Bulge RGB Abundance Trends

Light, Alpha, and Fe–Peak Element Abundances in the Galactic Bulge

Abstract

We present radial velocities and chemical abundances of O, Na, Mg, Al, Si, Ca, Cr, Fe, Co, Ni, and Cu for a sample of 156 red giant branch stars in two Galactic bulge fields centered near (l,b)=(5.25,–3.02) and (0,–12). The (5.25,–3.02) field also includes observations of the bulge globular cluster NGC 6553. The results are based on high resolution (R20,000), high signal–to–noise (S/N70) FLAMES–GIRAFFE spectra obtained through the ESO archive. However, we only selected a subset of the original observations that included spectra with both high S/N and that did not show strong TiO absorption bands. The present work extends previous analyses of this data set beyond Fe and the –elements Mg, Si, Ca, and Ti. While we find reasonable agreement with past work, the data presented here indicate that the bulge may exhibit a different chemical composition than the local thick disk, especially at [Fe/H]–0.5. In particular, the bulge [/Fe] ratios may remain enhanced to a slightly higher [Fe/H] than the thick disk and the Fe–peak elements Co, Ni, and Cu appear enhanced compared to the disk. There is also some evidence that the [Na/Fe] (but not [Al/Fe]) trends between the bulge and local disk may be different at low and high metallicity. We also find that the velocity dispersion decreases as a function of increasing [Fe/H] for both fields, and do not detect any significant cold, high velocity population. A comparison with chemical enrichment models indicates that a significant fraction of hypernovae are required to explain the bulge abundance trends, and that initial mass functions that are steep, top–heavy (and do not include strong outflow), or truncated to avoid including contributions from stars 40 M are ruled out, in particular because of disagreement with the Fe–peak abundance data. For most elements, the NGC 6553 stars exhibit nearly identical abundance trends to comparable metallicity bulge field stars. However, the star–to–star scatter and mean [Na/Fe] ratios appear higher in the cluster, perhaps indicating additional self–enrichment.

Subject headings:
stars: abundances, Galactic bulge: general, bulge: Galaxy: bulge, stars: Population II

1. Introduction

Understanding the formation and subsequent evolution of the Galactic bulge is important both for interpreting observations of extragalactic populations and for constraining Galaxy chemodynamical formation models. Recent large sample spectroscopic surveys, such as the Bulge Radial Velocity Assay (BRAVA; Rich et al. 2007a; Howard et al. 2008; Howard et al. 2009; Kunder et al. 2012), the Abundances and Radial velocity Galactic Origins Survey (ARGOS; Freeman et al. 2013; Ness et al. 2012; 2013b), the Apache Point Observatory Galactic Evolution Experiment (APOGEE; Majewski et al. 2010; Nidever et al. 2012), and the GIRAFFE Inner Bulge Survey (GIBS; Zoccali et al. 2014) provide a coherent view of the bulge as a barred, triaxial system exhibiting cylindrical rotation. Photometric and star count studies have also discovered a double red clump along some bulge sight lines that traces out an X–shaped structure (McWilliam & Zoccali 2010; Nataf et al. 2010; Saito et al. 2011). This structure appears to be dominated by stars with [Fe/H]–0.5 on bar–supporting orbits (Soto et al. 2007; Babusiaux et al. 2010; Ness et al. 2012; Uttenthaler et al. 2012; but see also Nataf et al. 2014).

Inclusive with these data are detailed composition analyses of field stars from moderate and high resolution spectroscopy (McWilliam & Rich 1994; Ramírez et al. 2000; Rich & Origlia 2005; Cunha & Smith 2006; Fulbright et al. 2006; Zoccali et al. 2006; Fulbright et al. 2007; Lecureur et al. 2007; Rich et al. 2007b; Cunha et al. 2008; Meléndez et al. 2008; Zoccali et al. 2008; Alves–Brito et al. 2010; Bensby et al. 2010a; Ryde et al. 2010; Bensby et al. 2011; Gonzalez et al. 2011; Hill et al. 2011; Johnson et al. 2011; Johnson et al. 2012; Rich et al. 2012; Uttenthaler et al. 2012; Barbuy et al. 2013; Bensby et al. 2013; García Pérez et al. 2013; Johnson et al. 2013a; Ness et al. 2013a; Jönsson et al. 2014) finding, at least in a general sense, that the bulge is composed of stars spanning more than a factor of 100 in [Fe/H]7, that bulge stars are uniformly enhanced in their [/Fe] ratios at low metallicity relative to the thin disk, and that the median [Fe/H] along bulge sight lines decreases as a function of increasing Galactic latitude (i.e., there is a metallicity gradient). The enhanced [/Fe] abundances, coupled with the low [La/Eu] ratios of bulge stars (McWilliam et al. 2010; Johnson et al. 2012), are consistent with the notion that the bulge formed rapidly (1–3 Gyr). In fact, the bulge appears uniformly old (10 Gyr) in age studies based on color–magnitude diagram analyses (e.g., Ortolani et al. 1995; Zoccali et al. 2003; Clarkson et al. 2008; Valenti et al. 2013; but see also Ness et al. 2014), and Clarkson et al. (2011) estimate from the blue straggler population in an inner bulge field that a truly young (5 Gyr) population should not constitute more than 3.4 of the bulge. In contrast, ages derived from microlensed dwarf studies (e.g., Bensby et al. 2013) find that while all metal–poor bulge stars are uniformly old, 5–25 of metal–rich stars, at least near the Galactic plane, may be only 2–8 Gyr in age.

While the observational data continue to grow, the difficult task of assembling the pieces into a fully self–consistent model of the bulge’s formation remains open. The chemodynamical bulge data are challenging to interpret. The bulge’s predominantly old age, enhanced [/Fe] ratios, vertical metallicity gradient, and the existence of possible “primordial building blocks” such as Terzan 5 (e.g., Ferraro et al 2009; Origlia et al. 2011; 2013) are more consistent with the classical, merger built formation scenario. However, the bulge’s boxy X–shape, similar composition characteristics to at least the thick disk, and cylindrical rotation profile suggest that the bulge formed via secular evolution from a buckling disk instability and may be a “pseudobulge” (e.g., Kormendy & Kennicutt 2004; but see also Zoccali et al. 2014). While Shen et al. (2010) rule out a classical bulge component that exceeds 8 of the disk mass, it may still be possible for a bar to form within a pre–existing classical bulge (e.g., Saha et al. 2012). Additionally, evidence such as the metallicity gradient may not be unique to the classical bulge scenario, and may be consistent with a secular evolution model in which a radial metallicity gradient in the buckling disk is transformed into a vertical gradient in the resultant bar (Martinez–Valpuesta & Gerhard 2013). The bulge may also be composed of at least two stellar populations with different composition and kinematics (Babusiaux et al. 2010; Hill et al. 2011; Bensby et al. 2011; Bensby et al. 2013; Ness et al. 2013a). However, at the moment the exact nature of these potentially distinct stellar populations is far from certain.

Although most of the chemical abundance work mentioned previously has focused on the [Fe/H] and [/Fe] ratios in comparison with the thin and thick disks, the light odd–Z and Fe–peak (and also neutron–capture) elements also provide discriminatory power between models and other stellar populations (e.g., see Kobayashi et al. 2011; their Figure 14). The Fe–peak elements in particular are useful as they may be sensitive to formation environment and metallicity. For example, the metallicity dependent yields and increased contributions of massive stars are predicted to produce enhanced [Cu/Fe] and [Zn/Fe] ratios in the bulge compared to the local disk. Similarly, if the bulge formed with a significantly flatter initial mass function (IMF) than the disk then bulge stars should exhibit very large [Co/Fe] and [Zn/Fe] ratios (Nomoto et al. 2013). Therefore, here we measure abundances of the Fe–peak elements Cr, Fe, Co, Ni, and Cu, in addition to the light odd–Z and –elements O, Na, Mg, Al, Si, and Ca, in 156 red giant branch (RGB) stars in two Galactic bulge fields at (l,b)=(5.25,–3.02) and (0,–12), and compare the abundance ratios with other bulge fields, the Galactic disk, and chemical enrichment models.

2. Observations, Target Selection, and Data Reduction

The FLAMES–GIRAFFE spectra for this project are based on data obtained from the ESO Science Archive Facility under request number 51251, which are based on observations collected at the European Southern Observatory, Paranal, Chile (ESO Program 073.B–0074). Details regarding the selection of targets and input parameters (e.g., photometry and astrometry) are given in Zoccali et al. (2008). To briefly summarize, fibers were placed on K giants approximately 1–2 magnitudes brighter than the bulge red clump, and the spectra were obtained in high resolution mode (R/20,000). The original program by Zoccali et al. (2008) included four fields centered at (l,b)=(1.14,–4.18), (0.21,–6.02), (0,–12), and (5.25,–3.02). While the (1.14,–4.18) and (0.21,–6.02) fields were observed in the HR 13, HR 14, and HR 15 setups (spanning 6100–6950 Å), the (5.25,–3.02) and (0,–12) fields were observed in the HR 11, HR 13, and HR 15 setups (5590–5835 Å; 6100–6400 Å; 6600–6950 Å). Since the HR 11 setup is the only one containing measurable copper lines, we have only analyzed GIRAFFE spectra from the (5.25,–3.02) and (0,–12) fields. We note that the (5.25,–3.02) field also includes the bulge globular cluster NGC 6553.

Figure 1 shows a 2MASS (Skrutskie et al. 2006) color–magnitude diagram of our final target selection from the archival data. The raw data set obtained from the ESO archive included observations of 205 RGB stars in the (5.25,–3.02) field and 109 RGB stars in the (0,–12) field. However, we only analyzed spectra for which the co–added signal–to–noise (S/N) ratio exceeded 70. We also discarded spectra that exhibited strong TiO absorption bands, for which a “standard” equivalent width (EW) analysis would be inappropriate. The final sample utilized here includes 75/205 stars (37) in the (5.25,–3.02) field and 81/109 stars (74) in the (0,–12) field. In Figure 1 we also identify stars that are likely members of NGC 6553 (see Section 3.5). In particular, note the broad dispersion in the color–magnitude diagram of cluster members, as well as with stars within 5 of the cluster center. This highlights the combined effects of differential reddening and population mixing along the line–of–sight toward the (5.25,–3.02) field. The star names and coordinates from the raw image headers and Zoccali et al. (2008), as well as available 2MASS photometry and star identifiers, are provided in Table 1.

Figure 1.— left panel: Color–magnitude diagram for the field centered near (l,b)=(5.25,–3.02). The filled red circles are all stars observed with the FLAMES instrument. The black outlined circles are those presented in this paper. The filled blue boxes indicate stars with radial velocities and metallicities consistent with belonging to the globular cluster NGC 6553. The small filled gray circles indicate all stars in the 2MASS catalog within 30 of the central coordinates. Similarly, the small filled magenta circles indicate all stars in the 2MASS catalog within 5 of NGC 6553. right panel: A similar plot but with the observed stars for the (l,b)=(0,–12) field shown in green.

The raw science and calibration data were downloaded and re–reduced using the GIRAFFE Base–Line Data Reduction Software (girBLDRS)8. In particular, the pipeline software was used to carry–out bias subtraction and overscan trimming, dark correction, fiber identification, flat–fielding, wavelength calibration, scattered light correction, and spectrum extraction. Sky subtraction was carried out using the IRAF9 skysub routine. Individual exposures were continuum normalized using a low order polynomial via the IRAF continuum routine, and the telluric band in the HR 13 spectra was removed using the IRAF task telluric and a set of FLAMES templates obtained during a different observing program with the same spectrograph setup. The individual spectra were shifted to a common velocity scale (i.e., the heliocentric velocity was removed) and co–added using IRAF’s scombine task.

3. Data Analysis

3.1. Model Stellar Atmospheres

The four primary model atmosphere input parameters of effective temperature (T), surface gravity (log(g)), metallicity ([Fe/H]), and microturbulence (vt) were determined via spectroscopic analyses. For stars in the (0,–12) field, we used the model parameters given in Zoccali et al. (2008) as a starting point before converging to a solution. However, the adopted model atmosphere parameters for stars in the (5.25,–3.02) field are not provided in Zoccali et al. (2008) nor Gonzalez et al. (2011). Therefore, we adopted the generic values T=4500 K, log(g)=2.0 cgs, [Fe/H]=–0.20 dex, and vt=1.5 km s before converging to a solution. The final parameters given in Table 1 were derived by enforcing Fe I excitation equilibrium for T, ionization equilibrium between Fe I/II10 for log(g), and removing trends in Fe I abundance versus line strength for vt. The final models were interpolated within the available grid of AODFNEW (–enhanced) and ODFNEW (scaled–solar) ATLAS9 model atmospheres11 (Castelli & Kurucz 2004). Stars with [/Fe]0.15 were measured using the –enhanced models, and we used the scaled–solar models for stars with [/Fe]0.15. However, the issue of an –enhanced versus scaled–solar model should not introduce an error in the abundance ratios that exceeds the 0.05–0.10 dex level (e.g., Fulbright et al. 2006; Alves–Brito et al. 2010; Johnson et al. 2013a).

Figure 2 shows our spectroscopically determined temperature and surface gravity values in comparison with the the spectroscopic T and photometric log(g) values given in Zoccali et al. (2008). As is evident in Figure 2, the spectroscopic determination of both parameters leads to a more extended distribution of surface gravities. This is likely due to the unavoidable problem that one must assume a distance (and mass) when deriving a photometric surface gravity. However, this is only a major issue when determining abundances of elements from transitions that are strongly sensitive to log(g). The model atmosphere parameters determined here are well–bounded by and follow the expected trends of the 10 Gyr isochrones with [Fe/H]=–1.5 (–enhanced) and [Fe/H]=0.5 ([/Fe]=0) shown in Figure 2.

Figure 2.— left panel: A plot of surface gravity versus effective temperature for all stars analyzed in this paper. The symbols are color–coded into rough metallicity bins. Metal–poor, –enhanced (blue) and metal–rich, –normal (red) 10 Gyr isochrone sequences (Dotter et al. 2008) are shown for guidance. right panel: The effective temperature (excitation equilibrium) and surface gravity (photometric) values employed by Zoccali et al. (2008) for the same stars presented here in the (0,–12) field. The literature model atmosphere parameters for stars in the (5.25,–3.02) field are not available for comparison.

We do note that 25/156 (16) stars in our sample converged to a solution in which log(g)3–3.5. The derived higher gravity values suggest some of these stars may be foreground lower RGB and subgiants rather than more evolved bulge RGB stars. A better measurement of surface gravity, either from the addition of more than 2–3 Fe II lines or the inclusion of more sensitive atmospheric pressure indicators, would better constrain the true nature of these stars. We do not find any strong systematic differences in the derived [X/Fe] ratios between stars of “low” and “high” gravity12, but it is unclear if the similar abundances should have any bearing when interpreting bulge versus thin/thick disk composition differences (see Section 4). However, the high gravity stars are also relatively metal–rich [Fe/H]=0.09 (=0.25), located preferentially on the blue half of the color–magnitude diagrams, and have a relatively small velocity dispersion (=55 km s for log(g)3). These data provide additional circumstantial evidence that the high gravity stars may be foreground, though possibly inner disk, contaminators (see also Zoccali et al. 2008; their Section 7).

In Figure 3 we compare the derived model atmosphere parameters between this study, Zoccali et al. (2008), and Gonzalez et al. (2011). We find good agreement in the derived T values with an average difference of only 2 K (=98 K). The dispersion of 100 K is reasonable given the different line lists and model atmospheres (but similar technique of excitation equilibrium) used. As mentioned previously, there is some discrepancy in log(g), especially for the highest gravity stars, between the present work and Zoccali et al. (2008). For stars with log(g)2.5, the average difference in log(g) is 0.01 dex (=0.29 dex), but for stars with log(g)2.5 the magnitude of the average gravity difference is 0.64 dex (=0.37 dex). Comparing the microturbulence values, which may be particularly sensitive to line choice and can vary as a function of gravity, we find an average difference of 0.18 km s (=0.27 km s).

Figure 3.— Derived model atmosphere parameters are compared between this work and Zoccali et al. (2008). Similar to Figure 2, the temperature, gravity, and microturbulence values are only available for the (0,–12) field in Zoccali et al. (2008). However, the metallicity panel compares our results to those in both the (0,–12) (Zoccali et al. 2008) and (5.25,–3.02) (Gonzalez et al. 2011) fields. In all panels the solid black line indicates perfect agreement.

When comparing derived [Fe/H] values, we find good agreement for [Fe/H]0.2 with an average difference of 0.03 dex (=0.13 dex). However, as is evident in Figure 3 our derived [Fe/H] values are systematically higher on average by 0.18 dex (=0.13 dex), for stars with [Fe/H]0.2. The source of this discrepancy may be related to the large 1 [Fe/H] uncertainties given in Zoccali et al. (2008) for stars with [Fe/H]0. This is illustrated in Figure 4 where we plot the 1 [Fe/H] uncertainties between our work and Zoccali et al. (2008) as a function of [Fe/H]. Ideally one expects to have measurement errors that are not correlated with metallicity, as is the case here. For the Zoccali et al. (2008) subsample in common with the present analysis, the line–to–line dispersions are comparable only for stars with [Fe/H]0.2.

Figure 4.— The 1 line–to–line dispersion values for [Fe/H] measurements in this work (filled red circles), the original Zoccali et al. (2008) stars selected for reanalysis (blue crosses), and the full Zoccali et al. (2008) sample that includes all four fields (filled gray circles).

3.2. Equivalent Width Abundance Determinations

The abundances of Fe I, Fe II, Si I, Ca I, Cr I, and Ni I were determined by measuring EWs via an interactive, semi–automatic code developed for this project. The measurement process followed the “standard” procedure of fitting single or multiple Gaussian profiles to the spectra for isolated and weakly blended lines, respectively. However, the measurement time frame was significantly reduced by implementing a simple machine learning algorithm that kept track of user input on a per–line basis to make an educated first guess for subsequent measurements in other stars of: the number of profiles to fit, profile fitting edges, and the central wavelength, width, and central depth of all associated nearby features. While all EW measurements were manually inspected, as was mentioned in Section 2 we selected stars from the archival data based primarily on S/N considerations in an effort to reduce measurement uncertainties. Sample spectra for stars of similar temperature but different metallicity are shown in Figure 5 to illustrate typical data quality in the three spectrograph setups.

Figure 5.— Sample spectra are shown to illustrate both data quality and the change in line strengths and continuum availability for stars of similar temperature but varying metallicity.

The line lists for this project were created by visually examining the high S/N spectra of cool metal–poor and metal–rich giants in the sample, finding all isolated and/or weakly blended features for elements of interest, and merging the two line list sets. This was done to ensure that a roughly equivalent number of lines could be used in metal–rich and metal–poor stars, and the manual inspection of each fit enabled us to discard prohibitively strong and weak lines. On average, the Fe I, Fe II, Si I, Ca I, Cr I, and Ni I abundances were based on the measurement of 70, 2, 8, 6, 6, and 16 lines, respectively. The log(gf) values were set via an inverse abundance analysis relative to Arcturus. We adopted the Arcturus model atmosphere parameters from Fulbright et al. (2006). Similarly, for Fe, Si, and Ca we adopted the Arcturus abundances from Fulbright et al. (2006), and for Cr and Ni we adopted the Arcturus abundances from Ramírez & Allende Prieto (2011). The final line list, including the adopted Arcturus and derived solar abundances (based on measurements of the Hinkle et al. 2000 Arcturus and solar atlases), are provided in Table 2. The derived solar abundances for Fe, Si, Ca, and Cr agree within 0.05 dex of the values given in Asplund et al. (2009).

The final abundances of Fe I, Fe II, Si I, Ca I, Cr I, and Ni I, determined using the abfind driver of the LTE line analysis code MOOG (Sneden 1973; 2010 version), are given in Table 3. Note also that the [Fe/H] values given in Table 1 are the average of the [Fe I/H] and [Fe II/H] abundances given in Table 3. However, the average difference in the sense [Fe I/H]–[Fe II/H] is 0.00 dex with a small dispersion (=0.02 dex).

3.3. Spectrum Synthesis Abundance Determinations

For the element abundances derived from transitions involving a small number of lines that are affected by significant blends from prevalent spectral features, such as molecules and Ca I autoionization, and/or broadened due to isotopes and/or hyperfine structure, we used spectrum synthesis rather than EW analyses. For the present work this list includes [O I], Na I, Mg I, Al I, Co I, and Cu I. The abundances were determined using the parallelized version of the synth driver for MOOG (Johnson et al. 2012). For O, Na, Mg, and Al, we adopted as a reference point the Arcturus abundances given in Fulbright et al. (2006). However, as described below the reference Arcturus abundances for Co and Cu are based on measurements using the Kurucz (1994) and Cunha et al. (2002) hyperfine structure line lists.

The specific reasons for using synthesis are slightly different for each element given above. The 6300.30 Å [O I] line is blended with both a Sc II feature at 6300.69 Å and a Ni I feature at 6300.33 Å. Additionally, for most stars in this sample the oxygen abundance is sensitive to the molecular equilibrium calculations set by the carbon and nitrogen abundances as well. Using the CN line list from the Kurucz (1994) database, we iteratively solved for the O and CN abundances in each star. For sodium, the 6154.23 Na I line is relatively clean, but the 6160.75 Na I line is partially blended with two relatively strong Ca I lines. The three Mg I lines at 6319 Å are strongly affected by a broad Ca I autoionization feature, which we set by fitting the slope of the pseudo–continuum from 6316–6318 Å. The 6696.02 and 6698.67 Å Al I lines are both affected by CN, particularly in cooler and more metal–rich stars. Therefore, as with [O I] we simultaneously fit the Al I doublet and nearby CN features. The odd–Z isotope Co constitutes almost 100 of the cobalt abundance. While the 5647.23 and 6117.00 Å Co I lines are relatively weak (EW 50 mÅ), we included the hyperfine structure components from the Kurucz (1994) line list in our syntheses. For copper, which is dominated by the two odd–Z isotopes Cu and Cu, we assumed a solar system mixture of 69.17 and 30.83, respectively. We adopted the hyperfine line list of Cunha et al. (2002) and derive a similar solar abundance of log (Cu)=4.04 but a slightly lower Arcturus abundance than McWilliam et al. (2013). Although the 5782.11 Å Cu I line is strong in most of our stars, the hyperfine broadening helps desaturate the line profile to some extent.

In Figure 6 we show sample syntheses of the O, Mg, and Cu features for a typical metal–rich spectrum. We note that the 5782 Å Cu I line is also sometimes affected by a nearby diffuse interstellar band (DIB). The width and depth of the DIB feature was found to be highly variable. The level of contamination depends on the relative velocity between the interstellar cloud and the individual star and also the reddening value. Therefore, stars in the (5.25,–3.02) field, which have an average E(B–V)=0.7, were more strongly affected than those in the (0,–12) field, which have an average E(B–V)=0.2 (Zoccali et al. 2008). Most of the stars listed in Table 3 that do not have a [Cu/Fe] abundance listed were omitted because of obvious contamination by the DIB feature.

Figure 6.— Sample spectrum synthesis fits are shown for the Cu I, [O I], and Mg I features. In all panels the solid black line indicates the best–fit value. The dashed red and blue lines indicate changes to the best–fit abundance by 0.3 dex, respectively.

3.4. Radial Velocities

Radial velocities were measured using the XCSAO code (Kurtz & Mink 1998) for each individual exposure of every star and in all three filters. For reference templates we generated synthetic spectra ranging in temperature from 4250 to 5000 K (250 K steps), log(g) from 0.5 to 3.5 cgs (0.5 dex steps), [Fe/H] from –1.5 to 0.5 dex (0.5 dex steps), and vt from 1 to 2 km s (0.25 km s steps). Radial velocities were determined relative to the nearest template. We found the average agreement between exposures to be 0.15 km s (=0.13 km s). The heliocentric corrections were taken from the headers of the pipeline reduced files, and the heliocentric radial velocities (RV) listed in Table 1 represent the average value of all exposures and filters for each star.

The kinematic properties of the bulge have been extensively discussed in dedicated survey papers (e.g., Rich et al. 2007a; Howard et al. 2009; Rangwala et al. 2009a; Babusiaux et al. 2010; Kunder et al. 2012; Ness et al. 2013b; Nidever et al. 2012; Babusiaux et al. 2014; Zoccali et al. 2014). Therefore, here we seek only to place our results in context with those surveys. Figure 7 shows velocity histograms for both fields, the velocity distribution as a function of [Fe/H], and the velocity dispersion as a function of [Fe/H]. While a detailed comparison between our measured velocities and those in Babusiaux et al. (2010) is not possible because their individual velocities were not published, for both fields we can compare our average results with those given in Figure 13 of Zoccali et al. (2008) and Table 3 of Babusiaux et al. (2010). For the (5.25,–3.02) field, ignoring NGC 6553 stars, we find average velocity and dispersion values of 4.55 km s and 95.51 km s, respectively. This compares well with the Zoccali et al. (2008) average velocity of 11 km s and velocity dispersion of 107 km s. Similarly, in the (0,–12) field we measured an average heliocentric radial velocity of –8.61 km s (=85.56 km s) compared to the Babusiaux et al. (2010) values of –14 km s (=80 km s). Additionally, as can be seen in Figure 8 our galactocentric radial velocity (V) distributions are similar to those of nearby fields from the BRAVA, GIBS, and APOGEE surveys.

Figure 7.— top left: The red histogram (20 km s bins) illustrates the heliocentric radial velocity distribution function for the (5.25,–3.02) field. The bulge globular cluster NGC 6553 is labeled. top right: The green histogram (20 km s bins) illustrates the heliocentric radial velocity distribution function for the (0,–12) field. bottom left: Heliocentric radial velocity is plotted as a function of [Fe/H] for the (5.25,–3.02) and (0,–12) stars, which are shown as filled red and filled green circles, respectively. The NGC 6553 stars (filled blue boxes) are particularly evident in this panel. bottom right: The heliocentric radial velocity dispersion is plotted as a function of (binned) [Fe/H], using the same color scheme as the other panels. For the middle [Fe/H] bin the blue box and red circle indicate the velocity dispersion with (blue) and without (red) the NGC 6553 stars included.

With the exception of the stars obviously related to NGC 6553, we find in agreement with previous bulge studies that, at least away from the Galactic plane, the velocity distributions are normal with no evidence for significant cold populations (but see also Rangwala et al. 2009b). This contrasts with Nidever et al. (2012) and Babusiaux et al. (2014), which find kinematically cold populations with V200 km s. However, their fields are significantly closer to the Galactic plane than those analyzed here. We do note however that these high velocity populations are also not found in the BRAVA, ARGOS, nor GIBS analyses, nor is there yet a satisfactory theoretical explanation for their origin (e.g., Li et al. 2014).

Figure 8.— top: The striped red, solid gray, and black lined histograms compare the galactocentric radial velocity distributions between the (5.25,–3.02) field analyzed here and nearby fields observed as part of APOGEE and GIBS, respectively. The narrow peak near the center of the distribution is due to NGC 6553. bottom: The striped red and gray histograms compare the galactocentric radial velocity distributions between the (0,–12) field analyzed here and the relatively nearby (0,–8) field from the BRAVA survey.

For the (0,–12) field we observe the same trend of a decrease in velocity dispersion with increasing [Fe/H] found by Babusiaux et al. (2010). However, while Babusiaux et al. (2010) find an increase in velocity dispersion with increasing [Fe/H] in the (1.1,–4) field of Baade’s window, our off–axis but similar Galactic latitude field at (5.25,–3.02) still exhibits a trend of decreasing velocity dispersion with increasing [Fe/H]. This further contrasts with recent fields observed close to the plane by Babusiaux et al. (2014; see their Figure 16) that also show a possible increase in velocity dispersion with increasing [Fe/H].13 Our result is more similar to studies of outer bulge fields that find a consistent decrease in velocity dispersion with increasing [Fe/H] (e.g., Johnson et al. 2011; Uttenthaler et al. 2012; Johnson et al. 2013; Ness et al. 2013b). There is weak evidence in Figure 7 that the trend in velocity dispersion and [Fe/H] may be more shallow for the (5.25,–3.02) field compared to the (0,–12) field. Note that the inclusion (or not) of NGC 6553 stars significantly affects the velocity dispersion of the [Fe/H] bin in which the cluster resides.

3.5. Identifying NGC 6553 Members

Members of the globular cluster NGC 6553 in the (5.25,–3.02) field are best identified in the velocity–metallicity diagram in Figure 7. The likely members (12 stars total) are clustered near [Fe/H]–0.10 and RV0 km s. Literature measurements of the cluster’s average [Fe/H] value vary considerably, with estimates that include: –0.55 (Barbuy et al. 1999), –0.16 (Cohen et al. 1999), –0.7 (Coelho et al. 2001), –0.3 (Origlia et al. 2002), –0.2 (Meléndez et al. 2003), and –0.2 (Alves–Brito et al. 2006). However, we find in agreement with the most recent estimates that [Fe/H]=–0.11 (=0.07). While the cluster is slightly iron–deficient relative to the Sun, the moderate enhancements of the cluster’s [/Fe] ratio (see Section 4.1) gives it an overall metallicity that is roughly solar. NGC 6553 is therefore one of the most metal–rich globular clusters in the Galaxy.

We find similar agreement with literature values for the cluster’s radial velocity, with RV=–2.03 km s (=4.85 km s). This is compared with recent values of: –1 km s (Coelho et al. 2001), 1.6 km s (Meléndez et al. 2003), and –1.86 km s (Alves–Brito et al. 2006). Finally, we note that the stars identified in Table 1 as possible cluster members have an average, projected radial distance from the cluster center of about 6 (=5). We have adopted a more lenient radial distance discriminator than the 2 limit used by Zoccali et al. (2008) and Gonzalez et al. (2011), and instead rely more on the [Fe/H] and velocity measurements to identify possible cluster members.

3.6. Abundance Ratio Comparisons with Previous Work

As noted previously, Zoccali et al. (2008) and Gonzalez et al. (2011) presented [Fe/H], [Si/Fe], [Ca/Fe], and [Ti/Fe] abundances based on the same GIRAFFE data utilized here. Therefore, in Figures 910 we compare our results with theirs for stars and elements in common. While a quantitative comparison of the individual [Fe/H] values is given in Section 3.1 (see also Figure 3), in Figure 9 we compare the general shapes and bulk properties of the metallicity distribution functions. For the (5.25,–3.02) field the average and median [Fe/H] ratios are similar, but the distribution from the present work is somewhat broader and extends to higher [Fe/H]. In contrast, there are no significant differences in the [Fe/H] distribution functions between the present work and the same stars from Zoccali et al. (2008), for the (0,–12) field. We also reconfirm one of the primary conclusions of Zoccali et al. (2008) that interior bulge fields have a higher average metallicity than outer bulge fields. Finally, we note that the distribution functions shown in Figure 9 do not provide strong evidence supporting the existence of multiple, discreet populations, as has been suggested in some studies (Bensby et al. 2011; Hill et al. 2011; Bensby et al. 2013; Ness et al. 2013a). However, the number of stars per field presented here is 100.

Figure 9.— top left: The red and gray histograms (0.1 dex bins) illustrate the derived metallicity distribution functions for the (5.25,–3.02) field in this work and Gonzalez et al. (2011), respectively. top right: The solid red and dashed gray lines illustrate the cumulative distribution functions for this work and Gonzalez et al. (2011), respectively. bottom left: The green and gray histograms (0.1 dex bins) illustrate the derived metallicity distribution functions for the (0,–12) field in this work and Zoccali et al. (2008), respectively. bottom right: The solid green and dashed gray lines illustrate the cumulative distribution functions for this work and Zoccali et al. (2008), respectively.

In Figure 10 we compare our derived [Mg/Fe], [Si/Fe], and [Ca/Fe] ratios to those given in Gonzalez et al. (2011). The average differences between the present work and that of Gonzalez et al. (2011) are [Mg/Fe]=0.00 (=0.14), [Si/Fe]=0.00 (=0.13), and [Ca/Fe]=0.06 (=0.14). The relatively consistent star–to–star scatter of 0.14 dex is a reasonable estimate of the attainable precision between the two studies, which derive –element abundances from different techniques (synthesis in Gonzalez et al. 2011 and EW measurements here). We note that the –elements oxygen (measured here) and titanium (measured in Gonzalez et al. 2011) were not both measured in each study.14

Figure 10.— A comparison between the [Mg/Fe], [Si/Fe], and [Ca/Fe] abundances derived here and in Gonzalez et al. (2011). The solid black line in each panel indicates perfect agreement.

3.7. Abundance Uncertainty Estimates

We investigated the sensitivity of derived abundances for each element in every star by taking the abundances given in Table 3, determining theoretical EWs using the line list in Table 2, and then varying the model atmosphere parameters T, log(g), [Fe/H], and vt individually while holding the other parameters fixed. We selected parameter changes of 100 K in T, 0.30 dex in log(g), 0.15 dex in [M/H], and 0.30 km s in vt, which are reasonable when comparing our derived parameters with those of the independent analysis by Zoccali et al. (2008; see also Section 3.1). The total uncertainty for each element ratio in each star resulting from this exercise is provided in Table 4.

In general, most elements are not affected by changes in T of 100 K at more than the 0.1 dex level. However, the two species presented here that reside in their dominant ionization states ([O I] and Fe II) are strongly affected by changes in surface gravity. For a change in log(g) of 0.3 dex, the log (O) and log (Fe II) abundances can change by 0.1–0.3 dex, but these effects are mitigated when the [O I/H] abundance is normalized with [Fe II/H]. These two species are also more strongly affected by changes in the model metallicity, and the larger [Fe II/H] measurement and sensitivity uncertainties are a contributing factor to the increased dispersion in the [O/Fe] ratios compared to other –elements (e.g., [Mg/Fe]). As expected, microturbulence sensitivity is correlated with a star’s overall metallicity (i.e., line strength). Among the transitions under consideration here, in metal–rich stars those of Na, Ca, and Cu typically have the strongest lines and are thus more strongly affected by the microturbulence uncertainty.

In Table 5 we also provide the 1 line–to–line dispersion values for all species measured here. These values should be mostly representative of the combined measurement error that includes effects such as: continuum placement, line deblending, synthesis fits via visual inspection, log(gf) uncertainties, and model atmosphere deficiencies. Typical line–to–line dispersion values are 0.08 dex. The measurement error of Cu may be underestimated because of the line’s large EW, non–negligible blending (see Figure 6), and possible contamination with a nearby DIBS feature. A more reasonable measurement uncertainty for Cu is, in most cases, 0.15–0.20 dex.

4. Results and Discussion

4.1. The Elements Oxygen, Magnesium, Silicon, and Calcium

The –elements have been the primary focus of detailed composition work in the Galactic bulge. To first order there is agreement among the various studies that: (1) the [/Fe] ratios are enhanced by 0.3 dex at [Fe/H]–0.3, (2) for stars with [Fe/H]0.3 there is a mostly monotonic decline in [/Fe] with increasing [Fe/H], (3) the bulge and thick disk may share similar chemistry over a wide range in metallicity, and (4) there are no significant variations in the [/Fe] trends between different bulge sight lines (McWilliam & Rich 1994; Cunha & Smith 2006; Zoccali et al. 2006; Fulbright et al. 2007; Lecureur et al. 2007; Meléndez et al. 2008; Alves–Brito et al. 2010; Bensby et al. 2010a; Ryde et al. 2010; Bensby et al. 2011; Gonzalez et al. 2011; Hill et al. 2011; Johnson et al. 2011; Uttenthaler et al. 2012; Bensby et al. 2013; García Pérez et al. et al. 2013; Johnson et al. 2013; Ness et al. 2013a; Jönsson et al. 2014). Additionally, there is evidence that the [O/Mg] ratio declines with increasing metallicity (Fulbright et al. 2007; Lecureur et al. 2007; McWilliam et al. 2008; Alves–Brito et al. 2010).

The new data presented here, and summarized in Figure 11, reinforce many observations from the previous studies mentioned above. In particular, we find that for [Fe/H]–0.3 all of the [/Fe] ratios are enhanced and exhibit minimal star–to–star scatter with [O/Fe]=0.54 (=0.10), [Mg/Fe]=0.33 (=0.08), [Si/Fe]=0.28 (=0.07), and [Ca/Fe]=0.34 (=0.09). For stars with [Fe/H]–0.3, we find that the [/Fe] ratios decrease with increasing [Fe/H]. However, Figure 11 illustrates the disparate trends for individual elements and highlights the information loss that can occur when averaging abundance ratios for multiple –elements. The [O/Fe] ratios are higher by 0.2 dex in metal–poor stars than those of other –elements, but this trend reverses for stars with [Fe/H]0 where [O/Fe] is, on average, lower by 0.2 dex. While both O and Mg are significant products of hydrostatic burning in massive stars (e.g., Woosley & Weaver 1995), the [Mg/Fe] trend exhibits a more shallow decline with increasing [Fe/H] than [O/Fe]. This is most clearly seen in Figure 12, which shows a sharply declining [O/Mg] ratio at [Fe/H]–0.1. Although massive star production of Si and Ca involves both hydrostatic and explosive burning (e.g., Woosley & Weaver 1995), the [Si/Fe] and [Ca/Fe] trends are more similar to [Mg/Fe] than [O/Fe]. Given the disparate trend of [O/Fe] compared to other –elements, and the low production of most –elements relative to Fe in Type Ia supernovae (SNe; e.g., Nomoto et al. 1997), we conclude in agreement with past work (e.g., McWilliam et al. 2008) that the strong decline in [O/Fe] at [Fe/H]–0.3 is likely a result of metallicity dependent yields in massive stars15. The influence of metallicity dependent yields on the bulge composition profile, especially from Wolf–Rayet stars, is further supported by fluorine measurements (Cunha et al. 2008; Jönsson et al. 2014), but it remains to be seen if this scenario can be reconciled with the observed carbon and nitrogen trends (Ryde et al. 2010; but see also Cescutti et al. 2009).

Figure 11.— [X/Fe] abundance patterns plotted as a function of [Fe/H] for all elements analyzed. The filled red circles, filled green circles, and filled blue boxes differentiate stars belonging to the (5.25,–3.02), (0,–12), and NGC 6553 populations. Note that the scale of the ordinate is identical in all panels.

When comparing the individual [/Fe] trends between the two fields analyzed here, Figure 11 shows no significant variations. Similarly, in Figure 13 we combine our two fields and compare with literature giant and dwarf [/Fe] data. A comparison between the present work and literature giant trends, which span a variety of bulge sight lines, leads us to find in agreement with Johnson et al. (2011; 2013) and Gonzalez et al. (2011) that no significant field–to–field [/Fe] variations exist over a broad region of the bulge. The microlensed dwarf data exhibit the same qualitative and quantitative distributions for [O/Fe] and [Mg/Fe], at least for [Fe/H]0, as the giant data, but there may be small systematic offsets with [Si/Fe] and [Ca/Fe]. In particular, the dwarf abundances are 0.1 dex lower for a given [Fe/H], when considering [Fe/H]0. At super–solar [Fe/H] values, the dwarf and giant data are in excellent agreement for [O/Fe], but the leveling–off or slight increase in [Mg/Fe], [Si/Fe], and [Ca/Fe] seems to be unique to the dwarf measurements. Unfortunately, the source of this discrepancy is not clear and may be related to analysis differences between dwarfs and giants.

Figure 12.— The [O/Mg] and [Cu/O] ratios are plotted as a function of [Fe/H]. The filled red circles, filled green circles, and filled blue boxes differentiate stars belonging to the (5.25,–3.02), (0,–12), and NGC 6553 populations.

In Figure 14 we compare the [O/Fe], [Mg/Fe], [Si/Fe], and [Ca/Fe] abundances between the bulge, thick disk, and thin disk. For stars with [Fe/H]–0.5, the bulge and thick disk stars exhibit similar abundance patterns for all four element ratios. However, we note that on average the bulge stars have [O/Fe] and [Mg/Fe] ratios that are slightly enhanced by 0.03 dex and [Si/Fe] and [Ca/Fe] ratios that are enhanced by 0.05 dex compared to similar metallicity thick disk stars. In contrast, the most metal–poor thin disk stars exhibit significantly lower [X/Fe] ratios for all of the –elements measured here. The bulge and thin disk stars with [Fe/H]0 are not strikingly different, but the star–to–star scatter, especially for [O/Fe], is significantly larger for the bulge giants. For the intermediate range of [Fe/H]–0.5 to 0, the bulge stars still exhibit significantly larger [/Fe] ratios than the thin disk and may remain enhanced to a higher [Fe/H] value than the thick disk.

Figure 13.— [X/Fe] ratios for the –elements O, Mg, Si, and Ca are plotted as a function of [Fe/H]. The filled red circles indicate abundances measured for this work (combining both fields and NGC 6553), the filled dark gray circles are abundances in bulge RGB and red clump stars from the literature, and the filled green triangles are abundances from bulge microlensed dwarfs (Bensby et al. 2013). The RGB and clump data are from: McWilliam & Rich (1994), Rich et al. (2005), Fulbright et al. (2007), Lecureur et al. (2007), Rich et al. (2007b), Meléndez et al.(2008), Alves–Brito et al. (2010), Ryde et al. (2010), Gonzalez et al. (2011), Hill et al. (2011), Johnson et al. (2011), Rich et al. (2012), García Pérez et al. (2013), and Johnson et al. (2013a,b).

The chemical similarities between especially the metal–poor bulge and thick disk found here have been documented in previous work (Meléndez et al. 2008; Alves–Brito et al. 2010; Bensby et al. 2010a; Ryde et al. 2010; Bensby et al. 2011; Gonzalez et al. 2011; Hill et al. 2011; Johnson et al. 2011, 2013). The apparent homogeneity between the most metal–poor bulge and thick disk stars lends credibility to the idea that the Galactic bulge formed in situ with the disk. However, there is not universal agreement in the literature that the metal–poor bulge and disk trends are identical. In particular, earlier work by Zoccali et al. (2006), Fulbright et al. (2007), and Lecureur et al. (2007) found that the bulge stars exhibited both larger [/Fe] ratios and remained enhanced to higher [Fe/H] values than the local thick disk16, which implies a more rapid formation timescale for the bulge. In contrast, purely differential analyses between thick disk and bulge giants (Meléndez et al. 2008; Alves–Brito et al. 2010; Gonzalez et al. 2011) find nearly identical [/Fe] versus [Fe/H] trends, at least for [Fe/H]–0.3. However, Bensby et al. (2013) noted in a similarly differential comparison of local thick disk dwarfs and bulge microlensed dwarfs that the inflection point at which [/Fe] declines may be 0.1–0.2 dex higher ([Fe/H]–0.3 to –0.2) in the bulge. While we compare bulge giants and thick disk dwarfs in Figure 14, our results are in agreement with Bensby et al. (2013). In particular, we find that the [Mg/Fe], [Si/Fe], and perhaps [O/Fe] ratios remain enhanced to a higher [Fe/H] value than those of the local thick disk17.

Figure 14.— A comparison of the O, Mg, Si, and Ca abundances for the bulge stars measured here (filled red circles) with those of the thick disk (open blue circles) and thin disk (open green boxes). The disk data are from: Bensby et al. (2003; 2005) and Reddy et al. (2006).

Finally, we note that combining the present data set with those available in the literature (e.g., see Figure 13) totals 10 bulge stars that have had [/Fe] measurements made from high resolution spectroscopy. Despite the large sample size, there is a paucity of stars with [/Fe] ratios that deviate significantly from the bulk trend. In agreement with work suggesting the Galactic bulge did not form predominantly from a build–up of merger events (e.g., Shen et al. 2010), we can effectively rule out significant contributions from the infall of objects with chemistry similar to those of many present–day dwarf galaxies (i.e., low [/Fe]; e.g., see Venn et al. 2004 and references therein). Additionally, as can be seen in Figure 11 (see also Gonzalez et al. 2011) the [X/Fe] abundance ratios of individual –elements for NGC 6553 stars are nearly identical to those of bulge field stars with similar [Fe/H]. Specifically, the average [X/Fe] values for NGC 6553 are: [O/Fe]=0.24 (one star), [Mg/Fe]=0.16 (=0.08), [Si/Fe]=0.17 (=0.10), and [Ca/Fe]=0.22 (=0.12), which compare well with the average abundances for nearby bulge field stars in the range [Fe/H]=–0.20 to 0.00: [O/Fe]=0.24 (=0.29), [Mg/Fe]=0.25 (=0.09), [Si/Fe]=0.15 (=0.08), and [Ca/Fe]=0.19 (=0.13). These values are in good agreement with past work that finds the cluster to be moderately –enhanced (Barbuy et al. 1999; Cohen et al. 1999; Coelho et al. 2001; Origlia et al. 2002; Meléndez et al. 2003; Alves–Brito et al. 2006). The similar [/Fe] abundances between the cluster and field stars suggests that NGC 6553 likely formed in situ with the bulge field population and is not a captured cluster.

4.2. The Light, Odd–Z Elements Sodium and Aluminum

In a similar fashion to the –elements, the light, odd–Z elements Na and Al provide clues of the processes that dominated the chemical enrichment of a stellar population. Furthermore, these elements are useful for “chemical tagging” analyses, and both the [Na/Fe] and [Al/Fe] ratios can vary significantly between stellar populations that have otherwise identical [/Fe] and [Fe/H] values. The large (0.5 dex) star–to–star [Na/Fe] and [Al/Fe] abundance variations present in metal–poor globular cluster but not halo/disk stars of the same metallicity are perhaps the most well–known example of this phenomenon (e.g., see reviews by Gratton et al. 2004; 2012 and references therein). While the production of Na and Al is dominated by hydrostatic helium, carbon, and neon burning in massive stars, the final yields are expected to grow significantly with increasing progenitor mass and metallicity (e.g., Woosley & Weaver 1995; Kobayashi et al. 2006; 2011). Intermediate mass (4–8 M) asymptotic giant branch (AGB) stars and the hydrogen–rich envelopes of massive stars can also produce significant amounts of Na and Al via the NeNa and MgAl proton–capture cycles (e.g., Decressin et al. 2007; de Mink et al. 2009; Ventura & D’Antona 2009; Karakas 2010). Since Na and Al are thought to result from similar production mechanisms, we expect their abundance patterns to reflect a comparable morphology.

While the bulge abundance patterns of [Na/Fe] and [Al/Fe] have not been investigated to the extent of the –elements, the combined literature sample now totals of order a few hundred stars. Interestingly, the agreement between studies regarding the [Na/Fe] and [Al/Fe] trends is worse than for the –elements. While all high–resolution analyses (McWilliam & Rich 1994; Fulbright et al. 2007; Lecureur et al. 2007; Alves–Brito et al. 2010; Bensby et al. 2010a, 2011; Johnson et al. 2012; Bensby et al. 2013) tend to agree that the average [Na/Fe] ratio rises with increasing metallicity, significant scatter is present at [Fe/H]–1 and [Fe/H]0. Similarly, there is general agreement that [Al/Fe] is enhanced in bulge stars at [Fe/H]–0.3. However, some studies find that [Al/Fe] remains enhanced at super–solar metallicities (McWilliam & Rich 1994; Fulbright et al. 2007; Lecureur et al. 2007; Alves–Brito et al. 2010) while others find a decline in [Al/Fe], similar to [/Fe] (Bensby et al. 2011, 2011; Johnson et al. 2012; Bensby et al. 2013). Additionally, there is general agreement that the [Na/Fe] and [Al/Fe] trends as a function of [Fe/H] are similar between the bulge and disk over a broad metallicity range, but differences could be present at the metal–poor and metal–rich ends of the bulge distribution. It is also not yet clear if any significant [Na/Fe] and [Al/Fe] abundance differences exist between different bulge sight lines.

Figure 11 shows our derived [Na/Fe] and [Al/Fe] abundances as a function of [Fe/H] for both fields and the possible NGC 6553 stars, and in Figure 15 we compare our results with those from previous work. For Na we find general agreement with literature values such that the average [Na/Fe] ratio rises with increasing [Fe/H]. However, we find only a small number of metal–rich stars with [Na/Fe]0.4 and do not reproduce the very large [Na/Fe] ratios of Lecureur et al. (2007). Additionally, we do not find significant evidence supporting large [Na/Fe] variations between the two bulge sight lines probed here. At [Fe/H]–0.5, the mean [Na/Fe] trend and star–to–star dispersion for our measured RGB stars is in good agreement with those of the microlensed bulge dwarfs (e.g., Bensby et al. 2013).

Figure 15.— A comparison plot of [Na/Fe] and [Al/Fe] ratios between the bulge RGB stars measured here (filled red circles), RGB and clump stars available in the literature (filled dark gray circles), and bulge microlensed dwarfs (filled green triangles). Additional dwarf and giant literature data are from: Johnson et al. (2007; 2008), Cohen et al. (2008; 2009), Epstein et al. (2010), and Johnson et al. (2012), in addition to those referenced in Figure 13.

The primary discrepancy between our work and some of the literature values occurs for stars with [Fe/H]–0.7, with the present work and Johnson et al. (2012) finding that the average Na trend decreases from [Na/Fe]0 at [Fe/H]=–0.5 to [Na/Fe]=–0.3 at [Fe/H]=–1.7. It is not immediately clear if the discrepancy, especially between the bulge RGB and dwarf data, is real or caused by analysis differences (e.g., NLTE, 3D, or spherical/plane–parallel effects between dwarfs and giants). The inclusion of NLTE corrections would minimize the differences at low metallicity between bulge RGB and dwarf stars, and also between bulge RGB and metal–poor thick disk dwarfs (see Figure 16), if the largely positive Na corrections for RGB stars from Gratton et al. (1999) were applied. However, more recent NLTE calculations (e.g., Lind et al. 2011) instead find that the sign of the Na correction is negative for the lines and atmospheric parameters used here. Similarly, the NLTE corrections for log (Fe I) appear to be positive (e.g., Lind et al. 2012; Bergemann et al. 2012) for most stars in our sample18, which would decrease the [Na/Fe] ratios. Further insights into this problem may be gained as more extensive NLTE calculations and 3D model atmosphere grids and codes become available.

Figure 16.— A plot of [Na/Fe] and [Al/Fe] ratios as a function of [Fe/H] for the bulge stars measured here (filled red circles), thick disk stars (open blue circles), and thin disk stars (open green boxes). The literature data are from the sources referenced in Figure 14.

When comparing the [Na/Fe] and [Al/Fe] trends in Figure 11, it is immediately clear that the two elements exhibit discrepant trends. While [Na/Fe] gradually rises with increasing [Fe/H], the [Al/Fe] trend is nearly indistinguishable from that of most –elements. In particular, we find in agreement with Bensby et al. (2010a; 2011; 2013) and Johnson et al. (2012) that [Al/Fe]0.3 in bulge stars until [Fe/H]–0.3 and then steadily declines at higher [Fe/H]. As mentioned previously, the decline in [Al/Fe] with increasing metallicity contrasts with other literature results that find [Al/Fe] remains enhanced even at [Fe/H]=0.5 (Fulbright et al. 2007; Lecureur et al. 2007; Alves–Brito et al. 2010). The data quality among the various studies is comparable, and it is not clear why the derived [Al/Fe] trends are in disagreement at high metallicity. We do note however that for cool, high metallicity stars the 6696 and especially 6698 Å Al I lines, as well as the continuum placement, can be affected by CN blending.

The discrepant [Na/Fe] and [Al/Fe] trends as a function of [Fe/H] are not limited to the bulge and may also be present in the disk, as can be seen in Figure 16. Despite nucleosynthesis models predicting similar production of Na and Al in massive stars (e.g., Woosley & Weaver 1995), Figure 16 shows that, at least in the metallicity range probed here, Al is over–produced relative to Na in both bulge and disk stars for [Fe/H]–0.3. The increased production of Na relative to Al in metal–rich stars, and especially in the bulge, suggests that metallicity dependent yields from massive stars vary more strongly for Na than Al. Contributions from intermediate mass AGB stars may also help explain the Na and Al trends since the AGB [Na/Fe] yields tend to increase at higher [Fe/H] while those of [Al/Fe] decline (e.g., Ventra & D’Antona 2009). Interestingly, we find that, unlike the case for [Na/Fe], the [Al/Fe] ratios are nearly indistinguishable between the bulge and thick disk at [Fe/H]0. Similarly, the [Al/Fe] ratios for bulge stars are identical to those in the thin disk at [Fe/H]0.

Given the similar behavior of [Al/Fe] to many of the –elements, in Figure 17 we provide a detailed comparison between [Al/Fe], [O/Fe], [Mg/Fe], [Si/Fe], and [Ca/Fe] for the bulge stars analyzed here. While the [O/Fe] trend is clearly different than that of [Al/Fe], there are no similarly strong discrepancies between [Al/Fe] and the other –elements. At [Fe/H]–0.8 both [Mg/Fe] and [Ca/Fe] are 0.10–0.15 dex enhanced compared to [Al/Fe], but those differences disappear at higher [Fe/H]. On the other hand, the [Si/Fe] and [Al/Fe] trends are essentially identical at all [Fe/H] with an average difference of 0.01 dex (=0.13 dex).

Figure 17.— [Al/Fe] ratios (open blue triangles) for all bulge and NGC 6553 RGB stars analyzed here are compared to the abundance trends of the –elements O, Mg, Si, and Ca (open red boxes).

Examining the NGC 6553 stars in Figure 11 shows that Na, and to a lesser extent Al, exhibit larger star–to–star [Na/Fe] and [Al/Fe] variations than similar metallicity field stars. In particular, the average Na and Al abundances for the cluster stars are [Na/Fe]=0.16 (=0.20) and [Al/Fe]=0.17 (=0.13), which can be compared to the similar metallicity fields stars having [Na/Fe]=0.03 (=0.11) and [Al/Fe]=0.16 (=0.10), respectively. The larger [Na/Fe] abundance and dispersion values for the cluster stars suggests NGC 6553 experienced some degree of self–enrichment. However, unlike low metallicity globular clusters, NGC 6553 does not exhibit a strong Na–Al correlation. This is in agreement with the observed trend that the Na–Al correlation is more mild and [Al/Fe] dispersions smaller in metal–rich as opposed to metal–poor globular clusters (e.g., Carretta et al. 2009; O’Connell et al. 2011; Cordero et al. 2014). Unfortunately, the 6300 Å telluric oxygen emission feature combined with NGC 6553’s relatively low radial velocity prohibited us from obtaining an [O/Fe] abundance for more than one star in NGC 6553. Therefore, we cannot comment further on the existence or extension of the likely O–Na correlation. Finally, we note that our mean [Na/Fe] and [Al/Fe] values and abundance dispersions are in excellent agreement with those found by Alves–Brito et al. (2006), but are considerably lower than the values (based on two stars) of Barbuy et al. (1999).

4.3. The Fe–Peak Elements: Chromium, Cobalt, Nickel, and Copper

Unlike the lighter elements, the abundance patterns of Fe–peak elements in the Galactic bulge are not well–explored. The production of Fe–peak elements occurs through a variety of processes in the late stages of massive star evolution, the resulting core collapse SNe, and also in Type Ia SNe. The Fe–peak abundance patterns can also be useful indicators of a stellar population’s IMF, with odd–Z elements in particular providing some diagnostic power (e.g., Nomoto et al. 2013). Some initial work on the bulge Fe–peak abundance distribution was included in McWilliam & Rich (1994), which found [V/Fe], [Cr/Fe], and [Ni/Fe] ratios near solar and a possible enhancement in [Co/Fe] and [Sc/Fe]. More recent work analyzing the Fe–peak abundance trends in the bulge has come from microlensed dwarf studies (Cohen et al. 2008; Johnson et al. 2008; Cohen et al. 2009; Bensby et al. 2010a; Epstein et al. 2010; Bensby et al. 2011; Bensby et al. 2013). The bulge [Mn/Fe] trend in RGB stars has also been investigated recently by Barbuy et al. (2013). The results of the these analyses indicate that the bulge Fe–peak trends are similar to that of the local disk, except that the bulge may have different [Mn/O] ratios than the thick disk for a given [O/H] value.

The general [X/Fe] versus [Fe/H] abundance trends derived here are shown in Figure 11. From these data we find that: (1) Cr is the element that most closely tracks Fe with [Cr/Fe]=0.00 (=0.11), (2) [Co/Fe] exhibits low level variations as a function of [Fe/H] but is generally enhanced with [Co/Fe]=0.14 (=0.11), (3) [Ni/Fe] shows similar variations to [Co/Fe] but at a much smaller amplitude and is slightly enhanced with [Ni/Fe]=0.09 (=0.06), (4) the Cu abundance increases monotonically from [Cu/Fe]=–0.84 in the most metal–poor star to [Cu/Fe]0.40 in the most metal–rich stars, and (5) there are no significant Fe–peak abundance variations between NGC 6553 stars and the field stars.

Although the exact nature of Cu nucleosynthesis is complex (e.g., see Mishenina et al. 2002 and references therein), the significant secondary (i.e., metallicity–dependent) production of Cu (and also Na) is evident in Figure 11. Additionally, Figure 12 shows that despite the larger measurement errors in both O and Cu abundances, the [Cu/O] ratio is strongly correlated with [Fe/H]. This trend has been noted previously and is prevalent in stellar populations with different star formation histories, such as the local disk and Sagittarius Dwarf Galaxy (e.g., McWilliam et al. 2013). The [Cu/O] trend is taken as evidence that a significant portion of Cu is synthesized in massive stars, perhaps via the weak s–process (e.g., Sneden et al. 1991). However, some component of Cu may also be produced by Type Ia SNe (Matteucci et al. 1993).

In Figure 18 we compare our derived Fe–peak abundance trends with those in the literature. For Cr there is general agreement between the bulge RGB stars analyzed here and the literature microlensed dwarf data. However, the small number of bulge literature data points for Co and Cu makes a direct comparison difficult. The [Ni/Fe] comparison also shows excellent agreement overall, but the RGB stars appear systematically enhanced by 0.1 dex in the range [Fe/H]=–0.3 to 0.1. Note also the similarly small star–to–star dispersion in especially [Ni/Fe] between the RGB and dwarf data.

Figure 18.— Plots comparing the [Cr/Fe], [Co/Fe], [Ni/Fe], and [Cu/Fe] abundances of the bulge stars measured here (filled red circles) with literature measurements of bulge microlensed dwarfs (filled green triangles) and field RGB/red clump stars (filled dark gray circles). The literature data are from the sources referenced in Figures 13 and 15.

A comparison between the bulge Fe–peak abundance trends and those of the thin/thick disk is shown in Figure 19. Interestingly, at least for [Fe/H]–1.5, the [Cr/Fe] distribution is seemingly independent of formation environment with the bulge, thick disk, and thin disk stars all having [Cr/Fe]0. For [Co/Fe], [Ni/Fe], and [Cu/Fe], there is significant overlap between the bulge and thick disk trends at [Fe/H]–0.5. At higher [Fe/H], the bulge may be enhanced in all three elements relative to both the thick and thin disks. This is especially evident in the Figure 19 panel showing [Ni/Fe] versus [Fe/H]; the low star–to–star scatter in [Ni/Fe] for all three populations highlights the possible composition difference between the local disk and bulge from [Fe/H]–0.4 to 0.2. While the strong rise in [Cu/Fe] with metallicity is, as mentioned previously, a common feature in many different stellar populations, the bulge stars at [Fe/H]–0.3 appear to extend to higher abundances than the local disk. However, the increased measurement uncertainty of Cu and paucity of disk [Cu/Fe] ratios at [Fe/H]0 prevents us from undertaking a more comprehensive analysis.

Figure 19.— Plots comparing the [Cr/Fe], [Co/Fe], [Ni/Fe], and [Cu/Fe] abundances of the bulge stars measured here (filled red circles) with literature data for the thick disk (open blue circles) and thin disk (open green boxes). The literature data are from the sources referenced in Figure 14.

4.4. Comparing Composition Data to Bulge Chemical Enrichment Models

Accurately modeling the chemical enrichment history of a stellar system requires solving for a variety of free parameters that may include the IMF, star formation rate, star formation efficiency, supernova/hypernova ratio19, inflow/outflow rate, binary fraction, stellar evolution time scales, mass loss rates, and stellar yields. While not all of the required input parameters are yet well–defined based on observed data, chemical enrichment models are effective tools for examining and interpreting chemical composition data. Therefore, in Figures 2021 we compare our derived abundance trends with those predicted by chemical enrichment models in which parameters such as the IMF, binary fraction, supernova/hypernova ratio, and outflow rate are varied.

Figure 20.— Chemical abundance trends are plotted as a function of [Fe/H] and compared to various chemical enrichment models. The solid black, blue, and green lines represent the baseline models from Kobayashi et al. (2006; 2011) for the Galactic bulge, thick disk, and thin disk, respectively. The dashed cyan and magenta lines illustrate how the bulge model changes if the hypernova fraction is 0 and 1, respectively, for masses 20M. Note that [Ni/Fe] in particular suffers from over–production from Type Ia SNe at [Fe/H]–1. Some other elements (e.g., Si) may also be better fit if systematic offsets were applied.

The baseline Galactic bulge model shown in Figures 2021 is from Kobayashi et al. (2006; 2011) and is designed to reproduce the metallicity distribution in Baade’s Window from Zoccali et al. (2008), assumes a Kroupa (2008) IMF, and assumes a star formation time scale of 3 Gyr (see Kobayashi et al. 2011; their Table 1 and Section 2.4 for more detail regarding model input parameters). In general, the baseline model does a reasonable job of reproducing the observed abundance trends of all abundance ratios, except [Na/Fe] and [Al/Fe]. All of the models shown in Figures 2021 predict large over–abundances of both [Na/Fe] and [Al/Fe] that are not observed, suggesting the massive star yields of both elements may be too high20. However, as can be seen in Figure 20, the enhanced Fe production from hypernovae (HNe) decreases the [Na/Fe] and [Al/Fe] yields and brings the baseline bulge model into better agreement with the light element data. The addition of HNe also provides better agreement between the models and observations for the Fe–peak elements, with a trade off of [/Fe] ratios that may be slightly too low. In contrast, Figure 20 also shows that a paucity of HNe generally leads to [X/Fe] ratios that are too high. It seems that a significant fraction of HNe are required to accurately reproduce the observed bulge abundance trends. Unfortunately, the HN fraction is best constrained at [Fe/H]–1, where data are scarce.

Figure 21.— Similar to Figure 20, the solid black line is our adopted baseline bulge model from Kobayashi et al. (2006; 2011). The solid blue line is the baseline bulge model with a top–heavy (flatter) IMF, and the dashed cyan line is the same model but with outflow and an increased Type Ia SN rate (10). The solid green line is the baseline bulge model with a steep IMF. The solid magenta line is the baseline bulge model with the IMF truncated at an upper mass limit of 40 M.

In Figure 21 we examine how changes in the IMF could affect the expected abundance trends. Compared to the Kroupa (2008) IMF adopted in our baseline bulge model, a steep IMF (x=1.6) is completely ruled out by the data. Additionally, adopting a Kroupa (2008) IMF that truncates at 40 M, and thus ignores contributions from the most massive stars, is inconsistent with the [Cu/Fe] abundances, and to a lesser extent those of [Co/Fe]. While a flatter, top–heavy IMF (x=0.3) alone leads to [X/Fe] ratios that are too high for nearly every element, a reduction in the yields from outflow and/or slow star–formation combined with a high Type Ia SN rate, artificially enhanced with a ten times larger binary fraction, could bring such a model into agreement with the data. However, bulge formation models with slow star–formation are likely unrealistic, and the observed [Co/Fe] and [Cu/Fe] data appear to rule out these models. Based on the present data it does not appear that the bulge required a uniquely “non–standard” IMF to reach its present–day composition (but see also Ballero et al. 2007, for example).

Finally, in Figure 20 we also compare the measured bulge abundance trends with our adopted baseline model and similar models representing the composition distributions of the local thick disk and thin disk. Comparing the three predicted trends indicates that in the range –0.8[Fe/H]–0.3 the bulge [/Fe] ratios should be similar or modestly enhanced and remain enhanced to higher [Fe/H] than the thick disk. Similarly, at [Fe/H]0 the bulge and thin disk should exhibit similar, if not identical, [/Fe] ratios. Both of these predictions match our observations (see Section 4.1). The predicted enhancements in the bulge for [Na/Fe] and [Al/Fe] compared to the local disk are not supported by observations, but this could be related to the previously mentioned possible over–production issues of the adopted stellar yields. However, in addition to Na and Al, Figure 20 shows that Co and Cu may also exhibit some discriminating power between the bulge and local disk populations. In particular, the data support bulge stars with [Fe/H]–0.5 having [Co/Fe] and [Cu/Fe] ratios that are higher than the local disk. Therefore, the data and models presented here provide some supporting evidence that the bulge experienced a different chemical enrichment path than the thick disk.

5. Summary

We have measured radial velocities and chemical abundances of O, Na, Mg, Al, Si, Ca, Cr, Fe, Co, Ni, and Cu in a sample of 156 RGB stars located in Galactic bulge fields centered near (l,b)=(5.25,–3.02) and (0,–12). The (5.25,–3.02) also includes 12 stars identified as likely members of the bulge globular cluster NGC 6553, based on their radial velocity and [Fe/H] values. The results are based on high resolution archival spectra obtained with the FLAMES–GIRAFFE instrument, and originally used to derive [Fe/H] and [/Fe] abundances in Zoccali et al. (2008) and Gonzalez et al. (2011). We culled the original target list and selected only those stars with co–added S/N70 that also lack strong TiO bands. The abundance analysis was carried–out using standard EW and spectrum synthesis techniques.

Our derived heliocentric radial velocity distributions for both fields are in good agreement with past surveys (BRAVA, GIBS, and APOGEE) covering nearby fields. We do not confirm the existence of a significant population of high velocity stars noted by Nidever et al. (2012) and Babusiaux et al. (2014). However, our targeted fields are farther away from the plane than most of those in which Nidever et al. (2012) and Babusiaux et al. (2014) observe the cold, high velocity stars. For both fields analyzed here we also find that the velocity dispersion monotonically decreases with increasing [Fe/H]. This is not unexpected for the outer bulge field at (0,–12), but the similar trend in the (5.25,–3.02) field appears to contradict the findings of Babusiaux et al. (2010; 2014) that the velocity dispersion of bulge stars with [Fe/H]0 increases at lower Galactic latitude. The reason for this discrepancy is not clear, but we note that previous analyses finding increased velocity dispersion at low Galactic latitude for metal–rich stars have all focused on minor–axis fields. The inner bulge field included here is several degrees off–axis.

The composition data reconfirm the already well–documented metallicity gradient in the bulge. Similarly, we find good agreement between our derived [Mg/Fe], [Si/Fe], and [Ca/Fe] abundances and those of Gonzalez et al. (2011). Additionally, we confirm that there are no significant field–to–field [/Fe] abundance variations among various bulge sight lines. Our new –element measurements also reinforce the previously held notion (e.g., McWilliam et al. 2008) that the decline in [O/Mg] with increasing metallicity is likely the result of metallicity dependent yields from massive stars. While we find that the bulge and thick disk exhibit nearly identical [/Fe] ratios at [Fe/H]–0.5, the bulge stars appear to remain enhanced in [/Fe] by up to 0.1–0.2 dex higher in [Fe/H] than the local thick disk. The bulge [/Fe] ratios at [Fe/H]0 are well–matched to the local thin disk trends. These results are in agreement with recent differential abundance analyses of microlensed bulge dwarfs (Bensby et al. 2013), and suggest the bulge experienced faster enrichment than the local thick disk. However, similar differential analyses comparing bulge and thick disk giants find no significant differences between the two populations (Meléndez et al. 2008; Alves–Brito et al. 2010; Gonzalez et al. 2011).

Combining the new data set of [/Fe] abundances with those available in the literature now totals several hundred stars. However, the combined data set does not reveal any significant population with “anomalous” chemistry, such as the low [/Fe] ratios reminiscent of many present–day dwarf galaxy stars. Therefore, we can effectively rule out these types of objects as major contributors to any portion of the present–day Galactic bulge field population. This further supports the idea that the Galactic bulge is not a merger–built system. Similarly, the [/Fe] ratios of the NGC 6553 stars are identical to those of similar metallicity field stars. This suggests NGC 6553 formed in situ with the bulge and is not a captured system.

With regard to the light, odd–Z elements, we find that Na and Al exhibit discrepant trends as a function of metallicity. In particular, bulge stars exhibit a steady increase in [Na/Fe] with increasing [Fe/H], but the [Al/Fe] trend almost exactly matches that of the –elements (except oxygen). While we do not find any significant field–to–field variations in either [Na/Fe] or [Al/Fe], our results indicate that the bulge and thick disk have different [Na/Fe] abundances at [Fe/H]–0.5 but similar [Al/Fe]. Interestingly, the “–like” behavior of [Al/Fe] contrasts with several previous bulge studies that found [Al/Fe] was enhanced up to [Fe/H]=0.5. Instead, our results are in agreement with the abundance patterns of microlensed bulge dwarfs (e.g., Bensby et al. 2013). The discrepant behavior of Na and Al suggests metallicity dependent yields from massive stars, and perhaps intermediate mass stars, leads to significantly more production of Na than Al at high metallicity. We also find that the NGC 6553 stars have nearly identical [Al/Fe] ratios as similar metallicity field stars, but both the average [Na/Fe] abundance and star–to–star dispersion of cluster stars are higher. This suggests NGC 6553 experienced some light element self–enrichment, which is typical for globular clusters.

The abundance trends of the Fe–peak elements are distinctly different: (1) the average [Cr/Fe] ratio is essentially solar over the full range in [Fe/H] and shows no variations over the metallicity range probed here, (2) both [Co/Fe] and [Ni/Fe] are enhanced by 0.1 dex at nearly all [Fe/H] and exhibit some low amplitude, metallicity–dependent variations, and (3) [Cu/Fe] exhibits a large increase from the metal–poor to metal–rich end of the distribution. In a similar fashion to [Na/Fe], the strong secondary (metallicity–dependent) production of Cu is evident in bulge stars, and the correlation between [Cu/O] and [Fe/H] suggests massive stars produce significant portions of Cu. However, Cu production from another source (e.g., Type Ia SNe) seems required to explain the high [Cu/Fe] abundances at super–solar metallicities. Interestingly, at [Fe/H]–2 the [Cr/Fe] trend is identical between the bulge, thick disk, and thin disk, but the heavier Fe–peak [X/Fe] ratios appear to all be enhanced in the bulge relative to the local disk. Additionally, the NGC 6553 Fe–peak abundance trends are in agreement with similar metallicity field stars.

Despite predicting [Na/Fe], [Al/Fe], and [Ni/Fe] ratios that are too high, our adopted baseline bulge chemical enrichment model from Kobayashi et al. (2006; 2011) does a reasonable job fitting the abundance trends of the and other Fe–peak elements. However, better agreement between the data and model is found when a significant fraction of HNe, which produce more Fe, are included. Unfortunately, setting the HN fraction is best constrained using abundance patterns at [Fe/H]–1, where the bulge data are sparse. While a Kroupa (2008) IMF provides a reasonable fit to the observed abundance trends, a top–heavy IMF including strong outflow cannot be ruled out. In contrast, the Fe–peak abundance data strongly rule out IMFs that are truncated to exclude the contributions of stars 40 M, steep IMFs (e.g., x=1.6), and top–heavy IMFs that do not include outflow. We conclude that the bulge likely does not require a particularly unusual IMF to explain its present–day abundance patterns, and that its enhanced abundances for several and Fe–peak elements match model predictions in which the bulge experienced a different enrichment history than the local disk.

We thank the anonymous referee for a careful reading of the manuscript and helpful comments that lead to improvement of the manuscript. This research has made use of NASA’s Astrophysics Data System Bibliographic Services. This publication makes use of data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. CIJ gratefully acknowledges support from the Clay Fellowship, administered by the Smithsonian Astrophysical Observatory. RMR acknowledges support from NSF-AST-1212095. AK acknowledges the Deutsche Forschungsgemeinschaft for funding from Emmy–Noether grant Ko 4161/1.
ID21 2MASS RA (J2000) DEC (J2000) V V–I J H K T log(g) [Fe/H] vt RV RV22
(degrees) (degrees) (mag.) (mag.) (mag.) (mag.) (mag.) (K) (cgs) (km s) (km s) (km s)
(l,b)=(5.25,–3.02)
119799C4 180810262547354 272.042708 25.793167 12.856 11.986 11.704 4400 2.05 0.39 1.90 57.96 0.11
129499C4 180816172543192 272.067375 25.721972 12.596 11.836 11.608 4900 2.75 0.13 2.10 27.10 0.04
176772C5 180809362556168 272.039417 25.938028 11.710 10.794 10.483 4225 2.15 0.06 1.90 37.82 0.08
181349C5 180804652554356 272.019292 25.909917 12.924 12.253 11.906 4775 2.75 0.26 2.10 38.51 0.04
183783C5 180809122553420 272.037958 25.895028 12.876 12.119 11.848 4700 2.75 0.10 1.75 97.45 0.01
184088C5 180807262553356 272.030250 25.893222 12.789 11.869 11.640 4500 2.50 0.47 1.80 173.45 0.22
184618C5 180804142553236 272.017208 25.889917 12.732 11.946 11.690 4600 2.15 0.59 1.55 247.54 0.34
185169C5 180805232553111 272.021708 25.886389 12.367 11.587 11.360 4775 2.50 0.32 1.85 84.03 0.25
185357C5 180817782553068 272.074083 25.885250 12.933 12.152 11.841 4550 2.60 0.28 1.90 60.42 0.16
185541C5 180809782553024 272.040708 25.884000 12.865 12.106 11.865 4750 3.10 0.34 1.80 12.63 0.02
187067C5 272.056958 25.874083 4850 2.30 0.95 1.35 115.30 0.31
193190C5 180812652550046 272.052625 25.834556 12.248 11.461 11.172 4675 2.75 0.24 1.85 7.60 0.34
197366C5 180814032548290 272.058417 25.808028 12.767 11.959 11.700 4550 2.90 0.32 1.75 73.59 0.03
215681C6 180844462557568 272.185292 25.965778 12.938 12.160 11.944 4800 2.75 0.31 1.95 54.15 0.20
216922C6 180845142557293 272.188083 25.958139 12.598 11.652 11.472 4400 2.10 0.20 2.20 135.08 0.16
218198C6 180821202557000 272.088333 25.950028 12.506 11.579 11.271 4225 1.95 0.32 2.10 158.44 0.02
219909C6 180832622556215 272.135917 25.939306 12.618 11.745 11.520 4650 2.25 0.73 1.75 62.33 0.02
221537C6 180850022555460 272.208375 25.929417 12.881 12.057 11.835 4750 2.75 0.29 1.60 60.86 0.04
223113C6 180840722555101 272.169667 25.919472 12.833 12.074 11.845 4800 2.55 0.15 1.75 25.52 0.32
223343C6 180821272555048 272.088625 25.918028 12.834 12.008 11.774 4525 2.65 0.15 1.45 61.02 0.44
223621C6 180852632554587 272.219208 25.916306 12.677 11.938 11.749 5150 3.75 0.31 1.35 88.54 0.18
223722C6 272.183958 25.915722 4525 1.35 0.75 1.75 114.88 0.18
224206C6 180846902554462 272.195417 25.912833 12.857 12.069 11.853 4800 3.00 0.46 1.65 142.44 0.13
224866C6 180833922554311 272.141417 25.908667 12.489 11.604 11.333 4350 2.10 0.22 1.75 96.89 0.01
224951C6 180851282554294 272.213667 25.908167 12.660 11.877 11.583 4650 2.75 0.17 1.80 87.33 0.32
225531C6 180828782554167 272.119958 25.904639 12.613 11.670 11.452 4425 1.75 0.73 1.80 135.60 0.20
226450C6 180834642553563 272.144333 25.898972 12.918 12.061 11.862 4500 2.50 0.27 1.80 22.47 0.13
226850C6 180823402553475 272.097542 25.896500 12.548 11.753 11.506 4800 2.15 0.85 1.85 21.21 0.03
227867C6 180841602553245 272.173333 25.890167 12.699 11.894 11.634 4850 3.15 0.09 1.60 62.08 0.14
228466C6 180829172553116 272.121542 25.886583 12.665 11.762 11.548 4700 1.55 0.98 2.15 24.42 0.12
229507C6 180855302552488 272.230375 25.880222 12.490 11.622 11.408 4650 1.75 0.29 1.60 71.66 0.13
230424C6 180850242552281 272.209292 25.874500 12.319 11.344 11.098 4400 2.10 0.84 1.75 0.90 0.06
230483C6 180825852552269 272.107708 25.874139 12.643 11.725 11.465 4450 2.10 0.65 1.60 19.64 0.09
231379C6 180822902552064 272.095417 25.868444 12.787 11.906 11.675 4500 2.30 0.30 1.80 36.71 0.36
231618C6 180844242552014 272.184250 25.867028 12.540 11.776 12.711 4550 2.60 0.04 1.65 22.46 0.10
232493C6 272.090500 25.861694 4625 2.00 1.14 1.20 160.24 0.05
233121C6 180830492551279 272.127000 25.857722 12.614 11.684 11.468 4500 1.90 0.50 1.90 21.69 0.12
233560C6 180820992551174 272.087417 25.854833 12.900 11.970 11.747 4375 2.10 0.21 1.85 170.65 0.27
233708C6 180841232551146 272.171750 25.853972 12.393 11.437 11.064 4500 2.50 0.14 1.70 57.52 0.07
240059C6 180831332548449 272.130542 25.812472 12.410 11.452 11.258 4400 1.65 0.69 1.60 124.52 0.52
240083C6 180824592548444 272.102500 25.812306 12.255 11.395 11.130 4425 1.65 0.58 1.70 33.61 0.10
259050C7 180913762557523 272.307250 25.964528 12.231 11.277 11.012 4275 1.55 0.65 1.65 124.20 0.16
259377C7 180859272557448 272.246958 25.962472 12.735 11.939 11.723 5000 3.25 0.29 1.85 25.50 0.18
260308C7 180908582557243 272.285792 25.956750 12.659 11.825 11.611 4600 2.50 0.16 1.90 43.66 0.10
262018C7 180914052556473 272.308583 25.946472 12.937 12.109 11.925 4700 2.75 0.08 2.00 16.13 0.31
266442C7 180901182555154 272.254917 25.920917 12.855 11.908 11.678 4350 2.00 0.42 1.60 62.73 0.02
270316C7 180916282553550 272.317917 25.898667 12.675 11.766 11.538 4400 1.90 0.37 1.65 1.74 0.47
275181C7 180910222552134 272.292583 25.870389 12.478 11.578 11.366 4375 1.70 0.59 1.75 253.05 0.05
277490C7 180906292551235 272.276208 25.856500 12.814 11.861 11.606 4250 1.55 0.56 1.65 76.93 0.38
278419C7 180859592551037 272.248292 25.850972 12.777 12.019 11.714 4725 2.75 0.20 1.55 68.67 0.05
282804C7 180916532549261 272.318917 25.823889 12.468 11.614 11.351 4575 2.50 0.10 1.70 135.83 0.04
286252C7 180900522548070 272.252167 25.801861 12.579 11.777 11.563 4700 3.10 0.43 1.40 76.01 0.06
45512C2 180904732545129 272.269792 25.753556 12.623 11.821 11.552 4500 2.15 0.27 1.75 96.20 0.03
47188C2 180906882544307 272.278667 25.741833 12.258 11.324 11.057 4350 1.70 0.52 1.85 128.90 0.34
77186C3 180828352547382 272.118125 25.793889 12.942 12.153 11.890 4750 3.15 0.41 1.85 5.75 0.10
77707C3 180836372547256 272.151542 25.790306 12.479 11.623 11.448 4825 3.15 0.44 1.65 4.81 0.14
80582C3 180851892546213 272.216208 25.772556 12.765 11.945 11.686 5100 3.55 0.07 1.60 39.87 0.37
81644C3 180854542545569 272.227208 25.765806 12.631 11.753 11.430 4600 2.85 0.21 1.70 52.17 0.11
82227C3 180841412545439 272.172500 25.762139 12.676 11.828 11.642 4900 3.25 0.25 1.65 14.02 0.18
83531C3 180843812545145 272.182500 25.754000 12.535 11.692 11.431 4725 3.25 0.45 1.80 26.11 0.22
84255C3 180821502544574 272.089583 25.749278 12.583 11.668 11.423 4350 1.50 0.68 1.45 221.43 0.03
86757C3 180830112544007 272.125458 25.733500 12.467 11.541 11.300 4400 2.20 0.02 1.90 72.12 0.01
88522C3 180835822543214 272.149250 25.722611 12.463 11.613 11.383 4700 3.00 0.39 1.90 5.87 0.23
NGC 6553
225847C6 180855742554094 272.232292 25.902583 12.895 12.000 11.757 4500 2.25 0.11 1.45 3.26 0.24
227379C6 180828092553354 272.117083 25.893167 12.947 12.011 11.779 4350 2.25 0.05 1.55 2.50 0.13
228407C6 180853542553128 272.223042 25.886889 12.718 11.781 11.547 4415 1.95 0.22 1.75 0.21 0.16
230208C6 180854302552332 272.226208 25.875917 12.846 12.020 11.842 4900 3.05 0.00 1.40 1.43 0.46
239284C6 180832262549034 272.134417 25.817611 12.438 11.512 11.232 4300 2.00 0.14 1.80 9.48 0.43
265795C7 180920182555283 272.334125 25.924556 12.757 11.959 11.788 4600 2.50 0.07 1.55 12.45 0.03
268360C7 180913792554357 272.308000 25.909833 12.075 11.278 10.987 4750 3.15 0.08 1.35 3.63 0.71
268493C7 180908372554327 272.284917 25.909056 12.647 11.734 11.552 4500 2.30 0.20 1.90 1.86 0.11
271021C7 180906612553404 272.277667 25.894472 12.544 11.775 11.591 4750 2.50 0.06 1.65 0.22 0.06
271400C7 180913202553328 272.305083 25.892389 12.418 11.519 11.325 4525 2.50 0.19 1.90 4.03 0.22
77182C3 180849662547388 272.206875 25.794056 12.806 11.927 11.670 4700 2.50 0.09 1.75 3.44 0.17
85597C3 180823692544268 272.098708 25.740778 12.723 11.896 11.655 4400 2.25 0.15 1.55 0.24 0.17
(l,b)=(0,–12)
1156C2 183556013433364 278.983417 34.560111 15.93 1.534 13.092 12.380 12.280 4300 1.80 0.46 1.45 72.84 0.01
1407C3 183450533433226 278.710542 34.556306 15.76 1.155 13.490 13.008 12.995 5125 2.40 0.62 1.45 141.73 0.09
1491C7 183533523449238 278.889708 34.823306 15.80 1.419 13.184 12.562 12.461 4700 2.25 0.32 1.85 31.46 0.62
1554C7 183540373449160 278.918208 34.821139 15.11 1.166 12.887 12.390 12.333 5050 2.20 0.67 1.50 7.77 0.18
166C3 183445393436070 278.689125 34.602000 15.90 1.267 13.466 12.945 12.873 4850 2.20 0.83 1.30 54.10 0.12
1754C3 183448923432327 278.703833 34.542417 15.43 1.301 12.955 12.409 12.298 4900 2.50 0.25 1.65 18.44 0.06
1814C1 183611563432060 279.048250 34.535000 15.85 1.427 13.229 12.567 12.421 4650 2.60 0.30 1.50 57.13 0.01
1876C2 183549373431562 278.955750 34.532250 15.86 1.106 13.679 13.187 13.147 5000 2.50 1.03 1.10 202.57 0.03
1917C1 183612483431515 279.052083 34.530972 16.22 1.426 13.562 12.975 12.850 4675 2.70 0.18 1.50 53.87 0.01
1918C1 183607523431511 279.031375 34.530833 15.83 1.248 13.428 12.875 12.791 4900 2.35 0.44 1.60 42.10 0.05
201583C3 183451393431404 278.714125 34.527889 15.79 1.258 13.411 12.848 12.813 4800 1.90 1.11 0.90 59.07 0.08
2110C7 183536093447564 278.900417 34.799056 15.59 1.405 13.013 12.367 12.338 4675 3.00 0.02 1.30 6.53 0.10
2178C7 183541563447486 278.923167 34.796889 15.77 1.371 13.195 12.619 12.505 4700 2.80 0.00 1.80 12.34 0.10
2200C3 183457783431359 278.740708 34.526639 16.15 1.315 13.588 13.073 13.014 4975 2.70 0.10 1.80 6.43 0.11
2220C7 183530723447426 278.878000 34.795194 15.62 1.233 13.230 12.739 12.685 5150 2.75 0.17 1.60 81.33 0.06
222C3 183514883436002 278.812083 34.600111 15.50 1.295 12.984 12.449 12.329 4950 2.60 0.43 1.90 90.22 0.53
2335C2 183558843430461 278.995208 34.512778 15.81 1.270 13.385 12.778 12.718 4750 2.05 0.90 1.50 119.07 0.12
2407C2 183601983430369 279.008292 34.510222 15.49 1.173 13.232 12.653 12.605 4975 2.30 0.64 1.70 145.30 0.15
2422C7 183548053447154 278.950250 34.787667 15.89 1.481 13.176 12.551 12.476 4400 2.15 0.17 1.90 21.36 0.07
2470C3 183521513431018 278.839667 34.517167 16.25 1.375 13.661 13.048 12.952 4800 3.10 0.00 1.35 89.04 0.18
2502C3 183454423430566 278.726750 34.515694 15.97 1.330 13.424 12.833 12.762 4750 2.25 0.62 1.40 255.23 0.44
2532C6 183513183446482 278.804917 34.780139 15.71 1.345 13.123 12.610 12.467 4750 2.50 0.17 1.90 42.27 0.16
2580C6 183515243446412 278.813500 34.778167 16.06 1.566 13.184 12.595 12.406 4575 2.85 0.30 2.00 16.54 0.24
2769C3 183457523430213 278.739667 34.505917 15.78 1.295 13.311 12.797 12.657 4900 3.15 0.06 1.25 51.43 0.48
2772C7 183548073446257 278.950292 34.773833 15.10 1.258 12.692 12.137 12.072 4900 2.75 0.25 1.65 79.69 0.09
2812C8 183609273446233 279.038667 34.773194 15.63 1.458 12.948 12.274 12.101 4475 2.00 0.65 1.55 90.18 0.36
2947C3 183500593429557 278.752458 34.498778 16.02 1.204 13.660 13.157 13.120 5150 2.60 0.50 1.95 98.82 0.15
2948C7 183524523446010 278.852208 34.767000 15.53 1.679 12.485 11.838 11.684 4275 2.00 0.12 1.65 22.63 0.13
3018C3 183503583429453 278.764958 34.495889 15.49 1.404 12.836 12.190 12.033 4650 2.40 0.54 1.45 64.70 0.13
3035C7 183551353445453 278.964000 34.762639 15.68 1.191 13.412 12.868 12.808 5050 2.50 0.44 1.45 8.23 0.07
3091C8 183608863445437 279.036958 34.762167 15.32 1.236 12.938 12.390 12.306 4900 3.00 0.50 1.45 76.91 0.13
3101C7 183537133445344 278.904708 34.759611 15.86 1.233 13.523 12.939 12.880 4950 2.65 0.40 1.70 67.03 0.02
3142C3 183509483429282 278.789500 34.491139 15.95 1.287 13.481 12.949 12.831 4900 2.65 0.22 1.75 18.58 0.02
3161C3 183519423429266 278.830917 34.490722 15.46 1.306 12.968 12.367 12.313 5100 3.20 0.15 1.70 29.96 0.08
3191C7 183555603445217 278.981708 34.756083 15.40 1.243 13.051 12.479 12.431 4950 2.50 0.37 1.60 64.72 0.09
3201C6 183503013445053 278.762583 34.751528 15.47 1.211 13.078 12.624 12.467 5200 3.50 0.04 1.35 71.88 0.11
3238C6 183516983444595 278.820792 34.749889 15.73 1.322 13.174 12.632 12.523 4900 3.00 0.22 1.35 65.42 0.19
3267C3 183522063429112 278.841917 34.486444 15.07 1.349 12.525 11.923 11.806 4700 2.65 0.04 1.40 119.57 0.22
3515C3 183450723428380 278.711000 34.477200 12.658 12.099 11.965 4750 2.70 0.01 1.80 14.32 0.08
3558C6 183512933444170 278.803875 34.738139 16.03 1.341 13.526 12.901 12.780 4800 2.75 0.15 1.75 21.61 0.25
3690C7 183528343444085 278.868125 34.735750 14.97 1.208 12.606 12.088 12.009 4800 1.45 1.46 1.50 40.71 0.07
3711C7 183541053444059 278.921042 34.735028 15.71 1.097 13.526 13.038 13.004 5350 3.70 0.49 1.25 3.48 0.21
3733C3 183517993428093 278.824958 34.469250 15.10 1.423 12.436 11.795 11.722 4750 2.60 0.18 1.70 45.51 0.01
3796C6 183458253443440 278.742708 34.728944 15.53 1.454 12.839 12.133 12.041 4500 1.85 0.82 1.65 38.50 0.21
3965C6 183459173443219 278.746583 34.722806 15.46 1.663 12.442 11.705 11.570 4400 2.35 0.08 1.70 97.46 0.27
4085C3 183513573427240 278.806542 34.456667 16.13 1.513 13.330 12.675 12.564 4550 1.90 0.58 1.65 7.91 0.12
4217C6 183455123442462 278.729667 34.712861 15.78 1.461 13.065 12.390 12.213 4500 2.15 0.57 1.70 25.21 0.08
4263C6 183501803442404 278.757542 34.711278 15.87 1.329 13.333 12.756 12.650 5000 3.05 0.34 1.90 90.50 0.07
431C2 183553593435209 278.973333 34.589194 15.63 1.175 13.302 12.805 12.744 5100 3.85 0.03 1.20 32.21 0.07
4365C3 183521163426510 278.838208 34.447500 16.08 1.148 13.854 13.339 13.339 4950 1.75 0.80 1.65 202.51 0.25
4478C8 183616693442257 279.069583 34.707167 15.79 1.260 13.441 12.906 12.751 4900 2.25 0.36 1.65 61.90 0.17
455C1 183617843435209 279.074375 34.589194 15.67 1.174 13.413 12.878 12.798 5000 2.70 0.61 1.50 168.72 0.05
4612C6 183501153441540 278.754792 34.698361 16.02 1.264 13.580 12.968 12.908 4950 2.60 0.78 1.30 35.76 0.21
4740C8 183619283441470 279.080417 34.696417 15.80 1.208 13.515 13.011 12.826 5000 2.30 0.49 1.60 181.98 0.17
4876C6 183512283441180 278.801208 34.688361 16.08 1.303 13.583 12.978 12.887 4800 2.70 0.59 1.35 8.46 0.01
5319C6 183443243440187 278.680167 34.671889 15.58 1.447 12.877 12.249 12.110 4600 2.45 0.40 1.50 280.09 0.22
5351C8 183607993440158 279.033333 34.671111 15.83 1.382 13.355 12.641 12.494 4700 2.65 0.08 1.55 79.63 0.24
5400C8 183612363440076 279.051542 34.668806 15.96 1.113 13.820 13.357 13.236 5050 3.75 0.12 1.30 29.93 0.10
5487C8 183614613439566 279.060958 34.665750 16.16 1.339 13.645 13.067 12.971 4750 2.30 0.47 1.75 45.61 0.01
5543C6 183515843439495 278.816042 34.663778 16.07 1.459 13.311 12.694 12.519 4400 1.60 0.17 2.00 6.45 0.04
5588C6 183506423439420 278.776750 34.661694 15.87 1.282 13.371 12.812 12.718 4900 2.60 0.20 1.60 54.21 0.06
5664C6 183452603439317 278.719167 34.658833 15.43 1.330 12.862 12.341 12.241 4800 2.75 0.23 1.60 82.41 0.01
5908C6 183448013438540 278.700000 34.648333 15.66 1.151 13.444 12.982 12.844 5150 2.50 0.67 1.60 138.69 0.05
5977C6 183512223438453 278.800917 34.645917 15.31 1.334 12.714 12.125 12.026 4650 2.45 0.19 1.65 7.78 0.16
5980C6 183443973438441 278.683250 34.645611 15.93 1.242 13.459 12.912 12.849 4950 3.45 0.17 1.25 45.00 0.08
608C1 183613683434559 279.057042 34.582194 15.13 1.133 12.938 12.450 12.348 5000 2.45 1.74 1.00 30.98 0.10
6090C6 183520043438289 278.833542 34.641333 16.15 1.479 13.361 12.735 12.575 4500 2.25 0.02 1.90 42.30 0.16
6164C6 183458843438175 278.745167 34.638194 15.25 1.265 12.830 12.234 12.183 4900 3.25 0.13 0.95 60.24 0.06
6230C5 183433983438536 278.641583 34.648222 15.06 1.154 12.814 12.371 12.284 5050 3.60 0.20 1.25 36.65 0.10
6263C6 183514203438060 278.809208 34.635000 16.01 1.648 12.949 12.313 12.147 4250 2.30 0.48 1.95 7.85 0.03
6391C8 183623573437379 279.098250 34.627167 15.28 1.140 13.105 12.554 12.534 5000 2.20 0.63 1.55 92.33 0.09
6419C5 183430963438299 278.629000 34.641667 15.40 1.390 12.821 12.183 12.082 4700 2.60 0.24 1.55 9.31 0.04
6426C8 183610083437330 279.042083 34.625833 15.79 1.280 13.374 12.825 12.667 4950 2.95 0.40 1.60 30.84 0.28
6505C6 183518673437309 278.827833 34.625222 16.04 1.232 13.778 13.038 12.999 5000 2.70 0.38 1.50 41.77 0.11
650C2 183556793434481 278.986667 34.580083 15.29 1.530 12.462 11.745 11.632 4350 1.55 0.67 1.60 45.55 0.06
6549C6 183503423437243 278.764292 34.623417 15.70 1.320 13.270 12.614 12.506 4900 3.35 0.33 1.30 44.42 0.03
6637C8 183607573437029 279.031583 34.617472 16.07 1.296 13.599 13.056 12.978 4800 2.50 0.26 1.60 32.65 0.20
6717C6 183452383437018 278.718292 34.617167 15.58 1.283 13.127 12.532 12.465 4900 3.30 0.19 1.20 132.37 0.03
6828C7 183556253436545 278.984375 34.615111 16.09 1.360 13.467 12.896 12.808 4650 2.50 0.10 1.70 2.17 0.02
6913C7 183552673436435 278.969458 34.612028 15.95 1.244 13.520 13.003 12.893 5000 2.75 0.27 1.70 4.76 0.07
867C3 183509093434346 278.787917 34.576306 15.76 1.228 13.378 12.839 12.733 5000 2.60 0.47 1.60 136.48 0.25
Table 1Star Identifications, Coordinates, Model Atmosphere Parameters, and Radial Velocities
Species Wavelength E.P. log(gf)23 log (X) log (X) [X/Fe] or [Fe/H]
(Å) (eV)
[O I] 6300.30 0.00 9.750 8.69 8.63 0.44
Na I 6154.23 2.10 1.560 6.33 5.89 0.06
Na I 6160.75 2.10 1.210 6.33 5.89 0.06
Mg I 6318.71 5.10 2.010 7.58 7.38 0.30
Mg I 6319.24 5.10 2.250 7.58 7.38 0.30
Mg I 6319.49 5.10 2.730 7.58 7.38 0.30
Al I 6696.02 3.14 1.570 6.47 6.28 0.31
Al I 6698.67 3.14 1.890 6.47 6.28 0.31
Si I 5645.61 4.93 2.090 7.55 7.38 0.33
Si I 5654.92 5.61 1.714 7.55 7.38 0.33
Si I 5665.56 4.92 1.910 7.55 7.38 0.33
Si I 5666.68 5.62 1.805 7.55 7.38 0.33
Si I 5690.43 4.93 1.910 7.55 7.38 0.33
Si I 5701.10 4.93 2.080 7.55 7.38 0.33
Si I 6142.48 5.62 1.575 7.55 7.38 0.33
Si I 6145.02 5.62 1.460 7.55 7.38 0.33
Si I 6155.13 5.62 0.774 7.55 7.38 0.33
Si I 6155.69 5.62 2.352 7.55 7.38 0.33
Si I 6195.43 5.87 1.560 7.55 7.38 0.33
Si I 6237.32 5.61 1.115 7.55 7.38 0.33
Si I 6244.47 5.62 1.303 7.55 7.38 0.33
Si I 6721.85 5.86 1.016 7.55 7.38 0.33
Ca I 5594.46 2.52 0.370 6.36 6.07 0.21
Ca I 5601.28 2.53 0.463 6.36 6.07 0.21
Ca I 5715.82 2.71 3.386 6.36 6.07 0.21
Ca I 6122.22 1.89 0.466 6.36 6.07 0.21
Ca I 6156.02 2.52 2.637 6.36 6.07 0.21
Ca I 6161.30 2.52 1.246 6.36 6.07 0.21
Ca I 6162.17 1.90 0.210 6.36 6.07 0.21
Ca I 6166.44 2.52 1.262 6.36 6.07 0.21
Ca I 6169.04 2.52 0.837 6.36 6.07 0.21
Ca I 6169.56 2.53 0.628 6.36 6.07 0.21
Cr I 5628.64 3.42 0.832 5.67 5.09 0.08
Cr I 5642.36 3.86 0.840 5.67 5.09 0.08
Cr I 5648.26 3.83 0.980 5.67 5.09 0.08
Cr I 5674.17 3.56 1.507 5.67 5.09 0.08
Cr I 5712.77 3.01 1.107 5.67 5.09 0.08
Cr I 5719.82 3.01 1.660 5.67 5.09 0.08
Cr I 5729.21 3.85 1.038 5.67 5.09 0.08
Cr I 5783.06 3.32 0.510 5.67 5.09 0.08
Cr I 5784.97 3.32 0.440 5.67 5.09 0.08
Cr I 5787.92 3.32 0.183 5.67 5.09 0.08
Cr I 5788.38 3.01 1.524 5.67 5.09 0.08
Cr I 5790.65 1.00 4.033 5.67 5.09 0.08
Cr I 6330.09 0.94 3.000 5.67 5.09 0.08
Cr I 6630.01 1.03 3.560 5.67 5.09 0.08
Cr I 6729.73 4.39 0.753 5.67 5.09 0.08
Fe I 5595.06 5.06 1.490 7.52 7.02 0.50
Fe I 5607.66 4.15 2.260 7.52 7.02 0.50
Fe I 5608.97 4.21 2.240 7.52 7.02 0.50
Fe I 5611.36 3.63 3.010 7.52 7.02 0.50
Fe I 5614.28 5.09 1.298 7.52 7.02 0.50
Fe I 5615.30 2.59 2.268 7.52 7.02 0.50
Fe I 5615.64 3.33 0.170 7.52 7.02 0.50
Fe I 5618.63 4.21 1.456 7.52 7.02 0.50
Fe I 5619.22 3.69 3.170 7.52 7.02 0.50
Fe I 5619.60 4.39 1.420 7.52 7.02 0.50
Fe I 5622.94 3.64 2.986 7.52 7.02 0.50
Fe I 5624.02 4.39 1.230 7.52 7.02 0.50
Fe I 5624.54 3.42 0.440 7.52 7.02 0.50
Fe I 5627.08 4.18 2.920 7.52 7.02 0.50
Fe I 5633.95 4.99 0.310 7.52 7.02 0.50
Fe I 5635.82 4.26 1.640 7.52 7.02 0.50
Fe I 5636.70 3.64 2.630 7.52 7.02 0.50
Fe I 5638.26 4.22 0.820 7.52 7.02 0.50
Fe I 5641.43 4.26 0.890 7.52 7.02 0.50
Fe I 5646.68 4.26 2.440 7.52 7.02 0.50
Fe I 5649.99 5.10 0.770 7.52 7.02 0.50
Fe I 5650.71 5.08 0.810 7.52 7.02 0.50
Fe I 5651.47 4.47 1.850 7.52 7.02 0.50
Fe I 5652.01 4.22 3.010 7.52 7.02 0.50
Fe I 5652.32 4.26 1.870 7.52 7.02 0.50
Fe I 5653.86 4.39 1.480 7.52 7.02 0.50
Fe I 5655.18 5.06 0.600 7.52 7.02 0.50
Fe I 5661.35 4.28 1.856 7.52 7.02 0.50
Fe I 5661.97 4.26 2.770 7.52 7.02 0.50
Fe I 5662.52 4.18 0.563 7.52 7.02 0.50
Fe I 5677.68 4.10 2.640 7.52 7.02 0.50
Fe I 5678.60 2.42 4.770 7.52 7.02 0.50
Fe I 5679.02 4.65 0.900 7.52 7.02 0.50
Fe I 5680.24 4.19 2.330 7.52 7.02 0.50
Fe I 5686.53 4.55 0.626 7.52 7.02 0.50
Fe I 5691.50 4.30 1.540 7.52 7.02 0.50
Fe I 5698.02 3.64 2.790 7.52 7.02 0.50
Fe I 5699.41 4.96 2.044 7.52 7.02 0.50
Fe I 5701.54 2.56 2.046 7.52 7.02 0.50
Fe I 5704.73 5.03 1.319 7.52 7.02 0.50
Fe I 5705.46 4.30 1.565 7.52 7.02 0.50
Fe I 5707.70 4.10 3.148 7.52 7.02 0.50
Fe I 5714.55 5.09 1.715 7.52 7.02 0.50
Fe I 5715.47 4.15 2.990 7.52 7.02 0.50
Fe I 5717.83 4.28 1.090 7.52 7.02 0.50
Fe I 5720.89 4.55 1.750 7.52 7.02 0.50
Fe I 5723.67 4.47 2.250 7.52 7.02 0.50
Fe I 5724.45 4.28 2.610 7.52 7.02 0.50
Fe I 5731.76 4.26 1.210 7.52 7.02 0.50
Fe I 5732.30 4.99 1.440 7.52 7.02 0.50
Fe I 5734.56 4.96 1.784 7.52 7.02 0.50
Fe I 5738.23 4.22 2.240 7.52 7.02 0.50
Fe I 5741.85 4.26 1.744 7.52 7.02 0.50
Fe I 5750.03 5.01 2.323 7.52 7.02 0.50
Fe I 5752.03 4.55 1.077 7.52 7.02 0.50
Fe I 5759.26 4.65 2.040 7.52 7.02 0.50
Fe I 5759.54 4.30 2.179 7.52 7.02 0.50
Fe I 5760.34 3.64 2.590 7.52 7.02 0.50
Fe I 5762.99 4.21 0.460 7.52 7.02 0.50
Fe I 5767.97 4.29 3.236 7.52 7.02 0.50
Fe I 5773.45 3.57 3.704 7.52 7.02 0.50
Fe I 5775.08 4.22 1.238 7.52 7.02 0.50
Fe I 5778.45 2.59 3.590 7.52 7.02 0.50
Fe I 5784.66 3.40 2.672 7.52 7.02 0.50
Fe I 5793.91 4.22 1.750 7.52 7.02 0.50
Fe I 5809.22 3.88 1.630 7.52 7.02 0.50
Fe I 5811.91 4.14 2.460 7.52 7.02 0.50
Fe I 5814.81 4.28 1.910 7.52 7.02 0.50
Fe I 5816.37 4.55 0.681 7.52 7.02 0.50
Fe I 5821.89 4.99 1.676 7.52 7.02 0.50
Fe I 5827.88 3.28 3.260 7.52 7.02 0.50
Fe I 6120.25 0.92 6.020 7.52 7.02 0.50
Fe I 6127.91 4.14 1.499 7.52 7.02 0.50
Fe I 6136.61 2.45 1.480 7.52 7.02 0.50
Fe I 6136.99 2.20 2.900 7.52 7.02 0.50
Fe I 6137.69 2.59 1.343 7.52 7.02 0.50
Fe I 6145.41 3.37 3.770 7.52 7.02 0.50
Fe I 6151.62 2.18 3.349 7.52 7.02 0.50
Fe I 6157.73 4.08 1.110 7.52 7.02 0.50
Fe I 6159.37 4.61 1.850 7.52 7.02 0.50
Fe I 6165.36 4.14 1.614 7.52 7.02 0.50
Fe I 6171.01 4.73 2.244 7.52 7.02 0.50
Fe I 6173.33 2.22 2.870 7.52 7.02 0.50
Fe I 6180.20 2.73 2.666 7.52 7.02 0.50
Fe I 6187.40 2.83 4.168 7.52 7.02 0.50
Fe I 6187.99 3.94 1.740 7.52 7.02 0.50
Fe I 6191.56 2.43 1.367 7.52 7.02 0.50
Fe I 6200.31 2.61 2.407 7.52 7.02 0.50
Fe I 6213.43 2.22 2.542 7.52 7.02 0.50
Fe I 6219.28 2.20 2.353 7.52 7.02 0.50
Fe I 6226.73 3.88 2.210 7.52 7.02 0.50
Fe I 6229.23 2.85 2.955 7.52 7.02 0.50
Fe I 6232.64 3.65 1.323 7.52 7.02 0.50
Fe I 6240.65 2.22 3.333 7.52 7.02 0.50
Fe I 6246.32 3.60 0.953 7.52 7.02 0.50
Fe I 6252.56 2.40 1.697 7.52 7.02 0.50
Fe I 6270.22 2.86 2.704 7.52 7.02 0.50
Fe I 6322.69 2.59 2.376 7.52 7.02 0.50
Fe I 6330.85 4.73 1.290 7.52 7.02 0.50
Fe I 6335.33 2.20 2.187 7.52 7.02 0.50
Fe I 6336.82 3.69 0.966 7.52 7.02 0.50
Fe I 6380.74 4.19 1.326 7.52 7.02 0.50
Fe I 6392.54 2.28 4.090 7.52 7.02 0.50
Fe I 6393.60 2.43 1.562 7.52 7.02 0.50
Fe I 6608.02 2.28 4.070 7.52 7.02 0.50
Fe I 6609.11 2.56 2.602 7.52 7.02 0.50
Fe I 6609.68 0.99 5.700 7.52 7.02 0.50
Fe I 6648.08 1.01 5.824 7.52 7.02 0.50
Fe I 6699.14 4.59 2.081 7.52 7.02 0.50
Fe I 6705.10 4.61 1.122 7.52 7.02 0.50
Fe I 6710.32 1.49 4.890 7.52 7.02 0.50
Fe I 6713.74 4.79 1.530 7.52 7.02 0.50
Fe I 6726.67 4.61 1.183 7.52 7.02 0.50
Fe I 6733.15 4.64 1.550 7.52 7.02 0.50
Fe I 6737.27 3.27 4.339 7.52 7.02 0.50
Fe I 6750.15 2.42 2.681 7.52 7.02 0.50
Fe I 6806.84 2.73 3.180 7.52 7.02 0.50
Fe I 6810.26 4.61 1.086 7.52 7.02 0.50
Fe I 6820.37 4.64 1.130 7.52 7.02 0.50
Fe I 6842.69 4.64 1.270 7.52 7.02 0.50
Fe I 6843.65 4.55 0.960 7.52 7.02 0.50
Fe I 6857.25 4.08 2.230 7.52 7.02 0.50
Fe I 6864.31 4.56 2.410 7.52 7.02 0.50
Fe II 6149.26 3.89 2.681 7.52 7.02 0.50
Fe II 6247.56 3.89 2.405 7.52 7.02 0.50
Fe II 6369.46 2.89 4.141 7.52 7.02 0.50
Co I 5647.23 2.28 hfs 4.90 4.52 0.12
Co I 6117.00 1.78 hfs 4.90 4.52 0.12
Ni I 5593.73 3.90 0.960 6.25 5.81 0.06
Ni I 5614.77 4.15 0.698 6.25 5.81 0.06
Ni I 5625.31 4.09 0.750 6.25 5.81 0.06
Ni I 5628.34 4.09 1.301 6.25 5.81 0.06
Ni I 5638.74 3.90 1.670 6.25 5.81 0.06
Ni I 5641.88 4.11 1.080 6.25 5.81 0.06
Ni I 5643.07 4.17 1.260 6.25 5.81 0.06
Ni I 5682.20 4.11 0.510 6.25 5.81 0.06
Ni I 5694.98 4.09 0.760 6.25 5.81 0.06
Ni I 5748.35 1.68 3.160 6.25 5.81 0.06
Ni I 5760.83 4.11 0.790 6.25 5.81 0.06
Ni I 5796.08 1.95 3.752 6.25 5.81 0.06
Ni I 5805.21 4.17 0.720 6.25 5.81 0.06
Ni I 6128.96 1.68 3.400 6.25 5.81 0.06
Ni I 6130.13 4.27 1.040 6.25 5.81 0.06
Ni I 6175.36 4.09 0.619 6.25 5.81 0.06
Ni I 6176.81 4.09 0.270 6.25 5.81 0.06
Ni I 6177.24 1.83 3.550 6.25 5.81 0.06
Ni I 6186.71 4.11 0.890 6.25 5.81 0.06
Ni I 6191.17 1.68 2.233 6.25 5.81 0.06
Ni I 6223.98 4.11 0.960 6.25 5.81 0.06
Ni I 6322.16 4.15 1.190 6.25 5.81 0.06
Ni I 6635.12 4.42 0.750 6.25 5.81 0.06
Ni I 6767.77 1.83 2.100 6.25 5.81 0.06
Ni I 6772.31 3.66 1.010 6.25 5.81 0.06
Ni I 6813.60 5.34 0.354 6.25 5.81 0.06
Cu I 5782.11 1.64 hfs 4.04 3.71 0.17
Table 2Line List
ID [O/Fe]24 [Na/Fe] [Mg/Fe] [Al/Fe] [Si/Fe] [Ca/Fe] [Cr/Fe] [Fe I/H] [Fe II/H] [Co/Fe] [Ni/Fe] [Cu/Fe]
(l,b)=(5.25,–3.02)
119799C4 0.63 0.08 0.29 0.26 0.38 0.17 0.07 0.39 0.38 0.25 0.21 0.26
129499C4 0.25 0.22 0.24 0.04 0.30 0.26 0.13 0.32 0.06
176772C5 0.12 0.19 0.15 0.08 0.05 0.16 0.05 0.07 0.05 0.07 0.14
181349C5 0.06 0.58 0.15 0.20 0.11 0.08 0.00 0.30 0.21 0.25 0.10
183783C5 0.32 0.12 0.30 0.11 0.15 0.15 0.10 0.32 0.08
184088C5 0.74 0.03 0.30 0.37 0.14 0.37 0.09 0.45 0.48 0.37 0.15 0.37
184618C5 0.00 0.63 0.28 0.24 0.31 0.06 0.58 0.59 0.14 0.15 0.28
185169C5 0.22 0.37 0.42 0.12 0.53 0.21 0.32 0.38 0.07 0.39
185357C5 0.05 0.24 0.01 0.06 0.11 0.04 0.09 0.29 0.27 0.04 0.12
185541C5 0.05 0.21 0.01 0.01 0.06 0.06 0.08 0.34 0.33 0.18 0.06
187067C5 0.35 0.30 0.10 0.19 0.53 0.03 0.95 0.94 0.07 0.07 0.43
193190C5 0.37 0.03 0.08 0.04 0.25 0.16 0.25 0.22 0.22 0.06
197366C5 0.01 0.48 0.08 0.12 0.05 0.13 0.12 0.33 0.31 0.26 0.19
215681C6 0.18 0.33 0.09 0.06 0.02 0.21 0.07 0.31 0.20 0.05 0.51
216922C6 0.43 0.28 0.29 0.34 0.24 0.42 0.25 0.19 0.20 0.24 0.08
218198C6 0.50 0.32 0.31 0.27 0.30 0.32 0.16 0.31 0.33 0.23 0.16
219909C6 0.52 0.20 0.38 0.28 0.25 0.32 0.02 0.73 0.73 0.20 0.04
221537C6 0.35 0.31 0.01 0.09 0.20 0.14 0.12 0.29 0.08 0.06
223113C6 0.43 0.01 0.29 0.12 0.14 0.25 0.01 0.14 0.15 0.15 0.08
223343C6 0.11 0.00 0.10 0.10 0.12 0.16 0.05 0.15 0.14 0.15 0.10
223621C6 0.04 0.02 0.09 0.12 0.25 0.07 0.33 0.28 0.34 0.07
223722C6 0.40 0.05 0.44 0.28 0.33 0.43 0.05 0.76 0.74 0.01 0.02
224206C6 0.10 0.10 0.05 0.03 0.24 0.09 0.46 0.45 0.21 0.11
224866C6 0.29 0.17 0.22 0.31 0.16 0.25 0.03 0.22 0.21 0.17 0.11
224951C6 0.32 0.22 0.07 0.05 0.19 0.14 0.04 0.17 0.14 0.05
225531C6 0.66 0.20 0.46 0.31 0.35 0.40 0.02 0.72 0.73 0.29 0.10 0.28
226450C6 0.16 0.06 0.03 0.03 0.09 0.07 0.13 0.27 0.08 0.05
226850C6 0.25 0.35 0.18 0.33 0.35 0.04 0.85 0.84 0.15 0.00 0.18
227867C6 0.51 0.08 0.12 0.06 0.11 0.08 0.05 0.11 0.07 0.17 0.14 0.38
228466C6 0.55 0.22 0.36 0.34 0.29 0.32 0.31 0.96 0.99 0.18 0.06 0.03
229507C6 0.19 0.15 0.20 0.06 0.26 0.14 0.30 0.28 0.00 0.11 0.15
230424C6 0.57 0.19 0.38 0.18 0.28 0.24 0.02 0.83 0.85 0.19 0.13 0.20
230483C6 0.62 0.13 0.35 0.42 0.32 0.31 0.22 0.63 0.66 0.33 0.13
231379C6 0.54 0.15 0.39 0.28 0.37 0.32 0.01 0.30 0.29 0.32 0.16 0.45
231618C6 0.35 0.07 0.09 0.06 0.12 0.05 0.03 0.04 0.04 0.04 0.11 0.28
232493C6 0.68 0.26 0.29 0.14 0.04 0.45 0.31 1.14 1.13 0.03 0.16 0.55
233121C6 0.63 0.01 0.31 0.26 0.33 0.29 0.04 0.49 0.50 0.16 0.07
233560C6 0.44 0.05 0.30 0.14 0.34 0.25 0.13 0.20 0.22 0.19 0.16 0.40
233708C6 0.13 0.15 0.03 0.01 0.13 0.14 0.11 0.12 0.16 0.09 0.17 0.15
240059C6 0.64 0.11 0.45 0.27 0.29 0.45 0.03 0.70 0.68 0.11 0.02
240083C6 0.38 0.08 0.35 0.27 0.24 0.34 0.02 0.59 0.56 0.04 0.00
259050C7