OBSERVATIONSOFPHYSICALPROCESSESINCLUSTERCORES:CONNECTION BETWEENTHEINTRACLUSTERGASANDTHEBRIGHTESTCLUSTERGALAXY By Aaron ADISSERTATION Submittedto MichiganStateUniversity inpartialentoftherequirements forthedegreeof AstrophysicsandAstronomy DoctorofPhilosophy 2015 ABSTRACT OBSERVATIONSOFPHYSICALPROCESSESINCLUSTERCORES: CONNECTIONBETWEENTHEINTRACLUSTERGASANDTHE BRIGHTESTCLUSTERGALAXY By Aaron Thisdissertationexaminestherelationshipbetweengalaxyclustersandthebrightest clustergalaxies(BCGs)ofthoseclusters.Ithasbeenknownforawhilethatthestateofthe hotintraclustermedium(ICM)gasinthecoreofagalaxycluster,quanasthecentral entropyofthegas,canbefoundintwoparticularstates.Galaxyclusterswithcentralen- tropiesgreaterthan30keVcm 2 aretypicallydisturbedclusterswithnoradioactivityorline emissionintheirBCG.Ontheotherhand,thoseclusterswithlowcentralentropycanhost BCGswhichareconsidered\active"andcontainstrongcentralradiosourcesaswellasline emissionsuggestingstarformation.Whilethereisthisdichotomy,therelativeimportance ofphysicalprocesseswhichmayhelptocreatethisdichotomy,isnotwelldetermined. InChapter2,weexaminetheultravioletandinfraredpropertiesofBCGsinahetero- geneoussampleofclusters.Wethatthedichotomystillholdswheninvestigatingstar formationinboththeobscuredandunobscuredregimes.Intheselowentropyclusters ˘ 40% haveaBCGwithsomeformofstarformationsuchthatstarformationinBCGsisenabled bythedenseX-rayemittingICMgas.Theresultsweareconsistentwithotherstar formationindicators,suchasH ,butweareabletocreateamorecompletepictureofthe starformationoccurringintheBCG. InChapter3weconductanindepthinvestigationofthecoolcoregalaxyclusterRXJ 2014.8-2430.Basedon Chandra X-raydata,wetheclustercoreissloshing.However, theBCGisstilllocatedneartheX-raypeakandthemetallicityisstillcentrallypeaked, whichsuggestssloshingisarecentphenomenon.Also,wedonotX-raycavititeseven thoughtheyareexpectedinacoolcorewithradioemission.WesimulateX-rayimageswith variousbubblegurationsandsizestosetlimitsonwhatwecouldhavemissedinthe data.WeanalyzenarrowbandH imagingandopticalspectraandelongationofthe H tsalongthesameeast-westaxisofthesloshing.Theemissionlinespectrashow avelocitygradientacrossthecentralH region,suggestingthegalaxyisgettingpulledinto oroutofthecluster.TheweaksloshingaswellasthelimitonX-raycavitiessuggestswe maybeobservingRXJ2014.8-2430duringarareperiodwheresloshingandtheAGNare beginningtoheattheclustercore. InChapter4wepresentresultsfromourpolarimetrypilotstudyfortheopticalimageron theSouthernAstrophysicalResearch(SOAR)Telescope.Wediscussthemethodologyused tocollectthedataanddeterminedataqualityforappropriateanalysis.Weverifythatwe canreproducepolarizationfractionsandanglesinsourcesthatarepolarizationstandards. WeattempttomeasurepolarizationintheH tsintheBCGM87,butdonot astatisticallytmeasurementofpolarizationinthetsandplaceanupper limitontheirtotalpolarization.Thelimitonpolarizationofemissionfromthetsin M87limitstherolesaturatedthermalconductioncanplayattheinterfaceofthehotICM andthecoldts.WesummarizetheresultsofthedissertationinChapter5. Formygrandmother,LaurettaCohen. iv ACKNOWLEDGMENTS ThankstoMeganDonahueforhelpingmenavigatethedenselypackedforestwhichisthe worldofobservationalastronomyaswellassecuringfundingduringmytimeheresoIcould focusonmyresearch.Forfundingduringgraduateschool,IwouldliketothankMichigan StateUniversityDepartmentofPhysicsandAstronomyaswellastheCollegeofNatural Sciencesformyteachingassistantship,thesiscompletionfellowship,andconferencetravel grant.DuringmostofmytimehereasaresearcherIwouldliketothankthefollowingfor thefundingprovided:MarcPostmanfromtheSpaceTelescopeScienceInstitute(STScI) andCLASH, Spitzer throughJPL, Chandra ,andWilliamSparks(alsofromSTScI)forthe purchaseoftheSOIpolarizerspurchasedundertheSTScIDirectorsDiscretionaryResearch Fund. I'dalsoliketothankmycoauthorsandcollaborators:MarkVoit,Amalia(Molly)Hicks, DeborahHaarsma,WilliamSparks,onBarthelemy.I'dalsoliketothankothersIhave metatmythreetripstoAASaswellastheCIAOandSherpaattheX-raysummer schoolforprovidinghelpfuldiscussiontofurthermywork.Thanksgoesouttoallofthe SOARoperatorsDanielMaturana,PatricioUgarte,SergioPizarro,andAlbertoPastenfor theirgreatassistanceduringourobservingnightsaswellasSteveHeathcoteandSeanPoints tohelptroubleshootthebiggerissueswiththetelescope.I'dliketothankShawnaPrater andDebbieBarrattformakingsureIcrossedallofmy`i'sanddottedallofmy`t'ssoI wasn'tdis-enrolledbeforeIevengottoMSU.Andnoonecouldsurvivegradschoolwithout othergradstudentstocommiseratewith.SoI'dliketothankallofthosegraduatestudents pastandpresentinastronomyforbothsciendiscussionandcommiserationinclasses andresearchtosomeverynon-sciendiscussion. v Also,Iwouldliketothankmyparentsforprovidinganimmeasurableamountofsupport togetmetothispointinmylife.Thankstomybrothersforcontinuingtobetheirannoying selves.ThankstoEmiformakingitworkfromsofarawayandlettingmebeahouse husbandwhileIsearchforajob.Thankstoaformerstrangerwholetmeborrowhertent. ThankstoRaft,Toad,andtheSHCforbeingamazingplacestoliveandmeettrulyunique people.Thankstotherestofmyfamilyandfriendsbackhomewhomakeitfuntopick backupwhereweleft ThisresearchhasmadeuseofSAOImageDS9,developedbySmithsonianAstrophysi- calObservatory.ThisresearchhasmadeuseoftheSIMBADdatabase,operatedatCDS, Strasbourg,France.ThisresearchhasmadeuseoftheNASA/IPACExtragalacticDatabase (NED)whichisoperatedbytheJetPropulsionLaboratory,CaliforniaInstituteofTech- nology,undercontractwiththeNationalAeronauticsandSpaceAdministration.Basedon observationsobtainedattheSouthernAstrophysicalResearch(SOAR)telescope,whichisa jointprojectoftheMinisteriodaCi^encia,Tecnologia,eInova˘c~ao(MCTI)daRepublicaFed- erativadoBrasil,theU.S.NationalOpticalAstronomyObservatory(NOAO),theUniversity ofNorthCarolinaatChapelHill(UNC),andMichiganStateUniversity(MSU). SupportforthisthesiswasprovidedbyMichiganStateUniversitythroughProf.Don- ahue'sstartupfunds,fromthePhysicsandAstronomyDepartmentandCollegeofNatural Sciences.Thisworkisbasedinpartonobservationsandarchivaldataobtainedwiththe SpitzerSpaceTelescope,whichisoperatedbytheJetPropulsionLaboratory,California InstituteofTechnologyunderacontractwithNASAanddataobtainedwiththeChandra X-rayObservatoryalsoundercontractwithNASA.Supportforthisworkwasprovidedby NASAthroughanawardJPL1353923(MSURC065195)andJPL1377112(MSURC065166) issuedbyJPL/Caltech.SupportforthisworkwasalsoprovidedbytheNationalAeronau- vi ticsandSpaceAdministrationthroughChandraAwardNumberSAOGO0-11018Xissued bytheChandraX-rayObservatoryCenter,whichisoperatedbytheSmithsonianAstrophys- icalObservatoryforandonbehalfoftheNationalAeronauticsSpaceAdministrationunder contractNAS8-03060.Finally,partialsupportforthisthesiswiththeprogram#12065.07 (CLASH)wasprovidedbyNASAthroughagrantfromtheSpaceTelescopeScienceInsti- tute,whichisoperatedbytheAssociationofUniversitiesforResearchinAstronomy,Inc., underNASAcontractNAS5-26555. vii TABLEOFCONTENTS LISTOFTABLES .................................... x LISTOFFIGURES ................................... xi Chapter1Introduction ................................ 1 1.1GalaxyClusters..................................1 1.2IntraclusterMedium...............................3 1.3ClusterEntropy..................................7 1.3.1CoolingFlowProblem..........................9 1.4BrightestClusterGalaxies............................11 1.5ClusterSamples..................................15 1.5.1ArchiveofChandraClusterTables...............15 1.5.2ClusterLensingAndSupernovasurveywithHubble..........15 1.5.3RepresentativeXMM-NewtonClusterStructureSurvey........16 1.6TelescopesandInstruments...........................17 1.6.1 Chandra X-rayObservatory.......................18 1.6.2GalaxyEvolutionExplorer........................21 1.6.3SouthernAstrophysicalResearchTelescope...............21 1.6.3.1GoodmanSpectrograph....................21 1.6.3.2SOAROpticalImager.....................22 1.6.4TwoMicronAllSkySurvey.......................24 1.6.5SpitzerSpaceTelescope.........................24 1.6.5.1InfraredArrayCamera.....................24 1.6.5.2MidIRPhotometerSystem..................26 1.7OutlineofDissertation..............................26 Chapter2InfraredandUltravioletStarFormationinBrightestCluster Galaxies ................................... 29 2.1Introduction....................................30 2.2Observations....................................34 2.2.1 Chandra X-RayObservations......................34 2.2.2 2MASS Observations-BCGiden...............36 2.2.3 GALEX Observations...........................38 2.2.4 Spitzer Observations...........................38 2.3AperturePhotometryandColors........................39 2.3.1 GALEX UVPhotometry.........................39 2.3.2 GALEX UVUpperLimits........................40 2.3.3 Spitzer NearandMidIRPhotometry..................42 2.3.4 2MASS NearIRObservations......................46 2.4Discussion.....................................46 2.4.1UVExcessandColor...........................46 viii 2.4.2IRColor..................................51 2.4.3StarFormationRates(SFRs)......................59 2.4.4StarFormationandClusterEntropy..............64 2.4.5ICMGasCoolingandStarFormationinBCGs............67 2.5Conclusions....................................68 Chapter3MultiwavelengthStudyoftheExtremelyCoolCoreCluster RXJ2014.8-2430 ............................. 71 3.1Introduction....................................72 3.2ObservationsandDataReduction........................77 3.2.1ChandraX-rayObservation.......................77 3.2.2SOARH ImagingandSpectra.....................82 3.2.3X-rayAGNLimits............................91 3.3Discussion.....................................92 3.3.1RadioBubbleLimits...........................92 3.3.2X-rayCavityToyModel.........................94 3.3.3SloshingintheClusterCore.......................98 3.3.4VelocitystructureintheBCGOpticalEmissionLines.........99 3.4Summary.....................................103 Chapter4PolarizationPilotProjectfortheSOARTelescope ........ 106 4.1Introduction....................................106 4.1.1SourcesofAstrophysicalPolarization..................107 4.1.2ObservingAstrophysicalPolarization..................108 4.1.3CoolCoreClustersandPolarization...................109 4.2Observations....................................113 4.3CalibrationandDataAnalysis..........................116 4.3.1StokesParameters............................116 4.3.2ImageReduction.............................119 4.3.3DomeFlats................................120 4.3.4UnpolarizedCalibrationTargets.....................121 4.3.5PolarizedCalibrationTargets......................125 4.4PolarizationLimitsofM87Filaments......................132 4.5ImplicationsfortheFilamentsinM87......................139 4.5.1ThermalConduction...........................140 4.6Conclusions....................................145 Chapter5Summary .................................. 147 APPENDIX .................................. 150 BIBLIOGRAPHY .................................... 157 ix LISTOFTABLES Table2.1.SummaryofObservationsandDetections...............35 Table3.1.Observations...............................76 Table3.2.H regionX-rayBoxes.........................81 Table3.3.GoodmanSpectralLineFits.......................88 Table4.1.PolarizationObservations........................115 Table4.2.ExtendedPolarizationStandards....................130 TableA.1.BrightestClusterGalaxyIden...............151 TableA.2.PhysicalProperties...........................152 TableA.3.FluxesMatchedtoUVAperture.....................153 TableA.4. Spitzer ApertureFlux..........................154 TableA.5. 2MASS ApertureFlux..........................155 TableA.6.StarFormationRates:..........................156 x LISTOFFIGURES Figure1.1ModelX-raySpectra...........................5 Figure1.2ClusterEntropy..............................10 Figure1.3BCGSpectraEnergyDistribution....................13 Figure1.4 Chandra ACISChipPlane........................19 Figure1.5 Chandra eArea..........................20 Figure1.6GALEXBandpass............................22 Figure1.7SOIPolarimeterSetup..........................23 Figure1.8IRACBandpass..............................25 Figure2.1BCGCentroidDistance.........................37 Figure2.2UVMagnitudeUpperLimits.......................41 Figure2.3KbandLuminosity............................47 Figure2.4NUV-KColor...............................49 Figure2.5FUV-NUVNUV-KColor........................50 Figure2.68.0-3.6InfraredRatio...........................53 Figure2.74.5-3.6InfraredRatio...........................54 Figure2.8InfraredRatioCorrelation........................55 Figure2.9Mid-IRColor...............................57 Figure2.10SINGSGalaxyComparison.......................58 Figure2.11UVandIRSFR..............................60 Figure2.12ComparisonofmodelIRSFRtosingleband70 mIRSFR......61 Figure2.13ComparisonofmodelIRSFRtosingleband24 mIRSFR......62 xi Figure2.14IRexcessandUVcolor..........................63 Figure2.15Relationbetween70micronSFRandcentralentropy.........66 Figure3.1X-rayImageofCluster..........................78 Figure3.2ACCEPTstyle.........................83 Figure3.3JACOMetallicity.............................84 Figure3.4ContinuumsubtractedH imageoftheBCGofthecluster......86 Figure3.5X-rayCavityToyModel.........................95 Figure3.6ToyModelFits..............................97 Figure3.7GoodmanSpectraVelocities.......................100 Figure3.8GoodmanSpectraVelocities.......................102 Figure4.1PolarimeterDesigns............................110 Figure4.2SOIPolarimetrySetup..........................111 Figure4.3StokesParameters............................119 Figure4.4SingleNightDomeFlatComparison..................122 Figure4.5MultipleNightDomeFlatComparison.................123 Figure4.6GlobularClusterPolarziationComparison...............126 Figure4.7CrabNebulaPolarizationImages....................128 Figure4.8CrabNebulaPolarizationVectors....................131 Figure4.9HicksonandvandenBergh[1990]CrabNebulaMagneticFieldVectors.133 Figure4.10RMonocerotisPolarizationImage....................134 Figure4.11M87AGNPolarizationComparison...................135 Figure4.12M87H PolarizationFields.......................138 Figure4.13M87ContinuumStokesParameters...................141 xii Figure4.14M87H StokesParameters.......................142 Figure4.15M87H RegionPolarization.......................143 xiii Chapter1 Introduction 1.1GalaxyClusters Galaxyclusters,whichcontainhundredstothousandsofgalaxies,arethelargestgravita- tionallyboundstructuresintheUniverse.Astheirnameimplies,galaxyclusterswere discoveredfromthespatialclusteringofgalaxies[Zwicky,1937,1938].Massiveellipticaland S0galaxiesarethedominantgalaxytypesseeninclusters(Zwickyetal.[1961];Bautzand Abell[1973]).Galaxyclustersaretfromsparsegalaxygroups,suchasourLocal Group,whichtypicallyhavetensofgalaxies.Also,galaxygroupstypicallyhostlowermass spiralsandirregulartypegalaxiesastheirprimarygalaxytypes.Clustersformhierarchi- cally[WhiteandFrenk,1991]suchthatthegravitationalattractionofnearbysubclusters builduptoformtheselargeself-gravitatingsystems.Whilegalaxyclustersaredeveloping, individualgalaxiesarealsoforming.Duringthistimetherearemanymergersoftheseproto- galaxiesandsubclusters,whichheatthegasandpreventsomeofthegasfromassociating withindividualgalaxies.Thesemergereventscausesomeofthegravitationalpotentialen- ergyintheclustertoconvertintokineticenergy,heatingupthegasparticles.However, earlyobservationsbeliethetruenatureofgalaxyclusters.Themassofstarsinindividual clustergalaxiesprovidesonlyasmallfractionofthetotalmassinacluster.Galaxyclusters, whicharetypically10 14 10 15 solarmasses(asolarmassisthemassoftheSun,1.99 10 30 1 kg),aremostlydarkmatter( ˘ 85%).Perhapsmoreintriguing,however,isthenextlargest masscontributionisthelowdensity(10 3 cm 3 e.g.[Davidetal.,1990b])gasthat permeatestheentirecluster.Itis,byfar( ˘ 80%),thedominantformofbaryonicmass.This gasisknownastheintraclustermedium(ICM)anditisverybright,withatypical luminosityof10 43 10 46 ergs 1 (10 9 10 12 solarluminosities).Fromthethermalmotions inthecluster,thetypicalICMvirialtemperatureisT virial ˘ 10 7 -10 9 K( kT virial ˘ 1-10 keV).Withthishighvirialtemperature,itwasn'tdiscovereduntilthe20thcenturybecause X-rayprobeslocatedabovetheEarth'satmospherearerequiredtoobservethehotgasof theICM.In1971,the UHURU X-raysatelliteedthereisX-rayemissioncomingfrom galaxyclusters[Gurskyetal.,1971].TheICMisopticallythinatX-raywavelengthssuch thathighenergyphotonemissionfromtheICMcanstreamfreelyfromthecluster. Thesemassive,gravitationallyboundboxeswhichcontaindarkmatter,galaxies,and multiphasegasareextremelyinterestingastronomicalsources.Inthisdissertation,weuse multi-wavelength(fromX-raystothemid-IR)observationstostudytphenomena presentingalaxyclusters.Eachwavelengthregimepresentsacomplementarysetofobser- vationsusedtobetterunderstandthephysicalprocessesinclusters.Inparticular,wefocus ontheinteractionbetweentheICMnearthecenteroftheclusterandthecentralbrightest galaxy,knownasthebrightestclustergalaxy(BCG),inthecluster.Theunderlyingphysical processeswhichfuelinteractionsbetweentheICMandtheBCGhavebeenastronglycon- tentioustopicwithavarietyofproposedtheoriesandmodels.Multi-wavelength observationsthatcoverawidevarietyofclustertypescanhelpconstraintheoreticalpredic- tionsand,hopefully,beusedtoresolvethesedebates. Inthenextsection,wereviewmoredetailsabouttheintraclustermedium.Then,we reviewtheconceptofentropyasitrelatestoclusters.Afteranintroductiontoentropy,we 2 introducethecoolingwproblemandtheroleofBCGsinclusters.Thenwediscussthe clustersampleswhereourdataarefromaswellasthetelescopesusedtocollectthedata. Finally,wesummarizeeachofthesubsequentchaptersofthethesis. 1.2IntraclusterMedium TheintraclustermediumisamixtureofHandHewithatypicalabundanceofheavier elements(abundanceishereastheratioofnumberofparticlesofelementXtothe numberofHydrogennuclei)thatisabout30%thatoftheSun'sheavyelementabundance, andthemassinheavyelementscomparedtothemassinhydrogenistermed\metallicity". ThesourceoftheICMislikelyfromintergalacticgaswhichwaspartoftheclusterduring itscreationbutisn'tgravitationallyboundbyindividualgalaxiessincethemassoftheICM isabout7timesthatofthestarsingalaxies.Despiteitsrelativelyhighheavyelement abundance,verylittleoftheICMcouldhavebeenprocessedbyastar,andisthereforethe sourceoftheHandHeisprimordial,pristine,intergalacticgas.Withvirialtemperatures, kT virial =1 15keV,thetypicalthermalmotionsofthemostlyionizedHgasare500 1500 kms 1 ,nearlythesamevelocitiesasthegalaxiesmovinginthegravitationalpotentialof thecluster.Thereforethegasisbythegravitationalpotentialandthetemperature ofthegasislargelydictatedbythedepthofthepotentialwell.SincetheICMisahighly ionizedplasma,theemissionisprimarilyfree-free(i.e.thermalbremsstrahlung)whichtakes theform: e ff =1 : 4 10 27 T 1 = 2 n e Z 2 g B ergs 1 cm 3 : (1.1) Thisspectralshapehasanexponentialcutathightemperaturesatthefrequency c = kT=h [RybickiandLightman,1979]. 3 Inadditiontothisthermalcontinuumemission,thereislineemissionshowninFigure1.1 fromSarazinandBahcall[1977].Therelativeline-to-continuumemissionstrengthis bybothclustertemperatureaswellasmetalabundance.Inparticular,thestrongestline emissioninclustersisfromtheFeK(7keV)linecomplex,whichisablendofKlines (dominatedbyFe +24 andFe +25 bound-freeemission)[PetersonandFabian,2006].These linestrengthsareusedtomeasurethemetallicity(typicallydescribedastheFe/Hcontent relativetothesolarabundanceofFe/H)oftheclusterwhichistypicallyathird[Bahcall andSarazin,1977]ofthesolarvalue[GrevesseandSauval,1998].Sincetheinitialelemental distributionintheUniverseissettobeabout75%hydrogen,25%heliumandtraceamounts oflithium,beryllium,andboron,roughly5minutesaftertheBigBang,carbonandheavier elementsareformedasaresultofmassivestarswhichhaveexplodedasTypeIIcorecollapse supernovaearlyinthehistoryofthecluster,aswellasTypeIasupernova(Arnaudetal. [1992],MushotzkyandLoewenstein[1997]).Much,ifnotall,ofthesenewlyformedelements, eveniftheymanagetoescapetheirparentgalaxies,areretainedinthedeeppotentialwell oftheclusters. TheresolutionofcurrentX-raytelescopescan'tdistinguishbetweenthelineswithinthe blendsofFeKandFeLshells.Thefuturegenerationofhighlyenergyresolvedtelescopes suchasAstro-H[Takahashietal.,2010]willberequiredtodisentangletheselinesandgive moreprecisevelocitiesandabundancesofthesegases. TheX-raysurfacebrightnessemission(photonsperunitarea)nearthecenterofarelaxed, isolatedclustercanusuallybewelldescribedbya modelasafunctionoftheprojected 4 Figure1.1ModelX-raySpectra .Plotsofnumericalmodelswhichassumeanisothermalgas arereproducedfromSarazinandBahcall[1977].Theupperpanelsareforhottertemperature clustersandfeatureonlythehighestionizedlinesontopofthethermalbremsstrahlung continuumspectrum.Themodelsallassumespheresofgasinsidearadiusof0.5Mpcand withaprotonnumberdensityof0.001cm 3 . 5 radius b [Sarazin,1988]: I X ( b ) / 1+ b r c 2 ! (3 1) = 2 : (1.2) Insomecases,galaxyclusterswithdenseX-raycoresrequireasecond modeltothecore ofthecluster.UsingtheX-raysurfacebrightnessasanemissionmeasure,(EM R n p n e dl ) wecananalyticallydeprojecttheobservedsurfacebrightnessdistributiontorecoverthetrue X-raygasdistributionasafunctionofthephysicalradius r : ˆ g ( r ) / 1+ r r c 2 ! 3 = 2 : (1.3) Theuseofa modelisnotderivedfromphysicalprinciples,itisausefulformtodescribe thesurfacebrightnessinasmuchasitaccuratelyrepresentstheshapeandcanbeused asareferencemodeltodetectasymmetriesandothersubstructure.Butbecauseitisnot aphysicalmodel,usingittoextrapolatebeyondthedetectedsurfacebrightnesscanlead toerrors.Forourwork,wetheanalysistothedetectedregimes(e.g.Leaetal. [1973],KelloggandMurray[1974]). PhotonenergiesarealsorecordedinanX-raydataset.Usingtheenergyinformation,X- rayspectracanbecreatedbycollectingphotoneventsfromspatialregionsandbinningthem asafunctionofphotonenergy.Fromthecomparisonofthatspectrumwithplasmaemission models,thetemperatureoftheICMgascanbeestimatedfromthethermalbremsstrahlung frequency,suchasthoseshowninFigure1.2,tothespectrum.Withsomeadditional assumptions,thedensitiesandtemperaturesmeasuredcanbeusedtocalculatethepressure andentropyofthesystem,whichleadtoimportantinsightsaboutthestateoftheICMand 6 thethermodynamichistoryofthecluster. 1.3ClusterEntropy Entropyisanimportantandfundamentalthermodynamicquantity.Onamacroscopicscale, itrepresentstheamountofenergyavailableinthesystemrelatedtoheattransferintheform dS = dQ=T .OurtwomainICMobservables,temperatureanddensity,arecombinedtomake anestimateofthegasentropy.Temperatureissensitivetothedepthofthegravitational potential,whilethedensityoftheICMissetbytheentropydistribution.Forexample,if thegasentropyislow,theICMcangetquitedense;ifitishigh(sayithasbeenhighly shockedbeforeitfallsintothecluster)itwillremainatalowdensity,itcan'tbecompressed veryeasily. Wecancomparetheadiabaticequationofstateforanidealmonatomicgas, P = Kˆ where K istheadiabaticconstant, =5 = 3foramonatomicgastotheidealgaspressure, P = ˆkT H where isthemeanmolecularweightofthegasand m H isthemassof hydrogen.Fromherewecansolvefortheadiabaticconstant K tobe K = kT H ˆ 2 = 3 : (1.4) IntermsofourX-raymeasurables,electrondensity, n e ,andX-raytemperature, T X (mea- suredinkeV),theadiabaticconstantbecomes: K = T X n 2 = 3 e : (1.5) BysimplymeasuringthesurfacebrightnessandingthespectraoftheX-rayimageof 7 thecluster,wecansolvefortheadiabaticconstantofthecluster.Theadiabaticconstant K isrelatedtothespentropy s = k lnK 3 = 2 +s 0 ,whichwecanusetocomparethe relativethermodynamicstatesbetweenclusterswithsimpleX-rayobservables.Becauseof thisrelationship,inthisdissertationweusetheterms\entropy"and\clusterentropy"when referringtotheadiabaticconstant K .Weexpectthatlowentropygaswillsinkintothe deepgravitationalpotentialwellatthecenteroftheclusterandanyhighentropygasthat iscreatedneartheclustercorewillbuoyantlyrisetotheoutskirtsofthecluster. Asyouapproachtheclustercenter,theentropytoanon-zerominimum entropyandtheminimumentropythatisobtainedatthecentervariesdrasticallyacross tgalaxyclusters.Cavagnoloetal.[2009]calculatethecentralentropybythe radialentropytothefunctionalform K ( r )= K 0 + K 100 ( r 100kpc ) (1.6) where K 0 istheentropyatthecenter,K 100 istheentropyat100kpc,andristhedistance fromthecenteroftheclusterinkpc.Figure1.2,fromCavagnoloetal.[2009],includesa heterogeneoussampleofgalaxiesentropyIntheapurecoolingmodelfrom Voitetal.[2002]isplottedasablacklineforcomparison.Atlargeradii,nearlyalltheclusters areconsistentwiththiswhichisintheterm K 100 ( r 100 kpc ) .Therefore,while thecentersofclustershavesomemechanismthatpreventstheclustersfromcatastrophic cooling,theoutskirtsoftheclustersarewellrepresentedbypurecooling.Wecanestimate acentralcoolingtime,whichisanestimateofthetimeittakesforalltheenergyinthecore gastodissipate.Thecoolingtimefunctionisintheform t cool =5 nkT X = 2 n e N H T;Z ) where T;Z )isamodeledcoolingrateasafunctionoftemperature T andmetallicity Z . 8 Typicallythecoolingtimeattheclusteroutskirtsinthiswayislongerthantheage oftheUniverse.Thegashasshortercoolertimesneartheclustercenter.Insomeclusters, thecentralcoolingtimecanbemuchshorter( < 1Gyr)thantheageoftheUniverse.The histograminFigure1.2fromCavagnoloetal.[2009],showsatfractionofthese clustershaveashortcentralcoolingtime.Thiswasinitiallyperplexingbecauseitwas thoughtthatclustersshouldbecatastrophicallycoolingduetoalowcentralentropy,which wouldcausethegastoforma\coolingw"leadingtoanextremelyluminousclustercenter [CowieandBinney,1977,FabianandNulsen,1977,MathewsandBregman,1978]. 1.3.1CoolingFlowProblem Toquantifyhow\bad"thecoolingwproblemis,anestimateoftheamountofgasrequired tobecoolingisapproximatedas dM dt = 2 5 kT ; (1.7) whereMisthemassofcoolinggas, L isthecoolingluminosity, misthemeanmolecular mass,and T isthetemperatureinsidethecoolingradius,assumingtheX-raygasiscooling fromthevirialtemperatureataconstantpressure.Fromthiscalculation,someclustershave coolingwsapproaching1000solarmassesperyear[Edgeetal.,1992].Thiswasinitially aproblembecauseiftherewassomuchlowentropygascollectinginthecenteritshould berapidlycondensingandformingstarsatratemorerapidlythanthemostvigorousstar- forminggalaxies.Inaddition,therewerepredictionsthatthereshouldbestrongemission linesinthesoftX-ray( < 1keV)toproduceenoughcoolingtolowertemperatures,butthese emissionlinesweren'tseen[Petersonetal.,2003].Nowtheconsensusistheremustbesome physicalprocess(es)whichheatuptheclustercenterandpreventitfromcatastrophically 9 Figure1.2ClusterEntropy .TheupperisreproducedfromCavagnoloetal.[2009]. TheradialentropyarecolorcodedbytheaverageclustertemperatureinkeV.The blacklineisapurecoolingmodel.ThelowerisalsoreproducedfromCavagnolo etal.[2009].Thehistogramhasaconstantbinninginlogspace.Thecumulativehistogram, thoughlessobviousbyeye,alsoshowsthebimodalityseenintheupperhistogram. 10 cooling.Unliketheirhighcentralentropycounterparts,coolcoreclustersaretypicallyvery symmetric.Inparticular,thebrightestgalaxyoftheclustersitsnearthecenterwherethis coolingwwouldterminate[Dubinski,1998].Therefore,manyoftheproposedsolutionsof heatingofthecoreoftheintraclustermediumrelatetopropertiesofthisbrightgalaxy.The typesofusuallyconsideredtoheattheICMinclude:starformationandstarbursts, AGNheating[Burns,1990,BinneyandTabor,1995,Churazovetal.,2001],andconduction [TuckerandRosner,1983,BertschingerandMeiksin,1986,BregmanandDavid,1988,Sparks etal.,1989a] 1.4BrightestClusterGalaxies Interestingly,wethatthepropertiesofbrightestclustergalaxies(BCGs),typicallyone ortwogalaxiesintheclusterthatarethebiggestandthebrightest,correlatewiththecentral entropyofthecluster[Cavagnoloetal.,2009].Inmanycases,theBCGisnearthecenterof massoftheclusterandisknownasthecentrallydominantgalaxy[Sarazin,1988].BCGsin clusterswithhighentropyintheircoresdidnotshowanyactivityindicatingyoungstarsor anactivecentralblackhole[Cavagnoloetal.,2008b].Initiallythesegalaxieswerethought tobejustthebrightestclusterellipticalgalaxies,butBCGsappearmoreextendedthan largeclusterellipticals[Hoessel,1980].Ellipticalgalaxiesaretypicallythelargestgalaxies andaredominatedbyoldstarswithoutrecentstarformation.Giventhattheirpopulation isdominatedbyoldstars,themostmassivestars(whichbeginblue)evolveofthemain sequenceandmoveontotheirredgiantphase.Lowermassstarshavenotevolvedof themainsequenceyet,butthesestarsareredincolor.Thereforeellipticalgalaxiesare considered\redanddead"andaren'tformingmanystarscomparedtogalaxieslikespirals. 11 SpiralgalaxiesliketheMilkyWayformafewsolarmassesofstarsperyear[e.g.Kennicutt, 1983]whilethemoststarforminggalaxies,knownatstarburstgalaxies[Weedman etal.,1981],canhavestarformationratesofafewhundredsolarmassesperyear.Those galaxieswithstrongstarformationhavesomeUVcontinuum(i.e.emissionfromyoungstars whichhaveectivetemperatures > 10,000Kequivalenttoapeak < 300nm)[e.g.Salim etal.,2007]andinfraredcontinuumemissionfromcolddust(50K;100 m).Additionally, thereisstronglineemissionintheopticalandinfrared.Oneofthedominantmeasurementof starformationinastronomyisH (i.e.the3to2transitionofneutralhydrogen).H isseen inemissionwhentherearestarsthathaveUVphotonswithhighenoughenergytoionize H.Therecombinationoftheionizedhydrogentoneutralhydrogengivesthismechanismas themostpreferredtransition. BCGshaveathanlargeellipticalgalaxiesofsimilarsizebecausethesmooth transitionfromtheBCGtotheintraclusterlight(Caonetal.[1993],Gonzalezetal.[2005]). ThismakesittodeterminethetotalsizeofaBCGastheyextendsmoothlyintothe lightofthecluster.Therefore,areusuallymeasuredinametricaperturefor easycomparisonsbetweendataintheliterature.Typicalmeasurementsfromtheliterature lookatmeasurementsofthecoreoftheBCGinsideof10kpc[e.g.Hoesseletal.,1980, PostmanandLauer,1995]. Weplotaspectralenergydistributionofthenearbybrightestclustergalaxy,Centaurus, fromradiotogammaraysinFigure1.3[EbneterandBalick,1983]todemonstratethat, whilemostoftheemissioninabrightestclustergalaxyisinthenear-infraredandcomes fromtheoldstars,therearesomeBCGsthathaveatamountofemissioncoming fromthecolddustatlongermid-infraredwavelengths.Therefore,thereareBCGswhichcan bemodeledasanoldstellarpopulation(approximatelya6,000Kblackbodywithstellar 12 Figure1.3BCGSpectraEnergyDistribution .Thespectralenergydistribution(SED)is plottedforthenearbybrightestclustergalaxyCentaurusA(NGC5128)from http: //www.mpe.mpg.de/ ~ hcs/Cen-A/cen-a-facts.html .Therearentcontributions totheemissionintheX-ray,optical,andinfrared.SourceisacompositeSEDfromthe NASA/IPACExtragalacticDatabase 1 . atmospherephysics)plusainfraredstarburstgalaxy.Thisemissionsuggeststhat brightestclustergalaxiesaresomecombinationofamassiveellipticalgalaxywithongoing starformationfromadditionalcoldgas. Asidefromthethermalemissionofthedustintheinfrared,thereisadditionalemissionin thenear-infraredwhichistypicallyassociatedwithpolycyclicaromatichydrocarbons(PAH) andsilicatemolecules(DraineandLi[2007],Tielens[2008]).Thesefeaturesresultfromthe re-radiationofUVphotonsemittedbyyounghighmassstars. Mostgalaxieshaveacentralsupermassiveblackholeafewhundredmilliontoafew 1 http://ned.ipac.caltech.edu/ 13 billiontimesthemassofthesun(e.g.Magorrianetal.[1998]).Someoftheseblackholesare activelyaccretingmaterial.Ifablackholeisaccretingthematerialatahighenoughrate,it maysendsomeofthatmaterialbackoutintheformofacollimatedjet.Supermassiveblack holesthatejectthismaterialareknownasactivegalacticnuclei(AGN)aswellasquasars (QSOs)becausetheyweresobrightandsofaraway. Forthosegalaxyclustersthathavelowcentralentropieswenoticethatnotallofthe brightestclustergalaxiesassociatedwiththemareredanddeadellipticals.Whileallhave largeevolvedstarpopulations,similartoellipticalgalaxies,therearesome( ˘ 40%)thatalso havesomeformofnebularemission.Heckmanetal.[1989],yetal.[2008],Cavagnolo etal.[2008b]allidenbrightestclustergalaxieswithnebularemission.Theyareknown tobeassociatedwithcoolcoreclusters,whichhavecentralentropiesbelow30keVcm 2 , however,thereasonforthisbimodalstructureisunclear. WiththeAGNinthecenter,thislargeiofgaswouldcausetheAGNtolightup anditwouldincreasetheentropyofthegasandblowoutbubblestoregulatethesystem. However,giventhesmallsizeofAGN( ˘ 1pcandthesizeoftheclustercore(10sofkpc)it ishardtoreconcilethescales.Insomenearbygalaxies,closeenoughtoseethetary structure,someofthesoftestX-rayemissionisco-locatedwiththesents[Crawford etal.,2005].MostoftheworkcomparingsoftX-rayemissionandopticallineemissionhas beendonefornearbygalaxies,likeNGC1275,theBCGofthePerseusCluster(e.g.Ferland etal.[2009],Fabianetal.[2011]),andM87,thebrightestclustergalaxyoftheVirgoCluster (e.g.Sparksetal.[1989a],Werneretal.[2013]). 14 1.5ClusterSamples Whileclusterscanbeinterestingindividually,forclusterstobetrulyusedasacosmological testtheymustbeexaminedenmasse.Whiletherearemanyclustercatalogsandsamples thathavebeenconstructedtostudytaspectsofclusterphysicsandcosmology,we focusourstudyonthreeparticularsamples:ACCEPT,CLASH,andREXCESS.We reviewthepurposeofeachofthesesamplesandhowtheyarerelatedtothestudiesinthis dissertation. 1.5.1ArchiveofChandraClusterTables TheArchiveofChandraClusterTables(ACCEPT)examinedpropertiesofgalaxy clusterswhichhadbeenwellobservedintheChandraarchivepriortothedateofpublication [Cavagnoloetal.,2009].Therewasatotalof239galaxyclustersavailableforanalysis,which wereheterogenousintheirmorphology.However,thesamplewaslargeenoughtoobservea widevarietyofgalaxyclustersandtheirassociatedentropypWeusethissamesample tolookfortheassociatedbrightestclustergalaxiesinthe Spitzer and GALEX archivesto compareIRandUVstarformation,respectively,totheX-raypropertiesfoundinACCEPT. 1.5.2ClusterLensingAndSupernovasurveywithHubble TheClusterLensingAndSupernovasurveywithHubble(CLASH)isalarge(524orbit) HubbleSpaceTelescope(HST)Multi-CycleTreasuryProgram(i.e.thedatabecomespublic immediatelyafterithasbeentakenonthetelescopeinsteadofthenormaloneyearpro- prietaryperiod)tostudy25galaxyclustersPostmanetal.[2012].Ofthe25clusters,20 areX-rayselectedclusters,mostofwhicharerelaxedclusters.Thesewerechosentoget 15 accuratemeasurementsoftheclustergasmassandhelpmeasuretherelativeconcentration ofbaryonicanddarkmatterintheclustercores.Theeremainingclustersareselectedfor theirstronggravitationallensingcharacteristics.Thesewerechosentooptimizethelikeli- hoodofdiscoveringhighredshiftgalaxies(z > 7)byusinggravitationallensingtomagnify thesedistantgalaxies. Alloftheclustershavebeenobservedby Chandra andallbutone(MACSJ0416.1-2403) hadbeeninthe Chandra archivepriortopublicationoftheACCEPTsample.Newercali- brationsfor Chandra aswellasnewdatasetsforsomeoftheclustershavecomeoutsince theACCEPTanalysis,makingthisaninterestingscienpursuit.Forthisproject,Iwas responsibleforthereductionandanalysisofthe Chandra observationswhichwereusedto determinetheX-rayattributes,includingtheX-raygasmass,previouslyfoundinACCEPT. ThereductionandanalysistoolswerealsousedinChapter3fortheanalysisofRXJ2014.8- 2430(RXJ2014.8-2430)andtheyarediscussedinthatsection.Thisworkhasbeenincluded inaCLASHX-rayoverviewpaper[Donahueetal.,2014]. 1.5.3RepresentativeXMM-NewtonClusterStructureSurvey TheRepresentativeXMM-NewtonClusterStructureSurvey(REXCESS)wasaXMM- NewtonLargeProgrammetoinvestigateasmallsampleof33nearby(0.055 < z < 0.183) galaxyclustersohringeretal.,2007a].Thegoalofthesurveywastosampleclustersin thelocalUniverseoverawiderangeofX-rayluminosity(aproxyforclustermass)andinde- pendentonclustermorphology.Fromhere,itwouldbepossibletodostatisticsonwhatthe galaxyclusterdistributionshouldlooklikeforthelocalUniverse.TheclusterRXJ2014.8- 2430waspartofthissampleandfoundtobethestrongestcoolcoreclusterofthegroup. Unfortunately,theresolutionofXMMisttoexaminekiloparsec-scalestructure 16 intheclustercoretodeterminewhatsourcesofheatingwerepresentinthecoolcore.Dr. DonahuewasthePIfora Chandra projecttoexaminetheclustercoreatamuchhigher resolutionthanwouldbeavailabletoXMMandlookforstructurenearthecore.These resultsarepresentedinourChapter3analysis. 1.6TelescopesandInstruments Thisdissertationiscomposedofmultipleobservationalstudiesofgalaxiesandgalaxyclus- ters.Tobestaccomplishthis,avarietyoftelescopesandinstrumentswithbothimagingand spectroscopywereusedoversixordersofmagnitudeinwavelengthfromtheX-ray( ˘ 1 A) tothemid-infrared( ˘ 160 m).Whilesomedatawereexplicitlytakenforthisthesis,much ofitwastakenfrompubliclyavailablearchivesfundedandadministeredbyNASA.Someof thesedataarefromsurveyswhichsurveyalargefractionoreventhewholesky,otherdata arefrompointedobservationsfrompreviousstudiesrequestedbyindividualastronomersfor spprojects.Intherestofthissection,Iwillovervieweachofthetelescopesand instrumentsweused(oruseddatafrom)aswellasashortdescriptionofeachoftheinstru- mentstohelpmotivatethereasonsthesetelescopeswerechosentoaddressthequestions presentedinthisthesis. Manyofthetelescopesusedinthisthesisarespace-basedtelescopes.Therearetwo mainreasonsforplacingtelescopesinspace,bothofwhicharerelatedtoofthe atmosphere.Turbulenceintheatmospherecreatesdistortionintheimages,becauseastar, whichappearspoint-like,willappeartodancearoundasthelightisrefractedthroughthe atmosphere.Forhighqualitysites,thetypical\seeing"(i.e.FWHMofapointsource)is ˘ 0 : 5 00 butatmosphericturbulencedegradestheseeingtoover1 00 onnights.Additionally, 17 overmanywavelengthregimes,theatmosphereispartiallyorcompletelyopaque. Furtherdiscussionandspofthetelescopesandinstrumentsaregivenineach ofthechapterswherethosetelescopesandinstrumentsareusedfordatacollection.We reduceandanalyzedatafromGALEX,Spitzer,and2MASSinChapter2.InChapter3we usedatafrom Chandra andtheimagerandspectrographontheSOARTelescope.InChapter 4weusetheimagerfromtheSOARTelescope.Thetelescopesandtheirinstrumentsare orderedbywavelength,fromshortestobservingwavelengthtolongest. 1.6.1 Chandra X-rayObservatory The Chandra X-rayObservatory(abbreviatedas Chandra orCXO)isoneofthefourNASA GreatObservatoriesandhasbeenoperatingsinceitslaunchintonearearthorbitin1999 [Weisskopfetal.,2002].Weusedthetwoimagingdetectors,ACIS-IandACIS-S,which canbeseeninFigure1.4.TheACIS-Sdetectorisback-illuminatedcomparedtothefront- illuminatedACIS-Idetector. 18 Figure1.4 Chandra ACISChipPlane .Duringobservations,galaxyclustersarecenteredon eitherthecrossontheI3chiportheplusontheS3chipdependingontheobserver'sdesired Thegalaxyclusterswehaveobservedtypicallycoverpartofallfourofthe ACIS-IchipsorpartsofthechipsadjacenttotheS3chip.FigurecourtesyoftheChandra X-rayObservatoryCenter,whichisoperatedbytheSmithsonianAstrophysicalObservatory onbehalfofNASA 2 . Thegoalwasfortheback-illuminateddetectorstobemoresensitivetosoftX-rayemission asseenintheeareaplotsinFigure1.5.X-raytelescopeshavetheadditionalb thattheirimagesprovidedbothpositionalinformationaswellasenergyandtimeinformation foreveryevent(orphoton).Also, Chandra 'sprincipaladvantageoverotherX-raytelescopes, suchasXMM,isitsexcellentresolution(0.5 00 FWHMofthePSFonaxiswithpixels0.492 00 insize).Forgalaxyclusters, Chandra makesitpossibletoprobestructurenearthecenters ofcoolcoreclusters,wheredetailedstructures,suchasshockfronts,aswellasX-raycavities causedbyAGN,canbeobserved. 2 http://chandra.harvard.edu/graphics/resources/illustrations/acis_ schematic-72l.jpg 19 Figure1.5 Chandra eArea .Theeareaoftheback-illuminatedACIS-Sand front-illuminatedACIS-Idetectors.eareaasafunctionofwavelengthistheX-ray equivalenttotheshownforopticalandinfraredtelescopes.Themainerence istheadditionalsensitivitytheACIS-SchipshavetoobservesoftX-rays.Figurecourtesyof theChandraX-rayObservatoryCenter,whichisoperatedbytheSmithsonianAstrophysical ObservatoryonbehalfofNASA 3 . 3 Figure6.4from http://cxc.harvard.edu/proposer/POG/html/chap6.html 20 1.6.2GalaxyEvolutionExplorer TheGalaxyEvolutionExplorer(GALEX)isaspacetelescopebuiltbyNASA[Morrissey etal.,2007].SimilarlytoX-rays,UVlightisblockedbytheatmosphere,soUVtelescopes mustbespace-based.Whileitwasdesignedasanall-skysurveytelescope,therewastime dedicatedtoadditionalfollowupoftargetedsourcesaswellasgeneralregionsofthesky whichweredeemedtobeinteresting.Itoperatedintwobands,thefarUV(1350 1780 A)andnearUV(1770 2730 A)bandswhichwasprimarilyusedtomeasureyoungstar formationingalaxies.InFigure1.6weseetheGALEXband-passes 4 inrelationtoastandard setofopticalStarformationlessthan1Gyrisdominatedbytheblueststars,which arethemostluminous.Therefore,weuseGALEXRelease6(GR6)observationsofthe brightestclustergalaxiesinACCEPTetal.,2012a]. 1.6.3SouthernAstrophysicalResearchTelescope TheSouthernAstrophysicalResearch(SOAR)Telescopeisa4.1metertelescopelocatedon CerroPachon,ChileandMSUisapartner,receivinga15%fractionoftheavailablenights inexchangeforconstructionoftheSpartanInfraredCamera. 1.6.3.1GoodmanSpectrograph TheGoodmanSpectrograph[Clemensetal.,2004]wasbuiltfortheSOARtelescopebythe instrumentgroupattheUniversityofNorthCarolinaatChapelHill.Itisablue-optimized opticalspectrographwithimagingcapabilities.Itwasdesignedtomeasurepreciseradial velocitiesofstarsdownto1kms 1 precision.Weuseittomeasureemissionlineratios 4 http://galexgi.gsfc.nasa.gov/docs/galex/Documents/ERO_data_description_ 2_files/image038.jpg 21 Figure1.6GALEXBandpass .TheNUVandFUVband-passesfromGALEXareplottedin comparisontothebandpassesoftheSloanopticalu 0 g 0 r 0 i 0 z 0 set 5 . inthecoresandtsofbrightestclustergalaxies.TheGoodmanSpectrograph'sCCD isthinnerthananaverageopticalCCD,suchthattheGoodmanSpectrograph'sCCDis moreidealforobservingbluerwavelengths,however,itcausesinterferencefringingatlonger wavelengths.Thefringingisnoticeable( > 20%peaktotroughoscillations)above700nm, whichdoessomeofourobservationsanddescriptionsonthemethodstocorrectthis arepresentedinChapter3.Sincethereistfringinginthered,wehaveconcentrated ourobservationsonH tolowredshiftBCGs. 1.6.3.2SOAROpticalImager TheSOAROpticalImager(SOI)[Walkeretal.,2003]wastheinstrumentontheSOAR Telescope.Ithasanapproximateof5 0 5 0 .Itsharesnarrow-bandandbroad-band 5 http://galexgi.gsfc.nasa.gov/docs/galex/Documents/ERO_data_description_ 2.htm 22 Figure1.7SOIPolarimeterSetup .ThesetupoftheSOIinstrument 6 .Thepolarizersare placedinthewheelclosesttotheskytopolarizethelightpriortoapplyingwavelength dependentImageprovidedbytheSOARTelescope 7 . withtheCTIOBlancotelescope.Thevarietyofnarrow-band(typically50-100 Awide) allowsustoexaminelineemissionacrossthebrightestclustergalaxies.Giventhelimited numberofwehavefocusedourofmeasuringH tsforlowredshift z < 0.2brightestclustergalaxies.Becauseofthewidth,[NII] 6548,6583 Aemissionis alsoobservedwithH . Infall2010,Dr.WilliamSparksofSTScI,throughagrantfromtheDirector'sDiscre- tionaryResearchFundatthatsameinstitution,purchasedasetoffourlinearpolarizing (0 ; 45 ; 90 ; 135 )tointroduceapolarimetrymodeforSOI.InFigure1.7weshow theSOIimagerwiththelocationofthewheels.Whentakingpolarizedobservations, thepolarizersareplacedinthewheelclosesttothesky.Thenarrow-bandandbroad-band areplacedinthesecondwheel,directlybehindthewheel. 6 http://www.ctio.noao.edu/ ~ points/SOI/manual_software.htm 7 http://www.ctio.noao.edu/soar/ 23 1.6.4TwoMicronAllSkySurvey TheTwoMicronAllSkySurvey(2MASS)wasaground-basedsurveyinthenear-infrared wavebandsofJ,H,andK.Likeellipticals,BCGshavemostoftheirordinarymassaswell asluminosityintheirredstars,inparticular,theevolvedredgiants.Theblackbodyspectra ofthesestarspeakinthisbandmakingBCGsparticularlybrightinthesewavelengths.In additiontotheobservations,extendedsourcecatalogswerecreatedforsourcesfound[Jarrett etal.,2003]whichwewereabletousefornear-infraredmeasurementsofthestellarmassof BCGsinACCEPTclusters. 1.6.5SpitzerSpaceTelescope TheSpitzerSpaceTelescope[Werneretal.,2004],oneofNASA'sGreatObservatories,was originallytheSpaceInfraredTelescopeFacility(SIRTF)andlaunchedin2003.Spitzeris currentlyinitswarmmodebecauseitnolongerhascryogeniccoolant.Itcanonlyuseits twoshortestwavelengthdetectors,at3.6and4.5 m,becausethethermalnoiseistoohigh forlongerwavelengthdetections.Priortothat,itobservedbothnear-andmid-infrared emissionfrom3.6to160 m.Thetelescopeisfairlysmallwithaprimarymirrorofjust85 cmindiameter. 1.6.5.1InfraredArrayCamera TheInfraredArrayCamera(IRAC)operatesinfourseparatebands:3.6,4.5,5.8,and8.0 m [Fazioetal.,2004].ThesebandsareplottedinFigure1.8.Typicallyallfourbandsaretaken duringanobservation,soifagalaxywasobservedbyIRAC,wewillhavedataforallfour wavebands.Thetwoshortestwavelengthbandsarenearthepeakoftheblackbodyforared 24 Figure1.8IRACBandpass .ThesearetheSpitzerIRACnear-infraredband-passes.These curvesdescribethespectralresponseoftheentiretelescopesystemthroughputandquantum .Thesecurvescanbeusedtomakeandcolorcorrectionsbetweenwavebands. Figurefromthe Spitzer DocumentationandTools 8 . giantstarandareidealforobservingevolvedstellarpopulationsinellipticalgalaxies.These complementthenear-infrared2MASSobservationsandshouldtoasingleblackbody.The nexttwobandsarecenteredat5.8and8.0 mwherethereisalmostnoemissionfroma stellarblackbody.InsteadthesearecenteredonmolecularfeaturesofPAHswhichhavebeen observedinmanystar-forminggalaxies. 8 http://irsa.ipac.caltech.edu/data/SPITZER/docs/irac/calibrationfiles/ spectralresponse/ 25 1.6.5.2MidIRPhotometerSystem TheMid-IRPhotometerSystem(MIPS)hasthreedetectorsat24,70,and160 mallof whicharedlimited[Riekeetal.,2004].Inadditiontocolddustemission,the24 mbandisgoodfordetectingobscuredAGN,whichhavetheirsoftX-rayandUVemission re-radiatedinthemid-infrared[Donleyetal.,2008].Innearbygalaxies,the70 m coverstheblackbodypeakof ˘ 40K,nearapeakofcontinuumemissonformanystar formingregions.The160 mbandiscenteredonthestructuretransitionof[CII] 158 m(singlyionizedcarbon;thebracketsareaddedtoindicatethatitisforbidden,non- dipole,transition)whichisanimportantcoolantintheinterstellarmedium(ISM).The ofviewoftheMIPSdetectorsaremuchsmallerthanthatofIRAC.Also,the70and160 mobservationsrequiredspecialcoolingmodessotherearefewerobservationsofBCGsin thosewavebands.Whiletherearesomestarformationindicators,suchasPAHs,inthe IRACbands,themajorityofdustisseeninthemid-infraredcoveredbyMIPSandthese observationsgiveusabetterestimateofthetotalstarformation. 1.7OutlineofDissertation Thisthesisinvestigatesthephysicalmechanismsthatdrivetheinteractionsbetweenthehot ICMofthegalaxyclusterandthecoolcondensedgasinsomebrightestclustergalaxies.In eachofthesubsequentchaptersweexaminetwavelengthsandclustersamples. InChapter2wepresenttheresultsfromthepapereretal.[2012a].Inthispaper weinvestigatedthetotalstarformationinbrightestclustergalaxieswhichareingalaxy clustersintheACCEPTsample[Cavagnoloetal.,2009].Weusethe GALEX telescope, whichoperatesintheUV,tolookattheun-obscuredstarformationwhileweusethe 26 SpitzerSpaceTelescope inthenear-tomid-IRtoinvestigatestarformationthatisobscured andre-radiatedfromtheobscuredUVphotonsinthenear-tomid-IR. InChapter3wediveintogreaterdetailofasinglegalaxycluster,RXJ2014.8-2430, whichisthecoolestX-raycoreintheREXCESSclustercatalogohringeretal.,2007a]. Weexamineobservationsfrom Chandra tolookforpossibleX-raycavitiesandshocksnear theclustercore.WesloshinginthecorebutdonotX-raycavitiesandsimulate clusterobservationstosetthelimitonwhatX-raycavitieswemayhavemissedbasedonthe limitsofourobservations.WeuseSOARTelescopespectraandimagingtoinvestigatethe opticallineemissionintheBCG,fromSOIwithnarrowbandH imagingandGoodman foropticalspectroscopyacrosstheBCG.Relativeemissionlineratios,aswellasagradient intheemissionlinevelocities,presentapictureofthegasdynamicsoftheemissioninthe BCG,whichappeartobeeitherfallinginorgettingpulledoutofthecluster. InChapter4wemoveinaslightlyntdirectionanddiscusstheSOARpolarimetry pilotproject.Forthepilotproject,wepresentpolarizationmapsofsomefamouspolarized sources,includingnebulaandAGN.WeshowthatnotonlyisSOIcapableofmeasuring polarizationusingasetoffourpolarizers,butitcanplacetightlimitsonthepolarization ofextendedsourcesaswell.Tothisend,weinvestigatethepolarizationoftheH ts takenfromnarrowbanddataintheM87,thebrightestclustergalaxyinthenearbyVirgo cluster.Weplaceupperlimitsonthepolarizationintheucleustsanddiscussthe limitsofthephysicalmechanisms,primarilyconduction,allowedtoilluminatetheoptical ts.Thereisalsodiscussiononthestepsandcalibrationsnecessarytoobserveinthe polarimetrymodeonSOI. WeconcludetheresultsoftheworkinChapter5anddiscussimplicationsofthese studiesaswellassomeremainingquestionswhichstillexistandhowtheymaybeanswered. 27 WeincludeappendicesfortablesfromChapter2.Inthisthesisweassumea cosmologywith H 0 =70kms 1 Mpc 1 , =0 : 7,and M =0.3[Spergeletal.,2007]. 28 Chapter2 InfraredandUltravioletStar FormationinBrightestCluster Galaxies 1 Wepresentinfrared(IR)andultraviolet(UV)photometryforasampleofbrightest clustergalaxies(BCGs).TheBCGsarefromaheterogeneousbutuniformlycharacterized sample,theArchiveof Chandra ClusterEntropyTables(ACCEPT),ofX-raygalaxy clustersfromthe Chandra X-raytelescopearchivewithpublishedgastemperature,density, andentropyWeusearchival GalaxyEvolutionExplorer ( GALEX ), Spitzer Space Telescope,andTwoMicronAllSkySurvey( 2MASS )observationstoassemblespectral energydistributions(SEDs)andcolorsforBCGs.WethatwhiletheSEDsofsome BCGsfollowtheexpectationofred,dust-freeoldstellarpopulations,manyexhibitsignatures ofrecentstarformationintheformofexcessUVormid-IRemission,orboth.Weestablish ameannear-UV(NUV)to 2MASS Kcolorof6 : 59 0 : 34forquiescentBCGs.Weusethis 1 ThissectionhasbeenpreviouslypublishedintheAstrophysicalJournalSupplementun- derthesametitleetal.,2012a].Contributingco-authorsofthispaperareMeganDon- ahue,G.MarkVoit,andAmaliaHicksofMichiganStateUniversityandRamonBarthelemy ofWesternMichiganUniversity. 29 meancolortoquantifytheUVexcessassociatedwithstarformationintheactiveBCGs.We usebothtoatemplateofanevolvedstellarpopulationandlibraryofstarburstmodels andmid-IRstarformationrelationstoestimatetheobscuredstarformationrates.Weshow thatmanyoftheBCGsinX-rayclusterswithlowcentralgasentropyexhibitenhanced UV(38%)andmid-IRemission(43%)from8-160microns,abovethatexpectedfroman oldstellarpopulation.Theseexcessesareconsistentwithon-goingstarformationactivity intheBCG,starformationthatappearstobeenabledbythepresenceofhighdensity, X-rayemittingintergalacticgasinthethecoreoftheclusterofgalaxies.Thishot,X-ray emittinggasmayprovidetheenhancedambientpressureandsomeofthefueltotrigger thestarformation.ThisresultisconsistentwithpreviousworksthatshowedthatBCGs inclusterswithlowcentralgasentropieshostH emission-linenebulaeandradiosources, whileclusterswithhighcentralgasentropyexhibitnoneofthesefeatures. GALEX UV and Spitzer mid-IRmeasurementscombinedprovideacompletepictureofunobscuredand obscuredstarformationoccurringinthesesystems.WepresentIRandUVphotometry andestimatedequivalentcontinuousstarformationratesforasampleofbrightestcluster galaxies. 2.1Introduction Thebasicstoryunderlyingourcurrentmodelsfortheformationofgalaxiesandclustersof galaxiesisthatbaryonicmatterfallsintodarkmatterpotentialwells,coolstomakecold molecularclouds,whichthenformstarsandsupermassiveblackholes.Thestateofthegas asitfalls,themorphologyoftheaccretion,thesourceofthedustthatcatalyzesformation ofmolecularclouds,thephysicalprocessesdeterminingthegastemperaturesandphasesare 30 alluncertain.Simplyput,wedonotknowthefullstoryofhowintergalacticgaseventually formsstarsandblackholes. BrightestClusterGalaxies(BCGs)provideuniqueopportunitiesfortheinvestigationof theroleofhotintergalacticgasingalaxyformation,andinparticularitsrolein theevolutionofthestarformationandactivegalacticnucleus(AGN)activityinthecentral galaxyinthemostmassivedarkmatterhalosintheuniverse.Theintergalacticgasboundto amassiveclusterofgalaxies{itsintraclustermedium(ICM){outweighsthestarsinthose galaxiesbyafactorof5-10[e.g.,Davidetal.,1990a,Arnaudetal.,1992,Gonzalezetal., 2007].TheBCGsinthecentersofX-rayclusterswherethegashasashortcoolingtime (orequivalently,lowgasentropy)exhibitsignsofactivity(e.g.radiosources,emission-line nebulae,excessblueorultravioletlight)thatarerareinBCGsinotherclustersofgalaxies [Huetal.,1985,Burns,1990,Cavagnoloetal.,2008b,rtyetal.,2008,Sandersonetal., 2009,Sun,2009].TheactivityintheBCGsofthiscategoryofclustershasbeenpresented asevidencethathotICMcondensesintocolddustygasthatsubsequentlyformsstars.Such BCGsmaybehostingreal-lifeversionsoflate-time( z< 1)accretionontosupermassiveblack holesincentralgalaxies;buttheroleofthehotICMinAGNorstarformationactivityis notentirelyclear. Thesimplesthypothesisforhowhotgascoolswhenitistoamassivedark halofails.TheX-rayobservationsoftheICMingalaxyclustersindicatedthatsome clustershaveahighcentralgasdensitiesandcentralcoolingtimesshorterthantheageof theuniverse(e.g.,FabianandNulsen[1977];CowieandBinney[1977]).Inthisscenario,such gascoolsslowly,losespressuresupport,compresses,allowinggasfromtheouterpartsofthe clustertosettlegentlyintothecenter.Theinferredmassaccretionratescouldbeaslarge as1000solarmassesperyear[Fabian,1994a].Suchclustersweredubbed\coolingws." 31 However,higherresolutionX-rayspectroscopyshowedthattheluminousemissionlinesone wouldexpectfromgascoolingsmoothlyfrom10 8 Ktonon-X-rayemittingtemperatures werenotpresent[Petersonetal.,2003].Nevertheless,suchclustersdoexhibitcoolcores withradii ˘ 50 100kpc,where kT core ˘ 1 = 2 1 = 3ofthatfoundintheouterradii.These clustersarenowoftencalled\coolcore"clusters. WithspatiallyresolvedX-rayspectroscopy,coolcoreclusterscanbebythe distributionofgasentropyofthegalaxycluster.Theclusterentropyisathermodynamic quantity.Conveniently,inagasofpurehydrogenemittingthermalbremsstrahlungradia- tion,thecoolingtimecanbewrittendownsolelyintermsofthegasentropy.Thegasentropy S isproportionaltothelogarithmofthequanitity K =T X n 2 = 3 e ,conventionallyreported inunitsofkeVcm 2 .Donahueetal.[2006]radiallyentropywithafunctionalform K(r)=K 0 +K x (r = r x ) ,whereK 0 isthecentralentropyinexcessabovethepowerlawt. Cavagnoloetal.[2008b]extendedthisproceduretotheentire Chandra archive,creatingthe Archiveof Chandra ClusterEntropyTables 1 (ACCEPT).Galaxyclusterswithhigh centralentropyoftencontainquiescentbrightestclustergalaxies(BCGs)orexhibitevidence fortmergerorinteractions.Theempiricalboundarybetweenclusterswhichoc- casionallyhostactiveBCGsandclusterswhichneverhostthemisK 0 ˘ 30keVcm 2 ,an entropyassociatedwithanICMcoolingtimeof ˘ 1Gyr[Voitetal.,2008].Furthermore, about70%oftheBCGsinthosecoolcoreclustershostradiosources,andabouthalfof thosehostextendedemission-linenebulacharacteristicoflow-ionizationnuclearemission- lineregions(LINERs;butaremoreextended)[Heckmanetal.,1989,Crawfordetal.,1999, Donahueetal.,2010].Cavagnoloetal.[2008b]andyetal.[2008]haveshownthat onlythoseBCGsinhabitingclusterswithlowcentralgasentropies(shortcentralgascooling 1 http://www.pa.msu.edu/astro/MC2/accept/ 32 times,highcentralgasdensities)presentlow-ionizationemission-linenebulae(H ),blue gradients,orradiosources. Inthispaper,welookforsignaturesstronglyassociatedwithstarformation,ultraviolet (UV)excessesandmid-infrared(mid-IR)emissionfromdust,intheACCEPTsampleof well-studiedX-rayclusters.SinceevenanevolvedstellarpopulationemitssomeUV(and mid-IR),wecharacterizethestellarcontentoftheBCGusingTwoMicronAllSkySurvey ( 2MASS )K-bandphotometryandphotometryfromtheIRACinstrumentonthe Spitzer SpaceTelescope,short-wavelength3.6and4.5micronbands,whereavailable.Toestimate thecontributionofrecentstarformationwemeasuretheultraviolet(UV)emissionwiththe GalaxyEvolutionExplorer ( GALEX )observations.TheUVsamplesthepeakofemissionin short-livedOandBstars,thustrackingrecent,unobscuredstarformation.Mostofthestar formationintheuniverseoccurshiddenwithincold,dustymolecularclouds.Thedustin thesecloudsabsorbstheUVandopticallightofburiedstarsandre-emitsthislightasmid- IRthermalradiationtypicalofdustat ˘ 100K.Someofthisreprocessedemissionemerges intheformoffeatures,suchastheemissioncomplexesassociatedwithpolycyclicaromatic hydrocarbons(PAHs)[Donahueetal.,2011].Puzzlingly,powerfulH 2 featuresappeartobe nearlyubiquitousinsystemswithH nebulae,atlevelsunlikelytobeassociatedwithtypical starformationprocesses[ElstonandMaloney,1994,andBremer,1997,Donahueetal., 2000,2011,Egamietal.,2006].Evencolderdust(20-30K)inthefar-IRhasbeenseenwith Herschel [Edgeetal.,2010a][Edgeetal.,2010b],andEdge[2001]detectedtmasses ofCO. ThemeasurementsofstarformationinBCGsbasedonUVormid-IRinformationto datehavebeenrelativelylimited.Forexample,Catinellaetal.[2010]reportthatstar formationvarieslittleoverawiderangeofgalaxymassesinamassivegalaxy 33 sample.However,whilethatsampleincludes190massivegalaxiesobservedwith GALEX and Arecibo ,ithasveryfewBCGs.Donahueetal.[2010]assessedtheUVpropertiesof theBCGsinarepresentativesampleof30X-rayselectedclustersfromtheRepresentative XMM-NewtonClusterStructureSurvey(REXCESS)ohringeretal.,2007a],whilemost UVstudiesareofalimitedsetofthemostextremeemission-lineBCGs[e.g.,Hicksetal., 2010,O'Deaetal.,2010].Quillenetal.[2008]andO'Deaetal.[2008]studied62BCGs with Spitzer ,selectedfortheirluminousH .Toexpanduponthesestudies,wepresentan assessmentoftheUV,near-IRandmid-IRpropertiesofBCGsinawell-studiedsampleof X-rayclusters.Thissampleislargerandmorediversethanpreviousstudies,asitincludes quiescentBCGsalongwiththemostextremecool-coreBCGs.InSection2we describetheoriginalX-rayclustersample,andgiveanoverviewofthe GALEX and Spitzer observations.WedescribehowtheBCGsareidenInSection3wediscussthedata reductionprocessfortheimagesinthe Spitzer , GALEX ,and 2MASS archives.Ourdiscussion andanalysisofthedataisinSection4.Wepresentestimatesoftheequivalentcontinuous UVandIRstarformationratesinthissection.UVcolorsarecomparedtothoseinWang etal.[2010].Wepresentasummaryoftheobservations,detections,andemissionexcessesin Table2.1.WeconcludethepaperinSection5.Forallcalculationstheassumedcosmology is H 0 =70kms 1 Mpc 1 , M =0 : 3, =0 : 7. 2.2Observations 2.2.1 Chandra X-RayObservations TheoriginalgalaxyclustersampleisfromtheACCEPTdatabase[Cavagnoloetal.,2009], whichincludes239galaxyclusters.Thissampleisaselectionofallgalaxyclustersinthe 34 Table2.1.SummaryofObservationsandDetections WavebandObservationsObservations( K 0 30keVcm 2 )DetectionsExcess NUV1688411232 a 4.5micron76527613 8.0micron76527643 24micron98569424 b 70micron65463232 160micron33211616 a tobeNUV-Kcolorlessthan6.25,whichisatleastonesigmabluerthanthe meanofthequiescentBCGs. b tobea24microntoKbandratiogreaterthan0 : 113,whichisatleastone sigmagreaterthanthemeanofthequiescentBCGs. Note.|RefertoappropriatesectionoftextforofDetectionsandExcess. Chandra archiveasofAugust2008thatmetaminimumcriterion.Theclusterswere selectedtoconstructentropyandprovidecentralentropyestimates.Tobeable toaccuratelymeasuretheentropytemperaturegradientswererequiredtohavea precisionbetterthan X ˇ 1 : 0keV.Cavagnoloetal.[2009]thereforerequiredatleast threeconcentricannuliwithaminimumof2500countseach.Thesearchresultedin317 observationsof239galaxyclusters.SixgroupsfromtheHighestX-rayFlux GalaxyClusterSample(HIFLUGCS)sample[Reiprichandohringer,2002]wereaddedto thecollectionandanumberofclusterswithanalysiscomplicationswereremoved.(Alladdi- tionalobjectsarelistedinCavagnoloetal.[2009].)Thissampleisnotaformallycomplete sample,but,byandlarge,theseclusterswerenotselectedtobeincludedintheChandra programbecauseoftheUVandmid-IRpropertiesoftheirbrightestclustergalaxies.An interestinglylargefractionoftheseclustersnowhavebeenobservedbyGALEXandSpitzer, 35 andsothetimeisrightforauniformanalysisoftheX-ray,UV,andmid-IRpropertiesof theBCGsinthesample. 2.2.2 2MASS Observations-BCGiden Weusedthe 2MASS archiveandpreviousliteraturetodeterminethelocationsofthethe BCGsinthesegalaxyclusters(TableA.1).Thebrightestclustergalaxieswereinitially idenbytheir 2MASS position.ThelocationsoftheBCGsweredeterminedusinga visualinspection(includingsourcebrightnessandmorphology)with 2MASS J-bandimages 5 0 5 0 insizecenteredontheX-raycentroidtodeterminethebrightestgalaxyinthe cluster.ThisvisualinspectionwasfollowedupwithNASA/IPACExtragalacticDatabase (NED) 2 andtheSetofIdenMeasurements,andBibliographyforAstronomical Data(SIMBAD) 3 objectsearcheswithin2 0 oftheX-raycentroidtoverifytheredshiftsofthe candidateBCGs.Allobjectsinthe 2MASS ExtendedSourceCatalog[Jarrettetal.,2003] werecheckedforredshiftinformationandanyotherindicationthattheyarethebrightest galaxyinthecluster.SomeBCGsweretoodistanttohaveassociated 2MASS catalogentries. TheBCGsofthesedistantclusterswereidenusingaliteraturesearchforjournalarticles indicatingthelocationoftheBCGinthecluster,andarenamedbytheirrightascension anddeclination.FortheclustersintheSloanDigitalSkySurvey(SDSS)footprint,color informationandbrightnessintheoptical(u'g'r'i'z')fromthedatarelease7(DR7)wereused toverifytheBCGselections[Leismanetal.,2011].Inasmallnumberofcaseswerevised theoriginalselectionof 2MASS location(Abell2034,RXJ1022.1+3830,4C+55.16,Abell 2069,Abell368,andAbell2255).TableA.1giveseachclusterandthe 2MASS coordinate 2 http://nedwww.ipac.caltech.edu/ 3 http://simbad.u-strasbg.fr/simbad/ 36 Figure2.1BCGCentroidDistance .Theprojectedphysicaldistance,inkpch 1 70 ,between X-raycentroidandtheBCGweidenAllBCGswithanRAandDec,notjustthose with GALEX and Spitzer dataareplottedhere.ThetheshadedregionhighlightsBCGs inlowK 0 clusters.Inhighcentralentropysystems,37%ofBCGsliewithin10kpcofthe X-raycentroid,whilethepercentageisincreasedto74%forlowcentralentropysystems.All BCGswhichliegreaterthan40kpcawayfromtheirX-raycentroidareinhighK 0 clusters. forthebrightestclustergalaxy. InTableA.2welistthephysicalseparationoftheBCGsfromtheX-raycentroidoftheir hostgalaxyclusters.WhilemostBCGslieneartheX-raycentroidoftheirgalaxycluster, consistentwiththeiridenascDgalaxies,thereareafewthatareveryfarfrom thecenter.ThephysicaldistancebetweentheX-raycentroidandtheBCGisplottedasa histograminFigure2.1.TheBCGweidenistwiceaslikelytobewithin10projected kpcofitsX-raycentroidinlowentropyclusters(74%)comparedtohighentropyclusters (37%).NotethatallBCGsinalowK 0 systemarewithin40kpcoftheircluster'sX-ray centroid. 37 2.2.3 GALEX Observations TheGalaxyEvolutionExplorer( GALEX )obtainsimagesinthenearUV(NUV)at eff = 2267 A(bandpasswithafullwidthathalfmaximun(FWHM)of269 A)andthefarUV (FUV)at eff =1516 A(FWHMof616 A)[Martinetal.,2005].Thereareatotalof 168BCGsinourinitialsamplewith GALEX observationsinthe GALEX archiveasof2011 October.Wethensearchedthe GALEX ReleaseSix(GR6)catalogforaUVsourcewithin 5 00 ofthe 2MASS BCGlocation.Inthecaseswherethereweremultipleobservations,the observationwiththehighestsignaltonoisewasused.TableA.1givesthe GALEX object idenforeachBCGdetected.Notethatnotallobservationswillhaveanobjectiden astheBCGmayhavegoneundetectedinthe GALEX archive. 2.2.4 Spitzer Observations Weanalyzedarchival Spitzer InfraredArrayCamera(IRAC)andtheMultibandImaging PhotometerforSIRTF(MIPS)observations.IRAChasfournearinfraredwavebandsat3.6, 4.5,5.8,and8.0 m[Fazioetal.,2004].MIPS[Riekeetal.,2004]operatesinthemid-IR andhasthreewavebandsat24,70,and160 m.The Spitzer imagingobservationsselected foranalysiswereaimedwithin1 0 fromtheX-raycentroid.TheAstronomicalObserving Request(AOR)numbersaregiveninTableA.1.Thereare79brightestclustergalaxiesin ACCEPTwithIRACobservationsand100ACCEPTBCGswithMIPSobservationsasof 2010December. 38 2.3AperturePhotometryandColors 2.3.1 GALEX UVPhotometry Weused GALEX aperturephotometryprovidedinthe GALEX catalogandGALEXView 4 . Wechoseaperturestomatch GALEX measurementsandderivecolorswithphotometry fromothercatalogs(e.g. 2MASS ,SDSS)andwithour Spitzer aperturephotometry.The optimalaperturefortheUVisdeterminedbycomparingtheestimateofthetotalgiven inGALEXViewtothecircularapertureThecircularaperturechosenistheonewith themeasurementnearesttotheestimatedtotalvalue.FormostBCGsthetwo largestapertureradii(12.8 00 and17.3 00 )wereused.Theminimumallowedapertureradius was9.0 00 toavoidaperturecorrection(theFWHMof GALEX observationsare ˘ 4 : 5 00 6 00 ). The GALEX -detectedUVemissionisusuallycentrallyconcentratedsogenerally,theUV emissionlieswithinaradiusof9 00 evenwhentheangularsize,asseenintheoptical,of thegalaxyislarger.Therefore,the GALEX aperturesizeisanapproximateupperlimit onthesizeoftheUVstarformationregion.SomeoftheUVlightisproducedbyevolved stars[e.g.,O'Connell,1999]soweusetheNUV-KcolortoestimatehowmuchUVcomes fromrecentstarformation.Wemakephotometricmeasurementswithintlylarge aperturestominimizethedegreetowhichaperturecorrectionscouldourconclusions. Themagnitudesareconvertedfromthemagnitudesgiveninthe GALEX catalogtoAB magnitudesusingzeropointsof20.08magnitudesfortheNUVand18.82magnitudesforthe FUV[Morrisseyetal.,2007].TheGalacticextinctioncorrectionsareappliedfromSchlegel etal.[1998]assumingaratioof3.1forA V = E B V .TheNUVcorrectionassumedis3.25A V andtheFUVcorrectionis2.5A V .UVphotometryispresentedinTableA.3. 4 http://galex.stsci.edu/GalexView/ 39 2.3.2 GALEX UVUpperLimits Toestimatethedetectionthresholdfor GALEX observations,weevaluatedthecataloged ofallthewell-detectedsourceswithamagnitudeerror < 0 : 35( S=N & 3)within1 oftheBCGtargets.OurGALEXupperlimitsarebasedondetectionsofpeakedsources, i.e.pointsourcesandcompactemissionregions.Auniform,extendedsourcethatthe aperturewillhaveahigherdetectionthresholdthanthisestimate.Weplotthese asafunctionoftheirindividualexposuretimesinFigure2.2.Theestimateddetection thresholdisinferredfromtheupperenvelopeofthesepoints,whichisapproximatedhere bycurves / t 1 = 2 .Fortheexposuretimestypicaloftheall-skyimagingsurvey(AIS)the estimatefortheupperlimitinABmagnitudesis19+1 : 25 log t NUV foranexposure time t NUV inseconds.SimilarlythefunctionforABmagnitudeupperlimitfortheFUV is18 : 5+1 : 25 log t FUV .Thisrelationunderestimatesthe GALEX sensitivityforlonger exposuretimes,longerthan ˘ 500seconds.Thereare9BCGs(Abell2319,3C295,Abell 611,Abell665,Abell1942,Abell2631,CLJ1226.9+3332,HCG62,andAbell2219)which hadUVexposuretimesgreaterthan500secondsandhaveanondetection.Fortheseobjects welookedintheandsettheupperlimittobeequaltothedimmestsourcethatwas detected(withamagnitudeerrorlessthan0.35).Wereportthisestimated3 ˙ upperlimit forallcaseswheretheBCGwasundetectedandwhenthe GALEX sourcehadalarge error( > 0 : 35mag),indicatingahighlyuncertaindetection.ForBCGswithNUVupper limits,the 2MASS arematchedwitha7 00 aperturesuchthattheyaresimilarinsize totheGALEXPSF. 40 Figure2.2UVMagnitudeUpperLimits .TheUVmagnitudes(ABscale)forallUVsources within1oftheBCGlocations(regardlessofidentity)withuncertaintieslessthan0.35 magnitudes.Theupperenvelopeofthisdistributionservesasabasisforestimatingthe upperlimitesforundetectedorpoorly-detectedBCGsforexposuretimeslessthan400 seconds: FUV UL =18 : 5+1 : 25log 10 t and NUV UL =19 : 0+1 : 25log 10 t .Weconsideredall GALEX detectionswithmagnitudeerrors > 0 : 35tobepoorlydetected. 41 2.3.3 Spitzer NearandMidIRPhotometry Forthevastmajorityoftheobservations, Spitzer photometrywasmeasuredfromthe pipelineproductpost-BasicCalibrationData(pbcd).Thepipelinedatawereuxcalibrated inunitsofMJysteradian 1 [Reachetal.,2005]fromtheIRACpipelineversionS18.7.0and theMIPSpipelineversion16.1.0.ForthefourIRACwavebands,weremeasuredinside acircularaperturewitharadiusof r =14 : 3kpch 1 70 .WewroteanIDLprogramtoperform allaperturemeasurementsfor Spitzer [Donahueetal.,2010].Thecircularaperture iscenteredontheBCGlocationinTableA.1.Thebackgroundswerecomputedfroman annuluswithaninnerradiusof35 00 andanouterradiusof45 00 forobjectswhichhavean angularradiussmallerthan35 00 .Forobjectswithbeyondthenominalaperture,the backgroundwascomputedwithanannuluswithaninnerradiusof1.1 theradiusforthe objectandanouterradius1.3 theradiusoftheobject.Toestimatethemeanbackground countsweaGaussiantoahistogramofcountsperpixelinthebackgroundannulus.This procedureprovidesabackgroundestimatethatisrobusttopossiblesourcesofcontamination (e.g.foregroundstars)thatincreasethecountsinasmallnumberofbackgroundpixelsbut donottlyethemeanoftheGaussian. MostofthegalaxiesdetectedbyMIPSareessentiallypointsourcesbecausetheFWHMof thepointspreadfunction(PSF)for24,70,and160 mare6 00 ,18 00 ,and40 00 ,respectively.We measureMIPSusingthesameIDLcode.SincenotallofthefromthePSFfallsin theaperture,MIPSaperturearecorrectedusingthesameaperturecorrectionmethods in x 4.3.4oftheMIPSHandbook 5 .The24micronapertureradiusisat13 00 witha backgroundannulusof15-25 00 givingacorrectionfactorof1.167.Similarly,the70 5 http://irsa.ipac.caltech.edu/data/SPITZER/docs/mips/ mipsinstrumenthandbook/ 42 micronapertureradiusisat35 00 withabackgroundannulusof40-60 00 ,andacorrection factor,assuminga30Ksource,of1.22.The160micronobservationsweremeasuredatan apertureradiusof40 00 withabackgroundannulusof64-128 00 ,andacorrectionfactorof1.752 (alsoassuminga30Ksource).Wealsoderivedestimatesusingsoftwareprovidedby the SpitzerScienceCenter ,APEXinMOPEX[MakovozandMarleau,2005],tocross-check ouraperturemeasurements.Thestandardinputparameterswereusedandresidual imageswerecreatedtoassesswhetherthesourcewascompletelysubtracted.Forallsources withpropersubtraction,themeasurementfromAPEXwascomparedtotheaperture measurementandwevtheywereconsistentwithinthecitederrors.Onlythe valuescalculatedfromaperturesareincludedinTableA.4.InTable2.1detectionsand excessesareequivalentforthe70and160micronobservationsaswedonothavean apriori beliefthatquiescentBCGsshouldexhibit70and160micronemission. Fortheclosestand,likely,spatially-extendedBCGs,thefromAPEXweresystem- aticallylowerthantheapertureestimates.TodeterminewhetheranyBCGhadextended emissionorcontaminationfromunrelatedpointsources,wecomparedaperture-corrected measurementswith13 00 and35 00 apertures,andweinspectedthe24micronimagesforpoint sourcecontaminationwithinthe35 00 radiusaperture.Visiblecontaminationwas aseithert,becausethebetweenthetwoaperture-correctedestimates wassmallerthanthestatisticaluncertaintyofthoseort.Weinspectedall detectionsforpossiblecontaminationinside35 00 butweonlyfoundpotentialcontamina- tionintheannulusbetweenthe13 00 andthe35 00 radii(i.e.wesawnoobvioussources ofcontaminationinside13 00 ).Therefore,wedonotexpectcontaminationtothe24 micronpointsourcemeasurementslistedinTableA.4.However,theexistenceofany contaminatingsourceseenat24micronsisforour70and160micronphotometryin 43 TableA.4,whichuseslargerapertures.(RefertothefootnotesinTableA.4foradescription ofthecontaminationcategories.) BCGswhichdidnothavepointsourcecontaminationvisibleat24micronsbutshowed anincreaseinoverthatexpectedforapointsourceinthelarger35 00 apertureare consideredextended.Alltheobjectswhichhavebeenidenassuchare,unsurprisingly, nearbygalaxies.Insteadofcorrectingtheoftheseobjectsasiftheywerepointsources at24microns,theforthesegalaxiesarereportedforthelargeaperturesweusedfor theIRACphotometry.(Oneexception,theBCGNGC4636,wasmeasuredata35 00 radius insteadbecauseoftpointsourcecontaminationbeyondthisaperture.) Tomoredirectlyaccountfor70microncontamination,ifa70micronsourcewaslistedas adetectionandthe24micronmeasurementindicatedcontamination,the70micronimage wasinspectedforcontaminatingsources.Ifa70microndetectedsourceinsidetheaperture appearstocomefromanobjectotherthantheBCG(i.e.itscentroidisconsistentwiththat ofanon-BCGgalaxy)thenthedetectionwasdowngradedtoaconservativeupperlimit. However,these70micronupperlimitsarebasedonphotometryusingasmaller,16 00 aperture radiuswiththecorrespondingpointsourcecorrectionof1.94toavoidincludingfrom extraneouspointsourcesintheupperlimit.The70micronupperlimitsestimatedthrough thismethodarenotedinthetable.TherearetwoBCGs,Abell2744aandMS04516-0305, thatarecontaminatedat70micronsaswellas160microns.Upperlimitsfortheir160 micronphotometrywerefoundusingthesame16 00 apertureradiuswiththecorresponding pointsourcecorrectionof4.697. Thestandardphotometricerrorof5%isusedfortheIRACpointsasthesystematic errorswerealwaysmuchlargerthanthestatisticalerrors.ForMIPSthestandarderrorsare 10%,20%,20%for24,70,and160microns,respectively.Thesestandarderrorsareusually 44 goodestimatesexceptinthecaseoflower S=N detectionsforwhichstatisticaluncertanties areimportant(i.e. S=N =5 20).Wereportthetotalerrors(includingstatisticaland systematicuncertainties)forMIPSwiththemeasurementsinTableA.4. ForMIPS,upperlimitswereestimatedfordetectionsthatarebelow5 ˙ .Thestandard deviationoftheobservationwascalculatedinthesamemannerasDonahueetal.[2010]. Ifthestandardaperturehada S=N< 5thedatawereusedinstead.The backgroundonthesedataarebettercontrolled,buttheMIPSHandbookwarnsthatlow surfacebrightnessemissioninthedatawillbelost.Therefore,thedata wereonlyusedwhenthestandardsourcedetectionfellbelowthe5 ˙ limit.Those imagesthatarestillbelowthe5 ˙ detectionthresholdwereassigneda5 ˙ upperlimitforthat detection.IfaBCGisundetectedwiththestandardmosaicbutisdetected( > 5 ˙ )usingthe dataitisconsideredadetectionandislabelledassuchintheTableA.4.There weremanyobservationsthatwereconsidereddetectionsinourstpassthroughthe data,butwererevisedtoupperlimitsbecauseof70microncontaminationfromnon-BCG sources. ForafewofthenearestandbrightestBCGstherewasanissuewiththedata productsinthe Spitzer pipeline.Inthesecases,theBCGcontainedaspuriouspointsource thatwasmuchbrighterthantherestofthegalaxy.Theseverybrightartifactsprovedtonot bephysicalbecausetheanomalouslevelswerenotdetectedintheindividualBCDframes. WemosaickedtheindividualBCDframeswiththeMOPEXsoftwareusingthestandard mosiackingprocedureandsettings.Thenewmosaicimagesdidnotexhibitthespurious pointsources.Thewerethencalculatedfromthenewimagesandwereinagreement withtheoriginalimagesifthepointsourcewasmaskedout.ThoseAORswhichrequired remosaickingarenotedinTableA.1. 45 2.3.4 2MASS NearIRObservations 2MASS J,HandKanderrorsareextractedfromthe 2MASS ExtendedObjectCatalog [Jarrettetal.,2003].Thecatalogprovidesaperturephotometrybetween5 00 and60 00 in radius.Forafewlargegalaxies(e.g.M87,M49,NGC4696)theaperturephotometrywas takenfromthe 2MASS LargeGalaxyAtlas.Themeasurementswereconvertedfromthe system'sVegamagnitudestoJanskysusingtheABmagnitudeconversions(0.9,1.37,and 1.84magforJ,H,andKbands,respectively)providedinCohenetal.[2003].Wecorrect 2MASS magnitudesforGalacticextinction: A K =0 : 112 A V , A J =0 : 276 A V , A H =0 : 176 A V [Schlegeletal.,1998].Inordertoderiveratiosnormalizedtoemissiondominatedby theoldstellarpopulationsampledinthenear-infrared,wematchedaperturesinthenear-IR withthoseatotherwavelengths.Thereforeweestimated 2MASS photometry(presented inTableA.5)foreachsourceinthreeapertures:(1)the GALEX aperturefor NUV K , (2)theIRACapertureof r =14 : 3 h 1 70 kpcforIRACtonear-IRratios,and(3)the24 micronaperture(forK-bandonly).Afterextinctionandk-correction,Figure2.3showsthat theBCGshavenotrendintheirKbandluminosity(themeanis1 : 6 +0 : 7 0 : 4 10 44 ergs 1 h 2 70 ) asafunctionofredshiftor K 0 ofthesegalaxyclusters. 2.4Discussion 2.4.1UVExcessandColor TheUVexcessisdeterminedbycomparing NUV K colors,plottedinFigure2.4against excessentropy K 0 fromCavagnoloetal.[2009].ThebaselineforquiescentBCGsisvisiblein thisTheBCGswithexcessUVemission,overandabovetheUVfoundinquiescent 46 Figure2.3KbandLuminosity .TheKbandluminosityiscalculatedfromtheinside 14 : 3kpch 1 70 kpcradius.Theluminositiesarek-correctedassumingpassiveevolution.The solidhorizontallinerepresentsthemean(1.6 10 44 ergs 1 h 2 70 )ofthedatapointswhile thedottedlinesarethe1 ˙ error(+0.7 10 44 ergs 1 h 2 70 ,-0.4 10 44 ergs 1 h 2 70 )onthe mean. 47 BCGS,areonlyinthelowK 0 galaxyclustersintheACCEPTsample.Whileweno BCGswithexcessUVemissioningalaxyclusterswithhighcentralentropy,therearemany quiescentBCGsinlowcentralentropyclusters.Fromoursampleweestimatethetypical NUV K colorforquiescentBCGsfromthemeanandstandarddeviationofallBCGswith centralentropiesabove30keVcm 2 .WederiveameancolorofinertBCGsis6.59 0.34. Incontrast,themeancolorforBCGsinclusterswithcentralentropieslessthan30keVcm 2 is6.11 0.99.Weneacolorexcess c =6 : 59 (NUV K).Thisexcesswillbeusedin x 2.4.3toestimatetheequivalentcontinuousstarformationrate.Thecolorexcessissimply suchthatbluelightinexcessofquiescentBCGsinhighentropyclusterscaneasily betranslatedintoaUVluminosityassociatedwithcontinuousunobscuredstarformation. BCGsareconsideredtohaveaNUVexcessinTable2.1iftheirNUV-Kcolorisatleast1 ˙ bluerthanthemeancolorofinertBCGs.Weseethat38%oflowcentralentropyclustersin oursamplehaveaNUV-Kexcess.TheBCGswiththebluestcolorsareinAbell426,Abell 1664,andRXJ1504.1-0248whichhavecolorsaround3.0. Weplotthe FUV NUV and NUV K colorsforBCGsinFigure2.5.Contamination fromlineemissionfromLy mayoccuriftheredshiftedLy lineisincludedintheFUV bandpass(withintheFWHM(269 A)oftheewavelength(1516 A)oftheFUVr), atredshiftsbetween0 : 15 0 : 36.TherightplotsonlynearbyBCGs(z < 0 : 15)to addressthispossibleExcludingtheBCGswhichmaybecontaminatedbylineemission (z > 0 : 15),wedonotdetectaantFUV-NUVcolorbetweenbluerBCGs (withNUV-Kcolorslessthan6.3)andredderBCGs(withNUV-Kcolorsgreaterthan6.3). ThemeanoftheFUV-NUVcolorforbluerBCGsis0 : 73 0 : 57whilethemeanofredder BCGsis0 : 79 0 : 30. Wangetal.[2010]usesGALEXandSDSStomeasurecolorsonasampleof113nearby 48 Figure2.4NUV-KColor .NUV-Kasafunctionofclustercentralentropy.Thevertical dashedlineisat30keVcm 2 ,ourfortheoflowentropyclusters.Note thelargecolordistributionforlowK 0 objects,whilethehighentropyobjectshaveamore consistentreddercolor.TheKbandhavebeenk-correctedassumingpassiveevolution. ThehorizontaldashedlinerepresentsaNUV-Kcolorof6.59magnitudes,themeanofthe BCGsinclusterswithwithK 0 30keVcm 2 .Theremayappeartobeatrendwiththe lowentropyobjects,butthisisaselectionwherethelowestentropyobjectsthatare observedarealsothenearestobjects. 49 Figure2.5FUV-NUVNUV-KColor .ThebluetrianglesareBCGsinlowK 0 clusters( 30keVcm 2 )whiletheredasterisksareBCGsinhighentropyclusters.Theleftplotincludes alloftheBCGswhiletherightplotonlyincludesnearby(z < 0 : 15)BCGsdemonstratingthat thebluestFUV-NUVcolors,inthelefthandplot,arelikelyarisingbecauseofcontributions fromLy . 50 (z < 0 : 1)opticallyselectedBCGsandcomparethemtoasampleofeldgalaxies.Also, theycomparetheirresultstoasampleof21X-rayselectedBCGsfromyetal.[2008] whichincludedBCGsinbothcool-coreandnon-cool-coreclusters.FromFigure7inWang etal.[2010],thedistributionoftheFUV-NUVcolorisconsistentwithourswithameanthat bettermatchesthephotometryfromtheirouterapertures(radiuscovers90%ofthelight) thanthatmeasuredwithintheirinnerapertures(radiuscovers50%ofthelight).Similarly, theirNUV-rcolorsareconsistentwithourNUV-Kcolors,aftertransformationbetween SDSSrand2MASSKbands,assumingthosebandsareonlybyemissionfromthe oldstellarpopulation. 2.4.2IRColor Theratiosof8.0to3.6microntracktheratiosofinfraredemissionfrompolycyclic aromatichydrocarbons(PAHs),stochasticallyheatedhotdustgrains,andpossiblyrota- tionallyexcitedmolecularhydrogenandotheremissionlines[e.g.,Donahueetal.,2011]to emissionfromstars.WeplottheseratiosasafunctionofredshiftinFigure2.6.Theline showstheexpectationforapassivelyevolvingstellarpopulationwithanageof10Gyrat z =0.Afternormalizingtheratioforthestellarpopulation,wedeterminethetotalnumber thatareatleast1 ˙ abovethenormalizedmeanforBCGsinhigh K 0 clusters(1 : 014 0 : 061) andrefertothoseasBCGswithexcess8.0micronemissioninTable2.1.Thepointsthat liewellabovethislinearelikelytohavesomeformofhotdustand/orPAHemissionasthe observedIRAC8.0microncolorissensitivetoonlystrongPAHfeatures.InFigure2.7the IRACratioof4.5to3.6micronfrom Spitzer areplottedagainstredshift,similarlyto theplotfromQuillenetal.[2008].Similartowhatwehavedoneforthe8.0to3.6micron ratio,wenormalizethe4.5to3.6micronratioforapassivelyevolvingstellarpopulation 51 withanageof10Gyrat z =0.Wethendetermineameanofthenormalizedratiofor BCGsinhigh K 0 clusters(1 : 048 0 : 019).AllBCGsatleast1 ˙ inexcessofthemeanare consideredtohaveexcess4.5micronemission. Forboththe8.0to3.6micronratioandthe4.5to3.6micronratio,theonly BCGswithexcessesoverandaboveapassivelyevolvingoldstellarpopulationarethose thatinhabitclusterswithlowcentralentropies,asshowninFigure2.7andFigure2.6.In Figure2.8the8.0to3.6micronratioandthe4.5to3.6micronratioarestronglycorrelated ( r =0 : 92 ; 15 ˙ forobjectswithmid-IRdetectionsand/orNUV-Kexcesses),whichisexpected iftheexcess4.5micronemissionisgeneratedbyprocessesrelatedtothatproducingthe8.0 micronemission.Thefunctionalplottedis log 10 (F 8 : 0 m = F 3 : 6 m )=(0 : 153 0 : 002)+(5 : 422 0 : 021) log 10 (F 4 : 5 m = F 3 : 6 m ) : (2.1) Boththeratioshavebeennormalizedforpassiveevolution.AslongastheIRACcalibration wasconsistentovertime,thesearepreciserelativeratios,independentofthecali- bration.Theabsoluteratiosareprecisetoabout2%.The8.0and4.5micronbandpasses willincludePAHandmid-IRemissionlinefeaturesassociatedwithactivityseenincoolcore BCGs[Donahueetal.,2011].Theemissionofdust-free,evolvedstellarpopulationsinthese samebandpassesissimilartotheRayleigh-Jeanstailofablackbody,decreasingsteeplyto longerwavelengths.WeseetwoBCGsHCG62andAbell1644thatshowanexcessinboth normalizedratioshoweverneithershowsaNUV-Kexcess.Abell1644wasnotobservedin MIPSandweexpecttoseeadetectioninthe70micronwavebandbasedonthiscorrelation. HCG62hasa70micronupperlimitwhichmayberelatedtotheselectionthatitisa verylow K 0 galaxygroup.Weassessthepresenceofaluminousdustcomponent,likelyto 52 Figure2.68.0-3.6InfraredRatio .Redshiftdependenceof8.0 mto3.6 mratio.The dottedlinerepresentstheexpectedratioforpassivelyevolvingstellarpopulationthat is10Gyrat z =0.WhileIRSspectraofH -emittingBCGsshowPAHfeaturesthatwould fallinthe8.0micronbandpass[e.g.Donahueetal.,2011]theobservedIRAC8.0micron colorissensitivetoonlystrongPAHfeatures.Below,theratiohasbeennormalizedby thepassiveevolutionmodelandisplottedagainstthecentralentropyofthecluster.The dottedlineidenthethreshold30keVcm 2 .Thereappearstobeacitofexcess-IR emittersinthelowK 0 clusters,butthisislikelytobeaselectionctsincelow K 0 canonlyberesolvedinthemostnearbygroups. 53 Figure2.74.5-3.6InfraredRatio .Thelefthandshowstheratiobetweenthe 4.5 mandthe3.6 misplottedasafunctionofredshift.Thedottedlineindicates aStarburst99modelforapassivelyevolvingellipticalgalaxywithaprimarilyoldstellar populationdominatedbyredgiants,withanageofabout10Gyrat z =0.Formost oftheBCGstheIRAC4.5 mto3.6 mcolorsareconsistentwiththoseofapassively evolvingpopulation.Inthebelow,theratiohasbeennormalizedbythepassive evolutionmodelandisplottedagainstthecentralentropyofthecluster.Thedottedline againiden K 0 =30keVcm 2 .ItisinterestingthatthehandfulofBCGs(Abell426, Abell1068,Abell1835,andZwCl0857.9+2107)withlargeexcess4.5micronemissionare locatedonlyinclusterswith K 0 lessthanthethreshold. 54 Figure2.8InfraredRatioCorrelation .8.0micronto3.6micronratioand4.5micronto 3.6micronratio.Bothratioshavebeennormalizedforpassiveevolution.The8.0/3.6and the4.5/3.6ratiosarestronglycorrelated( r =0 : 92(15 ˙ )forobjectswithmid-IRdetections and/orNUV-Kexcesses(shownasintriangles),whichisexpectediftheexcess4.5 micronemissionisgeneratedbyprocessesrelatedtothatproducingthe8.0micronemission. Dashedlineisatothedata;seetext.Abell1644andHCG62donothaveblueNUV-K colors,HCG62isa70micronupperlimit,andAbell1644wasn'tobservedbyMIPS.Since theseratiosuseonlyIRACdata,theuncertaintiesintheabsolutecalibrationarenot included.AslongastheIRACcalibrationwasconsistentovertime,thesearepreciserelative ratios.Theabsoluteratiosareaccuratetoabout2%. 55 beobscuredstarformationbutalsocouldbecontributedbyanAGN,bylookingatthe24 microntoK-band(2.2micron)ratioplottedagainstthecentralentropyinFigure2.9. WenoteasimilarpatternhereasfoundintheUVexcessplots(Figure2.4),thatthelow K 0 galaxyclustersarefarmorelikelytohostBCGswithwarmdust.Thepossibleexception tothispatternisAbell521,whichisahighentropyclusterwithanelevated24micronto Kbandratio.However,asseeninFerrarietal.[2006]thereisalowentropy,compact, X-raycorona[Sunetal.,2007](i.e.a\mini-coolingcore")aroundtheBCGinAbell521, embeddedinaclusterwithotherwisehighentropy.Excess24micronemissionisestimated bydeterminingthemeanofthe24-KratioofBCGsinhigh K 0 systems(excludingAbell 521)andanyBCGwithatleast1 ˙ abovethismean(0 : 063 0 : 050)isconsideredtohave excess24micronemission.Weseethat43%ofthecoolcoresinoursamplehaveanexcess intheir24microntoKbandratio.TheBCGswiththemostextreme24microntoKband ratiosareinZwCl0857.9+2107andCygnusAwitharatioofabout20.BCGsinAbell 426andAbell1068alsohavelargeratiosaround10.AllfourobjectslikelyhavesomeAGN contribution.Weseethescatter(i.e.standarddeviation)intheratiolog 10 (F 24 m = F K )is 0.81forBCGsinlowcentralentropyclusters. WecancompareIRratiosinBCGstothoseofnormalstar-forminggalaxiesandstar- bursts,similartoFigure1inJohnsonetal.[2007a].TheratiosfortheBCGsareplottedin Figure2.10aswellastheSINGSgalaxies[Kennicuttetal.,2003].Similartotheirsample ofawiderangeofgalaxies,theBCGsinoursamplehavethesamecolorsasstar-forming galaxiesinSINGS.WenotethatsomeofthenearbyBCGshaveahigherratioof8micronto 24micronemissionbyafactorof2.Thisratiomayindicatearelativelylargercontribution fromPAHemissionoververywarmdust.Also,thisbandpassmayincludecontributions fromtheS(4)transitionofmolecularhydrogen.Rotationallyexcitedmolecularhydrogen 56 Figure2.9Mid-IRColor .24microntoKbandratiowithcentralentropyofthe cluster.BCGswithexcess24microninhabitclusterswithlow K 0 ,withtheexception ofAbell521.EventhoughitisahighK 0 cluster,theBCGinAbell521isinalowentropy, compact,X-raycorona(i.e.a\mini-coolingcore")whichcanbeassociatedwithBCGswith radiosourcesandstar-formationactivity,likeBCGsinlowK 0 clustersofgalaxies[Ferrari etal.,2006]. 57 Figure2.10SINGSGalaxyComparison .ThisplotissimilartoFigure1inJohnsonetal. [2007a].TheblackdotsareourBCGsandtheSINGSgalaxies[Kennicuttetal.,2003]are overplottedasredtriangles.RatiosforobjectswithMIPSupperlimitsat24or70microns arenotplottedbutareconsistentwiththedistributionofthedetectedgalaxies.Anerror bar,representingthestandardIRACandMIPSsystematicerrors,isplottedtorepresenta typicalerrorbar.SomenearbyBCGshaveslightlyhigher8.0/24micronuxratiosthan SINGSgalaxies,butsimilar24/70micronratios. linesareextremelyluminousinsomeBCGs,andthesesamelinesarenotbrightinstar forminggalaxies(Donahueetal.2011).Itispossiblethatsomeoftheexcessemissionat the8.0micronmaybecontributedbymolecularhydrogen. 58 2.4.3StarFormationRates(SFRs) TheUVcolorexcess, c in x 2.4.1,canbeusedtoestimatetheexcessUVluminosity duetounobscuredstarformation: L SFR =L (1 10 c = 2 : 5 ) ; (2.2) wherethespluminosityL isconvertedfromtheNUVABmagnitude,correctedfor Galacticextinction.TheNUVk-correctionforastarformingspectrumisnegligibleoutto moderateredshifts[Hicksetal.,2010].Theunobscuredstarformationrateisthenestimated fromtherelationinKennicutt[1998]andlistedinTableA.6.ThetotalUVluminosityis estimatedtobeL UV ˘ L using =c = 2267 A.Upperlimitsarebasedon3 ˙ uncertainties inUVexcesses. Theobscuredstarformationrateisestimatedintwoways,(1)byngGrovesetal. [2008]starburstmodelsandamodelofanoldstellarpopulationtothe 2MASS and Spitzer IRAC/MIPSinfrareddatapoints,and(2)fromusingcalibratedconversionsofIRluminosity (mostly24and70micronluminosities)toSFRs.Inthecase,wepresentasumofthe twomodels,withindependentnormalizations.Starformationratesweredeterminedforall BCGswithdatafromatleast 2MASS andthe24 mbandofMIPS.Toestimaterest-frame IRluminositiesbasedonthe24and70micronk-correctionswereappliedsuchthat L rest =kL obs .ThecorrectionswerefoundusingthebGrovesmodelforthatindividual galaxyandconvolvingitwiththeMIPSbandpass,bothintherestframeandtheobserved frameofthegalaxy.Theactual70microncorrectionsdonotdependverymuchonthesp Grovesstarburstmodel.However,the24micronpointusuallyfallsaroundaminimumin thespectrum,whichcausesalargerscatterintherelationforagiveredshift(upto30%) 59 Figure2.11UVandIRSFR .Thedottedlinerepresentsalineofunity.TheUVSFR assumesaconstantrateofstarformation.Themodel-derivedIRstarformationratesare consistentwithstarformationratesmeasuredwithaMIPS70 mSFRestimateasshownin Figure2.12.Thoseobjectswhichfallbelowtheline,BCGswithexcessIRstarformation,are similartostarburstgalaxies,inthesensethatforthemostluminousstar-forminggalaxies, mostofthestar-formationisobscured. The24micronk-correctionsareintherange(0.125-1.056),the70micronk-correctionsare intherange(0.738-1.879).ThetotalIRluminosity,L dust ,isestimatedbyintegratingthe totalscaledstarburstmodelover 8-1000 m.WeplottheUVstarformationrateagainst theIRstarformationrateinFigure2.11. Calibratedconversionsforstarformationratesfrom24and70micronluminositieswere usedfromCalzettietal.[2010].The70micronluminosityconversiontoaSFRwasfrom Equations(21)and(22)fromthispaper,dependingontheluminosityofthatgalaxy.The24 micronSFRrelationwasfromEquation(6)whichisfromWuetal.[2005].FromFigure2.12 60 Figure2.12ComparisonofmodelIRSFRtosingleband70 mIRSFR .Wecomparethe estimatesfromtheGrovesmodelstarformationratestothestarformationrateestimates usingthe70micronluminosityintheleftplot.Thelinerepresentsalineofunity,nota Thedottedlinesrepresenttheboundaryforaofafactoroftwoinstarformation rate.TheblacktrianglesrepresentthehighluminosityrelationgiveninCalzettietal.[2010] whilethereddiamondsusetheirrelationforgalaxieswithlowIRluminosities(andtherefore alargeramountoftheIRisproducedbydustheatedbyevolvedstarsratherthanhot stars). andFigure2.13wehaveacomparisonbetweentheseconversionsandthemodelcalibrated starformationrate.The24micronluminosityisnotasgoodatpredictingthebolometric IRluminosity(andtheintegratedstarformation)becauseitdoesnotsampleasclosetothe colddustmid-IRemissionpeakasthe70micronluminosity.The70micronismuch closertothepeakandislikelyabetterestimateoftheIRluminosityandtheobscuredSFR. WeestimatetheIRexcessIRX=log 10 (L dust = L UV )andplotitagainstthe FUV NUV 61 Figure2.13ComparisonofmodelIRSFRtosingleband24 mIRSFR .Theplotisasimi- larplottoFigure2.12relatingthe24micronluminositytotheGrovesmodelstarformation rates.The24micronSFRstendtohavelowerestimatesasthe70micronluminositiesare foundnearertothepeakofthedustblackbodyandthereforemorerepresentativeofthe totalIRluminosityandSFR.Forthemostluminous70micronsgalaxies,itappearsthat themodeltendstooverpredicttheSFRandluminosityrelativetothe70micron estimate. 62 Figure2.14IRexcessandUVcolor .TheIRexcess(IRX)isinJohnsonetal.[2007a] tobetheratioofIRtoUVluminosity.ObjectsfromthecoolcoresampleofHicksetal.[2010] aremarkedwithX's.Figure6afromJohnsonetal.[2007a]isplottedinthebackgroundon thisgraph. color(Figure2.14),similartoFigure6intheJohnsonetal.[2007a]paper,whichpresents UVandIRdataforasampleofstar-formingdiskgalaxiesandstarburstgalaxies.Inan earliercomparisonofBCGswithstar-forminggalaxies,Hicksetal.[2010]foundthatthe coolcoreBCGsintheirsampletendedtobebluerinUVcolorandhavealargescatterin IRXcomparedtothosepropertiesinthegalaxiesinJohnsonetal.[2007a].Wedoseethe largerscatterinIRXforthoseBCGsthathaveabluerUVcolor.Wenotethatmostofthe BCGsinourplotarefoundinlowcentralentropyclustersbecausethosearetheonlyBCGs withFUV,NUV,and Spitzer mid-IRdetections. 63 2.4.4StarFormationandClusterEntropy Wehaveshownhereandinpreviousworks[e.g.,Cavagnoloetal.,2008b,2009,y etal.,2008],thatBCGsinclusterswithlowcentralentropy( K 0 )aretheonlyBCGsto exhibitsignsofvigorousstarformation.TheupperthresholdforactivityinBCGsappears tobearound30keVcm 2 .Table2.1presentsthesubsampleswithexcessemission.We investigateheretoseewhetherthestrengthofthesignaturesofactivity,theUVandmid-IR excess,exhibitedanytrendwiththecentralentropyororotherclusterproperty. Herewetakethederivedstarformationratesassimplyindicativeofthelevelofstar formationactivity.Byassumingthestarformationisconstant,wehavetakenanominal assumptionabouttheconversionfactorsandthestarburstmodels,andtranslatedluminosi- tiesintoSFRs.Wearenotclaimingthatthestarformationiscontinuous.Distinctions betweencontinuousstarformations,simplesingle-burstmodelsofasingleage,andconvolu- tionsofmorecomplicatedstarformationhistoriesarewellbeyondthescopeofbroad-band photometricdataandglobalmeasurements.Forexample,extremelyrecentstarformation isbesttrackedwithH ,buttheH availablefromtheliteraturearetypicallyfrom long-slitspectra,andthereforecanunderestimateemissionlineifsomeofitislocated outsidethecentral2 00 orso.H canalsobebydustextinctioninheavilyobscured regions;H canbeproducedbymechanismsotherthanbyrecombinationinstarformation regions.Mid-IRemissionprovidesaprettyreliableassessmentoftheobscuredstarformation energyoutput,sinceitislikeabolometricmeasureofluminosityemittedbydust.Atlow starformationrates,thecolderdust,heatedbyevolvedstarscancontributetothelonger wavelengthemission,sothelowestIRSFRsinoursample(belowabout0.1solarmassesper year)mayberegardedasupperlimits.TheUVlightfromagalaxyisverysensitivetothe 64 presenceofhotstarsifsomeoftheirlightescapesthegalaxy.Wedonotattempttocorrect theUVlightforinternalextinction,sotheUVandthemid-IRaresamplingcomplementary componentsofanystarformation-relatedlight. AsumoftheUVandIRSFRsisthereforeabestestimateofsomethingakintothetotal starformationrateoftheBCG,andeventhemostconservativeinterpretationisthatthey indicatethecurrentluminosityofstarformationintheBCG.Wedonotseeanycorrelation betweentheentropyandthestrengthofstarformationsignatures(e.g.theUVor themid-IRluminositiesoftheBCGswithvariousX-raygasquantities, K 0 orthevalueof theentropyat20kpc( K ( r =20kpc))).InFigure2.15weplotthequantitiesofSFR and K 0 .Uponglance,theremayseemtobeatrendforthedetectedlowestentropy systemstohavetheloweststar-formationluminosities.However,thesearelowestredshift groupsintheACCEPTsample,withlowerluminositiesandmassesoverall.Theyarequite nearby,sotheonesthatarewell-observedbyChandrahaveentropythatprobethe subkpc-scales.Excludingthegroups(orincludingtheupperlimitsforBCGsingroups withoutevidenceforstarformationactivity)erasesanysemblanceofatrend.Totestthat wewerenotmissingatrendbecausethebest K 0 couldbebiasedhighforthemoredistant clusters(seeCavagnoloetal2009),weplot K 0 andSFRfortheBCGswithzbetween0.05 and0.15.Inthissubsample,notrendisvisible.Furthermore,theexpectedtrendwouldbe thatthelowestentropysystemswouldhavethelargeststarformationluminositiesbecause thegashastheshortestdetectedcoolingtimes.Thereforeweseenoevidenceforasimple relationbetweencentralgasentropyorcoolingtimeandtheestimatedSFR. 65 Figure2.15Relationbetween70micronSFRandcentralentropy .Theobjectsarecolor codedandsizedbasedonredshift(i.e.higherredshift,largersize).Thecolorcoderanges fromredshiftof0.0to0.9.The\trend"thatisseeninthelowestcentralentropyclusters isnotaphysicaltrend,butitisthesameselectionnotedinFigure2.4.Notethatall BCGsinclusterswith K 0 > 30keVcm 2 inthisplothaveonlyupperlimits. 66 2.4.5ICMGasCoolingandStarFormationinBCGs Whilethepresenceofhighdensity,highpressureintraclustergasseemstobeaprerequisite foraBCGtohostsomestarformation,roleoftheintraclustergasisnotquiteclear.The currentparadigmsuggeststhatsomeofthehotgascoolsandformsstars,butagasthathas beenatX-raytemperaturesforsometimehaslikelysputteredawayanygrainsitmayhave had.ThelifetimeofatypicalGalacticdustgrainin10 7 Kgasisoforder10millionyears [DraineandSalpeter,1979].Dust-freegasformsmolecularhydrogenonlyveryslowly[e.g., Brommetal.,2009].VoitandDonahue[2011]showthatforBCGswithmeasuredreservoirs ofCO(andH 2 ),thegasresidencetime(= M ( H 2 ) =SFR )forBCGsisverysimilartothat ofstar-formingdiskgalaxiesatSFR < 10M yr 1 ,aroundaGyr.ForBCGswithrapid SFRs,theresidencetimeissimilartothatofstarburstswithsimilarSFRs( ˘ 10 7 10 8 yrs).TheycalculatethatifmuchofthestellarwindsandejectaofevolvedstarsintheBCG areretainedbytheBCG,perhapsasaconsequenceofthehigherintraclusterpressures,this gascouldfuelmuchoftheexistingstarformationoccuringatasteadyrate.Certainlyfor BCGswithSFR ˘ 10solarmassesperyearorless,thestellarejectaisasourceofmaterial thathasmassofsimilarorderofmagnitudetoanysourceofcooledICMgas. However,forgalaxieswithgasreservoirsof10 10 solarmassesormore,cooledICMappears toberequiredtosupplythemolecularclouds.Thestellarejectaorcontributionsfromthe ISMofdustygalaxies[e.g.Sparksetal.,1989b]mayprovidedustyseedsthatmaymixwith theICMandcantlyaccelerateitscooling.ThelargerSFRscannotbesustainedata steadyrate,giventhegassupply,andjustasinstarburstgalaxies,mustbeashort-term situation.Thegasmayaccumulateoveralongerperiod.Giventhat ˘ 1 = 3oflowredshift coolcoregalaxiesexhibitH ,asimilarfractionofcoolcoreBCGs(orpossiblyfewer,if 67 someoftheH emissionisnotrelatedtoSF)areinthestar-formingstate.Therefore,such galaxiescouldaccumulatetheejectaoftheirstellarinhabitantsintomolecularcloudsfor Gigayears,thenexperienceaburstonceathresholdsurfacedensityofmolecularhydrogen wasachieved. Theempiricalcorrelationbetweenthepresencelow-entropyICMandthestarformation inthecentralBCGisincontrovertible.However,thecommoninterpretationofthiscor- relationthatcooledICMfuelsthestarformationhasnotbeenbackedupbyaphysically plausibletheoryforhowthehotICMcoolsandmakescoldanddustymolecularclouds.The residentstellarpopulationisanobvioussourceofdust(andcoolgas)thatshouldnotbe neglected. 2.5Conclusions WepresentphotometryforbrightestclustergalaxiesintheACCEPTclustersample,derived from GALEX , Spitzer ,and 2MASS archivalobservations.Thissampleincludes239clusters whichwerewell-observedbyChandraupuntillate2008,withhotgasentropyuni- formlyextracted[Cavagnoloetal.,2009].WeidentheBCGsinalloftheclusters. InourBCGidenitistwiceaslikelytobewithin10projectedkpcofitsX-ray centroidinlowentropyclusters(74%)comparedtohighentropyclusters(37%). Similartowhathasbeenseeninotherstarformationindicators(e.g.H ),galaxyclusters withlowcentralgasentropies(alsoknownas"coolcore"clusters)aretheonlyclustersto hostBCGswithinfraredandUVexcessesabovethosefromtheoldstellarpopulation.The entropythresholdof30keVcm 2 isconsistentwiththeentropythresholdidenbyother work[Cavagnoloetal.,2008b,yetal.,2008,Cavagnoloetal.,2009].Wefound168 68 observationsbythenearUVimagingby GALEX ,ofwhich112BCGsweredetected.We foundameanNUV-K(6.59 0.34)colorseeninquiescentBCGsandusethattoquantify excessUVemissioninindividualBCGs.Ofthe84clusterswithlowcentralgasentropy, 32(38%)hostedBCGswithaUVexcess,whilenoneoftheclusterswithhighcentralgas entropydid.Thescatter(i.e.standarddeviation)intheNUV-KABcolorofBCGsinlow entropyclustersisconsiderablyhigherat0.99.Wedidnotdetectabetweenthe meanUVcolor(FUV-NUV)ofBCGs(notincludingthosewithpossibleLy contamination), withintheerror,forlowandhighentropyclusters. Similarly,wedetectedexcessinfraredemissioninsomeBCGsinlowgasentropyclusters overalargerangeofinfraredbands(e.g.4.5,8.0,24,and70microns)andnoexcessin BCGsinhighcentralentropyclusters.Themid-IRemissionratiosforBCGs(including quiescentBCGswithmid-IRdetections)areconsistentwith,andspanasimilarrangeto, galaxiesstudiedintheSloanDigitalSkySurvey(SDSS)galaxieswitharangeofstarforming propertiesbyJohnsonetal.[2007b].Forexample,24oftheobeserved56BCGs(43%)in lowentropyclustersshowexcess24microntoKbandemission.Thestandarddeviationof theratiolog 10 (F 24 m = F K )is0.81intheseBCGs.Wealsoseeastrongcorrelationbetween excess4.5micronand8.0micronthatmayindicatecorrelatedPAHemissioninboth ofthesebands,whenthePAHemissionisstrong. TheexcessemissionseenintheUVcanbeusedinconjunctionwiththeIRemission toestimateatotalstarformationrate,accountingforbothobscuredandunobscuredstar formation.TheUVandIRestimatesgivecomplementaryinformationwhereasH maybe bycontaminatingcontributionsfromothersources(e.g.dustextinction,shocks) orlimitedbytechnique(e.g.incompletespatialcoverageinlongslitspectroscopy,contam- inationbyN II innarrowbandimaging).Additionally,themulti-wavelengthcoverage(as 69 opposedtosinglebandmeasurements)canhelptofurtherconstrainpossiblesourcesofthe excessemission.Weseethatthenear-IRtofar-IRemissionisconsistentwithacombination ofastarburstmodelandanoldstellarpopulation.Clearsignsoftheseempiricalcorrelations andtdustemissioninsomelowentropyclusterscanhelpconstrainstarformation estimatesintheseBCGs.Asidefromthepreviouslynotedupperthresholdforactivityat K 0 =30keVcm 2 ,wedonotdetectacorrelationbetweenthelevelofluminositiesorexcesses with K 0 (orequivalently,centralcoolingtime.)However,whetherthegasfuelingthisactiv- itycomesfromcoolingoftheICMorotherprocesses,isnotsoclear.At,massive evolvedstellarpopulationinthesegalaxiesmayproducedustygaswhichmaybe bythehotgasanditmayprovidetheseedsofcondensationforthegasfromthehot,and presumablydust-free,intraclustermedium[VoitandDonahue,2011]. Supportforthisresearchwasprovidedby Spitzer contractsJPLRSA1377112(MSU RC065166)andJPL1353923(MSURC065195).M.DonahueandA.Hickswerepar- tiallysupportedbyaLongTermSpaceAstrophysicsgrantNASANNG05GD82G(MSU RC062757).WewouldalsoliketothankDeborahHaarsmaandLukeLeismanfortheir helpfuldiscussiononBCGidenandMarkVoitforhiscommentsonthetext.This researchhasmadeuseoftheSIMBADdatabase,operatedatCDS,Strasbourg,France.This researchhasmadeuseoftheNASA/IPACExtragalacticDatabase(NED)whichisoperated bytheJetPropulsionLaboratory,CaliforniaInstituteofTechnology,undercontractwith theNationalAeronauticsandSpaceAdministration. 70 Chapter3 MultiwavelengthStudyofthe ExtremelyCoolCoreClusterRXJ 2014.8-2430 WepresentX-rayandSouthernAstrophysicalResearch(SOAR)Telescopespectroscopic andnarrow-bandimagingforthemostextremecool-coreclusterintheRepresentativeXMM- Newton ClusterStructureSurvey.Surprisingly,the Chandra imagingobservationsdidnot revealbi-lateralX-raycavitiesonemightexpecttoseeinanextremecoolcorewithapowerful radiosource;cavitiesthatcommonlyappearinothersimilarsources.Wediscussthelimits onthepropertiesofaputativeradiobubbleassociatedwithanyundetectedX-raycavities. WeplacelimitsonanytX-raypointsourceinthebrightestclustergalaxy(BCG) wheretheX-raypeakisfromcentralradiosource.Thedataareconsistentwitha possiblecavitysystemalongthelineofsighttothecenteroftheclusterorwithapossible sloshingsignature.Theseobservationsleadustoconcludethateitherweareseeingayoung radiosourceinashort-livedphaseofactivityortheradiosourceanditscavitiesintheX-ray gasarenearlyalignedalongthelineofsight.TheimagingandspectroscopyofSOARreveal anextended,luminousopticalemissionlinesource.FromournarrowbandH imagingof theBCG,thecentralH peakisslightlyfromthe Chandra data,consistentwitha 71 sloshinghypothesis.However,wearguethatanysloshingmustberathergentleinnature giventheco-locationoftheH andstellaremissionpeak,theconcentrationoftheX-ray emissionpeakandthedistributionofmetals. 3.1Introduction TheevidencethataradioAGNcoulddisturbtheX-rayatmosphereofclustersofgalaxies wasseenwiththeHighResolutionImageronboardROSAT[Boehringeretal.,1993],in ahistoricobservationofthePerseuscluster.Thisimageshowedtwodepressionsinthe X-raysurfacebrightnessmap,bracketingthecentralAGN.Early Chandra observationsof nearbyclustersrevealedsimilarpatternsofcavitiesaroundradiosourcesinbrightestcluster galaxies(BCGs)[McNamaraetal.,2000,2001].Thesecavitieshavegenerallybeenfound tobewithradioemission,thoughoftenitistoofainttobedetectedinshallowradio surveysandothershortobservationsat1.4GHz[B^etal.,2012].Long-wavelengthradio observationsanddeeperobservationsofthePerseusclusterbyFabianetal.[2000]revealed olderandlargercavitiesfartheroutinthecluster,bylow-frequencyradioemission presumablyfromanagingpopulationofrelativisticelectrons. Previoustotheseobservations,astronomershadassumedthattherateofkineticenergy emergingfromradioAGNwouldbesimilartotheirradiativeluminosities,thatis,low.The assumptionthattherewasnosourceofenergytocounterbalancetheprodigiousreleaseof radiativeenergyfromthegasledtothenotionofa\coolingw",wheretheentiregas atmosphereslowlycompresses[FabianandNulsen,1977,CowieandBinney,1977,Fabian, 1994b].Thelackofhugereservoirsofcoldgasandtheconstraintsonstarformationrates atleastanorderofmagnitudelowerthanthecoolingwrateputthesimplecoolingw 72 modelintodoubt.Thedoorwasclosedonthismodelbyhighresolutionspectroscopymade bytheGratingSpectrographonboardthe Newton X-rayMultiMirror(XMM) telescope[Petersonetal.,2003].TheseobservationsshowedthatthestrongX-rayemission linesfromgasaround10 7 K( ˘ 0.9keV)predictedbythesimplecoolingwmodelwerenot presentattheirpredictedstrength. Nevertheless,theproblemremained:howcanthegasradiatesobrightlyandsocom- monly(abouthalfofallX-rayluminousclustershavecoolcores)?Thediscoveryofcavities commonlyassociatedwithAGNintheatmospheresofcoolcoresposeda\smoking-gun" answer.Thisscenariowasbolsteredbythediscoverythatthesizeofthecavityandthe X-raygaspressuregthecavitywereconsistentwiththeenergylostbythecluster coolingcore,usingthebuoyantrisetimeastherelevanttimescale[McNamaraetal.,2005, McNamaraandNulsen,2007,B^etal.,2008].TopreventtheICMinacoolcorefrom catastrophicallycoolingitispossibleforthecentralsupermassiveblackholeofthebright- estclustergalaxy(BCG)initsactivegalacticnucleus(AGN)phasetoquenchcooling[e.g BinneyandTabor,1995,Churazovetal.,2001].TheAGNmightaccomplishthisheating throughthecreationofX-raycavities(\bubbles")whichbuoyantlyrisefromthecluster center(e.g.Bruggen[2003],BruggenandKaiser[2002]).B^rzanetal.[2004]foundastrong correlationofAGNjetpowerwiththeX-raycoolingrateinclustersandgroups,withsome scatter.ThiscorrelationsuggeststhatmechanicalenergiesfromX-raycavitiesareinthe rangeof1 pV to16 pV percavityinordertoquenchthecoolingsuggestedbythecentralen- tropyofthecluster.ThecorrelationbetweenAGNjetpowerandX-rayluminositysuggests thatradiobubbleformationscalesstronglywiththeamountofcooling. Theemission-linenebulaeassociatedwiththesecoolcoresprovideanotherdiagnostic forthephysicalprocessesoccurringthere.Theoriginofthegasinthesedusty,optically- 73 luminoustsisnotclear[VoitandDonahue,2011],andtheprocessesthatexcitethe opticalemissionaresimilarlymysterious.Forsomets,photoionizationbyhotstars maywellbethedominantsourceofexcitation,butsomeadditionalsourceofheatmightbe requiredtoexplainthebrightestforbiddenlineemissionsimultaneouslywiththelackofHeII recombinationlines[VoitandDonahue,1997,DonahueandVoit,2004,Ferlandetal.,2009, Sparksetal.,2009,Werneretal.,2013].Nevertheless,theemissionlinenebulaeareproviding tcluestounlockingthecoolcoremystery:theyonlyappearinclusterswithcool cores,thatis,clusterswithlowcentralgasentropy,shortcoolingtimes[Cavagnoloetal., 2009]Coolcoreclusternebulaextendupto70kpcfromtheclustercore[e.g.McNamara etal.,1996].Typically,themorphologyoftheBCG'sH correlateswellwiththemorphology seeninthesoft( < 1keV)X-rayemission[e.g.Sparksetal.,2004,Fabianetal.,2006,Werner etal.,2010]. Inthispaperwepresentobservationstakenwiththe Chandra X-rayObservatoryandthe SouthernAstrophysicalResearch(SOAR)telescopeofthegalaxyclusterRXJ2014.8-2430 (RXJ2014.8-2430),whichisthestrongestcoolcoreclustersinREXCESS.TheREXCESS sampleohringeretal.,2007b]isarepresentativez ˘ 0.1sampleof31clustersspanninga widerangeinluminosity,mass,andtemperature.ItwasdesignedtoavoidbiasinX-ray morphologyorcentralsurfacebrightness.Theseobservationswerefollowedupwithashort (20ks) Chandra observationtocomplementtheXMMdatabystudyingtheclustercore witharcsecondresolution.FromtheXMMdata,Crostonetal.[2008]dRXJ2014.8-2430 isstronglypeakedandhastheshortestcentralcoolingtime(computedwithin0.03r 500 )of thesampleat0.550 0.026Gyrandis15%shorterthanthecentralcoolingtimeforthe nextcoolestcluster.Donahueetal.[2010]determinedanH luminosityof6.4 10 41 h 2 70 ergs 1 ),whichistwiceaslargeasthatofanyotherclusterinREXCESS.Additionally,it 74 hasamoderate(inrelationtootherREXCESSclusterswithradiosources)radiosourcein thecenter[Condonetal.,1998]anditsminimumstarformationratebasedonUVXMM opticalmonitordatauncorrectedfordustextinctionis8-14M yr 1 [Donahueetal.,2010]. Forallcalculations,theassumedcosmologyis H 0 =70kms 1 Mpc 1 , M =0 : 3, =0 : 7.Prattetal.[2009b]obtainedaredshiftestimateofRXJ2014.8-2430fromitsX- rayspectra, z =0 : 1538withnouncertaintyreported.Weusetheredshiftdeterminedfrom opticalspectroscopy,z=0.1555 0.0003,byDonahueetal.[2010].Usingthatupdated redshift,theangularscaleis2.694kpc = 00 andtheluminositydistanceis741.8Mpc[Wright, 2006]. 75 76 Table3.1.Observations TelescopeFilterExposureDateIDPI SOAR/SOICTIO7580/853 1200s2010September6thDonahue CTIO7384/843 720s SOAR/Goodman600l/mmgrating3 1200s2012July25thDonahue 3 1200s Chandra/ACIS-S20ks2009August25th11757Donahue XMM/MOS1+MOS226.7ks2004October8th201902201Boehringer Note.|Summaryofobservationsusedinthiswork. 3.2ObservationsandDataReduction 3.2.1ChandraX-rayObservation The20ksec Chandra observationofRXJ2014.8-2430wastakenonAugust25,2009(ObsID 11757)aspartofthe Chandra GuestObserversprogram.Observationdetailsarelistedin Table3.1.Observationsweretakenwiththe Chandra AdvancedCCDImagingSpectrometer centeredontheback-illuminatedACIS-S3chipinVFAINTmode.Thedatawerereprocessed withCALDB4.4.6andCIAO4.3.Weuseddeepbackgroundscaledtomatchthehigh energy(9.5-12.0keV)backgroundcountratemeasuredinalargesource-freeregioninthe data[HickoxandMarkevitch,2006].Pointsourcesaroundtheclusterwereidendby handandremoved.Toestimatethecentroid,weusedtheprocedurefromCavagnoloetal. [2008a]andwitharelaxedcool-coreclusterwithaslightasymmetryintheclustercore,this methoddeterminedanX-raypeakthatiscoincidentwiththepeakpixel(0.492 00 0.492 00 ) inthecleanandpointsource-subtractedimage.TheX-raycenterisRA20 h 14 m 51.65 s ,Dec -24 d 30 m 21.1 s . Wea2 2binnedcluster0.5-7.0keVmapwithellipticalusingthe ellipse [Jedrzejewski,1987]functioninIRAF[Tody,1993].Weestimatedthecentroidshift w = 1 R max r i < > ) 2 N 1 whereNisthetotalnumberofaperturesconsideredand i isthe separationofthecentroidscomputedwithin R max andwithinthe i th aperture.Usingour ellipticalourcentroidshiftparameter w =0.018 0.508,isstatisticallyconsistent withnocentriodshift.Forcomparison,Maughanetal.[2012]chose w =0.006(withrespect to R 500 )toseparaterelaxedfromunrelaxedclustersintheirsampleof114clusterswith Chandra ACIS-Iobservations.Thiscentroidshiftcutwasfoundtodistinguishwellthecool- core(CC)fromnon-CCclusters,withonly3CCclustersintheirsamplehaving w> 0 : 006. 77 Forourellipsethe R max is74kpc,muchsmallerthanthecluster's R 500 measuredby XMM. Tosearchforstructure,weconstructedanunsharpmaskedimageusingthesameprocess asRandalletal.[2009]wherewetooktheexposure-corrected2 2binnedbroadbandimageof thecluster,smootheditbyaonepixel(0.98 00 )Gaussian,anddividedthatbythesameimage smoothedbyatenpixel(9.8 00 )Gaussian.Theunsharpmaskedimage,seeninFigure3.1, showsacleardecrementofaspiralstructurearoundclustercenterandanexcessatthe clustercenterwithnoobviousX-raycavities. Figure3.1X-rayImageofCluster . Left: Chandra X-rayimageofRXJ2014.8-2430shown onasquarerootscale,roughlyfollowingprojecteddensity.Rightiswestandnorthisup. Right: UnsharpmaskedX-rayimageofRXJ2014.8-2430shownonthesamescaleasthe imageonleft. Wea2Delliptical model I ( r )= I 0 1+ r r 0 3 0 : 5 (3.1) 78 fromthe2 2binnedimageontheenergyrange0.5-7.0keV.Themodelwasin Sherpaandweallowedalltheparameters,includingthecenter,tovary;ngthemodel withtheCashstatistic.Theparametersforthe2D model(with1 ˙ errors)are: r 0 =5.53 00 0.08 00 ,I 0 =83.5 1.6countspixel 2 ,ellipticityof0.098 0.006, =0.998 0.002.The positionangleofthemajoraxisoftheellipsepointsdirectlynorthat-0.8 1.7 .The majoraxisangleisequivalenttoa0 positionanglefortheIRAFellipsewhichagrees overmostofthecluster.Weusetheresidualimagetolookforanyadditionaldeviations fromasmooth model X-rayspectrawereextractedusing specextract intheenergyrange0.3-11.0keVover annulicenteredattheX-raypeakwithatleast2500counts.Analysisisrestrictedtothe ACIS-S2andACIS-S3chips.WeusedtheMewe-Kaastra-Liedahl(MEKAL)model[Mewe etal.,1985,1986,KaastraandMewe,1993,Liedahletal.,1995]forhot,X-rayemitting plasmas,highlyionized,inthermalandionizationequilibriumtothetemperatureand metalabundanceinXSPEC12.6.0[Arnaud,1996].Forconcentricannuli,themetallicity parameterforthecentralbinisindependentandallsubsequentmetallicityparametersare tiedacrosspairsofconsecutiveannuli.ForallspectraltheGalacticforegroundextinction isatN H =7.4 10 20 cm 2 fromDickeyandLockman[1990].TheGalacticforeground columndensityaswellastheGrevesseandSauval[1998]relativesolarabundancesared parametersforthePHABSmodelusedforGalacticextinction.Wetheunbinnedspectral datawiththeCashstatistics[Cash,1979]implementationinXspec(moc-stat). Wetheprojectedtemperatureandmetallicityoftheannuliandtheresultanttem- perature,deprojectedelectrondensitypressureandentropyInthis procedurewehavemadetheapproximationthattheprojectedtemperatureisapproximately equaltothedeprojectedtemperature.Wecalculatetheclusterentropyfromthetempera- 79 tureanddensitythenusethefunctionalformfromDonahueetal.[2006],K(r)=K 0 +K 100 (r/100kpc) ,totheentropyThecentralentropy,K 0 ,foundforthecluster RXJ2014.8-2430is11.6+/-3.9keVcm 2 ,K 100 =159.0 57.2,and =1 : 284 0 : 002( ˜ 2 red =0.725).Inadditontoextractingannuli,wealsoextractedandX-rayspectraofregions onandoftheH ts.Thecounts,temperatues,andmetallicitiesfortheX-ray regionsoverlappingtheH regionsareshowninTable3.2. 80 81 Table3.2.H regionX-rayBoxes RegionRADecBoxSizeNetCountsT X Metallicity (|)(|)(|)(arcsec 2 )(|)(keV)(|) H center20 h 14 m 51.610 s -24 d 30 m 21.24 s 20.0012043.51 +0 : 45 0 : 37 0.91 +0 : 57 0 : 41 H ter120 h 14 m 51.820 s -24 d 30 m 18.51 s 30.9614334.15 +0 : 64 0 : 49 0.37 +0 : 39 0 : 29 H ter220 h 14 m 51.364 s -24 d 30 m 24.63 s 39.2619623.21 +0 : 29 0 : 26 1.10 +0 : 46 0 : 35 H 120 h 14 m 51.391 s -24 d 30 m 16.39 s 28.267474.04 +0 : 87 0 : 64 0.36 +0 : 49 0 : 31 H 220 h 14 m 51.791 s -24 d 30 m 26.81 s 41.1018614.81 +0 : 73 0 : 57 0.50 +0 : 42 0 : 33 H 1+2tied||69.3726084.45 +0 : 48 0 : 41 0.47 +0 : 27 0 : 21 terH 1+2tied||70.2233953.63 +0 : 26 0 : 25 0.98 +0 : 32 0 : 28 terH 1+2+centertied||90.2245993.65 +0 : 23 0 : 20 1.13 +0 : 29 0 : 25 WeusedthesameproceduresofDonahueetal.[2014]totheclusterwiththeJoint AnalysisofClusterObservations(JACO)[Mahdavietal.,2007,2013].FromtheJACO theXMMparametersare:M 2500 =1.76 0.05 10 14 M ,c 2500 =4.8 0.15,fgas 2500 =0.126 0.002,r 2500 =473 5kpc,r 500 =0.94 0.01Mpc.TheChandraparametersare:M 2500 =3.0 0.2 10 14 M ,c 2500 =2.3 0.13,fgas 2500 =0.098 0.003,r 2500 =564 11kpc;r 500 =1.19 0.03Mpc.Theradialprfortemperature,electronpressure,entropy,andelectron densityarepresentedinFigure3.2.Theradialforthemetallicityispresentedin Figure3.3.Uncertaintieslistedare1 ˙ (68%),statisticalonly,errors.Wenotethatthe JACO-ChandraHSEmassestimateforr 500 issomewhatclosertother 500 of1.15Mpcesti- matedforthisclusterbyPrattetal.[2009a]assumingtheM 500 -Y X relationfromArnaud etal.[2007].Thedataarebinnedby25countsperenergybinand ˜ 2 minimizationstatistic isusedfortheJACOThereare24radialbinsfor Chandra and35radialbinsforXMM. JACOdoesn'tdeprojectintheclassicsense,itforward-models,thatisitassumesa3-d parametricmassmodel,and,separately,anon-parametricmodelofconcentricshellsof constant-Tanddensitygas,thenprojectsthepredictedspectrainringsofconcentricannuli [Mahdavietal.,2007,2013]. 3.2.2SOARH ImagingandSpectra NarrowbandopticalimagingwastakenontheSouthernAstrophysicalResearch(SOAR) TelescopewiththeSOAROpticalImager(SOI)[Walkeretal.,2003]onSeptember6,2010. Aseriesofthreeexposureseachwithanexposuretimeof1200secondswastakenwiththe narrowband(7580/85)thatwascenteredontheredshiftedH (7572 A).Asecond setofthreeexposuresof720secondseachwastakenwithanarrowbandcontinuum (7384/84)todeterminecontinuumemissioncontributiontothe\onband"image.We 82 Figure3.2ACCEPTstyle .Weplotthetemperature(projectedforthenonJACO analysislabeled\Chandra"),electrondensity,electronpressure,andentropyforthecluster asafunctionofphysicaldistancefromtheclustercenter.Thehorizontalerrorbardescribes thewidthoftheannuluswheretheXspecmodelwasTheerrorbarsfromJACOare 68%uncertaintiesassumingstatistical-onlyerrorsdeterminedwithMCMCprocedurewhile thenonJACOChandradataare90%uncertainties. 83 Figure3.3JACOMetallicity .WeplottheJACOanalysisforthemetallicityofthecluster usingboththeXMMand Chandra dataasafunctionofphysicaldistancefromthecluster center.ThehorizontalerrorbardescribesthewidthoftheannuluswheretheXspecmodel wasTheerrorbarsfromJACOare68%uncertaintiesassumingstatistical-onlyerrors determinedwithanMCMCprocedure. 84 calibratedthecombinedimagewiththespectrophotometricstarLTT7379[Hamuyetal., 1994].Theimageswerealignedtotheworldcoordinatesystem(WCS)usingstarsfromthe TwoMicronAllSkySurvey(2MASS).Theresultingimagesarealignedtowithin0.2 00 astro- metrictoleranceinrightascensionand0.3 00 toleranceindeclination.Tocorrecttheimage fromforegroundGalacticextinctionweusedanE(B-V)=0.1491andassumedA V /E(B-V) =3.1[Schlegeletal.,1998].WecalculatedatotalH +[NII]luminosityof189 6 10 40 erg s 1 fortheclusterinsideacircularaperturewitharadiusof8 00 (21.5kpc)andcenteredit on20 h 14 m 51.57 s -24 30 0 22.3 00 toavoidapoorlysubtractedstarnearthegalaxy.Wetook abackgroundfroman8 00 radiuscircleofblankskycenteredon20 h 14 m 54.29 s -24 30 0 13.8 00 . ThenetH imageispresentedinFigure3.4withthe Chandra X-raycontoursoverlayed. OpticalspectraoftheBCGweretakenwiththeGoodmanspectrograph[Clemensetal., 2004]onJuly25th,2012.TheGoodmanspectraweretakenwiththe600 ` /mmgrating ( ˘ 2600 Acoverage)centeredon6500 Awiththe1.68 00 wideslit,whichcorrespondstoan approximaterestwavelengthrangeof4510 A 6760 A.WeobservedtheBCGataposition angle110 eastofnorth,alignedwiththeelongationofthecentralH region,andcentered onthebrightestpixeloftheBCG. Wereducedthe2DspectrausingthestandardIRAFspectralreductionroutinesinthe NOAO onedspec and twodspec packages.FeArandquartzlampswereobservedbeforeand aftereachobservation.TheCCDonGoodmanexhibitsseverespectroscopic(multiplicative) fringingfromtheinterferencepatternsofthemonochromaticlight.Thefringingisapprox- imately20%peak-to-peakinwavelengthsbeyond7000 A,withspacingof ˘ 35 Abetween thepeaksofthefringes.Tomakeafringecorrectionframewenormalizedtheoverallre- sponseinthequartztoathirdorderspline.Wedidnotdetectvariationsinthefringe patternbetweenthebeforeandafterquartzes:normalizedfringeframevariationswere < 85 Figure3.4ContinuumsubtractedH imageoftheBCGofthecluster .The Chandra X-ray surfacebrightnesscontours(0.5-7.0keV)areplottedontheSOIH ,continuumsubtracted, image.TheX-raycontoursarecompressedperpendiculartotheH wingsofthe BCG.TheseeinghasaFWHM ˘ 0.8 00 .Thescaleof3 00 isapproximately8kpcforthecluster. Thefeatureintheupperlefthandcornerisaresidualfromanimperfectly-subtractedstar image. 86 0.5%.Wewereabletoreduce20%peak-to-peakfringingdownto2%bydividingbythe normalizedfringeimage.ForwavelengthcalibrationweidenlinesintheFeArlamp (contaminatedwithHelium)spectra.Wevthecentersofthenightskylinesinthe objectframes( 5577,5889,6300,6363,6863,6923,7276,7316,and7340 A)werewithin1 Aofourwavelengthsolution.Wecalibratedthespectrausingthe APALL super-task withobservationsofthespectrophotometricstandardstarLTT9491.Toexamineemission featuresinthespectraweextracted3pixel(0.45 00 )wide1Dspectra.Theresultsofthese spectraareinTable3.3. 87 88 Table3.3.GoodmanSpectralLineFits. SpectralLineRegionCenter a LineCenterEQWFluxFWHM (|)(Pixels)( A)( A)(10 16 ergs 1 cm 2 )( A) H 8005616.82 0.3-24.7 1.72.367 0.168.882 0.9 OIII50075786.24 2.28-7.7 2.90.7731 0.312.44 5.89 NI51996006.91 1.03-10.9 2.20.9945 0.29.565 2.88 OI63007279.13 0.4-22.2 1.51.974 0.149.464 0.99 H 8035616.55 0.38-20.0 1.62.381 0.199.636 0.88 OIII50075784.63 2.05-8.3 2.11.007 0.2613.68 4.03 NI51996006.66 1.25-9.0 1.71.023 0.1910.61 2.61 OI63007279.21 0.58-15.4 1.51.735 0.1711.22 1.29 H 8065616.07 0.45-13.9 1.12.056 0.169.529 0.62 OIII50075784.11 0.94-6.8 1.80.9873 0.2710.95 3.62 NI51996006.35 1.92-7.4 1.50.9811 0.1912.14 2.94 OI63007278.44 0.55-14.0 1.51.76 0.1810.98 1.02 H 8095615.08 0.6-11.6 0.91.974 0.169.94 1.08 OIII50075783.43 0.91-9.3 1.41.546 0.2413.76 2.75 NI51996006.43 1.05-7.2 1.41.075 0.2211.65 2.22 OI63007277.94 0.83-12.8 1.31.83 0.1912.67 1.26 H 8125614.27 0.52-12.3 1.12.264 0.2110.11 0.93 OIII50075782.22 0.73-11.2 1.02.014 0.1913.27 1.9 NI51996005.64 1.53-5.1 0.90.8571 0.159.897 2.94 OI63007276.64 0.75-13.8 1.22.104 0.1813.85 1.35 H 8155613.0 0.59-12.8 0.92.37 0.1610.53 1.05 OIII50075781.22 0.49-12.5 1.22.228 0.2212.23 1.0 NI51996005.62 1.08-5.8 1.70.9444 0.2712.98 4.4 OI63007276.48 0.81-11.5 1.41.714 0.213.53 2.05 H 8185612.06 0.32-14.6 1.22.415 0.219.398 0.93 OIII50075779.17 0.37-13.2 1.02.168 0.168.645 0.89 NI51996004.6 2.76-5.0 1.40.7397 0.213.97 5.03 OI63007275.74 1.63-8.3 1.71.148 0.2313.45 2.92 H 8215611.1 0.28-13.3 1.41.895 0.218.033 1.04 89 Table3.3(cont'd) SpectralLineRegionCenter a LineCenterEQWFluxFWHM (|)(Pixels)( A)( A)(10 16 ergs 1 cm 2 )( A) OIII50075778.52 0.33-12.3 1.11.791 0.166.875 0.76 NI51996005.0 3.3-4.6 2.30.5812 0.312.49 9.32 OI63007275.75 2.5-10.0 1.81.056 0.1918.93 5.98 H 8245611.33 0.63-10.2 1.41.189 0.178.039 1.42 OIII50075778.55 0.45-10.4 1.11.239 0.136.552 0.95 NI51996005.31 3.61-6.2 2.20.6171 0.2215.48 9.44 OI63007274.81 3.82-9.7 3.10.7907 0.2522.96 13.93 H 8275612.19 1.51-8.9 2.30.7779 0.211.85 3.72 OIII50075779.16 1.37-7.2 1.40.6745 0.138.567 1.92 NI51996006.36 7.97-2.9 5.50.2399 0.468.206 10.71 OI63007272.52 15.44-3.3 5.40.2379 0.3815.7 27.14 DeblendedH +NII NII65487787568.33 2.01-33.4 6.70.6347 0.139.687 3.17 H 7586.6 0.71-111.0 16.62.168 0.3313.43 1.89 NII65487817568.33 0.91-27.0 5.60.7375 0.159.106 2.52 H 7585.78 0.39-86.9 9.02.414 0.2510.35 0.88 NII65487847568.54 0.47-31.7 5.71.029 0.189.581 1.94 H 7585.56 0.29-96.4 5.03.102 0.169.341 0.53 NII65487877568.5 1.26-35.2 5.01.237 0.1810.63 2.66 H 7585.2 0.21-110.4 6.53.983 0.248.732 0.59 NII65487907568.14 0.45-45.8 3.91.816 0.159.71 1.12 H 7584.96 0.13-129.2 3.05.236 0.129.063 0.31 NII65487937567.73 0.35-47.6 3.42.293 0.1710.26 0.75 H 7584.68 0.14-131.5 3.56.353 0.179.238 0.31 NII65487967567.25 0.46-39.3 3.22.641 0.2210.3 0.86 H 7584.32 0.09-112.1 2.97.643 0.29.663 0.29 NII65487997567.18 0.38-29.9 2.12.742 0.1910.2 0.63 H 7584.15 0.11-92.1 2.28.255 0.210.12 0.24 NII65488027567.38 0.58-30.5 2.23.098 0.2213.99 1.43 H 7584.17 0.17-86.0 2.48.539 0.2410.77 0.24 90 Table3.3(cont'd) SpectralLineRegionCenter a LineCenterEQWFluxFWHM (|)(Pixels)( A)( A)(10 16 ergs 1 cm 2 )( A) NII65488057566.03 0.68-25.0 3.13.171 0.3913.99 1.37 H 7583.59 0.19-67.8 2.18.501 0.2712.53 0.5 NII65488087565.3 1.29-26.7 4.13.737 0.5817.35 2.15 H 7582.56 0.28-67.0 3.89.128 0.5213.66 0.66 NII65488117563.2 1.25-22.8 2.23.621 0.3516.26 2.14 H 7581.34 0.35-70.2 3.210.83 0.4915.16 0.33 NII65488147560.82 0.42-8.7 2.11.327 0.328.786 1.3 H 7579.15 0.23-59.6 2.08.754 0.314.39 0.48 NII65488177558.29 0.79-7.8 2.21.038 0.298.08 2.12 H 7576.12 0.1-33.8 2.14.387 0.276.527 0.3 NII65488207558.12 0.59-9.6 2.51.12 0.298.056 1.92 H 7575.67 0.08-38.2 2.34.319 0.265.815 0.28 NII65488237558.58 1.02-13.4 4.91.167 0.4311.34 2.98 H 7575.63 0.19-34.5 4.92.957 0.426.483 0.67 NII65488267557.72 2.0-5.6 1.90.4371 0.158.918 3.99 H 7575.72 0.5-11.6 3.90.8754 0.295.318 1.46 a Thetableisorganizedby0 00 : 45regionstartingwiththeeasternmostregion. Allandequivalentwidths,computedusing splot ,arecalculatedintheobserver's framesuchthatwecanalsoestimateredshiftsineachofthelinestotrackvariationofthe velocitiesoftelementsinthecluster.WeuseaGaussiantoeachofthe linesandthe splot bootstrapresampling(100realizations)methodtocomputeerrorsonthe GaussianWeestimatedourbackgroundandcontinuumsubtractionfromalinear oftwocontinuumpoints ˇ 20 Afromtheouteredgesofeachmeasuredemissionline.The erroronthecontinuumwasfromtherootmeansquarefromanemission-freeregion.We thisat ˙ 0 =4.196 10 18 ergcm 2 s 1 A 1 overtherange5600-7600 A. WeestimatedtheredshiftofthestellaremissionbyFouriercross-correlatingthecontin- uumemissioninRXJ2014.8-2430tothespectrumofanellipticalgalaxySDSSJ120028.87- 000724.8(z=0.0813 0.0002).Usingthe fxcor taskinIRAF,weshiftedtheSDSSspectrum tothebaselineestimatedredshift(0.1555)ofRXJ2014.8-2430.WebinnedtheSDSSspec- trumtomatchthelowerspectralresolutionSOARspectrum.Weextracteda40pixelwide spectrumforRXJ2014.8-2430centeredonournominalcenter.Thecorrelationresultwas basedonanemission-line-freerangebetween6100-7100 Afromstellarabsorptionlinesand theerrorisbasedonresultsfrom1000randomlyselectedre-sampledsections200 Awide. Wendavelocityshifttothebaselineestimateof,-10.57kms 1 29.10kms 1 ,whichis statisticallyconsistentwithzeroshiftfromthenominalemissionlineredshift. 3.2.3X-rayAGNLimits WeplaceanupperlimitontheX-rayofanAGNpointsourceusingtheCIAOtool celldetect .Theeventswererestrictedtothe0 : 5 7 : 0keVenergyrangeandthesearchwas limitedtocellsizesof1pixeland3pixelsneartheclustercenter.Thelocalcluster backgroundwasestimated(13.67 4.54countspixels 2 )froma5 00 circlenearthecenter 91 ofthecluster.Wedidnotany < 3 ˙ detectionsofcompactsourceswithin10 00 ofthe clustercenter.UsingtheCIAO aprates taskwecalculateda3 ˙ upperlimitonthecounts expectedonanAGNpointsource.Thealgorithm 1 determinesintervalsbasedon aBayesianbackground-marginalizedposteriorprobabilitydistributionfunctionofpossible sourcecounts.Weassumethepossiblepointsourcewillnearlyhaveallofitsina singlecellandthebackgroundisthemeanvaluecomputedwithin5 00 ofthecenter.Our3 ˙ upperlimitof17countsforapointsourceabovethelocalbackgroundofextendedcluster emissioninthecoreofthecluster,correspondingtoalimitof < 1.55 10 14 ergcm 2 s 1 (at E=2keV),correspondingtoan0.5-7.0keVX-rayluminosityof < 5.39 10 42 ergs 1 for apower-law( =1)pointsourceattheclusterredshift. 3.3Discussion 3.3.1RadioBubbleLimits Giventhecentralentropyoftheclusterandtheexistenceofastrongcentralradiosource, theapparentlackofabubblemaybesurprising.Theadditionallackofradiolobeswould suggestthatanypotentialcavitiesmaybesmallandneworalongthelineofsight.Inorder toestimatethesizeofabubbleexpectedinsuchasource,weusedtherelationiny etal.[2006].Asinyetal.[2006],wetheX-raycoolingradiustobethatwithin whichthegashasacoolingtimelessthan7.7 10 9 years,whichisthecosmictimeelapsed since z =1tothepresentepochforthecosmologyadoptedinthispaper.yetal. [2006]considerthiscoolingtimerepresentativeofthetimeithastakentheclustertorelax 1 Thedetailsarefoundin http://cxc.harvard.edu/csc/memos/files/Kashyap_ xraysrc.pdf . 92 andestablishacoolingw.WeusethesameXspecmodelinSection2andincludea withtheMKCFLOWmodeladdedandthelowtemperatureto0.1keV.Similartothe modelingfortheMEKALmodelthemetallicityistiedacrosspairsofannuli,outsideof thecentralregion.Thehightemperature,metallicity,andtemperaturefortheMKCFLOW modelaretiedtothesimultaneouslyMEKALmodel.However,outsideofthecooling radiustheMKCFLOWnormalizationissettozero.Wecalculatetheluminosityofeach annulusafterusing lumin inXspecintheextrapolatedrangeof0.1to100keVto estimatethebolometricluminosity.WeestimatethebolometricluminosityL bol =19 : 7 10 44 ergs 1 intherange0.11555-100.0keVrestframeoverthetotalMEKALmodelinsideall annuli(675kpc).Thisluminosityisreasonablyconsistentwiththebolometricluminosity calculatedfromXMM(L X =21 : 06 0 : 07 10 44 ergs 1 within R 500 =1155.3 4kpc) whichwasoveraslightlylargerarea.UsingthetechniquefromB^etal.[2004]we thecoolingtimeforeachoftheannuliusing t cool =3 nkT X = 2 n e n H T;Z ),where T;Z ) istheX-rayemissivityasafunctionoftemperatureandmetallicity.Wesolvefor T;Z ) assumedbytheMEKALmodelusingthenormalizationfromtheMEKALmodelandthe bolometricluminosity.Weassumeafullyionizedplasmasuchthatthetotalnumberdensity n=2 : 3 n H .Wecomputethecoolingradiusatadistancefromtheclustercentroidsuchthat thecoolingtimeislessthan7.7 10 9 years.Forthisclusterthecoolingradiusis105kpc (39 00 ).InsidethisaperturewehavealuminosityL X ( 10kpcfromthegalaxyclusterpotentialminimumapproximately every500Myrs.Also,theythatinthepresenceofviscosityaswellastheadditionof alargeBCGpotential(theirFigure26)intheclustercorewilldecreasethecoreheating expectedfromsloshing.Basedonthesuggestionsofthesesimulations,wemaybeobserving RXJ2014.8-2430duringaspecial,butnotnecessarilyearly,periodofaminorsubcluster merger.Alternatively,arecentsloshingeventintheclustercorecouldgivethestructure seeninthecore,butthesloshingwastoogentletodispersethemetalsinthecore. 3.3.4VelocitystructureintheBCGOpticalEmissionLines TheH +[NII]isbrighteroveralargerareasowewereabletoextract17regionsof3 pixelseachfromthe2Dcalibratedspectralimagediscussedearlier.FortheH +[NII] complex,wedeblendthefeaturesbysimultaneouslytheH linewiththetwo[NII] linesusing deblend insplot.Howeverthe[NII] 6584iscontaminatedbyabrightskyline and,similartoDonahueetal.[2010]weusethelinemeasurementsforthe[NII] 6548 lineandmultiplybyafactorofthree(theconstant[NII]lineratiobyatomicphysics) todeterminethe[NII] 6584valueforcomputingabundanceratios.Usingthesevalues, theratios,likeratiosofsimilarregionsinnearbyclustersofgalaxies(e.g.Heckmanetal. [1989]),fallintothelowerrightsideoftheBPTdiagram[Baldwinetal.,1981]whichisused todiagnosethebetweenionizationbyhotstarsandaLowIonizationNuclear EmissionRegion(LINER).Wenote,however,thatunlikeaLINER,whichisunresolved point-like,thisemissionlineregionisextendedandunlikelytobeheatedbytheradiation comingfromanAGN,basedonargumentssimilartothosepresentedinHeckmanetal. 99 [1989]:alackofionizationgradientthatwouldindicateacentralionizationsource,presence ofextendedemissionwithnearlyconstantlineratios,relativelyconstantvelocitywidthsof fairlymodestwidth.Fromasampleofbrightestclusterandbrightestgroupgalaxiesinthe SloanDigitalSkySurvey,vonderLindenetal.[2007]foundthatmosthaveemissionline ratioswhichplacethegalaxiesontotheLINERregionoftheBPTdiagram.The[OIII]/H ratioisacrossthecluster,butthe[NII]/H ratio,inFigure3.7,dipsdownbyafactorof twoatthecenteroftheBCG.ThisincreaseintheH to[NII]ratiocouldindicateaweak radiosourceinthecore,inlinewiththeideathatthe\young"radiosourceintheclusteris atanearlytimeinitskineticenergyoutput. Figure3.7GoodmanSpectraVelocities .WeplotthevelocitywidthsoftheH ,whichhave thecomponentofinstrumentalvelocityremovedinquadrature.Totherightofcenter,the velocitywidthsareunresolvedupperlimitswithwidthslessthantheinstrumentalvelocity. FortheextractedX-rayspectraintheregionsonandH ,presentedinTable3.2the regionswiththeH emissionaresurroundedbyX-rayemittinggasthatisslightlycooler 100 andthathotgashasahighermetallicitythanthoseregionsthatarenotcoincidentwiththe H wings.Ingeneral,weseeinFigure3.3thattheclusterhasastrongmetallicitygradient andametalrichcoreintheX-raygas.Unlikewhathasbeenseeninsomeotherclusters (e.g.Sparksetal.[2004]),thestructureinthesoft( < 1keV)X-ray,whichhasamostly symmetriccoretothebroadbandX-rayimage,doesnotalignwiththeH structureofan elongatedcorewithperpendicularwings. Wecalculatedaninstrumentalvelocitywidthof294kms 1 basedonthewidthofun- saturatedFeArlinesinthecalibrationspectra.Theinstrumentalvelocityweissimilar totheexpectedspectralresolutionforourinstrumentsetup 2 .Toestimatethetruevelocity width,wesubtractedtheinstrumentalvelocityfromtheobservedvelocitywidthinquadra- ture.Wecomparedthespatialpositionofthecontinuumemissiontothespatialposition oftheH byextracting25 AregionsaroundthecenteroftheH aswellasanequivalent widthareaofemissionbluewardoftheH +[NII]complex.ThecontinuumandH have asimilarpeak,within0 00 .3of812pixelswhichwasthenominalcenterpointing.Whilethe emissionintheH isnotasspatiallysymmetricasthecontinuumlight,wedoseethatthe peakoftheH emissionandthecontinuumemissionareco-locatedwiththeX-raypeakto within1 00 . InFigure3.8wecomparetheredshiftsoftemissionandabsorptionlines.Relative tothecenter,thereisavelocitygradientalongthecentralH .Wecomparethattothe 90%interval(0.1531 0.0017)onafortheredshiftfromthe Chandra spectra usingthesameXspecmodelsbutfreeingtheredshiftparameterwhilerequiringittobethe sameforallclusterspectra.Giventhepatternsinthevelocitystructure,thegasislikely infallingorwing.Theopticalvelocitystructuredoesn'tlooklikearotatingdisksince 2 http://www.goodman-spectrograph.org/observers.html 101 itisone-sidedandthediskwouldhavetobeoterfromthepeakoftheemission,even thoughtheemissionpeaksarealigned.However,wecan'ttellwhetherthevelocitygradient isindicativeofgasinfallorw. Figure3.8GoodmanSpectraVelocities .Redshiftsandequivalentlineofsightvelocitieswith respecttotheredshift0.1555.Thegreendatapointswithcirclesinthepointsareaverages fromH ,[OIII]5007,[NI]5199,[OI]6300.ThebluelineistheaveragefromH and[NII] 6548.Thegraybandmarksthe90%rangefortheChandraredshiftfromusing thesameXspecmodelsbutfreeingtheredshiftparameterwhilerequiringittobethesame forallclusterspectra.Thenarrowredbandmarksthecrosscorrelationcenterforthestellar absorptionspectra. VoitandDonahue[2011]estimatethatthestellarmasslossofstarsinBCGsmaybeas highas8M peryear,whichisatleastaslargeasstarformationratesinmostBCGs.This gasfromstellarmasslossisalsopredictedtoremaincoolandmaybeasourceofemission linegas,suchthatsomeoftheemissionlinegasseenintheBCGofRXJ2014.8-2430could originatefromstellarmassloss.Whilethistheorypredictstheemissionlinegaswouldhavea similarvelocitytothestarsintheBCG,sloshingintheclustercorecouldtherelative motionsofthegasandstarsintheBCG.Inthepresenceofsloshingcausedbyaminor mergerthroughthecenterofthecluster,theISMoftheBCGwouldbebythe interactionsintheICMmorethanthestars.Therefore,theemissionlinegasfromstellar 102 masslossmayachieveaditvelocitydistributionthanthestellarvelocitydistribution. Withavelocitygradientthatspansalmost 200kms 1 andspanstherangeofthebt meanvelocitiesofboththeICMandtheBCGstars,thetruesloshingspeedislikelyhigh. Weareonlymeasuringradialvelocitieswhilethesloshingsignaturesweobserveareinthe planeofthesky,thereforethetrue,threedimensionalsloshingspeedislikelymuchhigher thantheradialvelocity. 3.4Summary Weconductamulti-bandanalysisofthecoolcoreclusterRXJ2014.8-2430.Priorobserva- tionsshowthisclusterisastrongcoolcoreandalsodemonstratesstrongemissioninother wavelengthregimes.Thedecrementsweseeinthe modelaswellastheunsharpmask imageareindicativeofapossiblepairofcavitiesjustfromourlineofsightaswell assloshinginthecore.InmodelspresentedbyZuHoneetal.[2010]itislikelythatthe sloshingintheclusterhasonlybegunrecently.Ontheotherhand,weseeaverystrong metallicitypeakinthecoreoftheclustersuchthatthemetallicityinthecentraltiedbinis supersolarwhichisinconsistentwiththesuggestionthatsloshingwilltransportmetalsfrom thecenteroftheclustertotheoutskirts,elytheofthemetallicity [Simionescuetal.,2010,dePlaaetal.,2010].However,itispossibletodecreasethe ofsloshingwithadditionalviscosityinthecoreaswellaslargerpotentialfromthe inclusionofamassiveBCG.Ifthesloshingdistanceatthispointisnotmuchlargerthan thecouplebinsanyofthesepossiblectswouldbecontained. Weseeevidenceforsloshingintheclustercorewhich,givenitslowcentralentropy,may beobscuringbubbles.Theeast-westsloshingcompressionisinthesameorientationasthe 103 elongationofthecentralH andthefactthatthenorthandsouthwingsoftheH areboth behindthebrightestknotinthecenteroftheH ,nearthecentroidoftheX-raysbutahead oftheX-raypeak,suggeststhattheX-raygasismovingpastthegalaxyorthegalaxyis movingthroughtheX-raygas.Thereisatvelocitygradientalongtheelongated centralH regionandthisemissionlinegasiseitherfallingintoorgettingpulledthrough thegalaxy. WedonotevidenceforanX-rayAGN,basedonthelackofapointsourceintheX- rayimagesconsistentwithwhatwasfoundbyDonahueetal.[2010]intheUV.Ouranalysis oftheopticalline-emissionresults,andthebrightcentralsourceintheradiosuggestaweak AGN. FromtheexpectationofourX-raycoolingwewouldexpectatpairof bubbleswhichcouldbewashedoutbythesloshinginthecorebutwemayalsobeina uniquepositioninthiscluster'sevolutionwheretheradiosourcemightjustbeturningon andonlystartingtocreateX-raycavities. Lookingattheopticalemission,weseethattheH iselongatedinthesamedirection theclusteristhoughttobesloshing.Fromtheopticalspectraweseeagradientof ˘ 400 kms 1 acrossthecentralellipsoidalH region.Thereisavelocitygradientthatindicatesa likelihoodtheemissionlinegasiseitherinfallingorgettingpulledoutoftheBCG,depending onwhichsideofthegalaxywearelookingfrom.TherearealsoH \wings"whicharenorth andsouthofthecentralH gas,pointingawayfromthecenteroftheemission. WewouldalsoliketothanktheCIAO/SherpaX-rayschoolandtechsfortheirassistance duringthe Chandra SummerSchoolaswellasthroughtheCXOhelpdesk.Wealsoacknowl- edgetheSOARoperatorsDanielMaturana,PatricioUgarte,SergioPizarro,andAlberto Pastenwhohelpedusduringournights.SupportforthisworkwasprovidedbytheNa- 104 tionalAeronauticsandSpaceAdministrationthroughChandraAwardNumberGO0-11018X (MSURC065171)issuedbytheChandraX-rayObservatoryCenter,whichisoperatedby theSmithsonianAstrophysicalObservatoryforandonbehalfoftheNationalAeronautics SpaceAdministrationundercontractNAS8-03060. 105 Chapter4 PolarizationPilotProjectforthe SOARTelescope 4.1Introduction Electromagneticradiationconsistsofelectricandmagneticcomponentsperpendicular tothedirectionofmotion.Intheplaneperpendiculartothedirectionofmotion,theelectric vectoristypicallyorientedatarbitraryangles,however,itispossiblefortheelectric vectortohaveapreferredorientation.Thesituationwhereweobservelightwithan electricvectorwithapreferredorientationisknownaspolarization. Whiletherearemanytypesofastrophysicalsourceswhichmaybepolarized,theirsources ofpolarizationfallintotwocategories:intrinsicpolarizationoflightandpolarizationof lightinducedbyexternalmedia.Alightsourceisintrinsicallypolarizedifthelightisbeing generatedbyaprocesswhichcreatesanelectricwithapreferredorientation.Onthe otherhand,alightsource,whichproducesrandomelectricanglescanappeartothe observeraspolarizedifthelightisscatteredorpassesthroughcertainmaterials. 106 4.1.1SourcesofAstrophysicalPolarization Manypolarizationstudiesareconductedinthenear-tofar-infraredbecauselightatthese wavelengthsarenotasattenuatedbydust,whichpreferentiallyabsorbsandscattershigher energyphotonssuchasstarlight.Infraredlightcangetthrough,butcarrywithitthe polarizationsignatureofhavingbeenscatteredbyaligneddustgrains.Youngstellarsources, whichareenshroudedbygasanddustrichnebula,suchasTTauriandWolfRayetstars,can showtamountsofpolarization( > 20%)[Bastien,1982].Evenweaklymagnetized regionswithchargedparticlesandchargeddustgrainsmayexhibitstrongpolarizationif randomthermalmotionsarettoscramblethealignmentofthegrains[Davisand Greenstein,1951].Atlongerwavelengths(sub-millimetertoradio)polarizationofthelight fromcoldermoleculargasaswellaspolarizationofthelightfromactivegalacticnuclei (AGN)canbeobserved[AngelandStockman,1980].InradioobservationsofAGN,the power-lawspectrumcharacteristicsofsynchrotronradiationisobservedfromhighenergy particlesspirallingalongthemagneticlinesoftheAGNjet.Thesejetsofintrinsically polarizedradioemissioncanextendseveralkiloparsecsinlength. Manyprocessesalsoproducepolarizedlightinoptical( ˘ 4000 8000 A)wavelengths. Thepolarizationofemissionofdustobscuringandscatteringlightaroundyoungstarscan easilybemeasuredintheopticalaswell.TheopticalsynchrotronfromAGNjetsaswell assupernovaremnantsarealsoobservedopticallyandshowasimilarlystrongpolarization signal(albeit,theorientationoftheopticalandradiosynchrotronmaybet)[e.g. Perlmanetal.,1999].Saturatedthermalconduction[CowieandMcKee,1977]inanionized plasma,whichisseeninsolar[Henouxetal.,1983a]hasintrinsicallypolarizedemission. Theprocessofsaturatedthermalconduction,whichwewilldiscussinSection4.5.1,has 107 beensuggestedasasourceofopticalemissionlinetsofbrightestclustergalaxies. Therefore,ifsaturatedconductionispresentintheemissionlinetsofBCGswecan expecttheemissiontobepolarized. 4.1.2ObservingAstrophysicalPolarization Therearethreecommontypesofsetupsfordeterminingopticalpolarization.Onetypeof polarizer,shownintheupperpanelofFigure4.1,isarotatingwaveplatepolarimeterwhich hasarotatingretarderwithalinearpolarizer.Theintensityismeasuredasafunction ofthepositionangleandthepositionoftheretarder.Sincethereisasmallnumberofcom- ponents,calibrationiseasyanditispossibletodeterminetheangleofmaximalpolarization tohighprecision.However,tomaximizethepolarizationsignalthroughrotation,thesese- tupsmustfocusonsingleobjectsinsmallelds.Thesecondtype,showninthelowerpanel inFigure4.1usespolarizersandbeamsplitterstosimultaneouslymeasureseparateStokes parameters.TheabilitytomeasureStokesparameterssimultaneouslytlyreduces thepolarizationvariationcausedbyvariabilityintheskyandweatheracrossobservations; howeveritisamoreinvolvedsetup,requiresobservationofbrigherobject,andadditional calibrationisneededbetweenseparatedetectors. Thethirdmethod,employedbytelescopessuchastheHubbleSpaceTelescope(HST) andnowtheSOARTelescope,istheuseofasetofpolarizingTheHSThasasetof threeconstructedfrompolarizedmatanglesof0 ,60 ,and120 .Asystemwith polarizationismuchcheaperandmuchmoreportablethansystemswhichrequiresro- tatingretardersorbeamsplitters.However,whilethesystemworkswellinspace,there areaddedcomplicationsforgroundbasedobservations.Duetovariationsinthesky,taking observationsthataremorespreadoutintimecanleadtolessconsistentresults.Variations 108 intheskycanbemitigatedbyusingasetoffourat45 incrementsbecausetheStokes parameters(Q,U,andI)canbedeterminedfromtwoseparatepairsofobservations,unlike athreesystem,whichrequiresallthreetomeasureeachStokesparameters.We presentoursetupforSOIinFigure4.2. 4.1.3CoolCoreClustersandPolarization Somebrightestclustergalaxies(BCGs)incoolcoreclustershaveshownntemission ofcool( < 10 4 K)gasts,butthesourcesofsuchemissionarestillunderdebate. Additionally,deepobservationsbytheChandraX-raytelescopehaveshownthatsomeofthe softX-ray( < 1keV)emissionintheintraclustermedium(ICM)correlateswiththeoptical H tsintheseclustersfurthersupportingtheideathattheH temissionis powered,inpart,bytheX-raygas.OnelargesourceofH emissionisphotoionizationof hydrogenbyrecently-formedmassivestars.H iscorrelatedwithothersignaturesofstar formation,suchasexcessUVlightandexcessfarinfraredlightfromdustheatedbyhot stars.Inrecentyears,thereisaconsensusthatsomeamountofthermalconductioncan reproducetheobservationsseeninBCGsinsomecoolcoreclusters.Thermalconduction istheprocessbywhichheatistransferredthroughparticlecollisionsduetoatemperature gradient.TheinterfacebetweentheICM(10 7 K)andtheintragalacticmediumH ts ( < 10 4 K)isastrongtemperaturegradientwhichwouldbeanobviouscandidateforthis typeofinteraction[Sparksetal.,1989a,Ferlandetal.,2009]. Therehavebeenpreviousstudiesinsolarphysicswhichobservepolarizedemissionin regionswithstrongtemperaturegradientswheretheprocessofthermalconductionismost t.Polarizationlevelsforsolarareemission,ahotX-raygassimilarintemperatureto 1 http://www.ctio.noao.edu/soar/ 109 Figure4.1PolarimeterDesigns .Theupperisastandardsinglebeamrotatingretarder (typicallyahalfwaveplate)withlinearpolarizer.Thelowerureisabeamsplitter designwhichhasfourdetectorstomeasurefullStokesparameterssimultaneously[Sheehan etal.,2010]. 110 Figure4.2SOIPolarimetrySetup .ThesetupoftheSOIinstrument.Thepolarizersare placedinthewheelclosesttothesky.ThetelescopefeedsSOIsittingatNasmyth focus.Variationsintheskycanbemitigatedbyusingasetoffourat45 increments becausetheStokesparameters(Q,U,andI)canbedeterminedastwoseparatepairs,unlikea threesystemwhichrequiresallthreetomeasureeachStokesparameters.Image providedbytheSOARTelescope 1 . 111 theICM,indicatelevelsupto20-30%[Henouxetal.,1983a].Inparticular,polarizationof H emissionlines,giveeasilydetectablepolarizationfractionsof > 5%intheregimewhere conductionissaturated[Henouxetal.,1983b]. NearbyBCGslikeM87(intheVirgoCluster)havebrightH tswhichcanbe easilyobservedwiththeSOARTelescopeandareexcellentcandidatesformeasuringpos- siblepolarizationingalaxyclusterenvironments.Polarizationmeasurements,orlimitson polarizationonthesenearbyBCGswillhelpbetterconstrainthephysicalprocesses whicharerelevantinthecoresofgalaxyclusters. InSection2,wediscusstheobservationsaswellasintroducethehardwareusedinthe newpolarimetrymode.InSection3wepresentthecalibrationanddataanalysis.We takebothstandardinternalCCDcalibrations(biasframesanddomeaswellason skycalibrationsofpolarizedandunpolarizedsources.Theunpolarizedsourcesareused tocharacterizethezerolevelpolarizationcorrectionbetweenthewhilethepolarized sourcesareusedtomeasuretheprecisionofthepolarizationfractionandangle.Herewe alsodiscussthevariabilitywesawbetweennightsandexplainhowwecanobtainreliable polarimetryovermultiplenightsandthemethodweusedtoobservationswhichlikely hadunreliablephotometry.InSection4wepresentoursciencetargets,brightestcluster galaxies(BCGs)withH ts,andsetupperlimitsontheamountofpolarizationin thesents.InSection5wediscusstheimplicationsforthelowlevelsofpolarization intheH tsinBCGs,whichcanhelpplacelimitsontheconductionseeninthe BCGsaswellaswhatwemaybeabletoexpectfrommagneticinBCGts.We concludethepaperinSection6. 112 4.2Observations TheSouthernAstrophysicalResearchTelescope(SOAR)isa4.1maperturetelescopelocated onCerroPaconatanaltitudeof2,700meters(8,775feet)abovesealevel,atthewestern edgeofthepeaksoftheChileanAndes.TheSOAROpticalImager(SOI)iscomposedoftwo 2048pixel 4096pixeldetectors.Theofviewis5 0 5 0 (whichcorrespondstoaphysical sizeof10.16cm 10.16cm)withaphysicalchipgap7.8 00 widebetweenthetwodetectors. Targetswereinallexposures,fromthechipgap.Weditheredthetargettositonat leastthreelocationsinagivenobservingsequence.Wedidnotfromthecenterofthe ofviewforourobservationsofglobularclustersbecausetheglobularclusterscovered thewholeofview.However,evenfortheseobservations,wemadesmall(5-10 00 )dithers betweeneachofthreeobservationstominimizesmall-scaleaterrorsandthe ofbadpixels.Ifpartoftheemission(e.g.extendedgalaxyemission)wasinthechipgap, wemadesuretoditheratleast10 00 eastorwesttopreventthoselocationsfromendingup inthechipgapforall3observations.ForSOI,areinstalledintwowheelswhich holduptoeineachwheel.WeusedthestandardSOIpixelbinningof2 2because thebinnedpixelsize(0.15 00 pixel 1 )isstillmuchsmallerthanthenaturalseeing(0.7-1 00 ,so thepointspreadfunction(PSF)remainswellsampled. Infall2010,WilliamSparksofSTScI,throughagrantfromtheDirector'sDiscretionary ResearchFundatthatsameinstitution,purchasedasetoffourlinearpolarizing(0 , 45 ,90 ,135 )tointroduceapolarimetrymodeforSOI.The0 and90 arecut fromthesamepieceofglassasarethe45 and135 Sincetherearetwo (i.e.apolarizerandanarrowbandorbroadbandinthelightpathinsteadofone, thetelescopefocusistthanthefocuswithasingleTypically,thefocus 113 isabout-100 20unitsfromthefocussettingforasinglewhenapolarizeris inplace.Thefocuswasregularlymonitoredthroughoutthenightandtypicallyre-tuned twiceduringthehalfofthenightwhilethetemperatureinsidethedomechanged morerapidly,andathirdtimeduringthesecondhalfofthenight.PriortoourNovember 23 rd 2011night,wedidnotdetermineaseparatefocuswhichcausedimageswhichare slightlyoutoffocus.Thefocusingerrorwasoriginallymissedastheseeingduringthose nightswaspoor( > 1 00 )whichmadeittodeterminethatafocuswasneeded. Datausedinthischapterwereacquiredduringthreeobservingnights,onlyoneofwhichis priortoNovember23 rd 2011.Thedatesoftheindividualobservations,selection,and exposuretimesarelistedinTable4.1.Sincetheexperimentisamatterofdeterminingsmall betweenpairedobservations,observationswerereducedandanalyzedseparately foreachnighttoindependentlyverifypolarizationmeasurementsandavoidbeing bytchangesinobservingconditions. 114 115 Table4.1.PolarizationObservations. Target (2000) (2000)FilterExposureTime a NightObserved (|)(|)(|)(|)(seconds)(|) UnpolarizedStandards NGC185105 h 14 m 06.7 s -40 02 0 48 00 V30 3January4 th ,2013 R30 3January4 th ,2013 V60 3April6 th ,2013 R60 3April6 th ,2013 NGC186605 h 13 m 39.1 s -65 27 0 56 00 V60 3April6 th ,2013 R60 3April6 th ,2013 PolarizedStandards HD11098412 h 46 m 44.83 s -61 11 0 11.58 00 V2 3April7 th ,2011 V2 3April6 th ,2013 CrabNebula05 h 34 m 32.0 s +22 00 0 52 00 V120 3January4 th ,2013 RMonocerotis06 h 39 m 09.95 s +08 44 0 09.7 00 V60 3January4 th ,2013 BCGScienceTargets M8712 h 30 m 49.4 s +12 23 0 28 00 6600-75 b 600 2April7 th ,2011 6129-140 c 600 2April7 th ,2011 6600-75 b 600 2April6 th ,2013 a ExposuretimesareintheformN M,whereNisthetimeinsecondsforeachexposureandMisthenumberofdithers. b ThiswasusedforanarrowbandH forthistarget. c Thiswasusedforanarrowbandcontinuumforthistarget. Thefourpolarizedwereplacedinthewheelclosesttothesky.The wheelwiththefourpolarizershadoneopenpositiontopreservetheabilitytotake observationswithoutapolarizer.Thesecondwheelwasusedforuptoenarrowor broadbandWeusedthebroadbandBessellVandRregularlyavailableonSOI aswellasadditionalfromtheCTIOnarrowbandcollection 2 .Foragivenscience andditherposition,exposuresweretakeninaconsistentorder(0 ,90 ,45 ,then135 ). Additionally,allobservationsforapositionweretakenbeforeditheringtothenextposition. Forexample,agalaxyrequiringaH andacontinuumhadaseriesoffourimages takenwiththeH andeachofthepolarizers,thenthecontinuumterwitheachofthe polarizers,andthenmovedtothenextditherpositiontorepeatthesequence.Thepaired observationsthroughpolarizationsgby90degreesweretakenascloseaspossible intimesequencetoreducethelikelihoodofintermittentweatheringthemeasurement oftheStokesparameters.Statisticallytvariationsintheofunpolarizedstars isanindicationofclouds;theseobservationswerenotusedinouranalysis.Additionally, Stokesparametersweredeterminedindependentlyfromeachpairofobservationsandcould becomparedtoStokesparametersderivedfromsubsequentdithersets.Ifonepairwas tlybycloudsitwasrejectedfromanalysis. 4.3CalibrationandDataAnalysis 4.3.1StokesParameters WeusedStokesparameterstocalculatethepolarizationpercentageaswellasdirection oftheelectricvector.ThelinearpolarizationvectordrawingsinFigure4.3aswell 2 http://www.ctio.noao.edu/instrumen 116 asthefollowingpolarizationequationsarereproducedfromCollett[2005].S LHP ,S LVP , S L +45 P ,S L 45 P areequivalenttotheimagestakenwith0,90,45,and135degreepolarizers, respectively.Usingthesefourimageswecancalculatethefollowingquantities: Q = S LHP S LVP 2 (4.1) U = S L +45 P S L 45 P 2 (4.2) I 0 = S LHP + S LVP + S L +45 P + S L 45 P 2 : (4.3) Withthefourimages,theStokesparametersQandUcanbemeasuredindependently. Similarly,independentmeasurementsof I 0 canbemadefrom: I 0+90 = S LHP + S LVP (4.4) I 45+135 = S L +45 P + S L 45 P : (4.5) Thiscomparisonisimportanttotestwhetherchangesintheobservingconditionshap- penedonashortenoughtimescaletotherelativefromthesourcebetweentheset offourimages,independentofthesourcespolarization.FromQ,U,and I 0 thepolarization degreePaswellasthepolarizationangle are: P = Q 2 + U 2 I 0 (4.6) 117 Q = I cos(2 )(4.7) U = I sin(2 )(4.8) =1 = 2tan 1 (U = Q) : (4.9) Sincetheinversetangentfunctionisboundedbetween(- ˇ /2, ˇ /2),therelativesigns oftheStokesparametersQandUcanbeusedtoconvertthecalculatedanglestoangles between0 and180 .Wealsocomputethestandarddeviationsforeachoftheseparameters byusingstandardpropagationoferror: ˙ Q = q ˙ 2 S LHP + ˙ 2 S LVP 2 (4.10) ˙ U = q ˙ 2 S L +45 + ˙ 2 S L 45 2 (4.11) ˙ I 0 = q ˙ 2 Q + ˙ 2 U 2 (4.12) ˙ P = q ( Q˙ Q ) 2 +( U˙ U ) 2 +( ˙ I 0 P 2 I 0 ) 2 PI 2 0 : (4.13) Additionallywecanmeasurethestandarderroronthemeanofeachofthesequantities usingtheform ˙= p n asallStokesparametersandregionsaretakenoveridenticalareas.For 118 Figure4.3StokesParameters .ThefourlinearpolarizationStokesparametersareshown. S LHP ,S LVP ,S L +45 P ,S L 45 P areequivalenttotheimagestakenwith0,90,45,and 135degreepolarizers,respectively.UsingthesefourimageswecancalculatetheStokes parametersQ,U,andI 0 .Where Q = S LHP S LVP 2 , U = S L +45 P S L 45 P 2 ,and I 0 = S LHP + S LVP + S L +45 P + S L 45 P 2 : lowvaluesofpolarization,degreeofpolarizationisbiasedbecauseitisapositive quantity.Tocompensateforthisfact,abiasterm P canbesubtractedfromthemeasured polarizationdegreesuchthat: P unbias = P P (4.14) P = q ˙ 2 Q + ˙ 2 U : (4.15) Weallowthisunbiasedmeasuretobenegativetokeepourdistributionssymmetric. 4.3.2ImageReduction TheimageswereinitiallyreducedthroughtheSOARpipeline 3 writtenbyNathanDeLee. TheSOIpipelineisacollectionof \IRAFCL" 4 scriptsthatwillsubtractthezero-length 3 http://khan.pa.msu.edu/www/SOI/ 4 IRAF(NOAO)V2.16 http://iraf.noao.edu/ 119 exposureCCDbiasthennormalizetheimagebasedonthevariationofthepixel-to-pixel response.Theoriginalimageisamulti-extensionFITSfromtwoCCDsegments(four totalextensions),whichiscombinedintoasinglecorrespondingtoasimpleFITS withaphysicalchipgap.Weeditedthescriptsfromapreviousversiontoallowforat usingadomeatwithapolarizerandascienceWeeditedtheimageheaderWorld CoordinateSystemtermsusingaTwoMicronAll-SkySurvey(2MASS)starintheas anastrometricreferencestarandtheusingother2MASSstarsinthe IndividualexposureswerealignedandtrimmedusingtheIRAFtaskimalign.Cosmicrays wereremovedusingtheL.A.CosmicLaplacianedgedetectionroutine[vanDokkum,2001], whichtypically < 1%ofstarsascosmicrayswhilecorrectlyidentifying ˘ 98%ofcosmic rays.Priortotakingobservationsofscientargets,wetakecalibrationsfromavariety ofsources.Inadditiontobiasframesanddomeweacquireskycalibrationsincluding: unpolarizedstandardstars,globularclusters,polarizedstandardstars,andpolarized nebula.Weusepolarizedandunpolarizedstandards(bothstarsandextendedsources)from Turnsheketal.[1990]. 4.3.3DomeFlats Foradomealightisprojectedontoawhitescreensuchthatthelightisscatteredand shouldbeunpolarized.Twilightontheotherhand,canbehighlypolarized(especially iftakenwiththetelescopepointedoppositethedirectionofthesettingorrisingsun,asoften isdone)andthereforewillbeunreliableforuseasacalibrationsource.Thedomewere takenforallcombinationsofastronomicalandpolarizationused.Wereducethe imagesbyatwithallmatchingcombinations.Toverifythecorrectionand quantifytheerrorsonthecorrection,weobserveunpolarizedstandardstarsandglobular 120 clusters.Wetesttheratiosofthe0and90degreepolarizersandthe45and135degree polarizersusingnormalizeddomeinFigure4.4fromtheJanuary4th,2013night.The normalizeddomearecomparedtoverifythereisnopixel-to-pixelvariationinsensitivity topolarizedlight.WealsocomparetheratioofnormalizedVband0degreeimagefrom January4thandJune4th2013inFigure4.5.Inbothandbothsetsofthere isaspreadof ˘ 0.6%intheperpixelratioandtheratiosarealsostatisticallyconsistent withone,similartothespreadintheindividual,pre-combinedimages.However,weare unabletousethecurrentdometoverifythetelescopeanddomelightdonotcause polarizationatanoverallthroughputlevel. Totestwhetherthereisanoverallthroughputpolarizationcausedbythetelescopesor domewewillneedtotakewiththesamewhilerotatingthe cameraassembly90 andthenbacktotheoriginalorientation.The90 rotationwillverify whetherthethroughputofthetelescopeispolarizedasthethroughputforeachpolarizer shouldbeequaltoitselfwitha90 rotation.Takinganadditionalobservationbackinthe originalorientationwillverifythatanyintherotateddomearenotdueto stability/variabilityinthebrightnessofthedomelampitself.Weplantoconductthis lasttestpriortopreppingthisworkforpublication,beyondthetimescaleforthisthesis. Forthiswork,wehavecontrolledforanyintrinsicpolarizationbyobservingunpolarizedstar asanullcontrol,whichwedescribeinthenextsection. 4.3.4UnpolarizedCalibrationTargets Weobservedunpolarizedstandardstarsaswellasglobularclusterstoestimatethezero pointcorrectionforthepolarizers.Weusedtheglobularclustersasexamplesofunpolarized extendedsources.Eventhoughglobularclustersaremostlyunpolarized,someofthestars 121 Figure4.4SingleNightDomeFlatComparison .Wetesttheratioofnormalizeddome tocheckthepotentialpixeltopixelvariationinpolarization.Theupperhistogramisaratio ofthenormalized0 and90 andthelowerhistogramisaratioofthenormalized45 and135 fortheVbandrfromJanuary4 th ,2013.Thecurveoverlayingtheshaded histogramistheGaussianforthedistribution.Themean( ),andstandarddeviation ( ˙ )areincludedineachofthe 122 Figure4.5MultipleNightDomeFlatComparison .Wetesttheratiodomeatsbetween January4thandJune4th2013fortheVbandpolarizationTheuppercompares thenormalized0degreeimagesandthemiddlecomparesthenormalized45degree images.Thelowercomparestwoindividual0degreedomebeforetheyhavebeen averaged. 123 orregionsoftheglobularclustermayhaveasmallamountofpolarizationbothintrinsicto theglobularclusteraswellaspartofthelineofsightinterstellarmediumsignal.Therefore, inourinvestigationofthevariabilityofthepolarizationsignalasafunctionofpositionon theweusestarsextractedfrommultipleglobularclusters.Starsarefoundineach oftheimagesusingtheIRAFDAOPHOTpackageforunsaturateddetections > 100 ˙ ( < 1% error).Fortheglobularclustersweobserved,typically500-2000starswerefoundineach imageandamajorityofthosestarshadmatchingpairs. Weplotthedistributionofthe0 90 and45 135 relativeinFigure4.6.We multiplicativecorrectionfactorswithrespecttothe0 imageof0.9725 0.00036,0.9488 0.00039,and1.0085 0.00095.ThebetweentheVandRbandcorrectionswas foundtobeatmost1.5%. Forourscienceobservationsweusebrightunsaturatedstarsinthetocomputethe correctionfactorbetweenthepolarizersastheseshouldbeamoreaccuraterepresentation ofthatcurrentincurrentconditions.Ifanycorrectionfactorsaretlyt thantheonesfoundinfromtheglobularclusters,thoseobservationswerenotusedbecause atwouldmeanthatcloudsorotherconditionsaoneorboth observationsinapair.Asacheckonthecorrectionaswellasphotometricerences, wecomputethetotalintensityinouraperturesbutseparatelysummingthe0 90 and 45 135 pairs.Wethearetypicallyinagreementto ˘ 1%.Wecompute ameanstandarderrorforthecorrectionfactorbasedonthemeancorrectionfrommultiple stars. TheinitialerrormapisthePoissonerroronthecounts(convertedtoelectronsbyapplying theSOIgainof2e ADU 1 )andaddingtheperpixelreaderrorof4.4e perexposure andtheerroronthecorrectionfactorinquadrature.Wethenchoseskybackgroundregions 124 whichwereinthesamepositionoftheskyforeachsetofpolarizers.Typicallyweused circularareasatleast30 00 acrosstoreducetheerroracrosstheimages.Theinitialerror(in electrons)forNpixelsis: ˙ = q (2 I ) 2 + N 4 : 4 2 + N 2 ˙ 2 corr =N stars + N 2 ˙ 2 bg =N bg ; (4.16) whereIisthetotalnumberofcounts, ˙ corr isthestandarddeviationinthecorrection factor, N stars arethenumberofstarsusedinthecorrectionfactor, ˙ bg isthestandard deviationinthebackgroundand N bg istheareausedtocomputethebackground. 4.3.5PolarizedCalibrationTargets Weanalyzedpolarizedstandardstarsusingthesameprocedureastheunpolarizedstandard starprocedure.Bycomparingourmeasurementstothepublishedmeasurementsforthe polarizationstandardprovidesanestimateoftheaccuracyofourestimatedpolarization fractionaswellastheaccuracyoftheangleofthepolarizationvector.Typically,these starsareshroudedbydustwhichcanchangeoveryeartimescalesandcreateavariable signalmakingboththemeasurementofpolarizationfractionaswellasangleto obtainconsistentmeasurements.Therefore,weusednebulaemissionandtheAGN jetinM87asadditionalpolarizedcalibrationstoensurewewereabletofaithfullyreproduce stronglypolarizedsources. WecreatedStokesparameterimageswithanIDLcodewrittenforthisproject.Thecode alsocalculateserrormapsofeachparameter.Wetypicallyappliedsmallbinning(2 2or 4 4abovethe0.15 00 pixels)tothedatawiththeCONGRIDIDLroutinetoimprovethe signaltonoiselevelofdetectionsanddidnotinterpolateacrosscellstoconserveand 125 Figure4.6GlobularClusterPolarziationComparison .Wepresentglobularclusterpolariza- tionhistogramsforNGC1851.Theupperplotisforthe0 90 comparisonand thelowerplotisforthe45 135 comparison.Thecurvesoverlayingtheshadedhistograms aretheGaussianforthedistribution.Thenumberofstars(#),mean( ),andstandard deviation( ˙ )areincludedineachofthe 126 tomaintainthenoisepropertiesfromcelltocellwhichworkswellonourpolarized emission.AswellasimagesforeachoftheStokesparametersI,Q,andU,wecreatemapsfor thefractionalpolarizationandandforthesignal-to-noiseratioofthefractionalpolarization. Forthecreationofpolarization-fractionvectors,weimposeaminimumS/N(i.e. I P =˙ I P ) ratioof5isused.WepresentimagesoftheCrabNebulaandshowtheoutputoftheBessell VimagesinFigure4.7. TheCrabNebula(M1)isapulsarwindnebulawithaveryhighpolarization( ˘ 25%) whichhashaditsopticalpolarizationmappedmanytimes[Woltjer,1957,Wilson,1974, Schmidtetal.,1979,HicksonandvandenBergh,1990].Weplotthepolarizationmaps createdwithourIDLcodefortheCrabNebulaVbandimagesinFigure4.7whichinclude: theintensityimage,theStokesQmap,theStokesUmap,thepolarizationfractionmap, andthesignaltonoisemap.Thechipgapisstillvisiblealongthecenterbecausethese weremadefromasingleobservations.Onlythescalebarforthesignaltonoiseplotis includedtogiveanifwhichareasthatarenotblackthathaveasignaltonoiseratioat least5.Wecompareourelectricpolarizationmaptothemagneticpolarization mapinHicksonandvandenBergh[1990]inourFigure4.8.Weonlyplotvectorswhich haveatleastasignal-to-noise(i.e.P/P err )greaterthan5.Ourvectorsareorthogonalto theirsbecauseoftheorthogonalityoftheelectricandmagneticcomponentsofthelight fromMaxwell'sequations.Wemakeaquantitativecomparisonusingtwonebularregionsin theCrabwhicharepolarizationcalibrationtargetsinTurnsheketal.[1990],presentedin Table4.2.FortheCrabNebula,astronglypolarizednebula,wethatthemeasurements fortheindividualobservationsareconsistent,butinacouplecasesarefourtoepercentage points( ˘ 20%)tthanthevaluesfromTurnsheketal.[1990],alargerthan themeasurementerror.Themeasuredanglesaretypicallywithinthemeasurederrorofthe 127 Figure4.7CrabNebulaPolarizationImages .WeincludeVbandimagesofthefull50 50 oftheCrabNebulafromourJanuary4th,2013night.Thestripedownthemiddleisthe chipgappresentinindividualobservations.Fromupperlefttolowerright,theare: Totalintensityimage,StokesparameterQimage,StokesparameterUimage,Polarization percentimage,andsignaltonoiseimage.InthePolarizationpercentimage,thegradient rangescalerepresentsapercentagepolarizationbetween2%and40%.Inthesignaltonoise image,thegradientrangescalerepresentsasignaltonoisebetween5and20. standardvalueandwithin2 ofthestandardvalue. RMonocerotisisaTTauristarwhichispartofthenebulaNGC2261.Giventhe youthofthestar(300,000years)thepolarizationvectorscanchangeonveryshorttimescales (afewyears)duetothevariabilityofthenebula[Johnson,1966].Polarizedstandardstars, likeRMonocerotis(RMon),typicallyhavecircumstellardustwhichinducespolarization ontheirlight.ThestarinthenebulaRMonisagoodtestbecausethepolarizationofthe emissioninthenebulaformsaringaroundthestar.InFigure4.10wecomparethepolar- 128 izationmapofRMontothemapinCloseetal.[1997]toshowthatthepolarizationangles whichmakearadialarcsnorthofthestarareconsistent.Wemakeaquantitativecompar- ison,presentedinTable4.2,usingthreenebularregionsinRMonwhicharepolarization calibrationtargetsinTurnsheketal.[1990]. 129 130 Table4.2.ExtendedPolarizationStandards SourceRADecApertureP V V Ref.P V Ref. V (|)(2000)(2000)( 00 )(%)(deg)(%)(deg) CrabNebula-15 h 34 m 33 : 01 s +22 00 0 40 : 0 00 5.3 00 16.61% 0.91%160.7 0 : 7 21.45% 0.50%160.7 CrabNebula-25 h 34 m 33 : 14 s +22 00 0 13 : 5 00 5.3 00 31.20% 0.32%168.6 0.4 29.68% 0.61%170.6 Rmon-16 h 39 m 09 : 98 s +8 44 0 41 : 4 00 5.3 00 17.91% 0.42%90.3 1 : 7 14.52% 0.47%89.9 Rmon-26 h 39 m 10 : 75 s +8 44 0 27 : 7 00 5.3 00 10.38% 1.22%133.1 1 : 0 13.78% 0.45%118.1 Rmon-36 h 39 m 09 : 98 s +8 44 0 23 : 7 00 5.3 00 10.95% 0.88%84.8 3 : 4 11.90% 0.47%87.7 Note.|WeuseTurnsheketal.[1990]forourreference. 131 Figure4.8CrabNebulaPolarizationVectors .Our5 0 5 0 intensityimagefortheCrabNebulae,overlayedwithpolarization electriclinescalculatedonthe2 00 2 00 bins.Theminimumpolarizationvectoris10%.WecomparetheSOARimageand thedirectionofthelinestothatofHicksonandvandenBergh[1990]inFigure4.9. Forthesetwostronglypolarizednebulawethatthemeasurementsfortheindividual observationsareconsistentbutinacouplecasesarefourtoepercentagepoints( ˘ 20%) tthanthevaluesfromTurnsheketal.[1990].Themeasuredanglesaretypically withinthemeasurederrorofthestandardvalueandwithin2 ofthestandardvalue.M87 (NGC4486)isthebrightestclustergalaxyinthenearbyVirgoCluster.Itcontainsan AGNwhichisknowntohaveastronglybeamedjetwhichhashaditspolarizationobserved opticallywithHST[Perlmanetal.,1999].Tomeasurethepolarizationofthejetwetooka seriesofV-bandimagesofM87andusetheIRAFroutineellipsetosubtractoutelliptical isophotesofthegalaxy.Wedonotmaskthejetandinsteadrelyonoutlierrejectionto avoidsubtractingthejet.ThesesurfacebrightnessisophotestrackwellforM87andsimilar largeellipticalsbecausetheirstarlightareverysmooth.Wereproducetwo fromPerlmanetal.[1999]andcompareittoourpolarizationmapforasimilarregionin Figure4.11.TheextractedregioninourSOARdataiscoincidentwiththeAGNw 11 00 -18 00 (whichareincludedintwoseparateinthePerlmanetal.[1999]paper)from thecenterofthegalaxyasseenintheHSTimage. 4.4PolarizationLimitsofM87Filaments WeinvestigatethepolarizationsignatureintheH tsofthebrightestclustergalaxy (BCG)M87.ForcontinuummeasurementsweobservedbroadbandVandRaswellasnarrow bandCTIO6129-140whichisofH aswellasothercommonemissionlinesfoundin BCGs.Similarlytotheotherpolarizationanalysis,weanalyzeeachnightseparatelytoverify theconsistencyofthemeasurements.Wetriedtobinthedata4 4toimprovethesignal tonoisewhilekeepingthementarystructureintacttomakeapolarizationmapsimilar 132 Figure4.9HicksonandvandenBergh[1990]CrabNebulaMagneticFieldVectors .Were- produceHicksonandvandenBergh[1990]Figure3foraslightlysmaller(3.9 0 3.7 0 ) oftheCrabNebula.Inthethemagneticlinesareplottedinsteadofelectric lines.Theirvectorsareplottedwithapolarizationaveragedfrom2 00 2 00 bins.The2 00 separationalsocorrespondstoa50%polarizationvector. 133 Figure4.10RMonocerotisPolarizationImage .Wecompareourmapofthepolarizedstan- dardstarRMonocerotistothemapinCloseetal.[1997].Noticethepolarizationvectors formacirclearoundthestarnearthebottomoftheimage.Therestofthepolarization vectorsinthenebulaformasimilarshape. 134 Figure4.11M87AGNPolarizationComparison .Weplotthepolarizationvectorsonthe intensityimageoftheM87jetwherewesubtractedoutellipticaltoremovethe backgroundgalacticstarlight.TheplottotherightisacombinationoftwoHSToptical polarizationmapsfromPerlmanetal.[1999],whichcorrespondtotherelativeregionsinour map. 135 totheM87jet,Crab,andRMon.However,evenatthatresolution,thelowlevelofsignal tonoisemadeitimpossibletomakeapolarizationmap.Instead,westrategicallyextracted 651.5 00 1.5 00 regionsonbrightregionsofthetsandcomputedStokesparametersfor eachoftheboxes.ThelocationsoftheseboxesontheH tsareshowninFigure4.12. Tocreateacontrolsample(i.e.regionswhichshouldn'thavepolarization)weextract234 1.5 00 1 : 5 00 regionsnorthofthecenterofM87inregionsthatarenotonH tsbut haveasimilarnumberofrawcountsperpixelsuchthatthesignaltonoiseforallregions aresimilar. Wedeterminethelimitsonthepolarizationofthetswithtwotests.First,intest 1,weexaminethecontinuumemissionasacontrolandverifyitisconsistentwithzeronet polarization.Intest2,weseparatelyextracttheH regionsfromthecontinuumregionsand runthetestwiththeH whichhasnotbeencontinuumsubtracted,toimprovethenumber ofcountsavailable.Fortest1,wevthatthecontinuumregionsareconsistentbetween dithersetswithinagivennightusingthenon-parametric,two-sampleKolmogorov-Smirnov (K-S)test.AsmallDstatistic(equivalentlyalargep-value)indicatedthatthedatafrom eachditheredpositionsareconsistentwithhavingbeendrawnfromthesamedistribution. However,eventhoughthedistributionsarethestatisticallysimilar,theremaybesome intherelativephotometrybetweeneachoftheimagessuchthattheQandU Stokesparametersmaynothaveadistributionconsistentwithzero.InFigure4.14weshow anexampleofthiswheretheQvaluesoftwotobservationshaveameanconsistent withzerogiventheerrorwhileoneoftheUcontinuummeasurementsisnotconsistent. Additionally,aK-Stestforthetwoobservationswhichhaveastandarddeviationlargerthan theirmeanmaynotcomeouttoanullresultbecausetheirmeansareontheoppositesides ofzero,likeinthecaseofFigure4.13.WiththeK-Stestweverifywhetherthecontinuum 136 polarizationdistributions,whichshouldhavenonetpolarization,lookstatisticallyidentical afterasmallmeanvalueissubtracted.Tocorrectfortheinthemeans,wesubtract eachofthemeansfromthedistributionssuchthattheyareidenticallyzeroandcanbe compared.TheresultsoftheK-Stestpairsfortest1indicateanullresult(i.e.p > 0.10) forallpairsofcontinuumcontrolsamples. Aftertheconsistenciesinthecontrolsamplewerevintest1,weexamined thedistributionoftheH boxesintest2,theshadedregionsinFigure4.13.Byeyeitap- pearsthat,whilenoisy,theH regionhistogramsalignwellwiththeirrespectivecontinuum histograms.TobeabletocomparetheH distributions,wecorrectedthemfortheircon- tinuumbysubtractingtheH histogramsbytheirrespectivecontinuummeans.Theresult forQandUforoneimagearedisplayedinFigure4.14.Afterthecorrection,thevaluesare nowconsistentwithzero.Therefore,wecandeterminethelevelofpolarizedemissioninthe H regionswithrespecttothecontinuumcontrolregions,whichshouldcontainnopolarized emission.Inthiscase,afterwecorrectforthecontinuum,wethatthedistributionof polarizationmeasurementsintheH regionsareconsistentwithnopolarization. Wearealsoinvestigatingifthereareanyspregionsthatmightcontainpolarization, sowemustverifythatthereisnospatialtrendeventhoughthetotaldistributionofthe H tshaveadistributioncenteredonzeropolarization.InFigure4.15weplotall H boxesforfourseparateimagesandcolorcodethembasedonpolarizationfraction.The polarizationfractionwhichiscomputedistheunbiasedpolarizationfractioncorrectedusing thecontinuumvalues;aprocesssimilartowhatwasdoneinforthecorrectedhistograms. Whiletheirscalesarenotidentical(typically-1%to1%inpolarizationperpixel)noindi- vidualpixel,norcombinationofclusteredpixels,hasastatisticallytpolarization level.WeestimatealimitonthepolarizationmeasuredoftheH tsusingthebiased 137 Figure4.12M87H PolarizationFields .Allthreepanelshavetheircolorsinvertedsuchthat darkerareasaretheareasofemission.Theupperpanelisthecontinuum-subtractedH image,whichisastackedimagethatincludes8observationsoffourpolarizedimagesfrom twotditherpositionsfromApril7 th ,2011.ForM87,the15 00 scalebarisequivalent to1.35kpc.Thetwolowerpanelsarezoomedinareasoftheupperimage.Theboxesare 1.5 00 1 : 5 00 boxesandweretheregionsusedtomeasureStokesparameters. 138 correcteduncertaintiesontheH QandUhistogramsinFigure4.14as: ˙ 2 = P N i =0 ˙ 2 Q i + ˙ 2 U i N ; (4.17) whereNisthenumberofobservations.Weda3 ˙ limitforpolarizationpercentage inanyoftheindividual1.5 00 squaresof2.4%.The3 ˙ upperlimitonthemeanpolarization overtheentiretcoveredbythethesemeasurementsis0.3%,usingtheerroronthe meanforthe65boxes. 4.5ImplicationsfortheFilamentsinM87 WhilewedodetectcantpolarizationalongthejetofM87aswellasatthenucleus,the ucleusH tarenoticeablypolarizationfree.Sparksetal.[2014]hasoptical(450 nmto900nm)spectropolarimetrydataforM87fromtheVLT.Emissionlines,including H fromtheucleartaryregionsofthegalaxyarepolarized < 0.1%.These measurements,aswellasourlimits( < 2.4%forindividualboxes, < 0.3%fortheentire region)ofthets,providetconstraintsonthepossiblemechanismsforthe illuminationoftheucleusH ts.FromdeepobservationofAbell2597,Voitand Donahue[1997]usedopticallineratiostoshowthatshockscan'tbeacantcontributor toionizationandthattheadditionofphotoionizationfromhotstarswouldalsofallshortof thetotalenergyrequirement.Anadditionalcontributionfromsomething,likeconduction fromthehotICMintothecoolerts,aprocesswhichoperatesinregimeswithstrong temperaturegradients,couldmakeupthisForM87,thecontributionfromhot starsislimitedaswellbecauseSparksetal.[2009]foundnohotstarsinthevicinityofthe south-easttsthatcouldaccountforUVneededtopowerthets.Instead, 139 theonlyUVemissionseenintheHSTimageswasfromatary10 5 Kgaswhichwas tobeC IV (1549 A)andHe II (1640 A)lineemission,byobservationswiththe CosmicOriginsSpectrographonHST[Sparksetal.,2012].Thatemissionispotentiallyan indicatoroftheintermediatetemperaturegasattheinterfacebetweenthehotICMand thecoolerts.Therefore,thermalconductionislikelyanimportantcontributorto thestructureandemissionseeninM87,aswellasotherBCGs.Howeverthereisadebate whethertheconductionisstandardeconductionorsaturatedconduction,adebate thatcanberesolvedbyexaminationofthepolarizationsignalinthelineemissionlaments. 4.5.1ThermalConduction Standarde(i.e.Spitzer)conduction[Spitzer,1962]istheprocessofthermalconduc- tionwheretheheatisconductedbyelectronsinafullyionizedhydrogenplasma.Theheat fromconductionis: q = ; (4.18) where ,thethermalconductivityisequalto: 1 : 84 10 5 t 5 = 2 e ergs 1 deg 1 cm 1 : (4.19) Thisformassumesthattherearemanycollisionsoverthescalelengthofthetemperature gradient,i.e.themeanfreepathismuchsmallerthanthescalelength.Undertheconditions ofSpitzerconduction,thehotelectronsareabletoheatthetsisotropically,therefore, theemissionisnotexpectedtopolarized.CowieandMcKee[1977]presentamodi ofSpitzerconductionintheregimewherethemeanfreepathoftheelectronsismuchlarger thanthesizeoftheinterface.InthiscasenormalSpitzerconductionreachesamaximum 140 Figure4.13M87ContinuumStokesParameters .ThehistogramsplottheQandUStokes parameterdistributionsforthe234scaled(i.e.dividedbythetotalcountsintherespective pairofpolarizers)continuumboxes(1.5 00 onasidesquares)aswellasthe65H boxesof identicalsizeinM87.Thecentersofthesedistributionsarenotshifted.Thetwot indexesrefertotwotpointingsinApril2011.Weseethatineachpointing,while thedistributionsforeachpointingaretheH andcontinuumdistributionsforeach pointingarealigned. 141 Figure4.14M87H StokesParameters .Thelefthistogramsincludesthescaled(i.e.divided bythetotalcountsintherespectivepairofpolarizers)QandUStokesparametersforthe continuum1.5 00 onasidesquares.Therightgraphpresentstheequivalentforthesampling of65H 1.5 00 onasidesquareswheretheeachofthedistributionshasbeenalignedby subtractingthemeansofthecontinuumQandU. 142 Figure4.15M87H RegionPolarization .Weplotphotometryrencesbetweenpolar- izersdividedbythemeanineachofthesquareaperturesasanestimateofpolarization fraction.ThetoptwoboxesarefromobservationsfromApril2011andthebottomtwoare fromApril2013.Thecalculatedfractionsareunbiasedbysubtractingthemeanofthedistri- butionforthecontinuum.Weallowthefractionstobenegative(whichareconsistentwith zeropolarization)todisplaythedynamicrange(i.e.standarddeviationoftheH boxes). Lookingacrossalltheimages,theboxeswhichareatleast2 ˙ abovezeropolarizationina givenimagearearenotseenconsistentlyabovezeropolarizationacrosstheimages. 143 heat q sat ˇ 3 = 2( n e kT e ) v char . Themaximumheatisknownassaturatedconduction.Thecharacteristicvelocity v char ismuchlessthantheelectronthermalvelocitybecauseotherwisetheelectrons,which arenearlyfreestreaming,wouldmovetooquicklycomparedtotheionsandbuildupanon- zeroelectricThereforethecharacteristicspeedismuchclosertheslowerionvelocity whichmeansthattheions,whicharemoremassive,transfermoreoftheheat.Fabian etal.[2011]suggestthattheH tsarepoweredthroughsaturatedconductionfrom theICMionsthroughreconnection[LazarianandDesiati,2010].Theprocessof reconnectionallowsthewofonegasintoanotherperpendiculartoamagnetic becauseoftheamountofturbulenceinthet.Werneretal.[2013]both turbulentandmagneticpressureinthetssuchthatthereconnectionis anapplicableprocess.UnderreconnectiontheICMgaswillowintothecold tsperpendiculartothets.Therefore,wewouldhaveananisotropicvelocity ofchargedparticles,whichproducesstrongpolarization,5-25%,inemissionlinessuch asLy andH [Laming,1990].However,wedonotseestrongpolarizationinthements ofM87sothereconnectionprocessisunlikelytobethecauseoftheH t emission.Whileitcouldbethecasethatourprojectionoftheentsisbiased,itwouldbe verytochangethegeometrytoproduceamoreisotropicvelocityandcreatean unpolarizedobservationinM87.Therefore,itismorelikelythatthermalconductionoccurs undertheSpitzerregimewherethehotelectronsinteractwiththetsisotropically. 144 4.6Conclusions Astronomicalpolarimetrycanprobethestructureandphysicsofmanytypesofastrophysical sourcesinnearlyallwavebands.Wepresentheretheimagingpolarizationobtainedwith theSOARTelescope,includingafullsuiteofsystematictestsandcalibrationsthatdemon- stratethesystemcanmeasurereproduciblefractionalpolarizationforextendedtargets.The additionofpolarimetrycapabilitiestotheopticalimagerontheSOARTelescopeprovides newcapabilitiesfortheSOARtelescope.Thecalibrationrequiresthestandardcalibration frames(i.e.biasframesanddomeforeachpolarizerandcombination)aswell asunpolarizedsources(standardstarsandglobularclusters)forzeroordercorrectionsand polarizedstandards(polarizedstarsandextendedsources)toverifypolarizationanglesas wellaspercentages.Wethatwecansetlimitsofpolarizationdownto0.3%asseenin theupperlimitsonthepolarizationoftheM87ts,measurepolarizationfractionsto within20%andmeasurethepolarizationangleofstronglypolarizedsourcestoaprecision of 2 . Thestudyofbrightsourcesaswellasadditionalcaretakentoincreasethenumberof ditherpositionsoncethecountrateishighenoughnottobedominatedbyinstrumental noisecouldgofarthertodecreasetheerrorsonpolarizationmeasurementswhicharerelative photometrymeasurementsinsteadofthemoreabsolutephotometrymeasurements neededforcalibration. Thelimitof2.4%polarizationfortheindividual(1.5 00 1 : 5 00 )H tregions,aswell asthelimitonthemeanofalloftheregions,couldbeimprovedwithadditionalobservations withabettererrorestimationonmeanoftheQandUStokesparameterhistograms.There isalimittomeasuringfainterstructures,becauseincreasingthelengthofobservationsis 145 limitedbyskyvariability.Forincreasedprecision,increasingthenumberofobservations andexcludingthemorevariableobservations,likelyduetobadweather. ThelackofpolarizationseenintheH tsmakesastrongcaseforrulingout saturatedconductionasthemechanismfortransportingenergyfromthehotICM,tothe emission-lineents,becausesaturatedconductionwouldresultinstronglypolarizedemis- sionofthements.Saturatedconductionmaystilloperateoversmallregionsbutitap- pearstoberuledoutasthedominantmethodforenergytransportation.However,standard Spitzerconductionisstillaviableoptionforenergytransport. WewouldliketothankDr.WilliamSparksforthepurchaseofthepolarizers(purchased undertheSTScIDirectorsDiscretionaryResearchFund)andhisfeedbackandsupporton designingthetestsandcalibrations.Also,wewouldliketothanktheSOARoperators DanielMaturana,PatricioUgarte,SergioPizarro,andAlbertoPastenfortheirhelpduring observingnightsandSOARDirectorDr.StephenHeathcoteandDr.SeanPointsfortheir helpwiththeinitialinstallation. 146 Chapter5 Summary Thestateoftheintraclustergasinagalaxyclusterisnotonlyindicativeofthecurrentstate ofthecluster,butcanalsoprovideinformationaboutthehistoryofthecluster.Therehave beenmanysuggestedmethodsofheatingwhichwouldpreventthecreationofa\coolingw" whichwouldcauseagalaxyclustertocatastrophicallycoolontimescalesmuchshorterthan theageofthecluster.Inthisdissertationwepresentedanalysisofbothbrightestcluster galaxiesandtheirhostgalaxyclusterswhichsupportsomeofthecommonmechanisms proposedforheatingthecoreofthecluster. InChapter2wesawthat ˘ 40%ofBCGsinlowcentralentropyclustershadUVor mid-infraredemissionthatwasconsistentwithstarformationsuchthatintheinfraredwe wereabletomodeltheemissioninthegalaxyasastarburstplusanoldstellarpopulation. Withsuchalargefractionofstarformationsignaturesinthesebrightestclustergalaxies, starformationcontinuestohaveanimportantroleintheemissioncharacteristicsofcluster cores.Withtheseobservationsoflargedustmassesinthemid-infrared,itwillbeinteresting toobservetheseBCGsatlongerwavelengthswithtelescopessuchasALMA,whichwillbe abletomeasuremolecularemissioninthecoresandprovidesourcesofevencoldergasand metals. InChapter3weexamineboththecoolcoregalaxyclusterRXJ2014.8-2430aswellas itsBCG.Thisisastrongcoolcoreclusterwhichshowssloshinginitscore,butcavitiesare 147 noticeablymissingintheX-rayimages.Usingtoymodelsaswellasanestimatefromthe X-rayluminosityinthecoreofthecluster,weestimatethelargestpossibleAGNcavities thatcouldbepartofthissystem.WealsomeasureopticalH emissionimagingwithoptical spectraoftheBCGandthatwhilethereisavelocitygradientacrossthecentralregion oftheBCG,whichmayindicatethatthegasisbeingpulledoftheclusterorfallingin, thecentersofthestellar,opticalline,andX-rayemissionareinfairlygoodagreement. Therefore,wemayhavecaughtthisclusteratapointintimewhenithasjustbegunto slosh.Atthispoint,RXJ2014.8-2430isknownasaradiosourcebuttheonlyobservations haveverypoorresolutionwhichpreventsusfrombetterinvestigatingtheimpactofradio bubblesinthecoreofthecluster.Ahigherresolutionradioimagewouldprovidee evidencefortheAGNactivityinthisclusterandhelpanswerthequestionofwhetherornot weareviewingthisclusterduringauniqueperiodwheretheAGNhasjustrecentlyturned on.Also,futureX-raymissions,suchasAstro-H,areofcriticalimportancetothestudy ofgalaxyclustersbecauseoftheimmenseimprovementsintheenergyresolution.With smallerenergybinsitwillbepossibletomorepreciselymeasuremetalabundancesaswell asseparateelementalabundances.Withadditionaliondiagnostics,itwillbeeasiertoprobe thestateofthegas.Improvedenergyresolutionwillalsobetterprobethevelocitywidths oftheX-raygaswhichcanbecomparedtotheopticalemission. Chapter4presentsourpilotstudytomeasureopticalpolarizationusingSOIontheSOAR Telescope.Wediscussthemethodaswellassomebestpracticestousewhileobservinginthis mode.WereviewthedatareductionandanalysisstepstocalculateStokesparametersand polarizationvectorsforbothstronglypolarizedtargetsaswellastocomputeupperlimits onsources.WeareabletosetlimitsonthepolarizationoftheH tsinM87which ruleoutsaturatedconductionasamechanismforconductingenergyfromtheintracluster 148 mediumintotheH tsoftheBCG.Withadditionaltime,itispossibletomeasure evenmoreprecisepolarizationsusingthisnewmodeontheSOARTelescope. 149 APPENDIX 150 TableA.1.BrightestClusterGalaxyIden ClusterName GALEX IAUName2MASSIDIRACAORMIPSAOR 1E065756a{06 h 58 m 16 s : 04 ; 55 56 0 3695 00 1267404821981440 a ;b 2308992023089664 a 1E065756b{06 h 58 m 35 s : 16 ; 55 56 0 558 00 1267404821981440 a ;b 2308992023089664 a 2A0335+096033840.6+095812 c 03384056+09581191864652818636544 a ;b a MIPSAOR(AstronomicalObservationRequest)observationincludesa24 mobserva- tion. b MIPSAORobservationincludesa70 mobservation. c GALEX observationonlyincludesFUVmeasurement. d GALEX observationonlyincludesNUVmeasurement. e GALEX observationincludesbothNUVandFUVmeasurement. f MIPSAORobservationincludesa160 mobservation. g Spitzer observationsweremadebyDonahueaspartoftheDDTprogram488. h Duetoananomalouspointsourceinthepbcdreduction,theimagesneededtobe reducedfromthebcdframes. i GALEX observationwastakenasaGuestInvestigatorforHicksetal.[2010]. j IRACAORobservationonlyhasbands1and3ontarget. k IRACAORobservationonlyhasbands2and4ontarget. Note.|Thistableisavailableinitsentiretyinamachine-readableformintheonline journaletal.,2012b].Aportionisshownhereforguidanceregardingitsformand content. 151 152 TableA.2.PhysicalProperties. ClusterNameRedshiftCentralEntropy a K 100 a alpha a ScaleIRRadiusUVRadiusCentroid a (keVcm 2 )(keVcm 2 )(kpc/ 00 )( 00 )( 00 )(kpc) 1E065756a0.2960307.4518.551.884.413.24{503.92 1E065756b0.2960307.4518.551.884.413.24{216.72 2A0335+0960.03477.14138.641.520.6920.712.81.00 2PIGGJ0011.5-28500.0753101.98214.680.841.4310.0117.32.10 2PIGGJ2227.0-30410.072917.13113.951.371.3910.3117.30.59 3C28.00.195223.85107.821.793.244.4212.83.55 3C2950.464114.5481.951.475.862.44916.98 3C3880.091717.03214.30.761.718.3817.31.44 a QuantitiesinCavagnoloetal.[2009]arefromradiallyentropypwithafunctionalform K(r)=K 0 +K x (r = r x ) ,whereK 0 isthecentralentropyinexcessabovethepowerlawK 100 istheentropy normalizationat100 h 1 70 kpcfromtheX-raycentroid,andalphaisthebpowerlawindex. b UVdataistakenfromHicksetal.[2010] Note.|RedshiftinformationandcentralentropiesarereproducedfromtheACCEPTdatabase.Thesizescale iscalculatedastheangulardistancesizeassumingthestandardcosmologyinthepaper.TheIRradiiareset at14.3 h 1 70 kpcinsizeandareusedfor 2MASS andIRACaperturemeasurementsexceptinthecasewherethe apertureisbelow5 00 .Inthiscasetheapertureissettobe5 00 tominimizelargeaperturecorrections.TheUV radiusissetbytheaperturephotometryinthe GALEX databasewhichmostcloselymatchestheGALEXView totalmeasurement.WecomparetheBCGpositionsthatwederivefromthe 2MASS locationsinTableA.1 andcomparethattotheX-raycentroidintheACCEPTdatabase.Thesedistancesarepresentedinahistogram inFigure2.1. Note.|Thistableisavailableinitsentiretyinamachine-readableformintheonlinejournaletal., 2012b].Aportionisshownhereforguidanceregardingitsformandcontent. 153 TableA.3.FluxesMatchedtoUVAperture. NameFluxErrorFluxErrorFluxErrorFluxErrorFluxError ClusterNameNUVNUVFUVFUVJJHHKK (mag)(mag)(mag)(mag)(mag)(mag)(mag)(mag)(mag)(mag) 1E065756a{{{{{{{{{{ 1E065756b{{{{{{{{{{ 2A0335+09618.05018.50013.480.013013.130.016013.090.0160 2PIGGJ0011.5-285020.070.109320.650.210813.150.008012.900.010013.050.0131 2PIGGJ2227.0-304118.400.017619.850.056812.980.007012.750.009012.900.0106 3C28.020.510.345320.560.291415.540.066015.350.072015.110.0697 3C29523.17023.800.324516.380.157016.180.223015.760.1470 3C38819.950.050720.970.187813.290.010013.100.012013.280.0139 a BCGsarealsointheHicksetal.[2010]sample. b BCG GALEX aretakenfromHicksetal.[2010]. Note.|Fluxesreportedwitherrorsequalto0are3 ˙ upperlimits.ForNUVupperlimits,the 2MASS ismatchedwitha7 00 aperturesuchthatitissimilarinsizetotheGALEXPSF. Note.|Thistableisavailableinitsentiretyinamachine-readableformintheonlinejournal etal.,2012b].Aportionisshownhereforguidanceregardingitsformandcontent. 154 TableA.4. Spitzer ApertureFlux. NameFluxErrorFluxErrorFluxErrorFluxErrorFluxErrorFluxErrorFluxError ClusterName3.6 3.6 4.5 4.5 5.8 5.8 8.0 8.0 24 24 70 70 160 160 (mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy) 1E065756a0.470.020.380.020.220.010.140.010.150.025.290{{ 1E065756b0.570.030.460.020.260.010.170.010.110.021.43 a 0{{ 2A0335+09626.121.3115.930.8014.790.749.940.502.40 b 0.2477.1020.36{{ 2PIGGJ0011.5-2850{{{{{{{{{{{{{{ 2PIGGJ2227.0-3041{{{{{{{{{{{{{{ 3C28.00.880.040.670.030.350.020.430.020.55 c 0.086.780{{ a Originally,aperturemeasurementindicatedadetection.However,24micronmeasurementandvisualinspection indicatedcontaminationintheaperturewhereislikelyfromanunrelatedsource.Thereportedmeasurementisnow anupperlimitcomputedusingthepointsourceestimateat16 00 . b Sourcenotextendedbuthastcontamination.Pointsourcemeasurementat35 00 radiusisgreaterthan10% errormargin. c Sourcenotextendedbuthasmildsourcecontamination.Pointsourcemeasurementat35 00 radiusiswithin10%error. d Fluxmeasurementisadetection. e IRACmeasurementsderivedfromaBCDimagethatwasremosaicked. f SourceisextendedintheMIPS24micronimage.24micronmeasuredwithintheaperture(seeTableA.2).Note anapertureradiusof35 00 wasusedforNGC4636becauseoftpointsourcecontaminationoutsideof35 00 . Note.|Fluxesreportedwitherrorsequalto0are5 ˙ upperlimits. Note.|Thistableisavailableinitsentiretyinamachine-readableformintheonlinejournal[Hoetal.,2012b].A portionisshownhereforguidanceregardingitsformandcontent. 155 TableA.5. 2MASS ApertureFlux. NameFluxErrorFluxErrorFluxErrorFluxError ClusterNameJJHHKKK24K24 (mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy)(mJy) 1E065756a{{{{{{{{ 1E065756b{{{{{{{{ 2A0335+09665.690.5472.490.9457.940.8340.230.50 2PIGGJ0011.5-285012.830.2215.560.3414.530.41{{ 2PIGGJ2227.0-304115.750.2619.440.3717.220.41{{ 3C28.01.260.121.640.171.680.193.350.51 3C2950.760.111.030.181.310.17{{ 3C38810.310.1812.480.2811.570.3215.380.49 Note.|Thistableisavailableinitsentiretyinamachine-readableformintheonlinejournal etal.,2012b].Aportionisshownhereforguidanceregardingitsformandcontent. 156 TableA.6.StarFormationRates: NameSFRErrorSFRSFRErrorSFRErrorMass ClusterNameUVUVIR70 70 24 24 StellarMass ( M yr 1 )( M yr 1 )( M yr 1 )( M yr 1 )( M yr 1 )( M yr 1 )( M yr 1 )(10 10 M ) 1E065756a{{2.125.2200.910.115.52 1E065756b{{1.641.6500.760.126.68 2A0335+096{{0.720.940.250.210.025.58 2PIGGJ0011.5-28500.920{{{{{{ 2PIGGJ2227.0-30411.820.33{{{{{{ 3C28.02.300.3713.512.1502.050.294.40 3C295{{{{{{{{ 3C3881.1401.781.610.361.470.157.06 4C55.16{{{{{{{{ Note.|Astarformationrateuncertaintyof0identhequotedrateasa3 ˙ upperlimit. Note.|UVSFRarecalculatedforallobjects,includingthoseinhighK 0 systems.Upperlimitsarethencalculatedusing theuncertaintiesontheNUV,K,andinertBCGcolor.IfastarformationrateisconsistentwithaSFRofzerowithin3 ˙ ,a 3 ˙ upperlimitisreported. Note.|TheIRSFRisestimatedbyatotheGrovesetal.[2008]models.The24and70micronSFRsareestimated fromempiricalrelationshipsreportedinCalzettietal.[2010]andtheuncertaintiesarecalculatedusingthe24and70micron uncertainties,respectively. Note.|Thistableisavailableinitsentiretyinamachine-readableformintheonlinejournaletal.,2012b].Aportion isshownhereforguidanceregardingitsformandcontent. BIBLIOGRAPHY 157 BIBLIOGRAPHY J.R.P.AngelandH.S.Stockman.Opticalandinfraredpolarizationofactiveextragalactic objects. AnnualReviewofAstronomyandAstrophysics ,18:321{361,1980.doi:10.1146/ annurev.aa.18.090180.001541. K.A.Arnaud.XSPEC:TheFirstTenYears.InG.H.JacobyandJ.Barnes,editors, As- tronomicalDataAnalysisSoftwareandSystemsV ,volume101of AstronomicalSociety oftheConferenceSeries ,page17,1996. M.Arnaud,R.O.Boulade,L.Vigroux,andE.Vangioni-Flam.SomeCon- straintsontheOriginoftheIronEnrichedIntraclusterMedium. AstronomyandAs- trophysics ,254:49,February1992. M.Arnaud,E.Pointecouteau,andG.W.Pratt.CalibrationofthegalaxyclusterM f 500 g - Y f X g relationwithXMM-Newton. AstronomyandAstrophysics ,474:L37{L40,Novem- ber2007.doi:10.1051/0004-6361:20078541. J.N.BahcallandC.L.Sarazin.ParametersandpredictionsfortheX-rayemittinggasof Coma,Perseus,andVirgo. AstrophysicalJournal,LetterstotheEditor ,213:L99{L103, May1977.doi:10.1086/182418. J.A.Baldwin,M.M.Phillips,andR.Terlevich.parametersfortheemission- linespectraofextragalacticobjects. PublicationsoftheASP ,93:5{19,February1981. doi:10.1086/130766. P.Bastien.AlinearpolarizationsurveyofTTauristars. AstrophysicsandSpaceScience, SupplementSeries ,48:153{164,April1982. L.P.BautzandG.O.Abell.ConcerningtheFormsoftheVelocity-DistanceRelation Clusters. AstrophysicalJournal ,184:709{718,September1973.doi:10.1086/152363. E.BertschingerandA.Meiksin.Theroleofheatconductioninthecoolingwsof galaxyclusters. AstrophysicalJournal,LetterstotheEditor ,306:L1{L5,July1986.doi: 10.1086/184692. J.BinneyandG.Tabor.EvolvingCoolingFlows. MonthlyNoticesoftheRAS ,276:663, September1995. L.B^D.A.y,B.R.McNamara,M.W.Wise,andP.E.J.Nulsen.ASystematic StudyofRadio-inducedX-RayCavitiesinClusters,Groups,andGalaxies. Astrophysical Journal ,607:800{809,June2004.doi:10.1086/383519. L.B^B.R.McNamara,P.E.J.Nulsen,C.L.Carilli,andM.W.Wise.Radiative andContentofExtragalacticRadioSources:TowardaUniversalScaling RelationbetweenJetPowerandRadioPower. AstrophysicalJournal ,686:859{880, October2008.doi:10.1086/591416. 158 L.B^D.A.y,P.E.J.Nulsen,B.R.McNamara,H.J.A.ottgering,M.W. Wise,andR.Mittal.Thedutycycleofradio-modefeedbackincompletesamplesof clusters. MonthlyNoticesoftheRAS ,427:3468{3488,December2012.doi:10.1111/j. 1365-2966.2012.22083.x. H.Boehringer,W.Voges,A.C.Fabian,A.C.Edge,andD.M.Neumann.AROSAT HRIstudyoftheinteractionoftheX-ray-emittinggasandradiolobesofNGC1275. MonthlyNoticesoftheRAS ,264:L25{L28,October1993. H.ohringer,P.Schuecker,G.W.Pratt,M.Arnaud,T.J.Ponman,J.H.Croston, S.Borgani,R.G.Bower,U.G.Briel,C.A.Collins,M.Donahue,W.R.Forman, A.Finoguenov,M.J.Geller,L.Guzzo,J.P.Henry,R.Kneissl,J.J.Mohr,K.Mat- sushita,C.R.Mullis,T.Ohashi,K.Pedersen,D.Pierini,H.Quintana,S.Raychaud- hury,T.H.Reiprich,A.K.Romer,P.Rosati,K.Sabirli,R.F.Temple,P.T.P.Viana, A.Vikhlinin,G.M.Voit,andY.-Y.Zhang.TherepresentativeXMM-Newtoncluster structuresurvey(REXCESS)ofanX-rayluminosityselectedgalaxyclustersample. As- tronomyandAstrophysics ,469:363{377,July2007a.doi:10.1051/0004-6361:20066740. H.ohringer,P.Schuecker,G.W.Pratt,M.Arnaud,T.J.Ponman,J.H.Croston, S.Borgani,R.G.Bower,U.G.Briel,C.A.Collins,M.Donahue,W.R.Forman, A.Finoguenov,M.J.Geller,L.Guzzo,J.P.Henry,R.Kneissl,J.J.Mohr,K.Mat- sushita,C.R.Mullis,T.Ohashi,K.Pedersen,D.Pierini,H.Quintana,S.Raychaud- hury,T.H.Reiprich,A.K.Romer,P.Rosati,K.Sabirli,R.F.Temple,P.T.P.Viana, A.Vikhlinin,G.M.Voit,andY.-Y.Zhang.TherepresentativeXMM-Newtoncluster structuresurvey(REXCESS)ofanX-rayluminosityselectedgalaxyclustersample. As- tronomyandAstrophysics ,469:363{377,July2007b.doi:10.1051/0004-6361:20066740. J.N.BregmanandL.P.David.Heatconductionincoolingws. AstrophysicalJournal , 326:639{644,March1988.doi:10.1086/166122. V.Bromm,N.Yoshida,L.Hernquist,andC.F.McKee.Theformationofthestars andgalaxies. Nature ,459:49{54,May2009.doi:10.1038/nature07990. M.Bruggen.SimulationsofBuoyantBubblesinGalaxyClusters. AstrophysicalJournal , 592:839{845,August2003.doi:10.1086/375731. M.BruggenandC.R.Kaiser.Hotbubblesfromactivegalacticnucleiasaheatsourcein cowclusters. Nature ,418:301{303,July2002. J.O.Burns.TheradiopropertiesofcDgalaxiesinAbellclusters.I-anX-rayselected sample. AstronomicalJournal ,99:14{30,January1990.doi:10.1086/115307. D.Calzetti,S.-Y.Wu,S.Hong,R.C.Kennicutt,J.C.Lee,D.A.Dale,C.W.Engelbracht, L.vanZee,B.T.Draine,C.-N.Hao,K.D.Gordon,J.Moustakas,E.J.Murphy, M.Regan,A.Begum,M.Block,J.Dalcanton,J.Funes,A.GildePaz,B.Johnson, S.Sakai,E.Skillman,F.Walter,D.Weisz,B.Williams,andY.Wu.TheCalibration ofMonochromaticFar-InfraredStarFormationRateIndicators. AstrophysicalJournal , 714:1256{1279,May2010.doi:10.1088/0004-637X/714/2/1256. 159 N.Caon,M.Capaccioli,andM.D'Onofrio.OntheShapeoftheLightlesofEarly TypeGalaxies. MonthlyNoticesoftheRAS ,265:1013,December1993. W.Cash.Parameterestimationinastronomythroughapplicationofthelikelihoodratio. AstrophysicalJournal ,228:939{947,March1979.doi:10.1086/156922. B.Catinella,D.Schiminovich,G.S.Fabello,J.Wang,C.Hummels,J.Lemo- nias,S.M.Moran,R.Wu,R.Giovanelli,M.P.Haynes,T.M.Heckman,A.R.Basu- Zych,M.R.Blanton,J.Brinchmann,T.Budaari,T.Gon˘calves,B.D.Johnson,R.C. Kennicutt,B.F.Madore,C.D.Martin,M.R.Rich,L.J.Tacconi,D.A.Thilker, V.Wild,andT.K.Wyder.TheGALEXAreciboSDSSSurvey-I.Gasfractionscaling relationsofmassivegalaxiesanddatarelease. MonthlyNoticesoftheRAS ,403: 683{708,April2010.doi:10.1111/j.1365-2966.2009.16180.x. K.W.Cavagnolo,M.Donahue,G.M.Voit,andM.Sun.BandpassDependenceofX-Ray TemperaturesinGalaxyClusters. AstrophysicalJournal ,682:821{834,August2008a. doi:10.1086/588630. K.W.Cavagnolo,M.Donahue,G.M.Voit,andM.Sun.AnEntropyThresholdforStrong H andRadioEmissionintheCoresofGalaxyClusters. AstrophysicalJournal,Letters totheEditor ,683:L107{L110,August2008b.doi:10.1086/591665. K.W.Cavagnolo,M.Donahue,G.M.Voit,andM.Sun.IntraclusterMediumEntropy foraChandraArchivalSampleofGalaxyClusters. AstrophysicalJournal, SupplementSeries ,182:12{32,May2009.doi:10.1088/0067-0049/182/1/12. E.Churazov,M.Bruggen,C.R.Kaiser,H.ohringer,andW.Forman.Evolutionof BuoyantBubblesinM87. AstrophysicalJournal ,554:261{273,June2001.doi:10.1086/ 321357. J.C.Clemens,J.A.Crain,andR.Anderson.TheGoodmanspectrograph.InA.F.M. MoorwoodandM.Iye,editors, SocietyofPhoto-OpticalInstrumentationEngineers (SPIE)ConferenceSeries ,volume5492of SocietyofPhoto-OpticalInstrumentation Engineers(SPIE)ConferenceSeries ,pages331{340,September2004.doi:10.1117/12. 550069. L.M.Close,F.Roddier,J.L.Hora,J.E.Graves,M.Northcott,C.Roddier,W.F. mann,A.Dayal,G.G.Fazio,andL.K.Deutsch.AdaptiveOpticsInfraredImagingPo- larimetryandOpticalHSTImagingofHubble'sVariableNebula(RMonocerotis/NGC 2261):ACloseLookataVeryYoungActiveHerbigAe/BeStar. AstrophysicalJournal , 489:210,November1997.doi:10.1086/304777. M.Cohen,W.A.Wheaton,andS.T.Megeath.SpectralIrradianceCalibrationinthe Infrared.XIV.TheAbsoluteCalibrationof2MASS. AstronomicalJournal ,126:1090{ 1096,August2003.doi:10.1086/376474. E.Collett. FieldGuidetoPolarizedLight .FieldGuideSeries.SPIE{theInternational SocietyforOpticalEngineering,2005.ISBN9780819458681.URL http://books. google.com/books?id=5lJwcCsLbLsC . 160 J.J.Condon,W.D.Cotton,E.W.Greisen,Q.F.Yin,R.A.Perley,G.B.Taylor,and J.J.Broderick.TheNRAOVLASkySurvey. AstronomicalJournal ,115:1693{1716, May1998.doi:10.1086/300337. L.L.CowieandJ.Binney.Radiativeregulationofgasowwithinclustersofgalaxies-A modelforclusterX-raysources. AstrophysicalJournal ,215:723{732,August1977.doi: 10.1086/155406. L.L.CowieandC.F.McKee.Theevaporationofsphericalcloudsinahotgas.I-Classical andsaturatedmasslossrates. AstrophysicalJournal ,211:135{146,January1977.doi: 10.1086/154911. C.S.Crawford,S.W.Allen,H.Ebeling,A.C.Edge,andA.C.Fabian.TheROSAT BrightestClusterSample-III.Opticalspectraofthecentralclustergalaxies. Monthly NoticesoftheRAS ,306:857{896,July1999.doi:10.1046/j.1365-8711.1999.02583.x. C.S.Crawford,N.A.Hatch,A.C.Fabian,andJ.S.Sanders.TheextendedH -emitting tssurroundingNGC4696,thecentralgalaxyoftheCentauruscluster. Monthly NoticesoftheRAS ,363:216{222,October2005.doi:10.1111/j.1365-2966.2005.09463.x. J.H.Croston,G.W.Pratt,H.ohringer,M.Arnaud,E.Pointecouteau,T.J.Ponman, A.J.R.Sanderson,R.F.Temple,R.G.Bower,andM.Donahue.Galaxy-cluster gas-densitydistributionsoftherepresentativeXMM-Newtonclusterstructuresurvey (REXCESS). AstronomyandAstrophysics ,487:431{443,August2008.doi:10.1051/ 0004-6361:20079154. L.P.David,K.A.Arnaud,W.Forman,andC.Jones.EinsteinobservationsoftheHydra Aclusterandtheofgalaxyformationingroupsandclusters. Astrophysical Journal ,356:32{40,June1990a.doi:10.1086/168812. L.P.David,W.Forman,andC.Jones.Theevolutionoftheinterstellarmediumin ellipticalgalaxies.I-Theearlywindphase. AstrophysicalJournal ,359:29{41,August 1990b.doi:10.1086/169030. L.Davis,Jr.andJ.L.Greenstein.ThePolarizationofStarlightbyAlignedDustGrains. AstrophysicalJournal ,114:206,September1951.doi:10.1086/145464. J.dePlaa,N.Werner,A.Simionescu,J.S.Kaastra,Y.G.Grange,andJ.Vink.Cold frontsandmulti-temperaturestructuresinthecoreofAbell2052. Astronomyand Astrophysics ,523:A81,November2010.doi:10.1051/0004-6361/201015198. J.M.DickeyandF.J.Lockman.HIintheGalaxy. AnnualReviewofAstronomyand Astrophysics ,28:215{261,1990.doi:10.1146/annurev.aa.28.090190.001243. M.DonahueandG.M.Voit.CoolGasinClustersofGalaxies. ClustersofGalaxies: ProbesofCosmologicalStructureandGalaxyEvolution ,page143,2004. 161 M.Donahue,J.Mack,G.M.Voit,W.Sparks,R.Elston,andP.R.Maloney.Hubble SpaceTelescopeObservationsofVibrationallyExcitedMolecularHydrogeninCluster CoolingFlowNebulae. AstrophysicalJournal ,545:670{694,December2000.doi:10. 1086/317836. M.Donahue,D.J.Horner,K.W.Cavagnolo,andG.M.Voit.Entropyinthe CoresofCoolingFlowClustersofGalaxies. AstrophysicalJournal ,643:730{750,June 2006.doi:10.1086/503270. M.Donahue,S.Bruch,E.Wang,G.M.Voit,A.K.Hicks,D.B.Haarsma,J.H. Croston,G.W.Pratt,D.Pierini,R.W.O'Connell,andH.ohringer.StarForma- tionandUVcolorsoftheBrightestClusterGalaxiesintheRepresentativeXMM- NewtonClusterStructureSurvey. AstrophysicalJournal ,715:881{896,June2010.doi: 10.1088/0004-637X/715/2/881. M.Donahue,G.E.deMessieres,R.W.O'Connell,G.M.Voit,A.B.R.McNamara, andP.E.J.Nulsen.PolycyclicAromaticHydrocarbons,IonizedGas,andMolecular HydrogeninBrightestClusterGalaxiesofCool-coreClustersofGalaxies. Astrophysical Journal ,732:40,May2011.doi:10.1088/0004-637X/732/1/40. M.Donahue,G.M.Voit,A.Mahdavi,K.Umetsu,S.Ettori,J.Merten,M.Postman, A.A.Baldi,D.Coe,N.Czakon,M.Bartelmann,N.Benitez,R.Bouwens, L.Bradley,T.Broadhurst,H.Ford,F.Gastaldello,C.Grillo,L.Infante,S.Jou- vel,A.Koekemoer,D.Kelson,O.Lahav,D.Lemze,E.Medezinski,P.Melchior, M.Meneghetti,A.Molino,J.Moustakas,L.A.Moustakas,M.Nonino,P.Rosati, J.Sayers,S.Seitz,A.VanderWel,W.Zheng,andA.Zitrin.CLASH-X:AComparison ofLensingandX-RayTechniquesforMeasuringtheMassofGalaxyClusters. AstrophysicalJournal ,794:136,October2014.doi:10.1088/0004-637X/794/2/136. J.L.Donley,G.H.Rieke,P.G.Palez,andG.Barro.Spitzer'sContribution totheAGNPopulation. AstrophysicalJournal ,687:111{132,November2008.doi: 10.1086/591510. B.T.DraineandA.Li.InfraredEmissionfromInterstellarDust.IV.TheSilicate- Graphite-PAHModelinthePost-SpitzerEra. AstrophysicalJournal ,657:810{837, March2007.doi:10.1086/511055. B.T.DraineandE.E.Salpeter.Onthephysicsofdustgrainsinhotgas. Astrophysical Journal ,231:77{94,July1979.doi:10.1086/157165. J.Dubinski.TheOriginoftheBrightestClusterGalaxies. AstrophysicalJournal ,502: 141,July1998.doi:10.1086/305901. K.EbneterandB.Balick.CentaurusA. PublicationsoftheASP ,95:675{690,October 1983.doi:10.1086/131239. A.C.Edge.Thedetectionofmoleculargasinthecentralgalaxiesofcoolingwclusters. MonthlyNoticesoftheRAS ,328:762{782,December2001.doi:10.1046/j.1365-8711. 2001.04802.x. 162 A.C.Edge,G.C.Stewart,andA.C.Fabian.Propertiesofcoolingwsina sampleofclustersofgalaxies. MonthlyNoticesoftheRAS ,258:177{188,September 1992. A.C.Edge,J.B.R.Oonk,R.Mittal,S.W.Allen,S.A.Baum,H.ohringer,J.N. Bregman,M.N.Bremer,F.Combes,C.S.Crawford,M.Donahue,E.Egami,A.C. Fabian,G.J.Ferland,S.L.Hamer,N.A.Hatch,W.R.M.Johnstone,B.R. McNamara,C.P.O'Dea,P.Popesso,A.C.Quillen,P.Salome,C.L.Sarazin,G.M. Voit,R.J.Wilman,andM.W.Wise.HerschelobservationsofFIRemissionlines inbrightestclustergalaxies. AstronomyandAstrophysics ,518:L46,July2010a.doi: 10.1051/0004-6361/201014569. A.C.Edge,J.B.R.Oonk,R.Mittal,S.W.Allen,S.A.Baum,H.ohringer,J.N. Bregman,M.N.Bremer,F.Combes,C.S.Crawford,M.Donahue,E.Egami,A.C. Fabian,G.J.Ferland,S.L.Hamer,N.A.Hatch,W.R.M.Johnstone,B.R. McNamara,C.P.O'Dea,P.Popesso,A.C.Quillen,P.Salome,C.L.Sarazin,G.M. Voit,R.J.Wilman,andM.W.Wise.Herschelphotometryofbrightestclustergalaxies incoolingwclusters. AstronomyandAstrophysics ,518:L47,July2010b.doi:10. 1051/0004-6361/201014572. E.Egami,G.H.Rieke,D.Fadda,andD.C.Hines.ALargeMassofH 2 intheBrightest ClusterGalaxyinZwicky3146. AstrophysicalJournal,LetterstotheEditor ,652:L21{ L24,November2006.doi:10.1086/509886. R.ElstonandP.Maloney.MolecularHydrogenEmissionfromCentralClusterGalaxiesin CoolingFlows.InI.S.McLean,editor, AstronomywithArrays,TheNextGeneration , volume190of AstrophysicsandSpaceScienceLibrary ,page169,1994. A.C.Fabian.CoolingFlowsinClustersofGalaxies. AnnualReviewofAstronomyand Astrophysics ,32:277{318,1994a.doi:10.1146/annurev.aa.32.090194.001425. A.C.Fabian.CoolingFlowsinClustersofGalaxies. AnnualReviewofAstronomyand Astrophysics ,32:277{318,1994b.doi:10.1146/annurev.aa.32.090194.001425. A.C.FabianandP.E.J.Nulsen.Subsonicaccretionofcoolinggasinclustersofgalaxies. MonthlyNoticesoftheRAS ,180:479{484,August1977. A.C.Fabian,J.S.Sanders,S.Ettori,G.B.Taylor,S.W.Allen,C.S.Crawford,K.Iwa- sawa,R.M.Johnstone,andP.M.Ogle.ChandraimagingofthecomplexX-raycore ofthePerseuscluster. MonthlyNoticesoftheRAS ,318:L65{L68,November2000.doi: 10.1046/j.1365-8711.2000.03904.x. A.C.Fabian,J.S.Sanders,G.B.Taylor,S.W.Allen,C.S.Crawford,R.M.Johnstone, andK.Iwasawa.AverydeepChandraobservationofthePerseuscluster:shocks, ripplesandconduction. MonthlyNoticesoftheRAS ,366:417{428,February2006.doi: 10.1111/j.1365-2966.2005.09896.x. 163 A.C.Fabian,J.S.Sanders,R.J.R.Williams,A.Lazarian,G.J.Ferland,andR.M. Johnstone.TheenergysourceofthetsaroundthegiantgalaxyNGC1275. MonthlyNoticesoftheRAS ,417:172{177,October2011.doi:10.1111/j.1365-2966. 2011.19034.x. G.G.Fazio,J.L.Hora,L.E.Allen,M.L.N.Ashby,P.Barmby,L.K.Deutsch,J.-S. Huang,S.Kleiner,M.Marengo,S.T.Megeath,G.J.Melnick,M.A.Pahre,B.M. Patten,J.Polizotti,H.A.Smith,R.S.Taylor,Z.Wang,S.P.Willner,W.F. mann,J.L.Pipher,W.J.Forrest,C.W.McMurty,C.R.McCreight,M.E.McKelvey, R.E.McMurray,D.G.Koch,S.H.Moseley,R.G.Arendt,J.E.Mentzell,C.T. Marx,P.Losch,P.Mayman,W.Eichhorn,D.Krebs,M.Jhabvala,D.Y.Gezari,D.J. Fixsen,J.Flores,K.Shakoorzadeh,R.Jungo,C.Hakun,L.Workman,G.Karpati, R.Kichak,R.Whitley,S.Mann,E.V.Tollestrup,P.Eisenhardt,D.Stern,V.Gorjian, B.Bhattacharya,S.Carey,B.O.Nelson,W.J.Glaccum,M.Lacy,P.J.Lowrance, S.Laine,W.T.Reach,J.A.J.A.Surace,G.Wilson,E.L.Wright,A. man,G.Domingo,andM.Cohen.TheInfraredArrayCamera(IRAC)fortheSpitzer SpaceTelescope. AstrophysicalJournal,SupplementSeries ,154:10{17,September2004. doi:10.1086/422843. G.J.Ferland,A.C.Fabian,N.A.Hatch,R.M.Johnstone,R.L.Porter,P.A.M.van Hoof,andR.J.R.Williams.Collisionalheatingastheoriginoftemission ingalaxyclusters. MonthlyNoticesoftheRAS ,392:1475{1502,February2009.doi: 10.1111/j.1365-2966.2008.14153.x. C.Ferrari,M.Arnaud,S.Ettori,S.Maurogordato,andJ.Rho.Chandraobservation ofthemultiplemergerclusterAbell521. AstronomyandAstrophysics ,446:417{428, February2006.doi:10.1051/0004-6361:20053946. A.H.Gonzalez,A.I.ZabluandD.Zaritsky.IntraclusterLightinNearbyGalaxy Clusters:RelationshiptotheHalosofBrightestClusterGalaxies. AstrophysicalJournal , 618:195{213,January2005.doi:10.1086/425896. A.H.Gonzalez,D.Zaritsky,andA.I.ZabluACensusofBaryonsinGalaxyClusters andGroups. AstrophysicalJournal ,666:147{155,September2007.doi:10.1086/519729. N.GrevesseandA.J.Sauval.StandardSolarComposition. SpaceScienceReviews ,85: 161{174,May1998.doi:10.1023/A:1005161325181. B.Groves,M.A.Dopita,R.S.Sutherland,L.J.Kewley,J.Fischera,C.Leitherer, B.Brandl,andW.vanBreugel.ModelingthePan-SpectralEnergyDistributionof StarburstGalaxies.IV.TheControllingParametersoftheStarburstSED. Astrophysi- calJournal,SupplementSeries ,176:438{456,June2008.doi:10.1086/528711. H.Gursky,E.Kellogg,S.Murray,C.Leong,H.Tananbaum,andR.Giacconi.AStrong X-RaySourceintheComaClusterObservedbyUHURU. AstrophysicalJournal,Letters totheEditor ,167:L81,August1971.doi:10.1086/180765. 164 M.Hamuy,N.B.SunS.R.Heathcote,A.R.Walker,P.Gigoux,andM.M.Phillips. Southernspectrophotometricstandards,2. PublicationsoftheASP ,106:566{589,June 1994.doi:10.1086/133417. T.M.Heckman,S.A.Baum,W.J.M.vanBreugel,andP.McCarthy.Dynamical, physical,andchemicalpropertiesofemission-linenebulaeincoolingws. Astrophysical Journal ,338:48{77,March1989.doi:10.1086/167181. J.-C.Henoux,G.Chambe,D.Heristchi,M.Semel,B.Woodgate,D.Shine,andJ.Beckers. Conductiveinthechromospherederivedfromlinelinearpolarizationobservation. SolarPhysics ,86:115{122,July1983a.doi:10.1007/BF00157180. J.C.Henoux,D.Heristchi,G.Chambe,B.Woodgate,R.Shine,J.Beckers,and M.Machado.Conductiveinsolarchromospheresdeducedfromthelinear polarizationobservations. AstronomyandAstrophysics ,119:233{238,March1983b. R.C.HickoxandM.Markevitch.AbsoluteMeasurementoftheUnresolvedCosmicX-Ray Backgroundinthe0.5-8keVBandwithChandra. AstrophysicalJournal ,645:95{114, July2006.doi:10.1086/504070. A.K.Hicks,R.Mushotzky,andM.Donahue.DetectingStarFormationinBrightest ClusterGalaxieswithGALEX. AstrophysicalJournal ,719:1844{1858,August2010. doi:10.1088/0004-637X/719/2/1844. P.HicksonandS.vandenBergh.CCDobservationsofthepolarizationoftheCrab Nebula. AstrophysicalJournal ,365:224{229,December1990.doi:10.1086/169473. J.G.Hoessel.Thephotometricpropertiesofbrightestclustergalaxies.II-SITandCCD surfacephotometry. AstrophysicalJournal ,241:493{506,October1980.doi:10.1086/ 158364. J.G.Hoessel,J.E.Gunn,andT.X.Thuan.Thephotometricpropertiesofbrightest clustergalaxies.I-Absolutemagnitudesin116nearbyAbellclusters. Astrophysical Journal ,241:486{492,October1980.doi:10.1086/158363. A.S.HoM.Donahue,A.Hicks,andR.S.Barthelemy.InfraredandUltraviolet StarFormationinBrightestClusterGalaxiesintheACCEPTSample. Astrophysical Journal,SupplementSeries ,199:23,March2012a.doi:10.1088/0067-0049/199/1/23. A.S.M.Donahue,A.Hicks,andR.S.Barthelemy.IRandUVstarformation inACCEPTBCGs2012). VizieROnlineDataCatalog ,219:90023,March 2012b. E.M.Hu,L.L.Cowie,andZ.Wang.Long-slitspectroscopyofgasinthecoresofX- rayluminousclusters. AstrophysicalJournal,SupplementSeries ,59:447{498,December 1985.doi:10.1086/191081. W.andM.N.Bremer.Molecularhydrogenemissionincoolingws. MonthlyNotices oftheRAS ,284:L1{L5,January1997. 165 T.H.Jarrett,T.Chester,R.Cutri,S.E.Schneider,andJ.P.Huchra.The2MASSLarge GalaxyAtlas. AstronomicalJournal ,125:525{554,February2003.doi:10.1086/345794. R.I.Jedrzejewski.CCDsurfacephotometryofellipticalgalaxies.I-Observations,reduc- tionandresults. MonthlyNoticesoftheRAS ,226:747{768,June1987. B.D.Johnson,D.Schiminovich,M.Seibert,M.Treyer,D.C.Martin,T.A.Barlow, K.Forster,P.G.Friedman,P.Morrissey,S.G.T.Small,T.K.Wyder,L.Bianchi, J.Donas,T.M.Heckman,Y.-W.Lee,B.F.Madore,B.Milliard,R.M.Rich,A.S. Szalay,B.Y.Welsh,andS.K.Yi.Ultraviolet,Optical,andInfraredConstraintson ModelsofStellarPopulationsandDustAttenuation. AstrophysicalJournal,Supplement Series ,173:377{391,December2007a.doi:10.1086/522932. B.D.Johnson,D.Schiminovich,M.Seibert,M.Treyer,D.C.Martin,T.A.Barlow, K.Forster,P.G.Friedman,P.Morrissey,S.G.T.Small,T.K.Wyder,L.Bianchi, J.Donas,T.M.Heckman,Y.-W.Lee,B.F.Madore,B.Milliard,R.M.Rich,A.S. Szalay,B.Y.Welsh,andS.K.Yi.UltravioletthroughInfraredSpectralEnergyDistribu- tionsfrom1000SDSSGalaxies:DustAttenuation. AstrophysicalJournal,Supplement Series ,173:392{403,December2007b.doi:10.1086/522960. H.M.Johnson.Hubble'svariablenebulaNGC2261. AstronomicalJournal ,71:224{227, April1966.doi:10.1086/109910. J.S.KaastraandR.Mewe.X-rayemissionfromthinplasmas.I-MultipleAugerionisation andprocessesforBetoZn. AstrophysicsandSpaceScience,Supplement Series ,97:443{482,January1993. E.KelloggandS.Murray.StudiesofclusterX-raysources-Sizemeasurements. Astrophys- icalJournal,LetterstotheEditor ,193:L57{L60,October1974.doi:10.1086/181630. R.C.Kennicutt,Jr.Therateofstarformationinnormaldiskgalaxies. Astrophysical Journal ,272:54{67,September1983.doi:10.1086/161261. R.C.Kennicutt,Jr.StarFormationinGalaxiesAlongtheHubbleSequence. Annual ReviewofAstronomyandAstrophysics ,36:189{232,1998.doi:10.1146/annurev.astro. 36.1.189. R.C.Kennicutt,Jr.,L.Armus,G.Bendo,D.Calzetti,D.A.Dale,B.T.Draine,C.W. Engelbracht,K.D.Gordon,A.D.Grauer,G.Helou,D.J.Hollenbach,T.H.Jarrett, L.J.Kewley,C.Leitherer,A.Li,S.Malhotra,M.W.Regan,G.H.Rieke,M.J. Rieke,H.Roussel,J.-D.T.Smith,M.D.Thornley,andF.Walter.SINGS:TheSIRTF NearbyGalaxiesSurvey. PublicationsoftheASP ,115:928{952,August2003.doi: 10.1086/376941. J.M.Laming.Emission-linepolarizationasanequilibrationdiagnosticfornonradiative shockfronts. AstrophysicalJournal ,362:219{225,October1990.doi:10.1086/169257. 166 A.LazarianandP.Desiati.MagneticReconnectionastheCauseofCosmicRayExcess fromtheHeliosphericTail. AstrophysicalJournal ,722:188{196,October2010.doi: 10.1088/0004-637X/722/1/188. S.M.Lea,J.Silk,E.Kellogg,andS.Murray.Thermal-BremsstrahlungInterpretationof ClusterX-RaySources. AstrophysicalJournal,LetterstotheEditor ,184:L105,Septem- ber1973.doi:10.1086/181300. L.Leisman,D.B.Haarsma,D.A.Sebald,andACCEPTTeam.BrightestClusterGalaxy IdenIn AmericanAstronomicalSocietyMeetingAbstracts#217 ,volume43 of BulletinoftheAmericanAstronomicalSociety ,page149.08,January2011. D.A.Liedahl,A.L.Osterheld,andW.H.Goldstein.NewcalculationsofFeL-shellX-ray spectrainhigh-temperatureplasmas. AstrophysicalJournal,LetterstotheEditor ,438: L115{L118,January1995.doi:10.1086/187729. J.Magorrian,S.Tremaine,D.Richstone,R.Bender,G.Bower,A.Dressler,S.M.Faber, K.Gebhardt,R.Green,C.Grillmair,J.Kormendy,andT.Lauer.TheDemographyof MassiveDarkObjectsinGalaxyCenters. AstronomicalJournal ,115:2285{2305,June 1998.doi:10.1086/300353. A.Mahdavi,H.Hoekstra,A.Babul,J.Sievers,S.T.Myers,andJ.P.Henry.Joint AnalysisofClusterObservations.I.MassofAbell478fromCombinedX-Ray, Sunyaev-Zel'dovich,andWeak-LensingData. AstrophysicalJournal ,664:162{180,July 2007.doi:10.1086/517958. A.Mahdavi,H.Hoekstra,A.Babul,C.Bildfell,T.Jeltema,andJ.P.Henry.Joint AnalysisofClusterObservations.II.Chandra/XMM-NewtonX-RayandWeakLensing ScalingRelationsforaSampleof50RichClustersofGalaxies. AstrophysicalJournal , 767:116,April2013.doi:10.1088/0004-637X/767/2/116. D.MakovozandF.R.Marleau.Point-SourceExtractionwithMOPEX. Publicationsof theASP ,117:1113{1128,October2005.doi:10.1086/432977. D.C.Martin,J.Fanson,D.Schiminovich,P.Morrissey,P.G.Friedman,T.A.Barlow, T.Conrow,R.Grange,P.N.Jelinsky,B.Milliard,O.H.W.Siegmund,L.Bianchi,Y.-I. Byun,J.Donas,K.Forster,T.M.Heckman,Y.-W.Lee,B.F.Madore,R.F.Malina, S.G.R.M.Rich,T.Small,F.Surber,A.S.Szalay,B.Welsh,andT.K.Wyder. TheGalaxyEvolutionExplorer:ASpaceUltravioletSurveyMission. Astrophysical Journal,LetterstotheEditor ,619:L1{L6,January2005.doi:10.1086/426387. W.G.MathewsandJ.N.Bregman.Radiativeaccretionwontogiantgalaxiesinclusters. AstrophysicalJournal ,224:308{319,September1978.doi:10.1086/156379. B.J.Maughan,P.A.Giles,S.W.Randall,C.Jones,andW.R.Forman.Self-similarscal- ingandevolutioninthegalaxyclusterX-rayluminosity-temperaturerelation. Monthly NoticesoftheRAS ,421:1583{1602,April2012.doi:10.1111/j.1365-2966.2012.20419.x. 167 B.R.McNamaraandP.E.J.Nulsen.HeatingHotAtmosphereswithActiveGalactic Nuclei. AnnualReviewofAstronomyandAstrophysics ,45:117{175,September2007. doi:10.1146/annurev.astro.45.051806.110625. B.R.McNamara,M.Wise,C.L.Sarazin,B.T.Jannuzi,andR.Elston.OpticalStructure intheAbell1795ClusterCentralGalaxy:EvidenceforStrippingandof RadioJets. AstrophysicalJournal,LetterstotheEditor ,466:L9,July1996.doi:10. 1086/310162. B.R.McNamara,M.Wise,P.E.J.Nulsen,L.P.David,C.L.Sarazin,M.Bautz, M.Markevitch,A.Vikhlinin,W.R.Forman,C.Jones,andD.E.Harris.Chandra X-RayObservationsoftheHydraACluster:AnInteractionbetweentheRadioSource andtheX-Ray-emittingGas. AstrophysicalJournal,LetterstotheEditor ,534:L135{ L138,May2000.doi:10.1086/312662. B.R.McNamara,M.W.Wise,P.E.J.Nulsen,L.P.David,C.L.Carilli,C.L.Sarazin, C.P.O'Dea,J.Houck,M.Donahue,S.Baum,M.Voit,R.W.O'Connell,andA.Koeke- moer.DiscoveryofGhostCavitiesintheX-RayAtmosphereofAbell2597. Astrophysical Journal,LetterstotheEditor ,562:L149{L152,December2001.doi:10.1086/338326. B.R.McNamara,P.E.J.Nulsen,M.W.Wise,D.A.y,C.Carilli,C.L.Sarazin, andE.L.Blanton.TheheatingofgasinagalaxyclusterbyX-raycavitiesandlarge- scaleshockfronts. Nature ,433:45{47,January2005.doi:10.1038/nature03202. R.Mewe,E.H.B.M.Gronenschild,andG.H.J.vandenOord.CalculatedX-radiation fromopticallythinplasmas.V. AstrophysicsandSpaceScience,SupplementSeries ,62: 197{254,November1985. R.Mewe,J.R.Lemen,andG.H.J.vandenOord.CalculatedX-radiationfromoptically thinplasmas.VI-Improvedcalculationsforcontinuumemissionandapproximation formulaefornonrelativisticaverageGauntfactors. AstrophysicsandSpaceScience, SupplementSeries ,65:511{536,September1986. P.Morrissey,T.Conrow,T.A.Barlow,T.Small,M.Seibert,T.K.Wyder,T.Budaari, S.Arnouts,P.G.Friedman,K.Forster,D.C.Martin,S.G.D.Schiminovich, L.Bianchi,J.Donas,T.M.Heckman,Y.-W.Lee,B.F.Madore,B.Milliard,R.M. Rich,A.S.Szalay,B.Y.Welsh,andS.K.Yi.TheCalibrationandDataProductsof GALEX. AstrophysicalJournal,SupplementSeries ,173:682{697,December2007.doi: 10.1086/520512. R.F.MushotzkyandM.Loewenstein.LackofEvolutionintheIronAbundanceinClusters ofGalaxiesandImplicationsfortheGlobalStarFormationRateatHighRedshift. AstrophysicalJournal,LetterstotheEditor ,481:L63,June1997.doi:10.1086/310651. R.W.O'Connell.Far-UltravioletRadiationfromEllipticalGalaxies. AnnualReviewof AstronomyandAstrophysics ,37:603{648,1999.doi:10.1146/annurev.astro.37.1.603. 168 C.P.O'Dea,S.A.Baum,G.Privon,J.Noel-Storr,A.C.Quillen,N.Zufelt,J.Park, A.Edge,H.Russell,A.C.Fabian,M.Donahue,C.L.Sarazin,B.McNamara,J.N. Bregman,andE.Egami.AnInfraredSurveyofBrightestClusterGalaxies.II.Whyare SomeBrightestClusterGalaxiesFormingStars? AstrophysicalJournal ,681:1035{1045, July2008.doi:10.1086/588212. K.P.O'Dea,A.C.Quillen,C.P.O'Dea,G.R.Tremblay,B.T.Snios,S.A.Baum, K.Christiansen,J.Noel-Storr,A.C.Edge,M.Donahue,andG.M.Voit.HubbleSpace TelescopeFar-ultravioletObservationsofBrightestClusterGalaxies:TheRoleofStar FormationinCoolingFlowsandBCGEvolution. AstrophysicalJournal ,719:1619{1632, August2010.doi:10.1088/0004-637X/719/2/1619. E.S.Perlman,J.A.Biretta,F.Zhou,W.B.Sparks,andF.D.Macchetto.Optical andRadioPolarimetryoftheM87Jetat0.2"Resolution. AstronomicalJournal ,117: 2185{2198,May1999.doi:10.1086/300844. J.R.PetersonandA.C.Fabian.X-rayspectroscopyofcoolingclusters. PhysicsReports , 427:1{39,April2006.doi:10.1016/j.physrep.2005.12.007. J.R.Peterson,S.M.Kahn,F.B.S.Paerels,J.S.Kaastra,T.Tamura,J.A.M.Bleeker, C.Ferrigno,andJ.G.Jernigan.High-ResolutionX-RaySpectroscopicConstraintson Cooling-FlowModelsforClustersofGalaxies. AstrophysicalJournal ,590:207{224,June 2003.doi:10.1086/374830. M.PostmanandT.R.Lauer.Brightestclustergalaxiesasstandardcandles. Astrophysical Journal ,440:28{47,February1995.doi:10.1086/175245. M.Postman,D.Coe,N.BenL.Bradley,T.Broadhurst,M.Donahue,H.Ford, O.Graur,G.Graves,S.Jouvel,A.Koekemoer,D.Lemze,E.Medezinski,A.Molino, L.Moustakas,S.Ogaz,A.Riess,S.Rodney,P.Rosati,K.Umetsu,W.Zheng,A.Zitrin, M.Bartelmann,R.Bouwens,N.Czakon,S.Golwala,O.Host,L.Infante,S.Jha, Y.Jimenez-Teja,D.Kelson,O.Lahav,R.Lazkoz,D.Maoz,C.McCully,P.Melchior, M.Meneghetti,J.Merten,J.Moustakas,M.Nonino,B.Patel,E.os,J.Sayers, S.Seitz,andA.VanderWel.TheClusterLensingandSupernovaSurveywithHubble: AnOverview. AstrophysicalJournal,SupplementSeries ,199:25,April2012.doi:10. 1088/0067-0049/199/2/25. G.W.Pratt,J.H.Croston,M.Arnaud,andH.ohringer.GalaxyclusterX-raylumi- nosityscalingrelationsfromarepresentativelocalsample(REXCESS). Astronomyand Astrophysics ,498:361{378,May2009a.doi:10.1051/0004-6361/200810994. G.W.Pratt,J.H.Croston,M.Arnaud,andH.ohringer.GalaxyclusterX-raylumi- nosityscalingrelationsfromarepresentativelocalsample(REXCESS). Astronomyand Astrophysics ,498:361{378,May2009b.doi:10.1051/0004-6361/200810994. A.C.Quillen,N.Zufelt,J.Park,C.P.O'Dea,S.A.Baum,G.Privon,J.Noel-Storr, A.Edge,H.Russell,A.Fabian,M.Donahue,J.N.Bregman,B.R.McNamara,and 169 C.L.Sarazin.AnInfraredSurveyofBrightestClusterGalaxies.I. AstrophysicalJournal, SupplementSeries ,176:39{58,May2008.doi:10.1086/525560. D.A.y,B.R.McNamara,P.E.J.Nulsen,andM.W.Wise.TheFeedback- regulatedGrowthofBlackHolesandBulgesthroughGasAccretionandStarbursts inClusterCentralDominantGalaxies. AstrophysicalJournal ,652:216{231,November 2006.doi:10.1086/507672. D.A.y,B.R.McNamara,andP.E.J.Nulsen.TheRegulationofCoolingandStar FormationinLuminousGalaxiesbyActiveGalacticNucleusFeedbackandtheCooling- Time/EntropyThresholdfortheOnsetofStarFormation. AstrophysicalJournal ,687: 899{918,November2008.doi:10.1086/591240. S.W.Randall,C.Jones,M.Markevitch,E.L.Blanton,P.E.J.Nulsen,andW.R. Forman.GasSloshingandBubblesintheGalaxyGroupNGC5098. Astrophysical Journal ,700:1404{1414,August2009.doi:10.1088/0004-637X/700/2/1404. W.T.Reach,S.T.Megeath,M.Cohen,J.Hora,S.Carey,J.Surace,S.P.Willner, P.Barmby,G.Wilson,W.Glaccum,P.Lowrance,M.Marengo,andG.G.Fazio. AbsoluteCalibrationoftheInfraredArrayCameraontheSpitzerSpaceTelescope. PublicationsoftheASP ,117:978{990,September2005.doi:10.1086/432670. T.H.ReiprichandH.ohringer.TheMassFunctionofanX-RayFlux-limitedSampleof GalaxyClusters. AstrophysicalJournal ,567:716{740,March2002.doi:10.1086/338753. G.H.Rieke,E.T.Young,C.W.Engelbracht,D.M.Kelly,F.J.Low,E.E.Haller,J.W. Beeman,K.D.Gordon,J.A.Stansberry,K.A.Misselt,J.Cadien,J.E.Morrison, G.Rivlis,W.B.Latter,A.Noriega-Crespo,D.L.Padgett,K.R.Stapelfeldt,D.C. Hines,E.Egami,J.Muzerolle,A.Alonso-Herrero,M.Blaylock,H.Dole,J.L.Hinz, E.LeFloc'h,C.Papovich,P.G.Palez,P.S.Smith,K.Y.L.Su,L.Bennett, D.T.Frayer,D.Henderson,N.Lu,F.Masci,M.Pesenson,L.Rebull,J.Rho,J.Keene, S.Stolovy,S.Wachter,W.Wheaton,M.W.Werner,andP.L.Richards.TheMultiband ImagingPhotometerforSpitzer(MIPS). AstrophysicalJournal,SupplementSeries ,154: 25{29,September2004.doi:10.1086/422717. G.B.RybickiandA.P.Lightman. Radiativeprocessesinastrophysics .1979. S.Salim,R.M.Rich,S.Charlot,J.Brinchmann,B.D.Johnson,D.Schiminovich,M.Seib- ert,R.Mallery,T.M.Heckman,K.Forster,P.G.Friedman,D.C.Martin,P.Morrissey, S.G.T.Small,T.K.Wyder,L.Bianchi,J.Donas,Y.-W.Lee,B.F.Madore, B.Milliard,A.S.Szalay,B.Y.Welsh,andS.K.Yi.UVStarFormationRatesin theLocalUniverse. AstrophysicalJournal,SupplementSeries ,173:267{292,December 2007.doi:10.1086/519218. A.J.R.Sanderson,A.C.Edge,andG.P.Smith.LoCuSS:theconnectionbetween brightestclustergalaxyactivity,gascoolinganddynamicaldisturbanceofX-raycluster cores. MonthlyNoticesoftheRAS ,398:1698{1705,October2009.doi:10.1111/j. 1365-2966.2009.15214.x. 170 C.L.Sarazin. X-rayemissionfromclustersofgalaxies .1988. C.L.SarazinandJ.N.Bahcall.X-raylineemissionforclustersofgalaxies.II-Numerical models. AstrophysicalJournal,SupplementSeries ,34:451{467,August1977.doi:10. 1086/190457. D.J.Schlegel,D.P.Finkbeiner,andM.Davis.MapsofDustInfraredEmissionforUsein EstimationofReddeningandCosmicMicrowaveBackgroundRadiationForegrounds. AstrophysicalJournal ,500:525,June1998.doi:10.1086/305772. G.D.Schmidt,J.R.P.Angel,andE.A.Beaver.Thesmall-scalepolarizationoftheCrab Nebula. AstrophysicalJournal ,227:106{113,January1979.doi:10.1086/156708. B.Sheehan,G.Kyne,P.Collins,M.Redfern,andA.Shearer.Detectorsfor High-SpeedPolarimetricObservationsUsingGASP.In ProceedingsofHigh TimeResolutionAstrophysics-TheEraofExtremelyLargeTelescopes(HTRA- IV).May5-7,2010.AgiosNikolaos,CreteGreece.Publishedonlineat