSystem,ASPConf.Series,107,173-191.Levison,H.F.,Terrell,D.,Wiegert,P.A.,Dones,L.,Duncan,M.J.2006a.OntheOriginofComet 2P/Encke.Icarus182,161-168.Levison,H.F.,Duncan,M.J.,Dones,L.,&Gladman,B.2006b.TheScatteredDiskasaSource ofHalley-TypeComets.Icarus,inpress.Liller,W.,Chaisson,L.J.,andMarsden,B.G.1977.1977UB.IAUCirc.,3151. Marchis,F.,Boehnhardt,H.,Hainaut,O.R.,andLeMignant,D.1999.AdaptiveOpticsObser-vationsoftheInnermostComaofC/1995O1.AreTherea“Hale”anda“Bopp”inCometHale-Bopp?A&Ap,349,985-995.McNaught,R.H.,and16colleagues2002.2002CR46.MinorPlanetElectronicCirculars2002-C84.Millis,R.L.,Buie,M.W.,Wasserman,L.H.,Elliot,J.L.,Kern,S.D.,andWagner,R.M.2002. TheDeepEclipticSurvey.AJ,123,2083-2109.Noll,K.2006.SolarSystemBinaries.IAUSymp.229;Asteroids,Comets,Meteors2005,inpress; astro-ph0603122.Noll,K.S.,Grundy,W.M.,Levison,H.F.,&Stephens,D.C.2006.(60458)2000CM –15– Spahr,T.B.,Berlind,P.,andMcNaught,R.H.2002.CometsC/2002A1(LINEAR)andC/2002 A2(LINEAR).IAUCirc.,7788.Stansberry,J.A.,Cruikshank,D.P.,Grundy,W.G.,Margot,J.L.,Emery,J.P.,Fernandez, Y.R.,andRieke,G.H.2005.Albedos,Diameters(andaDensity)ofKuiperBeltandCentaurObjects.AAS/DivisionforPlanetarySciencesMeetingAbstracts,37,52.05.Stephens,D.C.,andNoll,K.S.2006.DetectionofSixTrans-NeptunianBinarieswithNICMOS: AHighFractionofBinariesintheColdClassicalDisk.AJ,131,1142-1148.Stoss,R.,andMarsden,B.G.2002.2002CR46.MinorPlanetElectronicCirculars,2002-H53.Tegler,S.C.,andRomanishin,W.2000.ExtremelyredKuiper-beltobjectsinnear-circularorbits beyond40AU.Nature,407,979-981.Tegler,S.C.,Romanishin,W.,andConsolmagno,S.J.2003.ColorPatternsintheKuiperBelt: APossiblePrimordialOrigin.ApJ,599,L49-L52.Tiscareno,M.S.,andMalhotra,R.2003.TheDynamicsofKnownCentaurs.AJ,126,3122-3131.Weaver,H.A.,andLamy,P.L.1999.EstimatingtheSizeofHale-Bopp’sNucleus.EarthMoon andPlanets,79,17-33.Weaver,H.A.,Stern,S.A.Mutchler,M.J.Steffl,A.J.Buie,M.W.Merline,W.J.Spencer,J.R. Young,E.F.Young,L.A.2006.DiscoveryofTwoNewSatellitesofPluto.Nature,439,943-945. –16– Fig.1.—Combinedimageof(42355)2002CR46obtainedwiththeACS/HRCisshown.Fourseparate300secintegrationshavebeencombinedasdescribedinthetext.Thesecondaryisclearlyresolvedtothelowerleftoftheprimary.Thepixelsinthisimageare25milliarcseconaside;onlyasmallportionofthefullimageisshownfordetail.Thesecondarylies0.109±0.002arcsecfromtheprimaryatapositionangleof226.8±0.8degreesEastofNorth.North,inthisimage,liesapproximatelytotheright. –17– 40200-20-40-60-40-200Plan view (distances in AU) 2040Fig.2.—Aplanviewoftheorbitof(42355)2002CR46comparedtotheorbitsofthegas-giantplanets.Dotsshowthelocationofeachobjecton1March2006.Theorbitof(42355)2002CR46crossesboththeorbitsofNeptuneandUranus.Itisnotinanidentifiedresonancewitheitherplanet,willhavecloseencounterswithboth,andwilleventuallybescatteredoutofthisregionofthesolarsystem. –18– 80 60Distance (AU)40 20 0 0246 810Time (Myr) 12141618 Fig.3.—Thesemimajoraxis(solid),periheliondistance(lowerdotted),andapheliondistance(upperdotted)ofa(42355)2002CR46-liketestobjectasitevolvesover18Myrthroughrepeatedencounterswithgiantplanetsisshown. –19– 1 Probability of an Encounter Closer than D.1 tuSarnnusarUJuepitr.01.001 .01.1Closest Distance, D (AU) Neptune1 Fig.4.—Thecumulativeprobabilityofencountersatagivendistancewitheachofthegiantplanets.Thearrowsalongthex-axisshowtheapproachdistance,D,forwhichbinarydisruptionbecomesprobableinasingleencounter.Asshownbythecurves,JupiterandUranusarethemostlikelytodisruptaCR46-likeobject. –20– Table1:CentaursObservedwithHST 1Uncertaintyinfinaldigitshowninparentheses.2Detectionlimitforfaintcompanions=2.5log(p/3σ)wherep=peakpixelinthesourceandσisthermsbackgroundvariationinthecombined(multidrizzled)image.For(42355)2002CR46thepeakpixelreferstotheprimary.3Tegleretal.2003;4TeglerandRomanishin2000;5MinorPlanetCenter 因篇幅问题不能全部显示，请点此查看更多更全内容