AcceptedforPublicationin；APreprinttypesetusingLTE；ONTHEFRACTIONOFQUASARSWI；RajibGanguly&MichaelS.Br；AcceptedforPublicationin；arXiv:[astro-；Out?owsfromactivegalacti；Subje
AcceptedforPublicationinTheAstrophysicalJournalAPreprinttypesetusingLTEXstyleemulateapjv.10/09/06ONTHEFRACTIONOFQUASARSWITHOUTFLOWSRajibGanguly&MichaelS.Brotherton1AcceptedforPublicationinTheAstrophysicalJournalarXiv:
[astro-ph]
2 Oct 2007ABSTRACTOut?owsfromactivegalacticnuclei(AGNs)seemtobecommonandarethoughttobeimportantfromavarietyofperspectives:asanagentofchemicalenhancementoftheinterstellarandintergalacticmedia,asanagentofangularmomentumremovalfromtheaccretingcentralengine,andasanagentlimitingstarformationinstarburstingsystemsbyblowingoutgasanddustfromthehostgalaxy.Tounderstandtheseprocesses,wemustdeterminewhatfractionofAGNsfeatureout?owsandunderstandwhatformstheytake.Weexaminerecentsurveysofquasarabsorptionlines,reviewingthebestmeanstodetermineifsystemsareintrinsicandresultfromout?owingmaterial,andthelimitationsofapproachestakentodate.Thesurveysrevealthat,whilethefractionofspeci?cformsofout?owsdependsonAGNproperties,theoverallfractiondisplayingout?owsisfairlyconstant,approximately60%,overmanyordersofmagnitudeinluminosity.Weemphasizesomeissuesconcerningclassi?cationofout?owsdrivenbydatatyperatherthannecessarilythephysicalnatureofout?ows,andillustratehowunderstandingout?owsprobablyrequiresmoreacomprehensiveapproachthanhasusuallybeentakeninthepast.Subjectheadings:quasars:general―quasars:absorptionlines―galaxies:active―accretion1.INTRODUCTIONTheroleofout?owsfromquasarsandactivegalacticnuclei(AGN)hasrecentlybecomeanimportantfeatureintheoverallframeworkofhowgalaxiesandstarfor-mationprocessesevolveovercosmictime.MergersandotherinteractionstriggeringAGNseemtoprovidefeed-backa?ectingthelargerscaleenvironment.Recentef-fortstoincludethee?ectsofthisso-calledAGNfeedbackfocusontwomodes:a“radio”modewherebyarelativis-ticjetheatsthesurroundinginterstellarandinterclustermedia(e.g.,Best2007),anda“quasar”modewherebyalowervelocitybuthighermassout?owalsohelpstoclearoutpost-mergershroudinggasandquenchesstarforma-tion(e.g.,DiMatteo,Springel,&Hernquist2005).Wefocusonthissecondmodeinthispaper.Forthismode,anumberofquestionsrequireaddressing.Howcom-monareout?ows?DoallAGNhaveout?ows?Whatdrivesout?ows?Isthereasingleall-governingstructureofAGN?Answeringthesequestionswillhelpustoun-derstandtheroleAGNout?owswithrespecttoissueoffeedback,andotherimportantissueslikechemicalen-richmentandaccretion.Intheensuingsectionsweaimtoachieveseveralgoals:(1)toreviewthewaysinwhichout?owsaredetectedinAGNove(2)tocommentonthemeritsofvariouscatalogsofout?and(3)toarriveatthetrue(possiblyproperty-dependent)observedfre-quencyofout?ows.Initsmostbasicinterpretation,theobservedfrequencyofout?owscanbeequatedwiththefractionofsolidangle(fromtheviewpointofthecentralblackhole)subtendedbyout?owinggas.Thisinterpre-tationassumesthatallAGNfeatureout?owsandthatnotallsight-linestotheemittingregionsareoccultedbytheout?ow.Alternatively(andequallysimplistic),thefrequencycanbeinterpretedasthefractionoftheDepartmentofPhysics&Astronomy,TheUniversityofWyoming(Dept.EastUniversityAve.,Laramie,WY,820711dutycycleoverwhichAGNfeatureout?ows(assumingtheout?owsubtends4πsteradians).Theactualcon-versionofthefractionofAGNfeaturingspectroscopicevidenceofout?owtothesolidanglesubtendbysuchout?owshasbeentreatedbyCrenshawetal.(1999)andCrenshaw,Kraemer,&George(2003).Thiscomputa-tioninvolvesfurtherknowledgeoftheline-of-sightcover-ingfactor(thatis,thefractionoflines-of-sightthatreachtheobserverthatareoccultedbytheout?ow)aswellasanunderstandingofrangeofsolidanglesampledbytheAGNused(e.g.,Type1versusType2AGN).Thetruesituationislikelyinbetweenthesetwoextremes,andmaydependalsoonpropertieswecannotcurrentlycon-strain,suchasthetimesincetheAGNwastriggered.Additionally,westriveheretobuildacasethatmoree?ortshouldbemadetoconsiderout?owsofalltypesto-gether.Oftendatalimitationsofonesortoranotherhaveledtothestudyoflimitedpartsofparameterspace(e.g.,out?owvelocityorvelocitydispersion),creatingarti?cialoratleastbiaseddivisions.Thereappearstobeacon-tinuousrangeinpropertiesofout?owsandtheseshouldonlyberegardedasfundamentallydi?erentwhenthereisclearevidencetoreachsuchaconclusion.Belowwediscusstheidenti?cationofout?ows(§2)andthedata-drivensubcategories(§3).Weshowanillustrativeex-ampleofhowcombiningthedi?erentout?owsubclassesmayleadtoamoreuni?edphysicalunderstandingofout?ows(§4).Finally,webringtogetherthedi?erentsurveymethodologiestodetermineanoverallfractionofAGNdisplayingthesignaturesofout?ows(§5)andsum-marizethecaseformoreglobalstudiesoftheout?owphenomenon.Weadoptacosmologywith?M=0.3,?Λ=0.7,andH0=70kms?1Mpc?1.2.THEIDEALWAYTOSELECTQUASARSEXHIBITINGOUTFLOWSOut?owsfromAGNareprimarilydetectedinultravi-oletandX-rayabsorptionagainstthecompactcontin-uumsource(i.e,theinnerportionsoftheaccretiondisk)2and/orthemoreextendedbroademissionlineregion.Inafewcases,out?owshavebeendemonstrativelyobservedinemissionbothfromthebroadlineregioninnarrow-lineSeyfert1galaxies(e.g.,Leighly&Moore2004;Leighly2004;Yuanetal.2007)andfromthenarrowlineregionofSeyfert1galaxies(e.g.,Dasetal.).[Ar-guably,thefactthatbroademissionlinesinmostAGNhaveonlyasinglepeakisalsoasignatureofout?ow-inggas(e.g.,Murrayetal.1995).]Foremission-linegas,reverberationmappingprovidesadirectmeansates-tablishingthelocationofthegas.Forabsorption-linegas,placingadistancebetweenthegasandtheioniz-ingcontinuumreliesonusingabsorption-linediagnosticstoassessthephotoionizationparameter(U),andhavingotherinformationthatconstrainsthedensity(n)ofthegas.Thedistance,r,isr=??relatedtothesequantitiesvia4πhcnU,(1)whereνLL=3.3×1015HzisthefrequencyoftheLy-manlimit.Constraintsonthedensitycancomefromtime-variability(ifionization/recombinationdominatesthevariabilitytimescale),orthepresenceofexcited-statelines.Densityinformationisnottypicallyavailableforintrinsicabsorbers,sosecondaryindicatorsmustbeem-ployedtoseparateintrinsicabsorbersfromabsorptionbyinterlopingstructures(e.g.,IGM?laments,galaxyhalosanddisks).Inorderofdecreasingutilityandimportance,thesec-ondaryindicatorsofanintrinsicoriginforanabsorption-linesystemare:(1)velocitywidth,(2)partialcoverage,(3)timevariability,(4)highphotoionizationparameter,(5)highmetallicity,(e.g.Barlow&Sargent1997).Notallintrinsicabsorbersexhibitalloftheseproperties,buttheprobabilityofanintrinsicoriginishigherifanab-sorberexhibitsmorethanoneproperty.Likewise,withtheexceptionofthe?rsttwoindicators(andthe?rstonlyinitsmostextreme,see§3.1),eachoftheseindica-torshavebeenobservedininterveningmaterial.Thus,bythemselves,nooneindicatorshouldbetakentoimplyanintrinsicorigin.Historically,the?rstcriterionhasledtothreedivi-sionsintheclassi?cationofintrinsicabsorbers.Out-?owswiththelargestvelocitydispersionsaretermed“broadabsorptionlines”(e.g.,1Weymannetal.1991,BALs,FWHM≥2000kms?).Ontheotherex-treme,intrinsicabsorberswherethevelocitydisper-sionissu?cientlysmallastocleanlyseparatetheCIVdoubletarecalled“narrowabsorptionlines”(e.g.,Hamann&Ferland1999,NALs,FWHM≤500kms?1).Sincethereisawholecontinuumofvelocitywidths,thishasledtoanin-betweenclassknownas“mini-BALs”(e.g.,Hamannetal.1997;Churchilletal.1999).Belowweexamineeachoftheseclassesintermsoftheirob-servedfrequency,andnotevariousissuesindeterminingthisnumber,includingdependenciesonquasarphysicalproperties.3.OBSERVATIONSOFTHEINCIDENCEOFVARIOUSFORMSOFOUTFLOWS3.1.BroadAbsorptionLines(BALs)Broadabsorptionlinesinthespectraofquasarsarethemosteasilyidenti?ableformsofout?ows.Thelargeve-locitywidthisveryreadilyassociatedwithaccelerated,out?owinggas.Assuch,thesegarneredmoreattentionhistoricallythantheirsmallervelocity-widthkin(e.g.Weymannetal.1985;Turnsheketal.1988;Turnshek1988;Weymannetal.1991;Voitetal.1993).Weymannetal.(1991)establishedcriteria,summa-rizedinanumbercalledtheBALnicityindex(BI),fordeterminingifanabsorptionlineconstitutedaBAL.TheBIwasamodi?edformofanequivalentwidthwherebyonecountedabsorptionthatfellbelow90%ofthetruequasarcontinuumthatwas?1contiguouslybelowthislevelformorethan2000kms.Moreover,noabsorptionwithin3000kms?1ofthequasarredshiftwascountedinordertoremovepossiblecontaminationbyabsorption-linegasnotphysicallyassociatedwiththequasarcentralengine(e.g.,interstellargasfromthequasarhostgalaxy,orintergalacticmaterialfromthehostcluster).[Note:Withaminimumvelocityof3000kms?1andaminimumcontiguouswidthof2000kms?1,thismeansthatnoab-sorptionfallingentirelywithin5000kms?1ofthequasarredshiftiscounted.]Thisindexwasestablishedusinglow-dispersiondataofhigh-redshift(1.5≤z≤3.0)ob-jectsfromtheLargeBrightQuasarSurvey,orLBQS(Foltzetal.;Hewettetal.),andwasdesignedtoyieldapuresampleofobjectswithbona?deout?ows.WenoteherethattheutilityofBIwasdrivenpurelybythedataquality(signal-to-noiseratioandresolution)oftheLBQSspectrainconjunctionwiththedesiretoremovefalsepositives(attheexpenseoflosingsometrueBALquasars).WhiletheuseofBItode?nesamplesofBALquasarshasutility,especiallyincomparingresultsbetweendatasetsofvaryingquality,itexcludessomefractionofrealhigh-velocitydispersionout?owsthatqualitativelyappeartobeBALquasarsbutjustfailtohavepositiveBI.AnimprovementontheBI,termedtheintrinsicab-sorptionindex(AI),wasdevelopedbyHalletal.(2002)toalleviatetheinadequaciesofBIinselectingobjectswherehighvelocityout?owswereclearlyobservedbutwerenotincludedasBALquasarsbytheBIcriteria(e.g.,UM660,PG).TheAIwasdesignedtobemore?exibleandinclusiveandhasbeenveryuse-fulinitsapplicationtonewerandbetterqualitydatasetsliketheSloanDigitalSkySurvey(SDSS).This?exibility,whilegoodatincludingobjectsnotpreviouslyselectedbyBI,hasincreasedthecontaminationofsamplesofin-trinsicabsorptionwhilestillnotincludingotherformsofintrinsicabsorption(e.g.,Gangulyetal.2007).TheincidenceofBALshasprimarilybeendeterminedusingopticalspectrawhere,historically,largesamplesofhigh-redshiftquasars(togetrest-frameUVcoverage)coulde?cientlybeselected(e.g.,withcolor-selection).Insuchsurveys(Hewett&Foltz2003;Reichardetal.2003;Trumpetal.2006;Gangulyetal.2007),roughly10-25%ofobjectsareobservedtohostBALs.Anissuewithoptical/UVsurveys,however,ispotentialbiasesintheselectionofquasarsagainstthosehostingBALsduetothefactthatmuchofthecontinuumisabsorbed(e.g.,Goodrich&Miller1995;Goodrich1997;Krolik&Voit1998)andintrinsicallyreddened(e.g.,Reichardetal.2003).UsingtheLBQS,wheretheobservedfrequencyofBALquasarsintheredshiftrange1.5≤z≤3is15%usingaBIcriterion,Hewett&Foltz(2003)esti-matedatrueBALfrequencyof22%fromcomparisonsinthek-correctionsofBALandnon-BALquasars.TherecentcatalogofBALquasarsusinganAIcriterionfromTrumpetal.(2006)foundaBALfrequencyof26%(intheredshiftrange1.7≤z≤4.38).Bothoftheseesti-matesarebasedontheCIVλdoublet,whichisthemostcommonlyusedspeciesinselectingintrinsicabsorptionowingtotherelativelyhighabundanceofcar-bon,thehighionizationfractionofC3+inmoderately-ionizedgas,andtheresonantabsorptionofthedoublet.Tocombatpossibleselectionbiasesintheoptical,onecanexaminequasarcatalogsselectedinotherbands.Beckeretal.(2000)examinedradio-selectedquasarsfromtheFIRSTBrightQuasarSurveyandfoundaBALquasarfrequencyofabout18%(thoughitisonly14%ifonlyBI>0objectsarecounted,comparabletootheres-timatesbasedonoptical-selection).Thisagainpredom-inantlyusedtheCIVdoubletandobjectsatz>1.7.Incidentally,severalstudies(e.g.,Brothertonetal.1998)havenowdismantledthemyththatBALsareonlyobservedinformallyradio-quiet(i.e.,fν(5GHz)/fν(3000?A)<10)objects,thoughtheirfrequencydoessigni?cantlydecreaseamongthemostradio-loudquasars(Beckeretal.2001;Gregg,Becker,&deVries2006).Wenotethatasubsetofradio-selectedBALquasarscanbeidenti?edaspolarout?ows(e.g.,Zhouetal.2006;Brotherton,deBreuck,&Schaefer2006;Ghosh&Punsly2007).Atleastoneoftheseob-jects,FIRSTJ+351758,appearstobeanopticallyreddenedandbeamedradio-quietquasar(Berringtonetal.2007).ThepresenceofBALout?owsinsuchobjectsaswellinasedge-onFRIIBALquasars,(e.g.,Greggetal.2006)indicateshigh-velocityout?owsarepresentinavarietyofgeometries.Thereisasyetnoobservationalsignatureintheabsorptionspectrathatiscorrelatedwithorientationindicators,soanygeometricallyrestrictivemodelsuchasthoseidentifyingBALout?owssolelywithequatorialwindsareeitherwrongorincomplete.Anycompletepictureofout?owsmustre?ectarangeofgeometries.Ithasyettobeestablishedobservationallyhowoftenpolarout?owsoccurcomparedtoequatorial,orifthelocationordynamicsdi?er.Inadditiontoradio-selection,onecanexaminethefrequencyofBALquasarsfrominfraredselection.Re-cently,Daietal.(2007)comparedthecatalogofBALquasars(Trumpetal.2006)fromtheThirdDataRelease(DR3)ofSDSSandtheparentsampleofDR3quasars(Schneideretal.2005)withtheTwoMicronAll-SkySur-vey(Skrutskieetal.2006,2MASS)Point-SourceCatalog(PSC).Withsomevariationwithredshift,theyreportedanoveralltrueBALquasarfractionof43±2%,markedlyhigherthanestimatesbasedonUV/opticaldataalone.Presumably,thisdi?erenceaccountsforthee?ectsofdustandabsorptionthatmaybiasUV/opticalselectiontechniquesagainst?ndBALquasars.Wepointoutthatthisestimatereliesheavilyontheau-tomatedtechniquesemployedin?ndingBALquasarsinalargedatasetsuchasSDSS.Fromacriticallookat50881.7<z<2quasarsfromSDSSDR2,Gangulyetal.(2007)notedseveralinstancesoffalse(andmissed)clas-si?cationsintheTrumpetal.(2006)catalog.Acompar-isonoftheGangulyetal.(2007)samplewiththe2MASSPSCrevealsaBALfractionof66/287(23%),completely3consistentwiththeanalysisofHewett&Foltz(2003).BlindlyusingtheTrumpetal.(2006)catalogyieldsaBALfractionof96/287(33%),consistentwiththez<2pointsfromDaietal.(2007,seetheirFigure4).Atfacevalue,thisimpliesthatnearly30%oftheTrumpetal.(2006)-2MASScross-matchedsampleconsistsoffalse-positives.Wereturntotheissueoffalse-negativesbelow.3.2.NarrowAbsorptionLines(NALs)andmini-BALsIntrinsicNALsandmini-BALshave,withinthelastdecade,cometolightasaverypowerfulandcomple-mentarymeansofstudyingout?ows.Unliketheirverybroadkin,theseabsorbersaregenerallynotblendedand,therefore,o?erameanstodetermineionizationlevelsandmetalicitiesusingabsorption-linediagnostics.Thus,NALsandmini-BALsaremoreusefulasprobesofthephysicalconditionsofout?ows.Thedrawback,however,isthattrulyintrinsicNALsandmini-BALsaremoredi?culttoidentify,sinceinterlopingstructuressuchasthecosmicweb,galaxyclusters,andgalacticdisksandhalosalsohavecomparablevelocityspreads(??800kms?1).Historically,progresswasmadebysta-tisticallyidentifyinganexcessofabsorbersoverwhatisexpectedfromrandomlydistributedinterveningstruc-tures(e.g.,Weymannetal.1979).Withimprovedtech-nologies(suchashigh-resolutionspectroscopywithlargetelescopes),wecannowtakeadvantageoftheothersecondaryindicatorstoseparateintrinsicfrominterven-ingabsorption.Inthefollowingsubsections,wediscussthefrequencyoftwosubclassesbasedonbothhistor-icalandmorerecentstudies.Wedistinguishbetweenabsorbersthatappearnearthequasarredshift(associ-atedabsorbers),andthosethatappearatlargevelocityseparations.3.2.1.Associated(zabs～zem)Absorbers(AALs)Theterm“associated”referstonarrowvelocity-dispersionabsorption-linesystemsthatlienearthequasarredshift.Ithasbeenshownthatthefre-quencyofsuchsystemsismuchlargerthanthoseatlargevelocityseparations(Weymannetal.1979;Foltzetal.1987b;Andersonetal.1987;Aldcroftetal.1994;Richardsetal.1999;Richards2001).Typically,associatedabsorbersarede?nedasthoselyingwithin5000kms?1ofthequasarredshift(Foltzetal.1986).Assuch,theywerehistoricallyverycomplementarytoBALquasarsselectedusingBI.UpdatingBALclassi?cationtore?ectthebetterdataqualityusuallyavailabletodaydoesallowforsomeconfusionamongclasses,atleastinsomecases,andthisshouldbekeptinmind.TheissueofwhattypesofquasarshostedAALswasthesubjectofmuchscrutinywithsomestudiesclaimingtoseeanex-cessofAALs(e.g.,Foltzetal.1987b),whileotherstudiesclaimednoexcess(e.g.,Sargent,Steidel,&Boksenberg1988).ItwassurmisedthatstrongAALs(i.e.,thosewithalargeCIVequivalentwidth)werepreferentiallyfoundinoptically-faint,steepradiospectrumquasars(M?ller&Jakobsen1987;Foltzetal.1988).However,morerecentstudieshavefoundthatAALsarefound(withvaryingfrequency)inallAGNsubclassesfromSeyfertgalaxies(e.g.,Crenshawetal.1999;Kriss2006)tohigherluminosityquasars(e.g.Gangulyetal.2001;Laor&Brandt2002;Vestergaard2003;Misawaetal.42007),andfromsteepto?atradiospectrumsources(Gangulyetal.2001;Vestergaard2003).AnimportantissueintheconsiderationofAALsasitrelatestoout?owsiswheretheabsorbinggasoriginates.Wenotehereafewargumentsforadirectassociationwithout?owsfromthecentralengine.Whiledetailedstudiesofindividualobjectshaveshownabsorption-linecomponentsthatmustresideinthehostgalaxyfarfromthecentralengine(e.g.,Hamannetal.2001;Scottetal.2004;Gangulyetal.2006),onthewholetherehavebeennodocumentedcasesofAALsthataretrulyredshiftedwithrespecttotheactualsystemicvelocity.IfAALsweretooriginateinthehostgalaxy,onewouldexpectsomefractionoftheabsorberstoarisefrominfallingmaterial.Infact,thevelocitydistributionofCIVAALsissharplypeakedwiththeCIVemissionred-shift(Gangulyetal.2001),implyingaclosedynami-calconnectionbetweenAALsandthebroademission-lineregion.Inaddition,blindstudiesofAALsusingsecondaryindicators?ndthat≥20%aretime-variable(Wiseetal.2004),andthat～33%showpartialcoverage(Misawaetal.2007).Fromananalysisof59z<1quasars,Gangulyetal.(2001)showedthattheoverallfrequencyofAALswas25±6%,withsomevariationwithbroad-bandspectralproperties.SimilarfrequencieshavebeenestablishedathigherredshiftbyVestergaard(%)andMisawaetal.(2007,23%),bothofwhichmadeattemptsto?lteroutcontaminationbyinterveningabsorbers.Oddly,thesefractionsarelowerthantherecentstudyofGangulyetal.(2007),who?ndanAALfrequencyof%),althoughthe5000kms?1velocitycuto?fortraditionalAALswasnotstrictlyadheredtointhatsurvey.Wenotethat%)AALsinthatstudyweremissedbytheAIselectionusedbyTrumpetal.(2006).Thesecertainlyconstitutefalse-negativesfromthestandpointof?ndingintrinsicabsorp-tion,thoughnotfromthestandpointof?ndingonlyBALquasars.WhileVestergaard(2003)didnotethatquasarswithAALsareredderonaverage,acomparisonoftheGangulyetal.(2007)samplewiththe2MASSPSCre-vealsthatthefrequencyofAALsissimilartotheparentsample(107/287,37%).Thus,theselectionofAALsquasarsisnota?ectedbyopticalbiases(e.g.,reddeningorlargeopticalabsorption)likeBALquasars.3.2.2.HighVelocityNALsThe?rstobservationalevidenceforintrinsicnar-rowvelocity-dispersionabsorptionappearingathighejectionvelocity(manytensofthousandsofkilo-meterspersecond)camenearlyadecadeagoandinclude:PG(Jannuzietal.1996),Q(Hamann,Barlow,&Junkkarinen1997),andPG(Hamannetal.1997).Modelsofquasarwindsgenerallyareabletoexplainout?owswith?v/v～1,butarechallengedbythese?v/v<<1sys-tems.Oneideaisthatthesight-linecutsacrosstheout-?owthatwouldproduceaBALunderadi?erenceori-entation(e.g.,Elvis2000;Gangulyetal.2001),butthishasyettobedemonstratedtheoretically.Thesesystemsarealsointerestingbecausetheyonlyabsorbphotonsfromthecompactcontinuum.Thus,partialcoveragein-dicatorsprovidesevereconstraintsonthegeometryofthe?ow.Fig.1.―Wepresentasummaryplotofthemaximumveloc-ityDataofabsorptionCrenshawfromseveralversusstudiesthearemonochromaticincluded:Seyfertluminosity1at3000?A.quasarsetal.(1999,?lledbluecircles);z<0.5galaxiesfromDR2LBQSBALfromquasarsLaorfrom&Brandt(2002,opengreentriangles)Palomar-GSDSSintrinsicBALquasarsfromGangulyGallagheretetal.al.(2007,?lledyellowcircles);trinsicNALsfromVestergaard(2003,?lled(2006,blueopenpentagons)in-);lar675,BALNALsfromintrinsicquasarsfromNAL/mini-BALsfromMisawaGhoshetal.&(2007,opencyanpentagons);po-inPunslyQ07,andpinkPGstars);UMPGHamannfrometal.Jannuzi(1997,etblackal.(1996,stars)andstar)themini-BAL.inIntermsofdemographics,the?rstassessmentofthefrequencyofthesesystemscamefromRichardsetal.(1999)andRichards(2001).Fromastatisticalanaly-sisexaminingthevariationinthevelocitydistributionofCIVNALswithquasarradio-loudnessandspectralindex,Richardsetal.(1999)estimatedthatasmanyas36%ofCIVNALsmayarisefromout?owinggas.Re-cently,Misawaetal.(2007)reportthatonly10-17%ofCIVNALsinthevelocityrangekms?1showevidenceofpartialcoverage.(ThusitispossiblethatRichardsetal.(1999)overestimatedthefractionofhigh-velocityNALs,orthat50-70%ofintrinsicCIVNALsdonotshowpartialcoverage.)Thisisnotastatement,how-ever,onthefractionofquasarsthathostsuchout?ows.Vestergaard(2003)reportedthathighvelocityintrin-sicNALsappearedin18±4%of1.5<z<3.6quasarsinthevelocityrangekms?1,withaboutafactoroftwovariationbetweenradiocore-dominated(17±10%)andradiolobe-dominated(33±15%)mor-phologies.Inarecentsurveyof1.8<z<3.5SDSSsources,RodriguezHidalgoetal.(2007)?ndabout12%ofquasarshavehigh-velocity?1NALsinthevelocityrangekms?,1and～2.3%inthevelocityrangekms.Thislattervelocityrangeisoftenmissedbysurveysduepurelytoobservationalcuto?s.Overthisvelocityrange,absorptionbyCIVcanbecomeconfusedwithSiIVabsorption.4.ANEXAMPLEILLUSTRATINGTHEMERITSOFCOMPREHENSIVEOUTFLOWSTUDIESWhenparameterspaceistruncated,eitherintention-ally(e.g.,throughsubclasssegregationorthedesiretoavoidfalsepositives/negatives)orunintentionally(e.g.,bydatalimitations),realcorrelationsthatcouldleadtophysicalunderstandingmaybemissed.Thewidevari-etyofobservationaltechniquesandtheimprovementsinsamplesizenowmakeitpossibletostudytheout?owphenomenoninamorecompletemannerthaneverbe-forepossible.5TABLE1DemographicsofOutflowsRanges三亿文库包含各类专业文献、生活休闲娱乐、外语学习资料、文学作品欣赏、应用写作文书、高等教育、93On the Fraction of Quasars with Outflows等内容。　
　What is the internal rate of return on the bond’s cash flows? (a) (...[NOTE: this should be formatted as a stacked fraction] (d) PV = FV/(...}