REDSHIFTRADIO-LOUDQUASARS1
MatthewD.Lehnert2,3
SterrewachtLeiden,Postbus9513,2300RALeiden,TheNetherlands
WilJ.M.vanBreugel
InstituteofGeophysics&PlanetaryPhysics,LawrenceLivermoreNationalLaboratory,
L-413P.O.Box808,Livermore,CA94550TimothyM.Heckman
DepartmentofPhysicsandAstronomy,JohnsHopkinsUniversity,Baltimore,MD21218
and
GeorgeK.Miley
SterrewachtLeiden,Postbus9513,2300RALeiden,TheNetherlands
Received........................;accepted........................
–2–ABSTRACT
Wepresentrest-frameUVandLyαimagesofspatially-resolvedstructures(‘hosts’)aroundfivehigh-redshiftradio-loudquasarsobtainedwiththeWFPC2cameraontheHubbleSpaceTelescope.ThequasarswereimagedwiththePC1throughtheF555W(‘V’-band)filter,whichattheredshiftsofthequasars(2.1 TheUVcontinuummorphologiesofthehostsappearcomplexandknottyattherelativelyhighsurfacebrightnesslevelsofourexposures(about24Vmagsarcsec−2).Intwoquasarswefindevidenceforforegroundgalaxieswhichconfusethehostgalaxymorphologiesandwhichareresponsibleforsomeoftheperceivedradio/opticalmis-alignmentsobservedinground-basedimagingdata.WedofindgoodalignmentbetweentheextendedLyαandtheradiosources,strongevidenceforjet-cloudinteractionsintwocases,againresemblingradiogalaxies,andwhatispossiblythemostluminousradio-UVsynchrotronjetinoneofthehostsatz=2.110.Wediscussthesignificanceofjet-cloudcollisionsinradio-loudquasarsandtheirinfluenceonradiomorphologiesintheframeworkofsimpleorientation-basedquasar/radiogalaxyunificationschemes. Ourobservationssuggestthatthehostgalaxiesofradio-loudsteepspectrumquasarsaresimilartothoseofradiogalaxiesandstrengthenpreviousconclusions –3– basedonground-baseddatathatbothtypesofobjectsareprobablymembersofthesameparentpopulation. Subjectheadings:Galaxies:evolution—galaxies:jets—quasars:hostgalaxies—radiocontinuum:galaxies –4– 1. Introduction Itisnoexaggerationtosaythatbetweenredshiftsof≈2-3tothepresent,highlyluminous(M<−26)quasarshavebecomevirtuallyextinct.Theco-movingspacedensityofluminousquasarshasfallenbyaboutafactorof1000betweenthesetwoepochs(e.g.,Hartwick&Schade1990;Boyle1993).Whatprocessesledtosuchastrongdensityevolutionofhighlyluminousactivegalacticnuclei(AGN)?TheanswertothisquestionnotonlyhasimportantramificationsforourunderstandingoftheAGNphenomenon,butwillalmostcertainlygiveusimportantinsightintotheprocessesthatcontrolledtheevolutionofgalaxiesingeneralsincethe“quasarepoch”. ItisanintriguingpossibilitythatthereisastronglinkbetweenthefuelingofluminousAGNandgalaxyformation.Indeed,thequasarepochoccursatthetime“ColdDarkMatter”modelsidentifyaswhentypicalpresent-daygalaxieswherehierarchicallyassembledviadissipativemergers.Rees(1988),withoutregardtoanyspecificcosmogony,arguesthattheprocessofgalaxyformationwasstillcontinuingattheepochz=2.Thisisespeciallytrueofthetheoreticalmodelswheregalaxyformationisaslow,diffuse,low-luminosityprocess(e.g.,Baron&White1987;Kauffmann,White,&Guiderdoni1993;Baughetal.1998).ThesegeneralspeculationsthereforegiveaspecialprominencetohighredshiftAGNbyhypothesizingthatonlywhentheproto-galacticmaterialisenergizedbytheluminosityofanAGNwillithavehighenoughsurfacebrightnesstobereadilydetectableinemission-linesurveys.Inturn,theprocessofgalaxyformationmaybesubstantiallyalteredbytheeffectoftheAGN. Observationsofradio-loudquasarsarealsovaluableforwhatwecanlearnabouttheAGNphenomenoningeneral.Inparticular,therelationshipbetweenradio-loudquasarsandradiogalaxiesisofconsiderableinterest,especiallyinlightofeffortsto“unify”thesetwoclassesthroughdifferencesinviewingangle,environment,orevolutionarystate(e.g.,Barthel1989;Norman&Miley1984;Neff&Hutchings1990).Inparticular,the“viewingangle”schemeofBarthel(1989)inwhichradio-loudquasarsandradiogalaxiesaredrawnfromthesameparentpopulationbutviewedpreferentiallyatsmallorlargeangles,respectively,totheradioaxis,predictsthattheluminosityandcolorofthequasarhostshouldbeverysimilar,ifnotidentical,tothoseoftheradiogalaxiesatsimilarradiopowersandredshifts.Moreover,radiogalaxiesathighredshifts(z>0.6)exhibittheso-called“alignmenteffect”(McCarthyetal.1987;Chambersetal.1987)wheretheradio,andtherest-frameUVandopticalaxesareallroughlyco-linear.Recently,throughtheuseofbroad-bandHSTimagingdata,ithasbecomeclearthatquasarsdoindeedexhibitthe“alignmenteffect”(Lehnertetal.1999a)andthattheeffectinquasarsisweakerthanthatseeninradiogalaxies,asexpectedinsimpleorientation-basedunificationschemes(Lehnertetal.1999binpreparation). –5– Radio-loudquasarsatredshiftsof2-3haveimpressivelylargeandluminouscontinuumandemission-linenebulae(“hosts”;Heckmanetal.1991a,b;Lehnertetal.1992).Thesestructurescompriseabout3%toover40%ofthetotalfluxfromthequasarandhaveLyαluminositiesof≈1044−45ergss−1,absolutevisualmagnitudesofabout−25,bluespectralenergydistributionsconsistentwiththoseofnearbylate-typegalaxies(ScandIrr),andsizesoftheorderofmanytensofkiloparsecs.Ground-basedimagesofhigh-zradioloudquasarscontainonlylimitedmorphologicalinformation(1arcsecondseeingcorrespondstoabout11kpcattheseredshifts)andtheirmorphologycanbestbedescribedasasymmetric.Hostgalaxiesoflow-redshiftquasars(e.g.,Boroson&Oke1984;Boroson,Persson,&Oke1985;Smithetal.1986)haveemission-lineluminositiesmuchlowerthanthis(about1042ergss−1in[OII]),spectralenergydistributionssimilartolate-typegalaxies,andabsolutemagnitudesofaround−22.However,Bahcalletal.(1994,1995a,b)haverecentlyconcludedusingHSTWFPC2imagesthatthehostgalaxiesofsomeofthebrightest,low-redshiftQSOshavewiderangeofhostluminosities,withperhapsalargefractionofquasarshostshavingluminosities Inthispaperwerepresentananalysisoftheseobservations,usingseveraldifferentmethodstotesttherobustnessofourresults,andcomparethequasarhostpropertieswiththoseofradiogalaxies.ThroughoutthispaperweadoptH0=50kms−1Mpc−1,q0=0.1,andΛ=0foreasycomparisontopreviouslypublishedworkonhighredshiftradiogalaxies. –6– 2. ObservationsandReduction TheHSTobservationsofoursampleweremadefromSeptember1994throughMay1995usingtheWideFieldPlanetaryCamera2(WFPC2).EachquasarwascenteredonthePCandwasobservedforatotalof2100seconds(3×700seconds)throughtheF555Wfilter(whichcloselycorrespondstoaground-basedVfilter).Inaddition,weobtainedexposuresof5000seconds(5x1000seconds)usingtheWideFieldCamera(WFC)andoneoftheQuad[OII]filters.Thenarrow-bandfilterwasselectedsuchthatitscentralwavelengthcorrespondedcloselytothewavelengthofredshiftedLyαfromeachofthequasars.TheobservationsaresummarizedinTable1.Thepixelsizeis0.046′′pixel−1inthePCand0.1′′pixel−1intheWFC.Theindividualexposureswerereducedusingthestandardpipelinereduction.Thefinalimageswereproducedbyaveragingtheindividualexposureswithsigma-clippingtoremovecosmicrays. Thebandpassofthetelescope+WFPC2+F555Wcombinationcoversthewavelengthrangeofapproximately4500˚Aand6000˚A.ThestrongUVlinesofCIVλ1550,HeIIλ1640,andCIII]λ1909areincludedinthisbandpassandinprincipleitispossiblethattheselinescontributetothedetectedextendedemission.However,Heckmanetal.(1991b)andLehnert&Becker(1998)haveshownthattheselinesaregenerallyweakinquasarhosts,withthepossibleexceptionofPKS0445+097.ForthisquasarHeckmanetal.(1991b)foundthatthehosthasaHeIIλ1640equivalentwidthofseveral˚A.FromthisandthebrightnessofthehostinourPCimageweestimatethatHeIIemissioncontributesmuchlessthan1%tothetotalemissionseeninthePCimage.ThusthecontributionoflineemissiontothetotalemissionobservedfromthehostintheF555Wfilterisnegligible,forPKS0445+097andallotherquasarsinoursample. Thebroad-banddatawereflux-calibratedassumingtheinversesensitivityfortheF555Wfilterof3.459×10−18ergss−1cm−2˚A−1dn−1andazeropointof22.563(Whitmore1995).Thisputstheresultantmagnitudesonthe“Vegasystem”.ToconverttotheSTMAGsystem,whichassumesaflatspectralenergydistributionandhasazeropointof22.543,0.020magnitudesshouldbesubtractedfromthemagnitudesgivenhere. 3. 3.1. ImageReduction ConstructionandSystematicsofthePSF WehaveattemptedtoquantifytheshapeandconstancyoftheHSTPSF.First,wehavecollectedimagesofthestandardstarsusedtocalibratetheF555Wfilterandobservations –7– ofbrightstarsintheouterregionsofωCenthatweremadewithinaperiodofdaysofourobservations.Wefoundabout20starsthatweresuitablycloseintimeandwithinabout100pixelsofthecentralpixelofthePC(allofquasarimageswherecenterednearthemiddleofthePCCCD).Wethenconstructedanempiricalpoint-spreadfunctionusingthesedatabyaddingtheindividualexposuresaftertheyhadbeenalignedtoacommoncenter.ThisempiricalPSFwasthencomparedwiththemodelPSFconstructedusingthePSFmodelingprogram,“TinyTim”.However,thecloseagreementislimitedtoazimuthalaverages–TinyTimdoesnotreproducethedetailed2-dimensionalstructureofthePSF(thereisanasymmetryintheintensityofthediffractionspikes,especiallyinthe+U3direction,whichTinyTimdoesnotreproducewell).Next,wealsomeasuredtheencircledenergydiagrams(EEDs),i.e.,thefractionoffluxfromapointsourceinteriortoaradiusr,asafunctionofr.Wetheninter-comparedalltheEEDstakenthroughagivenfiltertodeterminethereproducibilityoftheEEDandcomparedtheindividualstellarEEDswiththatofthemodelPSFproducedbyTinyTim.WefoundverygoodagreementbetweentheshapeoftheEEDfromthesumoftheobservationsofstandardstarsandthatoftheTinyTimPSF.WethencomparedindividualstarEEDswiththatoftheTinyTimPSF.Thisinter-comparisonofapproximately20starsshowedthatwecandetecthostthatcontributesmorethanabout5%asmuchlightasthequasaritself(withinaradiusofabout1.4′′).ThislimitisconsistentwiththeknowntemporalvariationsintheHSTPSFduetoeffectslikethegentlechangeinfocusovertimesscalesofmonthsandshortertimescalevariationsduetotheso-called“breathing”ofthetelescope(seeBurrowsetal.1995).Wehaverestrictedourselvestoradiilessthanabout1.5′′,toavoidtheeffectofthepoorlyunderstoodlargeanglescatteringwhichbecomesimportantbeyondaradiusofabout2′′. 3.2.HostMeasurements:PSFSubtraction Measuringthepropertiesofthecircum-nuclearemissionisimportantifwearetogaintrueinsightintothepropertiesofquasarhostgalaxies.Onearcsecondattheredshiftofquasarscorrespondstoroughly11kpc(inthecosmology:H0=50kms−1Mpc−1andq0=0.1whichwillbeusedthroughoutthispaper)–whichissimilartothesizesofpresent-daygalaxies.Thusbeingabletoaccuratelysubtractthenuclearemissionofthequasarfromthemoreextendedemissioniscriticalifwearetoobtaininformationaboutthehostofhighredshiftquasaronscalesafractionofthepresent-daygalaxysize.WeattemptedthissubtractionusingtwodifferentPSFs.Inthefirst,wemodeledthePSFusingtheHSTPSFmodelingprogram,TinyTim.ThesecondmethodwastoconstructaPSFbyaveragingseveralimages(40sintegrations)ofabrightstarintheouterregionsofωCentauriusingtheF555WandthePC.Thesestellarimageswereallwithinabout100pixelsofthecenter –8– ofthePCchipandthusnearthelocationofthequasarimageonthePCchip.Theimageswerealignedtothenearestpixelbeforeaveragingandtheindividualimagescoveredthecentral200pixelsoftheplanetarycamera.Thecountsinthecentralpixelwasabout2500DN,comparabletothatintheimagesofthequasars.Werestrictedourselvestotheseobservationsandnottheentire20thatwereusedintheEEDanalysis.ThesubsampleweusedwerechosentosampletherangeofPCpositionscoveredbytheF555Wquasarexposuresaspartofthisprogram. ThesetwoPSFswereiterativelysubtracteduntilemissionduetothediffractionofthesecondarysupportwaszero.ThisprocedureallowedustoestimatetheuncertaintyinourfractionofextendedemissionbyobservingthepointswerethediffractionspikesbecamenegativeduetooversubtractionofthePSForwereobviousinthePSFsubtractedimageduetoundersubtraction.WeestimatethattherangeofpossibleacceptableamountsofPSFsubtractionleadtoafactorof30%or±0.3magnitudesinthequotedfluxfromthehost.Wenotethatthesediffractionspikeswereseenonlyouttoabout1′′inthePCframesandtherewasverylittleunderlyingnuclearemissionfromthequasarovermostofthisarea.UsingthemodelfromTinyTimortheimagesofthestarinωCengaveverysimilarresults(i.e.,verysimilarfractionoftheemissionthatwasextendedandsimilarmorphologyoftheunderlyingemission).Thereforethroughouttheremainderofthepaper,wewillquoteonlytheresultsofPSFsubtractionobtainedbysubtractingtheimageofthestarthroughtheF555WandthePC. Thenarrow-bandimagesofthequasarsgenerallyonlyrevealmodestextendedemission.Usingtheefficienciesofthetelescope,detectorandfiltercombinationmadeavailablebySTScIandtheintegrationtime,wecalculatethatthescalingfactornecessarytoremovethecontinuumcontributiontothenarrow-bandfilterusingtheF555Wfiltertobeabout50–100.Thereforecontinuumsubtractionhasonlynegligibleeffectonthepropertiesoftheextendedemission.Unfortunately,wecouldnotempiricallymeasurethisscalingfactor.ThisisbecausetheQuad[OII]filtersonlycoveroneoftheWFCsatatimeandineachoftheimagesofthequasars,thestarsavailableforsuchanestimatewereeithertoofainttobeuseful,orweretoobrightandweresaturatedintheF555Wimage.Tocontinuumsubtractthenarrow-banddata,weblock-averagedthePCcontinuumimages2×2tomakethescaleofthePCcontinuumimagesmatchthatoftheWFCnarrow–bandimages.ThetwoimageswerethenalignedandtheF555Wimagethenscaledbythefactorcalculatedabove.ThefinalimageisreferredtoastheLyαimage.ThefluxesmeasuredintheLyαimagesareinreasonableagreement(alwayswithinafactorof2)withthoseobtainedfromthegroundbyHeckmanetal.(1991a). ThenumberofstarsavailablefortheconstructionofempiricalPSFforthenarrow-band –9– frameswasverylimited.However,eachindividualstarusedfortheconstructionhadmanytimesthecountsintheimagesofthequasarsandtherefore,evenwithjustafewstars,thesignaltonoiseinthefinalPSFwassufficienttomakeareasonablesubtractionofthePSF.WenotethatinnoneofthecasesdidastrongpointsourceappearintheimagesandthusthemethodofPSFsubtractionwasalittlebitdifferentfromthesubtractionofthecontinuumimages.Noneofthenarrow-bandimagesofthequasarshadobviouslyvisiblediffractionspikesandthusmostoftheextendedemissionislittleeffectedbythestructureofthePSF.Wethereforesubtracteduntiltheemissionfromthecentralpeakroughlyblendedinwiththemoreextendedemission(i.e.,thesubtractiondidnotcauseaholeinthepositionofthenucleus).InTable2,wequotetheresultsofthePSFsubtraction. 3.3.HostMeasurements:OtherMethods Weattemptedseveralothertechniquesforestimatingtheamountofextendedemissionfromeachquasar.OnereliedonscalingthePSFsuchthatitsfluxinthecentral2pixelsmatchedthatofthequasarimage.Wethentooktheratioofthetotalenergyencircledinaperturesofincreasingradius.Thismethodprovidesanestimateoftheminimumamountofextendedfluxineachquasar.Theothertechniqueweusedtomakethisestimatewastocross-correlateaseriesofgalaxymodelsplusPSF(torepresenttheunderlyinggalaxyandquasar)withtheimageofthequasar.Thistechniquewasdevelopedinaseriesofpapers(Phillips&Davies1991;Boyce,Phillips,&Davies1993).Completedescriptionsofthetechniqueanditsrobustnesscanbefoundintheoriginalpaperscitedabove.Wemadethiscomparisonwithaseriesofmodelellipticalgalaxiesofvaryinghalf-lightradiiandellipticities.Sinceweandtherefereefoundthatgiventhecomplexityofthehostmorphologyrevealedinthepoint-spread-functionsubtractedimages,theseothermethodswereperhapsnotasconvincingasthePSFsubtraction.Thus,wewillnotgointodetailoftheresultsoftheseothermethods.ItissufficienttosaythattheygaveestimatesverysimilartothoseobtainedthroughPSFsubtraction. 4.Results ContourplotsoftheHSTimagesofall5quasarsareshowninFigure1.TocomparetheopticalandradiostructureswealsooverlaidhighresolutionradiomapsfromLonsdale,Barthel,&Miley(1993)ontheHSTimages,asshowninFigure2.Fortheseoverlays,thecoordinatesfromtheHSTimagesareinsufficientlyaccuratetodirectlyoverlaytheimagesusingthecoordinatesgivenbythestandardpipelinereduction.Thereisanuncertaintyof –10– 0.5′′to1′′betweentheabsoluteHSTpositionsasgivenbythepipelinereductionandtheradiocoordinatesystem.Therefore,wemadesomeassumptionsabouthowtopositiontheradioimagesrelativetotheHSTimages.Usingpublishedspectralinformationavailableforthesequasars,weidentifiedtheflattestspectrumcomponentineachradioimage,whichispresumablyassociatedwiththeradioAGN,andalignedthatcomponentwiththemostintensepixel(i.e.,theopticalquasarnucleus)intheHSTimage.Second,wecheckedthatourcoredeterminedusingthemethodjustoutlinedagreedwiththeastrometryfromBarthel(1984)betweenhighresolutionVLAradiomapsandthepositionofthequasarandineverycasewefoundgoodagreement. 4.1.NotesIndividualObjects4.1.1.PKS0445+097 ThecontinuumimageofPKS0445+097isratherasymmetricandonecandiscerntwocomponents:anasymmetriccircum-nuclearhostwithin∼1′′ofthenucleus,elongatedalongP.A.≈240◦,andadetached“blob”at∼2′′fromthenucleusatP.A.≈120◦Thehosthasatotalmagnitudeofabout22.1±0.3magnitudesthroughtheF555Wfilter.This“blob”,locatedabout1.5′′(∼17kpc)totheeast-southeastofthequasarnucleus,hasatotalmagnitudeof23.1±0.2andiscomposedofseveralbrightclumpsimmersedinamorediffusestructure.Atsurfacebrightnesslevelsof≈23mF555Warcsec−2,itstotalextentisapproximately2.5×1.5arcseconds(alongPAsof35◦±10◦and125◦±20◦).Thereishintofafainttailofemissionleadingfromtheregionaboutthequasarnucleusouttowardthisblobofemissiontotheeast-southeast.Anotherinterestingfeatureinthecircum-nuclearcontinuumemissionisa∼1′′long“arc”ofemissionthatcurvesouteastwardofthenucleusandbendstothesoutheast.Thetotalextentofthecircum-nuclearemission,downtosurfacebrightnesslevelsof23mF555Warcsec−2,isabout2.5×0.8arcseconds(roughlyalongPA=105◦andPA=15◦;30kpc×9kpcforz=2.110).Thetotalextendedemission(boththecircum-nuclearhostandblobtothesoutheast)comprisesabout25%ofthetotalemissionseenfromPKS0445+097intheF555Wfilter.Thecircum-nuclearhostcontributesabout17%ofthetotalandtheblobtothesoutheastofthenucleusabout7%. TheF555Wcontinuumimageisqualitativelyandquantitativelyconsistentwiththebroad-bandcontinuumimagesfromLehnertetal.(1992).Theseground-basedimagesshowthattheextendedemissionisblue(consistentwiththatofanearbyIrregularorScgalaxy)andextendedonspatialscalesofabout8arcseconds(∼100kpc)tothesoutheast(Lehnertetal.1992).TheHSTimagedoesnotshowaslargeofanextent,onlyabout2–3arcseconds,buttheorientationofthehostissimilarandtherelativefractionofextendedemissionis –11– thesimilar(15–40%intheground-baseddata).TheHSTimageismoresensitivetothehighsurfacebrightnessandmorecompactstructureswhiletheground-basedimageswiththeirinferiorspatialresolutionandlargerprojectedpixelsizes,aremoresensitivetothelowsurfacebrightnessandmorediffuseemission.Thisbeingthecase,inspectionoftheHSTF555Wandourpreviouslypublishedground-basedimages(Lehnertetal.1992)suggeststhatbothhighsurfacebrightnessareas–thecircum-nuclearhostandtheblobofemission–areembeddedinadiffuseareaofemissiononscalesoftensofkpcpreferentiallyorientedalongtheaxissoutheasttonorthwestbutalsohavingsomelowsurfacebrightnesstothenorth,northeast,andsouthwestofthenucleus. Thenarrow-bandimageofPKS0445+097ispeculiar.Thefluxofthebrightestsourceintheimageislowerbyafactorof∼100thanwhatwewouldhaveexpectedcomparedtothefluxmeasuredinaground-basedimage(Heckmanetal.1991a).Thefilterweusedforthenarrow-bandimagingofPKS0445+097isunusualinonerespectcomparedtotheotherfilters.Inthefilterholder,theFQUVN-Afilterislocatedwithinthebeamoftheplanetarycamera.Tomovethefilterontooneofthewide-fieldcameraarrays,requiredthefilterholdertoberotated33◦.Weattributethediscrepancyinthemeasuredfluxtoanunknownerrorinthepositioningofthefilterorofthetargetandfortheremainderofthepaper,wewillnotconsiderthenarrow-bandimageofPKS0445+097further. 4.1.2.MRC0549-213 ThecontinuumimageofMRC0549-213revealsacomplexstructuresurroundingthequasarnucleus.Theemissionissymmetricaboutthenucleus,withtheprincipalaxisoftheemissionchangingfromabout90◦withinafewtenthsofanarcsecondofthepositionofthenucleus,toapositionangleofabout150◦atadistanceof1′′.Thetotalmagnitudeofthequasar(nucleusplushost)isabout19.7andthemagnitudeofthehostisabout21.3±0.3magnitudesintheF555Wfilter.Thefractionofthetotalbrightnesscontributedbythehostisabout23%.Inaddition,weseeacomplexstructureofcontinuumemissionabout3.2arcsecondstothewestofthenucleus.Thisemissionregionhasacurvedarcshapeandisapproximately1′′insize,downtosurfacebrightnesslevelsof≈23mF555Warcsec−2.Ithasatotalmagnitudeofabout24.0. TheF555WcontinuumimageisqualitativelyandquantitativelyconsistentwiththeUbandcontinuumimageinHeckmanetal.(1991).Theground-basedUimageshowsthathostisveryextended,withthenear-nuclearemission(withinafewarcsecondsofthenucleus)beingpreferentiallyorientedalongPA≈150◦.Moreover,thereisa“tail”ofemissionthatextendsabout5arcsecondstothewest.Thismorphologyagreesquitewellwiththatseen –12– inourF555Wcontinuumimage.Comparingtheground-basedandHSTimageindetailsuggeststhattheemissionfromthenear-nuclearenvironmentofthequasarandtheblobtothewestmustactuallybephysicallyconnected(deepR-bandimagesofMRC0549-213obtainedaspartofanotherground-basedprogramshowsaverysimilarmorphology).OurHSTimageisnotsufficientlydeeptodetectthisconnection.Also,theground-baseddatasuggestedthatabout20%oftheUbandfluxisextended,consistentwiththe23%estimatedfromourHSTdata. Thenarrow-bandimageofMRC0549-213showsextendedstructure.Withinafewtenthsofanarcsecondofthenucleus,theemissionisextendedalongPA≈45◦.Onascaleofafewarcsecondstherearetworegionsofsignificantemission.OneisalongPA≈170◦andextendsoutaboutanarcsecondfromthenucleus.About1.6arcsecondtotheeastofthequasarnucleus,thereisafaintblobofemissionthatseemstobeconnectedtothequasarnucleusproper.ThisfaintblobofLyαemissionisapproximately1arcsecondlonginthenorth-southdirectionandabout0.5arcsecondswideintheeast-westdirectiondowntoourdetectionlimit.Thefluxfromthisdistinctregionofemissionis9.7×10−16ergss−1cm−2.Inaddition,weseeseveralareasoflowsurfacebrightnessemissionnearthisobject.Tworegionsareparticularlynoteworthy.Oneareacorrespondstotheblob3.2arcsecondstothewest.Thereisa≈4σregionofLyαemissionovertheregionofthisblob.Also,thereissomeLyαemissionroughlycorrespondingtoaareaofcontinuumemissionabout3arcsecondsawayfromthenucleusalongPA≈325◦.Unfortunately,theredoesnotexistaground-basedLyαimageofthisquasarwithwhichtocompare. 4.1.3.PKS1318+113 ThecontinuumimageofPKS1318+113showstwoconcentrationsofemission,oneimmediatelysurroundingthequasarnucleusandtheotherabout2arcsecondstotheeastofthequasarnucleus.Immediatelysurroundingthequasarnucleus(withinabout1′′),theemissionisasymmetric,withthebrighterisophotesorientedalongPA≈135◦andthefainterisophotesaremostextendedalongPA≈180◦to200◦.Thetotalmagnitudeofthequasarisabout19.0intheF555Wfilterandthehosthasamagnitudeofabout20.1.Thisimpliesthatthehostmakesupabout38%ofthetotalemissionfromthequasar(nucleus+host).Wenotethatperhapsthisissomewhatover-estimatedinlightofthefactthatthecross-correlationtechniqueimpliesthatonlyabout19%ofthequasarlightisextended.Thegalaxytotheeastofthenucleushasamagnitudeofabout21.9(measuredina2′′×2′′box,whichisaslargeascanbeusedduetotheproximityofthisgalaxytotheline-of-sightofthequasar).Downtosurfacebrightnesslevelsof≈24mF555Warcsec−2,theextentofthecircumnuclear –13– hostisabout1to1.5arcseconds. TheHSTF555WimageissimilartotheB-bandimagepresentedfromHeckmanetal.(1991a).Theground-basedimageshowsbrightextendedemissiontotheeastofthenucleus,withfainteremissiontothesouthandwest.Thetotalextentoftheground-basedBimageisabout6-10′′fromthenucleus.TheHSTimagedoesnotrevealemissionquiteasextendedasthis,onlyabout2–3arcseconds,butthegrosscharacteristicsofthehostissimilar.Therelativefractionofextendedemissionbetweentheground-basedandHSTdataarenotverysimilar(16%intheground-basedBdataversusabout38%intheHSTF555Wdata;althoughwenotethatthecross-correlationanalysisgives19%;seeTable2).Thissuggeststhatrelativelyspeaking,thelightfromthehostisconcentratedwithinan1′′ofthenucleus(scalesnotavailablefromtheground).Althoughagain,wenotethatthecross-correlationanalysisgivesaresultmuchmoreconsistentwithourpreviousground-basedresults. Thenarrow-bandimageofPKS1318+113showsextendedemission(Figure1).MostoftheextendedLyαemissionistothenorthandeastofthenucleus,primarilyalongPA≈45◦andisextendedon1′′to2′′fromthenucleus(downtosurfacebrightnessesof6.3×10−16ergss−1cm−2arcsec−2.Thereareseveralfaintregionsofemissionswithinafewarcsecondsofthequasar.Theseregionshavefluxesofroughly1.7to7.1×10−16ergss−1cm−2.Moreover,wefindreasonableagreementwiththemorphologyoftheground-basedLyαimagepresentedinHeckmanetal.(1991a).InthegroundbasedLyαimagesthehostwasextendedalongPA≈45◦withthemostextendedemissionbeingonthesouthwestsideofthenucleus.Detailedcomparisonbetweentheground-basedandHSTLyαimagessuggestthatthehighestsurfacebrightnessemissionisonthenortheastsideofthenucleuswithseveralbrightclumpsofthesouthwestsidethatisthenembeddedinahaloofdiffuseLyαemission. OneofthemostremarkableresultsofthissmallHSTsurveyofthehostgalaxiesofhighredshiftquasarsistheinterestingspatialrelationshipbetweenextendedLyαandradiojetemission.InFigure2,weshowanoverlayoftheLyαHSTimageandaVLAA-arraymapfromLonsdaleetal.(1993).WeseethatthejetpassesbetweentwoLyαemittingblobssouthwestofthequasarnucleus.Thisinteractionappearsatthepointwherethejetappearstobend.Thetwoblobshavetotalfluxesof1.67×10−16ergss−1cm−2and2.01×10−16ergsscm−2fortheeastern-mostandwestern-mostemissionregions.Measuringthesizesoftheseblobs,wefindthattheeasternmostblobisapproximatelycircularwithadiameterof0.3′′.Thewesternmostblobofthetwoisapproximately1′′×0.35′′(longversusshortaxisorientedPA≈150◦).WealsonotethatthereseemstobeanotherregionofLyαemissionalongthe“counter-jet”sideofthequasarbetweenthenucleusandthenorth-easternradiolobe. –14– 4.1.4.1658+575(4C57.29) Thecontinuumimageof1658+575(4C57.29)showsarelativelycompact(about1′′across)regionofextendedemission.Thetotalmagnitudeofthequasar(nucleus+host)is18.3andthemagnitudeofthehostis20.0.Wefindthatabout21%ofthetotalemissionfromthequasarisextended.WeseealinearfeaturealongPA≈150◦intheextendedemission.ThisfeatureisverylikelytobearesidualemissionthatwasnotaccountedforduringPSFsubtraction.Thisisnotsurprisingsince1658+575isthebrightestquasarimagedduringthisprogramandhencehadthemostextendedandintensediffractionspikescomparedtotheotherquasarimages.Ignoringthelinearfeatureweseethatthebrightestemissionisnorthandtothesouthwestofthenucleus.Thediameterofthehostisonlyabout1′′downtosurfacebrightnessesof22mF555Warcsec−2. Theground-basedBbandimageof1658+575presentedinHeckmanetal.(1991a)showsastructureroughlysimilartothatobservedusingtheHST.Thequasarisextendedonscalesofabout10′′intheground-basedimageandhasahighsurfacebrightnessregiontotheeast-southeastofthenucleuswithlowersurfacebrightnessemissionalsotothenorth-northeastandsouthofthenucleus. Thenarrow-bandimageof1658+575showsanextendedplumeofemissiontothenorth-eastwithsomeverysignificantemissionalsoextendedtothenorthwestofthenucleus.Thereisalsoalowersurfacebrightnessextensiontothesouthofthenucleus.Themostextendedemissionreachesaradiusofabout2′′fromthenucleus.Comparisonwiththeground-basedLyαimageofHeckmanetal.(1991a)againrevealsaverystrongsimilaritybetweenthetwoimages.Intheground-baseddataLyαisextendedonscalesofupto6′′fromthenucleus.Themostsignificantofthisextendedemissionistothenorthwestthroughthesouthsideofthenucleus. 4.1.5.PKS2338+042 ThehostgalaxyofthequasarPKS2338+042comprisesnearly40%ofthetotalcon-tinuumemissionfromthequasar.Thehostisasymmetricwithan“arm”ofemissionthatemanatesfromthenucleustothesouthandthenbendsaroundtotheeast.Inaddition,thereisa“plume”ofemissiontothenortheastofthenucleus.Thetotalextentofthecontinuumnebulaisabout1.5′′downtoasurfacebrightnessof24.0mF555Warcsec−2.Thecircularcontourseen0.4′′totheeastofthenucleusinthecontourplotisalocalminimumintheemission. The15GHzradiomapofLonsdaleetal.(1993)showsa“bent”core,jet,double –15– lobesourceorientedpreferentiallyinaneast-westdirection.The“jet”emanatesfromthenucleusalongPA≈90◦,withthehotspotoftheeasternlobeisatPA≈120◦.ThewesternhotspotisatPA≈270◦.OverlayingthismapontothePSFsubtractedHSTimage,weseeaclosecorrespondencebetweenfeaturesintheHSTimageandthatoftheradioimage.Themaximumintensityseeninthecontourplotoftheradioemissionjusttotheeastofthenucleusalongthe“jet”correspondstothelocalminimumweseeinthecontourplotoftheF555Wimage.Thislocalminimumgivestheimpressionthatwearelookingdowntheendofahollowtubeofemissionwhichcontainstheradioemission.Farthertotheeastweseethatthesurfacebrightnessofthecontinuumemissionincreasesatroughlythesamepointwherethejetseemsto“bend”towardsthehotspot.Moreover,wenoticethattherest-frameUVisophotesoftheF555Wimageseemtobendoutwardstothewestofthenucleusapproximatelyalongthesamepositionangleasthattothewesternradiolobe(i.e.,PA≈270◦). TheF555Wcontinuumimageisroughlyconsistentwiththeground-based4mimagetakenthroughtheB-filterbyHeckmanetal.(1991).Thisground-basedimageshowsthemostsignificantemissionistothesoutheastofthenucleusandisextendedabout6′′acrossdowntoasurfacebrightnessof27.3mBarcsec−2.TheHSTimagedoesnotrevealemissionquiteasextendedasthis,onlyafewarcseconds,buttheorientationofthehostisroughlysimilar.TheHSTimagehoweversuggeststhatamuchgreaterfractionoftheemissionisextended,namely∼40%versus∼16%fromtheground-basedBimage.TheHSTimageisobviouslymoresensitivetothehighsurfacebrightnessandmorecompactstructureswhiletheground-basedimageswiththeirinferiorspatialresolutionandlargerprojectedpixelsizes,aremoresensitivetothelowsurfacebrightnessandmorediffuseemission.ThiswouldsuggestthatperhapsHeckmanetal.(1991)over-subtractedtheground-basedimageofPKS2338+042andthattheUVcontinuumhostofPKS2338+042isverycompactcomparedtotherestofthissmallsample. TheLyαimageofPKS2338+042isalsoveryextended,revealingahostgalaxyapprox-imately2′′acrossdowntoasurfacebrightnessof5.4×10−16ergss−1cm−2arcsec−2.ThemorphologyofthehostgalaxyinLyαissimilartothatofthecontinuumemission.ThebasicorientationoftheLyαhostiseast-west.Themostsignificantpieceofthemorphologyofthisimageisan“arm-like”structurethatextendsfromthenucleusouttoabout1arcsecondtotheeastwhereitterminatesinarelativelyhigh.surfacebrightnessregionofemission.Tothenorthofthis“arm”thereisaregionofrelativelylowsurfacebrightnesscomparedtoitsimmediatesurroundings(a“hole”intheextendedemission).Therearefainter“plumes”ofemissiontothenortheastandthesouth-southeast.AtthelowestsurfacebrightnessvisibleintheLyαimage,thereisalsoafaintextensiontothenorthwest.Atthelowestsurfacebright-nesslevels,theorientationoftheLyαispreferentiallyinthesoutheast-northwestdirection –16– (PA≈140◦)asopposedtothegeneraleast-westorientationathighersurfacebrightnesses.ThemorphologyoftheHSTLyαimageissimilartothatseenintheground-basedLyαimageofHeckmanetal.(1991).Theground-basedLyαimageshowsageneralsoutheast-northwestorientationwithLyαemissionbeingseenoverabout9′′.Theground-basedLyαimageisofcourseamuchdeeperimagethanourHSTLyαimage,reachingdowntoasurfacebrightnessof1.5×10−17ergss−1cm−2arcsec−2. Overlayingthe15GHzradioimageofLonsdaleetal.(1993)ontheHSTLyαimageweagainseeagoodcorrespondencebetweenthehighestsurfacebrightnessLyαemissionandtheradio“jet”.TheLyαshowsahighsurfacebrightnessextensionabout0.8′′totheeastofthenucleus.Overthissameregionthejetofradioemissionisobserved.ItisinterestingthatovertheregionofthemostintenseextendedLyαemissionisalsotheregionwhereweseethe“jet”ofradioemissionandwheretheradioemissionundergoesitsmostseverebending(inprojection).Moreover,weagainseethattothewestofthenucleus,theisophotes“bend”outwardsfromthenucleusoveraregionabout0.5′′outfromthenucleus.Thisisalongthesamepositionanglefromthenucleusthatweseethemostdistantradiohotspot ToelucidatetherelationshipbetweentheradioandUVcontinuumandLyαlineemis-sion,wehavemadeasinglecutthroughthe15GHzradioimagefromLonsdaleetal.(1993),andboththePSFsubtractedF555Wandnarrow-bandLyαimage(whichhasnotbeenPSFsubtracted).Thesecutsweremadefromthehighestsurfacebrightnesspeakinthenucleusoftheradioimage,inthePSFsubtractedF555W,andintheLyαimageandthenincludingallthepixelstotheeastandwestofthenucleusouttoaradiusofroughlyanarcsecondinbothdirections.WehavenormalizedandoverlayedthesecutsinFigure3.Thedirectionofthecutwasselectedsuchthatitpasseddirectlyalongtheradiojetthatpointsdirectlytotheeastofthenucleussothatwemaydirectlycomparetheone-dimensionalspatiallyextendedradio,UVcontinuum,andLyαradialdistributions. AscanbeseeninthisFigure,theradioandUVcontinuumemissionarestronglycorrelated,whiletheradioandLyαemissionarenot.SincethisprojectedcutliesalongthedirectionoftheminimuminthespatialdistributionoftheUVcontinuum(seeFigure1),thetoppanelofFigure3showsthatthejetmustpassthroughthelocalminimum(bestdescribedasa“hole”)inthedistributionoftheUVcontinuumemission.Ontheotherhand,theanti-coincidenceoftheradioandLyαdistributionissuchthatatthepositionwherethejetisbendingawaytothesoutheast,whichiswhytheradiosurfacebrightnessinthebottompanelofFigure3isdecreasing,theLyαsurfacebrightnessisincreasingandreachingalocalmaximum.ThespatialrelationshipbetweentheLyαandradioemissionisverysuggestiveofa“jet-cloud”interactioninthattheareaofhighsurfacebrightLyαemissionisresponsibleforbendingthejet. –17– WhilethestructureoftheLyαimageisrathercomplex,toaidusininterpretingthedatainrelationshiptoapossible“jet-cloud”interaction,wewishtoestimatethefluxfromtheregionofrelativelyhighsurfacebrightnessintheLyαimageatthepointwherethejetbendstothesoutheast.Isolatingthepixelsoverthisregion(approximately0.3′′×0.3′′regionabout0.8′′fromthenucleus),wefindatotalLyαfluxof2.1×10−16ergss−1. 5.Discussion Inthissectionwediscusstheresultsandtheirimplicationsforourunderstandingofthecircum-nuclearenvironmentsofhighredshiftquasars.Oursampleistoosmallforadetailedstatisticalanalysis.Thus,wewillfocusourattentiononafewcommonalitiessharedbythequasarhostsandcomparethesepropertieswiththoseofhighredshiftradiogalaxiesandfieldgalaxies,andlowredshiftstarburstgalaxies. 5.1. TheRadio-AlignedUVContinuumandConfusionbyIntervening AbsorberGalaxies Heckmanetal.(1991a)andLehnertetal.(1992),fromground-basedimagesofquasars,foundweakevidenceforalignmentbetweentherest-frameoptical/UVandtheradioemission.InaHST/WFPC2snapshotstudyof43quasarsfromthe3CRcatalog,Lehnertetal.(1999b)arguethatquasarshostsindeedexhibitthe“alignmenteffect”inthecontinuumpluslineemission(allofthebroad-bandHSTimagesinthatstudyhavesomecontributionduetoemissionlines)butthattheeffectisslightlyweakerthaninradiogalaxiesatsimilarredshifts.Asdiscussedbelow,theHSTdatashowthatinterveninggalaxiesalongthelineofsighttothequasarmayconfusethemorphologiesofthehosts:2ofthe5quasarsfromoursampleappeartohavenearbygalaxiesseeninprojection(PKS0445+097andPKS1318+113).Interestingly,thesetwoquasarswerepreciselythosewhichshowedthegreatestmis-alignmentbetweentherest-frameUVcontinuumhostandradioemissioninthestudyofHeckmanetal.(1991a). TheevidencethatPKS0445+097hasanearbyinterveningsystemSEofthenucleusisbasedonKeck10mspectroscopy(Lehnert&Becker1998),aswellasmorphologicalandluminosityconsiderations.TheKeckspectrumshowsthattheSEblobisataredshiftof0.8384±0.0002,whichissimilartothatoftheMgIIabsorptionseenagainstthenuclearcontinuumofPKS0445+097(Barthel,Tytler,&Thompson1990).Therefore,itisnotsurprisingthatwefoundnoLyαemissionfromthis“blob”attheredshiftofthequasar. –18– Inaddition,Lehnert&BeckeralsofoundthatthisgalaxyislikelytocontainaSeyfert2nucleus.Ifweadoptz=0.84fortheredshiftofthisgalaxy,thecentralwavelengthoftheF555Wfiltercorrespondstoabout2930˚Aintherest-frameofthegalaxy.ThiswavelengthisclosetothewavelengthsoftheUandBfiltersandthusextrapolationsfromfluxdensitymeasuredintheF555WtoestimatethefluxdensitiesoftheUandBfiltersintherest-frameofthequasarhostaresmall.UsingthespectralenergydistributionfromArmusetal.(1997)forthisgalaxy(approximatelythatofthealate-typespiral)toextrapolatethemeasuredfluxdensityintheF555WfiltertothefluxdensityatthewavelengthsoftheUandBfiltersintherest-frameoftheinterveninggalaxyandcorrectingforGalacticextinction,wefindthattheUandBabsolutemagnitudeoftheblobtotheSEofthenucleusisMU=−21.8andMB=−21.3.Thusthisinterveninggalaxyisapproximatelyafactorof2moreluminousthanafiducialSchecterL∗galaxy. Moreover,thisinterveninggalaxyappearstohaveaverydistortedmorphology.TheHSTF555WimageshowsagalaxywiththreeknotsofemissionelongatedalongPA≈45◦withthebrightestregionnotroughlyinthecenteroftheemissionbuttowardsthenortheasternendofthegalaxy.Thegalaxyappearstobenearlyedge-on.ComparingthemorphologyofthisgalaxywithotherinterveningabsorbersobservedwiththeHST(e.g.,Dickinson&Steidel1996;Steideletal.1997),wefindthatthegalaxyalongthelineofsighttoPKS0445+097ispeculiar.MostMgIIabsorbinggalaxyhavepropertiesconsistentwiththegeneralfieldpopulationofgalaxiesandhence“normal”morphologiesanddistributionsofluminositysimilartofieldgalaxies(Steideletal.1997;Bergeron&Boisse1991;Steideletal.1994).However,manyofthepeculiarmorphologiesappeartobeassociatedwithgalaxiesthatareviewednearlyedge-on(Dickinson&Steidel1996).ThusweconcludethateventhoughthemorphologyappearstobepeculiarcomparedtomostMgIIabsorbinggalaxies,itsapparentlyedge-onorientationimpliesthatextinctioninthediskmayaccountforitsseeminglypeculiarmorphology.ThetwocoloranalysisofArmusetal.(1997)suggeststhatthecolorofthisgalaxyisconsistentwithScspiralgalaxyatz=0.84withabout0.5magnitudesofadditionextinctioncomparedtolowredshiftScspiralgalaxies.Thisadditionalreddeningisconsistentwithourclaimherethatextinctionmayaccountforthisgalaxy’sseeminglypeculiarmorphology.However,itcouldalsobethatsincethisgalaxyappearstobeharboringaSeyfertnucleus(Lehnert&Becker1998),itmightalsobethatthecomplexmorphologyisassociatedwithamergereventthatinitiatedtheSeyfertactivity. InthecaseofPKS1318+113nodirectspectroscopicevidenceexiststhatitscompaniontotheeastisalsoaforegroundobject.Ifthisobjectisattheredshiftofthequasar,itwouldhaveanimplausiblyhighluminosity(>25L∗andmoreluminousthanthequasarhost).Thereareseveralothermoreplausiblepossibilitiesfortheredshiftofthisobject.Thetwomostplausibleredshiftsforthisobject,0.8388and1.0541,whichareassociatedwithMg –19– IIabsorbersalongthelineofsighttoPKS1318+113(Bartheletal.1990;Jankarkarinen,Hewitt,&Burbidge1991).Sincethisgalaxyisbright(about21.9inF555W),itismorelikelythatthisgalaxyisassociatedwiththeMgIIabsorberatz=0.8388,whichisverysimilartothecaseofPKS0445+097.Theabsolutemagnitudeofthisgalaxyunderthesameassumptionsmadepreviouslyfortheabsorberalongthelineofsightto0445+097impliesMU≈−23andMB≈−22.5.ThesemagnitudesaremanyL∗andthusweconsiderassociatingthisgalaxywiththeMgIIabsorberatz=0.8388veryimplausibleandthatitisassociatedwiththeabsorberatz=1.0541evenlesslikely(seee.g.,Bergeron&Boisse1991;Steideletal.1994).However,suchaspeculationwillhavetoawaitspectroscopicobservationstodeterminetheredshiftofthisnearby(inprojection)galaxy. 5.2.TheNatureoftheUVContinuum WewillcenterourdiscussionoftheUVcontinuuminthehostsofquasarsontwoaspects:theoriginofradio-alignedUVcontinuumandthestellarpopulationoftheunderlyinggalaxy. 5.2.1.TheRadio-AlignedComponent TherehavebeenafewhypothesesforthephysicalcausesofextendedUVcontinuumemissioninhighredshiftquasarsandradiogalaxies.Themostviableonesare:1)starformationstimulatedbytheradiojetasitpropagatesoutwardsfromthenucleus(McCarthyetal.1987;Chambersetal.1987;DeYoung1989;Rees1989;Begelman&Cioffi1989),2)scatteringoflightfromahiddenquasarbyelectronsordust(e.g.,Fabian1989;Cimattietal.1997),3)inverseComptonscatteringofmicrowavebackgroundphotonsbyrelativisticelectronsintheradiojetsorlobes(Daly1992a;b),and4)selectioneffectsrelatedtothepossibleenhancementofradioemissionbydensegaswhichispreferentiallylocatedalongthegalaxy’smajoraxis(Eales1992).ObservingthatinfacttheradioandUVcontinuumonsmallscalesareanti-correlated(seealsoLehnert1996),meaningthatthehighsurfacebrightnessradioemissionfromthejetisactuallyinaminimumintherest-frameUV,providesatestofthesevariousproposedschemes. Inthemodelofjetinducedstar-formationandscatteringhypothesis,wemightexpecttoseesuchanti-correlationsonsmallscalesinadditiontothe“alignment”betweentheradioandUVcontinuumemission.Thismightcomeaboutforthesamereasoninbothcases.Thepressurefromthejetwouldpushmaterialoutwardsbothalongthejetandperpendiculartoit.Thehighperpendicularpressuremightcausecloudstobecomeunstableandcollapse –20– anditwouldalsoclearmaterialfromtheregionofthejet.Incaseofthejet-inducedstar-formationhypothesis,overpressureduetothepassingjetmightcausethesecloudstoformstarsandinthescatteringhypothesis,theover-pressurizedcloudswouldprovideformoreefficientscatteringofthequasarlight.Inbothhypotheses,thiswouldexplainthelargescalerelationship(i.e.,the“alignmenteffect”)butonsmallscales(i.e.,thewidthofthejet)aanti-coincidencewhethergeneratedbystar-formationorscattering.However,sinceitisdifficulttounderstandhowthejetcaninhibitstarsfrommovingintotheregionsthroughwhichitpasses,thejet-inducedstar-formationscenarioonlyworksifthecrossingtimeofthehighmassstarsismuchlongerthantheirevolutionarytimescale.Otherwise,themassivestarsthatareformedattheedgeofthejetwillfillinthejetregionwithhighsurfacebrightnessUVemission.Ifthetimescaleforthemassivestarstopenetrateintotheregionofthejetislongenough,themassivestarswilldieout,thuspreservingthe“hole”inthelightthroughwhichthejetispassing.Interestingly,theradiosourceinPKS2338+042islikelytobeyoung;itsobservedsmallradiosize(roughlya10-20kpc,moduloprojectioneffects)suggeststhatitisonlybetween106and107yearsoldwhichisroughlythesameorderofmagnitudeastheevolutionarytimescaleofhighmassstars.Forexample,ifthestarsareorbitingat100kms−1,theywilltransverse1kpc(0.1′′attheredshiftofthequasarintheadoptedcosmology)inabout107years.Thescatteringhypothesisdoesnotsufferfromsuchadrawbackandisfeasiblewiththeonlycaveatthatthejetmustbefairlyefficientatremovingpossiblescatterersfromtheregionsthroughwhichitispassing.GiventhatthepressureinthejetisestimatedtobemanyordersofmagnitudehigherthanthereasonablepressureoftheISMinanormalgalaxy(liketheMilkyWayforexample),suchapossibilityseemshighlyplausible. ThesuggestionofInverseComptonscatteringseemstoberuledoutbytheseobserva-tions.InverseComptonscatteringofmicrowavebackgroundphotonsbyrelativisticelectronsintheradiojetsorlobes(Daly1992a;b),wouldinfactrequirethattheregionsofthehighestelectrondensity(likelytobethejets)shouldhavethehighestUVsurfacebrightnesses.Thisisexactlytheoppositeofwhatweobserved. Assigningthe“alignmenteffect”topossibleselectioneffectsrelatedtotheenhancementofradioemissionbydensegaswhichispreferentiallylocatedalongthegalaxy’smajoraxis(Eales1992)isaninterestingsuggestionthatseemsplausiblegiventhecurrentdataset.Wehavefoundevidencefor“jetcloud”interactionsin2outofthe4sourcesimagedatLyα(see§5.3.1).Thereforestronginteractionsbetweentheradioandambientinterstellarmediumarecertainlynotrareinradiogalaxiesorquasarhosts(§5.3andreferencestherein).Withinthiscontext,itmaybethatthe“hole”intheUVcontinuummayberelatedtotheincreasedpressureintheregionsurroundingthejetduetothepassageofthejetthatinfactpreventsitfromde-collimating.However,inordertogaugewhetherornotthisspeculationisplausible –21– willrequiremoreextensiveobservations. Tomakethisdiscussionmoregeneric,wenotethatotherquasarhostsappeartohavethisgeneralalignment,butexhibitdetailedspatialanti-coincidencebetweentheradioandrest-frameUVemission(althoughperhapsnotasdramaticasthatseeninPKS2338+042).InHSTsnapshotdataonalargesampleofquasarsselectedfromthe3CRsample,Lehnert(1996)foundevidenceforsubtleanti-correlationbetweenradioandrest-frameUVcontinuumandlineemissioninthesesources(alsoseedeVriesetal.1996)eventhoughgenerallytheradioemission“aligned”withtherest-frameUVcontinuumandlineemission.Theseanti-coincidencesweremainlyseeninthesourceswithcomplexcompactmorphologies,roughlysimilartotheradiomorphologyofPKS1318+113andPKS2338+042. 5.2.2.APossibleRadio-UVSynchrotronJetinPKS0445+097 Ashasbeenemphasizedpreviously,animportantissueinthestudyofhighredshiftradio-loudAGNistherelationshipbetweentherelativisticradio-emittingplasmaandthelineandcontinuumemissionfromthehostgalaxies.Totestthishypothesis,wehaveoverlaida0.16′′resolutionimageofPKS0445+097fromLonsdaleetal.(1993)ontheF555Wimage(Figure2).Herewehaveassumedthatthebright,compacteasterncomponentisidentified 15GHz withthequasarnucleusonthebasisofitsinvertedradiospectrum(α5GHz=−0.6±0.2;Barthel1984).Theoverlayshowsthatthereisacurvedopticalfeaturesouthwestofthequasarwhichcorrespondstotheradiojet.Theopticalandradiofluxdensitiesina0.5×0.5arcsecondareacenteredonthisfeatureare0.18±0.04µJy(λobs=5398˚A,thecenteroftheF555Wfilter)and13.5±0.7mJy(λobs=2.0cm),implyingaspectralindex(Sν∼ν−α)ofO =1.1±0.2.Thisisconsistentwithasteepeningradio-opticalsynchrotronspectrumsinceαR 15GHz theradiospectralindexofthejetbetween5GHzand15GHzisα5GHz=0.8±0.2(Barthel1984)andsuggeststhattheemissionmayindeedbeassociated.Ifthisistrue,anddeepHSTimagingpolarimetrywouldberequiredtoprovethis,thentheopticaljetinPKS0445+097wouldbethemostluminousandhighestredshiftjetknown(intherest-frameofthequasar:logLB=29.0ergss−1Hz−1andlogP1.4GHz=34.6ergss−1Hz−1,morethananorder-ofmagnitudemoreluminousthananypreviouslydiscoveredsynchrotronjet;seeDeyandvanBreugel1994foradiscussionofknownoptical/radiosynchrotronjets).However,consideringthecomplexandfilamentarystructureofthecircum-nuclearhost,thereisofcoursealsothe O possibilitythattheoptical/radioassociationisaccidental,andtheαRfortuitouslyclosetothevalueexpectedforsynchrotronemission.Wenotealsothatthe5GHzmapofPKS0445+097publishedbyBarthel(1984)showsasmallextensionnortheastfromthecore,i.e.,oppositetothesouthwestradiojet,andcoincidentwiththeopticalextensioninthatsame –22– directionseenintheHSTimage.Noradiospectralindexinformationforthisfeatureisavailable. 5.2.3.StarFormingRegionsintheQuasarsHosts? Ifthecircum-nuclearUVcontinuumfromthehostsareduetorecentstarformation,thanitisofinteresttocomparetheUVluminositieswithlowredshiftstarburstandnormalgalaxies(Kinneyetal.1993;Donasetal.1987;Treyeretal.1998).Wefindthatthetypicalluminosityofthecircum-nuclearhostisabout1012L⊙at≈1700˚A(λPλ;Table3), 11 comparedto<∼few×10L⊙fornormalandstarburstgalaxiesatlowredshift(H0=50kms−1Mpc−1).ThusthehostgalaxiesofquasarsareatleastanorderofmagnitudemoreluminousthanthemostluminouslowredshiftgalaxiesintheUV.However,Calzetti,Kinney,&Storchi-Bergmann(1994)andMeureretal.(1997)findthetypicalUVextinctionintheKinneyetal.sampleandstarburstsgenerallytobeabout1-3magnitudesat≈1700˚A.IfwecorrectthemostextremeUV-emittinggalaxiesinthelocaluniverseforthisamountofUVextinction,theycanindeedreachtheUVluminositiesobservedinthecircum-nuclearhostsofthequasars.Thereforeanalogsforthecircum-nuclearhostassociatedwiththesequasarsmayexistinthelocaluniverse,butmustbethemostextremeUV-luminousgalaxiesand,moreover,thequasarhostsmustberelativelyunobscured.Thereisnogoodreasontoassumewhythelattershouldbethecase. Toaddsomefurtherperspectiveonthenatureofthehostsofradioloudquasars,wenotethattheUVluminosityofthehostsmeasuredinthisstudyaremuchmoreluminousthanthe“Lymandrop-out”fieldgalaxiesstudiedbybySteidelandcollaborators(Steideletal.1996;Giavalisco,Steidel,&Macchetto1996).UsingtheUV(1500˚A)luminosityfunctionforLymandrop-outgalaxiespresentedbyDickinson(1998),wefindthattheaverageUVluminosityofthese5quasarhostsisaboutafactorof10moreluminousthanthemostluminousLymandrop-outgalaxies.Thiscomparisonwasmadewithoutcorrectingeithersamplefortheeffectsofdustextinction.Therefore,itisdifficulttoassociatequasarhostswiththefieldpopulationofstarburstgalaxiesathighredshift–eventheextremelyluminousones. 5.3.TheExtendedLyαEmission OurquasarstypicallyhaveLyαluminositiesof≈few×1044ergss−1.Lyαluminositiesthishigharetypicalofwhatisobservedinhighredshiftradiogalaxies(e.g.,vanOijket –23– al.1997;McCarthy1993).Thissuggeststhatquasarsandhighredshiftradiogalaxieshavesimilargalaxy-scaleenvironmentsandionizingsources,andsupportsmodelswhichattempttounifyradiogalaxiesandquasarsthroughorientation,evolution,andenvironment.Inaddition,eveninourrelativelyshortHSTexposures,theLyαemissionisextendedovertensofkpc;againverysimilartowhathasbeenobservedinhighredshiftradiogalaxies.Theseresultsareinagreementwiththeconclusionsofourotherinvestigationsofhighredshift(z>2)quasarhosts(e.g.,Heckmanetal.1991a,b;Lehnert&Becker1998). However,quasarsofferusanadvantageovertheradiogalaxies.Sincewecanobservethenucleusmoredirectly,wecanestimatetheobservedionizingfluxemittedbythequasarnucleus.UsingthemeasuredUVfluxesfromtheHSTdataandusingthescalingrelationsbetweenthefluxofUVwavelengthsandthetotalionizingenergyinquasarsfromElvisetal.(1994),weestimatethatinordertoproducetheLyαemissionobserved,thehostgalaxyinterceptsonlyafewpercentofthetotalionizingluminosityofthequasar.Thisisnotsurprisingconsideringtheobservationsofthe“proximityeffect”intheLyαforestlines.The“proximityeffect”,wherethenumberdensityofLyαforestlinesislowernearthequasarthanawayfromit(e.g.,Weymann,Carswell,&Smith1981;Bechtold1994),hasbeeninterpretedasthehydrogenbecomingmorehighlyionizedwithinthesphereofinfluenceofthequasar.Observingthiseffectimpliesthatmuchoftheionizingradiationfromquasarsmustescapetoclusterscalesandthusthehostgalaxyofthequasarcannotbecompletelyopticallythicktoionizingphotonsinalldirections.Thisisinagreementwithourestimatethatthehostgalaxyandimmediateenvironmentofthequasarnucleusonlyinterceptsafewpercentofthetotalionizingluminosity.Ofcourseitcouldbelesssincetherecouldbelocalsourcesofionizationsuchasyoungstarsand/orshocksgeneratedbythemechanicalpowerfromtheradiojets.Todeterminetherelativeimportanceoflocalionizationsourceswouldrequirespatiallyresolvedspectroscopy. Theradio/Lyαoverlays(Fig.2)showthatthereisagoodassociationbetweentheradioemissionandthestructureoftheLyαemission.First,theprincipalaxisoftheradioemissionisgenerallyalongthesamedirectionastheextendedLyαemission.Thisisverysimilartotheradio-alignedextendedemission-lineregionsseeninhighredshiftradiogalaxies(e.g.,McCarthy1993andreferencestherein).Second,thesurfacebrightnessofthelineandradioemissionappearanti-correlatedtosomedegree(seee.g.,Fig.2).Thisisgenerallyseenatthepointwherethejetandradioemissionarecurved.Third,itseemsthatthebrightestLyαandradioemissionareonthesamesidewheretheradiolobeisclosesttothequasars.Thisresemblestheradio/emission-lineasymmetrycorrelationfoundforradiogalaxiesbyMcCarthy,vanBreugel,&Kapahi(1991).Allofthesepropertiesmaybebestunderstoodasbeingduetostronginteractionoftheradiosources(jetsandlobes)withdense,asymmetricallydistributedambientgas.Numerousexamplesofthisareknowninnearby –24– radiogalaxies(e.g.,3C277.3:vanBreugeletal.1985;3C405and3C265:Tadhunter1991;3C356:Eales&Rawlings1990;PKS2152-699:Fosburyetal.1998;PKS1932-464:Villar-Martinetal.1998;3C171:Clarketal.1998). 5.3.1.“Jet-CloudCollisions” Ourobservationsshowtwogoodexamplesinwhichjet-cloudcollisionsseemtooccur:PKS1318+113andPKS2338+042.InthecaseofPKS1318+113weobservetwoemission-lineregionsbetweenwhichtheradiojetispassing.Atthissamelocationtheradiojetbends.Thisradio/opticalmorphologyisverysimilartothatseeninsomenearbyradiogalaxies,especiallyMinkowski’sObject(vanBreugeletal.1985)andisstronglysuggestiveofajetcloudinteraction.InPKS2338+042thespatialresolutionisinsufficienttoallowasimilarlydetailedexaminationbuttheco-spatialbrightLyαandradioknotseastofthequasarandtheradiojetcurvaturedownstreamfromthislocationsuggestasimilarjet-cloudcollisionoccursinthisobject. Byanalogytotheradiogalaxieswewillbrieflyexaminewhethertheobservedcloudpropertiesareconsistentwithsuchaninterpretation.Withthelimiteddatainhand(i.e.,withouthighspatialresolutionspectroscopy)wecanexploreonlyfewoftheconsequencesexpectedfromaviolentjet-cloudcollision.Themainquestionswecanaddressare1)arethecloudsmassiveenoughtodeflectthejets,and2)cantheysurvivethecollisionsontime-scalescomparabletotheradiosourceages? ThetwoLyαblobsSWofthequasarnucleusinPKS1318+113haveLLyα=1.1×1043ergss−1andLLyα=1.3×1043ergss−1totheeastandwestoftheradioemissionrespectively.Assumingnodust,purecaseBrecombinationfor10,000K(Osterbrock1989),cylindricalorsphericalsymmetry,andusingtheprojectedisophotaldimensionsapparentintheLyαimageasdiscussedpreviously,wethenfindthatn2efV∼0.5forbothblobs,whereneistheelectrondensityandfVisthevolumefillingfactor.SincewedonothavedatatoestimateeitherneorfVindependently,wemustrelyonestimatesobtainedforotherobjects.RoughestimatesofthevolumefillingfactorsfortheextendedemissionlineregionsinvariousradiogalaxiessuggestfV∼10−4to10−6andne∼10-1000cm−3(seee.g.,Baumetal.1992;McCarthy1993;Lacy&Rawlings1994).Ifforconvenienceweassumene=100cm−3,thenthiswouldimplyavolumefillingfactorofabout5×10−5andthusconsistentwithvaluesfoundbypreviousstudies.Theseestimateswouldthenimplyamassofionizedmaterialinthesecloudsofaboutfew×107(ne/100cm−3)(fV/10−4)M⊙.MakingsimilarassumptionsfortheLyαemission-lineregionsinPKS2338+042,wefindthatn2efV∼1andwouldthusestimatethatthemassofthecloudsmustbe≈108(ne/100cm−3)(fV/10−4)M⊙.Wenote –25– thattheaboveestimateswouldbesimilarifweassumedthatthegaswereshockheatedinsteadofimplicitlyassumingthatthegasisinrecombinationequilibriumsincemostoftheHydrogenlineemissioncomespredominatelyfromthepost-shockrecombinationzone(e.g.,Dopita&Sutherland1992). Aresuchmassescapableofdeflectingtheradiojetsemanatingforthenuclei?Theo-reticalmodelingsuggeststhatjetscanbedeflectedbydiscreteobjects,butonlyifcertainminimalcriteriaaremet.Firstandforemost,thedeflectormustbesufficientlymassiveassonottobepushedoutofthewayoftheradiojettooquickly.FollowingtheargumentsinIcke(1991)andMcNamaraetal.(1996),weestimatethatdeflectingcloudsmusthaveamass, tjet0.1c4 Mcloud>6×10M()()(42−1⊙∼10ergssljet –26– ofdensecloudsand2)thatthesecanbeveryefficientatdeflectingradiojetsduringaasignificantfractionofthetotalageofthesource(i.e.,roughly107yrs).Tomakethisargu-mentsmoregeneral,wenotethatweonlyobservedgoodevidenceforjet-cloudinteractionintwosources,PKS1318+113andPKS2338+042.Intwoothersources,PKS1658+575andMRC0549-213wedidnotseeevidenceforajetcloudinteraction,andinPKS0445+097wesuspectthatthereissomethingwrongwiththenarrow-bandobservation(§4.1.1.).Thus,weseestrongcloud-jetinteractionintwoofthefoursources.However,PKS1658+57andMRC0549-213(andalsoPKS0445+097)exhibitlinearprojectedradiomorphologiessuggestingthatnodensecloudsinterceptthejetsintheseobjects;althoughthereisaregionofLyαemissionbeyondtheedgeoftheeasternradiolobeofMRC0549-213perhapssuggestingalargeamountofconfiningmaterialalongthatdirection.ObviouslyamuchlargersampleofquasarswithhighresolutionLyαimagesarenecessarybeforeasstatisticallysignificantconclusioncanbemade.However,ourobservationsstronglysuggestthatthebentradiostructuresinradioquasarsmayverywellbeduetotheinteractionoftheirjetswithdenseambientgas(Barthel&Miley1988)andthatsuchinteractionsmaybeverycommonandmayaffectalargepercentageofthetotalradio-loudhighredshiftquasarpopulation. 5.3.2.RelevancetoUnificationSchemes TheLyαandcontinuumimagesofPKS2338+042andtheLyαimageofPKS1318+113showrelativelyobvioussignsofinteractionbetweentheradioemittingplasmaandambientemissionlinegasandperhapsevenwiththestellarpopulation(asprobedbytheUVcontin-uumemission).Theseresultssuggestthattheinteractionbetweentheradioandtheambientinterstellarmediumofthehostgalaxyandcluster-scaleenvironmentmustbeimportant.Infact,ifwetakeourdataliterally,theyimplythatinteractionwiththeambientmediumisimportantindeterminingtheradiomorphologyin≈1/2ofthequasars.Clearly,amuchlargersampleofhighredshiftquasarsneedtobeobservedtodeterminetheexactstatisticsoftheimpactofthestructureofthehostgalaxyandcluster-scaleenvironmentininfluencingtheradioemission. However,evenforalimitednumberofquasars,thisobservationisimportant.Giventhatmuchoftheevidencefororientationbasedunificationreliesonvariousaspectsoftheradiomorphologyofthesources(linearsizes,lobearmlengthasymmetries,predominanceofjets,etc)thisresultallowsustospeculatethatonewouldnotexpecttheretobemuchevidencefororientation-basedunificationbasedonradioobservationsalone.IFinteractionsbetweentheradioandinhomogeneitiesintheISMofradioloudobjectsareimportantindetermining(andlimiting)thesizeandlobeasymmetriesinradiosources,thensuchinteractionmight –27– dominateoverthesimplegrowthoftheradioemissionwithtimeandorientationeffects.Studiesthatuseobservationsoflinearsizes,lobearmlengthasymmetriesandbendsandtwistsintheradiojetsandlobestotestunificationschemes(e.g.,Gopal-Krishna,Kulkarni,&Whitta1996;Kapahietal.1995)mayinfactgetstatisticallyinsignificantresultsnotbecauseorientation-basedunificationisincorrect,butbecauseinteractsbetweentherelativisticradioplasmaandtheambientISMandIGMeitherdominatesorprovidesasignificantsourceof“noise”intheobservations.Thismaypartiallyexplainwhytheresultsoftestsofunificationschemesusingradiodatahavebeensomixedandthatorientation-basedunificationseemsmostappropriateforaratherlimitedrangeofredshiftsandgenerallyonlyforsamplesofrelativelylowredshiftradiosources(seeBarthel1989forexample). 6.SummaryandConclusions Inthispaper,wepresentedHSTWFPC-2imagesofspatially-resolvedstructures(‘hosts’)aroundfivehigh-redshiftradio-loudquasars.Thequasarswereimagedusingtheplanetarycamerawiththebroad-bandF555W(‘V’)filterandinthewidefieldcamerawithanarrow-bandfilterwhosecentralwavelengthisapproximatelythatofredshiftedLyαineachofthequasars.Theseradio-loudquasarswereselectedfromtheearlierimagingsurveyofquasar“host”byHeckmanetal.(1991a)andLehnertetal.(1992).TheseHSTimagesconfirmandextendourearlierground-basedresults. FromananalysisoftheimagesandacomparisonwithhighresolutionVLAradioimagesweconcludethat: 1)Allofthehighredshiftquasarsareextendedinboththerest-frameUVcontinuumandinLyα.Wefindextendedfractionsthatrangefromabout5%to40%ofthetotalcontinuumwithinaradiusofabout1.5′′.Inspiteofthefactthattheseimageshavehigherspatialresolutionsandrelativelyshortintegrationtimesonasmalltelescope,themorphologicalagreementbetweentheground-basedimagesandtheseHSTimagesisquitegood.More-over,thereisreasonableagreementwithourestimatesofthefractionofthequasarlightcontributedbythehostgalaxiesinboththeHSTandground-baseddata.SucharesultissurprisinggiventhefactthattheHSTimagesrevealawealthofstructurewithinanarcsecondofthenucleuswhichiscurrentlyunattainablefromtheground. 2)ThetypicalintegratedmagnitudeofthehostisV∼22±0.5.ThetypicalUVluminosityisroughly1012L⊙(λPλ,uncorrectedforinternalextinction),whichisaboutafactorof10higherluminositythanthatobservedforthe“Lymandropout”fieldgalaxiesstudiedbySteidelandcollaboratorsandthemostUVluminouszeroredshiftstarburstgalaxies.The –28– Lyαimagesarealsospatially-resolved.ThetypicalluminosityoftheextendedLyαisaboutfew×1044ergss−1.Theseluminositiesrequireroughlyafewpercentofthetotalionizingradiationofthequasar. 3)Quasarhostgenerallyshow“alignment”betweentheradio,Lyα,andUVcontinuumemis-sion.Thereisclearevidencethatthegas“knows”abouttheradiosource.Thismanifestsitselfinthe“alignment”betweentheradio,Lyα,andUVcontinuumemission,indetailedmorphologicalcorrespondenceinsomeofthesourceswhichsuggests“jet-cloud”interactions,andinthefactthatthebrightestradioemissionandthesideoftheradioemissionwiththeshortestprojecteddistancefromthenucleusoccursonthesamesideofthequasarnucleusasthebrightest,mostsignificantLyαemission.Theseobservationsofjet-cloudinteractioninfluencingtheradiomorphologiesisachallengetosimpleorientationbasedquasar/radiogalaxyunificationschemes.Thisisperhapswhytheuseoftheradiomorphologyhasgener-allyleadtoconflictingresultswhenusedtojudgetheappropriatenessoforientationbasedquasar/radiogalaxyunificationschemes. 4)ThehighspatialresolutionoftheHSThasrevealedthatobjectsalongthelineofsightbutnearthequasarsinprojectionhavemadeasignificantcontributiontothecontinuumlightfromtheseobjects.Wenoteinparticularthat0445+097and1318+113weretwoquasarswithstrongmis-alignmentbetweentheprincipalemissionaxesintheradioandtheground-basedimagesatUVrestwavelengths(Heckmanetal.1991a).Itisnowclearthatthismis-alignmentwaspartiallyduetothecontaminatingeffectsofnearby(inprojection;lessthan2arcsecondsforthequasarnucleus)foregroundgalaxies. TheauthorswishtothankRayLucasforhisconsiderablehelpinmakingsurethatourprogramwentsmoothly.ConversationsaboutthecomplexitiesoftheHST/WFPC2PSFwithChrisBurrowsandJohnKristwereparticularlyhelpfulinmakingthemostofthedata.Wethankthereferee,Dr.EricSmith,whosecommentsleadtoasubstantialimprovementinthestyleandpresentationofthispaperandDr.GregBothunforhisconscientioushandlingofthemanuscriptinhisroleasthescientificeditor.TheworkofMDLandWvBatIGPP/LLNLwasperformedundertheauspicesoftheUSDepartmentofEnergyundercontractW-7405-ENG-48andtheworkofMDLattheSterrewachtLeidenwassupportedbyfundsprovidedbytheDutchOrganizationforResearch(NWO).ThisworkwassupportedinpartbygrantnumberGO-5393fromtheSpaceTelescopeScienceInstitute,whichisoperatedbytheAssociationofUniversitiesforResearchinAstronomy,Inc.,underNASAcontractNAS5-26555.WealsoacknowledgesupportfromaNATOresearchgrant. –29–REFERENCES Alexander,P.&Leahy,J.P.1987,MNRAS,225,1 Armus,L.,Neugebauer,G.,Lehnert,M.D.,&Matthews,K.1997,MNRAS,289,621Bahcall,J.N.,Kirhakos,S.,&Schneider,D.P.1994,ApJ,435,L11Bahcall,J.N.,Kirhakos,S.,&Schneider,D.P.1995,ApJ,447,L1Bahcall,J.N.,Kirhakos,S.,&Schneider,D.P.1995,ApJ,450,486 Bachall,J.N.,Kirhakos,S.,Saxe,D.H.,&Schneider,D.P.1997,ApJ,479,642Baron,E.,&White,S.D.M.1987,ApJ,322,585Barthel,P.1984,Ph.D.thesis,LeidenUniversity Barthel,P.,Miley,G.,Schilizzi,R.,&Lonsdale,C.1988,A&AS,73,515Barthel,P.,&Miley,G.1988,Nature,333,319Barthel,P.1989,ApJ,336,606 Barthel,P.,Tytler,D.,&Thompson,B.1990,A&AS,82,339 Baugh,C.M.,Cole,S.,Frenk,C.S.,&Lacy,C.G.1998,ApJ,498,504Baum,S.A.,Heckman,T.M.,&vanBreugel,W.1992,ApJ,389,208Bechtold,J.1994,ApJS,91,1 Begelman,M.,&Cioffi,D.1989,ApJ,345,L21Bergeron,J.,&Boisse,P.1991,A&A,243,344Boroson,T.A.,&Oke,J.B.1984,ApJ,281,535 Boroson,T.A.,Persson,S.E.,&Oke,J.B.1985,ApJ,293,120Boyce,P.J.,Phillips,S.,&Davies,J.J.1993,AA,280,694 Boyle,B.J.1993,inTheEnvironmentandEvolutionofGalaxies,eds.H.A.Thronsonand J.M.Shull,(Dordrecht:Kluwer),p.433Calzetti,D.,Kinney,A.L.,&Storchi-Bergmann,T.1994,ApJ,429,582 –30– Chambers,K.,Miley,G.,&vanBreugel,W.1987,Nature,329,604 Cimatti,A.,Dey,A.,vanBreugel,W.,Hurt,T.,&Antonucci,R.1997,ApJ,476,677Clark,N.E.,Axon,D.J.,Tadhunter,C.N.,Robinson,A.,&O’Brien,P.1998,ApJ,494, 546Daly,R.A.1992a,ApJ,386,L9Daly,R.A.1992b,ApJ,399,426 deKoff,S.,Baum,S.A.,Sparks,W.B.,Golombek,D.,Biretta,J.,Macchetto,D.,McCarthy, P.,&Miley,G.K.1996,ApJS,107,621DeYoung,D.S.1989,ApJ,342,L59 Dey,A.&vanBreugel,W.J.M.1994,AJ,107,1977 deVries,W.H.,O’Dea,C.P.,Baum,S.A.,Sparks,W.B.,Biretta,J.,deKoff,S.,Golombek, D.,Lehnert,M.D.,Macchetto,F.,McCarthy,P.,&Miley,G.K.1997,ApJS,110,191Dickinson,M.,&Steidel,C.C.1996,inNewLightonGalaxyEvolution,eds.R.Bender andR.L.Davies,(Dordrecht:Kluwer),p.265Dickinson,M.1998,preprint.Donasetal.1987,A&A,180,12Eales,S.A.,1992,ApJ,397,49 Eales,S.A.,&Rawlings,S.1990,MNRAS,243,1 Elvis,M.,Wilkes,B.J.,McDowell,J.C.,Green,R.F.,Bechtold,J.,Willner,S.P.,Oey,M. S.,Polomski,E.,&Cutri,R.1994,ApJS,95,1Fabian,A.1989,MNRAS,238,41P Fosbury,R.A.E.,Morganti,R.,Wilson,W.,Ekers,R.D.,diSeregoAlighieri,S.,&Tad-hunter,C.N.1998,MNRAS,296,701Giavalisco,M.,Steidel,C.,&Macchetto,F.1996,ApJ,470,189Gopal-Krishna,Kulkarni,V.K.,&Witta,P.J.1996,463,20 –31– Hartwick,F.,&Schade,D.1990,ARA&A,28,437 Heckman,T.M.,Lehnert,M.D.,vanBreugel,W.J.M.,&Miley,G.M.1991a,ApJ,390, 100Heckman,T.M.,Lehnert,M.D.,Miley,G.K.,&vanBreugel,W.J.M.1991b,ApJ,381, 373Icke,V.1991,inBeamsandJetsinAstrophysics,edP.A.Hughes,(Cambridge:Cambridge UniversityPress),p.232Junkkarinen,V.,Hewitt,A.,&Burbidge,G.1991,ApJS,77,203 Kapahi,V.K.,Athreya,R.M.,Subrahmanya,C.R.,Hunstead,R.W.,Baker,J.C., McCarthy,P.J.,&vanBreugel,W.1995,A&AS,16,125Kauffmann,G.White,S.D.M.,&Guiderdoni,B.1993,MNRAS,264,201 Kinney,A.L.,Bohlin,R.C.,Calzetti,D.,Panagia,N.,&Wyse,R.F.G.1993,ApJS,86,5Kotilainen,J.K.,Falomo,R.,&Scarpa,R.1998,A&A,332,503Lacy,M.&Rawlings,S.1994,MNRAS,270,431 Lehnert,M.D.,Heckman,T.M.,Chambers,K.C.,&Miley,G.K.1992,ApJ,393,68Lehnert,M.D.1996,in“TheSecondWorkshoponGigahertzPeakedSpectrumandCompact SteepSpectrumRadioSources”,eds.I.Snellen,R.Schillizi,H.Rottgering,andM.Bremer(Leiden:LeidenObservatory),Lehnert,M.D.,&Becker,R.H.1998,A&A,332,514 Lehnert,M.D.,Miley,G.K.,Sparks,W.B.,Baum,S.A.,Biretta,J.,Golombek,D., deKoff,S.,Machetto,F.D.,&McCarthy,P.J.1999a,ApJSsubmittedLehnert,M.D.,deKoff,S.,McCarthy,P.J.,Miley,G.K.,Baum,S.A.,&Sparks,W.B., 1999b,inpreparationLonsdale,C.,J.,Barthel,P.D.,&Miley,G.K.1993,ApJS,87,63Lowenthal,J.D.1991,Ph.D.thesis,UniversityofArizona McCarthy,P.,vanBreugel,W.,Spinrad,H.,&Djorgovski,S.,1987,ApJ,321,L29McCarthy,P.J.,vanBreugel,W.,&Kapahi,V.K.1991,ApJ,371,478 –32– McCarthy,P.1993,ARA&A,31,639M McCarthy,P.,Miley,G.K.,deKoff,S.,Baum,S.A.,Sparks,W.B.,Golombek,D.,Biretta, J.,&Macchetto,D.1997,ApJ,112,415McLeod,K.K.,&Rieke,G.H.1995,ApJ,454,77 McNamara,B.R.,Wise,M.,Sarazin,G.L.,Jannuzi,B.T.,&Elston,R.1996,ApJ,466, L9Meurer,G.Heckman,T.M.,Lehnert,M.D.,Leitherer,C.,&Lowenthal,J.1997,AJ,115, 54Neff,S.,&Hutchings,J.1990,AJ,100,1441Norman,C.,&Miley,G.K.1984,A&A,141,85 Osterbrock,D.E.1989,AstrophysicsofGaseousNebulaeandActiveGalacticNuclei,(Mill Valley,CA:UniversityScienceBooks)Phillips,S.,&Davies,J.J.1991,MNRAS,251,105Rees,M.1988,MNRAS,231,91PRees,M.1989,MNRAS,239,1p Smith,E.,Heckman,T.,Bothun,G.,Romanishin,W.,&Balick,B.1986,ApJ,306,64Steidel,C.C.,Dickinson,M.,Meyer,D.M.,Adelberger,K.L.,&Sembach,K.R.1997, ApJ,480,568Steidel,C.C.,Giavalisco,M.,Dickinson,M.,&Adelberger,K.L.1996,AJ,112,352Steidel,C.C.,Pettini,M.,Dickinson,M.,&Persson,S.E.1994,AJ,108,2046Tadhunter,C.N.1991,MNRAS,251,46p Treyer,M.A.,Ellis,R.S.,Milliard,B.,Donas,J.,&Bridges,T.J.1998,MNRAS,300,303Weymann,R.J.,Carswell,R.F.,&Smith,M.G.1981,ARA&A,19,41 Whitmore,B.1995,inCalibratingHubbleSpaceTelescope:PostServicingMission,ed.A. KoratkarandC.Leitherer(Baltimore:SpaceTelescopeScienceInstitute),p.269WideFieldandPlanetaryCamera2InstrumentHandbook,C.J.Burrows(Ed.),STScI Publication(June1995) –33– vanBreugel,W.,Miley,G.,Heckman,T.,Butcher,H.,&Bridle,A.1985,ApJ,290,496vanBreugel,W.,Filippenko,A.V.,Heckman,T.,&Miley,G.1985,ApJ,293,83vanOjik,R.,R¨ottgering,H.J.A.,Miley,G.K.,&Hunstead,R.W.1997,AA,317,358Villar-Martin,M.,Tadhunter,C.,Morganti,R.,Clark,N.,Killeen,N.,&Axon,D.1998, A&A,332,479 –34– FIGURECAPTIONS Fig.1.—Foreachrowofthisplot,weshowtheF555WimagewiththePSFremovedasdescribedinthetextontherightandtheLyαimageisontheleft,exceptforPKS0445+097forwhichweonlydisplaytheimagetakenthroughtheF555Wfilter.Eachimagehasbeenrotatedsothatnorthisatthetopandeastistotheleftandeachhasbeensmoothedusinga4×4medianfilter.ThefaintestcontoursarelistedinTable2andeachcontourisafactoroftwoincreaseinsurfacebrightness.Eachimageisapproximately8′′×8′′inprojectedsizeandthesolidbarindicatestheangularsizeof10kpcattheredshiftofthequasarassumingH0=50kms−1Mpc−1,q0=0.1,andΛ=0. Fig.2.—WeshowthePSFsubtractedF555Wcontinuumorthenarrow-band(grey-scale)withcontoursofthehigh-resolutionVLA2cmradiomapfromLonsdaleetal.(1993)overlaid.Ineachoftheplotstheradionucleusasdeterminedin§4hasbeencenteredtothepositionoftheopticalnucleusofthequasar.TheFiguresarespecifically,a)F555WimageofPKS0445+097,b)F555WimageofMRC0549-213,c)narrow-bandLyαimageofMRC0549-213,d)F555WimageofPKS1318+113,e)narrow-bandLyαimageofPKS1318+113,f)F555Wimageof1658+575(4C57.29),g)narrow-bandLyαimageof1658+575(4C57.29),h)F555WimageofPKS2338+042,andi)narrow-bandLyαimageofPKS2338+042.Fig.3.—Atthetop,weshowthecomparisonofaone-dimensionallightprofileofthePSFsubtractedF555W(rest-frameUVcontinuum)imageofPKS2338+042comparedtotheone-dimensionalradiointensityprofileoftheradiomapfromLonsdaleetal.(1993).Below,weshowasimilarlyconstructedcomparisonofthenarrow-bandLyαlightprofilecomparedtotheradiomapofLonsdaleetal.(1993).Both1-dimensionalprofilesarealongrightascensionwhichwaschosenbecauseitliesdirectlyalongtheradiojet.TheradioandHSTimageswerealignedandscaledforthiscomparison.Thefluxscaleisarbitraryandchosenforconvenienceinmakingthecomparisonofthelightdistributionsateachwavelength. –35– Table1.ObservationLog Quasar(1) Filter(2) FOV(3) N(4) Int(5) Date(6) Note.—Col.(1)—Sourcedesignation.Col.(2)—Filterusedfortheobservation.Col.(3)—Theapertureusedfortheobservation.“PC”istheplanetarycamera,and“WF2”,“WF3”,“WF4”arethewidefieldcameras2,3,and4.Col.(4)—NumberofseparateexposureseachwiththeexposuretimeinlistedinCol.(5).Col.(5)—Integrationtimeinsecondsperexposure.Col.(6)—Dateofobservationsintheformofday/month/year. –36– Table2.PhotometryoftheSourcesQuasar(1) mtotal(2) f(PSF-)RotationSBlimit(3)(4)(5) Note.—Col.(1)—Sourcedesignation.Col.(2) —TotalmagnitudeintheF555Wfilterortheloga-rithmofthetotalLyαflux(inunitsofergss−1cm−2)ofthequasar.ThetotalmagnitudeofthequasarintheF555Wfilterislistedinthefirstrowforeachobject,whilethelogarithmofthetotalLyαfluxisenumeratedinthesecondrow.Col.(3)—FractionofthetotalquasarlightthatisextendedasindicatedbyPSFsubtraction.Col.(4)—Angle(indegrees)thattheimagewasrotatedtomakenorthatthetopandeasttotheleftineachimage.Positivevaluesim-plyacounter-clockwisedirectionoftherotation.TocalculatethePAofthebrightestdiffractionspikeoneusesthefollowingformula.Iftherotationispositive,itsrotation−45◦,iftherotationisnegative,itisrotation−45◦+360◦.Col.(5)—Surfacebright-nesslimitofthelowestcontouroftheplotsshowninFigure1inunitsofmagnitudesarcsec−2forthecon-tinuumimagesandinunitsofthelogarithmofergss−1cm−2arcsec−2fortheLyαimages. –37– Table3.UVPowerandLyαLuminosities Quasar(1) z(2) λrest(3) fλ,total(4) fλ,fuzz(5) logλPλ,tot (6) logλPλ,fuzz (7) logLLyα,total (8) logLLyα,fuzz (9) Note.—Col.(1)—Sourcedesignation.Col.(2)—Redshiftofthesource.Col.(3)—ThecentralwavelengthoftheF555Wfilterintherest-frameofthequasarusing5397˚AasthecentralwavelengthoftheF555Wfilter.Col.(4)—Fluxdensityofthetotalemissionfromthequasarsinunitsofthelogarithmofergss−1cm−2˚A−1atthewavelengthgivenincol.(3).Col.(5)—Fluxdensityofthe“fuzz”inunitsofthelogarithmofergss−1cm−2˚A−1atthewavelengthgivenincol.(3).Col.(6)—ThelogarithmoftheUVpowerofthequasartakentobeλPλinunitsofsolarluminositiesusingtherestwavelengthgivenincol(3)andtheassumedcosmologyofH0=50kms−1Mpc−1andq0=0.1.WehavealsocorrectedforgalacticextinctionusingtheextinctionintheB-bandgiveninNEDandusingthestandardextinctioncurvefromOsterbrock(1989).ThevalueofsolarluminosityusedtomaketheconversionisLsun=3.83×1033ergss−1.Col.(7)—TheUVpowerofthe“fuzz”(λPλ)inunitsofthelogarithmofsolarluminositiesatthewavelengthgivenincol.(3).TherelativefractionofextendedemissionusedinthiscalculationisfromthePSFsubtractionanalysis(seeTable2andtextfordetails).Col.(8)—ThetotalLyαluminosityofthequasarinunitsofthelogarithmofergss−1.Col.(9)—ThetotalLyαluminosityofthe“fuzz”inunitsofthelogarithmofergss−1.TherelativefractionofextendedemissionusedinthiscalculationisfromthePSFsubtractionanalysis(seeTable2,cols.(2)and(3)forthefluxesusedandtextfordetails). This figure \"fig1a.gif\" is available in \"gif\" format from: http://arXiv.org/ps/astro-ph/9904114v1 This figure \"fig1b.gif\" is available in \"gif\" format from: http://arXiv.org/ps/astro-ph/9904114v1 This figure \"fig1c.gif\" is available in \"gif\" format from: http://arXiv.org/ps/astro-ph/9904114v1 GREY: 0445+097 HST PC F555W CONT: 0445+097 Radio 15 Ghz 09 45 41 40 DECLINATION (B1950)39 38 37 36 35 3404 45 37.3 37.2 37.137.036.9RIGHT ASCENSION (B1950) 36.8 GREY: MRC0549-213 HST PC F555WCONT: 0549-213 Radio 15 Ghz -21 20 26 27 28 DECLINATION (B1950)29 30 31 32 33 05 49 50.850.7 50.650.550.4RIGHT ASCENSION (B1950) 50.3 GREY: MRC0549-213 HST WFC Narrow-bandCONT: MRC0549-213 VLA A-array 15 Ghz -21 20 26 27 28 DECLINATION (B1950)29 30 31 32 33 05 49 50.850.7 50.650.5 RIGHT ASCENSION (B1950) 50.450.3 GREY: PKS1318+113 HST PC F555WCONT: 1318+113 Radio 15 GhZ 11 22 35 34 33 DECLINATION (B1950)32 31 30 29 28 13 18 49.8 49.749.649.5RIGHT ASCENSION (B1950) 49.4 GREY: PKS1318+113 HST WFC Narrow-band CONT: PKS1318+113 VLA A-array 15 Ghz 11 22 35 34 33 DECLINATION (B1950)32 31 30 29 28 13 18 49.8 49.749.649.5RIGHT ASCENSION (B1950) 49.4 GREY: 4C 57.29 HST PC F555WCONT: 4C 57.29 15 Ghz 57 35 56 55 54 DECLINATION (B1950)53 52 51 50 49 16 58 53.8 53.653.453.2RIGHT ASCENSION (B1950) 53.0 GREY: 1658+575 HST WFC Narrow-bandCONT: 1658+575 VLA A-array 15 Ghz 57 35 56 55 54 DECLINATION (B1950)53 52 51 50 49 16 58 53.8 53.653.453.2RIGHT ASCENSION (B1950) 53.0 GREY: PKS2338+042 HST PC F555W CONT: PKS2338+042 VLA A-array 15 Ghz 04 14 41 40 39 DECLINATION (B1950)38 37 36 35 34 23 38 24.924.8 24.724.6 RIGHT ASCENSION (B1950) 24.524.4 GREY: PKS2338+042 HST PC Narrow-band CONT: PKS2338+042 VLA A-array 15 Ghz 04 14 41 40 39 DECLINATION (B1950)38 37 36 35 34 33 23 38 24.9 24.8 24.724.6 RIGHT ASCENSION (B1950) 24.524.4 10UV continuumRadio continuum86420-1.5 -1 -0.5 RA Offset [’’] 00.51 10Radio continuum86420-1.5 -1 -0.5 RA Offset [’’] 00.51 因篇幅问题不能全部显示，请点此查看更多更全内容