ebook img

Chemical similarities between Galactic bulge and local thick disk red giants: O, Na, Mg, Al, Si, Ca and Ti PDF

0.53 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Chemical similarities between Galactic bulge and local thick disk red giants: O, Na, Mg, Al, Si, Ca and Ti

Astronomy&Astrophysicsmanuscriptno.alvesbrito13jan10 (cid:13)c ESO2010 January14,2010 Chemical similarities between Galactic bulge and local thick disk red giants: O, Na, Mg, Al, Si, Ca and Ti A.Alves-Brito1,2,J.Mele´ndez3,M.Asplund4,I.Ram´ırez4,andD.Yong5 1 Universidade de Sa˜o Paulo, IAG, Rua do Mata˜o 1226, Cidade Universita´ria, Sa˜o Paulo 05508-900, Brazil; e-mail: [email protected] 2 CentreforAstrophysicsandSupercomputing,SwinburneUniversityofTechnology,Hawthorn,Victoria3122,Australia 3 CentrodeAstrof´ısicadaUniversidadedoPorto,RuadasEstrelas,4150-762Porto,Portugal;e-mail:[email protected] 4 MaxPlanckInstitutfu¨rAstrophysik,Postfach1317,85741Garching,Germany 0 5 ResearchSchoolofAstronomyandAstrophysics,TheAustralianNationalUniversity,CotterRoad,Weston,ACT2611,Australia 1 0 Received:;accepted: 2 n ABSTRACT a J Context.The formationand evolution of theGalacticbulge and itsrelationship withtheother Galacticpopulations isstillpoorly understood. 4 Aims.Toestablishthechemicaldifferencesandsimilaritiesbetweenthebulgeandotherstellarpopulations,weperformedanelemen- 1 talabundanceanalysisofα-(O,Mg,Si,Ca,andTi)andZ-odd(NaandAl)elementsofredgiantstarsinthebulgeaswellasoflocal thindisk,thickdiskandhalogiants. ] R Methods.Weusehigh-resolutionopticalspectraof25bulgegiantsinBaade’swindowand55comparisongiants(4halo,29thindisk and22thickdiskgiants)inthesolarneighborhood. Allstarshavesimilarstellarparametersbutcoverabroadrangeinmetallicity S (−1.5 <[Fe/H]< +0.5).Astandard1DlocalthermodynamicequilibriumanalysisusingbothKuruczandMARCSmodelsyielded . h theabundancesofO,Na,Mg,Al,Si,Ca,TiandFe.Ourhomogeneous anddifferentialanalysisoftheGalacticstellarpopulations p ensuredthatsystematicerrorswereminimized. - Results.Weconfirmthewell-establisheddifferencesfor[α/Fe]atagivenmetallicitybetweenthelocalthinandthickdisks.Forall o theelementsinvestigated,wefindnochemicaldistinctionbetweenthebulgeandthelocalthickdisk,inagreementwithourprevious r studyofC,NandObutincontrasttoothergroupsrelyingonliteraturevaluesfornearbydiskdwarfstars.For−1.5<[Fe/H]<−0.3 t s exactlythesametrendisfollowedbyboththebulgeandthickdiskstars,withastar-to-starscatterofonly0.03dex.Furthermore, a bothpopulationssharethelocationofthekneeinthe[α/Fe]vs[Fe/H]diagram.Itstillremainstobeconfirmedthatthelocalthick [ diskextendstosuper-solarmetallicitiesasisthecaseforthebulge.Thesearethemoststringentconstraintstodateonthechemical 1 similarityofthesestellarpopulations. v Conclusions.Ourfindingssuggestthatthebulgeandlocalthickdiskstarsexperiencedsimilarformationtimescales,starformation 1 ratesandinitialmassfunctions,confirmingthusthemainoutcomesofourprevioushomogeneousanalysisof[O/Fe]frominfrared 2 spectrafornearlythesamesample.Theidenticalα-enhancementsofthickdiskandbulgestarsmayreflectarapidchemicalevolution 5 takingplacebeforethebulgeandthickdiskstructuresweseetodaywereformed,oritmayreflectGalacticorbitalmigrationofinner 2 disk/bulgestarsresultinginstarsinthesolarneighborhoodwiththick-diskkinematics. . 1 Keywords.Stars:abundances–Galaxy:abundances–Galaxy:bulge–Galaxy:disk–Galaxy:evolution 0 0 1 1. Introduction Genzeletal.2008)formedthisway,andthatthebulgeandthick : v diskmayhaveformedatthesametime.Thus,thenatureofour The Galactic bulge is the least understood stellar popula- i Galacticbulgecanbeunveiledbydetailedchemicalcomposition X tion in the Milky Way, as even its classification (classical or analysisandbycarefulcomparisonswiththethickdisk. r pseudo-bulge;Kormendy&Kennicutt2004)seemsunclear.The a Galactic bulge has signatures of an old (Ortolani et al. 1995; Althoughall recentworks agreein enhancementsof the α- Zoccalietal.2003)classicalbulgeformedrapidlyduringinten- elementsrelativetosolarabundancesinbulgefieldKgiants,the sivestarformationasreflectedintheenhancementofα-elements levelofenhancementiscurrentlyunderdebate.Basedonacom- (e.g.McWilliam&Rich1994;Cunha&Smith2006;Zoccaliet parisonofbulgegiantstarswiththickdiskdwarfstars,Zoccaliet al. 2006;Lecureuret al. 2007;Fulbrightet al. 2007;Mele´ndez al.(2006),Lecureuretal.(2007)andFulbrightetal.(2007)sug- et al. 2008; Ryde et al. 2009a,b). On the other hand its boxy gestedthatthebulgeandthethickdiskhavedifferentchemical shapeisconsistentwithapseudo-bulgeindicativeofformation compositionpatterns,andthattheα-elementsareoverabundant bysecularevolutionthroughdynamicalinstabilityofanalready inthebulgecomparedwiththethickdisk.Therefore,theyargued establishedinnerdisk. forashorterformationtimescaleandhigherstarformationrate Recently, Elmegreen et al. (2008) have shown that bulges fortheGalacticbulgethanthatforthethickdisk.Balleroetal. formed by coalescence of giant clumps can have properties of (2007)alsoconcludedthattheinitialmassfunctionsmusthave bothclassicalandpseudo-bulges,becausesecularevolutioncan been different between the two populations based on both the take place in a very short timescale (< 1 Gyr). They suggest high[Mg/Fe]andmetallicitydistributionofthebulge(see also that our Galactic bulge (and many z ∼ 2 early disk galaxies; Cescuttietal.2009).Nevertheless,thosecomparisonsshouldbe 2 Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk analysis of ∼180 clump giants by Mishenina et al. (2006), the studyof∼300nearbygiantsbyLuck&Heiter(2007),thesurvey of∼380giantsbyHekker&Mele´ndez(2007),andtheanalysis of∼320giantsbyTakedaetal.(2008).Furthermore,theUVES libraryofstellarspectra(Bagnuloetal.2003)wassearchedfor suitablediskandhalogiants. Ouranalysisof thin disk,thick disk andhalo starsis based mostly on high-resolution (R= 65,000) optical spectra taken in April 2007 with the MIKE spectrograph (Bernstein et al. 2003)ontheClay6.5mMagellantelescope,andcomplemented with observations using the 2dcoude´ spectrograph (Tull et al. 1995, R = 60,000) on the 2.7m Harlan J. Smith telescope at McDonald Observatory, the upgraded HIRES spectrograph (Vogt et al. 1994, R = 100,000)on the Keck I 10m telescope, the UVES library1 (Bagnulo et al. 2003, R = 80,000), and the ELODIEarchive2(Moultakaetal.2004,R=42,000).Themag- nitudes, population membership and instrumentation used for thedisk/halosampleareshowninTable1. Thedatawerereducedwith IRAFemployingstandardpro- cedures: correction for bias, flat field, cosmic rays and back- ground light, then optimal extraction of the spectra (using a Fig.1. H-R diagram showing our program stars. The symbols brightstartotracetheorders),wavelengthcalibration,barycen- are described in the plot. The two most luminous stars (filled tric and Doppler correction, and continuum normalization. In trianglesenclosedbycircles)arebulgegiantswhichshowabun- somecases,asdescribedbelow,avariationtothereductionpro- danceanomalieslikeO-deficiencyandNa-enhancementsimilar cedurewasnecessary.ThetiltofthelinesintheMIKEdataisse- tothoseobservedinsomeglobularclusterstars.Atypicalerror vereandvariesacrosstheCCD(e.g.Yongetal.2006),therefore barinT andloggisshown. itmustbecarefullycorrectedtoavoiddegradationofthespectral eff resolution.ThetiltwascorrectedusingMTOOLS3,specifically developedbyJ.BaldwintoaccountforthetiltedslitsinMIKE taken with care as systematic errors may be present due to the spectra. On the other hand, our HIRES spectra were extracted verydifferentstellarparameters,modelatmospheres,andNLTE using a new version of MAKEE4, an optimal extraction pack- effectsofdwarfandgiantstars.Indeed,inourconsistentanalysis age developed by T. Barlow specifically for data reduction of ofhighresolutioninfraredspectraofbothbulgeand thickdisk the improved HIRES spectrograph. MAKEE also performs an giantswithsimilarstellarparameters(Mele´ndezetal.2008),we automaticwavelengthcalibrationcross-correlatingtheextracted have shown that the bulge is in fact chemically very similar to ThAr spectra with a database of wavelength calibration solu- thethickdiskin[C/Fe],[N/Fe]and[O/Fe].Here,weextendthis tions. Both the UVES and ELODIE archive data were already worktootherα-elements(Mg,Si,Ti,Ca),andshowthatallthe extractedandwavelengthcalibrated.Theextractedspectrawere α-elementsinbulgeandlocalthickdiskgiantshaveessentially shiftedtotherestframeandcontinuumnormalizedusingIRAF. identicalchemicalabundancepatterns. Thesignal-to-noiseratio(S/N)perpixelofthereducedspectra rangesfromS/N ∼45−100forthebulgegiants(Fulbrightetal. 2006),whereasforthediskandhalogiantstheS/Nistypically 2. Observations ∼200 per pixel, ranging from ∼150 (2dCoude/McDonald) to The sample consists of 80 cool giant stars (Fig. 1) with ef- ∼200(MIKE/Magellan,ELODIE/OHP)to∼250(HIRES/Keck, fective temperatures 3800 ≤ T ≤ 5000K, surface gravities UVES/VLT),asestimatedfromrelativelyline-freeregionsofthe eff 0.5 ≤ logg ≤ 3.5, and metallicities −1.5 <[Fe/H]< +0.5. spectra. Similarnumberofthindisk(29),thickdisk(22)andbulge(25) giantswere selected, anda few (4) metal-richhalogiantswere alsoincluded. 3. AbundanceAnalysis All of our bulge giants are located in Baade’s window and are taken from Fulbright et al. (2006), who have cleaned the We have homogeneously performed all the equivalent width samplefromnonbulgegiants.Forthesebulgestars,wehaveal- (EW) measurements for the disk and halo sample. In order readypublishedanabundanceanalysisofC,N,OandFebased to check that the EW measurements of the bulge giants by onIR spectra (Mele´ndezetal. 2008).For the presentstudy we Fulbrightetal. (2006,2007)areconsistentwithoursystem for make use of the equivalent widths measured in optical spectra the disk and halo giants, we have observed one bulge star (IV- usingtheHIRESspectrograph(atR=45,000or67,000)onthe 203) with the MIKE spectrographand comparedthe EW mea- Keck-I10mtelescopebyFulbrightetal.(2006,2007). suredbyuswiththoseobtainedbyFulbrightetal.(2006,2007). To enableapropercomparisonwe havecompileda sample Theagreementissatisfactory,withameandifference(Thiswork of thin disk, thick disk and halo stars for which we have ob- -Fulbrightetal.2006,2007)of−2.0mÅandaline-to-linescat- tained our own optical spectra. The assignment of population membershipwasbasedonUVWvelocities(Bensbyetal.2004; 1 http://www.sc.eso.org/santiago/uvespop/ Reddyetal.2006).Thesampleselectionwasbasedonevaluat- 2 http://atlas.obs-hp.fr/elodie/ ingpopulationmembershipin morethan 1500giantstars from 3 http://www.lco.cl/telescopes-information/magellan/instruments- the literature, in particularan updatedversion of the Cayrel de 1/mike/IRAF tools/iraf-mtools-package/ Strobel (2001) catalog (see Ram´ırez & Mele´ndez 2005a), the 4 http://www2.keck.hawaii.edu/inst/hires/hires.html Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk 3 Table1.Programstarsdata Table 2.Sensitivitiesintheabundanceratiosbyemployingthe Kurucz models (Castelli et al. 1997). The atmospheric param- eters and α-enhancement were changed by ∆T = ±75K, eff Star V[mag] P∗[%] Instrument ∆logg = ±0.30dex, ∆v = ±0.20kms−1, and ∆[α/Fe] = t (1) (2) (3) (4) ±0.10dex. The total internal uncertainties are given in the last column Halo HD041667 8.533 00:01:99 MIKE/Magellan Abundance ∆T ∆logg ∆v ∆[α/Fe] (Px2)1/2 HD078050 7.676 00:00:100 ELODIE/OHP eff t (1) (2) (3) (4) (5) (6) HD114095 8.353 00:29:71 MIKE/Magellan HD210295 9.566 00:00:100 HIRES/Keck HD078050 ThickDisk [FeI/H] -0.09 0.01 0.07 0.00 0.11 [FeII/H] 0.02 -0.12 0.06 -0.02 0.14 HD023940 5.541 03:96:01 2dcoude/McDonald [O/Fe] -0.01 -0.13 0.00 -0.02 0.13 HD032440 5.459 08:91:01 MIKE/Magellan [Na/Fe] -0.06 0.01 0.01 0.00 0.06 HD037763 5.178 15:84:01 MIKE/Magellan [Mg/Fe] -0.06 0.05 0.04 0.00 0.09 HD040409 4.645 25:74:01 MIKE/Magellan [Al/Fe] -0.06 0.00 0.00 0.00 0.06 HD077236 7.499 00:58:42 MIKE/Magellan [Si/Fe] -0.03 -0.03 0.01 0.00 0.04 HD077729 7.630 25:74:01 MIKE/Magellan [Ca/Fe] -0.08 0.02 0.06 0.00 0.10 HD080811 8.35 00:97:03 MIKE/Magellan [Ti/Fe] -0.11 0.01 0.06 0.01 0.13 HD083212 8.335 00:94:06 2dcoude/McDonald HD099978 8.653 00:99:01 MIKE/Magellan IV203 HD107328 4.967 43:57:01 2dcoude/McDonald HD107773 6.355 20:78:02 MIKE/Magellan [FeI/H] 0.01 -0.07 0.06 -0.01 0.09 HD119971 5.454 14:85:01 MIKE/Magellan [FeII/H] 0.18 -0.18 0.04 -0.02 0.26 HD124897 -0.049 13:85:01 MIKE/Magellan [O/Fe] -0.01 -0.11 0.01 -0.02 0.11 HD127243 5.590 00:96:04 ELODIE/OHP [Na/Fe] -0.08 0.04 0.05 0.01 0.10 HD130952 4.943 01:98:01 MIKE/Magellan [Mg/Fe] 0.01 -0.04 0.03 -0.01 0.05 HD136014 6.195 24:75:01 MIKE/Magellan [Al/Fe] -0.06 0.02 0.03 0.01 0.07 HD145148 5.954 11:88:01 MIKE/Magellan [Si/Fe] 0.11 -0.09 0.03 -0.01 0.14 HD148451 6.564 00:64:36 UVES/VLT [Ca/Fe] -0.09 0.02 0.09 0.01 0.13 HD180928 6.088 00:74:26 MIKE/Magellan [Ti/Fe] -0.14 0.00 0.03 0.00 0.14 HD203344 5.570 00:97:02 ELODIE/OHP HD219615 3.694 29:70:01 ELODIE/OHP HD083212 HD221345 5.220 14:85:01 ELODIE/OHP [FeI/H] -0.09 -0.01 0.04 0.00 0.09 ThinDisk [FeII/H] 0.05 -0.12 0.05 -0.02 0.14 [O/Fe] 0.00 -0.13 0.01 -0.02 0.13 HD000787 5.255 98:02:00 2dcoude/McDonald [Na/Fe] -0.07 0.01 0.01 0.00 0.07 HD003546 4.361 78:22:00 ELODIE/OHP [Mg/Fe] -0.07 0.05 0.05 0.00 0.09 HD005268 6.163 86:14:00 HIRES/Keck [Si/Fe] -0.01 -0.04 0.00 -0.01 0.04 HD029139 0.868 98:02:00 ELODIE/OHP [Ca/Fe] -0.09 0.02 0.05 0.01 0.10 HD029503 3.861 96:04:00 2dcoude/McDonald [Ti/Fe] -0.16 0.00 0.07 0.02 0.17 HD030608 6.362 49:51:00 2dcoude/McDonald HD045415 5.543 99:01:00 UVES/VLT HD045415 HD050778 4.065 87:13:00 2dcoude/McDonald HD073017 5.673 86:14:00 ELODIE/OHP [FeI/H] -0.04 -0.02 0.09 -0.01 0.10 HD099648 4.952 99:01:00 MIKE/Magellan [FeII/H] 0.08 -0.14 0.08 -0.03 0.18 HD100920 4.301 99:01:00 MIKE/Magellan [O/Fe] 0.00 -0.14 0.00 -0.03 0.14 HD115478 5.333 99:01:00 MIKE/Magellan [Na/Fe] -0.06 0.08 0.05 0.00 0.11 HD116976 4.753 99:01:00 MIKE/Magellan [Mg/Fe] -0.02 0.02 0.03 -0.01 0.04 HD117220 9.010 95:05:00 MIKE/Magellan [Al/Fe] -0.05 0.02 0.04 0.01 0.07 HD117818 5.205 99:01:00 MIKE/Magellan [Si/Fe] 0.03 -0.05 0.03 -0.02 0.07 HD128188 10.003 98:02:00 MIKE/Magellan [Ca/Fe] -0.08 0.04 0.10 0.00 0.13 HD132345 5.838 97:03:00 MIKE/Magellan [Ti/Fe] -0.11 -0.01 0.01 0.00 0.11 HD142948 8.024 94:06:00 MIKE/Magellan HD171496 8.501 98:02:00 2dcoude/McDonald HD172223 6.485 91:09:00 MIKE/Magellan HD174116 5.24 98:02:00 2dcoude/McDonald terofσ 5 =5.5mÅ(Fig.2a).Additionally,A.McWilliamhas QD HD175219 5.355 99:01:00 MIKE/Magellan kindly made available to us the HIRES/Keck spectrum of an- HD186378 7.21 97:03:00 2dcoude/McDonald other bulge giant (I-322) for comparison purposes. Again, the HD187195 6.022 99:01:00 MIKE/Magellan agreementisgoodwithadifference(Thiswork-Fulbrightetal. HD211075 8.190 99:01:00 HIRES/Keck 2006,2007)of+2.2mÅandσ =5.5mÅ(Fig.2b).Sincemost HD212320 5.938 99:01:00 UVES/VLT QD oftheemployedlinesarerelativelystrong,thetypicalimpacton HD214376 5.036 99:01:00 HIRES/Keck HD215030 5.92 98:02:00 ELODIE/OHP HD221148 6.252 93:07:00 HIRES/Keck 5 weuseherearobuststandarddeviationbasedonthequartiledevi- ationQD(=Q3-Q1),σ =QD/1.349;seeforexampleAbu-Shawiesh Notes.— (*): The membership probabilities of the thin disk, QD etal.(2009) thickdiskandhalogiantsaregivenasthin:thick:halo. 4 Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk Fig.2.Comparisonofourequivalentwidthmeasurementswith Fig.4.Microturbulentvelocityasafunctionofeffectivetemper- linesincommonwithFulbrightetal.(2006,2007)fortwobulge ature (upper panel) and log g (lower panel). Giant stars with giant stars: (a) IV-203 and (b) I-322. The median (solid lines) [Fe/H]<–0.70and[Fe/H]≥–0.70are,respectively,represented andarobustproxyofstandarddeviation(σ ;dashedlines)are byfilledandopencircles.Thesolidlineisalinearleastsquares QD alsodisplayed. fittothedata,whoseresultsarelabeledinthefigure. pleofgiants.However,wedonotexpectanysignificantdiffer- enceswithrespecttoRam´ırez&Mele´ndez(2005b)fortherele- vantstellarparameters,exceptperhapsforasmall(∼1%)zero- pointoffsetintheT scale,whichisirrelevantheresinceweare eff performingadifferentialstudy. Reddeningforthebulgestarswasestimatedfromextinction maps(Stanek1996),while forthecomparisonsamplebothex- tinction maps (Mele´ndezetal. 2006b) and Nai D ISM absorp- tion lines were used. The E(B-V) values based on the D lines wereobtainedasfollows.Intheopticalthincasetherelationbe- tweencolumndensityN (unitscm−2)andequivalentwidthEW (unitsmÅ)is: N =1.13×1017EW/(f λ2), (1) (Spitzer1968).The f valuesare0.64and0.32,respectively,for the5889.950and5895.924Ålines(NISTdatabase6).Notethat theaboverelationbetween N(NaI)andequivalentwidthholds onlyforlinesonthelinearpartofthe curveofgrowth,i.e., for smallvaluesofE(B-V);forreddeninglargerthanafew0.01mag the interstellar lines must be modeled in detail (e.g. Welty et Fig.3. Comparison between evolutionary gravities from Y2 al. 1994) to avoid underestimation of the column densities. In particular we use the profile fitting program FITS6P (Welty et (filled circles) and Padova (open circles) isochrones, and trigonometricgravitiesforgiantstarswithgood(σ(π)/π≤10%) al.1994). The N(NaI)densitywastransformedto N(H)usingthere- Hipparcosparallaxes.Thesolidanddashedlinesdepictperfect agreementandvariationsof±0.1dex,respectively. lationfoundbyFerletetal.(1985): logN(HI+H )=(logN(NaI)+9.09)/1.04, (2) 2 abundancesfrom these EW differencesis negligible. Thus, the wherebothN(NaI)andN(H)areincm−2.Finally,E(B-V)was analysis of the faint bulge giants, the bright disk and the halo computedfromthetotalhydrogendensity(Bohlinetal.1978): giants,isessentiallyinthesamesystem. E(B−V)= N(HI+H )/5.8×1021, (3) Photometric temperatures were obtained using optical and 2 infrared colors and the infrared flux method Teff-scale of whereN(H)isincm−2andE(B-V)inmagnitudes.Althoughthis Ram´ırez&Mele´ndez (2005b). We note that the new improved relationseemsnotwellestablishedforE(B-V)<0.1,Ramirezet IRFM calibration of Casagrande et al. (2010) only applies to dwarfandsubgiantstarsandthuscannotbeappliedtooursam- 6 http://physics.nist.gov/PhysRefData/ASD/lines form.html Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk 5 al.(2006)haveshownittobeveryaccurateforaE(B-V)=0.01 havestellarparametersinthesamesystem,i.e.intheRam´ırez& star. Mele´ndez(2005b)temperaturescaleandloggintheHipparcos Albeitnotusedin thepresentwork,we shouldmentionfor scale. completeness that in addition to reddening maps and Na D in- Themicroturbulencewasobtainedbyflatteninganytrendin terstellar lines, E(B-V) can also be estimated from other inter- the [FeI/H] versus reduced equivalent width diagram. The mi- stellar features such as the diffuse interstellar band at 862 nm croturbulencefollow tight relations with temperature and logg (Munarietal.2008),aswellasmulticolorphotometry(e.g.Sect. (Fig.4),withascatterofonly0.20kms−1: 4.2ofMele´ndezetal.2006b;Ram´ırezetal.2006)andpolariza- tion(e.g.Fosalbaetal.2002). vt(Teff)=3.33−4.23×10−4Teff (MARCS)(8) The stellar surface gravities were derived from improved v(T )=3.40−4.41×10−4T (Kuruczovershooting)(9) Hipparcos parallaxes (vanLeeuwen 2007) for the sample of t eff eff nearbygiantstarsandassumingadistanceof8kpcforthebulge v(logg)=1.82−0.186logg (MARCS)(10) t giants.Inaddition,Yonsei-Yale(Y2;Demarqueetal.2004)and Padova isochrones (daSilvaetal. 2006) were employed to de- v(logg)=1.84−0.202logg (Kuruczovershooting)(11) t termine evolutionarygravities, as well as the inputmasses that The lines used for analysis (presented as online material) were adopted for the trigonometric gravities. In order to esti- mate the Y2 gravities, we generated a fine grid of isochrones, have been carefully selected to minimize the impactof blends. assuming[α/Fe]=0and+0.3for[Fe/H]> 0and[Fe/H]< −1, Completelyavoidingblendsisalmostanimpossibletaskincool, relatively metal-rich giants as in our sample, since their spec- respectively,andlinearlyinterpolatedinbetween.Allsolutions tra are heavily blended with many atomic and molecular lines allowed by the error bars were searched for, adopting as final (e.g. Coelho et al. 2005), in particular due to CN. We have result the median values. The Padova gravities were obtained usingtheBayesiantoolPARAM7.AsshowninFig.3,both(Y2, tried to avoid blendingby performingspectralsynthesisof CN (using the line list of Mele´ndez & Barbuy 1999) and discard- Padova) evolutionary gravities are in excellent agreement with ing the atomic lines whose equivalentwidths are contaminated thetrigonometricgravitiesofournearbygiantswithreliable(un- certainties≤10%)Hipparcosparallaxes.Theevolutionarylogg by more than 10% by CN. The cool giant Arcturus (Hinkle et values required small zero-pointcorrectionsof −0.04 (Y2) and al. 2000) was also carefully inspected to discard lines that are +0.07dex (Padova),to be on the same scale as the Hipparcos- severely contaminated with other features. In some cases even lineswhichareblendedbymorethan10%havetobeincluded, basedresultsforoursamplegiantstars(Fig.3).Bolometriccor- especiallyforelementsotherthanironbecauseonlyafewuse- rectionsfromAlonsoetal.(1999)wereadopted. fullineswereavailable.Forheavilyblendedlineswehaveper- We use iron lines to check our T and log g, but we do eff notassumeapriorithatouradoptedeffectivetemperatures,sur- formedthe measurementsby fitting onlythe unblendedpartof theprofile,ordeblendingthefeatureusingtwoormorecompo- face gravities, 1D model atmospheres, gf-values, selection of nents.Apreliminaryversionofourlinelist(Hekker&Mele´ndez lines, equivalent width measurements and LTE line formation, would result in absolute excitation (zero slope of Fei abun- 2007)hasbeentestedin∼380field(Hekker&Mele´ndez2007) dances vs. excitation potential) and ionization (A = A ) and39opencluster(Santosetal.2009)giants,andthefinallist FeI FeII equilibria.We use the nearbydisk/halo giants to determinethe hasbeenalreadyusedinfieldbulge(Rydeetal.2009b)andglob- slopes (d(A )/d(χ )) and differences between Fei and Feii, ularcluster(Mele´ndez&Cohen2009)giants. FeI exc The stellar chemical abundances were obtained from an followedbymoststars.Ourtestsoftheionizationandexcitation balancesofFeiandFeiilinesrevealedthatmostofthesample equivalent width analysis using the 2002 version of MOOG (Sneden 1973). The same transition probabilities were applied giants(58stars)satisfyourrelativespectroscopicequilibriumof to both the bulge and comparison samples. In the present iron lines within the uncertainties, therefore the overall agree- work, we employedboth Kuruczmodelswith convectiveover- ment is encouraging. Nevertheless, the photometric stellar pa- shooting(Castellietal.1997)andspeciallycalculatedMARCS rametersof22stars(8thindisk,5thickdisk,1halo,and8bulge (Gustafssonetal.2008)1Dhydrostaticmodelatmospheres.For stars) required some adjustmentsto be on our relative spectro- theMARCSmodels,bothα-enhanced([α/Fe]=+0.2and+0.4) scopicequilibriumscale.Thecorrectionsbasedonthetrendfol- lowedbythebrightdisk/halogiants,forwhichthephotometric and scaled-solar abundances models were constructed; for the Kurucz models, adjustments of [Fe/H] were applied to simu- stellar parameters (and stellar spectra) were more reliable than late the effects of α-enhancement on the model atmospheres forthebulgesample,is: (Salarisetal.1993): d(A )/d(χ )=0.008dexeV−1 (Kuruczovershooting), (4) FeI exc ∆[Fe/H]=log(0.64×10[α/Fe]+0.36) (12) d(A )/d(χ )=0.003dexeV−1 (MARCS), (5) FeI exc Theeffectsoffailingtoaccountforthevariationsin[α/Fe] stars within 2-σ (σ = 0.011dex eV−1) were consideredto ful- arerelativelysmallforadifferenceof[α/Fe]=+0.1dex(Table fill our relative excitation equilibrium. Ionization balance was 2), but could be important(∼ 0.1 dex) for the typicalenhance- achievedif ment of [α/Fe] ∼ +0.3-0.4 seen in thick disk, halo and bulge stars. A(FeII)−A(FeI)=0.08dex(Kuruczovershooting), (6) We estimate that our stellar parameters have typical un- certainties of ∆T ≈ ±75K, ∆logg ≈ ±0.3dex and ∆v ≈ A(FeII)−A(FeI)=0.00dex(MARCS), (7) eff t 0.2kms−1. The impactof these uncertaintieson the abundance and stars within ±0.07dex were considered to fulfill our rel- ratios,aswellasthetotalabundanceerrorsduetouncertainties in T , logg, v and [α/Fe] added in quadrature, are shown in ative ionization equilibrium. After these corrections were per- eff t Table 2, butnote thatsome uncertaintiesare likelyto becorre- formedthedeviatingthindisk,thickdisk,haloandbulgestars, latedtosomedegree(see,e.g.,Fulbrightetal.2007).Theuncer- 7 http://stev.oapd.inaf.it/cgi-bin/param taintiesgiveninTable2areprobablyconservativeinsomecases, 6 Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk Table 3. Internal zero-point abundances adopted for our giant starsusingMARCSandKuruczmodels. Sun Giants Specie Literaturea Thisworkb MARCSc Kuruczd (1) (2) (3) (4) (5) Fe 7.50,7.45,7.56,7.50 7.49±0.04 7.53 7.54 O 8.83,8.73.8.71,8.69 8.74±0.04 8.83 8.84 [OI] Na 6.33,6.27,6.27,6.24 6.24±0.04 6.24 6.24 Mg 7.58,7.54,7.58,7.60 7.56±0.04 7.65 7.66 Al 6.47,6.28,6.47,6.45 6.39±0.04 6.56 6.56 Si 7.55,7.62,7.54,7.51 7.54±0.03 7.60 7.63 Ca 6.36,6.33,6.36,6.34 6.34±0.02 6.32 6.30 Ti 5.02,4.90,4.92,4.95 4.94±0.05 4.83 4.81 Notes.—(a):SolarphotosphericabundancesfromGrevesse& Sauval(1998),Reddyetal.(2003),Bensbyetal.(2003,2004) Halo (4) Bulge (24) andAsplundetal.(2009);(b):solarabundancesbasedonour Thick disk (22) Thin disk (29) previous work (Mele´ndez et al. 2006a; Mele´ndez & Ram´ırez 2007; Mele´ndez et al. 2009) using different (e.g. McDonald, Keck, Magellan) solar spectra;(c,d):Ourinternal zero-points forgiantsrepresent thethin-diskabundances at[Fe/H]=0.0. Fig.5.[O/Fe]vs.[Fe/H]forthesamplestarsemployingMARCS Thesezero-pointsarenotabsoluteabundancesandshouldonly (top)and Kurucz (bottom)modelatmospheres.Symbolsare as beusedwhenbothoursamegf-valuesandanalysistechniques explainedin the figure. Note, however,that hereafter the bulge areadopted starsI-264andIV-203areomittedfromallfigures(refertothe textfordetail). asshownbytherelativelylowscatter(asafunctionofmetallic- veresothatat[Fe/H]=−1our1D-basedabundancescouldbein ity)ofourabundanceratios;the uncertaintiesintheabundance ratios[X/Fe]areprobably≤0.10dex. errorby<∼ 0.2dex.However,giventhesimilarityinparameters The uncertainty of ±75K in T is the based on the up- betweenthe bulgeanddisk giants,the relative abundanceratio eff differences – which we are primarily interested in here – will per and lower envelopes(excludingoutliers) of the differences besignificantlysmallerandthusinconsequentialforourconclu- between the slopes of iron abundance vs. excitation potential sions. (d(A )/d(χ ))oftheinitialphotometrictemperaturesandthe FeI exc adopted zero-points (relations 4 and 5). Note that these differ- Of particularimportanceto our work are the adoptedzero- encesareduenotonlytoerrorsinthetemperaturecalibrations, pointsofourabundancescale.MostworksusetheSuntodefine photometric errors and the quality of the spectra, but also due the zero-point of the thin disk at [Fe/H] = 0.0, but due to the to errors in E(B-V), which although for the nearby giants are differencesbetweendwarfsandgiants,thisapproachmayintro- low,forthebulgegiantsmaybehigher.Nevertheless,sincewe ducesystematicerrors.Instead,inthepresentworkweuseseven correctalloutliersfromouradoptedzero-points(whichinsome thindiskgiantswith−0.1 <[Fe/H]<+0.1dex(HD29503,HD cases may be due to incorrect reddenings), we are inmune to 45415, HD 99648, HD 100920, HD 115478, HD 186378, HD large errors in E(B-V). Our error of 0.3 dex in log g is based 214376)to define our zero points, which are shown in Table 3 onthedifferencesbetweenFeIIandFeIfromtheinitialtrigono- for both the Kurucz and MARCS models. In Table 3 we also metricloggandtheadoptedzeropointinFeII-FeI(relations6 show for comparison different abundance analysis of the Sun. and7).Notethatsincewearebasingouruncertaintiesontheup- Ascanbeseen,ourzero-pointsforFe,NaandCaareroughlyin perandlowerdiscrepanciesoftheinitialinputstellarparameters agreementwiththesolarabundances,butforO,Mg,andSithe and the adoptedzero-points,our uncertaintiesin T and log g giantsshowahigherzero-pointby∼+0.1dex,andforAldiffer- eff areconservative.Forthebrightdisk/halostarsinternalerrorsof encesashighas+0.15arefound.Ontheotherhand,Tiislower 50KinT and0.2dexinloggmaybemoreadequate.Rydeet by∼0.1dex.Thesezero-pointoffsetsof−0.1to+0.15dexshow eff al.(2009b)suggeststhattheuncertaintiesadoptedinthestellar thatitisnotstraightforwardtocompareabundancesobtainedin parametersofourmethod(whichwasusedtodeterminetheat- giantswiththosefoundindwarfs. mosphericparametersintheirsample)aresoundfortheirbulge These zero points we have found for giants are internal giants. In particular,uncertaintiesin T higher than ∼75K are for ourparticularset of gf-valuesandanalysis techniques.For eff excludedbasedontherelativelylowstar-to-starscatteroftheir comparisonwithchemicalevolutionmodels,theabsolutezero- [O/Fe]ratios. pointsshouldbe adoptedfromanalysis of the Sun (Asplundet No predictions of the effects of 3D hydrodynamical mod- al. 2009),which representswell the localthindisk at[Fe/H] = elsinsteadofclassical1Dmodelsusedhereareavailableasyet 0.0,exceptforsmallpeculiaritiesof a few 0.01dex(Mele´ndez for the exact stellar parameters of our targets (Asplund 2005). etal.2009;Ram´ırezetal.2009). Colletetal.(2007)haveperformedsuchcalculationsforslightly ThefinalstellarparametersaregiveninTable4andTable5 less evolved red giants (T ≈ 4700 and logg ≈ 2) and found fortheMARCSandtheKuruczmodels,respectively,whilethe eff that the the 3D abundance corrections for the species consid- abundanceratiosaregiveninTable6andTable7.Theequivalent ered herein are expected to be modest: |∆logǫ| <∼ 0.1dex at widthmeasurementsaregiveninTables8-15,whichisavailable [Fe/H]∼0.Atlowermetallicitythe3Deffectsbecomemorese- onlyintheelectronicversionofthearticle. Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk 7 Halo (4) Halo (4) Bulge (23) Bulge (23) Thick disk (22) Thick disk (22) Thin disk (29) Thin disk (29) Fig.6. [Mg/Fe] as a function of [Fe/H] for MARCS model at- Fig.8. [Ca/Fe] vs. [Fe/H] for MARCS model atmospheres. mospheres.Symbolsasexplainedinthefigure. Symbolsasexplainedinthefigure. Halo (4) Halo (4) Bulge (23) Bulge (23) Thick disk (22) Thick disk (22) Thin disk (29) Thin disk (29) Fig.7. [Si/Fe] vs. [Fe/H] for MARCS model atmospheres. Fig.9. Plot of [Ti/Fe] against [Fe/H] for the sample stars em- Symbolsasexplainedinthefigure. ployingMARCSmodelatmospheres.Symbolsareasexplained inthefigure. 4. Results In Fig. 5 we show the [O/Fe] ratios obtained in this work for in excellentagreement.The mean differenceis only −0.04dex bothMARCSandKuruczovershootingmodelatmospheres.As (OH - [Oi]) with a scatter of 0.10 dex. Since this is identical canbeseen,thereisagoodoverallagreementbetweenMARCS totheestimatederrorin[O/Fe]fromOHfoundinMele´ndezet and Kurucz models. In particular, the difference in iron abun- al. (2008),the uncertaintiesin our stellar parametersare likely dance (MARCS - Kurucz) is only −0.02 dex (σ = 0.03 dex). somewhatoverestimated,asalreadydiscussedabove.Although Thus, in the following figures, we present results based on the theoxygenabundancesobtainedfrom[Oi]andOH linesagree MARCS models only. Even though both set of models give well, the results obtainedfromOH lines have less scatter, pos- similar chemical abundance ratios (see Tables 6,7), for com- sibly becauseseveralOH lineswere usedinstead of relyingon parison with chemical evolution models we suggest to adopt onlyoneor two forbiddenlines. Therefore,forcomparisonsof theMARCSresults,sincetheywerecomputedwiththecorrect oxygenabundanceswithdetailedchemicalevolutionmodelsof [α/Fe] ratio. Our disk/halo comparison sample shows that the thethindisk,thickdiskandbulge,webelievethattheOHlines oxygenabundancesobtainedherefromthe[Oi]630and636nm arepreferable(Mele´ndezetal.2008;Rydeetal.2009a,b).The lines and in Mele´ndez et al. (2008)from infraredOH lines are [Oi]-basedoxygenabundancesconfirmsthesimilaritybetween 8 Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk Halo (4) Bulge (22) Halo (4) Thick disk (22) Bulge (22) Thin disk (29) Thick disk (22) Thin disk (29) Fig.10. Mean α-elements abundance ratio Fig.12. [Na/Fe] vs. [Fe/H] for MARCS model atmospheres. ([(O,Mg,Si,Ca,Ti)/Fe]) as a function of [Fe/H] for MARCS Symbolsasexplainedinthefigure. modelatmospheres.Symbolsasexplainedinthefigure. Halo (4) Bulge (20) Thick disk (22) Thin disk (29) Fig.13. [Al/Fe] vs. [Fe/H] for MARCS model atmospheres. Fig.11.Fitof[α/Fe]vs.[Fe/H]formetal-poor([Fe/H]< −0.5) Symbolsasexplainedinthefigure. and metal-rich ([Fe/H] ≥ −0.5) bulge (top),thick disk (center) andbothbulgeandthickdisk(bottom)stars.Bothstellarpopu- lationscanbefittedbysimilarrelations,withascatteraslowas al.2005;Carrettaetal.2009).Fulbrightetal.(2007)foundthat σ=0.03dex. these two giants have high Na and Al abundances reminiscent oftheabundanceanomaliesseeninglobularclusters,whichwe confirm. Thus, the oxygen abundances of these two stars most thebulgeandthelocalthickdisk,whichwepreviouslydemon- likelydonotreflectthetypicalbulgecomposition. strated basedonOH lines. Thisis in contrastto someprevious Theresultsfortheotherα-elements(Mg,Si,Ca,Ti)studied works on the topic (Zoccalietal. 2006; Fulbrightetal. 2007; hereareshowninFigs.6-9.Ascanbeclearlyseen,thechemi- Lecureuretal. 2007), which argue that [O/Fe] in the bulge is calpatternsofthebulgeandthickdiskareindistinguishablealso higherthaninthethickdiskbasedonacomparisontodiskdwarf for those elements, reinforcing our previous findings based on stars. oxygenabundances.TheaverageofourMgabundancesforthe As pointed out by Fulbright et al. (2007), two O-deficient 7starswith[Fe/H]≥0is[Mg/Fe]=0.1±0.1,ingoodagreement stars(I-264andIV-203)at[Fe/H] ≈ −1.25havepeculiarabun- withtheresultsfrommicrolensedbulgedwarfs,whichtypically dancessimilartotheO-Naanti-correlationseeninglobularclus- have [Mg/Fe]≈ +0.1 (Cohen et al. 2008, 2009; Johnson et al. ters(e.g.Grattonetal.2004;Cohen&Mele´ndez2005;Yonget 2008;Bensbyetal.2009a).Thelatestpreliminaryresultsbased Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk 9 This work Fulbright et al. Fig.14. [C/O] vs. [O/H] for the bulge (triangles), thick disk Fig.16.[O/Mg]asafunctionof[Fe/H]forbulgegiantsaccord- (filledcircles)andthindisk(opencircles)withdatatakenfrom ingtoFulbrightetal.(2007)(opentriangles)andourwork(filled Mele´ndez et al. (2008) and Ryde et al. (2009b). The original triangles). Both studies show a shallow trend up to aboutsolar (undepleted)C abundanceswere estimated fromC+N (referto metallicity and a step decrease in [O/Mg] for higher metallici- the text). Overplotted, we show model predictions as given in ties. Recent predictions by Cescutti et al. (2009) are shown as Cescuttietal.(2009;c.f.theirFigure5). solid(WW95model),shortdashed(WW95+M92),longdashed (WW95+MM02)anddot-shortdashed(MM02)lines.Allmod- elshavebeenshiftedby−0.2dexin[O/Mg](seeCescuttietal. 2009for a descriptionof the modelsand an explanationof the empiricaloffset). and high [α/Fe] to [Fe/H]>0.0 and low [α/Fe], but Reddy et al. (2006) and Ram´ırez et al. (2007) did not find any evidence of sucha knee.A re-examinationofthe latter results including newobservationsofkinematicallyselectedthick-diskmetal-rich objects is underway (Reddy et al., in preparation)and will ad- dressthisdiscrepancy.Theproblemisthatathigh[Fe/H]there is a significant numberof thick-disk candidatesthat follow the thin disk abundance pattern. That suggests that hot kinematics alone cannot be used to separate the thin from the thick disk, especially athigh[Fe/H]. Atsuper-solarmetallicities the prob- lemmaybeunsolvablebecausetheabundancepatternsofboth Halo (4) Bulge (22) disksmerge.Interestingly,thechemicalsimilaritiesbetweenthe Thick disk (22) Galacticbulgeandthelocalthickdiskgiantstarswefindinthis Thin disk (29) workinfactextendtosuper-solarmetallicities.Yet,asexplained above,itremainstobedemonstratedthatthefewselectedthick diskstarsarebonafidethickdiskmembersratherthankinemat- icallyheatedthindiskstars. Fig.15.[O/Mg]asafunctionof[Fe/H](toppanel)and[Mg/H] FromTable 1 we see that three giants, which are kinemati- (bottompanel).Symbolsasexplainedinthefigure. cally classified as thick disk (HD 77236,HD 107328)andthin disk(HD30608)members,presentambiguouskinematicalpop- ulation.ThestarHD77236couldbeeitherathickdisk/halostar, onmicrolensedbulgedwarfstars(Bensbyetal.2009b)alsoin- whilethestarsHD107328andHD30608bothhavesimilarlike- dicate similarities between the bulge and thick disk for Ti and lihoodofbelongingtothethinorthickdiskpopulations.These Mgatallprobedmetallicities(−0.8<[Fe/H]<+0.5). starshave[Fe/H]=(−0.67,−0.43,−0.28)and[α/Fe]=(+0.33, Evenmoreclearresultsarefoundwhenwecombinethere- +0.30, +0.08), respectively, which means that both HD 77236 sults for all α-elements, as shown in Fig. 10. It is clear that andHD107328couldbeindeedthickdiskstars,whilethestar the chemical patterns of the bulge and thick disk are indistin- HD30608hasanabundancepatternconsistentwithathindisk guishableintheirabundancepatternsuptothemetallicityrange starat[Fe/H]∼−0.3. where the thick disk is unambiguously identified, i.e., up to Linear fits of both the bulge and the thick disk [α/Fe] vs. [Fe/H] ≈ −0.3. Bensby et al. (2003, 2004) reported the exis- [Fe/H]relationsupto[Fe/H]=−0.3,showthatbothpopulations tence of a knee connectingthick-disk stars from[Fe/H]≈ −0.3 followidenticalpatterns,with a star-to-starscatter of onlyσ = 10 Alves-Britoetal.:ChemicalsimilaritiesbetweentheGalacticbulgeandthickdisk 0.03 dex. We thus set the most stringent constraints to date on the[Na/Fe]and[Al/Fe]trendsagreewellforthebulgeandthe the chemicalsimilarity of bulge and local thick disk stars. The local thick disk, although this is not too surprising given that metallicity of the bulge extends to significantly higher [Fe/H] there is no obvious offset between the thick and thin disk for than that, which remains to be convincingly demonstrated for thosetwoelements. thethickdisk,aspreviouslydiscussed. Previousworks(Fulbrightetal.2007;Lecureuretal.2007) Inordertoquantifyhowsimilararethebulgeandthethick have found high [Mg/Fe] ratios in bulge stars, as well as high disk at all metallicities, we have divided the stars in a metal- [X/Fe]ratiosinotherelementswithrespecttothethickdisk.It poor and a metal-rich sample with the division set somewhat may seem surprising that using the same equivalent widths as arbitrarily at [Fe/H] = −0.5, and we have performedlinear fits Fulbright et al. (2007) we find significantly lower [Mg/Fe] in of [α/Fe] vs. [Fe/H]. We find that the metal-poor part of both bulge giants at all metallicities. However, as mentioned in the stellarpopulationscanbefittedbyessentiallyidenticalrelations section 3, the zero-points we use in our analysis are based on (Fig.11):[α/Fe]=0.22-0.10×[Fe/H]forthebulge(σ=0.03 seventhindisksolarmetallicitygiants,which,asshowninTable dex)and [α/Fe] = 0.27- 0.07× [Fe/H] for the thick disk (σ = 3,arenotthesameasthezero-pointsbasedontheSun,whichis 0.03 dex). These relations are identical to within ±0.01 dex at ∼1400Khotterandhasasurfacegravity∼300timeshigherthan [Fe/H]=−1.5andtowithin±0.02dexat[Fe/H]=−0.5,hence our giants. The differences shown in Table 3 between the Sun bothdatasetscanbefittedbyasinglerelationfollowedbyboth and our giants could be due to the different impact of 3D and stellarpopulations: non-LTEeffectson giantand dwarfs,as well as problemswith lineblendingingiants.Furthermore,sincewecomparebulgegi- [α/Fe]=0.268−0.065×[Fe/H] ([Fe/H]<−0.5), (13) antstothickdiskgiants,ourconclusionsonthesimilarityofthe bulge and thick disk is independentof the adopted zero-point, withaverylowstar-to-starscatterofonly0.03dex(Fig.11). unlikethecomparisonsofZoccalietal.(2006),Fulbrightetal. A similar exercise for the most metal-rich bulge and thick (2007)andLecureuretal.(2007),thatcomparedbulgegiantsto disk stars with [Fe/H] ≥ −0.5, results in a single relation fol- diskdwarfs.AlthoughbothFulbrightetal.(2007)andLecureur lowedbybothstellarpopulations(Fig.11): et al. (2007) used the giants Arcturus ([Fe/H] ∼ −0.5) and µ Leo ([Fe/H] ∼+0.3)as referencestars, their zero-pointsare ul- [α/Fe]=0.104−0.381×[Fe/H] ([Fe/H]≥−0.5), (14) timately based on the Sun8, which as shown in Table 3, may be inadequate for the study of giants. It is important to men- with a low star-to-star scatter of only 0.06 dex (Fig. 11). This tion that Fulbright et al. (2007) also included 17 nearby disk scatter is higher than for the more metal-poor stars (σ = 0.03 giants (mostly from the thin disk) in order to check whether dex), but fully explained by the higher uncertainties in the theyfollowthesameabundancepatternasdiskdwarfs,butmost analysis of the more crowded spectra of the metal-rich stars. of their giants were observed at a lower resolving power, R Nevertheless,wedonotdiscardthattheremaybesomecontam- ∼ 30,000, implying higher errors in the abundances obtained inationinthemetal-richsamples.Wedemonstratethusthatatall fromthecrowdedspectraofmetal-richgiantstars.Furthermore, probedmetallicities(−1.5<[Fe/H]<+0.5),boththebulgeand their comparison sample of disk dwarfs included several stud- thickdiskstellarpopulationshaveastrikingchemicalsimilarity. ies which may have different systematic offsets between them. TheAlabundancesofthemostmetal-richbulgestars([Fe/H] For example the study by Reddy et al. (2003) used Stro¨mgren &0)seemenhanced([Al/Fe]∼+0.25),butNaseemssolar(al- photometryto estimate effectivetemperaturesusingthe (b−y) thoughwith a large scatter) at these metallicities, as illustrated calibrationbyAlonsoetal.(1996).However,asrecentlyshown in Fig.12-13. Thus,the enhancementin Alinmetal-richbulge byMele´ndezetal.(2010,inpreparation)usinguvby-βphotome- giants is probably not related to the Al-Na correlation seen in tryofsolartwins,thiscalibrationhasazero-pointerrorof130K. globularclusters,butmostlikelyduetothefactthatthetwoAli In turn, this implies abundancevariations (∆[X/H]) from -0.12 lines employed are blended and are more difficult to deblend dex(ObasedonOItriplet,whichwasthemainabundanceindi- in the relatively moderate S/N spectra of the metal-rich bulge catorusedbyReddyetal. 2003)to+0.13dex(Ti).Yet, dueto giants. This is reinforcedby recent studies of metal-rich bulge a compensatingchangein iron, for mostelements studied here dwarfsthroughmicrolensing(Cohenetal.2008;Johnsonetal. (exceptfor[O/Fe]thatisaffectedby-0.22dex;basedontheOI 2008; Bensby et al. 2009a), which find [Al/Fe] ∼ +0.10 dex. triplet),the∆[X/Fe]rangesfrom-0.08dex([Si/Fe])to+0.03dex Thus,ourAlabundancesforbulgestarswith[Fe/H]&0arethus ([Ti/Fe]).Ontheotherhand,Bensbyetal.(2003,2004)adopted likely affected by systematic errors, which should be borne in spectroscopictemperatures,thereforenotonlytheirTeffbutalso mindincomparisonswithchemicalevolutionmodels. their[X/Fe]abundanceratiosarelikelymoreaccurateforcom- parisonofthesereddenedbulgeregions. Thus, the main reason why our conclusions regarding the 5. Discussion abundancetrendsinthebulgeincomparisonwiththethickand Our previous homogeneous abundance analysis of OH lines thindiskdifferfromthefindingsofpreviousstudies(e.g.Zoccali in high resolution infrared Gemini+Phoenix spectra was the etal.2006;Fulbrightetal.2007;Lecureuretal.2007)isthatwe first to show that the Galactic bulge and the local thick disk perform a strictly differential analysis of red giants with very have indistinguishable[O/Fe] trendsup to at least metallicities similarparametersforallpopulationsratherthanrelyingonei- [Fe/H]≈ −0.3wherethethickdiskisunambiguouslyidentified ther literature data or using dwarf stars for the disk samples. (Mele´ndezetal.2008).Inthepresentworkwehaveanalyzedthe 8 similar problems were already recognized by Gratton & Sneden forbiddenoxygenlinesanddemonstratedthatthoseindeedgive (1990)intheirabundanceanalysisofµLeo,whichwasultimatelyrel- consistentresultswiththeOHlines.Importantly,wealsoextend ativetotheSunsincetheyusedsolargf-values.Theyremarkthattheir the conclusions of Mele´ndez et al. (2008) to other α-elements procedure might introduce inconsistencies due to the different atmo- (Mg,Si,Ca,Ti).Theα-elementsinthebulgeandlocalthickdisk sphericstructuresofµLeoandtheSun.Furthermore,independentlyof starsarethesameintherange−1.5<[Fe/H]<−0.3,showinga theadopted gf-values, bydefinition thesolar abundances areneed to verylowstar-to-starscatterin[α/Fe]ofonly0.03dex.Similarly, obtain[X/Fe].

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.