The cored distribution of dark matter in spiral ga

arXiv:astro-ph/0403154v15Mar2004Mon.Not.R.Astron.Soc.000,000–000(0000)Printed2February2008(MNLATEXstylefilev2.2)ThecoreddistributionofdarkmatterinspiralgalaxiesG.Gentile1⋆,P.Salucci2,U.Klein1,D.Vergani1,3andP.Kalberla11RadioastronomischesInstitutderUniversit¨atBonn,AufdemH¨ugel71,53121Bonn,Germany2SISSA,viaBeirut4,34014Trieste,Italy3ObservatoiredeParis,GEPI,CNRSUMR8111andUniversit´eParis7,5placeJulesJanssen,92195MeudonCedex,FranceAcceptedReceivedABSTRACTWepresenttheHidatafor5spiralgalaxiesthat,alongwiththeirHαrotationcurves,areusedtoderivethedistributionofdarkmatterwithintheseobjects.AnewmethodforextractingrotationcurvesfromHidatacubesispresented;thistakesintoaccounttheexistenceofawarpandminimisesprojectioneffects.Therotationcurvesobtainedaretestedbytakingthemasinputtoconstructmodeldatacubesthatarecomparedtotheobservedones:theagreementisexcellent.Onthecontrary,themodeldatacubesbuiltusingrotationcurvesobtainedwithstandardmethods,suchasthefirst-momentanalysis,failthetest.TheHirotationcurvesagreewellwiththeHαdata,wheretheycoexist.Moreover,thecombinedHα+Hirotationcurvesaresmooth,symmetricandextendedtolargeradii.Therotationcurvesaredecomposedintostellar,gaseousanddarkmattercontri-butionsandtheinferreddensitydistributioniscomparedtovariousmassdistributions:darkhaloeswithacentraldensitycore,ΛColdDarkMatter(ΛCDM)haloes(NFW,Mooreprofiles),HiscalingandMOND.Theobservationspointtohaloeswithcon-stantdensitycoresofsizercore∼roptandcentraldensitiesscalingapproximatelyasρ0∝r−2/3core.ΛCDMmodels(whichpredictacentralcuspinthedensityprofile)areinclearconflictwiththedata.HiscalingandMONDcannotaccountfortheobservedkinematics:wefindsomecounter-examples.Keywords:galaxies:methods:dataanalysis–kinematicsanddynamics–galaxies:spiral–darkmatter.1INTRODUCTIONOneofthetantalisingandyetunansweredquestionsofcon-temporaryastrophysicsisthatconcerningthenatureofdarkmatter(DM).Whileitsexistencehasbeeninferredforsev-eraldecades,onlyintherecentpastitsamounthasbeencarefullyassessedonvariousastrophysicalscales.Ongalaxyscales,theclassicalandstillmostpowerfultooltounveilDMisthatofrotationcurves.ComparedtothepioneeringworkofRubin,Thonnard&Ford(1980)andBosma(1981),modernstudiestakeadvantagefromgreatlyimprovedobser-vationaltechniques,facilitatingmuchmoresignificantvali-dationsofcosmologicaltheories,suchasColdDarkMatter(CDM).AlthoughtheprimeastrophysicalimportanceofthejuxtapositionofobservationsandmodelswasassessingtheamountofDMingalaxies(Persic&Salucci1988),thetimedquestionhasbecomeitsdistribution.Thisissuedirectlyper-tainstothenatureofthiscomponent,sinceapreciseknowl-⋆E-mail:[e-mail]fileρ(R)islikelytobedecisive:theDMnatureislikelytodeterminethehaloproperties.Yet,awealthofobservationshavenotbeenabletosettlebutthemostfundamentalissues,suchasthevalidityoftheCDMscenarioongalacticscales(e.g.Navarroetal.1996,Mooreetal.1998).Someoftheshortcomingsontheobservationalsidewererelatedtothelimitationsofopticalrotationcurves(internalrms,limitedradialextent,dustcontamination),whileradiostudieswereaffectedbythewell-knownlimitedspatialresolution.Nevertheless,progresshasbeenmade,duetotheincreasingnumberofinvestigationsondwarfspi-ralsandlow-surfacebrightness(LSB)galaxies.ThesestudiescastseriousdoubtsononeofthefundamentalpropertiesofCDMhaloes,namelytheir’cuspiness’inthegalaxycentres,andinferthepresenceoflargecoreradiiintheDMdensityprofiles(seee.g.Flores&Primack1994,Moore1994,Burkert&Silk1997,Kravtsovetal.1998,McGaughetal.1998,Salucci&Burkert2000,deBlok,McGaugh&Rubin2001,Salucci2001,deBlok&Bosma2002,Salucci,Walter&Borriello2003,Weldrake,deBlok&Walter2003,Simonetal.2003).c0000RAS2G.Gentile,P.Salucci,U.Klein,D.VerganiandP.KalberlaHowever,duetothemanystepsinthedataanalysis,therecanbesubtlesystematicerrorsthatcoulddistorttheresultsorinanycaserendertheresultsverypoorlyconstrained.Thishastriggeredofftherecentdebateconcerningthereliabilityofthedataandhowwellthemassmodelsarereallyconstrained.ThereareclaimsthattheobservationscouldactuallybeconsistentwiththedarkmatterdensityprofilespredictedbytheCDMsimulations,notonlybyconsideringtheHidataalone(vandenBoschetal.2000andvandenBosch&Swaters2001),butalsobycombiningHαandHidata(Primack2002andSwatersetal.2003a).Thisisthereasonwhyparticularcareshouldbetakeninchoosingasuitedsampleandinperformingthedataanalysis.Notethatrecentsimulations(e.g.Navarroetal.2003)donotconvergetoawell-definedvalueoftheinnerslopedowntotheresolutionlimit(about1kpc),eventhoughtheslopeofthedarkmatterdensityprofile(definedas−dlnρ/dlnr)at1%ofthevirialradiusisstillabout1.2foratypicalgalaxy.NoticealsothattheobservationalresultsonspiralgalaxiesshowadiscrepancywiththestandardΛCDMpredictionswellbeyond1kpc,i.e.wellbeyondtheresolutionlimitofthesimulations.Inthispaperwestudyasampleofgalaxyrotationcurvesidealforderivingthepropertiesofthedarkmatterhaloesaroundgalaxies;itconsistsoffivelate-typebulge-lessnormal(highsurfacebrightness)spiralgalaxies;comparedtothesampleofBorriello&Salucci(2001)wehavethesameresolutioninthecentralregions,butalargerspatialexten-sionallowi