On the Origin of Metal-poor Stars in the Solar Neighborhood
aa r X i v : . [ a s t r o - ph . S R ] J u l Draft version July 17, 2020
Typeset using L A TEX twocolumn style in AASTeX62
On the Origin of Metal-poor Stars in the Solar Neighborhood
Timur ¸Sahin and Sel¸cuk Bilir Akdeniz University, Faculty of Science, Department of Space Sciences and Technologies07058, Antalya, TURKEY Istanbul University, Faculty of Science, Department of Astronomy and Space Sciences34119, Beyazıt, Istanbul, TURKEY (Received January 1, 2020; Revised January 1, 2020; Accepted July 17, 2020)
Submitted to ApJABSTRACTWe determined the ages, the kinematic parameters and Galactic orbital parameters of six metal-poor(-2 . < [Fe/H] < − . Gaia
DR2 was used. Highresolution ELODIE spectra of the six dwarfs were also used to obtain accurate [Fe/H] abundances andup-to-date [ α /Fe] abundances. The calculations for stellar ages were based on Bayesian statistics, withthe computed ages falling in the range 9.5-10.1 Gyrs. On the basis of the metallicities and ages, sixHMP stars are members of halo (HD6755, HD84937, BD +42 3607) or members of the low-metallicitytail of the thick disk (HD 3567, HD 194598, HD 201891). However, Galactic orbital parameters suggestthin disk (HD 84937, HD 194598), thick disk (HD 3567, HD 201891), and halo (HD 6755, BD +42 3607)population. The dynamical analysis was also performed for the escape scenario from the candidateGCs. The tidal disruption of a dwarf galaxy was also considered to be as an alternative origin. HD6755, HD 194598, and HD 3567 with their retrograde orbital motions are likely candidate stars for atidally disrupted dwarf galaxy origin. However, HD 194598 relationship with NGC 6284 presents aninteresting case. Its encounter velocity is low (16 ±
28 km s − ) and their ages and metallicities arevery nearly consistent with each other at the 1 σ level. The rest of the HPM sample stars have a 4% to18% probability of encountering with selected GCs for 1.5 tidal radii. This indicates that a globularcluster origin for the program stars is unlikely. Keywords:
Stars: individual - Stars: kinematics and dynamics - Galaxy: solar neighbourhood - Galaxy:abundances INTRODUCTIONThe unevolved late-type stars of different Galacticpopulations with large proper motions play an impor-tant role for our knowledge of the Galaxy’s stellar con-tent, and for reconstructing the Galactic chemical evo-lution. Late-type, metal-poor stars (the F to K typedwarfs) with narrow and less blended spectral linesin their spectra are excellent probes at medium res-olution for the chemical history of the earliest stellarpopulations. It is equally important to note that theongoing surveys of such stars, such as GALactic Ar-chaeology with HERMES (GALAH) (Heijmans et al.2012, De Silva et al. 2015),
Gaia -ESO Public Spectro-scopic Survey (GES) (Gilmore et al. 2012), the LargeSky Area Multi-Object Fibre Spectroscopic Telescope (LAMOST) (Zhao 2012), the APO Galactic EvolutionExperiment (APOGEE) (Allende Prieto et al. 2008),Sloan Extension for Galactic Understanding and Explo-ration (SEGUE) (Yanny 2009), the RAdial Velocity Ex-periment (RAVE) (Steinmetz et al. 2006) greatly de-pend on accurately calibrated stellar parameters fromsuch stars. In addition, discovery of several halo sub-structures (Grillmair 2006; Sesar et al. 2007; Belokurovet al. 2007; Kepley et al. 2007; Klement 2010; Helmi etal. 2017; Li et al. 2019) in the literature also requiresup-to-date stellar abundances for halo stars. However,an almost distinct abundance pattern in α -elements ispresent for the thin and the thick disk member of theGalaxy (Bensby et al. 2003, 2005, 2014; Reddy et al.2003, 2006; Nissen & Schuster 2011; Adibekyan et al.2012; Karaali et al. 2016, 2019; Yaz G¨ok¸ce et al. 2017;Plevne et al. 2020). The latter group tend to be slightlymetal-poor but have enhanced α -element abundances.On the other hand, while the majority of the stars inthe solar neighborhood appear to be members of the thindisk, several of them are found to be associated with theGalactic thick disk. Furthermore, a small percentage ofthese nearby stars with very old ages and low metallici-ties are believed to belong to the halo component of theGalaxy (Bilir et al. 2006, 2012). Altogether, a concisedefinition of stellar populations is necessary not only intesting and constraining assumptions but also in model-ing early nucleosynthesis. For this, the metal-poor dwarfstars are certainly the best candidates.In this work, we provide fresh metallicities and α -element abundances for those F-type high proper mo-tion (here after HPM) stars from the ELODIE library(Prugniel & Soubiran 2001). We determined their ages,kinematic parameters and Galactic orbital parameters.Since a concise description of their origin is unknown,correlating metallicities and α -element abundances aswell as abundances for a large number of elements withtheir kinematics and computed Galactic orbital param-eters may shed light on the nature of those metal-poordwarf stars of HPM. Although most of the stars that aresubject of this work have been extensively studied in theliterature, the reported stellar parameters of some wereseen to present large variations. At the very least, a newmeasurements of metallicities was seen to be necessarysince the metallicity is also of importance for accuratedetermination of ages. Even a crude discussion of spacemotions with their chemical compositions in the contextof their origin could have been proven to be a useful tool.If far from descriptive, it might, at least, spur the discus-sion on the HPM nature of the stars. In fact, a recentstudy on s-process enriched metal-poor star HD 55496by Pereira et al. (2019) commented on the possible ori-gins of this field halo object as a second generation glob-ular cluster star. However their estimate was based onthe past encounter probabilities (noted to be very small,i.e. ≤ SELECTION OF PROGRAM STARSThe
ELODIE (Soubiran 2003) library contains 1953spectra of 1388 stars. Our first selection criterion wasspectral type. Among the 625 F spectral type stars, wechose the stars with an HPM designation in the library.As a second criterion, binarity was inspected using the
SIMBAD database . The spectroscopic binaries fromthe sample were removed. In consequence the wholeHPM sample consists of 54 F-type stars, together withthe ELODIE library (Prugniel & Soubiran 2001) witheffective temperatures 4900 < T eff < . < log g (cgs) <
5, metallicities -2 . < [Fe/H] < < d <
283 pc. Thefinal selection was done for halo candidate stars withlow metallicities. We specifically focussed on metal-poorcandidate stars with a metallicity in the range -2 . < [Fe/H] < − Observations
High resolution ( R = 42 000) and high signal-to-noiseratio (S/N; upto 140) spectra of the stars were obtainedon the 1.93 m telescope of the Haute Provence Ob-servatory equipped with the ELODIE fiber-fed cross-dispersed echelle spectrograph that provided a spectralcoverage from 3900 to 6800 ˚A. The log of observationsis given in Table 1. The spectra were continuum nor-malized, wavelength calibrated and the radial velocitycorrected by the data-reduction pipeline run at the tele-scope. Since some problems were seen in continuum nor-malization of the spectra from the library, the
ELODIE spectra were re-normalized, using a in-house developedinteractive normalization code LIME (¸Sahin 2017) in
IDL prior to the abundances analysis. The character ofthe spectra of the program stars is displayed in Figure 1.Many lines in the spectra were suitable for abundanceanalysis are apparently unblended.LIME was also employed for the line identification pro-cess. It provides the most probable identifications forthe line of interest and lists the recent atomic data (e.g.Rowland Multiplet Number-RMT, log gf , and LowerLevel Excitation Potential-LEP) that are compiled from No comments on binarity was reported in the
ELODIE archive. They cover a range from −
122 to +633 mas.yr − in RAand from −
899 to +340 mas.yr − in DEC.) Atmospheric pa-rameters are from the ELODIE archive at http://atlas.obs-hp.fr/
ELODIE /; Moultaka et al. (2004)
HD 6755BD+423607HD 201891HD 194598HD 84937HD 3567
Wavelength (Å)
HD 6755BD+423607HD 201891HD 194598HD 84937HD 3567 R e l a t i v e F l u x Figure 1.
A small region of the spectrum for six HPM sample stars. Identified lines are also indicated. the literature (e.g. from
NIST database). Equivalentwidths (EWs) are obtained using both SPECTRE (Sne-den 1973) and LIME codes. The results for a representa-tive sample of weak and strong lines agreed well within ± THE ABUNDANCE ANALYSISFor abundance analysis of halo sample stars in thisstudy, we have acquired
ATLAS9 model atmospheres(Castelli & Kurucz 2004) computed in local thermody-namic equilibrium (
LTE ) (ODFNEW). The elementalabundances for the program stars were computed by us-ing an
LTE line analysis code MOOG (Sneden 1973) .The details of the abundance analysis and the source ofatomic data are the same as in ¸Sahin & Lambert (2009),¸Sahin et al. (2011, 2016). In the following subsections,we discuss the adopted line list, atomic data, and thederivation of model parameters.3.1. The line list A systematic search was performed for unblendedlines. They were identified using MOOG by calculat-ing the synthetic spectrum for the observed wavelengthregion. Only for a very small percentage of the acceptedlines, the spectrum synthesis was preferred to a directestimate of EW. For the iron lines, one has to be cau-tious for their selection in metal-poor stars. Because, 3D(time dependent) hydrodynamical model atmospheresof cool stars highlight the importance of lower-level en-ergy dependence of 3D abundance corrections for neutraliron lines (e.g. Dobrovolskas et al. 2013). Also, low-excitation (i.e. E exc < i lines in metal-poorstars may provide higher abundances compared to linesof higher excitation energies (e.g. Lai et al. 2008). Ourinspection of the mean abundances from low-excitationlines of Fe i indicates ≈ T eff determination appears to benegligible. The gf -values for chosen lines of Fe i and Fe ii were taken from compilation of Fuhr & Wiese (2006).Our selection of iron lines included 254 Fe i lines and 29Fe ii lines. Chosen lines of Fe i and Fe ii are exhibited inTables A1-A3. The list of identified lines for elements Table 1. []Log of observations for the HPM sample stars. The S/N values in the raw spectra are reported near 5000 ˚A. Spectral typestaken from INCA.Stars α δ l b
Sp.Type Exposure S/N MJD V
HEL (h:m:s) ( ◦ : ’ : ” ) ( ◦ ) ( ◦ ) (s) (2400000+) (km s − )HD 6755 01:09:43 +61:32:50 125.11 − − − − − other than Fe with most up-to-date atomic data is pro-vided in Tables A4 - A6.To check the uncertainties in log gf -values and tominimize systematic errors, we have derived solar abun-dances using stellar lines. The lines were measured offthe solar flux atlas of Kurucz et al. (1984) analyzedwith the solar model atmosphere from the Castelli &Kurucz (2004) grid for T eff = 5788 K and log g = 4.40cgs. Our analysis gave a microturbulent of 0.8 km s − and the solar abundances in Table 2. As can be seen,the solar abundances were recreated fairly well. The so-lar abundances obtained are compared with those fromAsplund et al. (2009) in their critical review. In ref-erencing stellar abundances to solar values, we use oursolar abundances. Hence, the analysis in this study isperformed differentially with respect to the Sun. Such adifferential approach in the analysis will reduce the er-rors due to uncertainties in the oscillator strengths, theinfluence of the spectrograph characteristics and also de-viations from LTE. Errors in effective temperatures fromspectroscopic excitation technique can also stem fromsystematic errors in oscillator strengths as a function ofexcitation potential. The same is also true for errors inequivalent widths.A further health check on log gf - values was per-formed by comparing the log gf -values employed in thisstudy to those included in Gaia -ESO line list v.5 whichwas provided by the
Gaia -ESO collaboration. The
Gaia -ESO line list includes a list of recommended linesand atomic data (hyperfine structure-hfs- corrected gf -values) for the analysis of FGK stars. It is worth tonote that several lines in the spectra of FGK stars arestill unidentified (Heiter et al. 2015). Over 228 commonFe i lines, the difference in log gf -values (log gf Thisstudy -log gf Gaia − ESO ) is found to be 0.001 dex. There is nodifference in log gf - values for 25 common lines of Fe ii .Only two Fe ii lines showed differences at ≈ ii ii gf -values for elements other than Fe reported in A4 -A6 with the number of lines given in subsequent paren- Table 2.
Solar abundances obtained by employing the solarmodel atmosphere from Castelli & Kurucz (2004) comparedto the photospheric abundances from Asplund et al. (2009).The abundances presented in bold type-face are measured bysynthesis while remaining elemental abundances were calcu-lated using the line EWs. ∆log ǫ ⊙ (X) = log ǫ ⊙ (X) This study − log ǫ ⊙ (X) Asplund
This study AsplundSpac. log ǫ ⊙ (X) n log ǫ ⊙ (X) ∆ log ǫ ⊙ (X)(dex) (dex) (dex)C i ± ± i ± ± i ± ± i ± ± ii ± ± i ± ± i ± ± ii ± ± i ± ± ii ± ± V i ± ± Cr i ± ± ii ± ± i ± ± i ± ± ii ± ± Co i ± ± Ni i ± ± Cu i ± ± i ± ± ii ± ± ii ± ± i ± ± ii ± ± ii ± ± ii ± ± ii ± ± ii ± ± thesis are as follows: 0.004 dex for C i (1) and Mg i (7);-0.004 dex for Na i (4); -0.11 dex for Si i (6); -0.03 dexfor Si ii (2); -0.09 dex for Ca i (28), Ni i (46), and Y ii (4); 0.02 dex for Sc ii (8) and Ti i (37); no difference forTi ii (27), Cr i (25), Sr i (1), Zr i/ii (1,3), Ce ii (1), andSm ii (1); -0.07 dex for Cr ii (7); 0.06 dex for Sr ii (2);0.03 dex for Ba ii (3); 0.08 dex for Nd ii (3).Since Fe i and Fe ii abundances were employed to con-strain the model atmosphere parameters in this study,we must consider non-LTE effects on Fe. These effectswere considered to be insignificant for Fe ii lines (Berge-mann et al. 2012; Lind, Bergemann & Asplund 2012;Bensby et al. 2014). Lind et al. (2012) demonstratedthat the departures from LTE for Fe ii lines of low ex-citation potential ( < > -3dex was negligible. To account for non-LTE effects onFe i lines, we used 1D non-LTE investigation of Lind etal. (2012). For this, the non-LTE web tool INSPECTprogram v1.0 (see Lind et al. 2012) was employed. Forcommon Fe i (also see Tables A1-A3), the non-LTE cor-rections computed by us via INSPECT program was ofabout 0.1 dex with the exception of 4994 ˚A Fe i line.The line required a non-LTE correction (∆(log ǫ (Fe i )) of -0.6 dex for HD 3567. We obtained comparable depar-tures from the LTE for this line for other HPM stars. Forthe 5198 ˚A Fe i line in the spectrum of HD 201891, wepredict ∆(log ǫ (Fe i ))=0.4 dex, and it is 0.5 dex for 5216˚A Fe i line and HD 3567. Its effect on the model atmo-sphere parameters or equilibria of Fe is not discernible,i.e. the magnitudes of the slopes of the relationship be-tween the abundance of iron from Fe i lines and the ex-citation potential of each line (or the reduced EW ofeach line) showed no significant variation when this linewas excluded from the analysis. We also followed therecipe given by Lind, Bergemann & Asplund (2012; seealso their Figure 4, 5, and 7) to scrutiny the role of non-LTE effects in determining the atmospheric parametersfrom Fe i and Fe ii lines in the current analysis. ForHD 201891, HD 194598, and HD 3567, the non-LTE ef-fects on the excitation balance in 1D was less than 100K. The excitation balance for HD 6755 required a cor-rection of about 50 K. For the most metal poor HPMstars in our sample, BD+42 3607 and HD 84937, it wasfound to be <
150 K and <
200 K, respectively. Thecorrections estimated by Lind, Bergemann & Asplund(2012) was typically ≤
50 K.Bergemann et al. (2017) reported non-LTE abun-dance corrections for Mg i lines (e.g. 4571, 5528, and5711 ˚A). They were computed using 1D hydrostatic ∆(log ǫ (Fe i )) = log ǫ (Fe i ) NLTE - log ǫ (Fe i ) LTE ) model atmospheres. The corrections for a representa-tive model of ( T eff , log g , [Fe / H]) = (6000, 4.0, -2.0),were ranging from 0.04 dex for 5528 ˚A line to 0.07 dexfor 4571 ˚A line. Bergemann et al. (2017) used wave-lengths and oscillator strengths from Pehlivan Rhodinet al. (2017) who provides theoretical transition proba-bilities for Mg i lines. Comparison of log gf -values thatare common to both Bergemann et al. (2017) and Pehli-van Rhodin et al. (2017) provided a difference of only-0.10 ± gf -valuesfor the hfs components are obtained from Lawler et al.(2014) for V and from Den Hartog et al. (2011) for Mnand from Lawler et al. (2015) for Co. We assumed asolar system isotopic ratios for these elements. Moreonto this, the study by Bergemann et al. (2019) on 3Dnon-LTE formation of Mn lines in late type stars pro-vided new experimental transition probabilities for man-ganese lines. They noted that for some of the manganeselines, the new log gf - values were typically 0.05-0.1 dexlower than the old values. The difference between ourand Bergemann’s (hfs included) gf -values is small, i.e. − . ± .
08 dex.Over a sample of metal-poor stars, the study by Berge-mann & Gehren (2008) for the non-LTE effects onmanganese lines shows that non-LTE effects begin todominate with increasing T eff and decreasing metallic-ity. Surface gravity was noted to become important at[Fe/H] ≤ -2 dex and T eff > i lines regardless of theirmultiplet. It was also noted that, the Mn abundancescould be underestimated via 1D LTE analysis by ≈ -0.2dex for the models of dwarf stars and the change inmetallicity would not have changed this value. The linesof multiplets 23 (e.g. 4761, 4762, 4765, and 4766 ˚A) and32 (e.g. 6016 and 6021 ˚A) were listed to be reliable linesof Mn i for abundance analysis. We employed these mul-tiplet members for abundance analysis of the HPM starsin this study.Zhang et al. (2008) investigated non-LTE effects onthe scandium for the Sun and reported that the non- T e ff ( K ) l og g ( c g s ) −3.0−2.5−2.0−1.5−1.0−0.5 [ F e / H ] ( d e x ) Figure 2.
Model parameters reported in the literature for the program stars (filled circle). The new measurements of stellarparameters ( T eff , log g , [Fe/H]) as well as estimated errors in those was also indicated with a cross symbol (see also Table A9). LTE effects were negligible for Sc ii . However, the strongnon-LTE effects were observed for Sc i . For the Sun, thenon-LTE effect on solar Sc i abundance was +0.18 dexand the Sc i abundance did not present any change.Bergemann et al. (2010) reported non-LTE correc-tions for cobalt and noted the metallicity as the mainstellar parameter that determines the magnitude of non-LTE correction to be applied on cobalt abundance. Intheir Table 3, they listed non-LTE abundance correc-tions for the common Co i line at 4121 ˚A for selectedmodel atmospheres (see also their Figure 4). Inspectionof their table provided the non-LTE abundance correc-tion for the 4121 ˚A line. It is ranging from ≈ ≈ ≈ Model atmosphere parameters
Table 3.
Model atmosphere parameters of HPM stars fromthis study.
Stars T eff log g [Fe/H] ξ ( K) (cgs) (dex) (km s − )HD 6755 5175 ±
140 2.90 ± ± ±
130 3.72 ± ± ±
160 4.20 ± ± ±
165 4.10 ± ± ±
175 3.58 ± ± ±
140 3.50 ± ± Since the reported stellar parameters in the literaturefor some of the program stars were seen to present largevariations (see Figure 2), we decided to obtain new mea-surements of model atmosphere parameters spectroscop-ically in this study.
Fe I 4442.35
Å / EW = 47 mÅ
Fe I 4447.73
Å / EW = 34 mÅ
Fe I 4466.56
Å / EW = 45 mÅ
Fe I 4919.00
Å/ EW =52 mÅ
Fe I 4938.82
Å / EW = 22 mÅ
Fe I 5232.95
Å / EW = 70 mÅ
Fe I 5302.31
Å / EW = 17 mÅ
Fe I 5324.20
Å/ EW =52 mÅ
Fe I 5266.56
Å / EW = 42 mÅ
Fe I 5339.94
Å / EW = 23 mÅ
Fe I 5383.38
Å / EW = 33 mÅ
Fe I 5446.92
Å/ EW =79 mÅ
Fe I 5497.53
Å / EW = 32 mÅ
Fe II 5169.05
Å / EW = 79 mÅ
Fe II 4583.84
Å / EW = 34 mÅ
Fe II 5316.62
Å/ EW =23 mÅ
Wavelength (Å) R e l a t i v e F l u x Figure 3.
The observed (filled circles) and computed (full red line) line profiles for Fe i and Fe ii lines. The measured EWsand their wavelengths are indicated at the top of each panel. The computed profiles show synthetic spectra for the abundancesreported for BD+42 3607 in Table A1-A3. We determined model atmosphere parameters – effec-tive temperature, surface gravity, microturbulence, andmetallicity – by using neutral and ionized Fe lines (Ta-ble A1-A3), and the full line list can be found online. Aselection of observed and computed line profiles of thechosen Fe i and Fe ii lines for BD+42 3607 are shown inFigure 3.The effective temperature was estimated first throughimposing the condition that the derived abundancebe independent of the lower level excitation potential(LEP). If all lines have the same LEP and a similarwavelength, the microturbulence ( ξ ) is found by requir-ing that the derived abundance be independent of thereduced equivalent width (EW). The precision in de-termination of the microturbulent velocity is ± − . We determined the surface gravity ( log g ) by re-quiring ionization equilibrium, e.g. that Fe i and Fe ii lines produce the same iron abundance (Figure A1).Iron lines are quite numerous even in very metal-poorsample stars. Since these model atmosphere parametersare interdependent, an iterative procedure is necessary. Small changes are made to the model parameters be-tween each above mentioned step. We also verified thatno significant trend of iron abundances with wavelengthwas present. The resulting stellar model parametersalong with our determination of model parameters ofthe Sun are listed in Table 3.The program stars are also listed in the Gaia
DR2(Gaia collaboration, 2018). Our spectroscopic temper-atures are in excellent agreement with T eff values re-ported by the Gaia consortium for HD 6755, HD 201891,and HD 194598. KINEMATIC PROPERTIES OF THE HPMSTARSThe space velocity components of the six HPM starswere calculated using the algorithm of Johnson &Soderblom (1987) for the J2000 epoch. The U , V and W are the components of a space velocity vector for agiven star with respect to the Sun. The transforma-tion matrices use the notation of right-handed system.Hence, U is defined to be positive towards the Galacticcentre, V is positive in the direction of the Galacticrotation and W is positive towards the North GalacticPole.Equatorial coordinates ( α , δ ), radial velocities (RV),proper motions ( µ α cos δ , µ δ ) and trigonometric parallax( π ) of the program stars were taken into account in thecalculations (Table 4). The proper motions and trigono-metric parallaxes of the stars are taken from the Gaia
DR2 catalogue (Gaia collaboration, 2018) while the ra-dial velocities were obtained from the
ELODIE spectra.Since no astrometric data for HD 84937 was availablefrom the
Gaia
DR2 catalogue, the proper motions andtrigonometric parallax of the star were taken from the re-reduced
Hipparcos catalogue (van Leewuen, 2007). Er-rors in the proper motion components from two differ-ent catalogues were compatible with each other, how-ever, the errors in the trigonometric parallaxes werenot consistent. The relative parallax errors calculatedfor five stars from the
Gaia
DR2 catalogue are smallerthan 0.006, whereas the relative parallax error calcu-lated from the
Hipparcos catalog for HD 84937 is in theorder of 0.06. In space velocity calculations, the biggestuncertainty comes from the errors in distances. Even ifthe relative parallax error calculated for HD 84937 is ap-proximately 0.06, it is expected that the errors in spacevelocity components will not be large because HD 84937is close to the Sun (i.e. d = 73 pc).To obtain the precise space velocity components, thefirst order Galactic differential rotation correction byMihalas & Binney (1981) was applied. The correc-tions for U and V space velocity components were foundto be − . ≤ du ≤ .
73 and − . ≤ dv ≤ . − , respectively, while the W space velocity com-ponent is not affected by a first order approximation.Following the differential rotation correction, the spacevelocity components were also corrected for the pe-culiar velocity of the Sun (LSR), which is ( U ⊙ , V ⊙ , W ⊙ ) LSR = (8.50 ± ± ± − (Co¸skuno˘glu et al. 2011). The total space velocitiesof the stars ( S ) were calculated via the square root ofthe sum of the squares of the space velocity compo-nents. The uncertainties in the space velocity compo-nents ( U err , V err , W err ) were computed by propagatingthe uncertainties in proper motions, distances and radialvelocities using the algorithm by Johnson & Soderblom(1987). The corrected space velocity components, totalspace velocities and their errors for six HPM stars arelisted in Table 5.To determine their population types, we employedthe kinematic method of Bensby et al. (2003). Thekinematic method assumes a Gaussian distribution forthe Galactic space velocity components in the followingform. f = 1(2 π ) / σ U σ V σ W exp (cid:20) − U σ − ( V − V asym ) σ − W σ (cid:21) (1)Here, σ U , σ V , and σ W are the characteristic velocitydispersions for different Galactic populations: 35, 20 and16 km s − for thin disk (D); 67, 38 and 35 km s − forthick disk (TD); 160, 90 and 90 km s − for halo (H). V asym is the asymmetric drift velocity: -15, -46 and -220km s − for thin disk, thick disk and halo, respectively(Bensby et al. 2003; 2005).The probability for a star of a given population withrespect to another population is defined as the ratio ofthe Gaussian distribution functions (see Eq. 1) multi-plied by the ratio of the local space densities for thesetwo populations. For each star, the relative probabilitiesfor a certain Galactic population was calculated via thefollowing equations: T D/D = X TD X D f TD f D , T D/H = X TD X H f TD f H (2) X D , X TD and X H are the local space densities for thin-disk, thick-disk and halo, i.e. 0.94, 0.06, and 0.0015,respectively (Robin et al. 1996, Buser et al. 1999,Cabrera-Lavers et al. 2007). Bensby et al. (2003, 2005)proposed four categories to determine the Galactic pop-ulation memberships of the stars and they are as fol-lows: TD/D ≤ . < T D/D ≤ < T D/D ≤
10 for low probability thick-disk stars and
T D/D >
10 for high probability thick-disk stars. Thecomputed
T D/D and
T D/H values for each programstar are given in Table 5. If
T D/D is equal to a smallnumber, the probability of the star in question of beingthick-disk star relative to the thin-disk is relatively low. galpy , a python library developed by Bovy (2015) forGalactic dynamics, was used to calculate the Galac-tic orbital parameters of the HPM stars. We assumed R gc = 8 kpc (Majewski 1993) and Z ⊙ = 27 ± galpy potential MilkywayPo-tential2014 which is composed of three potentials thatmake up the gravitational field of the bulge, disk andhalo components of the Milky Way. The bulge com-ponent is represented as a spherical power law densityprofile by Bovy (2015): ρ ( r ) = A (cid:16) r r (cid:17) α exp " − (cid:18) rr c (cid:19) (3)Here r and r c present reference and the cut-off radius,respectively. A is the amplitude that is applied to the Table 4.
The equatorial coordinates, radial velocities, proper motion components, trigonometric parallaxes, and estimateddistances from the Sun for six HPM stars.
Star α δ RV ± err µαcosδ ± err µδ ± err π ± err d ± err d ± err ∗ Ref.( h : m : s ) ( ◦ : ’ : ” ) (km s −
1) (mas yr −
1) (mas yr −
1) (mas) (pc) (pc) (RV/ µ / π )HD 6755 01 09 43.06 61 32 50.29 -317.86 ± ± ± ± ± ± Gaia
BD+42 3607 20 09 01.41 42 51 54.93 -195.44 ± ± ± ± ± ± Gaia
HD 201891 21 11 59.03 17 43 39.89 -44.15 ± ± ± ± ± ± Gaia
HD 194598 20 26 11.92 09 27 00.43 -246.76 ± ± ± ± ± ± Gaia
HD 3567 00 38 31.95 -08 18 33.40 -46.79 ± ± ± ± ± ± Gaia
HD 84937 09 48 56.10 13 44 39.32 -15.17 ± ± ± ± ± Hipparcos *: Sch¨onrich et al. (2019).
Table 5.
The calculated
U, V, W space velocity components and S total space velocities of six HPM stars along with theirmembership probabilities with different Galactic populations. The Galactic orbital parameters and their errors are also presented.The space velocity components and total space velocities are given according to LSR. Star U ± err V ± err W ± err S ± err Prob. Z max R a R p e p – e v(km s −
1) (km s −
1) (km s −
1) (km s − TD/D TD/H (kpc) (kpc) (kpc)HD 6755 -220.78 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± potential in mass density units and α presents the innerpower. For the the Galactic disk component, we usedthe potential proposed by Miyamoto & Nagai (1975).Φ disk ( R gc , z ) = − GM d r R + (cid:16) a d + p z + b (cid:17) (4)Here z is the vertical distance from the Galactic plane, R gc is the distance from the Galactic centre, M d , is themass of the Galactic disk, G is the Universal gravita-tional constant, a d and b d are the scale length and scaleheight of the disk, respectively. The potential for thehalo component was obtained from Navarro et al. (1996)and it is as follows.Φ halo ( r ) = − GM s R gc ln (cid:18) R gc r s (cid:19) (5)where r s and M s are the radius and mass of the darkmatter halo of the Galaxy.The orbits of the six stars around the Galactic cen-ter were determined with 1 Myr time steps and for a10 Gyr integration time. We used the same input datathat are used in calculations of space velocity compo-nents in estimating orbital parameters. The apo andperi Galactic distances ( R a , R p ), the mean Galactocen-tric distance ( R m = ( R a + R p ) / e p , e v ) and maximum and minimumdistances above the Galactic plane ( Z max , Z min ) wereobtained for each program star. In the calculation of e p and e v eccentricities, e p = ( R a − R p ) / ( R a + R p ) and e v = ( | Z max | + | Z min | ) /R m were used, respectively. Z max values are very close to Z min values because the axisym-metric approach is applied in the solutions of the Galac-tic potentials. The calculated orbital parameters of thestars with galpy code are listed in Table 5 and Galacticorbits of six stars as projected onto X − Y and X − Z planes are shown in Figure 4.In order to study the radial and vertical kinetic ener-gies of the stars, a Toomre diagram was compiled (Fig-ure 5). In the same figure, the six stars are also marked.It is apparent that the total space velocity ( S ) of fivestars in the sample is greater than 250 km s − and is inthe range of 100 < S <
150 km s − for HD 201891. Ac-cording to Nissen (2004), thin-disk stars have total spacevelocities of S <
60 km s − , while the total space veloc-ities of thick-disk stars were reported to show a largerinterval in velocity, i.e. 80 < S <
180 km s − . Thetotal space velocity of the halo stars in the Solar neigh-bourhood is greater than 180 km s − . On the basis ofNissen’s (2004) kinematic criteria, five stars in the sam-ple are members of the halo population and one star is amember of the thick disk population. When consideringthe kinematic criteria of Benbsy et al. (2003)’s, T D/D values of all six stars are greater than 10 (see Table5), indicating that the stars in the sample are membersof the halo population. In the same diagram, one canalso distinguish between disk and halo stars and betweenretrograde and prograde halo stars. It is apparent fromFigure 5 that HD 6755, HD 3567, and HD 194598 haveretrograde motions. This may be indicating that theyoriginated in a satellite (Sakari et al. 2018). In terms oftheir metal abundances, these three stars are apparentlyin the metal-poor tail of the thick disk.0 −60 −40 −20 0 20 40 60X(kpc)−60−40−200204060 Y ( k p c ) −60 −40 −20 0 20 40 60X(kpc)−10−50510 Z ( k p c ) HD 6755 −15 −10 −5 0 5 10 15X(kpc)−15−10−5051015 Y ( k p c ) −15 −10 −5 0 5 10 15X(kpc)−3−2−10123 Z ( k p c ) BD+423607 −10.0 −7.5 −5.0 −2.5 0.0 2.5 5.0 7.5 10.0X(kpc)−10.0−7.5−5.0−2.50.02.55.07.510.0 Y ( k p c ) −10.0 −7.5 −5.0 −2.5 0.0 2.5 5.0 7.5 10.0X(kpc)−1.0−0.50.00.51.0 Z ( k p c ) HD 201891 −7.5 −5.0 −2.5 0.0 2.5 5.0 7.5X(kpc)−7.5−5.0−2.50.02.55.07.5 Y ( k p c ) −7.5 −5.0 −2.5 0.0 2.5 5.0 7.5X(kpc)−0.3−0.2−0.10.00.10.20.3 Z ( k p c ) HD 194598 −10 −5 0 5 10X(kpc)−10−50510 Y ( k p c ) −10 −5 0 5 10X(kpc)−1.5−1.0−0.50.00.51.01.5 Z ( k p c ) HD 3567 −10 −5 0 5 10X(kpc)−10−50510 Y ( k p c ) −10 −5 0 5 10X(kpc)−0.4−0.3−0.2−0.10.00.10.2 Z ( k p c ) HD 84937
Figure 4.
The computed meridional Galactic orbits and projected onto the Galactic X − Y and X − Z planes for six HPMstars. The filled red circles shows the present observed position for each program stars. In this study, Galactic orbital parameters were alsoused to check the population classes of the stars. Forthis, stars in the sample are marked on the Z max × e v plane where their Z max values increase towards the in-creasing vertical eccentricities as seen in Figure 6. Also,from Galactic structure studies (e.g. Karaali et al. 2003;Bilir et al. 2008; G¨u¸ctekin et al. 2019), stars locatedwithin a distance of 2 kpc from the Galactic plane aremembers of the thin disk while stars within 2 and 5 kpc are members of the thick disk. The stars with dis-tances larger than 5 kpc are considered to be membersof the halo population. When stars in the sample wereanalysed on the basis of their spatial distribution in theGalaxy, four stars were found to be members of the thindisk, one star belongs to the thick disk and one staris a member of the halo. As seen in Table 5, the fourstars with e v < .
55 have e p values are quite large, i.e. e p > .
50. Hence the assigned population class for these1
Figure 5.
Toomre diagram computed for six HPM stars.Solid lines (iso-velocity curves) represent constant total spacevelocities S = 100, 200, 300, 400, and 500 km s − . Theuncertainties are also presented; however, they are smallerthan the symbol size for five HPM stars. The dashed verticalline shows the boundary of the objects that make retrograde( V LSR < −
220 km s − ) and prograde ( V LSR ≥ −
220 kms − ) motion. Figure 6. Z max × e v diagram for six HPM stars. The un-certainties on Z max and e v are also indicated. In some cases,they are smaller than the symbol size. four stars as the thin-disk, is surprising. AGES OF THE HPM STARSFor computation of ages in this paper, we used thestellar age probability density function calculated us-ing Bayesian statistics to estimate a precise age withisochrone matching. The probability density functionsto be used for the age computations are provided byJorgensen & Lindegren (2005) and given in the follow-ing form. f ( τ, ζ, m ) ∝ f ( τ, ζ, m ) × L ( τ, ζ, m ) (6)where f presents the initial probability density functiongiven as f = ψ ( τ ) ϕ ( ζ | τ ) ξ ( m | ζ, τ ) (7) In Equation 6, τ , ζ and m are the theoretical modelparameters which represents age, metallicity, and mass,respectively and they are independent from each other.In Equation 7, ψ ( τ ), ϕ ( ζ ) and ξ ( m ) represent the starformation rate, the metallicity distribution and the ini-tial mass function, respectively. Inserting Equation 6 inEquation 7 when integrated, one obtains G ( τ ) functionwhich is the probability density function provided bycomparison of theoretical model parameters and obser-vationally obtained model atmosphere parameters. It iscalculated for large range of ages. In this calculation,the model parameters are compared with the availableisochrones. The maximum of G provides the most likelyage for a given star. The isochrones are from the PAR-SEC stellar evolution models for − . < [Fe / H](dex) < +0 . < τ (Gyr) <
13 with 0.05 dex and 0.1 Gyrsteps (in ζ ), respectively (Bressan et al. 2012). Figure 7shows PARSEC isochrones for six program stars alongwith the Bayesian age estimations. The further detailsof the method for age calculations can be found in ¨OnalTa¸s et al. (2018). The age determination via Bayesianapproach provided the following range in ages for theprogram stars, 9 . < τ (Gyr) < . PARSEC isochrones, we re-port following mass values: 1.06 ± ⊙ for HD 6755;0.801 ± ⊙ for BD +42 3607; 0.982 ± ⊙ forHD 201891; 0.935 ± ⊙ for HD 194598; 1.280 ± ⊙ for HD 3567 and 0.948 ± ⊙ for HD 84937which is a Gaia benchmark star.Sahlholdt et al. (2019) also used Bayesian isochronefitting method, as we did in this study, to derive age esti-mates for the benchmark stars including HD 84937. Thisclosest dwarf star with such a low metallicity has neitherinterferometric nor asteroseismic data that would be ex-tremely useful to provide a precise value for log g . Itis important to remind the reader that although Gaia benchmark stars have been widely analysed in the lit-erature, different authors reported different model at-mosphere parameters. In fact, this was the case for allHPM stars in this study and was also main source ofour motivation to derive their updated model parame-ters. We determined the age of HD 84937 as 9.86 ± +1 . − . Gyr for the star. Ear-lier studies (e.g. Casagrande et al. 2011; VandenBerg etal. 2014) reported somewhat higher values of age suchthat they were comparable to the age of the Universe(13.7 Gyr by WMAP; Bennett et al. 2013). For in-2 eff (Kelvin)012345 l o gg ( c m s − ) HD 84937 log(τ) = 9.86 0.17(Gyr) 300040005000600070008000 T eff (Kelvin)012345 l o gg ( c m s − ) HD 3567 log(τ) = 9.46 0.15(Gyr) 300040005000600070008000 T eff (Kelvin)012345 l o gg ( c m s − ) HD 194598 log(τ) = 10.00 0.09(Gyr)300040005000600070008000 T eff (Kelvin)012345 l o gg ( c m s − ) HD 201891 log(τ) = 9.97 0.11(Gyr) 300040005000600070008000 T eff (Kelvin)012345 l o gg ( c m s − ) BD+42 3607 log(τ) = 10.08 0.01(Gyr) 300040005000600070008000 T eff (Kelvin)012345 l o gg ( c m s − ) HD 6755 log(τ) = 9.81 0.29(Gyr)
Figure 7.
Hertzsprung-Russell diagram for the six HPM stars, plotted with PARSEC isochrones. LTE parameters are shownas cross symbols. The errors in computed ages are also indicated. stance, VandenBerg et al. (2014) reported 12.09 ± Gyr of age for the star. It is interesting to note that theages reported by Gehren et al. (2006) and for instance,by Ibukiyama & Arimoto (2002) was even larger than13.7 Gyr. Sahlholdt et al. (2019) reported an age of13.5 Gyr for HD 84937 and provided 11.0 - 13.5 Gyr asa range of age for the star. In this context, it should bementioned that the isochrones employed in Sahlholdt etal. (2019) and in this study were created with the sameresolutions in age and in metallicity. As a health check,we repeated the calculation by Sahlholdt et al. (2019)by using the model atmosphere parameters that they re-ported for the star. We confirmed their computed agefor HD 84937. However, by using up-to-date model pa-rameters obtained in this study via spectroscopy, thestar seemed to be ≈ PARSEC -isochroneswhen fitting to the magnitude or the log g (see also theirFigure A1) and the alternative grids of isochrones that The error in age is derived from the T eff error bar. they preferred provided somewhat larger values for age.In Figure 8, we provide age estimates for all HPM starsin this work and from the literature.Luck (2017) reported masses and ages for HD 201891and HD 194598. The reported mass and age by Luck(2017) for HD 201891 from BaSTI isochrones are 0.86M ⊙ and 9.50 Gyr, respectively. We reported a massof 0.98 ± ⊙ for the star from the PARSEC isochrones. Our determination of age for HD 201891from the
PARSEC isochrones is 9.97 ± ± ⊙ and 10.0 ± PARSEC isochrones. Luck’s (2017) reported age forthe star from the BaSTI isochrones is 1.5 Gyr younger.An alternative isochrones from Bertelli et al. (1994) byLuck (2017) provided 10.73 Gyr of age for HD 194598.To validate Luck’s (2017) result, we also computed massand age of the star using Bertelli’s isochrones but withthe model parameters determined in this study. The re- HD201891 HD194598 HD 6755 BD+423607 HD 3567 HD84937246810121416 A g e ( G y r )
25 12428272629,3023 120211922 11718 1 115101216891113 14 12346 57
1: This work;2: Bensby et al. (2014), 10.2 Gyr and 2.5 Gyr of age as uncertainty;3: VandenBerg et al. (2014) and 12.09±0.63 Gyr of age;4: Sahlholdt et al. (2018), 13.5 Gyr;5: Schonrich & Bergemann (2018), 13.57±0.41 Gyr;6: Ibukiyama Arimato (2002), 14.36 Gyr;7: Gehren et al. (2006), 15.5 Gyr; 8: Ramirez et al. (2013), 6.71 Gyr;9: Lambert (2004), 7.30 Gyr;10: Ge et al. (2016), 8.73±0.83 Gyr;11: Feltzing (2001), 9.34±3.63 Gyr;12: Ge et al. (2016), 9.50±0.50 Gyr;13: Bensby et al. (2014), 10.2 Gyr;14: Ramirez (2012), 10.16±3.01 Gyr; 14: Aquilera-Gomez (2018), 10.16±2.91 Gyr;15: Schuster et al. (1989), 12.0±0.7 Gyr;16: Chen (2001), 13.7 Gyr;17: Feltzing (2001), 2.3±0.7 Gyr;18: Mints (2017), 6.78±0.22 Gyr;19: Ibukiyama Arimato (2002), 7.66 Gyr; 20: Luck (2017), 8.5 Gyr;21: Bertelli et al. (1994), 10.73 Gyr;22: Bensby et al. (2014), 12.4 Gyr;23: Isaacson (2010), 3.37 Gyr;24: Ibukiyama Arimato (2002), 8.03 Gyr;25: Luck (2017), 9.50 Gyr; 26: Mints (2017), 11.12±0.12 Gyr;27: Ramirez (2013), 12.69 Gyr;28: Pace (2013), 12.85±0.90 Gyr;29: Bensby et al. (2014), 13.1 Gyr;30: Ramirez (2012),13.16±2.29 Gyr;30: Aquilera-Gomez (2018), 13.16±2.26 Gyr
Figure 8.
Ages estimated in this work based on Bayesian isochrone fitting to the PARSEC models with the ages collected fromthe literature. The uncertainties on ages are also presented for all available age measurements in the literature; however, theyare smaller than the symbol size in some cases (e.g. 0.01 Gyr for BD+42 3607). The horizontal dashed line indicates the age ofthe Universe of 13.7 Gyr as determined by WMAP (Bennett et al. 2013). sults are 0.82 +0 . − . M ⊙ and 8.46 +0 . − . Gyr, respectively.When Luck’s (2017) model atmosphere parameters used(agreed with ours within error limits) for the star, weobtained a mass of 0.86 M ⊙ which is in excellent agree-ment with Luck’s (2007) reported mass from Bertelli’sisochrones (1994) for HD 194598. RESULTS AND DISCUSSIONIn addition to kinematics and Galactic orbits, sincewe also aim to spur discussion on the origin of the HPMstars under scrutiny in the context of abundances, we de-rived up-to-date photospheric abundances of 29 speciesincluding not only α -elements, but also slow (s)- andrapid (r)-process elements from Y to Sm. The final el-emental abundances log ǫ (X) averaged over the sets of4measured lines for the stars are listed in Tables A7 andA8, where the first column shows species, the second- logarithmic elemental abundances and the third - el-ement over iron ratios . The number of lines used inthis analysis are also given. The last columns in thosetables present computed solar abundances by us in thisstudy. The errors reported in logarithmic abundancespresent 1 σ line-to-line scatter in abundances. The errorin [X/Fe] is the square root of the sum of the quadratureof the errors in [X/H] and [Fe/H]. The formal errors forthe abundances arising from uncertainties of the atmo-spheric parameters T eff , log g and ξ are summarized inTable A10 for changes with respect to the model. Fora further discussion on chemical abundances includingup-to-date photospheric abundances of 29 species in thespectra of HPM stars, the reader is referred to the ap-pendix section.Given an understanding of the chemical abundancesand kinematics of the HPM stars presented in this study,it is worthwhile to consider the origin of these stars.More specifically, how have they been accelerated tosuch high velocities? Tidal interactions in a globularcluster (GC) or Galaxy interactions could be a possiblesource for HPM stars. In order for stars to reach thehigh velocities observed in this study, it is likely theyhave been part of a complex three or more body inter-action. The dense cores of GCs represent one of themost likely sites for such interactions, as they typicallyhave both a relatively high binary fractions (Milone etal. 2012) and encounter rates (Leigh & Sills 2011). Infact, searching for a dynamic connection between HPMstars and Galactic GCs is now possible in the Gaia era,as the proper motions and parallaxes of most GCs andnearby dwarf spheroidal galaxies (dSph) are now known(Helmi et al. 2018, Vasiliev et al. 2019).The ages and metalicities of the HPM stars suggestthey are all old and metal-poor, also consistent with oncebeing members of a GC population. While its entirelypossible the stars were ejected from a cluster that hassince reached dissolution, it is interesting to note thatthe metallicity of the six HPM are comparable to sev-eral Galactic GC within uncertainty. Harris (1996; 2010edition) reported distances, velocities, metallicities, lu-minosities as well as dynamical parameters for 157 glob-ular clusters in the Milky Way. The up-to-date orbitalparameters reported for the common GCs from Helmi etal. (2018) with the help of these metallicities from Har-ris (1996) were used to estimate origin of six HPM stars [X/Fe]=[log ǫ (X)-log ǫ (Fe)] star - [log ǫ (X)-log ǫ (Fe)] ⊙ For GCs, the modest (photometric) binary fractions rangefrom 1-10%. in this study. For this aim, we compiled the Galacticorbital parameters (i.e. R a , R p , e p ), the metallicities,ages of each HPM star and cluster and listed them inTable 6. In this table, we only included GC candidatesthat have similar Galactic orbital parameters, metallic-ities and ages compared to those obtained in this studyfor the program stars. As it can be inspected from Table6, even for such a crude comparison, the similarities inGalactic orbital parameters, metallicities and ages areintriguing. In Table 6, the best matching GCs are in-dicated in bold-type face for clarity. This table alsoprovides encounter probabilities that happens to pro-vide an alternative diagnostic when it comes to testingdynamical origin of the program stars.The ages of the GCs listed in Table 6 were compiledfrom seven different studies (Massari et al. 2016, Van-denberg et al. 2013, Cezario et al. 2013, Koleva etal. 2008, de Angeli et al. 2005, Catelan et al. 2002,and Salaris & Weiss 1998). For these studies, the mosthomogeneous study of the age of the GCs belongs toVandenberg et al. (2013). VandenBerg et al. (2013) de-rived the ages for 55 GCs in the Milky Way and nicelypresented their variations with the metallicity. Theircomputations of mean ages for the GCs as derived bythe grouping of [Fe/H] < -1.7 dex and [Fe/H] > -1 dexprovided respective ages of 12.5 and 11 Gyr. They es-timated errors in age to be 0.25 Gyr, mainly caused byapplications of the isochrones to observed data. Also,the errors due to uncertainties in distances and clustermetal abundances ranged from 1.5 to 2 Gyr. The meanages of the globular clusters with the reported errors intheir ages by VandenBerg et al. (2013) show agreementwith the mean ages (10 Gyr) and their errors obtainedin this study for the program stars.In order to test the dynamical origin of the programstars, we followed the same procedure as in Pereira etal. (2017). The orbital evolution of the HPM stars weresimulated over 12 Gyr into the past with the orbits of16 candidate GCs from the literature. As mentioned inSection 4, MilkywayPotential2014 (Bovy 2015) was usedto calculate the Galactic orbits of stars and globularclusters. The initially determined orbital parameters ofthe HPM stars and the GCs are randomly varied. Thesimulations for each program star were repeated 5000times considering different orbital parameters for thestars and the clusters within their observational uncer-tainties. During these orbital simulations, errors in theirproper motions, heliocentric distances, and radial veloc-ity in the equatorial coordinates are considered for HPMstars (see Table 4). The parameters and errors used inthe calculation of the orbital parameters of the GCs arealso taken from Helmi et al. (2018). For each simula-5
NaI MgI SiI CaI ScII TiI TiII VI CrI MnI CoI NiI CuI YII BaII SmII−1.2−0.8−0.40.00.40.8 [ X / F e ] NGC5897HD201891NaI MgI SiI CaI ScII TiI TiII VI CrI MnI CoI NiI CuI YII ZrII BaII SmII−1.2−0.8−0.40.00.40.8 [ X / F e ] NGC4833HD194598NaI MgI SiI CaI ScII TiI TiII VI CrI MnI CoI NiI CuI YII ZrII BaII−1.2−0.8−0.40.00.40.8 [ X / F e ] NGC0362HD194598NaI MgI SiI CaI ScII TiI TiII VI CrI MnI CoI NiI YII ZrII BaII−1.2−0.8−0.40.00.40.8 [ X / F e ] NGC0362HD3567
Figure 9.
Abundances of NGC 5897, NGC 4833, and NGC 362 along with the abundances of HD 201891, HD 194598, andHD 3567. tion, the probability for a close encounter between starand the GC was calculated at a certain distance. For aclose encounter, the distance to any cluster was assumedto be smaller than 5 tidal radii. The tidal radii for theGCs were obtained from Moreno, Pichardo Velazquez(2014). Table 6 summarizes the encounter probabilitiesfor each HPM star and a candidate GC for distancesthat corresponds to 5 and 1.5 tidal radii. Moreover,we also computed the average encounter velocities forHPM program stars and they are also reported in Table6. Orbital simulations have shown that HPM stars havea 5% to 30% for 5 tidal radii and 4% to 18% probabil-ity of encountering with selected GCs for 1.5 tidal radii.The larger values of the listed encounter probabilitiesare likely to be related to a GC origin.To further constrain the list of possible GC progen-itors, we also compiled abundances of the GC candi-dates to check whether they also show similar abun-dances. Figure 9 (also Table A11) presents abundancesof HPM star/GC candidate. The discrepant abundances are shown in bold type-face in Table A11. There is a fairagreement for several elements between abundances ofHPM stars and assigned GC candidates. Such compari-son may also provide valuable information on identifyingelements and/or certain lines to trace a GC progenitorfor an HPM star.Further on the Galactic population classification ofthe HPM stars, in Figure 10, by considering the abun-dances determined in this study for magnesium, silicon,calcium, and titanium, we also provided [ α /Fe] for eachof the HPM stars for which the mean values of fourelements were used as estimates of [ α /Fe] (see bottompanel in Figure 10). The [ α /Fe] values for a representa-tive sample of the Galactic thick disk and halo stars fromBensby et al. (2014) as well as halo and disk stars (in-cluding metal-poor dwarf stars) from Fulbright (2000)are shown for comparison of α -element abundances. Itis apparent from Figure 10 that all HPM stars are richin α -elements. The over all agreement is satisfactoryin silicon, calcium, and titanium abundances. Between6 −3.0 −2.5 22.0 21.5 21.0 20.5 0.020.420.20.00.20.40.60.8 [ M g / F e ] ( d e ) This s−.d1F.lbrigh− (2000) Be(sb1 a− al. (2014) Hal)Be(sb1 e− al. (2014) Thi k Disk Be(sb1 a− al. (2014, avg.) HaloBensby at al. (2014, avg.) Thick Disk Fulbright (2000, avg.) −2.5 22.0 21.5 21.0 20.5 0.020.20.00.20.40.60.81.01.21.4 [ S i / F e ] ( d e )
1: HD 194598; 2: HD 6755; 3: HD 3567; 4: HD 84937; 5: BD+42 3607; 6:HD 201891 [ C a / F e ] ( d e ) [ T i / F e ] ( d e ) [Fe/H] (de0) [ A l p h a / F e ] ( d e ) This s−.d1 (a/g.)F.lb+igh− (2000, avg.) Bensby et al. (2014, avg.) HaloBensby et al.(2014, avg.) Thick Disk Bensby at al. (2014, avg.) HaloBensby at al. (2014, avg.) Thick disk Fulbright (2000, avg.)
Figure 10.
Elemental abundances for the four α -elements relative to Fe. A representative sample of stars from Bensby etal. (2014) for the Galactic thick disk and halo are also marked by small circles and downward triangles, respectively, and theabundances of the individual elements for the metal poor dwarfs from Fulbright (2000) and the HPM stars in this study bycross and diamond symbols, respectively. Individual error bars in [X/Fe] axis for the HPM stars are also indicated. The largeblack circles and large downward triangles are average abundances in 0.25 dex intervals of [Fe/H] for thick disk and halo stars,respectively. The large squares show the mean values for the metal poor dwarfs from Fulbright (2000) as indicated in the legendat the bottom. The bottom panel shows the mean values of the four elements for both sample stars and the HPM stars. Table 6.
Parent globular cluster candidates for the program stars. The Galactic orbital parameters, metallicities, ages as wellas their errors, the encountering probabilities ( P (%)) and the average encounter velocities for the program stars for 5 and 1.5tidal radii are also presented. The uncertainties available from the literature for the parent GCs candidates were also included(e.g. Helmi et al. 2018). For some, no uncertainty was reported. R apo R peri e p [Fe/H] Ages GC P (%) . V enc Star Star GC Star GC Star GC Star GC ∗ Star GC Ref. P (%) . (kpc) (kpc) (dex) (Gyr) (km s − )HD 6755 58.07 +3 . − . +0 . − . +0 . − . -1.39 -2.31 NGC 5466 15.92 11 ± (3.15) (0.01) (0.06) (0.18) (0.09) (0.29) (0.25) 5.4517.82 +0 . − . +0 . − . +0 . − . -1.96 10.32 NGC 2298 5.60 76 ± BD+423607 16.15 14.29 +0 . − . +0 . − . +0 . − . -2.43 -1.18 10.08 10.90 3 NGC 2808 5.55 110 ± +0 . − . +0 . − . +0 . − . -1.29 9.98 4 NGC 6864 2.26 167 ± +0 . − . +0 . − . +0 . − . -2.19 12.54 5 NGC 4372 14.68 12 ± +0 . − . +0 . − . +0 . − . -1.90 10.10 NGC 5897 31.88 24 ± HD 201891 9.16 7.55 +0 . − . +0 . − . +0 . − . -1.02 -0.35 9.97 11.35 7 NGC 6356 24.44 95 ± +0 . − . +0 . − . +0 . − . -1.70 12.50 NGC 6656 17.27 89 ± +0 . − . +0 . − . +0 . − . -2.33 6.98 2 NGC 7078 22.85 23 ± +0 . − . +0 . − . +0 . − . -1.31 10.21 NGC 6284 23.13 16 ± +0 . − . +0 . − . +0 . − . -1.29 10.56 2 NGC 5946 0.00 151 ± +0 . − . +0 . − . +0 . − . -1.17 -1.30 10.00 10.75 1 NGC 362 15.18 59 ± +0 . − . +0 . − . +0 . − . -1.89 12.50 1 NGC 4833 2.00 172 ± +1 . − . +0 . − . +0 . − . -1.90 10.10 6 NGC 5897 18.85 83 ± +0 . − . +0 . − . +0 . − . -1.30 10.75 NGC 362 7.56 171 ± HD 3567 10.78 10.61 +0 . − . +0 . − . +0 . − . -1.10 -2.35 9.46 12.75 1 NGC 6341 5.55 223 ± +0 . − . +0 . − . +0 . − . -2.00 12.75 1 NGC 6779 5.66 103 ± +0 . − . +0 . − . +0 . − . -2.35 12.75 NGC 6341 12.19 151 ± +0 . − . +0 . − . +0 . − . -1.32 11.50 1 NGC 288 14.83 89 ± +0 . − . +0 . − . +0 . − . -2.40 -1.30 9.86 10.75 1 NGC 362 3.24 265 ± +0 . − . +0 . − . +0 . − . -2.00 12.75 1 NGC 6779 12.14 96 ± α -elements, Ca is generally considered to betrue representative of the α -elements in the literaturesince O, Si, and Mg are often thought to be altered dueto the recycling of the products of an earlier generationof stars during subsequent star formation inside GCs(Kraft et al. 1997, Gratton et al. 2004, Carretta etal. 2010, Gratton et al. 2012). We also measured theoxygen abundance in the spectra of the six stars by us-ing the triplet oxygen lines at 6156 ˚A. To be more spe-cific, we employed O i line at 6156 ˚A for HD 194598 andHD 84937; the OI line at 6158 ˚A for HD 6755; the OIlines at 6156 ˚A and 6158 ˚A for HD 3567, however, wedid not include the oxygen lines in the current analysisbecause of the fact that they were weak and contributedby Ca and Fe in the wings of those lines although thecontribution of Ca and Fe were removed with success viaspectrum synthesis technique. The triplet lines were notobservable in the spectrum of BD+42 3607. The oxy-gen abundances were used to compile [Na/Fe] vs [O/Fe]plots for which the summary of the findings for some se-lected program stars are as follows: HD 194598 followsthe same abundance pattern as the stars of NGC 362,where anti-correlation is observed with [Na/Fe]: it in-creases as the [O/Fe] ratio decreases. Similar case wasalso seen for HD 194598 and NGC 4833. HD 201891 isalso seen to show similar abundance pattern as the starsof NGC 5897. HD 6755:
On the basis of its Galactic orbital param-eters, HD 6755 is classified as a halo star. The positionof the star in the Z max × e v plane is also agreed withthe membership status of the star as a halo star (i.e. Z max = 13.22 ± P ≈
16% for 5 tidal radii was obtained. Theencounter probability for the star for 1.5 tidal radii wasfound to be P ≈ BD+42 3607:
As one of the most metal-poor HPMstar in our sample with HD 84937, its Galactic orbitalparameters (and metallicity) indicate a halo membershiplike HD 6755 and three GCs are assigned as parent GCcandidates, i.e. NGC 2298, NGC 2808, and NGC 6864.The metallicity and age of the star clearly support theassumption of a GC origin, but the encounter probabil-ity for NGC 2298 is found to be relatively low ( P = 5%for 1.5 tidal radii). We exclude from consideration the latter two GCs that show ≈ HD 201891:
The calculated Galactic orbital param-eters of the star provided five matches as for GC can-didates and indicate a thick disk membership for thestar like HD 3567. The position of the star in theToomre diagram also corroborates the assessment as thethick disk (Figure 5). The best matching GC candiatesbetween among those listed in Table 6 are NGC 5897and NGC 6656 and they have relatively higher meanmetallicities compared to the metallicity of the star (i.e.[Fe / H] = − . ± .
16 dex). Although an excellentmatch was otained in orbital parameters, metallicity,and age for NGC 5897, even preferred candidate pol-luters, e.g. fast rotating massive stars or the enrichmentoccurred through equatorial disks of stars between 20M ⊙ and 100-120 M ⊙ (Decressin et al. 2007), or mas-sive AGB stars (Ventura et al. 2001) through slow windsfollowing the hot-bottom burning phase, may not pro-vide the stipulated enrichment ( ≈ P ≈
18% for 1.5 tidal radii.More onto this, our evaluation for the star’s origin froma parent GC candidate solely based on star’s elementabundances from the high-resolution spectroscopy per-formed in this study proved to be inconclusive althoughabundances for HD 201891 and NGC 5897 agreed wellwith the exception of V i , Cr I , and Cu I abundances (seeFigure 9). For the second best matching GC candidate,NGC 6656, the age of the cluster from Vandenberg et al.(2013) is far from being agreed with the star’s reportedage in this study. Therefore, assigned GC candidatesfor the star in Table 6 can be excepted for a GC originscenario. HD 194598:
The position of the star in the Toomrediagram (Figure 5) implies a retrograde motion likeHD 6755 and HD 3567 hence an alternative origin viatidal distruption of a dwarf galaxy can not be ruled out.Three best matching GC candidates for the star havemetallicities that are in accordance with the star’s metal-licity, NGC 6284, NGC 5946, and NGC 362, but thehighest encounter probability was obtained for the for-mer cluster ( P ≈ ≈ σ level in Galactic orbital parameters, metallicity, and agefor the star are contemplated. However, it is importantto note that the metallicity and Z max (Figure 6) of thestar are not in accordance with the star’s membershipstatus as a thin-disk star suggested by stars Galactic or-bital parameters. This with its retrograde motion, wecan not rule out the possibility that the star may haveoriginated from a tidally disrupted dwarf galaxy. HD 3567:
When the star’s position in Z max × e v planewas inspected, HD 3567 was seen to be located in a re-gion where thin-disk star are expected to reside. Thisfinding clearly contradicts with the assigned status forthe star as a thick disk star from its metallicity and age.Although the match in its Galactic orbital parameterswith three parent GC candidates listed in Table 6 is sat-isfactory, the best matching GC candidate for the star onthe basis of its metallicity and age is seen to be NGC 362.Figure 9 presents abundances of the star along with theabundances of the cluster. The abundances of severalspecies (e.g. Na i , Mg i , Si i , Ca i , Ti iii , Cr i , Co i , Ni i ,Y ii , Zr ii , and Ba ii ) are agreed well with those of thecluster. On the other hand, for the abundances of Sc ii ,V i , and Mn i the discrepancies up to ≈ HD 84937:
The Galactic orbital parameters of this
Gaia benchmark star suggest a thin-disk membershipstatus for the star. This is also in accordance with theposition of the star in the Z max × e v plane where thestar has Z max = 0 . ± .
25 kpc. However, on the ba-sis of its metallicity and age, the star is a member ofhalo. The best matching GC candidate for the star,NGC 6341, has a metallicity that is in accordance withthe star’s metallicity. Also, within the error limits theagreement in orbital parameters is satisfactory. On thebasis of its orbital parameters, metallicity, age, and thethe low encounter probability calculated for the star, thepossibility for a GC origin is the less probable case.To conclude, their orbital parameters, metallicities,ages, and also computed encounter probabilities allowed us to determine the origin of the HPM stars, e.g. prob-able candidate GCs as parent clusters. Tidal disruptionfrom a dwarf galaxy was also considered as an alterna-tive origin. However, this should not be evaluated asa strict assignment. HD 6755, HD 194598, and HD 3567with their retrograde orbital motions are likely candi-date stars for a dwarf galaxy origin. Despite the factthat a small encounter probability ( P ≈ P ≈ Gaia
Gaia
Gaia
Multilateral Agreement.
Software:
LIME (¸Sahin 2017), SPECTRE (Sneden1973), MOOG (Sneden 1973), INSPECT program v1.0(Lind et al. 2012), galpy (Bovy 2015), MilkywayPoten-tial2014 (Bovy 2015).REFERENCES
Adibekyan V. Zh., Santos N. C., Sousa S. G. et al., 2012,A&A, 543, A89Aguilera-G´omez C., Ram´ırez I., Chanam´e J., 2018, A&A,614, A55Allende Prieto C., Majewski S. R., Schiavon R., C. et al.,2008, AN, 329, 1018Allen D. M., Porto de Mello G. F., 2011, A&A, 525, A63Asplund M., Grevesse N., Sauval A. J., Scott P., 2009,ARA&A, 47, 481Balbinot E., Gieles M., 2018, MNRAS, 474, 2479Bastian N., Lardo C., 2018, ARA&A, 56, 83Battisini C., Bensby T., 2015, A&A, 577, A9Belokurov V. et al., 2007, ApJ, 657, 89Bennett C. L. et al., 2013, ApJS, 208, 20Bensby T., Feltzing S., Lundstr¨om I., 2003, A&A, 410, 527Bensby T., Feltzing S., Lundstr¨om I., Ilyin I., 2005, A&A,433, 185Bensby T., Feltzing S., Oey M. S., 2014, A&A, 562, A71Bergemann M., Gehren T., 2008, A&A 492, 823–831Bergemann M., Pickering J. C., Gehren T., 2010, MNRAS,401, 1334Bergemann M., Lind K., Collet R., Magic Z., Asplund M.,2012, MNRAS, 427, 27Bergemann M., Collet R., Sch¨onrich R., Andrae R.,Kovalev M., Ruchti G., Hansen C. J., Magic Z., 2017,ApJ, 847, 16Bergemann M. et al., 2019, A&A, 631, A80Bertelli G., Bressan A., Chiosi C., Fagotto F., Nasi E.,1994, A&AS, 106, 275Bertin G., Varri A. L., 2008, ApJ, 689, 1005Bilir S., Karaali S., Gilmore G., 2006, MNRAS, 366, 1295Bilir S., Cabrera-Lavers A., Karaali S., Ak S., Yaz E.,L´opez-Corredoira M., 2008, PASA, 25, 69Bilir S., Karaali S., Ak, S. et al., 2012, MNRAS, 421, 3362Binney, J., Tremaine, S. 1987, Galactic Dynamics(Princeton, NJ: Princeton Univ. Press)Blackwell-Whitehead R., Bergemann M., 2007, A&A, 472,L43Boeche C., Grebel E. K., 2016, A&A, 587, A2Boesgard A. M., Rich J. A., Levesque E. M., Bowler B. P.,2011, ApJ, 743, 140Bovy J., 2015, ApJS, 216, 29 Bressan A., Marigo P., Girardi L., Salasnich B., Dal CeroC., Rubele S., Nanni A., 2012, MNRAS, 427, 127Burris D. L., Pilachowski C. A., Armandroff T. E., SnedenC., Cowan J. J., Roe H., 2000, ApJ, 544, 302Buser R., Rong J., Karaali S., 1999, A&A, 348, 98Cabrera-Lavers A., Bilir S., Ak S., Yaz E.,L´opez-Corredoira M., 2007, A&A, 464, 565Carney B. W., Wright J. S., Sneden C., Laird J. B., AguilarL. A., Latham D. W., 1997, AJ, 114, 363Carretta E., Gratton R. G., Clementini G., Pecci F. F.,2000, ApJ, 533, 215Carretta E., Bragaglia A., Gratton R. et al., 2009, A&A,508, 695Casagrande L., Sch¨onrich R., Asplund M., Cassisi S.,Ramirez I., Melendez J., Bensby T., Feltzing S., 2011,A&A, 530, A138Castelli F., Kurucz R. L., 2004, preprint(arXiv:astro-ph/0405087)Catelan M., Borissova J., Ferraro F. R., Corwin T. M.,Smith H. A., Kurtev R., 2002, AJ, 124, 364Cayrel R. Depagne E., Spite M. et al., 2004, A&A, 416,1117Cezario E., Coelho P. R. T., Alves-Brito A., Forbes D. A.,Brodie J. P., 2013, A&A, 549, A60Chen Y. Q., Nissen P. E., Zhao G., Zhang H. W., Benoni T.2000, A&AS, 141, 491Chen H. W., Lanzetta K. M., Webb J. K., Barcons X.,2001, ApJ, 559, 654Colucci J. E., Bernstein, R. A., McWilliam, A., 2017, ApJ,834, 105Co¸skuno˘glu B. et al., 2011, MNRAS, 412, 1237De Angeli F. et al., 2005, AJ, 130, 116Decressin, T., Meynet, G., Charbonnel, C., et al. 2007,A&A, 464, 1029de la Fuente Marcos R., de la Fuente Marcos C., 2019,A&A, 627, 104De Marchi G., Pulone L., 2007, A&A, 467, 107De Silva G. M. et al., 2015, MNRAS, 449, 2604Den Hartog E. A., Lawler J. E., Sneden C., Cowan J. J.,2003, ApJS, 148, 543Den Hartog E. A., Lawler J. E., Sobeck J. S., Sneden C.,Cowan J. J., 2011, ApJS, 194, 35 Dobrovolskas V., Kucinskas A., Steffen M., Ludwig H. G.,Prakapavicius D., Klevas J., Caffau E., Bonifacio P.,2013, A&A, 559, A102Feltzing S., Holmberg J., Hurley J. R., 2001, A&A, 911, 924Fishlock C. K., Yong D., Karakas A. I. et al., 2017,MNRAS, 466, 4672Fulbright J. P., 2000, AJ, 120, 1841Fuhr J. R., Wiese W. L., 2006, J. Phys. Chem. Ref. Data,35, 4Fuhrmann K., Pfeiffer M., Frank C., Reetz J., Gehren T.,1997, A&A, 323, 909Fuhrmann K., 1998, A&A, 338, 161Gaia Collaboration, Katz D. et al., 2018, A&A, 616, 11Ge Z. S., Bi S. L., Chen Y. Q., Li T. D., Zhao J. K., Liu K.,Ferguson J. W., Wu Y. Q., 2016, ApJ, 833, 161Gehren T., Shi, J. R., Zhang H. W., Zhao G., Korn A. J.,2006, A&A, 451, 1065Gilmore et al., 2012, The Messenger, 147, 25Gratton R. G., Carretta E., Claudi R., Lucatello S.,Barbieri M., 2003, A&A, 404, 187Gratton, R., Sneden, C., & Carretta, E. 2004, ARA&A, 42,385Gratton, R. G., Carretta, E., & Bragaglia, A. 2012,A&A Rv, 20, 50Grillmair C. J., 2006, ApJ, 651, 29G¨u¸ctekin S., Bilir S., Karaali S., Plevne O., Ak S., 2019,AdSpR, 63, 1360Hannaford P., Lowe R. M., Grevesse N., Bi´emont E., 1982,ApJ, 261, 736Hansen C. J., Primas F., Hartman H. et al., 2012, A&A,545, A31Harris W. E, 1996, AJ, 112, 1487Hawkins K., Jofre P., Heiter U. et al., 2016, A&A, 592, A70Heijmans J., Asplund M., Barden S. et al., 2012,Proceedings of the SPIE, 8446, 84460WHeiter U., Jofre P., Gustafsson B., Korn A. J., Soubiran C.,Thevenin F., 2015, A&A, 582, A49Helmi A. et al., 2017, A&A, 598, 58Helmi A., van Leeuwen F. et al. 2018, A&A, 616, A12Ibukiyama A., Arimoto N., 2002, A&A, 394, 927Isaacson H., Fischer D., 2010, ApJ, 725, 875Ishigaki M. N., Chiba M., Aoki W., 2012, ApJ, 753, 64Ishigaki M. N., Aoki W., Chiba M., 2013, ApJ, 771, 67Israelian G., L´opez R. J. G., Rebolo R., 1998, ApJ, 507, 805Jehin E., Magain P., Neuforge C., Noels A., Parmentier G.,Thoul A. A., 1999, A&A, 341, 241Jofre P., Heiter U., Soubiran C. et al., 2014, A&A, 564,A133Jorgensen B. R., Lindegren L., 2005, A&A, 436, 127Johnson D. R. H., Soderblom D. R. 1987, AJ, 93, 864 Karaali S., Ak S. G., Bilir S., Karatas , Y., Gilmore G., 2003,MNRAS, 343, 1013Karaali S., Yaz G¨ok¸ce E., Bilir S., 2016, Ap&SS, 361, 354Karaali S., Bilir S., G¨ok¸ce E. Yaz, Plevne O., 2019, PASA,36, 40Kelleher D. E., Podobedova L. I., 2008, Journal of Physicaland Chemical Reference Data, 37, 267Kelleher D. E., Podobedova L. I., 2008, Journal of Physicaland Chemical Reference Data, 37, 1285Kepley A. A., Morrison H. L., Helmi A. et al., 2007, AJ,134, 1579Klement R. J., 2010, A&ARv, 18, 567Klose J. Z., Fuhr J. R., Wiese W. L., 2002, Journal ofPhysical and Chemical Reference Data, 31, 217Koch A., McWilliam A., 2014, A&A,565, A23Koleva M., Prugniel P., Ocvirk P., Le Borgne D., SoubiranC., 2008, MNRAS, 385, 1998Kraft, R. P., Sneden, C., Smith, G. H., et al. 1997, AJ, 113,279Kurucz R. L., Furenlid I., Brault J., Testerman L., 1984, inKurucz R. L., Furenlid I. Brault J. Testerman L., eds,Solar Flux Atlas from 296 to 1300 nm. National SolarObservatory, Sunspot, NMKruijssen J. M. D., Mieske S., 2009, A&A, 500, 785Lai D. K., Bolte M., Johnson J. A., Lucatello S., Heger A.,Woosley S. E., 2008, ApJ, 681, 1524Lambert D. L., Reddy B. E., 2004, MNRAS, 349, 757Larsen S. S., Brodie J. P., Strader J., 2017, A&A, 601, A96Lawler J. E., Den Hartog E. A., Sneden C., Cowan J. J.,2006, ApJS, 162, 227Lawler J. E., Sneden C., Cowan J. J., Ivans I. I., DenHartog E. A., 2009, ApJS, 182, 51Lawler J. E., Guzman A., Wood M. P., Sneden C., CowanJ. J., 2013, ApJS, 205, 11Lawler J. E., Wood M. P., Den Hartog E. A., Feigenson T.,Sneden C., Cowan J. J., 2014, ApJS, 215, 20Lawler J. E., Sneden C., Cowan J. J., 2015, ApJS, 220, 13Leigh N. W. C., Sills A., 2011, MNRAS, 410, 2370Leigh N. W. C., Stone N. C., Geller A. M., Shara M. M.,Muddu H., Solano–Oropeza D., Thomas Y., 2016,MNRAS, 463, 3311Ljung G., Nilsson H., Asplund M., Johansson S., 2006,A&A, 456, 1181Li H., Du C., Liu S., Donlon T., Newberg H. J., 2019, ApJ,874, 74Lind K., Bergemann M., Asplund M., 2012, MNRAS, 427,50Luck R. E., 2017, AJ, 153, 21Majewski S. R., 1993, ARA&A, 31, 575Mashonkina L., Gehren T., 2000, A&A, 364, 249 Mashonkina L., Zhao G., Gehren T. et al., 2008, A&A, 478,529Massari D., et al., 2016, A&A, 586, A51McLaughlin D. E., van der Marel R. P., 2005, ApJS, 161,304Melendez J., Casagrande L., Ramirez I., Asplund M.,Schuster W. J., 2010, A&A, 515, L3Mihalas D., Binney J. 1981, Galactic Astronomy: Structureand Kinematics (2nd ed.; San Francisco, CA: Freeman),608Milone A. P., Piotto G., Bedin L. R. et al., 2012, A&A,540, A16Mints A., Hekker S., 2017, A&A, 604, 108Mishenina T. V., Kovtyukh V. V., 2001, A&A, 370, 951Mishenina T. V., Kovtyukh V. V., Korotin S. A., SoubiranC., 2003, Azh, 80, 5Mishenina T. V., Soubiran C., Kovtyukh V. V., KorotinS. A., 2004, A&A, 418, 551Mishenina T. V., Gorbaneva, T. I., Basak N. Yu., SoubiranC., Kovtyukh V. V., 2011, Azh, 88, 8Miyamoto M., Nagai R., 1975, PASJ, 27, 533Moultaka J., Ilovaisky S. A., Prugniel P., Soubiran C.,2004, PASP, 116, 693Navarro J. F., Frenk C. S., White S. D. M., 1996, ApJ, 462,563Nissen P. E., Gustafsson B., Edvardsson B., Gilmore G.,1994, A&A, 285, 440Nissen P. E., Chen Y. Q., Schuster W. J., Zhao G., 2000,A&A, 353, 722Nissen P. E., Primas F., Asplund M., Lambert D. L., 2002,A&A, 390, 235Nissen P. E., 2004, in Carnegie Observatories AstrophysicsSer. Vol. 4: Origin and Evolution of the Elements, ed.McWilliam A. & Rauch M. (Cambridge: CambridgeUniv. Press)Nissen P. E., Schuster W. J., 2010, A&A, 511, L10Nissen P. E., Schuster W. J., 2011, A&A, 530, A15¨Onal Ta¸s ¨O, Bilir S., Plevne O., 2018, Ap&SS, 363, 35Pace G., 2013, A&A, 551, 8Pehlivan Rhodin, A., Hartman, H., Nilsson, H., Jonsson, P.2017, A&A, 598, A102Pereira, C. B., Smith, V. V., Drake, N. A., et al. 2017,MNRAS, 469, 774Pereira C. B., Drake N. A., Roig F., 2019, MNRAS, 488,482Peterson R. C., 2013, ApJ, 768, L13Plachowski C. A., Sneden C., Kraft R. P., 1996, AJ, 111,1689Plevne O., ¨Onal Ta¸s ¨O, Bilir S., Seabroke G.M, 2020, ApJ,893, 108, 21 Prochaska J. X., McWilliam A., 2000, ApJ, 537, L57Prugniel P., Soubiran C., 2001, A&A, 369, 1048Prugniel P., Soubiran C., Koleva M., Le Borgne D., 2007,yCat, III/251Ramirez I., Fish J. R., Lambert D. L., Allende Prieto C.,2012, ApJ, 756, 46Ramirez I., Allende Prieto C., Lambert D. L., 2013, ApJ,763, 78Reddy B. E., Tomkin J., Lambert D. L., Allende PrietoC. A., 2003, MNRAS, 340, 304Reddy B. E., Lambert D. L., Allende Prieto C., 2006,MNRAS, 367, 1329Robin A. C., Haywood M., Creze M., Ojha D. K.,Bienayme O., 1996, A&A, 305, 125Roederer I. U., Preston G. W., Thompson I. B. et al., 2014,AJ, 147, 136Roederer I. U., Thompson I. B., 2015, MNRAS, 449, 3889¸Sahin T., Lambert D. L., 2009, MNRAS, 398, 1730¸Sahin T., Lambert D. L., Klochkova V. G., TavolganskayaN. S., 2011, MNRAS, 410, 612¸Sahin T., Lambert D. L., Klochkova V. G., Panchuk V. E.,2016, MNRAS, 461, 4071¸Sahin T., 2017, Turkish Journal of Physics, 41, 367Sahlholdt C. L., Feltzing S., Lindegre L., Church R. P.,2018, MNRAS, 482, 895Sakari, C. M., Placco, V. M., Farrell, E. M., et al. 2018,ApJ, 868, 110Salaris M., Weiss A., 1998, A&A, 335, 943Sesar B. et al., 2007, AJ, 134, 2236Sch¨onrich R., Bergemann M., 2014, MNRAS, 443, 698Sch¨onrich R., McMillan P., Eyer, L. 2019, MNRAS, 487,3568Schuster W. J., Nissen P. E., 1989, A&A, 222, 69Shi J. R., Gehren T., Mashonkina L., Zhao G., 2009, A&A,503, 533Sitnova T., Zhao G., Mashonkina L. et al., 2015, ApJ, 808,148Sneden C., 1973, Ph.D. thesis, Univ. Texas, AustinSobeck J. S., Lawler J. E., Sneden C., 2007, ApJ, 667, 1242Soubiran C., Bienayme O., Siebert A., 2003, A&A, 398, 151Spite M., Peterson R. C., Gallagher A. J., Barbuy B., SpiteF., 2017, A&A, 600, A26Steinmetz M., Zwitter T., Siebert A. et al., 2006, AJ, 132,1645Valenti J. A., Fischer D. A., 2005, ApJS, 159, 141VandenBerg D. A., Brogaard K., Leaman R., CasagrandeL., 2013, ApJ, 775, 134VandenBerg D. A., Bond H. E., Nelan E. P., Nissen P. E.,Schaefer G. H., Harmer D., 2014, ApJ, 792, 110Vasiliev E., 2019, MNRAS, 484, 2832 van Leeuwen F., 2007, A&A, 474, 653Ventura, P., D’Antona, F., Mazzitelli, I., et al. 2001, ApJL,550, L65Webb J. J., Leigh N. W. C., 2015, MNRAS, 453, 3278Wood M. P., Lawler J. E., Sneden C., Cowan J. J., 2013,ApJS, 208, 27Wood M. P., Lawler J. E., Sneden C., Cowan J. J., 2014,ApJS, 211, 20Worley C. C., Cottrell P. L., 2010, MNRAS, 406, 2504Yan H. L., Shi J. R., Nissen P. E., Zhao G., 2016, A&A585, A102 Yanny B., Rockosi C., Newberg H. J. et al., 2009, AJ, 137,4377Yaz G¨ok¸ce E., Bilir S., Karaali S., Plevne O., 2017, Ap&SS,362, 185Zhao G., Gehren T., 2000, A&A, 362, 1077Zhang H. W., Zhao G., 2005, MNRAS, 364, 712Zhang H. W., Gehren T, Zhao G., 2008, preprint(arXiv:0802.2609)Zhao G., Zhao Y. H., Chu Y. Q., Jing Y. P., Deng L. C.,2012, Research in Astronomy and Astrophysics, 12, 723Zhao G., Mashonkina L., Yan H. L. et al., 2016, ApJ, 833,225 Table A1 . Fe i lines used in the analysis of ELODIE spectra. The equivalent widths are those obtained with the LIME code. HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i Table A1 continued Table A1 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i Table A1 continued Table A1 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i Note —References for the adopted gf values: F: Fuhr & Wiese (2006); N: NIST Atomic Spectra Database (http://physics.nist.gov/PhysRefData/ASD)
Table A2 . Fe i lines used in the analysis of ELODIE spectra. The equivalenth widths are those obtained with the LIME code. HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i i i Table A2 continued Table A2 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i Table A2 continued Table A2 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i i i i i i i i i i i Note —References for the adopted gf values: F: Fuhr & Wiese (2006); N: NIST Atomic Spectra Database (http://physics.nist.gov/PhysRefData/ASD).
Table A3 . Lines used in the analysis of ELODIE spectra. The EWs reported for individual lines are those obtained via the LIME code.
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i Table A3 continued Table A3 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i ii ii ii ii ii ii ii ii Table A3 continued Table A3 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Fe ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii i i i i i i i i i i i i i i i i i i ii ii i i i i i i i i i i i Table A3 continued Table A3 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Ca i i i i i i i i i i i i i i i i i i ii ii ii ii ii ii ii ii i i i i i i i i i i i i i i i i i i i i i i i i i i Table A3 continued Table A3 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Ti i i i i i i i i i i i i Note — References for the adopted gf values: B: Kelleher & Podobedova (2008a); K: LG:
Lawler et al. (2013); P: Kelleher & Podobedova (2008b); N: NIST Atomic Spectra Database(http://physics.nist.gov/PhysRefData/ASD);
Ast:
Astrophysical (computed).
Table A4 . Lines used in the analysis of ELODIE spectra. The EWs reported for individual lines are those obtained via LIME code.HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Ti i ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii ii Table A4 continued Table A4 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Ti ii ii ii ii ii ii ii ii ii ii ii i i i i i i i i i i i i i i i i i i i i i i i i i i i Table A4 continued Table A4 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Cr ii ii ii ii ii ii ii i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i Table A4 continued Table A4 (continued)
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Ni i Note —References for the adopted gf values – LW:
Wood, Lawler, Sneden, & Cowan (2013); N: NIST Atomic SpectraDatabase (http://physics.nist.gov/PhysRefData/ASD); G: Lawler et al. (2015); J: Sobeck, Lawler, Sneden (2007); K: BW:
Blackwell-Whitehead &Bergemann (2007);
LWD:
Lawler et al. (2014), hfs-included for 4379 ˚A but no hfs for 6090 ˚A V i line; W: Wood, Lawler,Sneden, & Cowan (2014).
Table A8 . Atmospheric parameters of HPM sample stars from thisstudy -I.Stars T eff log g [Fe/H] ξ Notes( K) (cgs) (dex) (km s − )HD 84937 6000 ±
140 3.50 ± ± ±
50 4.00 ± ± ±
90 3.70 ± ± ±
120 4.00 ± ± ±
80 4.03 ± ± ±
80 4.03 ± ± ±
40 4.10 ± ± ±
100 3.80 ± ± ±
100 4.00 ± ± ± ±
100 4.08 ± ± ±
36 4.15 ± ± ±
97 4.11 ± ± ±
145 4.23 ± ± ±
70 3.93 ± ± ±
85 4.09 ± ± ±
175 3.58 ± ± ±
40 3.90 ± ± ±
70 4.01 ± ± ±
100 3.80 ± ± ±
50 4.07 ± ± ±
50 4.16 ± ± ±
100 4.12 ± ± ± ? 4.47 -1.04 – Reddy, Lambert & Allende Prieto (2006)6177 ±
50 4.14 ± ± Table A5.
Lines used in the analysis of ELODIE spectra. The EWs reported for individual lines are those obtained via LIMEcode.
HD 194598 HD 201891 HD 6755 HD 3567 HD 84937 BD+ 423607Species λ L.E.P log(gf) EW log ǫ (X) EW log ǫ (X) EW log ǫ (X) EW log ǫ ( X ) EW log ǫ (X) EW log ǫ (X) ref(˚A) (eV) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex) (m˚A) (dex)Ni i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i ii ii ii ii ii ii i ii ii ii ii ii ii ii ii ii ii ii gf values – A: Allen & Porto de Mello (2011); W: Wood, Lawler, Sneden, & Cowan (2014); N: NIST Atomic Spectra Database(http://physics.nist.gov/PhysRefData/ASD); K: D: Ljung,Nilsson, Asplund, & Johansson (2006); H: Hannaford et al. (1982);
LW:
Lawler et al. (2006);
LS:
Lawler et al. (2009);
DH:
Dan Hartog, Lawler, Sneden, & Cowan(2003).
Table A8 . Atmospheric parameters of HPM sample stars from thisstudy -I.Stars T eff log g [Fe/H] ξ Notes( K) (cgs) (dex) (km s − )6051 ±
30 4.02 ± ± ±
62 4.29 ± ± ±
92 4.19 ± ± ± ±
165 4.10 ± ± Table A6.
Abundances of observed species for HD 194598, HD 201891, HD 6755, and the Sun. Solar abundances were obtainedby employing the Solar atmosphere from Castelli & Kurucz (2004).
HD 194598 HD 201891 HD 6755 SunElement log ǫ (X) [X/Fe] n log ǫ (X) [X/Fe] n log ǫ (X) [X/Fe] n log ǫ (X) nC i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ... ... 1.72 ± ± ± ± ± ii ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± ii ± ± ± ± ± ii ... ... ... ... ... ... 0.66 ± ± ± ii ± ± ± ± ± ± ± Table A8 . Atmospheric parameters of HPM sample stars from thisstudy -I.Stars T eff log g [Fe/H] ξ Notes( K) (cgs) (dex) (km s − )5981 4.26 -1.15 – Prugniel (2007)6050 ±
100 4.27 ± ± ±
120 4.15 ± ± ±
80 4.27 ± ± ±
20 4.20 ± ± ±
40 4.20 ± ± ±
80 4.33 ± ± ±
80 4.27 ± ± ±
100 4.00 ± ± ±
50 4.31 ± ± ±
20 4.00 ± ± ±
30 4.32 ± ± ±
80 4.31 ± ± Table A7.
Abundances of observed species for HD 3567, HD 84937, BD +42 3607, and the Sun. Solar abundances were obtainedby employing the Solar atmosphere from Castelli & Kurucz (2004).
HD 3567 HD 84937 BD +42 3607 SunElement log ǫ (X) [X/Fe] n log ǫ (X) [X/Fe] n log ǫ (X) [X/Fe] n log ǫ (X) nC i ± ± ± i ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ii ± ± ± ± ± i ± ± ± ± ± ± ± i ... ... ... ... ... ... ... ... ... 3.11 ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ± ± i ± ± ± ± ± ii ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± i ... ... ... ... ... ... ... ... ... 2.80 ± ii ± ± ± ± ± ± ± ii ± ± ± ± ± ± ± ii ... ... ... ... ... ... ... ... ... 1.60 ± ii ± ± ± ± ± ii ± ± ± Table A8 . Atmospheric parameters of HPM sample stars from thisstudy -I.Stars T eff log g [Fe/H] ξ Notes( K) (cgs) (dex) (km s − )HD 201891 5850 ±
160 4.20 ± ± ±
100 4.19 ± ± ±
80 4.24 ± ± ±
100 4.00 ± ± ±
80 4.24 ± ± ±
80 4.28 ± ± ±
40 4.30 ± ± ±
70 4.43 ± ± ±
100 4.45 ± ± ±
50 4.28 ± ± ±
44 4.39 ± ± ±
37 4.32 ± ± ±
41 4.10 ± ± ±
69 4.35 ± ± Table A8 . Atmospheric parameters of HPM sample stars from thisstudy -I.Stars T eff log g [Fe/H] ξ Notes( K) (cgs) (dex) (km s − )BD+423607 5480 ±
130 3.72 ± ± ±
60 5.00 ± ± ±
100 4.00 ± ± ±
50 4.58 ± ± ±
100 4.31 ± ± ±
140 2.90 ± ± ±
40 2.80 ± ± ±
100 2.70 ± ± NOTES ON ABUNDANCES ANALYSIS OFHD 6755From KPNO spectra of the star, the reported neutralsilicon and calcium abundances by Pilachowski, Sneden,Kraft (1996) are in excellent agreement with our mea-surement of those elements (i.e. ∆([X/Fe] TS -[X/Fe] PSK )= 0.00 dex and 0.09 dex, respectively). The nickelabundance is also agreed within 0.1 dex. The largestscatter was observed for abundances of neutral sodium(∆([Na/Fe] TS -[Na/Fe] PSK = 0.25 dex), magnesium (-0.24 dex), and single ionized scandium (-0.21 dex). Theyreported sodium abundance via spectrum synthesis of5682 ˚A and 5688 ˚A Na i lines. The abundance for mag-nesium by Pilachowski, Sneden & Kraft (1996) comesfrom 5711 ˚A Mg i line. Our magnesium abundance isprovided by five magnesium lines reported in Table A4 with log gf -values from Kelleher & Podobedova (2008).The difference in reported log gf -values for this commonmagnesium line is 0.11 dex. The scandium abundanceby Pilachowski et al. (1996) is determined from 5669˚A Sc ii line. The mean difference in the log gf -values ofthe scandium lines (∆(log gf PSK - log gf TS )) is 0.23 dex,hence the difference in scandium abundance is probablydue to the difference in the adopted log gf values for thiselement. Table A4 is available online.
The star was listed in the
SIMBAD database as aspectroscopic binary. We did not see any indication forthe binarity in its ELODIE spectrum.Several studies in the literature reported a sub-giant sta-tus for the star (i.e. Pilachowski et al. 1996; Burris etal. 2000). However, Carretta et al. (2000), with nogiven concrete evidence to support, noted that the starwas evolving to a giant stage in the H-R diagram.The overall agreement between reported abundancesby Fulbright (2000) and in this study for all common ele-ments are satisfactory. For instance, for neutral sodium,silicon, and single ionized barium abundances, the dif-ference is 0.2 dex. The agreement in abundances forcalcium, titanium, chromium, nickel, yttrium, and zir-conium is also excellent (i.e. ∆([X/Fe] TS -[X/Fe] F ) =0.00 dex, -0.11 dex, 0.06 dex, -0.06 dex, -0.07 dex, and0.02 dex, respectively). Magnesium abundance fromthe listed neutral magnesium lines in Table A4 pro-vided ≈ i lines havecomparable equivalent widths in both studies (see Ta-ble A4). Their log gf -values are also agreed well ex-cept the 4703 ˚A Mg i line,i.e. the difference is ≈ TS -[X/Fe] F )) is ≈ -0.17 dex for vanadium.Our determination of neutral vanadium abundance isfrom 4379 ˚A and 6090 ˚A V i lines. The lines were notincluded in the analysis by Fulbright (2000).0 Table A9.
Sensitivity of the derived abundances to the uncertainties of ∆ T eff , ∆ log g , and ∆ ξ in the model atmosphereparameters for HPM stars. ∆ log ǫ HD201891 HD194598 HD6755Species ∆Teff ∆logg ∆ ξ (-) ∆ ξ (+) ∆Teff ∆logg ∆ ξ (-) ∆ ξ (+) ∆Teff ∆logg ∆ ξ (-) ∆ ξ (+)(+160) (+0.33) (-0.50) (+0.50) (+165) (+0.33) (-0.50) (+0.50) (+140) (+0.33) (-0.50) (+0.50)( K) (cgs) (km s −
1) (km s −
1) ( K) (cgs) (km s −
1) (km s −
1) ( K) (cgs) (km s −
1) (km s − i i -0.07 0.00 -0.01 0.01 -0.07 0.00 -0.01 0.00 0.00 0.05 0.13 0.10Mg i -0.12 0.02 -0.06 0.08 -0.12 0.07 -0.03 0.04 -0.10 0.07 -0.05 0.06Si i -0.05 -0.02 -0.07 -0.02 -0.05 0.00 0.00 -0.02 -0.03 0.02 0.07 0.12Si ii i -0.10 0.06 -0.05 0.07 -0.11 0.06 -0.05 0.07 -0.11 0.05 -0.11 0.11Sc i -0.15 0.00 0.00 0.00 -0.10 0.00 0.00 0.00 -0.10 0.00 0.05 0.10Sc ii -0.05 -0.09 -0.05 0.08 -0.06 -0.10 -0.04 0.06 0.02 -0.03 0.05 0.12Ti i -0.13 0.01 -0.04 0.06 -0.14 0.01 -0.04 0.05 -0.17 0.02 -0.11 0.08Ti ii -0.04 -0.10 -0.09 0.10 -0.04 -0.11 -0.06 0.07 -0.03 -0.12 -0.19 0.16V i -0.10 0.00 0.00 0.04 -0.10 -0.10 0.00 0.03 -0.15 -0.03 -0.03 -0.03Cr i -0.14 0.02 -0.07 0.08 -0.12 0.01 -0.05 0.05 -0.15 0.02 -0.17 0.13Cr ii i -0.05 0.05 -0.02 0.13 -0.10 0.00 -0.01 0.02 -0.12 -0.10 0.00 0.03Fe i -0.12 0.04 -0.05 0.08 -0.13 0.03 -0.07 0.08 -0.14 0.04 -0.14 0.13Fe ii i -0.05 0.05 0.10 0.10 -0.15 0.00 -0.10 0.00 -0.05 -0.05 -0.20 0.20Ni i -0.10 0.00 -0.03 0.04 -0.10 0.00 -0.03 0.04 -0.11 0.00 -0.06 0.05Cu i -0.13 0.00 0.00 0.00 -0.10 -0.05 -0.02 -0.05 -0.15 -0.10 -0.10 -0.05Sr i .... .... .... .... .... .... .... .... -0.10 0.00 0.00 0.05Sr ii ii i -0.15 0.00 0.00 0.00 -0.10 0.00 -0.05 -0.05 -0.05 0.15 0.15 0.15Zr ii -0.07 -0.12 -0.07 0.00 -0.10 -0.15 -0.05 0.00 -0.10 -0.17 -0.20 -0.02Ba ii -0.12 -0.09 -0.17 0.05 0.00 0.00 -0.10 0.03 -0.06 -0.08 -0.16 0.12Ce ii .... .... .... .... 0.00 -0.05 0.00 0.05 0.00 -0.15 0.00 0.00Nd ii .... .... .... .... .... .... .... .... 0.00 -0.12 -0.02 0.05Sm ii -0.12 -0.02 -0.12 -0.12 -0.05 -0.05 0.00 0.05 0.00 -0.10 0.00 0.05HD3567 BD+423607 HD84937Species ∆Teff ∆logg ∆ ξ (-) ∆ ξ (+) ∆Teff ∆logg ∆ ξ (-) ∆ ξ (+) ∆Teff ∆logg ∆ ξ (-) ∆ ξ (+)(+175) (+0.27) (-0.50) (+0.50) (+130) (+0.45) (-0.50) (+0.50) (+140) (+0.18) (-0.50) (+0.50)( K) (cgs) (km s −
1) (km s −
1) ( K) (cgs) (km s −
1) (km s −
1) ( K) (cgs) (km s −
1) (km s − i .... .... .... .... .... .... .... .... .... .... .... ....Na i -0.05 0.00 0.05 0.00 -0.10 -0.07 -0.10 0.00 .... .... .... ....Mg i -0.06 0.16 0.11 0.21 -0.09 0.02 0.00 -0.01 -0.06 0.01 -0.05 0.03Si i -0.06 -0.01 -0.01 0.00 0.02 0.00 0.02 0.09 .... .... .... ....Si ii i -0.11 0.04 -0.08 0.11 -0.08 0.03 -0.04 0.04 -0.08 0.00 -0.04 0.02Sc i .... .... .... .... .... .... .... .... .... .... .... ....Sc ii -0.04 -0.12 -0.07 -0.04 -0.04 -0.14 -0.09 -0.01 -0.09 -0.14 -0.06 -0.06Ti i -0.15 0.00 -0.05 0.05 -0.13 0.00 -0.03 0.04 -0.11 0.01 -0.02 0.02Ti ii -0.05 -0.09 -0.10 0.10 -0.07 -0.13 -0.13 0.14 -0.06 -0.06 -0.14 0.09V i -0.10 0.05 0.05 0.05 .... .... .... .... 0.00 -0.10 -0.05 0.00Cr i -0.14 0.00 -0.06 0.05 -0.14 0.05 -0.07 0.10 -0.12 0.00 -0.07 0.04Cr ii i i -0.13 0.03 -0.10 0.11 -0.12 0.02 -0.07 0.06 -0.10 0.00 -0.08 0.05Fe ii -0.01 -0.09 -0.09 0.09 -0.02 -0.14 -0.06 0.08 -0.02 -0.07 -0.09 0.06Co i -0.15 -0.05 -0.10 0.05 -0.15 -0.05 -0.40 -0.05 -0.15 0.00 0.00 0.05Ni i -0.11 0.00 -0.03 0.04 -0.08 -0.03 -0.02 0.01 -0.10 0.00 -0.03 0.01Cu i .... .... .... .... .... .... .... .... .... .... .... ....Sr i -0.10 -0.05 -0.05 -0.05 .... .... .... .... -0.10 -0.15 -0.15 -0.10Sr ii ii i .... .... .... .... .... .... .... .... .... .... .... ....Zr ii ii ii .... .... .... .... .... .... .... .... .... .... .... ....Nd ii ii -0.15 -0.20 -0.10 -0.10 .... .... .... .... .... .... .... .... Table A10.
Abundances of NGC 5897, NGC 4833, and NGC 362 along with the abundances of HD 201891, HD 194598, andHD 3567. The reported errors in abundances are the standard errors.
NGC 5897 HD 201891 NGC 4833 HD 194598 NGC 362 HD 3567Element [X/Fe] Ref [X/Fe] N [X/Fe] Ref [X/Fe] N [X/Fe] Ref [X/Fe] NNa i ± ± ± ± ± ± i ± ± ± ± ± ± i ± ± ± ± ± ± i ± ± ± ± ± ± ii ± ± ± ± ± ± i ± ± ± ± ± ± ii ± ± ± ± ± ± i -0.14 ± ± ± ± -0.14 ± -0.50 ± i -0.20 ± ± ± ± ± ± i -0.43 ± ± ± ± ± ± i ± ± ± ± ± ± i -0.07 ± ± ± ± ± ± i -0.70 ± -0.11 ± -0.65 ± -0.31 ± ± ii -0.29 ± ± ± ± ± ± ii ... ... 0.10 ± ± ± ± ± ii -0.03 ± ± ± ± ± ± ii ± ± ± ± R ≈
20 000) reported ≈ ii and Ba ii abundances. Their stron-tium abundance is provided by 4077 ˚A and 4215 ˚A Sr ii lines. In this study, the reported strontium abundanceis from Sr ii gf -values in both stud-ies are adequate. For Y ii , Zr ii , and Nd ii abundances,the difference is < TS -[X/Fe] F )= -0.03 dex, 0.03 dex, and -0.08 dex, respectively). We alsosearched for La ii , Eu ii and Dy ii lines in the ELODIEspectrum of the star. The line profile for La ii line at4333 ˚A is seen to have contribution in its red wing. In-spection of line profiles for Eu ii lines at 4129 ˚A and 4205˚A via spectrum synthesis indicated blending for thoselines in the ELODIE spectrum. Also, the red wing of theDy ii line at 4077 ˚A is found to be highly contributedby Sr ii .The reported abundances by Mishenina & Kovtyukh(2001) for Mg i and Sr i are ≈ i abundance is excellent (i.e.∆([Ca/Fe] TS -[Ca/Fe] MK ) = 0.02 dex). The Si i , Ba ii ,and Nd ii abundances present ≈ ≈ ≈ B. NOTES ON ABUNDANCES ANALYSIS OFBD +42 3607Carney, Wright, Sneden et al. (1997) obtained abun-dances of Li, Mg, Ca, Ti, and Ba abundances forBD +42 3607. No silicon abundance was reported. Theirreported titanium ([Ti ii /Fe]=0.66 dex) and magnesiumabundances ([Mg/Fe]=0.25 dex) are agreed with ourabundances of those elements in this study within er-ror limits. The abundances for calcium and barium aresomewhat discrepant with our reported abundances forthose elements. For instance, the difference in elementover iron ratio for the calcium is ≈ gf values for the cal-cium lines employed in the analysis. For instance, when the atomic data for the common calcium lines at 5581 ˚A,5588 ˚A, and 5590 ˚A are scrutinized, their log gf valuesare seed to differ ≈ ≈ ii lines at 4554 ˚A and 5853 ˚A.The lines were synthesized for abundance determinationand the hfs were also considered for the lines.The nickel and calcium abundances reported by Grat-ton, Carretta, Claudi et al. (2003) are agreed with theabundances reported for the star in this study. Theagreement in nickel abundance in two studies is excel-lent, i.e. the difference in abundance is 0.04 dex. How-ever, calcium abundance by Gratton, Carretta, Claudiet al. (2003) is 0.35 dex lower. The difference can beexplained via differences in the adopted log gf values intwo studies. There are eleven common calcium lines toboth studies. The common calcium lines at 5261 ˚A, 5588˚A, 6166 ˚A, and 6169 ˚A show ≈ gf -values. The difference in log gf for commoncalcium lines at 5590 ˚A, 6439 ˚A, and 6717 ˚A is at 0.1dex level.Zhang & Zhao (2005) analyzed spectra of 32 metalpoor stars. Some discrepancies are present for the abun-dances of Si i/ii and Ca i .Boesgard et al. (2011) used HIRES spectrum andtheir reported abundances as for [Mg/Fe] and [Ti/Fe]show good agreement with our abundances reported forthese elements. The difference in [Mg i /Fe] ratios in bothstudy is 0.04 dex. Their reported titanium abundance([Ti/Fe]) is in excellent agreement with [Ti ii /Fe] ratioobtained in this study.Peterson (2013) used ultraviolet part of the HIRESspectrum. The reported abundances by Peterson (2013)for common elements in both studies as for Ca, Mn,Sr, Y, Zr, Nd show fair agreement for manganese andzirconium. The exceptions are for calcium, strontium,and yttrium abundances. The abundances for those el-ements in the ultraviolet present somewhat discrepantabundances. For instance, the abundance difference foryttrium is 0.57 dex while the difference in element overiron ratios between Peterson (2013) and this study forstrontium is 0.45 dex.Boeche & Grebel (2016) determined model parame-ters of the star from a moderate resolution spectrum.They did not report silicon abundances. Their reportedcalcium abundance of [Ca/Fe]=0.20 differs 0.4 dex fromour measurement in this study. The element over ironratios for Mg, Sc, Ti, and Cr show excellent agreement(i.e. the differences are within 0.1 dex). They reported ≈ HD 6755
Fe IIFe I
BD + 423607
Fe IIFe I
HD 201891
Fe IIFe I
HD 194598
Fe IIFe I
HD 3567
Fe IIFe I
HD 84937
Fe IIFe I
Figure A1.
An example for the determination of atmospheric parameters T eff and ξ using abundance (log ǫ ) as a function ofboth lower level excitation potential (LEP, upper and middle panels) and reduced EW (REW; log (EW/ λ ), bottom panel) forHD 6755, BD +42 3607, HD 201891, HD 194598, HD 3567, HD 84937. The solid line in the all panels is the least-square fit to thedata. Co 4.69Co 4.19Co 3.69
Sr 2.47Sr 1.97Sr 1.47
V 3.37V 2.87V 2.37
HD 201891 | CH A
Mn 4.76Mn 4.26Mn 3.76
C 8.13C 7.63C 7.13 | CrI
Y 1.51Y 1.01Y 0.51 | TiI R e l . F l u x Wavelength (Å)
Co 4.42Co 3.92Co 3.42
Zr 2.10Zr 1.60Zr 1.10
Y 1.39Y 0.89Y 0.39
HD 1945984378.75 4378.98 4379.22 4379.450.550.680.810.941.07
V 3.20V 2.70V 2.20 | CH A
Ba 1.60Ba 1.10Ba 0.60
C 8.21C 7.71C 7.21 R e l . F l u x Wavelength (Å)
Sr 2.60Sr 2.10Sr 1.60
Nd 1.50Nd 1.00Nd 0.50 | CH A
V 3.15V 2.65V 2.15
HD 3567 | CH A
Ba 1.73Ba 1.23Ba 0.73
Sr 2.25Sr 1.75Sr 1.25
Fe I
Co 4.07Co 3.57Co 3.07 R e l . F l u x Wavelength (Å)
Sr 1.65Sr 1.20Sr 0.65
Co 3.17Co 2.67Co 2.17
Sr 1.30Sr 0.80Sr 0.30
HD 849374553.65 4553.87 4554.08 4554.300.450.610.780.941.10
Ba −0.07Ba −0.57Ba −1.07
Y 0.44Y −0.06Y −0.56
Ca 4.96Ca 4.46Ca 3.96 R e l . F l u x Wavelength (Å)
V 2.71V 2.21V 1.71 | CH A
Ce 0.71Ce 0.21Ce −0.29 | Cr I
Sm 0.51Sm 0.01Sm −0.49
HD 6755 | Fe I | Cr I+Sm II | Sm II
Nd 0.96Nd 0.46Nd −0.04
Zr 1.91Zr 1.41Zr 0.91
Y 1.05Y 0.55Y 0.05
Ti I R e l . F l u x Wavelength (Å)
Ba 0.00Ba −0.50Ba −1.00
Sr 1.47Sr 0.97Sr 0.47
Ca 4.90Ca 4.40Ca 3.90
BD + 4236074246.5 4246.7 4246.9 4247.10.400.580.750.931.10
Sc 1.37Sc 0.87Sc 0.37
Ti 3.43Ti 2.93Ti 2.43
Ti 3.46Ti 2.96Ti 2.46 R e l . F l u x Wavelength (Å)
Figure A2.
The observed (filled circles) and computed (full red line) line profiles for some of the neutral metal lines used inthe analysis of all program stars. The computed profiles show synthetic spectra for the abundances reported in Table A4-A6. C. NOTES ON ABUNDANCES ANALYSIS OFHD 201891In a search for possible non-LTE affects on magnesiumabundances from magnesium lines at 4703 ˚A, 5528 ˚A ,and 5711 ˚A in the FOCES spectrum of metal poor dwarfstars, Zhao & Gehren (2000) reported LTE-non-LTE dif-ferences in abundances of these magnesium lines to be ≈ -0.1 dex. In the current study, the reported magne-sium abundance for HD 201891 in Table A7 are providedby the Mg i lines at 4571 ˚A and 5711 ˚A. The magne-sium over iron ratios for those lines from the ELODIEspectrum are 0.37 dex and 0.24 dex, respectively. Theabundance for the common magnesium line at 5711 ˚Ashow excellent agreement with that reported by Zhao &Gehren (2000) for this line (i.e. 0.24 dex vs. 0.29 dex).The abundances (e.g. [Element/Fe]) by Fulbright(2000) for Na i , Mg i , Si i , Ca i , Ti i , Cr i , V i and Ni i are in excellent agreement with the reported abundancesof those elements in this study. The difference is ≤ ≈ ii /Fe] to ≈ ii /Fe].Mishenina & Kovtyukh (2001) also used the ELODIEspectrum for their spectroscopic analysis but reportedsomewhat discrepant abundances for Mg i and Y ii . Thedifferences for those elements between two studies areat ≈ ii lines. Their adopted log gf -values by Mishenina & Kovtyukh (2001) are slightly dif-ferent. For instance, the log gf for the former line agreedwithin 0.04 dex, however, the difference in log gf for 5087˚A line is 0.12. The [Ba ii /Fe] ratio by Mishenina &Kovtyukh (2001) also differed 0.2 dex from the bariumabundance in this study. The element over iron ratiosfor Si i , Ca i , and Sr i are agreed within 0.1 dex.The reported abundances by Gratton et al. (2003)for Na i , Si i , Ca i , Ti i , Ti ii , V i , Sc ii , and Ni i showexcellent agreement with the abundance reported in thisstudy for those elements (i.e. ∆[X/Fe] GCC - [X/Fe] TS ≤ i abundance: thedifference is 0.08 dex. The difference in element overiron ratios for magnesium, single ionized chromium, andmanganese abundances is ≈ The tables for the summary of the model parameters andmean abundances by Fulbright (2000) listed two different [Fe/H]values for HD 201891 as -1.0 dex and -1.12 dex, respectively.
Mishenina et al. (2003) used model atmosphere pa-rameters from Mishenina & Kovtyukh (2001) and re-ported non-LTE Na abundance for star as [Na/Fe]=0.07dex from relatively weak Na i lines at 5682 ˚A (of 29.0m˚A), 6154 ˚A (of 5.0 m˚A), and 6160 ˚A (of 13.0 m˚A). It isin excellent agreement with our measurement of sodiumabundance, i.e. the difference is 0.04 dex.The abundances reported by Reddy, Lambert & Al-lende Prieto (2006) via high resolution spectroscopy forNa i , Mg i , Si i , Ca i , Sc ii , Ti i , V i , Cr i , Ni i , Y ii , andBa ii is agreed well within 0.1 dex. The [C/Fe] differed ≈ ≈ ≈ R ≈
40 000) FOCES spectrum of the star and reportedLTE and non-LTE abundances of Mg, Na, and Al.They’re as follows: ([Mg/Fe]
LTE ,[Mg/Fe]
NLTE ) = (0.32,0.43), ([Na/Fe]
LTE ,[Na/Fe]
NLTE ) = (0.21, 0.07), and([Al/Fe]
LTE ,[Al/Fe]
NLTE )= (0.24, 0.59). The reportedLTE abundance of magnesium by Gehren et al. (2006)showed excellent agreement (i.e. 0.01 dex different) withthe reported magnesium abundance in this study. Onthe other hand, they reported 0.1 dex higher sodiumabundance, however, they did not provide the atomicdata for the lines of those elements that are included intheir analysis.Mishenina et al. (2011) reported abundances for Na,Al, Cu, and Zn. They reported [Na/Fe] = -0.02 dex.We did not dedect lines of aluminum, copper, and zincin the ELODIE spectrum. The sodium abundance re-ported for the star in Table A7 is agreed with the sodiumabundance reported by Mishenina et al. (2011) within0.1 dex.The α -element abundances by Bensby, Feltzing andOey (2014) shows excellent agreement with the α -element abundances reported for the star in this study(i.e. ∆[X/Fe] BFO - [X/Fe] TS ≤ i /Fe]ratio in this study differs 0.17 dex from that of Bensby,Feltzing and Oey (2014). The differences in elementover iron ratios between two studies for Sc ii , Ti i , Cr i ,Ni i , Y ii and Ba ii is ranging between 0.02 and 0.07 dex.Battisini & Bensby (2015) adopted model parame-ters from Bensby, Feltzing and Oey (2014) and reportedabundances for the star for Sc ii , and Mn i as [Sc ii /Fe]=0.17 dex and [Mn i /Fe]= -0.26 dex, respectively. Theirreported manganese abundance is from 6016 ˚A Mn i line. The logarithmic abundance reported for the lineby Battisini & Bensby (2015) is 4.12 dex. This is inexcellent agreement with the logarithmic abundance re-ported for this line in this study (see Table A7). The6listed manganese abundance for the star in this study isfrom eight Mn i lines in Table A7 . The scandium abun-dance is also agreed well, i.e. the difference in scandiumabundances is 0.05 dex. For vanadium, the reportedabundance by Boeche & Grebel (2016) is agreed withour determination. The difference is negligible, i.e. 0.03dex. It is ≈ gf -values for their spectroscopicanalysis.The reported zirconium abundance by Battisini &Bensby (2015) via model parameters from Bensby, Feltz-ing and Oey (2014) for the Zr ii line at 4208 ˚A is inexcellent agreement with the calculated abundance forthis line in this study.On the basis of its computed abundances in thisstudy, the spectrum of this candidate benchmark starfor the Gaia (Hawkins et al. 2016) indicates a halo likechemisty. D. NOTES ON ABUNDANCES ANALYSIS OFHD 194598Zao & Gehren (2000) using FOCES spectrum, re-ported LTE and non-LTE magnesium abundances ofthe star over four neutral lines of magnesium: 4571 ˚A(LTE/non-LTE; 0.25/0.33), 4703 ˚A (0.20/0.32), 5528 ˚A(0.24/0.32) ve 5711 ˚A (0.19/0.31) with the model pa-rameters reported in Table A9 .The abundances (i.e. [element/Fe]) obtained by Ful-bright (2000) with the model atmosphere parametersreported in this study (Table A9) are in very good agree-ment (within 0.1 dex) for Na i , Si i , Ca i , Ti i , V i , Cr i ,Ni i , and Y ii . The vanadium lines employed by Ful-bright (2000) at 4389 ˚A, 6090 ˚A, and 6216 ˚A have mea-sured equivalent widths of 9.0 m˚A, 0.3 m˚A, and 0.2 m˚A,respectively. The lines are too weak to be measured inthe ELODIE spectrum. Their reported single ionizedzirconium and barium abundances are only ≈ i /Fe] is 0.25 dex.The abundances obtained from the ELODIE spectrumby Mishenina & Kovtyukh (2001) is in very good agree-ment with our measurements, i.e. the differences in [ele-ment/Fe] ratios are less than 0.1 dex for Mg i , Si i , Ca i ,Y ii , and Ba ii . The single ionized cerium abundance(i.e. [Ce ii /Fe]) shows 0.18 dex difference. Mishenina &Kovtyukh (2001) reported Ce ii lines at 4486, 4562, 4572,4628, and 5274 ˚A however, it was unclear whether these Table A7 is available online. Table A9 is available online. The tables for the summary of the model parameters andmean abundances by Fulbright (2000) present two different [Fe/H]values for HD 194598 as -1.1 dex and -1.23 dex, respectively. lines were used for cerium abundance determination byMishenina & Kovtyukh (2001) for HD 194598. The 4628˚A Ce ii line, apparently a single common cerium line inboth studies, and has a reported log gf -value which is0.04 dex less than the adopted log gf -value for the linein the present study.The reported abundances of, Na i , Si i , Ca i , Ti i ii ,Cr i ii , Mn i , and Ni i over model parameters reportedin Table A9 by Gratton et al. (2003) are in agreementwith the computed abundances of those elements in thisstudy within 0.1 dex. The difference for single ionizedscandium abundance (in [Sc/Fe]) is 0.2 dex. The highestdifference in abundance was observed for magnesium: itis 0.3 dex. For the three common Mg i lines to both stud-ies at 4703 ˚A, 5528 ˚A, and 5711 ˚A, their measured equiv-alent widths in the ELODIE spectrum as well as adoptedlog gf -values in this study are in accordance with thosereported by Gratton et al. (2003). Vanadium abun-dance by Gratton et al. (2003) is in excellent agreementwith the vanadium abundance in this study, [V i /Fe] TS -[V i /Fe] GCC ≈ i lines at 5727 ˚A, 6090 ˚A, and6216 ˚A. and the lines are too weak to provide reliableabundances. Their equivalent widths by Gratton et al.(2003) are 5.8 m˚A, 2.3 m˚A, and 3.2 m˚A, respectively.Mishenina et al. (2003) computed non-LTE abun-dances of weak sodium lines at 6154 ˚A with equiva-lent width of 4.2 m˚A and at 6160 ˚A of 5.9 m˚A. Theyreported [Na/Fe]=0.00 dex which is in excellent agree-ment with the sodium abundance reported for the starin Table A9 of this study.Gehren et al. (2006) used FOCES ( R LTE ,[Na/Fe]
NLTE )= (0.00, -0.14) and([Mg/Fe]
LTE , [Mg/Fe]
NLTE )= (0.16, 0.28). The agree-ment in abundances for those elements is good within0.1 dex.Nissen & Schuster (2010) used UVES and FIES spec-tra of the star for their abundance analysis. Their analy-sis was based on abundances computed with astrophysi-cal log gf -values. The calculated magnesium abundancein the LTE over MARCS model atmospheres and FIESspectrum slightly differs, i.e. +0.05 dex. The overallagreement in abundances for Na i , Mg i , Si i , Ca i , Ti i ,Cr i , and Ni i is noteworthy, i.e. the differences in abun-dance between two studies is less than 0.1 dex.Abundances reported for Mn ([Mn/Fe]=-0.1 dex), Sr([Sr/Fe]=-0.3 dex), and Zr ([Zr/Fe]=0.2 dex) by Peter-7son (2013) over HIRES spectrum agreed with the ob-tained abundances of those elements in this study. Thedifferences in abundances for cobalt and yttrium is alsoagreed within ≈ i .Our cobalt abundance is from 4121 ˚A Co i line (TableA5).Bensby, Feltzing, and Oey (2014) reported abun-dances for Na i , Mg i , Si i , Ca i , Ti i , Cr i , Ni i , Y ii , andBa ii . The over all agreement in abundances for thoseelements is satisfactory, i.e. the differences are < ii , V i ,Mn i and Co i . In the ELODIE spectrum, even for arelatively short wavelength coverage, we were able toreport abundances for those elements (Table A7).The reported element over iron ratios for Mg i , Si i ,Ti i , and Cr i are in excellent agreement with the abun-dances of these elements in this work. The differencesare seen to be within 0.1 dex. The [Ca/Fe], [Co/Fe],and [Ni/Fe] by Boeche & Grebel (2016) differed ≈ ≈ i lines at 5105 ˚A as [Cu i /Fe] = -0.46 (-0.35) and for 5218 ˚A as [Cu i /Fe] = -0.44 (-0.38). Theiratomic data for copper was from the NIST database.Fishlock et al. (2017) used MIKE spectrum ( λλ ii /Fe]) and cerium ([Ce ii /Fe]) abundances show ≈ wasdetermined via spectrum synthesis. The line provideda [Ce ii /Fe] of 0.17 dex. The zirconium abundance isobtained by 4208 ˚A Zr ii line. The line was observed inthe ELODIE spectrum (see Table A6). Its abundanceis determined via line synthesis. Fishlock et al.(2017)determined scandium abundance from 5526 ˚A Sc ii linevia spectrum synthesis. Table A6 is available online. E. NOTES ON ABUNDANCES ANALYSIS OFHD 3567The abundances (i.e. [element/Fe]) by Fulbright(2000) for neutral magnesium, silicon, and titaniumagreed within 0.2 dex. The same difference was also ob-served for single ionized yttrium, zirconium, and bariumabundances. The calcium abundance of the star in thisstudy is in excellent agreement with the calcium abun-dance by Fulbright (2000) (i.e. ∆([Ca/Fe]
T S -[Ca/Fe] F )= +0.03 dex). Our determination of vanadium abun-dance is obtained using 4379 ˚A V i line. There is nocommon lines of vanadium in the spectroscopic analysisby Fulbright (2000), however, the agreement is good,i.e. ∆([V/Fe] T S -[V/Fe] F ) = -0.13 dex. Our chromiumabundance from 19 neutral chromium lines shows agood agreement with their result (i.e. ∆([Cr i /Fe] T S -[Cr i /Fe] F ) = -0.07 dex). Nickel abundance also showsvery good agreement (i.e. ∆([Ni/Fe] T S -[Ni/Fe] F ) =+0.02 dex). Overall, the comparison is satisfactory.Gratton et al. (2003), over UVES and SARG spec-tra of the star, reported abundances for neutral sodium,magnesium, calcium, vanadium, and manganese thatare in agreement with elemental abundances in thisstudy within 0.1 dex. A similar difference was alsoseen for single ionized chromium (i.e. ∆([Cr ii /Fe] T S -[Cr ii /Fe] GCC )= -0.14 dex). The neutral silicon andneutral titanium abundances are in excellent agreementwith those reported by Gratton et al. (2003). The abun-dance for single ionized scandium differed only 0.15 dexand the difference in abundance for single ionized tita-nium was 0.25 dex. The Ti ii line at 4583 ˚A is commonin both studies (Table A5). Its log gf in this study isfrom Wood et al. (2013) and is agreed with the log gf from Gratton et al. (2003).Zhang & Zhao (2005) via medium resolution ( R ≈
37 000)spectra over λλ T S -[Si/Fe] ZZ ) = -0.02 dex). The chromium abundanceis also shows a very good agreement (i.e. ∆(ours-Zhang) = -0.01 dex). As it comes to abundances forneutral calcium and single ionized barium, the differ-ence in abundance is ≈ gf values for the scandium are takenfrom the NIST database (Table A4). Our log gf valuesfor manganese lines are from Blackwell-Whitehead &Bergemann (2007). The hfs is only included for 6016˚A Mn i line. The vanadium abundance in this study is8provided from the V i gf -value for the line in Table A5 is the astrophysical log gf .Figure A2 shows the best-fitting synthetic spectrum forthe 4379 ˚A V i line.Reddy, Lambert and Allende Prieto (2006) used ahigh resolution spectra ( R ≈ λλ T S -[X/Fe]
RLP )) for chromium, mag-nesium, silicon, and manganese abundances are rangingfrom -0.07 dex for Mn to +0.13 dex for Cr. The dif-ferences for sodium, calcium, and scandium abundancesare at 0.3 dex level. For cobalt and yttrium abundances,it was -0.5 dex and -0.19 dex, respectively. The onlycommon cobalt line with Reddy, Lambert and AllendePrieto (2006) is the 4792 ˚A Co i line. We adopted astro-physical log gf -value for the line from Reddy, Lambertand Allende Prieto (2003). It was obtained via invertingsolar and stellar spectra.Nissen & Schuster (2010), in a study for spectroscopicanalysis of 94 dwarf stars used UVES spectrum of thestar to obtain abundances for Na i , Mg i , Si i , Ca i , Ti i ,Cr i , and Ni i . Their reported abundances for those ele-ments are in excellent agreement with our measurements(less than 0.1 dex).The study by Hansen et al. (2012) over a mixture ofgiant and dwarf stars including HD 3567 reported abun-dances for common elements. These included Sr, Y, Zr,and Ba abundances for the star. The element over ironratios by Hansen et al. (2012) for Sr and Y is in ex-cellent agreement with our results (the difference is 0.06dex). The abundance ratio for Zr showed ≈ ≈ gf -values for the lines byHansen et al. (2012) is slightly lower (0.05 dex for 4208˚A; 0.07 dex for 4317 ˚A) compared to the log gf -valuesreported in Table A6 in this study.The reported abundances by Bensby, Feltzing & Oey(2014) for Na, Mg, Si, Ca, Ti, Cr, Ni, and Y show excel-lent agreement ( ≤ ≈ T S -[Co/Fe] BG = -0.36dex. Boeche & Grebel (2016) did not provide atomicdata for their spectroscopic analysis. Yan et al. (2016) used UVES spectrum and model at-mosphere parameters from Nissen & Schuster (2010) toobtain both LTE and non-LTE abundances for Cu fortheir halo sample stars. The atomic data was takenfrom NIST database. Using the Cu line at 5105 ˚Athey reported an LTE abundances of [Cu/Fe]=-0.73 dex.( T eff =6051, log g =4.02, [Fe/H]=-1.16 dex ξ =1.5 kms − ). The line is too weak to be measured in the ELODIE spectrum.Fishlock et. al.(2017) used MIKE spectrum of thestar ( λλ T S -[X/Fe]
F Y K )) for scandium andzirconium are -0.16 dex and +0.03 dex, respectively. Theobserved difference in neodymium abundance in bothstudies is 0.15 dex. The Nd ii abundance by Fishlock et.al.(2017) was based on 4706 Nd ii line RMT-3. How-ever, the line has a measured equivalent width of 4 m˚Ain their MIKE spectrum.Bergemann et al. (2017) adopted model parametersfrom Hansen et al. (2012) to report magnesium abun-dance for the star. The [Mg/Fe] ratio by Bergemann etal. (2017) is 0.1 dex higher (i.e.∆([X/Fe] TS -[X/Fe] B )= -0.1 dex). Our determination of magnesium abundance isfrom six neutral magnesium lines (see Table A4). Theseincluded the resonance line of Mg i at 4571 ˚A, and opticaltriplet lines at 5172 ˚A and 5183 ˚A. All these lines areknown to be sensitive to the model atmosphere struc-ture. However, we also included Mg i lines at 5528 ˚Aand 5711 ˚A. The calculated logarithmic abundances forlatter two are in excellent agreement with the reportedmean magnesium abundance over six magnesium linesin Table A8. F. A GAIA
BENCHMARK STAR: HD 84937Table A8 gives the results on equivalent width analy-sis of the spectrum for the following model parameters: T eff = 6000 ±
140 K, log g = 3.50 ± ± ξ = 1.60 km s − .Fulbright (2000), over a high resolution spectrum ofthe star, reported abundances for several species in-cluding Na, Mg, Si, Al, Ca, Ti, V, Cr, Ni, Y, Zr, Bave Eu. Their listed [element/Fe] ratios for HD 84937slightly differ for some of the species (i.e. they reported[Y/Fe]=0.03 ± ± i ), and titanium (Ti ii ) abundances in bothstudies are in excellent agreement. Our results for thecalcium and nickel is consistent with their results (i.e.within ≈ ELODIE spectra of the metal-poor stars,Mishenina & Kovtyukh (2001) reports [Y ii /Fe] = -0.02 ± ± i /Fe]=0.36 ± i /Fe] = 0.34 ± ii /Fe] = -0.10 ± ii line forthe listed abundance in Table A6 for the line.Over high resolution McD, UVES and SARG spec-tra of the selected metal-poor stars, the abundances re-ported by Gratton et al. (2003) for HD 84937 showedonly ≈ ≈ i /Fe] = 0.18 ± i lines at3905 ˚A and 4102 ˚A. The former was not observed inthe ELODIE spectra where the 4102 ˚A Si i line is seento be blended hence the line is not used in the siliconabundance determination in current study.Ishigaki et al. (2012, 2013) via high resolution( R ≈ λλ ≈ T S -[X/Fe]
BF O ) = +0.06 dex). Ti i , Ni i , andY ii abundances were agreed within 0.2 dex. Excep-tion is for neutral calcium and single ionized bariumabundances: the differences were +0.27 dex and -0.55 dex, respectively. The log gf values for Ba ii by Bensby,Feltzing & Oey (2014) showed 0.2 dex difference fromthe log gf -values reported in this study for 4554 ˚A and5853 ˚A Ba ii lines. The difference in log gf is ≈ ii line. Hence, it is apparent thatthe difference in barium abundance between two stud-ies may partially be due to atomic data differences inboth studies. For instance, for calcium, over five com-mon Ca i lines in both studies, the mean difference inloggf values is -0.05 ± i lines showed ≈ gf -values by Bensby, Feltzing & Oey (2014). Thelog gf values for Ca i in this study is from the NISTdatabase.The magnesium abundance by Boeche & Grebel(2016) showed a fair agreement, of about ≈ ≈ R =60 000; λλ i and Sc ii abundances for thestar were agreed within 0.1 dex. [Ca i /Fe] ratio differedby ≈ i /Fe] T S -[Ca i /Fe] Z ) = +0.28dex). Single ionized titanium and barium abundancesshowed excellent agreement (i.e.∆([X/Fe] T S -[X/Fe] Z )=0.0 dex and -0.03 dex, respectively). The differencein abundance for single ionized zirconium was 0.2 dex.The largest difference in abundance was observed for thestrontium, i.e. ∆([Sr/Fe] T S -[Sr/Fe] Z )= +0.6 dex. Theiradopted log gf values for the common Sr ii lines at 4077˚A and 4215 ˚A are in excellent agreement with thoseadopted from the NIST database in this study.More recent work by Spite et al. (2017) reported onlyMg and Ca abundances as for α -element (e.g. Mg, Si,and Ca) for this halo main-sequence (turnoff) star. Inaddition to the modest overabundance of α -elements,a solar value of [C/Fe] was reported. Our result formagnesium abundance is in excellent agreement withSpite et al. (2017). The difference in calcium abundancebetween two studies is only 0.05 dex. http://physics.nist.gov/PhysRefData/ASD C I Na I Mg I Si I Si II Ca I Sc I Sc II Ti I Ti II V I Cr I Cr II Mn I Fe I Fe II Co I Ni I Cu I Sr II Y II Zr I Zr II Ba II Ce II Sm II−0.50.00.5 H D [ X / F e ] This studyFulbright (2000)Mishenina at al.(2001)Gratton et al. (2003)Gehren et al. (2006)Nissen & Schuster (2010)Peterson (2013)Fishlock et al. (2017)C I Na I Mg I Si I Si II Ca I Sc I Sc II Ti I Ti II V I Cr I Cr II Mn I Fe I Fe II Co I Ni I Cu I Sr I Sr II Y II Zr I Zr II Ba II Sm II−0.50.00.5 H D [ X / F e ] Thi1 12ud5Fulb0igh2 (2000)Mishenina at al. (2001)Gratton et al. (2003)Mishenina (2011)Bensby at al. (2014)Gehren et al. (2006)C I Na I Mg I Si I Si II Ca I Sc I Sc II Ti I Ti II V I Cr I Cr II Mn I Fe I Fe II Co I Ni I Cu I Sr I Sr II Y II Zr I Zr II Ba II Ce II Nd II Sm II−1.0−0.50.00.5 H D [ X / F e ] This StudyFulbright (2000)Mishenina at al.(2001) Burris et al. (2000) Pilachowski at al. (1996)C I Na I Mg I Si I Si II Ca I Sc II Ti I Ti II V I Cr I Cr II Mn I Fe I Fe II Co I Ni I Sr I Sr II Y II Zr II Ba II Nd II Sm II01 H D [ X / F e ] This studyFulbright (2000)Gratton et al. (2003)Nissen & Schuster (2010)Zhang & Zhao (2005)Fishlock et. al.(2017)Bergemann et al. (2017)M( I Ca I Sc II Ti I Ti II V I C0 I C0 II M. I Fe I Fe II Co I Ni I S0 I S0 II Y II 0 II Ba II−0.50.00.5 H D [ X / F e ] This studyFulbright (2000)Mishenina at al. (2001)Be.1by (2014)Na I M( I Si I Si II Ca I Sc II Ti I Ti II C0 I C0 II M. I Fe I Fe II Co I Ni I S0 II Y II 0 II Ba II Nd II−0.50.00.51.01.5 B D + [ X / F e ] This studyCarney et al. (1997)