Non-Nucleosynthetic Constraints on the Baryon Dens

arXiv:astro-ph/9708016v212Jun1998May28,1998(revised)OSU-TA-14/97IASSNS-AST97/43Non-NucleosyntheticConstraintsontheBaryonDensityandOtherCosmologicalParametersGarySteigman1,2,NaoyaHata3,andJamesE.Felten4,1ABSTRACTBecausethebaryon-to-photonratioη10isinsomedoubt,wedropnucleosyntheticconstraintsonη10andfitthethreecosmologicalparameters(h,ΩM,η10)tofourobser-vationalconstraints:Hubbleparameterho=0.70±0.15,ageoftheuniverseto=14+7−2Gyr,clustergasfractionfo≡fGh3/2=0.060±0.006,andeffectiveshapeparameterΓo=0.255±0.017.Errorsquotedare1σ,andweassumeGaussianstatistics.WeexperimentwithafifthconstraintΩo=0.2±0.1fromclusters.Wesetthetiltpa-rametern=1andthegasenhancementfactorΥ=0.9.WeconsiderCDMmodels(openandΩM=1)andflatΛCDMmodels.WeomitHCDMmodels(towhichtheΓoconstraintdoesnotapply).WetestgoodnessoffitanddrawconfidenceregionsbytheΔχ2method.CDMmodelswithΩM=1(SCDMmodels)areacceptedonlybecausethelargeerroronhoallowsh0.5.BaryonicmatterplaysasignificantroleinΓowhenΩM∼1.OpenCDMmodelsareacceptedonlyforΩM∼0.4.ThecombinationofthefourotherconstraintswithΩo≈0.2isrejectedinCDMmodelswith98%confidence,suggestingthatlightmaynottracemass.ΛCDMmodelsgivesimilarresults.Inallofthesemodels,η10∼6isfavoredstronglyoverη10∼2.ThissuggeststhatreportsoflowdeuteriumabundancesonQSOlinesofsightmaybecorrect,andthatobservationaldeterminationsofprimordial4Hemayhavesystematicerrors.PlausiblevariationsonnandΥinourmodelsdonotchangetheresultsmuch.IfwedroporchangethecrucialΓoconstraint,lowervaluesofΩMandη10arepermitted.TheconstraintΓo=0.15±0.04,derivedrecentlyfromtheIRASredshiftsurvey,favorsΩM≈0.3andη10≈5butdoesnotexcludeη10≈2.1DepartmentofPhysics,TheOhioStateUniversity,174West18thAvenue,Columbus,OH432102DepartmentofAstronomy,TheOhioStateUniversity;[e-mail]3InstituteforAdvancedStudy,Princeton,NJ08540;[e-mail]4Code685,NASAGoddardSpaceFlightCenter,Greenbelt,MD20771;[e-mail](presentaddress)11.INTRODUCTIONInaFriedmann-Lemaˆıtrebigbangcosmology,theuniversalbaryonicmass-densityparameterΩB(≡8πGρB/3H20)maybecalculatedfromΩBh2=3.675×10−3(T/2.73K)3η10=3.667×10−3η10,(1)wherehisdefinedbythepresentHubbleparameterH0[h≡H0/(100kms−1Mpc−1)],Tisthepresentmicrowavebackgroundtemperature,andη10isthebaryon-to-photonnumberratioinunits10−10.Thelastmemberofequation(1)isobtainedbysettingT=2.728K(Fixsenetal.1996).Inprinciple,η10iswelldetermined(infactoverdetermined)bytheob-servedorinferredprimordialabundancesofthefourlightnuclidesD,3He,4He,and7Li,ifthenumberoflight-neutrinospecieshasitsstandardvalueNν=3.Forsomeyearsithasbeenarguedthatη10isknowntobe3.4±0.3(Walkeretal.1991;theseerrorbarsareabout“1σ”;cf.Smith,Kawano,&Malaney1993)oratworst4.3±0.9(Copi,Schramm,&Turner1995a;cf.Yangetal.1984),andthatequation(1)isapow-erfulconstraintonthecosmologicalparametersΩBandh.Inpractice,itseemsrecentlythatη10maynotbesowelldetermined,andeventhatthestandardtheoryofbigbangnucleosynthesis(BBN)maynotgiveagoodfit.Withimprovedabundancedata,itappearsthatthejointfitofthetheorytothefournuclideabun-dancesisnolongergoodforanychoiceofη10(Hataetal.1995).Theseauthorsofferseveraloptionsforre-solvingtheapparentconflictbetweentheoryandob-servation.Althoughtheysuggestthatsomechangeinstandardphysicsmayberequired(e.g.,areduc-tionintheeffectivevalueofNνduringBBNbelowitsstandardvalue3),theynotethatlargesystematicerrorsmaycompromisetheabundancedata(cf.Copi,Schramm,&Turner1995b).Thenatureofsucher-rorsisunclear,andthisremainscontroversial.Otherauthorshavereactedtotheimpendingcrisisinself-consistencybysimplyomittingoneormoreofthefournuclidesinmakingthefit(Dar1995;Olive&Thomas1997;Hataetal.1996,1997;Fieldsetal.1996).Thiscontroversyhasbeensharpenedbynewob-servationsgivingthedeuteriumabundancesonvari-ouslinesofsighttohigh-redshiftQSOs.ThesedatashouldyieldtheprimordialDabundance,butcur-rentresultsspananorderofmagnitude.Thelowvalues,D/Hbynumber≈2×10−5(Tytler,Fan,&Burles1996;Burles&Tytler1996),correspondingtoη10≈7inthestandardmodel,havebeenrevisedslightlyupward[D/H≈(3−4)×10−5(Burles&Tytler1997a,b,c);η10≈5],butitstillseemsim-possibletoreconciletheinferredabundanceof4He[YP≈0.234;Olive&Steigman1995(OS)]withstan-dardBBNforthislargevalueofη10(whichimpliesYBBN≈0.247)unlesstherearelargesystematicer-rorsinthe4Hedata(cf.Izotov,Thuan,&Lipovetsky1994,1997).SuchlowD/Hvalueshavealsobeenchal-lengedonobservationalgroundsbyWampler(1996)andbySongaila,Wampler,andCowie(1997),anddeuteriumabundancesnearlyanorderofmagnitudehigher,D/H≈2×10−4,havebeenclaimedbyCar-swelletal.(1994),Songailaetal.(1994),andRugersandHogan(1996)forotherhigh-redshiftsystemswithmetalabundancesequallyclosetoprimordial.Al-thoughsomeoftheseclaimsofhighdeuteriumhavebeencalledintoquestion(Tytler,Burles,&Kirkman1997),Hogan(1997)andSongaila(1997)arguethatthespectraofotherabsorbingsystemsrequirehighD/H(e.g.,Webbetal.1997).Ifthesehigherabun-dancesarecorrect,thenDand4Heareconsistentwithη10≈2,butmodellersofGalacticchemicalevolutionhaveamajorpuzzle:HowhastheGalaxyreducedDfromitshighprimordialvaluetoitspresent(local)lowvaluewithoutproducingtoomuch3He(Steigman&Tosi1995),wi