Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud
W. Narloch, G. Pietrzy?ski, W. Gieren, A. E. Piatti, M. Górski, P. Karczmarek, D. Graczyk, K. Suchomska, B. Zgirski, P. Wielgórski, B. Pilecki, M. Taormina, M. Ka?uszy?ski, W. Pych, G. Hajdu, G. Rojas García
AAstronomy & Astrophysics manuscript no. aa © ESO 2021February 1, 2021
Metallicities and ages for 35 star clusters and their surroundingfields in the Small Magellanic Cloud
W. Narloch (cid:63) , G. Pietrzy´nski , W. Gieren , A. E. Piatti , , M. Górski , , P. Karczmarek , D. Graczyk ,K. Suchomska , B. Zgirski , P. Wielgórski , B. Pilecki , M. Taormina , M. Kałuszy´nski , W. Pych , G. Hajdu and G. Rojas García Univesidad de Concepción, Departamento de Astronomia, Casilla 160-C, Concepción, Chilee-mail: [email protected] Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences, Bartycka 18, 00-716, Warsaw, Poland Instituto Interdisciplinario de Ciencias Básicas (ICB), CONICET-UNCUYO, Padre J. Contreras 1300, M5502JMA, Mendoza,Argentina Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Godoy Cruz 2290, C1425FQB, Buenos Aires, Argentina Nicolaus Copernicus Astronomical Center, Polish Academy of Sciences, Rabia´nska 8, 87-100 Toru´n, PolandReceived ; Accepted 14 January 2021
ABSTRACT
Aims.
In this work we study 35 stellar clusters in the Small Magellanic Cloud (SMC) in order to provide their mean metallicities andages. We also provide mean metallicities of the fields surrounding the clusters.
Methods.
We used Strömgren photometry obtained with the 4.1 m SOAR telescope and take advantage of ( b − y ) and m Results.
The spatial metallicity and age distributions of clusters across the SMC are investigated using the results obtained by Ström-gren photometry. We confirm earlier observations that younger, more metal-rich star clusters are concentrated in the central regions ofthe galaxy, while older, more metal-poor clusters are located farther from the SMC center. We construct the age–metallicity relationfor the studied clusters and find good agreement with theoretical models of chemical enrichment, and with other literature age andmetallicity values for those clusters. We also provide the mean metallicities for old and young populations of the field stars surround-ing the clusters, and find the latter to be in good agreement with recent studies of the SMC Cepheid population. Finally, the Strömgrenphotometry obtained for this study is made publicly available.
Key words.
Methods: observational – techniques: photometric – galaxies: individual: Small Magellanic Cloud – galaxies: starclusters: general – galaxies: abundances
1. Introduction
The Small Magellanic Cloud (SMC) is a dwarf irregular galaxy;together with the Large Magellanic Cloud (LMC) it forms a pairof interacting satellites of the Milky Way. Because of its proxim-ity, it is an ideal environment for various astrophysical studies, ofwhich the chemical evolution is one of the most crucial. Clustersserve as tracers of the chemical evolution of the galaxies. Thederived metallicities and ages of stellar clusters generally fol-low the age–metallicity relation (AMR) of a given galaxy, whichallows us to follow the chemical enrichment process of the en-vironment and draw conclusions about the galactic history. Thismakes the AMR an important tool for understanding the chemi-cal evolution of the galaxies.There are several methods for determining stellar metallici-ties. Spectroscopy is a very good tool for obtaining the metallic-ities of stars; the best for this purpose are high-resolution spectrawith wide spectral range and high signal-to-noise ratio. How-ever, it is challenging to get good-quality high-resolution spectraof stars from nearby galaxies. An alternative is the use of low-resolution spectra and the Ca II triplet (CaT). There have beenseveral spectroscopic studies of the SMC star clusters in the liter-ature based on CaT. Da Costa & Hatzidimitriou (1998) obtained (cid:63) contact author spectra of individual red-giant-branch (RGB) stars in seven SMCclusters and calculated the mean metallicities for six of them.They found abnormally low metallicities for Lindsay 113 andNGC 339, suggesting that they have di ff erent origins, possiblybeing formed from infalling unenriched gas, in contrast to therest of their studied clusters. Carrera et al. (2008) determinedmetallicities of over 350 RGB stars in 13 fields distributed acrossthe SMC, and for the first time found a spatial metallicity gradi-ent in this galaxy. The average metallicity of the innermost fieldswas about − − .
94 dex on theCG97 scale. Furthermore, they also found that the mean ageand metallicity of clusters older than 3 Gyr are 5.8 Gyr and − .
08 dex, while for clusters younger than 3 Gyr these val-ues were 1.6 Gyr and − .
85 dex, respectively, thus confirmingthe previous findings. They also refuted the hypothesis of DaCosta & Hatzidimitriou (1998) regarding the anomalous natureof Lindsay 113 and NGC 339. Parisi et al. (2014) further im-
Article number, page 1 of 22 a r X i v : . [ a s t r o - ph . GA ] J a n & A proofs: manuscript no. aa proved the AMR from their first work using more accurate pho-tometry obtained for clusters from their sample and utilized it todetermine the cluster ages.Mighell, Sarajedini & French (1998) used archival HubbleSpace Telescope (HST) data to study the color–magnitude di-agrams (CMDs) of seven SMC clusters. For this purpose theyapplied two methods, and adopted weighted mean metallicitiesfrom both approaches. The first technique was the simultane-ous reddening and metallicity (SRM) method, which takes asinput the magnitude of the horizontal branch (HB), the colorof the RGB at the level of HB, the shape described by eithera quadratic relation or higher order polynomial, and the positionof the RGB. As a result, the SRM provides simultaneous metal-licity and reddening determination. The second method uses thefact that the RGB slope steepens with decreasing metallicity.This dependency can be calibrated for specific colors. Mighell,Sarajedini & French (1998) present this calibration for V versus(B-V), while Mucciarelli et al. (2009) presented these relationsfor near-infrared
JHK bands in four SMC clusters. This methodreturns metallicity for a given reddening.The filters C and T of the Washington photometric sys-tem are very e ff ective for metallicity and age studies (Piatti etal. 2005, 2007a,b; Piatti 2011, 2012). The di ff erence in T magnitude between the red clump (RC) stars and the main se-quence turn-o ff point (MSTO) allows us to determine the agesof stellar populations. Metallicity can be estimated by compar-ing the shape of the RGBs of stellar clusters with published stan-dard fiducial globular cluster RGBs. However, this technique re-quires an age-dependent correction to metallicities derived forintermediate-age objects, which is the case for the majority ofSMC clusters (Parisi et al. 2009).Metallicity values of stellar clusters can be also obtained di-rectly by fitting theoretical isochrones to the CMDs. For exam-ple, Perren, Piatti & Vázquez (2017) developed the AutomatedStellar Cluster Analysis (ASteCA) package, which calculates thesynthetic CMD that best matches the observed cluster CMD fora given set of fundamental parameters (metallicity, age, distancemodulus, reddening, and mass).Metallicity also can be calculated based on calibration be-tween [Fe / H], ( b − y ) and m = ( v − b ) − ( b − y ) Strömgrencolors (Hilker 2000; Dirsch et al. 2000). This relation is welldefined for red stars within a certain range of ( b − y ) colors. Themain advantage of this method is that metallicities can be ob-tained for a large number of individual stars. An early successfulapplication of this method was performed by Grebel & Richtler(1992), among others, for studies of NGC 330. The relation usedin that work was later extended toward lower metallicities byHilker (2000), and adopted by Dirsch et al. (2000) for studiesof six LMC clusters and their surrounding fields obtained with1.54 m Danish Telescope placed in La Silla Observatory, Chile.Calamida et al. (2007) introduced a calibration for red giant starsfrom old Galactic globular clusters based on ( v − y ) and ( u − y )colors, which have stronger sensitivity to e ff ective temperaturethan ( b − y ). Livanou et al. (2013) presented metallicity and agedeterminations for 15 LMC and 8 SMC star clusters based onthe Strömgren data from the Danish Telescope. More recently,Piatti (2018) employed Strömgren photometry from the 4.1 mSOAR telescope (Cerro Pachón, Chile) to investigate four SMCintermediate-age clusters, in their search for hints of multiplestellar populations. The work of Piatti et al. (2019) presents de-rived metallicities of yellow and red supergiants in nine youngLMC and four SMC clusters. In a recent study, Piatti (2020)used metallicity calibration for ( v − y ) − m ff ect on Strömgren metallic- ities, in the sense that younger clusters appear to be more metalpoor, and the di ff erence is a quadratic function.Motivated by these results, we decided to determine metallic-ities, based on Strömgren photometry, for stars belonging to 35star clusters and their surrounding fields from the SMC. We alsoestimated the ages of stellar clusters in our sample using theoret-ical isochrones. We then constructed the AMR, which allowed usto trace the chemical evolution of the SMC. We also present herephotometric measurements for stars from our fields. The metal-licities of some clusters calculated from data presented in thiswork have already been published (e.g., Piatti 2018; Piatti et al.2019; Piatti 2020). We decided to reanalyze them for a severalreasons. In this work we apply the metallicity calibration pre-sented by Hilker (2000), as it is calibrated for a wide range ofmetallicities, and therefore can be used for variety of stellar clus-ters. To this end, we used direct color–color transformation equa-tions to the standard system, which lowers the errors of the co-e ffi cients compared to the calibration method presented by Piattiet al. (2019), among others. Moreover, we rephotometrized im-ages and standardized two of the chips of the camera separately,which is a further improvement. We also used proper motionsto reject foreground stars and adopted the new reddening valuescoming from the recently published reddening maps of the Mag-ellanic Clouds (Górski et al. 2020; Skowron et al. 2020), aswell as positions and sizes of clusters from the updated catalogof Bica et al. (2020).This paper is organized as follows. In Section 2 we describeour observations and data reduction pipeline, as well as the se-lection of stars used for cluster and field metallicity calculation,the adopted reddenings, the metallicity determination procedurebased on the two-color Strömgren diagram, and the age estima-tion procedure. In Section 3 we present the results obtained forthe distribution of the metallicities of clusters and their surround-ing fields in the SMC, as well as the distribution of the clusters’ages and the resulting AMR. In Section 4 we compare our AMRwith those found in the literature. Finally, in Section 5 we sum-marize our results and draw the conclusions of this work.
2. Observations and data reduction
The optical images of fields containing star clusters in the SMCin three Strömgren filters ( v , b and y ) were collected within theAraucaria Project (Gieren, Pietrzy´nski, & Bresolin 2005) duringsix nights on 4.1 m Southern Astrophysical Research Telescope(SOAR) placed in Cerro Pachón in Chile, equipped with theSOAR Optical Imager (SOI) camera (program ID: SO2008B-0917, PI: Pietrzy´nski). Observing nights were divided into tworuns. The first was on 17, 18, and 19 December 2008 and thesecond on 16, 17, and 18 January 2009. The SOI is a mosaiccamera composed of two E2V 2k ×
4k CCDs (read by four am-plifiers). The field of view is 5.26 × at a pixel scaleof 0.077 arcsec · pixel − . During observations 2 × · pixel − . Weobserved 29 fields with star clusters in the SMC in total, wherefields with NGC 330, NGC 265, and NGC 376 were observedtwice. Single images were taken in the air mass range 1 . − . / ALLSTAR package (Stetson 1987) us-ing a Gaussian function with spatial variability to define the pointspread function (PSF). In the case of dense fields, images wereadditionally divided into smaller overlapping subframes to fur-
Article number, page 2 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud ther reduce the PSF and background variability. The PSF modelwas constructed from about 30 to over 200 stars depending onthe stellar density of a given image. The master list of stars ina given frame was obtained iteratively by gradually decreasingthe detection threshold, and in the last iteration inspected by eyeto manually add stars omitted in the automatic procedure. Theaperture corrections for each frame were calculated using theDAOGROW package (Stetson 1990), and instrumental CMDswere constructed. The average errors of the photometry were0.02 mag in V and ( b − y ) and 0.04 mag in m V <
20 mag. Figure 1 illustrates the precision of ourphotometry on an example of the NGC 330 observed during thefirst and third night.The magnitudes and colors of stars were standardized foreach chip of the camera separately using the following trans-formation equations: y inst = V std + a + a · ( b − y ) std + a · ( X y − . , ( b − y ) inst = b + b · ( b − y ) std + b · ( X ( b − y ) − . , m inst = c + c · ( b − y ) std + c · ( X m − . + c · m std , where y inst , ( b − y ) inst , and m inst are instrumental magnitude andcolors; V std , ( b − y ) std and m std are standard magnitude and colorsfrom the Paunzen (2015) catalog; X is airmass; and a i , b i , c i aretransformation coe ffi cients summarized in Table 2. The typicalerror of the transformation is lower than 0.02 mag and is givenin Col. 8 of Table 2, also illustrated as the spread of points inFigure 2. The astrometric solutions for the images in the y filterwere obtained based on Gaia DR2 catalog (Gaia Collaboration2016, 2018b) with subarcsec accuracy.We tested the completeness of our photometry by perform-ing artificial star tests. We used the ADDSTAR routine of theDAOPHOT package to add randomly generated artificial starsto each image in y filter. Their number was about 5% of thenumber of stars found in a given frame. We created 20 such im-ages for each subframe of each field and performed the profilephotometry on them. Next, we added the same list of stars toimages in b and v filters and repeated the procedure. We cal-culated the retrieval rate of artificial stars. The results in everycluster are roughly similar. The retrieval rate for stars brighterthan V =
13 mag is about 86% in the y filter, 93% in b , andalmost 100% in v . The lower number of retrieved stars in the y filter might be due to the overexposition of stars. Finally, for starsbetween 13 and 19 mag, the range for which we perform metal-licity calculations, completeness is about 100% in all filters andthen starts to drop, reaching practically zero for stars fainter than V =
22 mag.
An e ffi cient way to separate cluster members from non-membersis via the proper motions (PMs) of stars. We cross-matched ourmaster lists of stars with the Gaia DR2 catalog. Unfortunately,the accuracy of Gaia PMs for stars from our fields turned out tobe insu ffi cient for reliable membership determination. The aver-age PM error is about 1 mas · yr − in RA and 0.93 mas · yr − inDEC, which correspond to about 296 km · s − and 275 km · s − ,if distance to SMC of 62.44 kpc is used (Graczyk et al. 2020).Nevertheless, we used Gaia data to reject galactic foregroundstars having significant values of PMs. To determine obviousSMC non-members we used a similar approach to that describedin Narloch et al. (2017), among others, where stars are rejectedbased on their location on the vector point diagram. For all stars
12 14 16 18 20 22V [mag]0.00.10.20.3 V [ m a g ] NGC330 (field 1)NGC330 (field 2)12 14 16 18 20 22V [mag]0.00.10.20.3 ( b y ) [ m a g ]
12 14 16 18 20 22V [mag]0.00.10.20.3 m [ m a g ] [ F e / H ] C [ d e x ]
17 18 19V [mag]101 [ F e / H ] F [ d e x ]
14 15 16V [mag]0.10.00.1 [ F e / H ] b F [ d e x ] Fig. 1.
Comparison of photometric precision in two fields of NGC 330captured during the nights of 17 and 19 December 2008. The three up-per panels show the photometric errors from DAOPHOT for V , ( b − y )and m ff erences in metallicitiesdetermined for the first and second field for clusters, and old and youngfield giants, respectively. N1: (b-y) std r e s V (b-y) std r e s ( b - y ) (b-y) std r e s m N2: (b-y) std r e s V (b-y) std r e s ( b - y ) (b-y) std r e s m N3: (b-y) std r e s V (b-y) std r e s ( b - y ) (b-y) std r e s m N4: (b-y) std r e s V (b-y) std r e s ( b - y ) (b-y) std r e s m N5: (b-y) std r e s V (b-y) std r e s ( b - y ) (b-y) std r e s m N6: (b-y) std r e s V (b-y) std r e s ( b - y ) (b-y) std r e s m Fig. 2.
Residuals of the transformation to the standard system. Filledcircles are for chip 1 and open circles for chip 2. in a given cluster field, we calculated mean values and standarddeviations of their PMs, µ (M α , M δ , S α , S δ ), and PM errors, σ µ (ME α , ME δ , SE α , SE δ ). We did not divide stars into magnitudebins because often there were too few stars in a given bin. Next,we selected only stars satisfying the conditions µ ≤ · S and σ µ ≤ ME + · S E and repeated the procedure. This way galac-tic foreground stars with high PMs were removed from the inputlists.
Article number, page 3 of 22 & A proofs: manuscript no. aa
In the next step, we selected stars enclosed in a certain ra-dius as cluster members. We adopted equatorial coordinates andradii of clusters from updated catalog of Bica et al. (2020). Starsoutside this cluster radius were classified as field stars.We decided to not perform a statistical subtraction of the fieldstars. The number of stars in most clusters is small and they arelocated in dense fields which makes it di ffi cult to do statisticalsubtraction correctly. The small field of view of the camera doesnot provide good statistics for the field stars, and we cannot besure that there are no cluster members among them. On the otherhand, in star clusters located farther from the SMC center insparse fields, the contamination of the field stars is negligible.Once the individual metallicities are derived, they can help todisentangle field and cluster members. In order to correct data for the reddening we used reddeningmaps published recently by Górski et al. (2020, hereafter G20),and Skowron et al. (2020, hereafter S20), both obtained by cal-culating the di ff erence of the observed and intrinsic color of theRC stars in the SMC; G20 used the OGLE-III data set while S20used OGLE-IV data with a much larger field of view. The red-dening values for our fields from S20 are systematically smallerthan those from G20 (see Table 5). We adopted the average ofboth maps ( E ( B − V ) GS ) as the reddening of the stellar clustersand their surrounding fields. The adopted reddenings are inde-pendent of our data. The E ( V − I ) from the S20 maps were con-verted into E ( B − V ) with E ( B − V ) = E ( V − I ) / . σ E ( B − V ) GS = .
016 mag, which was propagated into thesystematic error on the derived metallicities resulting from thereddening.We calculated the reddening values for magnitudes and col-ors using the following equations: A V = . · E ( B − V ), E ( b − y ) = . · E ( B − V ) , and E ( m = − . · E ( B − V )(Schlegel, Finkbeiner & Davis 1998). A convenient property of the Strömgren photometric system,which makes it very useful in stellar astrophysics, is the pos-sibility to obtain the metallicity for individual stars nearly inde-pendent of their age (e.g., Dirsch et al. 2000). For the calcula-tion of the metallicity we adopted a calibration of the Strömgren m − ( b − y ) two-color relation derived by Hilker (2000). This re-lation is valid only in the certain color range 0 . < ( b − y ) < . / H] = m + a · ( b − y ) + a a · ( b − y ) + a , (1)where a = − . ± . , a = . ± . , a = . ± . , a = − . ± . . Errors on the metallicity determination of individual starscan be calculated by performing full error propagation of Equa-tion 1 as (Piatti et al. 2019) σ [Fe / H] = (cid:20)(cid:32) ( b − y ) c σ a (cid:33) + (cid:32) c σ a (cid:33) + (cid:32) ( b − y ) · [Fe / H] c σ a (cid:33) + (cid:32) [Fe / H] c σ a (cid:33) + (cid:32) ( a − a · [Fe / H]) c σ ( b − y ) (cid:33) (2) + (cid:32) c σ m (cid:33) (cid:21) , where c = a · ( b − y ) + a . The first four terms in Equation 2 relate to the systematic er-ror, and the remaining two to the statistical error. The dominanterror in the Strömgren two-color diagram is σ m . Consequently,the metallicity error is larger for more metal-poor stars than forthe more metal-rich correspondents of the same color. The cali-bration of Hilker (2000) was based on stars, with spectroscopicmetallicities on the Zinn & West (1984, hereafter the ZW84scale).In the first step of the metallicity determination we selectedstars from the dereddened (see Section 2.2) color range of 0 . < ( b − y ) < .
1. Moreover, only stars having σ ( b − y ) < . σ m < . m − ( b − y ) relation someof the bluest stars from this color range deviate greatly from thewell-defined relation for redder stars. In addition, the applied cutin color on the blue edge introduces a bias toward metal-poorstars with larger metallicity errors. To work around this prob-lem, and to eliminate deviating stars, we followed Dirsch et al.(2000) and applied additional selection criteria by drawing a lineperpendicular to the base of the RGB and included only starsredder than this line. In the first iteration the mean and unbi-ased standard deviation for the remaining stars were calculated.Next, we applied 3 σ clipping to reject outliers and recalculatedthe previously obtained values. In the end, the resulting CMDsand m − ( b − y ) relations were examined by eye and singlestars deviating significantly from either of them were rejectedmanually and the final values of the mean and the unbiased stan-dard deviation were obtained. As a statistical error of the meanmetallicity we adopted unbiased standard deviation divided bythe square root of the number of stars used for the calculation.Dirsch et al. (2000) noted that metallicities measured pho-tometrically are very sensitive to assumed reddening, being themajor source of systematic metallicity error. This degeneracy isparticularly important in old clusters and field stars. In youngerclusters the reddening can be determined quite precisely becausethe color of the hot main sequence stars depends weakly on thetemperature and metallicity. Figure 3 shows how [Fe / H] dependson the E ( B − V ) for two star clusters: the older Lindsay 113 andthe younger the NGC 330. The errors shown in the figure werederived by dividing the unbiased standard deviation of the meanmetallicity of a cluster by the square root of the number of starsused for the metallicity calculation. The increase in assumed red-dening of 0 .
01 mag increases the derived metallicity by about0.06 dex for Lindsay 113 and 0.04 dex for NGC 330, so on av-erage by about 0.05 dex. For a typical error on the adopted red-dening (see Section 2.2), this corresponds to σ [Fe / H] ≈ .
08 dexfor all clusters and surrounding fields.
Article number, page 4 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud E ( B V ) [mag] [ F e / H ] [ d e x ] Fig. 3.
Metallicity derived from Equation 1 vs. assumed reddening inLindsay 113 (points) and NGC 330 (squares).
The second source of systematic uncertainty is the precisionof the m b − y ) calibration to the standard system. The cal-ibration errors cause a bias of the corresponding metallicity thatdepends on the color of a star. This e ff ect is more profound forbluer stars, which leads to larger metallicity errors (see Figure 1in Dirsch et al. 2000). As described in Section 2, the r.m.s. er-rors for the calibration of ( b − y ) and m b − y ) col-ors of stars used for the mean metallicity calculation of a givencluster or field. We do the same for m ff erential reddening acrossthe field, which might a ff ect stars with di ff erent colors in a dif-ferent way, resulting in broadening of the metallicity distribu-tion and consequently also the age distribution. Due to the smallfield of view of the SOAR telescope and the resolution of theG20 reddening maps of 3 arcmin, and of the S20 maps of about1.7 arcmin in the central parts of the SMC, decreasing in the out-skirts, we cannot precisely estimate this e ff ect. Consequently, weneglected it.Finally, the total systematic error of a given mean metallicityis composed of reddening and calibration errors added in squaresunder the square root. The statistical and systematic errors forthe cluster and field samples are given in Tables 3 and 4, respec-tively. For the age determination we employed isochrones from theDartmouth Stellar Evolutionary Database (Dotter et al. 2008,hereafter the Dartmouth isochrones) and the Padova database ofstellar evolutionary tracks and isochrones available through theCMD 3.3 interface (Marigo et al. 2017) calculated with thePARSEC (Bressan et al. 2020) and COLIBRI (Pastorelli et al.2019) evolutionary tracks (hereafter the Padova isochrones). TheDartmouth isochrones were available only for the 1 −
15 Gyrisochrones sets, so were too old for most of our clusters. Instead,the Padova isochrones covered all possible ages so most of ourresults are based on these isochrones. Still, in cases where it waspossible, we used both types of isochrones for the comparison.The Dartmouth isochrones seem to better reflect the shape ofred giant branches in older clusters, while the Padova isochronestend to flatten too much near the tip.During the isochrone fitting procedure, we used theisochrone of a specific metallicity determined from the Ström-gren data at a fixed distance to the SMC from Graczyk etal. (2020), where the distance modulus is ( m − M ) SMC = .
977 mag. In a few cases (Lindsay 6, Lindsay 113, NGC361,IC1611) the isochrones of the adopted metallicity did not fit wellto the CMD of the cluster, so we adopted new values for thereddening from the literature, recalculated the metallicity, andrepeated the procedure. The error in age was adopted as a halfof the age di ff erence between two marginally fitting isochronesselected around the best fitting isochrone. Table 6 presents the first five rows of the compiled catalog ofStrömgren photometry for stars from the fields studied in thiswork. The catalog contains stars for which all three Strömgren vby filters were available, and subsequently the color indices( b − y ) and m ffi cients from Table 2, given in a separate column. A com-plete version of Table 6 is available online on the AraucariaProject webpage and at the CDS via anonymous ftp to cdsarc.u-strasbg.fr (130.79.128.5) or via http: // cdsweb.u-strasbg.fr / cgi-bin / qcat?J / A + A /
3. Results
Figures 4–7 present two-color diagrams and CMDs for stellarclusters and their surrounding fields for three cases: clusters withwell-populated RGBs (Figure 4), and young clusters with a fewstars (Figure 5–6) and no stars (Figure 7) for metallicity calcula-tion. Figure 8 shows the spatial distribution of the mean metal-licities of clusters and fields, while Figure 9 similarly shows thedistribution of cluster ages. We summarized our measurementsin Tables 3 (clusters) and 4 (fields).There are seven intermediate-age stellar clusters in our sam-ple of 35 SMC clusters (Lindsay 1, 6, 19, 27, and 113; NGC 339and 361) with well-populated RGBs having between 18 and 93stars used for metallicity determination. These are well studiedobjects having many age and metallicity measurements in theliterature. For their age determination the Dartmouth and the http: // stellar.dartmouth.edu / models / isolf_new.html http: // stev.oapd.inaf.it / cgi-bin / cmd_3.3 https: // araucaria.camk.edu.pl / Article number, page 5 of 22 & A proofs: manuscript no. aa
Padova isochrones were both used. A further six clusters in oursample (NGC 330 and 265; OGLE-CL SMC 45, 69, and 88;IC 1611) are young stellar clusters with between 4 and 12 redgiants, which is su ffi cient for reliable metallicity determination.Only the Padova isochrones were used for their age determina-tion. We indicate these 13 clusters in Figure 10 with squares.Twelve of the young star clusters studied in this work had lessthan four, but there was at least one star lying within the clus-ter radius and fulfilling the criteria for metallicity determina-tion. To the best of our knowledge, the metallicities for OGLE-CL SMC 68, 71, 82, 126, 143, and [BS95] 123 are providedfor the first time. NGC 376, IC 1612, Bruck 39, and OGLE-CL SMC 32, 54, and 156 have at least one previous metallic-ity estimation. Only the Padova isochrones were suitable fortheir age determination. Clusters from this group are indicatedin Figure 10 with open circles. Finally, there were ten youngstar clusters (OGLE-CL SMC 49, 50, 61, 78, 99, 128, 129,142, 144, and 205) for which we have not found any suitablestars for cluster metallicity calculation. Moreover, these clus-ters have no published metallicities in the literature, the soleexception being OGLE CL-SMC-49 with one such determina-tion. However, we were able to calculate the mean metallicitiesof the field stars surrounding these clusters. We also estimatedthe ages of the clusters by employing the Padova isochrones for[Fe / H] = − .
70 dex, which is a value close to the mean metal-licity of the SMC (discussed in the next section).
The upper panel in Figure 8 presents a metallicity map of thestar clusters given in Table 3. The black cross indicates the cen-ter of the SMC determined based on the distribution of classicalCepheid variables by Ripepi et al. (2017) ( α = . ± .
01 deg; δ = − . ± .
01 deg). In the map it is clear that the most metal-poor clusters are located in the outskirts of the galaxy, while themore metal-rich ones group close to the SMC central region. Theaverage metallicity of the clusters most distant from the SMCcenter (Lindsay 1, 6, and 113, and NGC 336) is − .
05 dex witha standard deviation ( σ ) of 0.10 dex. The average metallicity ofthree clusters located a bit farther from the elongated central re-gion (NGC 361, Lindsay 19 and 27) is − .
90 dex ( σ = .
13 dex).The average metallicity of all seven outer clusters is − .
99 dex( σ = .
13 dex). Most of the clusters studied in this work arelocated along the denser central region of the SMC. Their av-erage metallicity is − .
70 dex ( σ = .
22 dex). Moreover, wecan distinguish two groups, one positioned close to the SMCcenter to the west (hereafter western group) and a more sparsegroup located where the HI super-shell 304 A is placed (here-after eastern group). We show these two groups in Figure 8 withdotted rectangles. Five clusters from the western group (OGLE-CL SMC 49, 50, 61, 78, and 205) had no suitable stars for themetallicity calculation, so no information about their metallic-ity is available. The average metallicity of ten remaining clus-ters NGC 265; OGLE-CL SMC 32, 45, 54, 68, 69, 71, 82, 88;and Bruck 39) is − .
70 dex ( σ = .
24 dex). In case of eightclusters from the eastern group (NGC 330 and 376; IC 1611;OGLE-CL SMC 126, 143, and 156) the average metallicity is − .
71 dex ( σ = .
19 dex) where clusters OGLE-CL SMC 99,128, 129, 143, and 144 were omitted because we had no infor-mation about their metallicities. The values for the eastern andwestern groups agree within the errors.The lower panel in Figure 8 presents a map of the meanmetallicities of the fields surrounding the studied star clusters given in Table 4. In five cases we studied more than one clus-ter in the field of view, so the field stars from these regions comefrom the area around all present clusters. The average metallicityof the old giants used for the metallicity calculation from the sur-rounding of the three outermost clusters (without Lindsay 113,where there are only a few stars most probably belonging to thecluster) is − .
93 dex ( σ = .
39 dex). For field stars around thethree other clusters located closer to the elongated SMC centralregion it is − .
80 dex ( σ = .
36 dex). The average metallic-ity of old giants surrounding all six intermediate-age clusters is − .
84 dex ( σ = .
37 dex). The average metallicity of the fieldstars used for the metallicity calculation lying in the central re-gions of the SMC is − .
73 dex ( σ = .
45 dex). For the west-ern group it is − .
74 dex ( σ = .
46 dex) and for the eastern − .
71 dex ( σ = .
43 dex), which is a statistically insignificantdi ff erence.The young helium burning giants (HBGs) belonging to thefield are common in the central regions of the SMC, but in theouter fields in our sample they are absent, or nearly so. In thispopulation of stars we can expect to observe Cepheid variables.Ripepi et al. (2017) calculated photometric metallicities for 462Cepheids and found a peak of their metallicity distribution atabout [Fe / H] = − .
60 dex. On the other hand, Romaniello etal. (2008) reported [Fe / H] = − . ± .
02 dex (with dispersionof 0.08 dex) based on high-resolution spectroscopic studies of14 stars from the SMC. Recently, Lemasle et al. (2017) pub-lished metallicities for four SMC Cepheids obtained from high-resolution spectroscopy with average of [Fe / H] = − .
74 dex.The average metallicity of all HBG stars measured by us in thefields around clusters is − .
70 dex ( σ = .
28 dex), which is inexcellent agreement with the mentioned studies. The di ff erencebetween the western ( − .
69 dex, σ = .
26 dex) and the eastern( − .
71 dex, σ = .
32 dex) group is statistically insignificant.
Figure 9 shows the age map of studied clusters in the SMCderived from the Dartmouth (upper panel) and Padova (lowerpanel) isochrones given in Table 3. The rectangles superim-posed onto the plot indicate the same spatial groups of clus-ters discussed in previous subsection. The outermost clustersin our sample are the oldest ones. The logarithm of their agesderived from the Dartmouth isochrones ranges between 9.40 to9.85 (2.5 to 7.0 Gyr) with the average of 9.65 with σ = . ∼ σ = . ∼ σ = .
26 ( ∼
129 Myr) and for the eastern 7.66 with σ = . ∼
46 Myr). The mean age for both groups of central young clus-ters is 7.90 with σ = .
38 ( ∼
79 Myr). The spatial age distributionwe obtained closely follows the distribution of young star clus-ters presented by Glatt et al. (2010, see their Figure 7) wheremost of the young clusters are located along the central overden-sity and to the east from the center of the SMC where the HIsuper-shell 304 A is placed.
The resulting age–metallicity relation for star clusters studied inthis work is illustrated in Figure 10.
Article number, page 6 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud b y ) m b y )1617181920212223 V NGC 339E(B-V) C = 0.041 mag[Fe/H] C = -1.10 dexlog(Age D ) = 9.78 +0.070.08 Gyrlog(Age P ) = 9.75±0.08 yr0.5 0.6 0.7 0.8 0.9 1.0 1.1( b y ) m b y )1617181920212223 V FieldE(B-V) F = 0.041 mag[Fe/H] F = -1.05 dex1.751.501.251.000.750.500.250.00 [ F e / H ] N [ F e / H ] Fig. 4.
Reddening corrected two-color diagrams and unreddened CMDs for NGC 339 (upper panels) and its surrounding field stars (lower panels).Left panels: Stars having photometry in vby filters (black points); stars excluded from metallicity determination (open squares); stars used tocalculate the mean metallicity of a cluster and field (color-coded points, where colors represent metallicity distribution); lines of constant metallicity(dashed lines); additional selection criteria drawn after visual inspection of the plot (dotted line); obtained mean metallicities of cluster andfield stars (black solid lines); the statistical and systematic errors of the mean metallicity of the cluster (darker and lighter shaded areas). Thearrows indicate the reddening vectors. Right panels: Dartmouth and Padova best-fitting isochrones (gray and turquoise dashed lines, respectively)superimposed onto the field CMD, in order to illustrate the position of the cluster against field stars.
In order to compare our results with theoretical predictions ofthe chemical evolution of the SMC, we considered a few modelspublished in the literature. The model of Pagel & Tautvaišvien˙e(1998, hereafter the PT98 model,) predicts intensive star for-mation and chemical enrichment in the SMC during the initialepoch that brought the metallicity of the galaxy up to about − . ffi cient. Thenext rise in the mean star formation rate occurred during recent3 Gyr with bursts at ages of 2.5 Gyr, 400 Myr, and 60 Myr,where the first two bumps are consistent with past perigalacticpassages by the SMC with the Milky Way. The major mergerscenario for the SMC was proposed by Tsujimoto & Bekki Article number, page 7 of 22 & A proofs: manuscript no. aa b y ) m b y )121416182022 V NGC 330E(B-V) C = 0.075 mag[Fe/H] C = -0.98 dexlog(Age P ) = 7.50±0.10 yr0.5 0.6 0.7 0.8 0.9 1.0 1.1( b y ) m b y )121416182022 V FieldE(B-V) F = 0.075 mag[Fe/H] F = -0.81 dex[Fe/H] bF = -0.89 dex1.501.251.000.750.500.250.000.25 [ F e / H ] N [ F e / H ] Fig. 5.
Reddening corrected two-color diagrams and unreddened CMDs for NGC 330 (upper panels) and surrounding field stars (lower panels).Triangles indicate the young field giants. The meaning of other symbols is the same as in Figure 4. (2009, hereafter the TB09 model). It predicts that a major mergeroccurred ∼ ff erent formation history is proposedby Cignoni et al. (2013) through the two Bologna models, (C13-B) and Cole (C13-C). They predict fast initial enrichment priorto 9 Gyr ago, then monotonic increase in metallicity between 9 to4 Gyr with no evidence of metallicity dips, followed by anotherenrichment at more recent times.Our AMR presented in Figure 10 in general is very consis-tent with the models. The older clusters Lindsay 19, 27, and113 closely follow the burst of star formation at ∼ ff er-ent from the distance we use in the present study. However, a dif-ferent distance would change the age, although not the metal-licity. Another possibility is that these clusters are located in re-gions richer in gas and could be chemically enriched during theirformation or evolution. The HZ04 model is on average too metalrich for the discussed clusters, which is also true for other litera-ture values derived from photometric or spectroscopic data (seeFigure 11). Only values provided by Perren, Piatti & Vázquez(2017) closely follow this model at ages older than 2.5 Gyr.Da Costa & Hatzidimitriou (1998) described Lindsay 113 andNGC 339 as anomalous, and suggested that they could have beenformed from the infall of unenriched, or less enriched, gas. How- Article number, page 8 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud b y ) m b y )121416182022 V SMC 68E(B-V) C = 0.06 mag[Fe/H] C = -0.97 dexlog(Age P ) = 7.74±0.14 yr0.5 0.6 0.7 0.8 0.9 1.0 1.1( b y ) m b y )121416182022 V FieldE(B-V) F = 0.06 mag[Fe/H] F = -0.73 dex[Fe/H] bF = -0.77 dex2.01.51.00.50.0 [ F e / H ] N [ F e / H ] Fig. 6.
Reddening corrected two-color diagrams and unreddened CMDs for OGLE-CL SMC 68 (upper panels) and surrounding field stars (lowerpanels). Triangles indicate the young field giants. The meaning of other symbols is the same as in Figure 4. ever, Parisi et al. (2009) showed that these two clusters behavewell, but in turn Lindsay 1 appears too metal rich with respectto the PT98 model. We do not see any anomalous behavior ofLindsay 113 or NGC 339, but Lindsay 1 indeed seems to be toometal rich. There is a discrepancy between the ages we deter-mined based on the Dartmouth and Padova databases. For clus-ters older than about 3 Gyr the Dartmouth isochrones give olderages than the Padova isochrones, and for clusters younger than3 Gyr the e ff ect is the opposite. Moreover, it seems that the olderthe cluster is, the larger the age di ff erence becomes for these twosets of isochrones.Clusters younger than 1 Gyr in our sample fit equally wellall models of chemical evolution, although they appear to besystematically less metal rich than predictions. The results forthese clusters are less reliable due to small number of stars usedfor the metallicity calculation. Parisi et al. (2009) indicated that their metallicity for NGC 330 is significantly more metal poorthan the PT98 model prediction. This is also true for our case.NGC 330 deviates greatly from the AMR, although the num-ber of stars used for its metallicity measurement is su ffi cientfor a reliable mean metallicity determination. Furthermore, themetallicity for NGC 376 is extremely low in our AMR. This re-sult however is based on only one star in the first field, and twostars in the second field. Apart from these two outliers, otheryoung star clusters are still characterized by quite large spreadin the metallicity. However, overall compliance with the theoret-ical predictions is satisfactory.
4. Discussion
Figure 11 shows our AMR with superimposed literature valuesfrom Table 5: metallicities derived from spectroscopy (low- and
Article number, page 9 of 22 & A proofs: manuscript no. aa b y ) m b y )121416182022 V SMC 99E(B-V) C = 0.052 mag[Fe/H] ad = -0.70 dexlog(Age P ) = 7.65±0.03 yr0.5 0.6 0.7 0.8 0.9 1.0 1.1( b y ) m b y )121416182022 V FieldE(B-V) F = 0.052 mag[Fe/H] F = -1.06 dex[Fe/H] bF = -1.03 dex2.01.51.00.50.0 [ F e / H ] N [ F e / H ] Fig. 7.
Reddening corrected two-color diagrams and unreddened CMDs for OGLE-CL SMC 99 (upper panels) and surrounding field stars (lowerpanels). Triangles indicate the young field giants. The meaning of other symbols is the same as in Figure 4. The turquoise dashed line is a Padovabest-fitting isochrone for the adopted metallicity and metallicity of young field giants. high-resolution), Strömgren photometry. and RGB slope methodall expressed on the ZW84 scale, as well as values obtained byPerren et al. (2017) with the ASteCA package, and from theo-retical isochrone fitting obtained by various authors. Figure 12presents the comparison between the derived metallicities andtheir literature counterparts.The metallicities and ages of intermediate-age star clustersin our sample are on average more metal rich and, consequently,younger than most literature values expressed on the same metal-licity scale. The possible explanation is the adopted calibra-tion of Strömgren colors with metallicity. Data points basedon Strömgren photometry from Livanou et al. (2013), Piatti(2018), and Piatti et al. (2019) were obtained with di ff erent cal-ibrations (Hilker, Richtler & Gieren 1995; Grebel & Richtler1992; Calamida et al. 2007, respectively). Hilker (2000) pre- sented an extended calibration of Grebel & Richtler (1992)for cluster and field red giants for a wide range of metallici-ties ( − . < [Fe / H] < . . − . − . < [Fe / H] < − . Article number, page 10 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud h m h m h m h m h m RA-74-73-72-71 D e c METALLICITY(CLUSTERS) -0.71 dex-0.70 dex -0.99 dex [ F e / H ] C [ d e x ] h m h m h m h m h m RA-74-73-72-71 D e c METALLICITY(FIELD POPULATION) -0.71 dex-0.74 dex -0.84 dex [ F e / H ] F [ d e x ] Fig. 8.
Metallicity map of the field (upper panel) and cluster (lowerpanel) stars in the SMC. The black cross indicates the center of theSMC at α = .
54 deg; δ = − .
11 (Ripepi et al. 2017). North is up;east is left. study as it is calibrated for the widest range of metallicities, andit can be used for both old metal-poor and young metal-rich gi-ants and supergiants. This choice motivated us to reanalyze thedata sets published by Piatti (2018) and Piatti et al. (2019). Tokeep the uniformity in the analysis, we redid the photometry andcalibrated the data using direct color–color transformation equa-tions for the two chips separately, which was an improvement asthe zero points of the two chips of the camera di ff er slightly. Forthe selection of the cluster members we used the updated catalogof Bica et al. (2020) and the applied additional step of rejectingforeground stars based on their Gaia proper motions. For most ofthe overlapping clusters (except two) we adopted di ff erent red-dening values from new reddening maps of G20 and S20. Thesechanges resulted in the selection of di ff erent stars for metallic-ity calculation, which is especially visible in the case of youngclusters. The resulting discrepancies between our findings andother studies based on Strömgren photometry are the smallestfor metal-poor clusters ( ∼ ∼ − .
70 dex.The metallicities from low-resolution spectroscopy from DaCosta & Hatzidimitriou (1998) are significantly more metal poorthan our results (by ∼ ∆ [Fe / H] ≈ − .
06 dex); Spite, Richtler & Spite 1991; Hill1997; Gonzalez & Wallerstein 1999). Data points from the RGB h m h m h m h m h m RA-74-73-72-71 D e c CLUSTER AGES(DARTMOUTH) ~9.65 (~4.5 Gyr) l o g ( A g e D ) [ y r ] h m h m h m h m h m RA-74-73-72-71 D e c CLUSTER AGES(PADOVA) ~7.66 (~46 Myr)~8.11 (~129 Myr) ~9.62 (~4.2 Gyr) l o g ( A g e P ) [ y r ] Fig. 9.
Age map of the cluster stars in the SMC derived from the Dart-mouth (upper panel) and Padova (lower panel) theoretical isochrones.The black cross indicates the center of the SMC as in Figure 8. North isup; east is left. slope method from Mighell, Sarajedini & French (1998) are alsosystematically more metal poor than our results (by ∼ ∆ [Fe / H] ≈− .
33 dex). The overall agreement of our results with literaturevalues from isochrones fitting and Washington photometry aresatisfactory ( ∆ [Fe / H] ≈ − .
11 dex and 0.09 dex, respectively).The conclusion of Livanou et al. (2013) was that thereis no indication of an AMR in the SMC. We do not confirmthat statement. Despite of a handful of low-metallicity youngstar clusters like NGC 330, NGC 376, and OGLE-CL SMC 68and 69, the average metallicity of young clusters is − .
70 dexwith σ = .
22 dex, while for older ones it is − .
99 dex with σ = .
13 dex.Piatti et al. (2007a) noticed the tendency that the clusterslocated in the inner regions of the SMC are younger than thosefrom outer regions, while their mean metallicity and its disper-sion are greater close to the SMC center. We confirm both ofthese observations (see Figure 8). Furthermore, Piatti (2012)pointed out that stellar populations younger than about 2 Gyrare more metal rich than [Fe / H] ≈ − . Article number, page 11 of 22 & A proofs: manuscript no. aa
Age [Gyr] [ F e / H ] [ d e x ] PT98 modelclosed box model, -0.6 dexclosed box model, -0.5 dexHZ04 modelC13-B modelC13-C modelTB09-1 modelTB09-2 modelTB09-3 model log(Age) [yr] [ F e / H ] [ d e x ] PT98 modelclosed box model, -0.6 dexclosed box model, -0.5 dexHZ04 modelC13-B modelC13-C modelTB09-1 modelTB09-2 modelTB09-3 model
Fig. 10.
Age–metallicity relation for clusters studied in this work. Clus-ters with ages derived using Darthmouth isochrones (black squares);clusters with reliable number of stars for metallicity determination(open squares), having ages derived from the Padova isochrones; clus-ters with 1 − − . − . axis (cid:46) ◦ . We find that on average the younger stellar popula-tions from this region have metallicities of about or less than − . < − ∼ Age [Gyr] [ F e / H ] [ d e x ] PT98 modelclosed box model, -0.6 dexclosed box model, -0.5 dexHZ04 modelC13-B modelC13-C modelTB09-1 modelTB09-2 modelTB09-3 model log(Age) [yr] [ F e / H ] [ d e x ] PT98 modelclosed box model, -0.6 dexclosed box model, -0.5 dexHZ04 modelC13-B modelC13-C modelTB09-1 modelTB09-2 modelTB09-3 model
Fig. 11.
Age–metallicity relation for clusters studied in this work com-pared with the literature (see Table 5). Values derived from two-colorStrömgren diagram (black diamonds); values derived by Perren, Pi-atti & Vázquez (2017) using the ASteCA package (open circles);low-resolution spectroscopic metallicities expressed in ZW84 scale(gray squares); high-resolution spectroscopic metallicities in ZW84scale (blue squares); values obtained from RGB slope method given inZW84 scale (open triangles); values derived from fitting of theoreticalisochrones to optical data (crosses); values derived from fitting of the-oretical isochrones to data in Washington system (open crosses). Redsquares and open circles indicate measurements from this work pre-sented in Figure 10 for comparison. Overplotted theoretical models areas in Figure 10.
In the conclusions Perren, Piatti & Vázquez (2017) indi-cated that the metallicities obtained with the ASteCA packageare on average ∼
5. Summary and conclusions
In this work we presented the analysis of Strömgren photometryof 35 star clusters from the SMC in order to obtain their meanmetallicities and ages. We also provided mean metallicities ofthe fields surrounding the clusters. Metallicities and ages werederived in a consistent manner by using the relation of photomet-ric metallicity and Strömgren colors calibrated by Hilker (2000),which allowed us to compare the obtained results and trace themetallicity and age distribution across the SMC. Moreover, weused for the calculations the most recent reddening maps of G20and S20, as well as the distance to the SMC derived by Graczyket al. (2020), which is precise to 2%.
Article number, page 12 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud [Fe/H] [dex] [ F e / H ] li t [ d e x ] Fig. 12.
Comparison of the metallicities for star clusters obtained inthis work with the literature values from Figure 11. The red solid linerepresents the 1:1 relation.
The metallicity distribution of the field stars in the SMCshows a trend typical of irregular galaxies. The more metal-richstars tend to accumulate close to the central region of the SMC.The farther away from the SMC main body, the more metalpoor the stars become. The average metallicity values of theyoung field giants and supergiants ( − .
70 dex) and old field stars( − .
73 dex) are similar within the errors. The average metallic-ity of HBG stars found by us is in close agreement with the re-sults reported by Ripepi et al. (2017, − .
60 dex), Romanielloet al. (2008, − .
75 dex), or Lemasle et al. (2017, − .
74 dex)for Classical Cepheids, a subset of such stars. The age distribu-tion of clusters in the SMC confirms earlier studies by Carreraet al. (2008) or Glatt et al. (2010), among others, that youngstellar clusters distribute along the SMC main body while theintermediate-age clusters are located farther from it.The two features described above are reflected in the AMRconstructed for studied stellar clusters. The majority of clus-ters analyzed in this paper are young, with very few stars forreliable metallicity and age calculation. Only seven clusters inour sample have well populated RGBs, where the majority ofstars for which Strömgren photometry provides metallicity es-timates are found. The overall results agree well with theoreti-cal models of chemical enrichment of the SMC, and with pre-vious literature studies. The metallicities for seven star clusters(OGLE-CL SMC 68, 71, 82, 88, 126, 143, and Bruck 39) areprovided for the first time, according to our knowledge. Threeintermediate-age clusters (Lindsay 19, 27, and 113) reproducewell the burst of chemical enrichment ∼ ∼ Acknowledgements.
We thank the anonymous referee for valuable commentswhich improved this paper. The research leading to these results has re-ceived funding from the European Research Council (ERC) under the Euro-pean Union’s Horizon 2020 research and innovation program (grant agreementNo 695099). We also acknowledge support from the National Science Center,Poland grants MAESTRO UMO-2017 / / A / ST9 / / / G / ST9 / / WK / /
09 grants of the Polish Ministry ofScience and Higher Education. We gratefully acknowledge financial support forthis work from the BASAL Centro de Astrofisica y Tecnologias Afines (CATA)AFB-170002 and the Millenium Institute of Astrophysics (MAS) of the Ini-ciativa Cientifica Milenio del Ministerio de Economia, Fomento y Turismo deChile, project IC120009. M.G. gratefully acknowledges support from FONDE-CYT POST- DOCTORADO grant 3170703.
References
Bica, E., Dottori, H., Pastoriza, M. 1986, A&A, 156, 261Bica, E., Westera, P., Kerber, L., et al. 2020, AJ, 159, 82Bressan, A., Marigo, P., Girardi, L. et al. 2012, MNRAS, 427, 127Calamida, A., Bono, G., Stetson, P. B. et al. 2007, ApJ, 670, 400Carrera, R., Gallart, C., Aparicio, A. et al. 2008, ApJ, 136, 1039Carretta, E., Gratton, R. G. 1997, A&AS, 121, 95Chantereau, W., Salaris, M., Bastian, N., Martocchia, S., 2019, MNRAS, 484, 4,5236-5244Chiosi, E., Vallenari, A., Held, E. V., et al. 2006, A&A, 452, 179Chiosi, E., Vallenari, A. 2007, ASPC, 374, 275Cignoni, M., Cole, A. A., Tosi, M., et al. 2013, ApJ, 775, 83Da Costa, G. S., Hatzidimitriou, D. 1998, AJ, 115, 1934Dias, B., Coelho, P., Barbuy, B. et al. 2010, A&A, 520, 85Dirsch, B., Richtler, T., Gieren, W. P., Hilker, M. 2000, A&A, 360, 133Dotter, A., Chaboyer, B., Jevremovi´c, D. et al. 2008, ApJS, 178, 89Gaia Collaboration et al., 2016, The Gaia mission. A&A, 595, A1Gaia Collaboration, Brown, A. G. A., Vallenari, A., Prusti, T. et al. 2018b, A&A,616, A1Gieren, W., Pietrzy´nski, G., Bresolin, F. 2005, The Messenger, 121, 23Glatt, K., Grebel, E. K., Sabbi, E. et al. 2008, AJ, 136, 4, 1703-1727Glatt, K., Grebel, E. K., Koch, A. 2010, A&A, 517, 50Glatt, K., Grebel, E. K., Jordi, K. et al. 2011, AJ, 142, 2Gonzalez, G., Wallerstein, G. 1999, AJ, 117, 2286Górski, M., Zgirski, B., Pietrzy´nski, G. et al. 2020, ApJ, 889, 179Graczyk, D., Pietrzy´nski, G., Thompson, I. B., et al. 2020, ApJ, 904, 13Grebel, E. K., Richtler, T. 1992, A&A, 253, 359Harris, J., Zaritsky, D. 2004, AJ, 127, 1531Hilker, M., Richtler, T., Gieren, W. 1995, A&A, 294, 648Hilker, M. 2000, A&A, 355, 994Hill, V. 1997, A&A, 324, 435Kuˇcinskas, A., Dobrovolskas, V., Lazauskait˙e, R., Tanabé, T. 2009, BaltA, 18,225Lemasle, B., Groenewegen M. A. T., Grebel, E. K. et al. 2017, A&A, 608A, 85Livanou, E., Dapergolas, A., Kontizas, M. et al. 2013, A&A, 554, 16Maia, F. F. S., Piatti, A. E., Santos, J. F. C. 2014, MNRAS, 437, 2, 2005-2016Marigo, P., Girardi, L., Bressan, A. et al. 2017, ApJ, 835, 77Martocchia, S., Dalessandro, E., Lardo, C. et al. 2019, MNRAS, 487, 4, 5324-5334Mighell, K. J., Sarajedini, A., French, R. S. 1998, AJ, 116, 2395Milone, A. P., Marino, A. F., Di Criscienzo, M. at al. 2018, MNRAS, 477, 2,2640-2663Mucciarelli, A., Origlia, L., Maraston, C., Ferraro, F. R. 2009, ApJ, 690, 1Nayak, P. K., Subramaniam, A., Choudhury, S., Sagar, R. 2018, A&A, 616, 187Narloch, W., Kaluzny, J., Poleski, R. et al. 2017, MNRAS, 417, 1446Pagel, B. E. J., Tautvaišvien˙e, G. 1998, MNRAS, 299, 535Parisi, M. C., Grocholski, A. J., Geisler, D., et al. 2009, AJ, 138, 51Parisi, M. C., Geisler, D., Carraro, G. et al. 2014, AJ, 147, 4Pastorelli, G., Marigo, P., Girardi, L. 2019, MNRAS, 485, 5666Paunzen, E. 2015, A&A, 580, 23Perren, G. I., Piatti, A. E., Vázquez, R. A. 2017, A&A, 602, 89Piatti, A. E., Sarajedini, A., Geisler, D. et al. 2005, MNRAS, 358, 1215Piatti, A. E., Sarajedini, A., Geisler, D. et al. 2007a, MNRAS, 377, 300Piatti, A. E., Sarajedini, A., Geisler, D. et al. 2007b, MNRAS, 381, 1, L84-L88Piatti, A. E., Geisler, D., Sarajedini, A. et al. 2008, MNRAS, 389, 1, 429-440Piatti, A. E. 2011, MNRAS, 416, 89-93Piatti, A. E. 2012, MNRAS, 422, 1109
Article number, page 13 of 22 & A proofs: manuscript no. aa T a b l e . S t a r c l u s t e r s i n t h e S M C . C l u s t e r : n a m e o f t h ec l u s t e r ; R A , D E C : e qu a t o r i a l c oo r d i n a t e s o f t h ec l u s t e rf o r e po c h J fr o m B i cae t a l . ( ) ; D a t e : d a t e o f ob s e r v a ti on ; T e xp : e xpo s u r e ti m e o f fi lt e r y , b , a nd v ; A i r m a ss : a i r m a ss o f ob s e r v a ti on s ; S ee i ng : a v e r a g e s ee i ng ; O t h e r n a m e : o t h e r n a m e o f t h ec l u s t e r s i nu s e . C l u s t e r R AD E C D a t e T ex p ( y , b , v ) A i r m a ss ( y , b , v ) S ee i ng ( y , b , v ) O t h e r n a m e ( hh : mm : ss . s )( dd : mm : ss . s )( s )( a r c s ec ) L i nd s a y11301 : : . - : : D ec ; ; . ; . ; . . ; . ; . E S O - NG C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y59 , E S O - , K r on36 NG C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y67 , E S O - , K r on46 L i nd s a y100 : : . - : : D ec ; ; . ; . ; . . ; . ; . E S O - L i nd s a y600 : : . - : : D ec ; ; . ; . ; . . ; . ; . E S O - , K r on4 NG C : : . - : : D ec ; ; . ; . ; . . ; . ; . E S O - , K r on35 , L i nd s a y54 , D ec , ; . ; . ; . . ; . ; . OG LE - C L S M C OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y35 , K r on25 L i nd s a y2700 : : . - : : D ec ; ; . ; . ; . . ; . ; . K r on21 , OG LE - C L S M C L i nd s a y1900 : : . - : : D ec ; ; . ; . ; . . ; . ; . OG LE - C L S M C NG C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y34 , OG LE - C L S M C D ec ; ; . ; . ; . . ; . ; . E S O - , K r on24 NG C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y72 , E S O - , K r on492008 D ec ; ; . ; . ; . . ; . ; . OG LE - C L S M C I C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y61 , E S O - , K r on40 OG LE - C L S M C I C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y62 , E S O - , K r on41 OG LE - C L S M C OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . NG C , L i nd s a y30 , E S O - K r on23 B r u c k3900 : : . - : : D ec ; ; . ; . ; . . ; . ; . OG LE - C L S M C OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . [ B S ] OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . NG C , L i nd s a y42 , E S O - OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . B r u c k79 OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y66 OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y74 , E S O - , K r on50 [ B S ] : : . - : : D ec ; ; . ; . ; . . ; . ; . OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . OG LE - C L S M C OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . B r u c k48 OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . L i nd s a y39 OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . [ H ] OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . B r u c k57 OG LE - C L S M C : : . - : : D ec ; ; . ; . ; . . ; . ; . [ H ] OG LE - C L S M C : : . - : : J a n16120 ; ; . ; . ; . . ; . ; . [ H ] OG LE - C L S M C : : . - : : J a n1690 ; ; . ; . ; . . ; . ; . [ B S ] OG LE - C L S M C : : . - : : J a n1690 ; ; . ; . ; . . ; . ; . L i nd s a y46 , K r on31 OG LE - C L S M C : : . - : : J a n1690 ; ; . ; . ; . . ; . ; . L i nd s a y65 , [ H ] OG LE - C L S M C : : . - : : J a n17100 ; ; . ; . ; . . ; . ; . B r u c k105 OG LE - C L S M C : : . - : : J a n1790 ; ; . ; . ; . . ; . ; . [ B S ] OG LE - C L S M C : : . - : : J a n18120 ; ; . ; . ; . . ; . ; . L i nd s a y80 Article number, page 14 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud
Table 2.
Transformation coe ffi cients. Night chip eq. coe ff coe ff coe ff coe ff r.m.s. y ± ± ± b − y ) 0.059 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.059 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.062 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.039 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.063 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.042 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.073 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.072 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.060 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.069 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.072 ± ± ± m ± ± ± ± y ± ± ± b − y ) 0.077 ± ± ± m ± ± ± ± Piatti, A. E. 2018, AJ, 156, 5Piatti, A. E., Pietrzy´nski, G., Narloch W. et al. 2019, MNRAS, 483, 4, 4766-4773Piatti, A. E. 2020, accepted in A&A, arXiv:2008.05270Pietrzy´nski, G., Udalski, A. 1999a, Acta Astron., 49, 157Pietrzy´nski, G., Udalski, A. 1999b, Acta Astron., 49, 435Ripepi, V., Cioni, M-R L., Moretti, M. I. et al. 2017, MNRAS, 472, 808Romaniello, M., Primas, F., Mottini M. et al. 2008, A&A, 488, 731Schlegel, D. J., Finkbeiner, D. P., Davis, M. 1998, ApJ, 500, 525Skowron, D. M., Skowron, J., Udalski, A. et al. 2020, preprintSpite, F., Richtler, T., Spite, M. 1991, A&A, 252, 557Stetson, P. B. 1987, PASP, 99, 191Stetson, P. B. 1990, PASP, 102, 932Tsujimoto, T., Bekki, K. 2009, ApJ, 700, 69Zinn, R., West, M. J. 1984, ApJS, 55, 45
Article number, page 15 of 22 & A proofs: manuscript no. aa
Table 3.
Star clusters in the SMC. Cluster: name of the cluster; D maj ;D min : major and minor axes from the updated catalog of Bica et al. (2020); E ( B − V ) C : reddening adopted for the cluster stars; [Fe / H] C : mean cluster metallicity calculated in this work (systematic errors are given in thebrackets); N C : number of stars used for metallicity calculation; log(Age P ): logarithm of age derived from the Padova isochrones; log(Age D ):logarithm of age derived from the Dartmouth isochrones. Cluster D ma j ;D min E ( B − V ) C [Fe / H] C N C log(Age P ) log(Age D )(arcmin) (mag) (dex) (yr) (yr)Lindsay 113 4.4;4.4 0.03 -1.14 ± ± ± ± ± + . − . NGC 361 2.6;2.6 0.03 -0.79 ± ± + . − . Lindsay 1 4.6;4.6 0.033 -1.06 ± ± + . − . Lindsay 6 1.7;1.7 0.03 -0.91 ± ± + . − . NGC 330 2.8;2.5 0.075 -0.98 ± ± ± ± ± ± ± + . − . Lindsay 19 1.7;1.7 0.062 -0.86 ± ± + . − . NGC 265 1.2;1.2 0.075 -0.75 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± Notes. * Metallicities of the clusters provided for the first time. Not used for the AMR because of the lower quality.
Article number, page 16 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud
Table 4.
Fields surrounding star clusters in the SMC. Field: name of the cluster in the field; E ( B − V ) F : Reddening adopted for the field stars;[Fe / H] F : mean metallicity of the field stars (systematic errors are given in the brackets); [Fe / H] bF : mean metallicity of the young field giants; N F ,N bF : number of stars used for the mean metallicity calculation of old and young giants, respectively. Field E ( B − V ) F [Fe / H] F N F [Fe / H] bF N bF (mag) (dex) (dex)Lindsay 113 0.058 - - - -NGC 339 0.041 -1.05 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± / IC 1612 0.085 -0.51 ± ± / Bruck 39 0.083 -0.52 ± ± ± ± ± ± ± ± ± ± / [BS95] 123 / SMC 144 0.070 -0.69 ± ± ± ± ± ± ± ± ± ± / SMC 205 0.062 -0.86 ± ± ± ± / SMC 88 0.057 -0.76 ± ± ± ± ± ± ± ± ± ± Table 6.
Strömgren photometry of fields in the SMC.
RA DEC X Y Field V σ V DAO σ V ( b − y ) σ ( b − y ) DAO σ ( b − y ) (deg) (deg) (pixel) (pixel) (mag) (mag) (mag) (mag) (mag) (mag)27.531676 -73.741133 2.534 666.291 Lindsay113 20.324 0.021 0.027 0.326 0.039 0.04227.530038 -73.752099 12.939 408.817 Lindsay113 21.318 0.045 0.049 0.439 0.064 0.06727.528188 -73.712040 26.633 1349.382 Lindsay113 14.134 0.006 0.015 -0.050 0.009 0.01427.527827 -73.758950 27.267 247.924 Lindsay113 22.112 0.092 0.093 0.081 0.113 0.115(...) m σ m DAO σ m CHI SHARP(mag) (mag) (mag)0.171 0.161 0.178 0.727 0.055-0.116 0.117 0.132 0.763 -0.0670.060 0.016 0.024 3.371 0.3510.408 0.174 0.194 0.807 -0.724(...)
Notes.
A complete table is presented in its entirety in the electronic form on the Araucaria Project webpage and the CDS. A portion is shown herefor guidance regarding its form and content. Article number, page 17 of 22 & A proofs: manuscript no. aa
Appendix A: Comments on individual clusters
There were a few cases when the reddening value adopted fora given star cluster as a mean of G20 and S20 was unsuitable forage determination. The isochrones for calculated metallicitieswere not fitting well the observed CMDs. In case of Lindsay 113, E ( B − V ) GS = .
058 mag from S20 turned out to be too high.The isochrone of [Fe / H] = − .
01 dex was not fitting well si-multaneously the subgiant and red giant branch, suggesting thata steeper isochrone for lower metallicity is needed. To that end,we adopted E ( B − V ) = .
03 mag often used in the literature.Similar situation was in case of NGC 361 and Lindsay 1 having E ( B − V ) GS = .
08 mag and 0.049 mag, respectively, for whichalso E ( B − V ) = .
03 mag was adopted. Also in case of IC 1611having E ( B − V ) GS = .
085 mag the isochrone of − .
19 dex didnot fit well the main sequence. We adopted E ( B − V ) = .
03 maginstead which resulted in [Fe / H] = − .
58 dex.Perren, Piatti & Vázquez (2017) reported[Fe / H] = − .
01 dex for OGLE-CL SMC 45 while Piatti etal. (2019) give − .
85 dex for E ( B − V ) = .
03 mag. For E ( B − V ) GS = .
063 mag we got [Fe / H] = − .
31 dex. Thedi ff erences between this work and Piatti et al. (2019) using thesame data could be caused by the choice of di ff erent stars forthe mean metallicity calculation, photometric and calibrationerrors, as well as the use of di ff erent two-color calibration.NGC 376 is too metal-poor compared to the literature. Thiscluster was observed twice during two subsequent nights. Forits metallicity calculation we have chosen the same two stars asused in Piatti et al. (2019). The reported mean metallicity of thecluster is − .
55 dex which is much higher value than what weobtained in this work ( − .
98 dex). One of the stars is lying in thevery center of the cluster and its photometry could be a ff ectedby the dense environment. Its metallicity agrees well betweenthe two nights though ( − . ± .
28 dex and − . ± .
27 dex inthe second and third night, respectively). The other chosen starwas measured only in the image from the third night, because onthe second it fell into the gap between the chips. We obtained forit [Fe / H] = − . ± .
25, which is also more metal-poor than inPiatti et al. (2019). The isochrone of − .
77 dex seems to fit wellthe CMD of NGC 376, suggesting that it could be more metal-rich and simultaneously younger (log(Age P ) ≈ .
50 yr) than theage of best-fitting isochrone of − .
98 dex. Livanou et al. (2013)have only one star in this region of the CMD and in their case itis very metal-poor.The S20 reddening map gives very small reddening valuesfor OGLE-CL SMC 78 and 82, very di ff erent than G20 and otherauthors. We used the average of S20 and G20 for metallicity cal-culations, but higher reddening values resulting in higher metal-licities and ages would be equally acceptable for these clusters. Article number, page 18 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud T a b l e . R e dd e n i ng s , m e t a lli c iti e s a nd a g e s o f s t a r c l u s t e r s fr o m t h e lit e r a t u r e . N a m e E ( B − V )[ F e / H ] A g e l og ( A g e )( m a g )( d e x )( G y r)( y r) L i nd s a y1130 . ± . , . ± . , , - . ± . , - . ± . . ± . , . ± . . ± . , . . , . , . , . - . ± . , - . ± . . ± . , . ± . . a , . b - . ± . , - . ± . . ± . , . ± . - . ± . , - . ± . . ± . , - . ± . , - . ± . - . , - . , - . NG C . , , , . ± . , . - . ± . , - . ± . > , . ± . , . ± . . ± . , . . ± . , . ± . - . ± . , - . ± . . ± . , . , . ± . . , . b - . ± . , , - . ± . . ± . - . , - . ± . - . ± . , - . ± . NG C . , , . ± . , - . ± . , - . ± . , , . ± . , . ± . . . , . , . , . b - . ± . , , - . ± . . ± . , . , . ± . - . ± . , - . ± . . ± . - . ± . L i nd s a y10 . ± . , . , , . - . ± . , , , - . ± . . ± . , . ± . . . b - . ± . , - . ± . , . ± . , . ± . - . ± . , - . ± . . ± . , . ± . , , , - . ± . , - . ± . L i nd s a y60 . , . ± . - . ± . , - . ± . , . ± . , . , . . ± . , . . b , . , . b - . ± . NG C . , , . , . , - . , - . , - . ± . . ± . , . . ± . , . , . . , . a , . b - . ± . , - . ± . . . , . b - . ± . , - . ± . , - . ± . , - . , - . ± . , - . ± . OG LE - C L S M C . , . , . , . - . , - . ± . , . . ± . , , . , . . ± . , . b , . , . b - . ± . . ± . L i nd s a y270 . , . , , . ± . - . ± . , - . ± . , . ± . , . ± . > , . ± . , . . , . b - . ± . . ± . L i nd s a y190 . , . , , . b , - . ± . , - . ± . , . ± . , > , . , . ± . . b , . , . b - . ± . . , . Article number, page 19 of 22 & A proofs: manuscript no. aa T a b l e . c on ti nu e d . N a m e E ( B − V )[ F e / H ] A g e l og ( A g e )( m a g )( d e x )( G y r)( y r) NG C . , . , . - . , - . - . . ± . . ± . , . , . ± . . ± . , . b , . b - . ± . , . ± . , . , . , . , . b . ± . NG C . , . , . , . - . ± . , - . ± . , . ± . , . . ± . , , . , . . ± . , . , . b - . ± . . ± . . , . . b , . , . b I C . , . , . , . - . , - . ± . . ± . . ± . , , . , . . b , . b , . , . b I C . , . b , . , . - . , - . ± . , . ± . , . , , . . ± . , . b , . , . b - . ± . , . , . . b OG LE - C L S M C . , . , . - . ± . , . ± . . ± . , . , , . , . ± . , . b , . b . . , . b B r u c k390 . b , . , . ± . - . ± . , - . ± . , . , . , . . b , . , . b . OG LE - C L S M C . , . , . b , . -- . ± . , . , . . b . OG LE - C L S M C . , . , . b , . b - . - . ± . , . , . , . b . ± . , . ± . , . , . , . OG LE - C L S M C . , . , . b -- . ± . , . , . . b , . , . b . , . OG LE - C L S M C . , . , . b -- . ± . , . , . . , . b . OG LE - C L S M C . , . , . b , . b -- . ± . , . , . . , . b . , . [ B S ] . b , . , . b , . -- . , . , . , . . b Article number, page 20 of 22. Narloch et al.: Metallicities and ages for 35 star clusters and their surrounding fields in the Small Magellanic Cloud T a b l e . c on ti nu e d . N a m e E ( B − V )[ F e / H ] A g e l og ( A g e )( m a g )( d e x )( G y r)( y r) OG LE - C L S M C . , . b , . , . -- . , , . , . . b OG LE - C L S M C . , . , . ± . - . ± . , - . ± . , . , . . b , . , . b . ± . , . OG LE - C L S M C . , . , . b -- . ± . , . , . . , . b . OG LE - C L S M C . , . , . , . ± . - . ± . , - . ± . , . , , . . b , . b , . , . b . ± . , . OG LE - C L S M C . , . b , . , . b -- . ± . , . , . OG LE - C L S M C . , . , . b , . b -- . ± . , . , . . , . b . , . OG LE - C L S M C . b , , . , . b -- . , . , . OG LE - C L S M C . , , . b , . -- . ± . , . , . . b . OG LE - C L S M C . , , . b , . -- . ± . , . , , . . b OG LE - C L S M C . b , . , . b -- . , OG LE - C L S M C . , . , . b -- . ± . , . , . . b , . , . b . , . OG LE - C L S M C . , . , . b -- . ± . , . , . . , . b . OG LE - C L S M C . , . b , . , . b -- . ± . , . , . OG LE - C L S M C . , . , . a , . b - . ± . , - . ± . . ± . . ± . , . , , . b , . , . b . Article number, page 21 of 22 & A proofs: manuscript no. aa N o t e s . * R e f e r e n ce s : ( ) B i cae t a l . ( )( DDO , H β pho t o m e t r y ) ; ( ) S p it e , R i c h tl e r & S p it e ( )( h i gh -r e s o l u ti on s p ec t r o s c opy ) ; ( ) G r e b e l & R i c h tl e r( )( S t r ö m g r e npho t o m e t r y ) ; ( ) D a C o s t a & H a t z i d i m it r i ou ( )( l o w -r e s o l u ti on s p ec t r o s c opy , C a T ) ; ( ) M i gh e ll , S a r a j e d i n i & F r e n c h ( )( H S T pho t o m e t r y , R G B s l op e ) ; ( ) G on za l ez & W a ll e r s t e i n ( )( h i gh -r e s o l u ti on s p ec t r o s c opy ) ; ( ) H ill ( )( h i gh -r e s o l u ti on s p ec t r o s c opy ) ; ( ) P i e t r z y ´ n s k i & U d a l s k i ( a )( op ti ca l pho t o m e t r y ) ; ( ) C h i o s i e t a l . ( )( op ti ca l pho t o m e t r y ) ; ( ) C h i o s i & V a ll e n a r i ( )( H S T pho t o m e t r y ) ; ( ) P i a tti e t a l . ( a , b )( W a s h i ng t onpho t o m e t r y ) ; ( ) G l a tt e t a l . ( )( H S T pho t o m e t r y ) ; ( ) P i a tti e t a l . ( )( W a s h i ng t onpho t o m e t r y ) ; ( ) M u cc i a r e lli e t a l . ( )( N I R pho t o m e t r y , R G B s l op e ) ; ( ) K u ˇ c i n s k a s ( )( N I R pho t o m e t r y , R G B s l op e ) ; ( ) D i a s e t a l . ( )( i n t e g r a t e d s p ec t r o s c opy ) ; ( ) G l a tt e t a l . ( )( op ti ca l pho t o m e t r y ) ; ( ) G l a tt e t a l . ( )( H S T pho t o m e t r y ) ; ( ) P i a tti ( )( W a s h i ng t onpho t o m e t r y ) ; ( ) L i v a nou e t a l . ( )( S t r ö m g r e npho t o m e t r y ) ; ( ) M a i a , P i a tti & S a n t o s ( )( W a s h i ng t onpho t o m e t r y ) ; ( ) P a r i s i e t a l . ( )( l o w -r e s o l u ti on s p ec t r o s c opy , C a T ) ; ( ) P e rr e n , P i a tti & V áz qu ez ( )( W a s h i ng t onpho t o m e t r y , A S t e C A ) ; ( ) M il on ee t a l . ( )( H S T pho t o m e t r y ) ; ( ) N a y a k e t a l . ( )( op ti ca l pho t o m e t r y ) ; ( ) P i a tti ( )( S t r ö m g r e npho t o m e t r y ) ; ( ) C h a n t e r ea u e t a l . ( )( H S T pho t o m e t r y ) ; ( ) M a r t o cc h i ae t a l . ( )( H S T pho t o m e t r y ) ; ( ) P i a tti e t a l . ( )( S t r ö m g r e npho t o m e t r y ) ; ( ) B i cae t a l . ( )( ca t a l og ) ; ( ) G ó r s k i e t a l . ( )( op ti ca l pho t o m e t r y ) ; ( ) S ko w r on e t a l . ( )( op ti ca l pho t o m e t r y ) a E ( b - y ) b E ( V -I)-I)