The Open Cluster Chemical Abundances and Mapping Survey: IV. Abundances for 128 Open Clusters using SDSS/APOGEE DR16
John Donor, Peter M. Frinchaboy, Katia Cunha, Julia E. O'Connell, Carlos Allende Prieto, Andres Almeida, Friedrich Anders, Rachael Beaton, Dmitry Bizyaev, Joel R. Brownstein, Ricardo Carrera, Cristina Chiappini, Roger Cohen, D. A. Garcia-Hernandez, Doug Geisler, Sten Hasselquist, Henrik Jonsson, Richard R. Lane, Steven R. Majewski, Dante Minniti, Christian Moni Bidin, Kaike Pan, Alexandre Roman-Lopes, Jennifer S. Sobeck, Gail Zasowski
DDraft version February 24, 2020
Typeset using L A TEX twocolumn style in AASTeX63
The Open Cluster Chemical Abundances and Mapping Survey: IV.Abundances for 128 Open Clusters using SDSS/APOGEE DR16
John Donor, Peter M. Frinchaboy, Katia Cunha,
2, 3
Julia E. O’Connell, Carlos Allende Prieto,
5, 6
Andr´es Almeida, Friedrich Anders, Rachael Beaton,
9, 10, ∗ Dmitry Bizyaev,
11, 12
Joel R. Brownstein, Ricardo Carrera, Cristina Chiappini, Roger Cohen, D. A. Garc´ıa-Hern´andez,
5, 6
Doug Geisler,
4, 17, 18
Sten Hasselquist, † Henrik J¨onsson,
19, 20
Richard R. Lane,
21, 22
Steven R. Majewski, Dante Minniti,
24, 25, 26
Christian Moni Bidin, Kaike Pan, Alexandre Roman-Lopes, Jennifer S. Sobeck, and Gail Zasowski Department of Physics & Astronomy, Texas Christian University, TCU Box 298840,Fort Worth, TX 76129, USA (j.donor, [email protected]) Steward Observatory, The University of Arizona, 933 North Cherry Avenue, Tucson, AZ 85721-0065, USA Observat´orio Nacional, Rua General Jos´e Cristino, 77, 20921-400 S˜ao Crist´ov˜ao, Rio de Janeiro, RJ, Brazil Departamento de Astronom´ıa, Universidad de Concepci´on, Casilla 160-C, Concepci´on, Chile Instituto de Astrof´ısica de Canarias, V´ıa L´actea S/N, 38205 La Laguna, Tenerife, Spain Universidad de La Laguna, Departamento de Astrof´ısica, 30206 La Laguna, Tenerife, Spain Instituto de Investigaci´on Multidisciplinario en Ciencia y Tecnolog´ıa,Universidad de La Serena, Benavente 980, La Serena, Chile Departament de F´ısica Qu`antica i Astrof´ısica, Universitat de Barcelona, IEEC-UB, Mart´ı i Franqu`es 1 08028 Barcelona, Spain Department of Astrophysical Sciences, Princeton University, 4 Ivy Lane, Princeton, NJ 08544 The Observatories of the Carnegie Institution for Science, 813 Santa Barbara St., Pasadena, CA 91101 Apache Point Observatory and New Mexico State University, P.O. Box 59, Sunspot, NM, 88349-0059, USA Sternberg Astronomical Institute, Moscow State University, Moscow, Russia Department of Physics & Astronomy, University of Utah, 115 S. 1400 E., Salt Lake City, UT 84112, USA Astronomical Observatory of Padova, National Institute of Astrophysics, Vicolo Osservatorio 5 - 35122 - Padova Leibniz-Institut fur Astrophysik Potsdam (AIP), An der Sternwarte 16, 14482 Potsdam, Germany Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA Departamento de Astronom´ıa, Universidad de La Serena, Avenida Juan Cisternas 1200, La Serena, Chile Instituto de Investigaci´on Multidisciplinario en Ciencia y Tecnolog´ıa, Universidad de La Serena. Benavente 980, La Serena, Chile Materials Science and Applied Mathematics, Malm¨o University, SE-205 06 Malm¨o, Sweden Lund Observatory, Department of Astronomy and Theoretical Physics, Lund University, Box 43, SE-22100 Lund, Sweden Instituto de Astrof´ısica, Pontificia Universidad Cat´olica de Chile, Av. Vicuna Mackenna 4860, 782-0436 Macul, Santiago, Chile Instituto de Astronom´ıa y Ciencias Planetarias, Universidad de Atacama, Copayapu 485, Copiap´o, Chile Department of Astronomy, University of Virginia, Charlottesville, VA 22904-4325, USA Departamento de Ciencias Fisicas, Facultad de Ciencias Exactas, Universidad Andres Bello, Av.Fernandez Concha 700, Las Condes, Santiago, Chile Millennium Institute of Astrophysics, Av. Vicuna Mackenna 4860, 782-0436, Santiago, Chile Vatican Observatory, V00120 Vatican City State, Italy Instituto de Astronom´ıa, Universidad Cat´olica del Norte, Av. Angamos 0610, Antofagasta, Chile Department of Astronomy - Universidad de La Serena - Av. Juan Cisternas, 1200 North, La Serena, Chile Department of Astronomy, University of Washington, Seattle, WA, 98195, USA
Submitted to Astronomical JournalABSTRACTThe Open Cluster Chemical Abundances and Mapping (OCCAM) survey aims to constrain keyGalactic dynamical and chemical evolution parameters by the construction of a large, comprehensive,uniform, infrared-based spectroscopic data set of hundreds of open clusters. This fourth contribution
Corresponding author: John [email protected] a r X i v : . [ a s t r o - ph . GA ] F e b Donor et al. from the OCCAM survey presents analysis using SDSS/APOGEE DR16 of a sample of 128 openclusters, 71 of which we designate to be “high quality” based on the appearance of their color-magnitudediagram. We find the APOGEE DR16 derived [Fe/H] abundances to be in good agreement withprevious high resolution spectroscopic open cluster abundance studies. Using the high quality sample,we measure Galactic abundance gradients in 16 elements, and find evolution of some of the [X/Fe]gradients as a function of age. We find an overall Galactic [Fe/H] vs R GC gradient of − . ± . − over the range of 6 < R GC < . Keywords:
Open star clusters (1160), Galactic abundances (2002), Milky Way evolution (1052), Chem-ical abundances (224) INTRODUCTIONIn this era of multi-fiber spectrographs, studies of tensof thousands of stars across the Galaxy are common.However, to derive critical parameters such as age anddistance, the importance of reliable calibration samplescannot be understated. Open clusters serve as reliableage, distance, and chemical tracers distributed aroundthe Galactic disk.Open clusters have been used to study Galactic chem-ical trends as far back as Janes (1979), where the authorshowed open clusters to be a reliable tracer of a Galacticradial metallicity gradient. More recently, this trend hasbeen consistently considered a 2-function gradient (e.g.,Sestito et al. 2008; Bragaglia et al. 2008; Friel et al. 2010;Carrera & Pancino 2011; Yong et al. 2012; Frinchaboyet al. 2013; Reddy et al. 2016; Magrini et al. 2017), withthe break falling between R GC ≈
10 kpc and R GC ≈ − .
05 dex kpc − (Reddy et al. 2016; Casamiquelaet al. 2019), and − . − (Jacobson et al. 2016).In addition, Donor et al. (2018) (henceforth OCCAMII)showed that this gradient could change by as much as40% depending on which distance catalog was used.Since open clusters can range in age from a few Myr tomore than 6 Gyr, they also provide a unique opportunityto study the evolution of Galactic abundance gradient s.A number of authors have measured metallicity gradi-ents for open clusters in various age bins (e.g., Carraro ∗ Hubble FellowCarnegie-Princeton Fellow † NSF Astronomy and Astrophysics Fellow et al. 1998; Friel et al. 2002; Jacobson et al. 2011; Car-rera & Pancino 2011; Cunha et al. 2016), and while allstudies agree that the gradient is shallower for youngerclusters, further comparison is difficult due to a some-what heterogeneous choice of age bins; there does notseem to be a consensus as to the measured gradient forclusters of any given age range.Indeed, there are indications the picture is even morecomplicated. While open clusters have the advantage ofprecise age estimates, there are complexities that mustbe considered when using them to probe Galactic evo-lution. Anders et al. (2017) suggest open clusters in theinner galaxy are more likely to be broken up, leading tosamples significantly biased towards younger clusters.Galactic trends in elements besides iron have been re-ported (e.g., Yong et al. 2005; Friel et al. 2010; Jacobsonet al. 2011). Trend lines are commonly fit for α -elements(e.g., Carrera & Pancino 2011; Yong et al. 2012; Reddyet al. 2016), and in some cases for other elements , suchas [Ni/Fe], [Cr/Fe], and [V/Fe] (Casamiquela et al. 2019)or [Na/Fe] and [Al/Fe] (Yong et al. 2012). There is agrowing consensus that there is a mild positive [ α /Fe]versus R GC trend in the inner galaxy, similar to somechemodynamical model predictions (see Minchev et al.2014). OCCAMII showed the value of studying trendsin other elements, finding strong evidence for a negativetrend in [Mn/Fe] vs R GC .In this paper we will present the expanded OC-CAM sample based on results from SDSS IV ApachePoint Observatory Galactic Evolution Experiment 2(APOGEE 2; Majewski et al. 2017) Data Release 16(DR16) (J¨onsson et al., in prep ). We discuss this samplein comparison to the previously studied sample of openclusters that used SDSS IV DR14 results (OCCAMII),as well as other results from the literature. We then ex- CCAM: IV. Cluster Abundances with APOGEE DR16 α elements, iron-peakelements, and all other elements reported by APOGEEas a function of Galactocentric distances. We finallybreak the sample in age bins to explore changes in ra-dial abundance trends over time. DATATo minimize the impact of calibration differencesand other systematic effects, and ensure uniformity,the OCCAM survey uses as much data from as fewsources as possible; therefore, the majority of this anal-ysis is based primarily on two large surveys,
Gaia andSDSS/APOGEE.Our primary source of chemical abundance and ra-dial velocity (RV) data is the Sloan Digital Sky Sur-vey’s (SDSS) sixteenth data release (DR16) (Ahumadaet al., submitted ; J¨onsson et al., in prep ; Blanton et al.2017) taken as part of the second, dual hemispherephase of APOGEE (APOGEE 2) (Majewski et al. 2017).APOGEE is a high resolution, near infrared spectro-scopic survey currently operating in both hemispheres,at Apache Point Observatory (APO; New Mexico, Gunnet al. 2006) and Las Campanas Observatory (LCO;Chile, Bowen & Vaughan 1973). The APOGEE/DR16dataset includes about 430,000 stars, collected betweenAugust 2011 and August 2018 using the two 300-fiberAPOGEE spectrographs (Wilson et al. 2019) and, forthe first time, the APOGEE survey has near-completecoverage in Galactic longitude, due to the first release ofdata from LCO. The APOGEE data reduction pipeline(Nidever et al. 2015; Holtzman et al. 2015, 2018, J¨onssonet al., in prep ) provides stellar atmospheric parametersand radial velocity measurements, while elemental abun-dances are provided from the ASPCAP pipeline (Garc´ıaP´erez et al. 2016; M´esz´aros et al. 2012; Zamora et al.2015; Holtzman et al. 2018, J¨onsson et al., in prep ).Copper, cerium (Cunha et al. 2017), neodymium (Has-selquist et al. 2016), and ytterbium abundances are re-ported from ASPCAP for the first time in DR16, al-though neodymium and ytterbium lines are so weak orblended that these ASPCAP abundances are consideredunreliable. Concerning cerium, the APOGEE regioncontains several Ce II lines (Cunha et al. 2017), how-ever, the current DR16 results are only based on one CeII line; future data releases will use the full sample ofcerium lines. Therefore we will postpone any discussionof cerium until future data releases.In the APOGEE DR16 allStar-file several types ofabundances are reported for every star and element:firstly the abundance reported by the analysis pipelineis supplied in the FELEM-array. Secondly, these abun-dances have been calibrated with a zero-point shift to ensure solar metallicity stars in the solar neighbor-hood have [X/M]=0; in practice these shifts are small, < .
05 dex, except for Al, K, V, and Mn. Finally,these calibrated abundances have been culled for par-ticular uncertain values by the ASPCAP-team (e.g., for[Y/Fe] or [Nd/Fe]). These final, “cleaned” and cali-brated abundances are supplied in the “named tags”;FE H, MG FE, CE FE, etc. More information, includ-ing what zero-point shifts have been applied, is providedin J¨onsson et al., in prep . In this paper we use theabundances of the “named tags” as is recommended inJ¨onsson et al., in prep .Targeting for APOGEE relied on input from two all-sky surveys: 2MASS (Cutri et al. 2003) and WISE(Wright et al. 2010). More details specifically aboutopen cluster targeting are provided in OCCAMII, anddetails about APOGEE targeting generally can be foundin Zasowski et al. (2013, 2017).Our secondary source of data is
Gaia
DR2 (Gaia Col-laboration et al. 2016, 2018; Lindegren et al. 2018); weuse photometric and astrometric data for 1,365,376
Gaia stars, radial velocity measurements for 16,084 stars, andparallax values for 886 stars in common with APOGEE.We use cluster coordinates and radii from Dias et al.(2002). For this study, we use the uniform distancedetermination from Kharchenko et al. (2013, generallyreferred to as the Milky Way Star Cluster, MWSC,catalog) when measuring galactic trends; however, webriefly compare to other uniform distance catalogs (e.g.,Cantat-Gaudin et al. 2018; Bailer-Jones et al. 2018) in § METHODS3.1.
Membership Analysis
The selection of cluster member stars utilizes the stel-lar radial velocities, proper motions (PM), spatial loca-tion, and derived metallicities as membership discrim-inators. For this study, we use the membership pro-cedure, fully described in OCCAMII with some minorimprovements. The method of OCCAMII first performsa PM analysis using
Gaia
DR2 to isolate likely clustermembers. If multiple APOGEE stars are selected forthe same cluster that have very different RVs, there isan inherent ambiguity and a “correct” systemic clus-ter velocity cannot be chosen. We now leverage the RVmeasurements from
Gaia , when available, for stars iden-tified as likely PM members to significantly increase thenumber of RV measurements in a cluster and more reli-ably determine the cluster system velocity.To be included as a cluster member, a star must fallwithin 3 σ of the cluster mean as established by the ker- Donor et al.
Table 1.
OCCAM DR16 Sample - Basic Parameters
Cluster Qual l b R a Age b R GC b µ α c µ δ c RV [Fe/H] Numname flag deg deg ( (cid:48) ) Gyr (kpc) (mas yr − ) (mas yr − ) (km s − ) (dex) starsHigh Quality ClustersRuprecht 147 1 21.0089 − − ± − ± ± ± − − ± − ± ± ± − − ± − ± ± ± − ± − ± ± − ± − ± − ± − ± ± − ± − ± ± ± − ± − ± ± − ± − ± − ± ± ± − − ± − ± − ± − ± − − ± − ± − ± ± a Radius from Dias et al. (2002) b Calculated using or taken from MWSC Catalog. c µ α and µ δ and their 1 σ uncertainties are those of the 2D Gaussian fit, as in OCCAMII.(This table is available in its entirety in machine-readable form.) Table 2.
OCCAM DR16 Sample - Detailed Chemistry
Cluster [Fe/H] [O/Fe] [Na/Fe] [Mg/Fe] [Al/Fe] [Si/Fe] [S/Fe] [K/Fe]name (dex) (dex) (dex) (dex) (dex) (dex) (dex) (dex)[Ca/Fe] [Ti/Fe] [V/Fe] [Cr/Fe] [Mn/Fe] [Co/Fe] [Ni/Fe] [Cu/Fe](dex) (dex) (dex) (dex) (dex) (dex) (dex) (dex)High Quality ClustersRuprecht 147 0 . ± . − . ± .
03 0 . ± . − . ± .
02 0 . ± . − . ± .
05 0 . ± .
06 0 . ± . − . ± . − . ± .
09 0 . ± .
07 0 . ± .
09 0 . ± .
03 0 . ± .
20 0 . ± . − . ± . . ± . − . ± .
02 0 . ± . − . ± . − . ± .
03 0 . ± .
01 0 . ± . − . ± . − . ± . − . ± . − . ± . − . ± .
04 0 . ± .
02 0 . ± .
03 0 . ± .
01 0 . ± . . ± . − . ± .
01 0 . ± . − . ± . − . ± .
02 0 . ± .
01 0 . ± . − . ± . − . ± . − . ± . − . ± . − . ± .
03 0 . ± .
01 0 . ± .
03 0 . ± . − . ± . − . ± .
01 0 . ± . − . ± . − . ± . − . ± .
02 0 . ± .
01 0 . ± . − . ± . − . ± . − . ± . − . ± . − . ± .
03 0 . ± .
01 0 . ± . − . ± .
01 0 . ± . . ± .
04 0 . ± .
03 0 . ± .
06 0 . ± .
03 0 . ± .
07 0 . ± . − . ± .
05 0 . ± . − . ± .
03 0 . ± . − . ± . − . ± .
09 0 . ± .
13 0 . ± .
08 0 . ± .
04 0 . ± . . ± . − . ± .
03 0 . ± . − . ± . − . ± .
03 0 . ± .
03 0 . ± . − . ± . . ± .
02 0 . ± .
03 0 . ± .
13 0 . ± .
03 0 . ± .
03 0 . ± .
06 0 . ± .
02 0 . ± . − . ± . − . ± .
02 0 . ± . − . ± . − . ± . − . ± .
01 0 . ± . − . ± . . ± .
01 0 . ± . · · · . ± .
03 0 . ± . − . ± . − . ± . − . ± . . ± . · · · − . ± . − . ± . − . ± . − . ± .
01 0 . ± . − . ± . . ± .
02 0 . ± . · · · . ± .
05 0 . ± . − . ± . − . ± .
01 0 . ± . − . ± . − . ± .
02 0 . ± . − . ± . − . ± . − . ± .
01 0 . ± .
07 0 . ± . . ± . − . ± . · · · . ± .
04 0 . ± . − . ± . − . ± .
01 0 . ± . . ± . · · · . ± . − . ± . − . ± . − . ± .
01 0 . ± . − . ± . − . ± . − . ± . · · · − . ± . − . ± .
01 0 . ± . − . ± . − . ± . CCAM: IV. Cluster Abundances with APOGEE DR16 all three spaces considered (RV, [Fe/H], and PM).3.2.
Visual Quality Check
A visual inspection of each cluster’s PM-cleaned color-magnitude diagram (CMD) was performed by multipleof the authors. Figure 1 shows five example CMDs. Thevisual assessment is meant to evaluate whether starsthat pass the combined RV, proper motion and metal-licity criteria also lie in a sensible position in the ob-served cluster CMD, considering their spectroscopicallydetermined log(g). This is an easy case when, for ex-ample, one or more APOGEE OCCAM candidates withhigh log(g) ( log ( g ) ≥ .
7) are found to lie along an eas-ily discernible photometric main sequence in the CMD(e.g., Melotte 22), thus providing a joint affirmation thatthe star is likely a main sequence member of the clus-ter. These clusters are flagged as “1” or “high quality”.However, most of the OCCAM stars from APOGEEturn out to be evolved stars – subgiants, giants andred clump stars, with log ( g ) < .
7. In this case, thestar is still considered a member if the star lies alongthe subgiant/giant branch of the cluster, which, how-ever, must generally be projected from the location ofthe main sequence and its turn-off, given that the sub-giant/giant sequences in most clusters are typically verypoorly populated (e.g., NGC 1664). These clusters arealso flagged as “1” or “high quality”. The latter pro-cess becomes more challenging when the main sequenceis also poorly populated (e.g., Chupina 5), or when thefield star contamination becomes so dominant as to ob-scure the cluster main sequence (e.g., ASCC 18). Theseclusters are flagged as “0” or “potentially unreliable”.Clusters where the APOGEE OCCAM candidate is nota part of any discernable sequence or where there is nodiscernable sequence (e.g., Basel 15) are rejected. Thesequality flags are included in the full version of Tables 1and 2 (available online), and in the value added catalog,described below.3.3.
Data Access - SDSS Value Added Catalog
The data this analysis uses are also available as aValue Added Catalog (VAC) that was released alongwith SDSS-IV DR16. The VAC consists of two tables.The first is a combination of Table 1 and Table 2, show-ing bulk cluster parameters derived here including PM,and RV, but also including abundances for all elementsreported in DR16. We note that cluster ages are not in- Elements such as Rb and Y that do not have calibrated valuesreported in DR16 are not included.
Table 3.
A summary of the individual star data included inthe DR16 OCCAM VAC
Label DescriptionCLUSTER The associated open cluster2MASS ID star ID from 2MASS surveyLOCATION ID a from APOGEE DR16GLAT Galactic latitudeGLON Galactic longitudeFE H a [Fe/H]FE H ERR a uncertainty in FE HVHELIO AVG a heliocentric radial velocityVSCATTER a scatter in APOGEE RV measurementsPMRA b proper motion in right ascensionPMDEC b proper motion in declinationPMRA ERR b uncertainty in PMRAPMDEC ERR b uncertainty in PMDECRV PROB membership probability based on RV (This study)FEH PROB membership probability based on FE H (This study)PM PROB membership probability based on PM (This study)CG PROB membership probability from Cantat-Gaudin et al. (2018) a Taken directly from APOGEE DR16 b From
Gaia
DR2 cluded in the VAC as only ages from the MWSC catalogare used in this work.Five measurements of R GC are also included. Wecalculate R GC using catalog distances from Dias et al.(2002) , Kharchenko et al. (2013, MWSC), and Cantat-Gaudin et al. (2018). We also calculate R GC based onmedian parallax from member stars and median distancefor member stars from Bailer-Jones et al. (2018), as inOCCAMII. In § × Radius
Dias of the cluster center).For each star, we reproduce relevant parameters (RV,[Fe/H], and proper motion) and provide our member-ship probability estimate based on each parameter. Forconvenience, we also provide the membership determina-tion from Cantat-Gaudin et al. (2018) (when provided).All columns available in the VAC are presented in Table3. The catalog is available from sdss.org here .Both tables are also available for exploration usingFiltergraph (Burger et al. 2013) at https://filtergraph.com/sdss apogee occam/. We acknowledge an error in our pipeline that populated R GC forsome clusters where no distance is reported by Dias et al. (2002).”R GC DIAS” values for the clusters ASCC 16, Chupina 3, 4,& 5, Collinder 95, FSR 0687, L 1241s, NGC 358, and Platais 4should be disregarded Donor et al. . . . . . BP-RP G NGC 1664: 1
BP-RP G Melotte 22: 1
BP-RP G ASCC 18: 0
BP-RP G Chupina 5: 0
BP-RP G Basel 15: rejected
Figure 1.
Five example color-magnitude diagrams of open clusters analyzed in the study, with cluster name and qualitydesignation from Table 1 .
Gaia stars within twice the cluster radius are shown; stars identified as PM members and inside thecluster radius are blue. Non-member stars are shown as a Hess diagram in grey except for Chupina 5 where actual stars areshown. The OCCAM pipeline-identified APOGEE members are shown as orange stars.
20 15 10 5 0 5
X (kpc) (R = 8.00 kpc) Y ( k p c ) R GC = R GC = R GC = [ F e / H ] Figure 2.
The full OCCAM DR16 sample plotted in theGalactic plane. Square points are “high quality” clusters,triangles are the lower quality clusters. The colorbar shows[Fe/H]. The concentric circles show R GC = 8, 16, & 24 kpc4. THE OCCAM DR16 SAMPLEOur final sample in this study consists of 128 openclusters with 914 member stars, out of 10,191 stars nearcluster fields considered in the analysis. Of those 128clusters, 83 clusters were designated as “high quality”based on a visual CMD inspection. For the Galacticabundance analysis in this study, we will only use thoseclusters flagged as high quality, as presented in Table 1.The other clusters with questionable quality, e.g., thosethat did not pass visual checks ( § ≤ R GC ≤
14 kpc, with good R GC coverage in that range. Twohigh quality clusters fall outside of this range: Berkeley20 at R GC ≈ . R GC ≈ . . Using age estimates from the MWSC catalog, oursample spans a range in age from ∼ ∼ ,with nearly half under 1 Gyr.4.1. Modifications to the High Quality Sample
Beyond those clusters excluded from analysis basedon our visual inspection of their PM-cleaned CMDs, wehave further excluded 12 clusters (ASCC 16, ASCC 19,ASCC 21, Briceno 1, Chupina 1, Chupina 3, Collinder69, Collinder 70, IC 348, NGC 1980, NGC 1981, NGC2264) because they are reported to be very young ( <
50 Myr) (Kharchenko et al. 2013) and previous studiesof young stars in APOGEE suggest the pipeline resultsmay be unreliable (e.g., Kounkel et al. 2018). Thus thefinal sample used for analysis consists of 71 clusters.There are additional affects within clusters that mayresult in unreliable abundance determinations. Soutoet al. (2018, 2019) showed that abundances in dwarfand giant stars in the old cluster NGC 2682 differed sig-nificantly due to atomic diffusion. For this reason, thedwarf stars in NGC 2682 are excluded from our abun-dance analysis. NGC 752 is also relatively old and maysuffer from diffusion effects, we therefore exclude the We note Dias et al. (2002) find Be 29 to be significantly fur-ther away at R GC ≈ . We note some studies of NGC 6791 (e.g. Brogaard et al. 2012) findit to be significantly older, however in the interest of a uniformanalysis we rely only on ages from the MWSC catalog
CCAM: IV. Cluster Abundances with APOGEE DR16 − . − . − . . . . . [Fe/H] − . − . . . ∆ ( d e x ) , D R - D R Figure 3.
The difference in reported [Fe/H] from DR14 toDR16 for the 19 clusters from OCCAMII. A characteristicerror-bar is shown. dwarfs in this cluster from abundance analysis as well.As a result, for both NGC 752 & NGC 2682 we only usethe giant stars to determine the cluster abundances.4.2.
Comparison to previous work
OCCAM PAPER II
For the 19 open clusters studied in OCCAMII, we plot∆ [Fe/H] vs DR16 [Fe/H] in Figure 3. Figure 3 showsthat the mean [Fe/H] for OCCAM clusters changed be-tween APOGEE DR14 and DR16; this is mostly due tochanges in the gf-values of the Fe I lines in the DR16line list (Smith et al. in prep ). There is a clear offsetfor all clusters, with a mean difference of 0.05 dex. InOCCAMII it was shown that APOGEE DR14 [Fe/H]values for six well studied open clusters were on averageapproximately 0.05 dex more metal-rich than the resultsin the literature. If we repeat the same literature com-parison using our DR16 values we find a mean offsetof [Fe/H] = 0.004. All of these offsets are within theirmeasured 1 σ dispersions.Figure 4 shows a similar plot for other elements. Be-yond the quoted uncertainties in each case, there are noobvious systematic trends for any of these elements.4.2.2. Open Clusters Observed by the LAMOST Survey
Zhang et al. (2019) published mean abundances foropen clusters using results from the LAMOST survey(Luo et al. 2015). Our sample includes 22 open clus-ters in common with Zhang et al. (2019) and we finda median offset in [Fe/H] (in the sense LAMOST -APOGEE) of -0.01 dex; however we note some signifi-cant outliers. Figure 5 shows the difference in [Fe/H] be-tween Zhang et al. (2019) and this work ([Fe/H]
LAMOST - [Fe/H]
AP OGEE ). There is fairly good agreement nearsolar metallicity, but towards lower metallicities (asmeasured by APOGEE), there are some clusters withhighly discrepant results. The three clusters with themost discrepant metallicities, (cid:38) . ∼ . − . . . .
10 [O/Fe] [Mg/Fe] − . . . .
10 [Si/Fe] [S/Fe] − . . . .
10 [Ca/Fe] [V/Fe] − . . . .
10 [Cr/Fe] [Mn/Fe] − . − . . . − . . . .
10 [Co/Fe] − . − . . . − . − . − . . . . . [ F e / H ] D R [X/Fe] ∆ ( d e x ) , D R - D R Figure 4.
Similar to Figure 3 but for other elements.Characteristic error bars are shown. Datapoints are coloredby their [Fe/H] as reported in APOGEE DR16 − . − . − . − . − .
05 0 . [Fe/H] APOGEE − . − . . . . . . . L A M O S T - A P O G EE NA P O G EE s t a r s Figure 5.
The difference between the metallicities in theLAMOST (from Zhang et al. 2019) and APOGEE surveysfor open clusters in common. The color bar indicates thenumber of APOGEE stars in the cluster (saturating at 5).The square symbols denote clusters with a single star inZhang et al. (2019). (Czernik 23 and NGC 2264) have only one star in theZhang et al. (2019) analysis, and Czernik 23 has onlyone star in APOGEE as well. NGC 2264 and ASCC 21are among the young clusters which were excluded fromour high quality sample. Removing these three mostdiscrepant clusters, the LAMOST values are much moreconsistent with APOGEE.A previous comparison of APOGEE DR14 to LAM-OST found an offset in [Fe/H] of 0.06 with a scatter of0.13 (Anguiano et al. 2018). Given the analysis in § Donor et al. − . . [ F e / H ] (a) Dias CatalogR GC < ± GC > ± − . . [ F e / H ] (b) MWSC CatalogR GC < ± GC > ± − . . [ F e / H ] (c) Inverse ParallaxR GC < ± GC > ± R GC (kpc) − . . [ F e / H ] (d) Cantat-Gaudin CatalogR GC < ± GC > ± N APOGEE stars
Figure 6. [Fe/H] vs R GC trends measured using differentdistance determinations. This is similar to an analysis per-formed in OCCAMII, but we have added measurements fromCantat-Gaudin et al. (2018) where available. The colorbarshows the number of APOGEE stars per cluster, saturatingat 5. 5. MEASURING GALACTIC TRENDS5.1.
Choosing a Distance Catalog
In OCCAMII, Galactocentric distances to open clus-ters were calculated using the average distance for mem-ber stars from Bailer-Jones et al. (2018). However, dueto the application of a geometric prior to each star in-dividually, this may not be an optimal solution for clus-ters (Bailer-Jones et al. 2018). Another uniform sourceof distances is therefore desired.Distances to open clusters are frequently recomputedby many groups. Some form of isochrone fitting hasbeen used by a number of studies (e.g., von Hippelet al. 2006; Kharchenko et al. 2013), however, onlyKharchenko et al. (2013, MWSC) have produced a cata-log using a uniform isochrone fitting method to measuredistances for a very large (over 1000) set of open clus-ters. Recently, the
Gaia survey has made it possible tocreate large catalogs of cluster distances based on paral-lax (e.g., Cantat-Gaudin et al. 2018). Of the two largecatalogs, the MWSC catalog covers significantly moreof our sample, but still, two clusters in our high qualitysample (BH 211 and Teutsch 12) are not included. Forthese clusters, we rely on stellar parallaxes from
Gaia
DR2. Since the MWSC catalog does not include dis- tance uncertainties, we assume an uncertainty of 10% ofthe distance.For completeness, and to highlight the significant in-fluence that choosing a particular distance catalog canhave on the measured gradient, Figure 6 repeats thebasic analysis of § ∼ ∼ Fitting Galactic Abundance Gradients
It has become common in the literature, when measur-ing Galactic metallicity gradients, to divide the samplesomewhere between R GC ≈
10 kpc and R GC ≈
13 kpcand fit two separate lines to the data (e.g., Twarog et al.1997; Sestito et al. 2008; Friel et al. 2010; Jacobson et al.2011; Carrera & Pancino 2011; Yong et al. 2012; Frinch-aboy et al. 2013; Reddy et al. 2016; Magrini et al. 2017),with a much shallower trend in the outer galaxy than inthe inner galaxy. Since the OCCAM sample includesopen clusters as far away as R GC ≈
19 kpc, we caninvestigate if the Galactic metallicity gradient becomessignificantly shallower at a given R GC .In this study, we fit two separate lines to the data,and impose the additional constraint that both mustmeet at some “knee”, although the location of the kneeis allowed to vary. If we let k be the x -coordinate of theknee, the equation describing the fit line is then: y = m · x + b x ≤ km · ( x − k ) + ( m · k + b ) x > k (1)We estimate the values of m , b , m , and k us-ing maximum likelihood estimation. Uncertainties ineach parameter are estimated using the emcee pack-age (Foreman-Mackey et al. 2013). For trends whichdo not appear to have multiple components (e.g., [ α /Fe]vs R GC trends), we perform a maximum likelihood fitand emcee error estimation for a single line. THE GALACTIC METALLICITY GRADIENTFitting to the overall [Fe/H] versus R GC gradient us-ing open clusters as probes is common in many Galacticstudies ( see e.g., Table 4 of OCCAMII). We fit the over-all [Fe/H] vs R GC trend using our high quality sampleof 71 open clusters, with a 2 line function fit (Figure7). We find an inner ( R GC < . − . ± .
004 dex/kpc and an outer ( R GC > . − . ± .
011 dex/kpc.
CCAM: IV. Cluster Abundances with APOGEE DR16
9A consensus on the apparent location of the “knee”has nearly been reached in the literature, with valuesconverging around R GC ≈
12 kpc. However, this loca-tion does not appear to have been rigorously tested ;that is, the position of the “knee” has never been in-cluded as a free parameter in the fit.We find the location of the break in the Galactic[Fe/H] vs R GC trend to be at R GC = 13 . R GC >
14 kpc and the effect this may have on the deter-mination of this parameter. Additional open clusters inthis R GC range have been targeted as part of APOGEE2 and should be observed soon.If we consider only the 19 open clusters studied in OC-CAM II and fit a single line as in that previous study,we find a gradient of − . ± .
005 dex/kpc if we in-clude NGC 6791 and − . ± .
005 dex/kpc if we donot include NGC 6791. OCCAM II found a gradient of − . ± .
003 dex/kpc using distances from the MWSCcatalog and excluding NGC 6791. We emphasize thatalthough we find a global offset of 0.05 dex in [Fe/H]between DR14 and DR16, this is not expected to havean effect on the slope of the [Fe/H] versus R GC trend asthe offset should be roughly similar at any given [Fe/H].Given the comparison between gradients derived fromDR14 and DR16 results, this appears to be the case.Table 4 of OCCAMII summarized recent measure-ments of the Galactic metallicity gradient from the liter-ature in the distance range considered of 6 (cid:46) R GC (cid:46) − . − .
085 dex/kpc. The result in this study of − .
068 dex/kpc sits neatly in the middle of this range.We can compare in more detail to the recent resultsfrom Carrera et al. (2019), which also used APOGEEdata (from DR14). The authors chose to split theirsample at R GC = 11 kpc, and find an inner gradientof − . ± .
007 dex/kpc. This is nearly in agreementwith our result. We note the authors used distancesfrom Cantat-Gaudin et al. (2018); in Figure 6d we mea-sure the metallacity gradient using the same distancesand find a slope of − . ± .
001 dex/kpc, in goodagreement with their result. GALACTIC TRENDS FOR OTHER ELEMENTS7.1.
Galactic Trends for α -Elements Figure 8 shows Galactic trends versus Fe for six α -elements (O, Mg, S, Si, Ca, and Ti). Since we find abreak in the [Fe/H] vs R GC trend at R GC ≈
14 kpc, welimit the sample to R GC <
14 kpc and measure trends for the inner clusters. For all α elements studied here,except for silicon and titanium, there is a statisticallysignificant slight positive trend from the inner galaxy tothe outer galaxy and the gradients in [ α /Fe] are consis-tent overall. However, for silicon and titanium we finda flat gradient. We note there is significant scatter for[S/Fe], and very little scatter for [Ca/Fe].Our results are consistent with Yong et al. (2012) whomeasured mild positive gradients for [O/Fe], [Si/Fe],and [Ca/Fe], of the order of 0.01 dex kpc − , but a flattrend for [Mg/Fe], although the uncertainties on all fourtrends are nearly as large as their measured gradients.Casamiquela et al. (2019) report slight positive gradientsfor [Si/Fe] (0 . ± . . ± . ≤ R GC ≤
11 kpc, although both slopes are muchshallower when they include more clusters from the lit-erature. Carrera & Pancino (2011) and Reddy et al.(2016) report [ α /Fe] vs R GC gradients of 0 . ± . − and 0 . ± .
005 dex kpc − , respectively.Our results are therefore in good agreement with theliterature, except perhaps for Si which appears to bealmost completely flat in our case.Recent work using APOGEE data showed a possi-ble temperature effect for silicon abundances (Zasowskiet al. 2019): cooler stars show lower abundances thanwarmer ones. The stars in more distant clusters tendto be cooler since only brighter, more evolved stars aredetectable farther away. Thus the flat [Si/Fe] trend maypartly reflect this effect in APOGEE data.7.2. Galactic Trends for Iron-Peak Elements
APOGEE DR16 reports abundances for six elementsthat are classified as “iron-peak” elements: vanadium,chromium, manganese, cobalt, nickel, and copper. Fig-ure 9 shows Galactic abundance trends for each of theseelements. The [Ni/Fe] vs R GC trend is completely flat;the abundances stay very near solar with small scatterfor the Galactic radii explored. Statistically significantslightly positive trends are measured for [V/Fe], [Cr/Fe],and [Cu/Fe], however there are some significant out-liers for [Cr/Fe] (Czernik 18 having a single star with[Cr/Fe] = +0 .
57) and [Cu/Fe] (Chupina 1 having a sin-gle star with [Cu/Fe] = − . R GC ≈
11 kpc and R GC ≈
13 kpc. Interest-ingly, Casamiquela et al. (2019) find a mildly significant negative trend for [V/Fe]. For [Cr/Fe] they find conflict-ing trends depending on which sample they use. Thissuggests a need for more observational data to betterconstrain the gradients in these elements.0
Donor et al. R GC (kpc) − . − . . . . [ F e / H ] R GC < ± GC > ± N APOGEE stars
Figure 7.
The full high quality sample Galactic [Fe/H] versus R GC trend, with a 2-line fit (described by Eq. 1). Clustersflagged with quality “0” are shown as light blue circles. The color bar indicates the number of member stars per cluster,saturating at 5. For [Mn/Fe], a significant negative trend of -0.015 ± ± R GC < ∼ “Odd-z” Gradients There are three other APOGEE elements that donot readily fall into the above categories: sodium, alu-minum, and potassium , often referred to as “odd-z” elements. We note that while [P/Fe] abundancesare reported in DR16, there are serious doubts aboutthe reliability of the abundances for this element (seeJ¨onsson et al. in prep ). Figure 10 shows the Galac-tic trends for [Na/Fe], [Al/Fe], and [K/Fe]. [Al/Fe] and[K/Fe] show nearly identical significant positive gradi-ents, while [Na/Fe] shows a significant negative gradient.All three trends have at least one significant outlier, butthe trends nevertheless appear fairly robust. Yong et al.(2012) find a similar trend for [Al/Fe] of 0.03 ± THE EVOLUTION OF GALACTICABUNDANCE GRADIENTSMinchev et al. (2019) discuss the effect that sampleselection can have on measured abundance gradients, inparticular the bias introduced by most samples contain-ing a majority of young clusters. To more accuratelycompare to previous work, and provide more meaning-ful comparisons for galactic evolution models, in thissection we compare mono-age samples.8.1.
Iron
Our sample is large enough that it can be split intofour age bins, which we divide at 400 Myr, 800 Myr, and2 Gyr, with all bins being reasonably well populated.Figure 11 shows the [Fe/H] versus R GC trend for clustersseparated in age bins. We use ages from the MWSCcatalog because they are derived in a uniform fashion,and should certainly be reliable enough to place clustersin the coarse bins we have chosen.The evolution of the [Fe/H] vs R GC trend has beenstudied extensively in the literature (e.g. Carraro et al.1998; Friel et al. 2002; Jacobson et al. 2011; Carrera &Pancino 2011; Yong et al. 2012). A summary of resultsfrom the literature is provided in Figure 12. Here we plotthe measured metallicity gradient for clusters in a givenage range vs the middle of that age range (for examplethe middles of our age bins are 0.2, 0.6, 1.4, 4 Gyr). It isimportant to note that the majority of clusters from allfour studies in Figure 12 fall in the range R GC <
14 kpc,with the exception of a few clusters from Carraro et al.(1998). Figure 12 shows a consistent trend of steeper
CCAM: IV. Cluster Abundances with APOGEE DR16 − . . . [ O / F e ] (a) ± − . . . [ M g/ F e ] (b) ± − . . . [ S i / F e ] (c) -0.001 ± − . . . [ S / F e ] (d) ± − . . . [ C a/ F e ] (e) ± R GC (kpc) − . . . [ T i / F e ] (f) -0.000 ± N APOGEE stars
Figure 8.
The [X/Fe] vs R GC trend for α elements. Asbefore the color bar indicates number of member stars, sat-urating at 5. metallicity gradients for older populations. There is onepoint in disagreement with this trend: the oldest clus-ters from Carraro et al. (1998) appear to reverse thistrend. This may be due to the inclusion of some clus-ters near R GC ≈
15 in their oldest bin. If we considerthe large uncertainties on the two oldest measurements,it is possible the trend levels out after 4 Gyr.It should be mentioned that the trend found here is op-posite that seen for field stars (e.g., Anders et al. 2017),where the oldest populations show a shallower gradi-ent. Radial migration is expected to cause this flatten-ing of the metallicity gradient on a long enough timescale (e.g., Minchev et al. 2018). To explain the absenceof this phenomenon in open clusters Anders et al. (2017)suggest that clusters that do not migrate or clusters thatmigrate towards the inner Galaxy preferentially breakup. − . . . [ V / F e ] (a) ± . . [ C r / F e ] (b) ± − . . . [ M n / F e ] (c) -0.015 ± − [ C o/ F e ] (d) -0.010 ± − . . . [ N i / F e ] (e) -0.002 ± R GC (kpc) − . . . [ C u / F e ] (f) ± N APOGEE stars
Figure 9.
The [X/Fe] vs R GC trend for iron-peak elements.Light blue circles are clusters that have an [X/Fe] abundancereported but σ [X/Fe] ≥ . In Figure 13 we show the OCCAM IV sample plottedwith the pure chemical evolution model of Chiappini(2009) and the chemo-dynamical simulation of Minchevet al. (2013, 2014, MCM), divided in the same age binsas Figure 11. There is good agreement between the mod-els and the OCCAM IV sample in the younger threebins. In the oldest bin the effects of radial migrationare clearly seen in the MCM points. Also in the old-est bin, there is a noticeable lack of clusters towardsthe inner galaxy, and a clear steepening of the gradient,which could be due to migration of inner old clusterstowards outer regions. This is consistent with the sug-gestion from Anders et al. (2017) that clusters migratinginward preferentially break up. Elsewhere, the clustersare roughly consistent with the MCM model.8.2.
Other Elements Donor et al. − . . . [ N a/ F e ] (a) -0.025 ± − . . [ A l / F e ] (b) ± R GC (kpc) − . . . [ K / F e ] (c) ± N APOGEE stars
Figure 10.
The [X/Fe] vs R GC trend for the ”odd-z”elements reported in APOGEE DR16. As before, the colorbar indicates number of members and light blue circles areclusters with very high uncertainty in that element. − . . . [ F e / H ] (a) a ≤ ± − . . . [ F e / H ] (b) < a ≤ ± − . . [ F e / H ] (c) < a ≤ ± R GC (kpc) − . . [ F e / H ] (d) > a Gyr-0.094 ± N APOGEE stars
Figure 11.
The Galactic [Fe/H] vs R GC trend in 4 agebins, showing the general decrease in steepness over time. Age trends in elements other than iron also provideinsight into the chemical evolution of the Galaxy. Thetop panel of Figure 14 provides a summary of abun-dance gradients for each element presented previouslyas a function of cluster age, measured in the same fourage bins as for iron (Figure 11). The top panel of Fig-ure 14 shows an overall similar behavior for all elements:the gradient for the oldest population (open clusters − . − . − . − . − . − . − . − . S l o p e ( ∆ [ F e / H ] / ∆ R G C ) Carraro et al. 1998Jacobson et al. 2011Carrera & Pancino 2011This study
Figure 12.
A summary of Galactic metallicity gradientsmeasured in mono-age populations from the literature. older than ∼ α -element abun-dance gradients have also been found for young B stars(e.g., Daflon & Cunha 2004) and H II regions (e.g., Es-teban et al. 2015). We note that for [V/H] the youngestbin is populated with only five clusters, while for [Co/H]the gradient in the youngest population is heavily influ-enced by a single very [Co/H]-poor cluster.8.3. The Evolution of [X/Fe] Gradients
To understand the differences in the evolution of ele-mental abundances better, it is also informative to studythe evolution of [X/Fe] gradients over time. The bottompanel of figure 14 is similar to the top panel but now weshow the evolution of [X/Fe] trends. A variety of trendscan be seen; some elements show a stable trend over time(e.g., [Ni/Fe], [Si/Fe]), some show an increasingly posi-tive trend (e.g., [Al/Fe]), and [Mn/Fe] shows an increas-ingly negative trend. All of these trends are worth dis-cussing and we do so below. We do not consider [V/Fe],[Cr/Fe], [Co/Fe], or [Cu/Fe] in detail, either because theuncertainties are larger than the trends or because oneor more age bins are poorly populated for that element.We do not discuss [Ni/Fe] further because, as stated in § α /Fe] gradients within the time spannedby this open cluster sample. It could be argued thatthere are mild trends with age for the [O/Fe], [Mg/Fe], CCAM: IV. Cluster Abundances with APOGEE DR16 Figure 13.
OCCAM IV clusters (red) plotted with the pure chemical evolution model of Chiappini (2009) (blue line) and theMCM chemo-dynamical simulation (Minchev et al. 2013, 2014), seperated into the age bins used previously. [ O / H ][ O / H ] [ N a/ H ][ N a/ H ] [ M g/ H ][ M g/ H ] [ A l / H ][ A l / H ] [ S i / H ][ S i / H ] [ S / H ][ S / H ] [ K / H ][ K / H ] [ C a/ H ][ C a/ H ] [ T i / H ][ T i / H ] [ V / H ][ V / H ] [ C r / H ][ C r / H ] [ M n / H ][ M n / H ] [ C o/ H ][ C o/ H ] [ F e / H ][ F e / H ] [ N i / H ][ N i / H ] [ C u / H ][ C u / H ] − . − . − . − . − . − . − . . G r a d i e n t( d e x / k p c ) a ≤ < a ≤ < a ≤ > a Gyr [ O / F e ] [ N a/ F e ] [ M g/ F e ] [ A l / F e ] [ S i / F e ] [ S / F e ] [ K / F e ] [ C a/ F e ] [ T i / F e ] [ V / F e ] [ C r / F e ] [ M n / F e ] [ C o/ F e ] [ F e / H ] [ N i / F e ] [ C u / F e ] − . − . − . − . . . . G r a d i e n t( d e x / k p c ) o f c l u s t e r s Figure 14.
Gradients measured in four age bins as for Figure 11 are plotted for each element. The points increase in sizefrom youngest to oldest; the color indicates number of clusters used to measure each gradient. and [Ca/Fe] gradients, but the changes between differ-ent aged populations are on the order of the uncertain-ties. For [Si/Fe], [S/Fe], and [Ti/Fe] there are morevariations, but also larger uncertainties and it is moreappropriate to consider the gradients as roughly con-stant for different aged populations. It was shown inFigure 8 that nearly all of the [ α /Fe] abundances ex-hibit mildly increasing radial trends and Figure 14 indi-cates that such trends appear to be fairly stable within the time spanned by our cluster sample. The flatten-ing of the abundance gradients in recent times suggestsmore recent chemical enrichment in the outer Galaxy,but, taken together with the stability of the increasing[ α /Fe] gradient, we might deduce that the enrichmentin the outer Galaxy had a more significant contributionfrom core-collapse supernovae. This is consistent withthe conclusions from § Donor et al. supernovae Ia dominated recent enrichment in the innerGalaxy.For [Na/Fe] and [Al/Fe], the gradients for the old-est clusters are clearly set apart, even considering thesizeable uncertainty. For [Al/Fe], in particular, thereappears to be a clear trend where we see the youngerpopulations showing an increasingly positive slope. Sig-nificantly larger Na and Al yields are expected from corecollapse supernovae than SNe Ia (Nomoto et al. 2013),so a flattening of the [Na/Fe] gradient and an increas-ingly positive [Al/Fe] gradient are both consistent witheither more recent star formation in the outer Galaxythan the inner Galaxy or higher SNe Ia efficiency in theinner Galaxy. This is also consistent with the explana-tion for the [Mn/Fe] gradient in § CONCLUSIONSWe present a sample of 128 open clusters, 71 of whichwe designate “high quality”, using APOGEE DR16. Wedemonstrate that DR16 cluster abundances are in goodagreement with those of other high resolution abundancestudies. Using the high quality sample, we measureGalactic abundance gradients in 16 chemical elements,and we measure how those gradients change for differentage samples.We find an overall Galactic [Fe/H] vs R GC gradientof − . ± .
004 dex kpc − for R GC < . R GC = 13 . α elements. We present further evidence forthe negative [Mn/Fe] vs R GC trend first found in OC-CAMII. We find significant Galactic trends in vanadium,chromium, and copper, although we are unable to sug-gest a strong explanation for these trends. We find verysignificant trends in sodium, aluminum, and potassium;so-called “odd-Z” elements. We recognize a need for fur- ther study of trends in these elements as they are notwell reported in the literature.We divide our sample into four age bins and inves-tigate changes in 16 elements over time. We show that[X/H] abundance gradients for all 16 elements follow thesame general trend, becoming more shallow over time, ashas consistently been found for iron. We further investi-gate age trends in [X/Fe] for 15 elements. A number ofthese trends seem to support a similar conclusion: eitherincreased SNe Ia efficiency towards the inner Galaxy orless recent star formation in the inner Galaxy comparedto the outer Galaxy.ACKNOWLEDGMENTSWe would like to thank Marina Kounkel for very help-ful discussions about young clusters. We would alsolike to thank Jos´e G. Fern´andez-Trincado and Borja An-guiano for helpful comments.JD and PMF acknowledge support for this researchfrom the National Science Foundation (AST-1311835& AST-1715662). KC acknowledges support for thisresearch from the National Science Foundation (AST-0907873).DAGH acknowledges support from the State Re-search Agency (AEI) of the Spanish Ministry of Sci-ence, Innovation and Universities (MCIU) and the Eu-ropean Regional Development Fund (FEDER) undergrant AYA2017-88254-P.D.G. and D.M. gratefully acknowledge support fromthe Chilean Centro de Excelencia en Astrof´ısica y Tec-nolog´ıas Afines (CATA) BASAL grant AFB-170002.D.G. also acknowledges financial support from the Di-reccin de Investigaci´on y Desarrollo de la Universidad deLa Serena through the Programa de Incentivo a la In-vestigaci´on de Acadmicos (PIA-DIDULS). D.M. is alsosupported by the Programa Iniciativa Cientifica Mileniogrant IC120009, awarded to the Millennium Instituteof Astrophysics (MAS), and by Proyecto FONDECYTregular No. 1170121.H. J. acknowledges support from the Crafoord Foun-dation, Stiftelsen Olle Engkvist Byggm¨astare, and Ruthoch Nils-Erik Stenb¨acks stiftelse.A. Roman-Lopes acknowledges financial support pro-vided in Chile by Comisi´on Nacional de Investi-gaci´on Cient´ıfica y Tecnol´ogica (CONICYT) throughthe FONDECYT project 1170476 and by the QUIMALproject 130001Funding for SDSS-III has been provided by the AlfredP. Sloan Foundation, the Participating Institutions, theNational Science Foundation, and the U.S. Department CCAM: IV. Cluster Abundances with APOGEE DR16
Gaia
Gaia
Gaia
Multilateral Agreement.This research made use of Astropy, a community-developed core Python package for Astronomy (AstropyCollaboration, 2018).
Facilities:
Sloan (APOGEE), FLWO:2MASS,
Gaia
Software:
Astropy ,REFERENCES
Anders, F., Chiappini, C., Minchev, I., et al. 2017, A&A,600, A70, doi: 10.1051/0004-6361/201629363Anguiano, B., Majewski, S. R., Allende-Prieto, C., et al.2018, A&A, 620, A76, doi: 10.1051/0004-6361/201833387Bailer-Jones, C. A. L., Rybizki, J., Fouesneau, M.,Mantelet, G., & Andrae, R. 2018, AJ, 156, 58,doi: 10.3847/1538-3881/aacb21Blanton, M. R., Bershady, M. A., Abolfathi, B., et al. 2017,AJ, 154, 28, doi: 10.3847/1538-3881/aa7567Bowen, I. S., & Vaughan, A. H., J. 1973, ApOpt, 12, 1430,doi: 10.1364/AO.12.001430Bragaglia, A., Sestito, P., Villanova, S., et al. 2008, A&A,480, 79, doi: 10.1051/0004-6361:20077904 Brogaard, K., VandenBerg, D. A., Bruntt, H., et al. 2012,A&A, 543, A106, doi: 10.1051/0004-6361/201219196Burger, D., Stassun, K. G., Pepper, J., et al. 2013,Astronomy and Computing, 2, 40,doi: 10.1016/j.ascom.2013.06.002Cantat-Gaudin, T., Jordi, C., Vallenari, A., et al. 2018,A&A, 618, A93, doi: 10.1051/0004-6361/201833476Carraro, G., Ng, Y. K., & Portinari, L. 1998, MNRAS, 296,1045, doi: 10.1046/j.1365-8711.1998.01460.xCarrera, R., & Pancino, E. 2011, A&A, 535, A30,doi: 10.1051/0004-6361/201117473Carrera, R., Bragaglia, A., Cantat-Gaudin, T., et al. 2019,A&A, 623, A80, doi: 10.1051/0004-6361/201834546 Donor et al.
Casamiquela, L., Blanco-Cuaresma, S., Carrera, R., et al.2019, MNRAS, 490, 1821, doi: 10.1093/mnras/stz2595Chiappini, C. 2009, in IAU Symposium, Vol. 254, TheGalaxy Disk in Cosmological Context, ed. J. Andersen,Nordstr¨oara, B. m, & J. Bland -Hawthorn, 191–196Cunha, K., Frinchaboy, P. M., Souto, D., et al. 2016,Astronomische Nachrichten, 337, 922,doi: 10.1002/asna.201612398Cunha, K., Smith, V. V., Hasselquist, S., et al. 2017, ApJ,844, 145, doi: 10.3847/1538-4357/aa7bebCutri, R. M., Skrutskie, M. F., van Dyk, S., et al. 2003,2MASS All Sky Catalog of point sources.Daflon, S., & Cunha, K. 2004, ApJ, 617, 1115,doi: 10.1086/425607Dias, W. S., Alessi, B. S., Moitinho, A., & L´epine, J. R. D.2002, A&A, 389, 871, doi: 10.1051/0004-6361:20020668Donor, J., Frinchaboy, P. M., Cunha, K., et al. 2018, AJ,156, 142, doi: 10.3847/1538-3881/aad635Esteban, C., Garc´ıa-Rojas, J., & P´erez-Mesa, V. 2015,MNRAS, 452, 1553, doi: 10.1093/mnras/stv1367Foreman-Mackey, D., Hogg, D. W., Lang, D., & Goodman,J. 2013, PASP, 125, 306, doi: 10.1086/670067Friel, E. D., Jacobson, H. R., & Pilachowski, C. A. 2010,AJ, 139, 1942, doi: 10.1088/0004-6256/139/5/1942Friel, E. D., Janes, K. A., Tavarez, M., et al. 2002, AJ, 124,2693, doi: 10.1086/344161Frinchaboy, P. M., Thompson, B., Jackson, K. M., et al.2013, ApJL, 777, L1, doi: 10.1088/2041-8205/777/1/L1Gaia Collaboration, Brown, A. G. A., Vallenari, A., et al.2018, ArXiv e-prints. https://arxiv.org/abs/1804.09365Gaia Collaboration, Prusti, T., de Bruijne, J. H. J., et al.2016, A&A, 595, A1, doi: 10.1051/0004-6361/201629272Garc´ıa P´erez, A. E., Allende Prieto, C., Holtzman, J. A.,et al. 2016, AJ, 151, 144,doi: 10.3847/0004-6256/151/6/144Gunn, J. E., Siegmund, W. A., Mannery, E. J., et al. 2006,AJ, 131, 2332, doi: 10.1086/500975Hasselquist, S., Shetrone, M., Cunha, K., et al. 2016, ApJ,833, 81, doi: 10.3847/1538-4357/833/1/81Holtzman, J. A., Shetrone, M., Johnson, J. A., et al. 2015,AJ, 150, 148, doi: 10.1088/0004-6256/150/5/148Holtzman, J. A., Hasselquist, S., Shetrone, M., et al. 2018,AJ, 156, 125, doi: 10.3847/1538-3881/aad4f9Jacobson, H. R., Pilachowski, C. A., & Friel, E. D. 2011,AJ, 142, 59, doi: 10.1088/0004-6256/142/2/59Jacobson, H. R., Friel, E. D., J´ılkov´a, L., et al. 2016, A&A,591, A37, doi: 10.1051/0004-6361/201527654Janes, K. A. 1979, ApJS, 39, 135, doi: 10.1086/190568 Kharchenko, N. V., Piskunov, A. E., Schilbach, E., R¨oser,S., & Scholz, R.-D. 2013, A&A, 558, A53,doi: 10.1051/0004-6361/201322302Kounkel, M., Covey, K., Su´arez, G., et al. 2018, AJ, 156,84, doi: 10.3847/1538-3881/aad1f1Lindegren, L., Hernandez, J., Bombrun, A., et al. 2018,ArXiv e-prints. https://arxiv.org/abs/1804.09366Luo, A. L., Zhao, Y.-H., Zhao, G., et al. 2015, Research inAstronomy and Astrophysics, 15, 1095,doi: 10.1088/1674-4527/15/8/002Magrini, L., Randich, S., Kordopatis, G., et al. 2017, A&A,603, A2, doi: 10.1051/0004-6361/201630294Majewski, S. R., Schiavon, R. P., Frinchaboy, P. M., et al.2017, AJ, 154, 94, doi: 10.3847/1538-3881/aa784dM´esz´aros, S., Allende Prieto, C., Edvardsson, B., et al.2012, AJ, 144, 120, doi: 10.1088/0004-6256/144/4/120Minchev, I., Chiappini, C., & Martig, M. 2013, A&A, 558,A9, doi: 10.1051/0004-6361/201220189—. 2014, A&A, 572, A92,doi: 10.1051/0004-6361/201423487Minchev, I., Anders, F., Recio-Blanco, A., et al. 2018,MNRAS, 481, 1645, doi: 10.1093/mnras/sty2033Minchev, I., Matijevic, G., Hogg, D. W., et al. 2019,MNRAS, 487, 3946, doi: 10.1093/mnras/stz1239Nidever, D. L., Holtzman, J. A., Allende Prieto, C., et al.2015, AJ, 150, 173, doi: 10.1088/0004-6256/150/6/173Nomoto, K., Kobayashi, C., & Tominaga, N. 2013,ARA&A, 51, 457,doi: 10.1146/annurev-astro-082812-140956Reddy, A. B. S., Lambert, D. L., & Giridhar, S. 2016,MNRAS, 463, 4366, doi: 10.1093/mnras/stw2287Sestito, P., Bragaglia, A., Randich, S., et al. 2008, A&A,488, 943, doi: 10.1051/0004-6361:200809650Souto, D., Cunha, K., Smith, V. V., et al. 2018, ArXive-prints. https://arxiv.org/abs/1803.04461Souto, D., Allende Prieto, C., Cunha, K., et al. 2019, ApJ,874, 97, doi: 10.3847/1538-4357/ab0b43Twarog, B. A., Ashman, K. M., & Anthony-Twarog, B. J.1997, AJ, 114, 2556, doi: 10.1086/118667von Hippel, T., Jefferys, W. H., Scott, J., et al. 2006, ApJ,645, 1436, doi: 10.1086/504369Wilson, J. C., Hearty, F. R., Skrutskie, M. F., et al. 2019,PASP, 131, 055001, doi: 10.1088/1538-3873/ab0075Wright, E. L., Eisenhardt, P. R. M., Mainzer, A. K., et al.2010, AJ, 140, 1868, doi: 10.1088/0004-6256/140/6/1868Yamaguchi, H., Badenes, C., Foster, A. R., et al. 2015,ApJL, 801, L31, doi: 10.1088/2041-8205/801/2/L31Yong, D., Carney, B. W., & Friel, E. D. 2012, AJ, 144, 95,doi: 10.1088/0004-6256/144/4/95
CCAM: IV. Cluster Abundances with APOGEE DR1617