The SCUBA HAlf Degree Extragalactic Survey (SHADES) - III. I(14)

Determining an accurate position for a submm galaxy (SMG) is the crucial step that enables us to move from the basic properties of an SMG sample - source counts and 2-D clustering - to an assessment of their detailed, multi-wavelength properties, their con

14Ivisonetal.

least,thedif cultyofmeasuringaccurateandconsistentradio uxdensitiesusingdatawithdifferentuvcoverage.

Fig.8showsalog–logplotofS24µm/S850µmversusS1.4GHz/S850µmforSHADESSMGs,withdifferentsymbolsrepresentingidenti cationsmadeindifferentwavebands(radioplusmid-IR;mid-IRonly;radioonly).Aswehavediscussed,S1.4GHz/S850µmissensitivetoredshift(andtemperature)andtheChapmanetal.parameterisationisshownasahorizontalbar.Itisapparentthat24-µm uxdensityiscorrelatedsigni cantlywithredshift,asexpectedfortheKcorrectionatthatwavelength.TheSMGwithS1.4GHz/S850µm>0.1isSXDF850.21,themostobvi-ousexampleofalocalgalaxyinthesample(z=0.044,Simpsonetal.2006;seeAppendixA,Fig.A2).

8THEDIAGNOSTICPOWEROFMID-IRCOLOURIvisonetal.(2004)usedacolour-colourplottoexploitthestrongdiagnosticpotentialofthemid-IRfordiscriminatingbetweengalaxiesdominatedbystarburstsandAGN.Keyspectralindicesforhigh-redshiftgalaxiesareavailablebetween3.6and24µmsincetherest-frame～3–10µmslopeforstarburstsissteeperthanforAGN,witha atterregionbetween1and3µm,whereasAGNexhibitpower-lawspectracoveringrest-frame～0.2–10µm(e.g.Mrk231).

Fig.9showsS24µm/S8µmversusS8µm/S4.5µm.Weexpectthelow-S8µm/S4.5µmportion–theleftside–tobeoccupiedbyz>～0.7starbursts,representedherebytheredshifttrackofArp220.High-redshiftstarburstsareexpectedinthelowerleftre-gionofFig.9,butspectralfeaturesinArp220’sSEDyieldsev-eralkinkswhichlimitthediagnosticpoweroftheplot;power-lawAGN,representedinFig.9byMrk231,trackleft-to-rightwithin-creasingredshiftacrossthelowerthirdoftheplot,returningtotheleftonlyz>～4.TheredshifttrackofNGC6240–aclassi-calCompton-thickAGNdisplayingmid-IRPAHfeaturesindistin-guishablefromthoseofastarburstgalaxy–overlapssigni cantlywiththecolour-colourspaceoccupiedbyArp220,atz～0.4andatmuchhigherredshifts,butmostoftheconfusingoverlapoccurswhereweexpectNGC6240-typeSEDsatz～0.6andMrk231-likeSEDsatz>6.

DoSMGsstandoutfroma24-µm-selectedSpitzersampleincolour-colourspace?Fig.9showsanindependentgalaxysam-pleselectedat24µmintheLH,atdepthscommensuratewithourSpitzeridenti cations,andwecanseethatthedataareclus-teredalongthetrackoccupiedbyArp220-likeSEDsforz 0.7,withasigni cantnumberofsourcesalongthetrackde nedbyaMrk231-likeSED.SMGsaresimilarlypositionedanddonotstandoutclearlyfrom24-µm-selectedgalaxies.However,thehatchedareasofFig.9–thosecolourcombinationswherewemightex-pectto ndSMGswiththehighestredshifts(z>～4)–arewellpopulatedwithSMGs.ThefractionofSMGsintheseregionsissigni cantlylargerthanforthecontrolsample:we ndonly14percentofthe4,457mid-IR-selectedgalaxiesinthehatchedregions.BasedontheChapmanetal.parameterisationofS850µm/S1.4GHz,theirmedianredshiftishigherthanthatoftheradio-detectedfrac-tionofSHADES,3.2versus2.8,althoughwenotethatsomeofthebestz<～1candidatesalsofallintheseregions,e.g.SXDF850.52.Nevertheless,itseemssensiblethatanysearchforahigh-redshiftpopulationofSMGsshouldbaseitstargetselectiononacombina-tionoftheS850µm/S1.4GHz,S1200µm/S850µm(Ealesetal.2003;Greveetal.2004),S24µm/S8µmandS8µm/S4.5µmcolours.

9CONCLUDINGREMARKS

Wehavedeterminedthemostlikelyradioand/ormid-IRidenti ca-tions,andhenceaccuratepositions,fortheSHADESSourceCata-loguepresentedbyCoppinetal.(2006).Wehaveidenti edrobustcounterpartstoovertwothirdsofthissample(54and46percentat1.4GHzand24µm,respectively),presentingoptical,24-µmandradioimagesofeachSMG.

Employingthesubmm/radio uxdensityratioasanindicatorofredshift,guidedbytheChapmanetal.(2005)parameterisation,we ndamedianredshiftof2.8fortheradio-identi edsample,somewhathigherthanthespectroscopicmedian.

Wepresentadiagnosticcolour-colourplot,basedonSpitzerdata,inwhichweidentifyregionscommensuratewithSMGsatveryhighredshift.

Wefurtherexploitouridenti cationstoshowthat:for pruningobservedthetrendsparentinSHADESidenti cationsamplerate(cf.giveIvisonnostrongetal.2002);rationaleexpectations, uncertaintieswithinnosubmmevidencepositionforsigni cantareconsistentadditionalwiththeoreticalsourcesofpositionalerror;

than signi cantlywouldbeexpectedmoreSMGsbychance,havemultipleindicativerobustofphysicalcounterpartsasso-ciations.ThesemultiplesystemsaremostcommonamongstthebrightestSMGsandaretypicallyseparatedby2–6arcsec,～15–50/sinikpcatz～2,consistentwithearlyburstsseeninmergersimulations.

ACKNOWLEDGEMENTS

AWBacknowledgessupportfromtheAlfredP.SloanFoundationandtheResearchCorporation.ISacknowledgessupportfromtheRoyalSociety.CSandSRacknowledge nancialsupportfromthePPARC.IAandDHHacknowledgesupportfromCONACYTgrants39548-Fand39953-F.

References

AlexanderD.A.,BauerF.E.,ChapmanS.C.,SmailI.,BlainA.W.,BrandtW.N.,IvisonR.J.,2005a,ApJ,632,736

AlexanderD.A.,SmailI.,BauerF.E.,ChapmanS.C.,BlainA.W.,BrandtW.N.,IvisonR.J.,2005b,Nature,434,738

AretxagaI.etal.,2007,MNRAS,submitted(astro-ph/0702503)AshbyM.L.N.etal.,2006,ApJ,644,778

BargerA.J.,CowieL.L.,SandersD.B.,FultonE.,TaniguchiY.,SatoY.,KawaraK.,OkudaH.,1998,Nature,394,248

BargerA.J.,CowieL.L.,SmailI.,IvisonR.J.,BlainA.W.,KneibJ.-P.,1999,AJ,117,2656

BargerA.J.,CowieL.L.,RichardsE.A.,2000,AJ,119,2092BertoldiF.etal.,2000,A&A,360,92

BiggsA.D.,IvisonR.J.,2006,MNRAS,371,963

BinneyJ.,Merri eldM.,1998,GalacticAstronomy,PrincetonBlainA.W.,IvisonR.J.,SmailI.,1998,MNRAS,296,L29

BlainA.W.,KneibJ.-P.,IvisonR.J.,SmailI.,1999,ApJ,512,L87BlainA.W.,ChapmanS.C.,SmailI.,IvisonR.J.,2004,ApJ,611,725

BondJ.R.,EfstathiouG.,1987,MNRAS,226,655

BorysC.,ChapmanS.,HalpernM.,ScottD.,2003,MNRAS,344,385

c0000RAS,MNRAS000,000–000