The Swift/UVOT Stars Survey. III. Photometry and Color-Magnitude Diagrams of 103 Galactic Open Clusters
Michael H. Siegel, Samuel J. Laporte, Blair L. Porterfield, Lea M. Z. Hagen, Caryl A. Gronwall
DDraft version May 14, 2019
Typeset using L A TEX preprint style in AASTeX62
The
Swift
UVOT Stars Survey. III. Photometry and Color-Magnitude Diagrams of 103 GalacticOpen Clusters
Michael H. Siegel, Samuel J. LaPorte, Blair L. Porterfield,
1, 2
Lea M. Z. Hagen,
1, 2, 3 andCaryl A. Gronwall
1, 3 Pennsylvania State University, Department of Astronomy525 Davey LaboratoryUniversity Park, PA, 16802 Space Telescope Science Institute3700 San Martin DriveBaltimore, MD, 21218 Institute for Gravitation and the CosmosThe Pennsylvania State UniversityUniversity Park, PA 16802
ABSTRACTAs part of the Swift/UVOT Stars Survey, we present near-ultraviolet (3000-1700˚A) point-source photometry for 103 Galactic open clusters. These data, taken overthe span of the mission, provide a unique and unprecedented set of near-ultravioletpoint-source photometry on simple stellar populations. After applying membershipanalysis fueled mostly by GAIA DR2 proper motions, we find that 49 of these 103 haveclear precise CMDs amenable to investigation. We compare the CMDs to theoreticalisochrones and find good agreement between the theoretical isochrones and the CMDs.The exceptions are the fainter parts of the main sequence and the red giant branch inthe uvw − uvw uvw a r X i v : . [ a s t r o - ph . S R ] M a y Siegel et al.
Keywords: open clusters:general; stars:general; stars: early-type INTRODUCTIONOpen clusters are gravitationally bound collection of 10 -10 stars born in the same star formationevent. They are, by far, the most abundant family of clusters in the Milky Way with over 1500cluster or cluster candidates having been cataloged (Dias et al. 2002) Galactic open clusters arepredominantly young and metal-rich – most of the older clusters having long since dissolved into theGalactic field. However, a number of old and metal-poor clusters have been detected and serve as“fossils” for studying the early Galaxy.As compact collections of stars of similar age, abundance, distance and reddening, individual starclusters present a snapshot of stellar evolution. Studying large samples of clusters allows us topiece together the narrative of stellar evolution and constrain the photometric properties of stars asfunctions of mass, age and chemistry. Much of our understanding of stellar evolution and photometrycomes from the study of Galactic open and globular star clusters. Cluster of all ages and chemistriesare known to host stars that are bright in the ultraviolet (UV), as shown in Siegel et al. (2014,hereafter Paper I). These UV-bright stars in the nearby universe have counterparts in the UV-brightstars seen in nearby galaxies and are, ultimately, parallels to the stars contributing to the UV lightof distant unresolved stellar populations. Any understanding of luminous hot UV-bright stars andstellar populations dominated by them will therefore have to lay its foundation on the study of nearbystar clusters. They are not only the best repositories in which to find such stars, but having foundthem, the properties can be immediately connected with stellar populations of known age, distance,reddening and chemistry.The study of the UV properties of star clusters is still in its early stages, as detailed in Paper I. Butthere a number of recent results that indicate this is a fertile field of study. For example, studies ofopen clusters in the Magellanic Clouds have shown complex structure in the main sequence turnoff(MSTO) – an extended MSTO (eMSTO) and split MSTO’s (Mackey & Broby Nielsen 2007; Miloneet al. 2009; Goudfrooij et al. 2011). This split has also recently been identified in a number ofGalactic clusters (Marino et al. 2018. Cordoni et al. 2018) This fine structure is thought to thebe the result of stellar rotation, which can affect the evolutionary lifetimes of stars, and may showup more clearly in the UV. A recent comprehensive study of M67 (Sindhu et al. 2018) identifiednumerous blue straggler stars (BSS), white dwarf binaries and stars that had excess far-UV emissionconsistent with either unresolved binarism or chromospheric activity.In addition to their utility for studying stellar evolution and the properties of rare phases of stellarevolution, Galactic open clusters can be used as the building blocks for studying larger astrophysicalquestions. Open clusters, for example, can be used to probe the chemodynamical properties of theGalactic disc (see, e.g., Friel 1995, Twarog et al. 1997, Yong et al. 2005, 2012; Frinchaboy &Majewski 2008). Old open clusters, with ages greater than 1 Gyr, are particularly suited to tracingthe chemical evolution of the disk through their age-metallicity relationship (AMR). Indeed, thesimilarity between the AMR of a small group of open clusters and those of merging dwarf galaxysystems has been key to confirming past Galactic mergers (Frinchaboy et al. 2006). Open clustersmay be particularly useful for probing the UV extinction properties of nearby dust. As we have notedbefore (Paper I, Hagen et al. 2017, 2019), the UV extinction curve is uncertain, both in slope and in VOT Stars: Open Clusters integrated light . By combining these two measures, one cancalibrate the study of more distant stellar populations that cannot be resolved on a star-by-star basisbut can only be studied though the properties of their integrated light (see an illustration of this in thecontext of the Magellanic Clouds in Searle et al. 1980). Studies of the integrated light of unresolvedstellar population have proven extraordinary useful for untangling the star-formation histories anddust extinction properties of both nearby and distance galaxies (see, e.g., Hoversten et al. 2011,Calzetti et al. 2015; Hagen et al. 2017, 2019). However, these studies have a limitation: the UVportion of the spectral-energy distribution is not as well constrained as the optical and infrared. Thelack of empirical information needed to calibrate synthetic spectra of unresolved stellar populationswas specifically cited by Bruzual (2009) as a limitation on the utility of the models.Despite the tremendous efforts made by investigators, most of the Milky Way’s open clusters havenot been studied in great detail at any wavelength. The sheer number of open clusters in the Galaxyprecluded most from being targeted for individual study. Indeed, many of the parameters usedby the WEBDA online database of open clusters come from global optical-infrared (OIR) surveysof hundreds of clusters such as that of Kharchenko et al. (2013 and 2016, hereafter K13 and K16,respectively). But there are additional issues that complicate the study of open clusters, most notablytheir preferred location within the Galactic midplane, with concomitant heavy foreground reddeningand field-star contamination.This deficit of individual study is particularly acute in the UV where very few clusters have beenstudied in detail (see discussion in Paper I). This lacuna is the natural apotheosis of the factors thatlimit their study in the optical: their location within the Galactic midplane makes study difficulteither due to crowding (most previous UV missions had coarse spatial resolution), field star contam-ination, foreground extinction (which is much stronger in the UV) or simple brightness constraintsthat preclude missions like GALEX from studying objects in the Galactic midplane.As a result of this, the spectral synthesis models used to study distant unresolved stellar populationsare still not well-constrained in the UV. And this is particularly problematic because young stellarpopulations – those most visible over the long distances in extragalactic astronomy and cosmology– emit a large fraction of their rest-frame light in the UV. The lack of thorough, systematic andempirical study of young UV-bright stellar populations is a significant and limiting problem for extantand planned ultraviolet surveys of the Milky Way, the Magellanic Clouds, Local Group objects, nearbygalaxies and distant galaxies . Understanding the star-by-star and integral photometry of all thesedistant and extragalactic systems – from the nearby to the deep universe – is critically dependent onstudies of nearby open clusters.In this contribution, we address this deficit in our knowledge of Galactic open clusters with asurvey of 103 open clusters observed in the near-ultraviolet (NUV) with the Neil Gehrels SwiftMission’s Ultraviolet-Optical Telescope (UVOT). By combining Swift’s wide-field moderate resolution Siegel et al. photometry with precise astrometry from the recent GAIA DR2 catalog, we are able to study theseclusters in unprecedented detail, testing the utility of theoretical isochrones and either confirming orrevising the literature parameters for these clusters.We present the observational details of the program in §
2. We then detail the limitations on ouranalysis produced by saturation of the brightest stars and contamination from the Galactic disk in §
3. Using some radial velocities but mostly GAIA DR2 proper motions to establish membership, weare able to identify 49 clusters that have clear and distinct color-magnitude sequences, many of whichhave never been the target of detailed study. We present these color-magnitude diagrams (CMDs)along with comparison to theoretical isochrones, with which we measure or revise their fundamentalproperties. The resultant test of the theoretical isochrones shows excellent agreement, indicatingthat the models perform well in the UV. We show that the isochrones have a tendency to not matchthe photometry of cooler stars – red giant branch (RGB) and late-type main sequence (MS) stars.We finally look ahead to other uses for this dataset, which we will provide to the community. OBSERVATIONS AND DATA REDUCTIONS2.1.
UVOT Data
UVOT is a modified Richey-Chretien 30 cm telescope that has a wide (17’ × u , b and v optical filters, uvw uvm uvw . (cid:48)(cid:48) Swift ’s X-Ray Telescope (XRT; Burrows et al. 2005) allow a broad range ofscience, including the study of hot or highly energetic stars. It is particularly ideal, in the contextof hot stars, for studying nearby star clusters. Its wide field can enclose most nearby open clustersin a single pointing and its resolution allows measurement of stars almost to the center of the mostcrowded fields. The NUV filters – with effective wavelengths of 2600, 2246 and 1928 ˚A for uvw uvm uvw
2, respectively, provide good coverage of the NUV wavelength range and sensitivityto the 2175 ˚A blue bump.The open clusters surveyed by UVOT are drawn from a variety of programs. The bulk wereobserved as part of M. Siegel’s approved Cycle 12 GI and team fill-in programs, observed optimallyfor 6 ks in UVOT’s three NUV filters (although, in practice, the observing time tended to be closerto 5 ks). Others were observed as calibration targets or coincidentally with other studies. We haveonly included clusters in the sample if they were observed for at least 0.8 ks in two of UVOT’s threeNUV filters and where the field center was within 5’ of the nominal cluster center, although someexceptions have been made to expand the sample. The clusters in our GI and fill-in program weregenerally selected to be comparable to the UVOT field-of-view, with low reddening ( E ( B − V ) < . u passband. All exposure times are given as the maximumexposure time on the combined image, which removes deadtime losses. Most of the cluster images VOT Stars: Open Clusters Table 1.
Swift/UVOT Observations of Open Clusters
Cluster u Exp uvw1 Exp uvm2 Exp uvw2 Exp ObservationTime (ks) Time (ks) Time (ks) Time (ks) DatesBlanco 1 0 3378 2461 3071 2010-11-10 – 2011-01-25NGC 188 1195 2842 3111 4555 2005-06-29 · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · · were created by combining any image within 7 . (cid:48) b and v when available.They have been filtered of objects with DAOPHOT (Stetson 1987, 1994) structural parameters of | SHARP | > . uvw uvm uvw uvm uvw t <
100 Myr). The reason for this is thatvery young clusters tend to be either far too bright or far too enshrouded in dust for UVOT study. ANALYSIS AND RESULTS3.1.
Bright Star Issues
Most of our cluster fields are close to or within the Galactic midplane. They also tend to be young( t < m − M < − , the limit at which UVOT can no longer count incident photons (corresponding toAB magnitudes of 12.52, 12.11 and 12.68 in the uvw uvm uvw Siegel et al. for incident photons will result in dark boxes around the saturated cores of bright stars (Poole et al.2010). This lowers the limit at which useful photometry can be measured to well below 372 sec − .This is particularly noticeable with PSF photometry as very high coincidence loss both changes theshape of the PSF for bright stars and can cause DAOPHOT to underestimate the background skylevel. In the discussions below, the phrase “saturated” stars will be applied to both contexts andhas the general meaning of a star whose coincidence loss is too strong for reliable PSF photometry.While Page et al. (2017) recently outlined a method to measure photometry from saturated starsusing readout streaks, this method was not appropriate for most of our saturated stars as it wouldnot produce the precision needed to constrain theoretical isochrones. We have noted, within the text,several clusters where the bright end of the MS is saturated and we can only provide upper limits onthe age of the clusters. Two clusters that were observed for our program – NGC 2547 and NGC 2516– are not included in the sample of 103 presented here because of bright star contamination problemsthat made photometry of any kind impossible.The other complication of bright stars is that they produce diffuse light which may affect photometry(see extended discussion in the context of M 31 in Hagen et al. 2019). This can include internalreflection, donuts and read-out lines. However, while scattered light is a problem for the photometryof extended sources, we have not found it to be a problem for point sources , which are measuredfrom a local sky. Our comparison of UVOT photon-counting photometry to ground-based CCDphotometry in Paper I showed that our methods produces reliable linear photometry for all but thebrightest stars. 3.2. Cluster Membership
While some of the clusters in our sample – such as M 67 – are above the Galactic plane, thevast majority are within the plane ( | b | < § VOT Stars: Open Clusters Figure 1.
GAIA DR2 vector-point (VPD) and Swift/UVOT color-magnitude diagrams of the old opencluster M 67 compared the PARSEC-COLIBRI isochrones fit in Paper I. The solid red squares show astro-metric members selected from GAIA DR2 while the open red squares show radial-velocity members selectedfrom Geller et al. (2015).
Figure 1 demonstrates the utility of the GAIA data in the context of M 67. The left panel showsa vector-point diagram (VPD) of the M 67 field. The dispersed population toward the center is theGalactic disk while the concentrated cluster of points off center is the cluster. Almost all of the radial-velocity members from the recent survey of Geller et al. (2015) land within the offset clump. Forthe majority of the open cluster the separation between cluster and field was not quite so dramatic.More often, the cluster proper motion was within the field star distribution. However, the GAIADR2 parallaxes and proper motions are so precise that the cluster members still show up a very tightclump of stars within the more dispersed Galactic background and can be easily disentangled fromthe field star population.After extensive analysis, we found that 49 of our 103 target clusters could both easily be distin-guished from the field star population and showed a clear sequence in the color-magnitude diagrams.We note that there are a number of clusters that show clear clumps of cluster stars in the GAIA dataitself but fail to cross-match enough stars to the UVOT data for analysis, either because the clus-ters are too faint or are too spatially dispersed. The photometry catalogs included all photometricmeasures for both members and non-members.3.3.
Isochrone Fitting and Analysis
The goal of this paper is to compare the point-source photometry of the clusters to theoreticalisochrones in order to test the utility of the isochrones (and, by extension, the atmospheric models
Siegel et al. underpinning them) in the context of young to intermediate age stellar populations. The isochroneschosen for this exercise are the PARSEC-COLIBRI isochrones of Marigo et al. (2017, hereafter M17).These isochrones cover the entire range of ages and metallicities of the open cluster sample. The mostsignificant revisions of this iteration are to the thermally pulsing asymptotic giant branch (TP-AGB)which is not relevant to our analysis as these stars are too cool for good constraint with UVOT. infact, the detailed treatment of this phase cause the isochrones to loop around as the theoretical stars’UV emission (actually, their red leak into the UV passbands, see Appendix to Paper I) waxes andwanes during their TP-AGB phases. For the figures below, the TP-AGB phase has been removedfrom the isochrones for the sake of clarity.Isochrones were initially laid down using literature values - from studies specific to the individualcluster if possible, from K16 where no previous focused study had been made. The isochrones werethen adjusted interactively to better overlap the GAIA-selected sequences, with variation allowed inage, metallicity, reddening, distance and reddening law (either Galactic, SMC or LMC, based on theformulations of Pei 1992). The main degeneracy seen in the UV isochrone fitting involved metallicity.With all the parameters allowed to be free, the metallicity of any cluster became unconstrained.Any change in metallicity could be accommodated by responsive changes in distance modulus andreddening. We therefore fixed the metallicity of the clusters to literature spectroscopic or photometricvalues, where available, and solar metallicity otherwise. For the most part, we ignored the RGB starssince very few are detected in the NUV and those that are tend to be faint and dominated by redleak. We do note below a few clusters such as NGC 2477 ( § uvm − uvw uvw − uvw uvm − uvw uvw − uvw given the assumed metallicity .Most of our program clusters were deliberately chosen to have low foreground reddening, whichprecluded any analysis of the reddening law itself (set, by default, to the Milky Way reddeninglaw). However, for a few clusters, we were able to explore this issue to a modest extent (see, e.g.,Collinder 220, § disk sequence, ratherthan the cluster. In this region of the sky, the reddening along the line of sight causes the “blue wall”of disk MSTO stars to veer redward as it gets fainter, making it resemble a MS. However, applyingthe GAIA astrometric membership (as well as radial velocity memberships, when available) made itclear that this was not the cluster sequence, which was well-defined in the GAIA astrometry. VOT Stars: Open Clusters Figure 2.
A comparison of different isochrone fits to the photometry of the globular cluster NGC 2360. Thethree lines are all set at a metallicity of [M/H]=-0.1 and a distance modulus of m − M = 10 .
05. However,they set at different reddening values, with age adjusted to better match the MSTO. Note that increasingor decreasing the reddening has opposite effects in the two CMDs. A lower (higher) reddening moves theisochrone below (above) the observed MS in the uvm − uvw uvw − uvw Notes on the individual clusters are detailed below, in order of ascending Galactic longitude. Be-cause this is a large survey program, we do not go into great detail on our program clusters, manyof which are worthy of a paper of their own. However, we do comment on discrepancies between ourderived values and those found in the literature, the presence of unusual stellar types in any clusterand any potential broadening in the MSTO.3.4.
Individual Clusters
NGC 752
NGC 752 is situated well below the Galactic midplane in the second Galactic quadrant. It is verywell-studied, with several cluster members established as photometric standards and numerous thor-0
Siegel et al.
Figure 3.
Color-magnitude diagrams of the open clusters NGC 752, M 34, NGC 2192 and NGC 2204 (leftto right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text. ough photometric and spectroscopic surveys in the literature. Its age (1.4 Gyr), detailed abundanceratios and projected orbit make NGC 752 a typical representative of the family of old disk clusters(Carraro & Chiosi 1994; Maderak et al. 2013, B¨ocek Topcu et al. 2015, Twarog et al. 2015). Itssimilarity to disk red giants is evidence that much of the disk population formed from disruptedclusters (Reddy et al. 2012a,b).The UVOT field only encloses the central regions of this dispersed cluster (K13 measure a radiusof 1.4 degrees but a core radius of 3 . (cid:48) M 34
M 34 is second quadrant cluster situated well below the Galactic midplane. It has been exten-sively studied and shown to be a solar or slightly super-solar abundance intermediate age (200 Myr)moderately-reddened ( E ( B − V ) ∼ .
10) cluster (Jones & Prosser 1996; Schuler et al. 2003). K13measure an outer radius of 43 . (cid:48)
8, indicating that M 34 fills the entire UVOT field and we are onlyable to image the central regions of the cluster (K13 measure a core radius of 9 . (cid:48) VOT Stars: Open Clusters uvw − uvw
1) color-magnitude diagram(lower panel). As shall be seen, this discrepancy is common in clusters that are close enough for UVOTto probe the lower MS. We should note that this is near the point at which the red leak begins totake over the isochrones as these stars are too cool ( T eff < K ) to produce significant UV light.Indeed, the only way to make the isochrone match the data on the lower MS is to significantly increase the reddening (which causes the isochrones to miss the upper MS in both panels). As thereddening increases, these faint stars become “bluer” because the red leak flux in the uvw uvw § NGC 2192
NGC 2192 is a second quadrant cluster located above the Galactic midplane. Photometry from Park& Lee (1999) and Tapia et al. (2010) indicate that it is old (1.3 Gyr) and metal-poor ([
M/H ] ∼ − . M/H ] = − .
3. We identify one potentialmember star brighter and bluer than the nominal MSTO. This star is identified as a 100% likelymember by CG18 and could represent a blue straggler. The MSTO of NGC 2192 does show somebroadening beyond what would be expected from photometric errors alone. That may indicate thatNGC 2192 has an extended main-sequence turnoff, although it is a bit at the older range for clustersthat would have potential eMSTOs. It could also represent either photometric scatter in an oldfaint cluster or differential reddening in a moderately-reddened cluster. Our analysis shows thatdifferential reddening of even 0.05 magnitudes would be enough to create the relevant scatter giventhe high sensitivity of the NUV filters to reddening. However, there are not enough bright stars inNGC 2192 to measure any spatial variation. Spectroscopic study could resolve this issue.3.4.4.
NGC 2204
NGC 2204 is situated 16 degrees below the Galactic plane in the third quadrant. It is a slightly-metal-poor ([
M/H ] = − .
20) old (1.6 Gyr) cluster (Kassis et al. 1997; Jacobson et al. 2011). Whileit has been surveyed spectroscopically, these surveys have focused on old red giant stars, most ofwhich are faint in the UV. Our data do not include the uvw u passband.The last column of figure 3 shows the CMDs compared to the M17 isochrones. The astrometricselection overlaps the member stars from the radial velocity surveys. Selecting stars as members basedon either astrometry or radial velocity reveals a clear sequence of stars in the CMD. Our isochronefits to the cluster favor a slightly older age (2.3 Gyr) than previous studies with minimal reddening2 Siegel et al. (although slightly larger reddening with an SMC-like extinction law would also be consistent with thedata). Note that a NGC 2204 has a prominent population of blue stragglers, which are all confirmedas astrometric members. This confirms the detection of BSS by Frogel & Twarog (1983).3.4.5.
NGC 2243
Figure 4.
Color-magnitude diagrams of the open clusters NGC 2243, NGC 2251, NGC 2281 and NGC 2287(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
NGC 2243 is a third quadrant cluster located well below the Galactic midplane. Previous studiesby Twarog et al. (1997) and Jacobson et al. (2011) have found it to be old (5 Gyr) and metal poor([
M/H ] = − . M/H ] = − .
5) and minimal reddening. We also identify a number of blue straggler stars that areastrometric members. 3.4.6.
NGC 2251
NGC 2251 is a third-quadrant cluster located near the Galactic midplane. Parisi et al. (2005) andReddy et al. (2013) have studied the cluster, revealing it to be slightly metal-poor ([
M/H ] = − . E ( B − V ) = 0 . VOT Stars: Open Clusters E ( B − V ) = 0 .
25) and age (Figure 4, second column). We do note that the MSTOis near the saturation limit but we do not find a large number of saturated stars in the image, so itlikely this represents the true MSTO and thus the true age of the cluster.3.4.7.
NGC 2281
NGC 2281 is positioned 16 degrees above the Galactic midplane in the second quadrant. Studies byGlaspey (1987) and Netopil (2017) indicate that it is intermediate age (600 Myr) and solar metallicity.NGC 2281 is also one of the few clusters to have been previously studied in the UV, using FUV datafrom GALEX (Smith 2018). The latter study found that the two-color FUV-optical sequence ofNGC 2281 was slightly offset from that of the Hyades and Coma Ber, suggesting that it was youngerthan those two comparison clusters. However, this conclusion was tentative given the uncertainty inreddening. K13 measure a radius of 24 . (cid:48) E ( B − V ) = 0 . E ( B − V ) = 0 . M 41 (NGC 2287)
NGC 2287 (M 41) is a bright nearby second quadrant cluster situated below the Galactic midplane.Extensive study has shown it to be young (200 Myr) minimally reddened and with a high binaryfraction (Harris et al. 1993, Dobbie et al. 2012). The UVOT field only covers the central regionof the cluster (K13 measure a radius of 26 . (cid:48)
4) and the data suffered from massive saturation fromits many bright stars, which prevented both adequate PSF fitting and photometry of some of thebrightest cluster members. It also created numerous false detections around the image artifacts butthese were easily removed by matching to the GAIA DR2 catalog.The astrometric selection includes a number of radial velocity members from MMU. Selecting thesestars and examining their aperture photometry shows a clear narrow main sequence which is well-fitby the M17 isochrones with parameters similar to those derived by previous investigators (Figure 4,fourth column). The saturation limits our analysis, allowing us to only place an upper limit (500Myr) on the age of M 41 and making it impossible to confirm the eMSTO report by Cordoni etal. (2018). We detect a number of confirmed astrometric members along a probable white dwarfsequence and show the deviation in the ( uvw − uvw
2) colors seen in several other clusters.3.4.9.
NGC 2301
NGC 2301 is a bright third quadrant cluster located within the Galactic midplane. The catalogs ofK16 and Paunzen et al. (2010) describe it as slightly metal-rich ([
M/H ] = +0 .
06) and with a smallamount of foreground reddening ( E ( B − V ) = 0 . Siegel et al.
Figure 5.
Color-magnitude diagrams of the open clusters NGC 2301, NGC 2304, NGC 2355 and NGC 2360(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text. plethora of bright saturated stars prevented PSF-fitting photometry. The data presented is aperturephotometry cleaned by matching to the GAIA DR2 and corrected to the same aperture as the PSFphotometry.The astrometrically-selected CMD (Figure 5, first column) shows clear loci for the cluster stars.The one member star catalogued by MMU is well within the proper motion locus and along theRGB, albeit significantly fainter than expected. The M17 isochrones fit the photometry quite well,providing an upper limit on the age of 400 Myr and favoring a slightly higher foreground reddening( E ( B − V ) = 0 . NGC 2304
NGC 2304 is a third quadrant cluster located above the Galactic midplane. Multiple photometricstudies have been done, producing a broad range of parameters as shown in Table 2 (Ann et al.2002, Hasegawa et al. 2008, Lata et al. 2010, Oralhan et al. 2015). We were unable to identify acomprehensive radial velocity catalog and the cluster was not surveyed by MMU.The astrometric selection produces a clean CMD (Figure 5, second column). The M17 isochronesfavor a cluster that is intermediate age (900 Myr), moderately metal-poor ([
M/H ] = − .
3) andslightly reddened ( E ( B − V ) = 0 . VOT Stars: Open Clusters
NGC 2343
NGC 2343 is a third quadrant cluster situated near the Galactic midplane. Pe˜na & Mart´ınez (2014)showed the cluster to be young (12 Myr) and metal-poor , with [
F e/H ] ∼ − . uvby − β photometry. It would be unusual for such a young cluster to be so metal-poor. Netopil et al. (2016),by contrast, estimate a photometric metallicity if [ F e/H ] = − .
03. Several catalogues, includingWEBDA, list the cluster at [
F e/H ] = − . F e/H ] = − . Figure 6.
Vector-point and color-magnitude diagrams of the cluster NGC 2343 compared to the best-fit isochrones of PARSEC-COLIBRI. The brightest stars in the cluster are saturated, preventing us frommeasuring more than an upper limit on the age of the cluster. The solid red squares show astrometricmembers selected from GAIA DR2 while the open red square is a spectroscopic member from MMU. Thestarred point marks the star used by Pe˜na & Mart´ınez (2014) to estimate the metallicity of the cluster.
For unknown reasons, the CG18 catalog of NGC 2343 was among a few third-quadrant clustersthat had very few stars with high cluster membership probabilities. We therefore created our ownmembership probability measures by fitting two Gaussians to the distribution of proper motions inRA-DEC space. Cluster membership was defined as the ratio of the two Gaussians and memberswith ratios greater than 9.0 were identified as likely members. This method was initially applied toall of our program clusters. For most, it produced member catalog almost identical to CG18 and wechose to utilized the CG18 membership as it used more information and its membership lists wereslightly more conservative. NGC 2343 was one of the few where the methods disagreed.6
Siegel et al.
Figure 6 shows the vector-point diagram of the cluster and shows a tight clump of stars towardthe edge of the field star distribution. Using these as cluster members, we find that the metallicityestimate of and Pe˜na & Mart´ınez are likely in error. The star they identify as 29 and use as theirmetallicity estimator is almost certainly not a member of the cluster, being placed well to the edgeof the field star distribution and off of the MS. It is most likely a metal-poor thick disk or halo star.HD 54387, used as the metallicity estimator by Claria (1985), is saturated in the UVOT images. Itsproper motion does place it near the NGC 2343 proper motion, however.The member stars trace a clear MS. Given the uncertainty in the metallicity, we adopt solarmetallicity. If the cluster is indeed slightly metal-poor, the result would be a shorter distance modulus( m − M = 9 . uvw − uvw NGC 2355
NGC 2355 is a third quadrant cluster located above the Galactic midplane. Previous investigationshave revealed it to be slightly metal-poor ([
M/H ] = − .
06) and of intermediate age (900 Myr, Donatiet al. 2015) with typical disk abundance ratios (Jacobson et al. 2011).The CMDs (Figure 5, third column) show a fit consistent with previous investigations, but favoringa slightly higher age of 1.1 Gyr. The MSTO of the cluster proved difficult to fit precisely with anycombination of parameters. Notably, we find that there is some broadening in the MSTO. Thiscould indicate an eMSTO but could also be the effect of differential reddening (to which the NUVis particularly sensitive at the level of E ( B − V ) = 0 . NGC 2360
NGC 2360 is third quadrant cluster located above the Galactic midplane. The literature on thiscluster contains widely different values for the parameters as shown in Table 2. Reddy et al. (2012)and Sales-Silva et al. (2014) favor a large distance ( m − M = 11 .
72) and young age (560 Myr)while Claria et al. (2008) favors a closer distance ( m − M = 10 .
09) and older age (1.8 Gyr).This discrepancy is likely due to the presence of two distinct color-magnitude sequences in the field,something we have observed in numerous low-latitude third quadrant clusters, as detailed in a numberof following clusters. One of these represents foreground disk; the other the cluster itself. WEBDAlists the cluster at the longer distance and younger age.Previous spectroscopic studies have been mostly focused on red giants, which our data cannot useto constrain the properties of the cluster, due to their lack of UV emission. However, the astrometricselection (Figure 5, last column) clearly aligns with the brighter/redder main sequence. This sequenceis favored independently by the proper motions, parallaxes and spatial distribution. Fitting the M17isochrones to this sequence produces a fit consistent with the Claria et al. result and inconsistent with
VOT Stars: Open Clusters
Figure 7.
A close-up of the MSTO region of NGC 2360. The panel shows GAIA- and spectroscopically-selected members. The MSTO region is very broad due to the combination of eMSTO, convective hook andthe beginning of the RGB.
NGC 2360 is one of the few clusters in our sample that is both bright enough and close enough toexplore the fine structure of the MSTO. Milone et al. (2018) identified it as a cluster with an eMSTO.This region of the CMD is tricky because it overlaps both the convective hook in the MSTO and thebeginning of the RGB. However, the uvm − uvw Berkeley 37
Berkeley 37 is a faint third quadrant cluster situated above the plane. The only photometric studyto date is that Oralhan et al. (2015) who showed to be distant ( m − M = 13 .
60) with light extinction( E ( B − V ) = 0 .
05) and an intermediate age (600 Myr). K16 derive a slightly shorter distance withmore reddening.The CG18 selection did not pick out many cluster members, likely due to the similarity of thecluster kinematics to that of the disk. We applied our own selection criteria using the double-Gaussian method described for NGC 2343 ( § NGC 2374 Siegel et al.
Figure 8.
Color-magnitude diagrams of the open clusters Berkeley 37, NGC 2374, NGC 2396 and NGC 2420(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
NGC 2374 is third quadrant cluster situated near the Galactic midplane. It has been poorlystudied with no spectroscopic metallicity available. The only recent photometric study is Carraro etal. (2015), who found it be young (250 Myr) and argued that its position near the Galactic warpmakes it consistent with either a thick disk cluster or an extended thin disk cluster.MMU only identified one member star – a red giant that is faint and well-removed from the mainsequence. CG18 only identify a handful of astrometric members. The VPD, however, shows a clearclump of stars well-removed from the disk sequence that defines narrow photometric sequences inthe CMDs (Figure 8, second column). We derive parameters similar to that of Carraro et al. (2015)albeit at slightly higher reddening ( E ( B − V ) = 0 . NGC 2396
NGC 2396 is a third quadrant positioned just above the Galactic midplane. It has never been thefocus of an individual study and the only parameters in the literature come from the comprehensivesurvey of K16, who showed the cluster to be nearby ( m − M = 8 . E ( B − V ) = 0 . E ( B − V ) = 0 .
15) and a dramaticallylarger distance ( m − M = 11 . uvw − uvw VOT Stars: Open Clusters
NGC 2420
NGC 2420 is a third quadrant cluster positioned 19 degrees above the Galactic midplane. Extensiveinvestigation (Von Hippel & Gilmore 2000, Anthony-Twarog et al. 2006, Souto et al. 2016) hasrevealed it to be old (2 Gyr) and slightly metal-poor ([
F e/H ] ∼ − . uvm − uvw uvw − uvw α -abundances (Suoto et al. 2016). Lowering the metallicity to [Fe/H]=-0.3 and decreasing the clusterdistance produces very good isochrones fits. However, the abundance of NGC 2420 is very-wellstablished through multiple spectroscopic studies. Previous research has indicated some variation inthe UV reddening law within external galaxies and within the Galaxy, particularly the R V value andbump strength (Siegel et al. 2014, Hagen et al. 2017) but the foreground reddening in NGC 2420 islow and modifying the reddening law produces only minor changes. We are unable, at this point, toexplain the slight discrepancy in NGC 2420 compared to other clusters that have better fits.We do note, however, that the MSTO does show a bit of broadening, which would be consistent withan eMSTO. NGC 2420’s 2 Gyr age is toward the lower end of the range where eMSTO is expectedto manifest before magnetic breaking evens out the stellar rotation rates (Georgy et al. 2019). Itwould be worth further investigation to determine if the MSTO stars shows differences in rotationrates and potentially set the age limits of the eMSTO phenomenon on a more empirical footing.3.4.18. NGC 2422
NGC 2422 is a bright nearby third quadrant cluster positioned near the Galactic midplane. Studiesby Rojo Arellano et al. (1997) and Prisinzano et al. (2003) have shown it to be very young (80Myr) with moderate foreground reddening ( E ( B − V ) = 0 . F e/H ] = − .
03) but within uncertainties of solar metallicity. The clusterproved difficult to study due to the large number of saturated stars and the concomitant number offalse detections in the diffraction spikes. The sample studied was only of stars the could be matchedthe GAIA DR2. The cluster covers a large area of sky – 37 . (cid:48) uvm uvw NGC 2423 Siegel et al.
Figure 9.
Color-magnitude diagrams of the open clusters NGC 2422, NGC 2423, NGC 2428 and NGC 2437(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
NGC 2423 is a third quadrant cluster positioned above the Galactic midplane. Prior study hasrevealed it be a intermediate age (1 Gyr) and metal-rich ([
F e/H ] = +0 .
10) cluster (Claria et al 2008,Santos et al. 2009, 2012, Paunzen et al. 2010, Conrad et al. 2014).Like NGC 2360 (3.4.13, the CMD shows a double sequence of the old thin disk and cluster (Figure9, second column). The astrometric selection picks the brighter sequence and contains the radialvelocity members identified by MMU. The M17 isochrones produces a fit with parameters consistentwith the literature. 3.4.20.
NGC 2428
NGC 2428 is a third quadrant cluster positioned 16 degrees below the Galactic midplane. Conradet al. (2014) show the cluster to be slightly metal-poor ([
F e/H = − .
15) with modest reddening( E ( B − V ) = 0 . NGC 2437
VOT Stars: Open Clusters . (cid:48) E ( B − V ) = 0 . m − M = 10 . m − M = 9 . E ( B − V ) = 0 .
35) would also fit the data quite well. Given thehigh uncertainy in the latter measure and the lower likelihood that such a young cluster would be sometal-poor, we list NGC 2437’s parameters in Table 2 as derived for the higher metallicity.3.4.22.
NGC 2447
Figure 10.
Color-magnitude diagrams of the open clusters NGC 2447, Berkely 39, NGC 2477 and ESO 123-26 (left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solidred squares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text. Siegel et al.
NGC 2447 is a third quadrant cluster situated within the Galactic midplane. Studies by Clariaet al. (2005), Santos et al. (2009, 2012), Conrad et al. (2014) and Reddy et al. (2015) describeas having slightly subsolar metallicity ([
F e/H ] = − .
10) and intermediate age (400 Myr). The K13radius of 30 . (cid:48) Berkely 39
Berkeley 39 is a third quadrant cluster situated 10 degrees above the Galactic midplane. Previousstudies (Kassis et al. 1997; Carraro et al. 1994, 1999; Frinchaboy et al. 2006) have shown the clusterto be massive, old (5.5 Gyr) and moderately metal-poor ([
F e/H ] = − .
20) The extensive spectro-scopic survey of Bragaglia et al. (2012) identifies 30 radial-velocity cluster members with very smallstar-to-star variations in abundance, consistent with a simple stellar population. Unfortunately, onlya handful of their program stars have counterparts in the UVOT data due to their cool temperaturesand subsequent low UV emission.Because the cluster is old, reddened and distant, we only detect the very top of the MS in the uvw − uvw uvm − uvw uvm NGC 2477
NGC 2477 is a third quadrant cluster located below the Galactic midplane. Extensive previousinvestigation has shown it to be of intermediate age (1 Gyr) with moderate ( E ( B − V ) = 0 . . (cid:48)
0, indicating we only observe the core.The astrometric and spectroscopic selection identify a clear MS in the CMDs (Figure 10, thirdcolumn). The M17 isochrones produce a fit consistent with the previous literature. However, we notethat the MS is fairly broad and that this breadth occurs along the entire observed MS. It is likely thisrepresents differential reddening as NGC 2477 is one of the few clusters we have that has moderateforeground reddening ( E ( B − V ) = 0 .
25) and even small variations can produce noticeable breadthin the CMD. We note at least three stars that are likely blue stragglers.As noted above, we mostly ignore the RGB stars in the clusters due to their faintness and thedominance of the red leak in their photometric measures. NGC 2477, however, is one of the thathas a prominent RGB. It is also well-surveyed spectroscopically. One of the most noticeable aspectsof its CMD is that the RGB stars tend to land significantly blueward of the isochrone prediction.This tendency of the isochrone to miss the few RGB stars can be seen in other clusters but this theclearest example. This suggests that the isochrones are not reproducing the NUV properties of theRGB accurately. The reasons for this could be many but the two most likely are that the atmosphericmodels do a poor job of predicting the intrinsically low UV flux for such cool objects (likely due toopacity issues at low temperatures) or that our accounting for the red leak is incorrect. Further NUV
VOT Stars: Open Clusters
ESO 123-26
ESO 123-26 is a fourth quadrant cluster positioned well-below the Galactic midplane. It has notbeen the subject of individual study, with the only parameter estimates coming from global surveyssuch as K13 and K16.This cluster is not included in the CG18 compilation and so we applied the double Gaussianselection method used for NGC 2343 ( § m − M = 10 .
20) and a slightly older age (500 Myr). We caution however, that the proximity ofthis cluster may mean that its brightest members are saturated, so this age should be regarded asan upper limit. 3.4.26.
NGC 2479
Figure 11.
Color-magnitude diagrams of the open clusters NGC 2479, NGC 2482, NGC 2506 and NGC 2509(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
NGC 2479 is above the Galactic midplane in the fourth quadrant. A recent photometric study byPiatti et al. (2010) revealed the cluster to be old (1 Gyr) and relatively distant ( m − M = 11 . Siegel et al.
NGC 2479 is also not included in the CG18 study and we applied our own astrometric selection(see § NGC 2482
NGC 2482 is a fourth quadrant cluster positioned just above the Galactic disk. Recent studieshave shown it to be an intermediate age (400 Myr) near-solar metallicity cluster (Reddy et al. 2013,Krisciunas et al. 2015, Conrad et al. 2014). The astrometric selection shows a clear sequence in thecolor-magnitude diagram (Figure 11, second column). The M17 isochrones produce a fit similar toprevious studies but with lower reddening ( E ( B − V ) = 0 . NGC 2506
NGC 2506 is a third quadrant cluster located above the Galactic midplane. Previous studies haveshown it to be an old (2 Gyr) metal-poor ([
F e/H = − .
3) cluster (Reddy et al. 2012b, Lee et al.2012. Anthony-Twarog et al. 2016). The cluster has also been the subject of extensive spectroscopicsurveys which have confirmed its low metallicity (MMU, Anthony-Twarog et al. 2018).The CMD shows a dominant locus of cluster stars, which contains the radial velocity members ofMMU and astrometrically-selected members (Figure 11, third column). The M17 isochrones producea fit consistent with prior literature given the assumed metallicity of [Fe/H]=-0.3. NGC 2506 alsoshows a very prominent blue straggler sequence, the brightest of which is both a proper motion andradial velocity member. There is quite a bit of breadth to the MS, which is unexpected given thelow reddening. The broadening occurs along the entire length of the MS rather than just the MSTO,making the nature of it unclear. 3.4.29.
NGC 2509
NGC 2509 is an older cluster positioned above the Galactic disk. The literature shows a numberof very different solutions to NGC 2509’s properties. Sujatha & Babu (2003) argue for a very oldage of 8 Gyr while Tadross (2055) and Carraro & Costa (2007) find a more intermediate age of1.2-1.6 Gyr, but with a significant difference in distance modulus between them ( m − M = 11 .
50 and12.50, respectively). No spectroscopic metallicity is available and those studies have assumed solarmetallicity.The CG18-selected CMD (Figure 11, last column) shows a faint MS. Assuming solar metallicity,the best fit to the photometry is consistent with the Tadross et al. (2005) results, showing an olderage (1.7 Gyr) and a shorter distance ( E ( B − V )=11.72).3.4.30. NGC 2527
NGC 2527 is a third quadrant cluster near the Galactic midplane. Reddy et al. (2013) foundit be of intermediate age (450 Myr) slightly metal-poor ([
F e/H = − .
09) and slightly reddened( E ( B − V ) = 0 . F e/H ] = +0 .
06) but based only on two stars.The CMD (Figure 12, first column) shows a clean sequence, well-separated from the disk. We findparameters similar to those of Reddy et al.. The age of the cluster is a bit uncertain as the brightest
VOT Stars: Open Clusters Figure 12.
Color-magnitude diagrams of the open clusters NGC 2527, NGC 2533, M 48 and NGC 2567(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
MSTO stars are saturated. However, the curvature of the upper parts of the CMD are inconsistentwith an age older than that found in Reddy et al. Note again that the uvw − uvw NGC 2533
NGC 2533 is a third-quadrant cluster positioned above the Galactic midplane. The cluster hasnever been targeted for individual photometric study. However, basic parameters come from the bluestraggler survey of Ahumada & Lapasset (2007) and the global survey of K16.The CG18 selection showed very few members stars so we applied our own proper-motion analysis,as detailed in § m − M = 12 .
20) and older age (700 Myr), although we emphasize thatthe analysis of this cluster is complicated by the somewhat high ( E ( B − V ) = 0 .
40) reddening andred leak causing the RGB to cross over the MSTO in the uvw − uvw M 48 (NGC 2548)
M 48 is one of the most well-studied open clusters in the Galaxy. While the specific parametersof the various studies vary, there is a general consensus that is close, of intermediate age (400 Myr)and with minimal reddening (Rider et al. 2004; Balaguer-N´u˜nez et al.2005; Wu et al. 2005; Sharmaet al., 2006). Most studies assume near solar metallicity although Balaguer-N´u˜nez et al. measure6
Siegel et al. a photometric metallicity of [Fe/H]-0.24. The massive size of the cluster (K13 measure a radius of43 . (cid:48)
2) indicates that we have only surveyed the central regions.The field contains many saturated stars and so only stars with GAIA DR2 counterparts are shownin the CMD (Figure 12). Assuming solar metallicity, we derive similar parameters to the consensusliterature, with slightly elevated reddening ( E ( B − V ) = 0 . ∼
300 Myr.3.4.33.
NGC 2567
NGC 2567 is a third-quadrant cluster positioned slightly above the Galactic midplane. It has notbeen the subject of individual study. The K16 study provides basic parameters, identifying theclusters as intermediate age (400 Myr) with modest foreground reddening ( E ( B − V ) = 0 .
13) whileConrad et al. (2014) identify it as slightly metal-poor ([
F e/H ] = − . § NGC 2571
NGC 2571 is a third quadrant cluster positioned above the midplane. It has been analyzed by Giorgiet al. (2002) and ¨Ozeren et al. (2014). The latter, based on 2MASS photometry, shows moderatereddening ( E ( B − V ) = 0 .
36) while the former shows a small amount of foreground reddening( E ( B − V ) = 0 .
10. Both agree that the cluster is quite young ( ∼
50 Myr).The CMD (Figure 13, first column) shows a clean MS well-separated from the disk. We find,however, that the parameters derived by previous investigators are somewhat inconsistent with theNUV photometry. The high reddening value derived by ¨Ozeren et al. (2014) would be inconsistentwith the color of the upper MS for any age, predicting a uvm − uvw NGC 2627
NGC 2627 is a third quadrant cluster situated just above the Galactic midplane. Prior photomet-ric studies have derived an old age of 1.6 Gyr (Piatti et al. 2003; Ahumada 2005). There is nospectroscopic metallicity but Piatti et al. estimate that is slightly subsolar ([Fe/H] ∼ − .
1) based onWashington indices.MMU surveyed the cluster but none of their four RGB stars are detected in the Swift data. Theastrometric selection, however, produces a clear CMD with a well-defined MS and MSTO (Figure13, second column). The best-fit isochrone roughly corresponds to the parameters from Ahumada.Note that because of the age, distance, and reddening of this cluster, the RGB/AGB region overlaps
VOT Stars: Open Clusters Figure 13.
Color-magnitude diagrams of the open clusters NGC 2571, NGC 2627, NGC 2658 and NGC 2669(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text. the MSTO in the uvw − uvw NGC 2658
NGC 2658 is a third quadrant cluster located six degrees above the Galactic midplane. The clusterhas not been the target of spectroscopic study. Paunzen et al. (2010) indicated that it was metal-poorbut the updated survey of Netopil (2016) indicate solar metallicity. Photometric studies (Ramsay &Pollaco 1992, Ahumada 2003) describe it as intermediate age (300 Myr), solar metallicity and withmoderate reddening ( E ( B − V ) = 0 . E ( B − V ) = 0 .
30. The MS does show a significant breadth and some extension beyond the MSTO.This could indicate that it has an eMSTO.3.4.37.
NGC 2669
NGC 2669 is a poorly-studied third quadrant cluster located six degrees below the Galactic mid-plane. The last focused study of the cluster was Vogt & Moffat (1973) but it has been included inthe global surveys of Paunzen et al. (2010), K13 and K16. These studies hint at a young (100 Myr)slightly metal-poor ([
F e/H ] ∼ − . E ( B − V ) = 0 .
18) cluster.8
Siegel et al.
The astrometric selection includes one MMU radial velocity member and traces the bluer sequencein the CMDs (Figure 13, last column). Assuming the Paunzen et al. metallicity, we find parameterssimilar to those in the literature. We note, however, that the brightest stars are saturated and so theage given is an upper limit. 3.4.38.
NGC 2818
Figure 14.
Color-magnitude diagrams of the open clusters NGC 2818, NGC 2825, Collinder 220 andNGC 3680 (left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in thetext. Solid red squares show astrometric members selected from GAIA DR2 by CG18 while open squaresare spectrosopic members from either MMU or sources given in the text.
NGC 2818 is a distant third quadrant cluster situated above the Galactic midplane. MMU has anextensive radial-velocity membership. This cluster was recently studied in great detail by Bastian etal. (2018) who found that it has a prominent eMSTO. Their complementary radial velocity studyshowed that stars on the red side of eMSTO were fast rotators compared to stars on the blue side,confirming the hypothesis of stellar rotation as the origin of the eMSTO.The CMDs (Figure 14, first column) show a well-defined MS with a prominent MSTO. We findthat the cluster is a bit older than the previous studies, with the MSTO corresponding to an age ofapproximately 1 Gyr with slightly less-reddening than Bastiat et al. (2018) of E ( B − V ) = 0 . VOT Stars: Open Clusters Figure 15.
A close-up of the MSTO region of NGC 2818. The left panel shows GAIA- and spectroscopically-selected members. The MSTO region is very broad due to the combination of eMSTO, convective hook andthe beginning of the RGB. The right panel shows deviations from a linear fit to the MS region and showsa broad pattern in the MSTO region, again consistent with an eMSTO in NGC 2818 likely resulting fromdifferences in stellar rotation among the MSTO stars.
As noted, Bastian et al. identified an eMSTO in this cluster. As with NGC 2360, the overlap ofconvective hook, RGB and MSTO make analysis tricky. However, the uvm − uvw NGC 2925
NGC 2925 is a fourth quadrant cluster situated just below the Galactic midplane. The onlyfocused study is that of Topaktas (1981) who concluded that the cluster was moderately reddened( E ( B − V ) = 0 .
08) based on photographic photometry. K16 estimate an intermediate age of 500Myr.While the astrometric selection produces a clean sequence in the CMD (Figure 14, second column)we note that the three RGB stars cross-identified from MMU lie outside of the astrometric selection(two of which are detected by UVOT). In proper-motion space, they span a range of 5 mas yr − ,which is a stark contrast to other clusters in which spectroscopically-confirmed members span atmost a few tenths of a mas yr − . Applying our own analysis to the GAIA data results in an almostidentical membership selection to CG18 and confirms that the three MMU members are outliers. Wetherefore find it likely that the MMU stars are not members of NGC 2925 but, given the high propermotions, are halo contaminants. The astrometric members favor the redder sequence in the field.Fitting the M17 isochrones produces a fit very similar to K16 but with an increase in reddening to E ( B − V ) = 0 .
15. 3.4.40.
Collinder 220
Collinder 220 is a fourth-quadrant cluster within the Galactic midplane. It has not been the subjectof any published individual study and our parameters in Table 2 are taken from K16. UVOT obtained0
Siegel et al. very deep imaging of this cluster due to its proximity to 1E1024.0-5732 (WR21a), a colliding windbinary near the Westerlund 2 cluster that was intensely monitored through periastron.Analysis of this cluster proved challenging. The VPD shows a clear tight clump of stars withinthe field star distribution. However, the field is so dense that our double-Gaussian modelling failed.Picking out stars in this clump by hand from within the clump showed (1) a tight sequence of stars inthe CMD (Figure 14, third column); (2) a concentration of stars at the nominal position of the cluster;(3) a mean proper motion roughly consistent with the K13 parameters (( µ α , µ delta ) ∼ ( − . , . m − M = 11 .
99) and a younger age (200 Myr). What is really interesting about thiscluster, however, is that it the first to require a change in the reddening law in order to properly fit thecolor-magnitude sequences. Most of our clusters were picked to have low foreground extinction andeven those with more extinction are usually well-fit by the standard Milky Way law. However, usingthe Milky Way law on Collinder 220 produces a discrepancy in the colors of the two main sequenceswhich no combination of parameters could fit. Adopting the LMC extinction law, by contrast, resultsin a consistent fit for both CMDs. Given the disagreement between our measured parameters andthose of K16, we put forward this interpretation with caution. However, spectroscopic study ormulti-wavelength study of Collinder 220 would seem warranted to confirm or refute any changes inthe UV extinction law toward the cluster.3.4.41.
NGC 3680
NGC 3680 is a fourth quadrant cluster placed slightly above the Galactic midplane. It has beenthe subject of detailed abundance studies (Santos et al. 2009; Pena Suarez et al. 2018) as well asphotometric investigations (Anthony-Twarog & Twarog 2004) which have described it as an old (1.75Gyr) near-solar metallicity disk cluster.The astrometric selection produces a clear main sequence that contains all of the MMU radialvelocity members (Figure 14, last column). The isochrones, however, favor a solution with minimalor no reddening, which produces a consistent fit with a slightly shorter distance ( m − M = 10 . Lynga 2
Lynga 2 is a small obscured cluster situated toward the Galactic Center. The cluster has beenpoorly studied, likely due to the high density of field stars against a relatively spartan cluster.Alter et al. (1970) and Bica et al. (2006) describe it as young (90 Myr), with moderate reddening( E ( B − V ) = 0 . NGC 6823
VOT Stars: Open Clusters Figure 16.
Color-magnitude diagrams of the open clusters Lynga 2, NGC 6823, NGC 6866, NGC 6939(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
NGC 6823 is a distant obscured first quadrant cluster situated near the midplane. It is the onlyvery young ( t <
20 Myr) cluster in our program that is well-observed although it still has a numberof saturated stars. It has not been the subject of individual study so parameters are taken from K16.The astrometric selection produces a clean, if spartan, MS (Figure 16, second column). Adjustedto lower reddening ( E ( B − V ) = 0 .
70) and longer distance ( m − M = 11 .
49) than given in K16,we get excellent agreement between the isochrone and the sequence. The MS is a bit broader thanthe photometric errors, possibly indicating differential reddening or the influence of stellar rotation.The brightest stars in this cluster are saturated and so only an upper limit on the age can be given.However, this upper limit is close to the K16 age. Given the high-quality data, young age andforeground extinction, this cluster would be a good candidate for future spectroscopic study.3.4.44.
NGC 6866
NGC 6866 is a first quadrant cluster situated within the Galactic midplane. Previous studies haveshown it to be of intermediate age (900 Myr), moderately reddened ( E ( B − V ) = 0 .
16) and slightlymetal-poor with [
F e/H ] = − .
10 (Janes et al. 2014, Bostanci et al. 2015)The CG18 selection provided to be too conservative – identified very few clusters members – and sowe applied our own astrometric selection ( § Siegel et al.
NGC 6939
NGC 6939 is a second quadrant cluster positioned 12 degrees above the Galactic midplane. Nu-merous studies have been done of the cluster and have found it to be of solar abundance (Jacobsonet al. 2007), heavily reddened ( E ( B − V ) = 0 .
33) and old, with an age of 1.3-1.6 Gyr (Andreuzzi etal. 2004, Rosvick & Balam 2002). However, there are significant disagreements between the exactparameters derived from the studies, as shown in Table 2.The astrometric selection identifies a faint MS in the CMDs (Figure 16, last column). We find thatwe can fit this sequence using parameters similar to those of Andreuzzi et al. (2004) but only withsome additional reddening ( E ( B − V ) = 0 . uvw − uvw bluer than the MSTO because we are only measuring the degreeof red leak in these cool and heavily reddened stars. The CMD also shows at least one star bluer andbrighter than the MSTO. This is a proper-motion member and crudely along the MS that a youngerstellar population would follow. It is likely that it is a blue straggler.3.4.46. NGC 6991
Figure 17.
Color-magnitude diagrams of the open clusters NGC 6991, NGC 7058, NGC 7063 and NGC 7209(left to right). The solid lines are PARSEC-COLIBRI isochrones set to the parameters in the text. Solid redsquares show astrometric members selected from GAIA DR2 by CG18 while open squares are spectrosopicmembers from either MMU or sources given in the text.
NGC 6991 is a poorly studied first quadrant midplane cluster. The only estimates of age anddistance come from the global survey of K16. Casamiquela et al.(2016) studied the radial velocity
VOT Stars: Open Clusters . (cid:48)
7) and thuswe only image the central regions.Th first column of gigure 17 shows the CMDs of the cluster. The GAIA astrometry matches aradial velocity member from Casamiquela et al. (2016). These stars define a narrow CMD sequence.We measure parameter similar to K16, but with an older age (1.6 Gyr).3.4.47.
NGC 7058
NGC 7058 is a nearby sparse second quadrant cluster that is within the Galactic midplane. It hasnot been the target of individual study, with the only estimates of its age, distance and reddeningfrom the global survey of K16.The CG18 selection selected very few member stars and we therefore applied our method ( § uvw − uvw NGC 7063
NGC 7063 is a first quadrant cluster situated nine degrees below the Galactic midplane. The onlycomprehensive study to date is that of Pena et al. (2007) who characterize it as young (140 Myr)and with low foreground reddening ( E ( B − V ) = 0 . § uvw − uvw NGC 7209
NGC 7209 is a second quadrant cluster positioned seven degrees below the Galactic midplane.No spectroscopic metallicity has been published but photometric studies describe the cluster as solarmetallicity, intermediate age (500 Myr) and moderately reddened, with E ( B − V ) = 0 .
17 (Vanseviciuset al. 1997, Paunzen et al. 2010, Netopil et al. 2016, K16).The astrometric and spectroscopic selection find a clear MS in the CMDs (Figure 17, last column).Fitting the M17 isochrones to this sequence yields parameters similar to the literature. However, wenote that the CMD shows significant structure that almost resembles a second turnoff. Increasing theage of the cluster to 700 Myr would match this second turnoff albeit with some slight discrepancy inthe curvature of the upper MS. This would relegate the brighter stars to being blue stragglers. If thebrighter stars represent the MSTO, then NGC 7209 would have an age of 500 Myr but an eMSTO.Spectroscopic investigation of the rotation rates could determine if this is the case.4
Siegel et al.
Table 2.
Parameters of Clusters in this Study
Cluster Distance E (B-V) log(Age) [Fe/H] Sourcem-M (yr)NGC 752 8.30 0.03 9.15 -0.03 B¨ocek Topcu et al. (2015), Twarog et al. (2015)NGC 752 8.30 0.05 9.15 +0.00 This StudyNGC 1039 8.38 0.10 8.35 +0.07 Jones & Prosser (1996)NGC 1039 8.71 0.07 8.25 +0.07 Schuler et al. (2003)NGC 1039 8.54 0.08 8.38 0.07 Kharchenko et al. (2013)NGC 1039 8.75 0.10 < a Using LMC extinction law.This table is available in its entirety in machine-readable and Virtual Observatory (VO) forms. DISCUSSION4.1.
Isochrones and Cluster Fits
The purpose of this program was to test the utility of standard isochrones in the UV. In that sense,we find that the isochrones perform quite well, successfully reproducing the upper main sequence forcluster up to several Gyr in age. Most of the CMDs are well-described using isochrones set to valuesat or near those in the literature. For clusters with significant revisions, the source of the discrepancyis either older studies or, in some cases, previous studies mistaking the disk sequence for the clustersequence.The consistency between the models and the data shows that even for fairly young stellar pop-ulations, the NUV emission is well-described by the existing atmospheric models. There is littleevidence of stochastic variation and little evidence of the excess emission produced by chromosphericactivity as seen in the FUV (see discussion in Smith 2018). However, it is important to note thatour study is focused on only limited range of parameter space. Almost all of our clusters are olderthan 100 Myr and the brightest stars in the youngest are few and saturated. Younger clusters willhost much more massive stars, in which factors like binarism, rotation and magnetic field can play amore important role and our understanding of convention and mixing are poor (see, e.g., Viallet et al.2013). In addition, our stars are primarily metal-rich (or assumed to be) so we are unable to test howthe isochrones perform at lower metallicity levels. Finally, almost all of our clusters are minimally
VOT Stars: Open Clusters < M yr ) stellar populations or populations with significant reddening would requiresurveys of the nearby stellar populations in the Magellanic Clouds. Swift/UVOT has also surveyedthese areas (Hagen et al., 2017) but its reach may exceed its grasp in such dense star-forming regionsas 30 Doradus where the younger stars tend to reside. That is more properly the domain of futuremissions that have the wide-field high-resolution multi-filter FUV imaging.The main deficiencies in the isochrones are for redder stellar types: the RGB, AGB and lower MSstars that are cool and have little emission in the UV. This is hardly a surprise given that for coolstars (those with T eff < K ), the “UV” signal in two of Swift’s filters becomes dominated by thered leak. This plays havoc with the effective wavelength (see, e.g., Siegel et al. (2012)) and causesthe reddening to have a non-linear effect (Paper I, Brown et al. 2010). For later stellar types thatare heavily extinguished, this will cause them to move blueward in uvw − uvw uvw uvw Hubble Space Telescope ’s upcoming HubbleUV Legacy Library of Young Stars As Essential Standards (ULLYSES) program.
Figure 18.
A comparison of photoemtric distance from the fit isochrones for a sample of the UVOT clusterstars, plotted against absolute T eff (bottom) and observationsl uvw − uvw Siegel et al.
To determine the nature of these discrepancies, we examined the deviations of stellar measuresfrom the fit isochrones. This deviation was calculated by measuring, for each star, the minimumquadrature color-magnitude distance from the isochrone. Figure 18 shows the residuals plottedagainst observational uvw − uvw observed color, where thered leaks takes over the signal. However, it the problems arises from atmospheric models, it shouldoccur at a consistent effective temperature, where the atmospheric models begin to fail. For clustersthat show the deviation at faint magnitudes, the discrepancies do not track observed stellar colorvery well. The deviations do, however, track effective temperature quite well, beginning to deviatetoward bluer colors at approximately logT eff ∼ . Reddening Law
As noted above, most of our program clusters have low intrinsic reddening. Those few that havesubstantial reddening seem to work fine with a “Milky-Way”-like extinction law (i.e., one with ashallower slope and a strong 2175 ˚A bump). Only one cluster – Collinder 220 – shows any sign ofdeviation from the standard reddening law. This is somewhat surprising as previous investigationshave shown some variation in the extinction law at high latitudes within the Galaxy (Peek & Schimi-novich 2013), at low latitude within the Galaxy (Siegel et al. 2012) and in nearby galaxies (Hagen etal. 2017, 2019). This suggests that future endeavors should both look at resolved stellar populationin nearby galaxies where the reddening law is known to vary and at more reddened clusters withinour own Galaxy where variations in the law will be amplified and more readily detectable. It mightalso be of benefit to make these studies multi-wavelength, as in Hagen et al. (2017, 2019) where themore well-understood OIR extinction law can be used to “nail down” the line-of-sight extinction,freeing the UV to focus more on variation in the law instead of variation in the total extinction levelalong the line of sight. 4.3.
Extended MSTOs
Two of our clusters – NGC 2360 and NGC 2818 – have previously been identified as having extendedMSTOs based on photometric and spectroscopic exploration. We confirm the photometric broadeningand further note that a number of other clusters (NGC 2355, NGC 2420 NGC 2658, NGC 6823,NGC 7209) show a similar broadening of the MSTO beyond what is expected from photometricerror. This is unlikely to be the effect of differential reddening because most of our clusters wereselected to have low reddening and differential reddening would broaden the entire MS. Spectroscopicinvestigation of rotation rates would confirm the nature of these extended MSTOs, as it has forNGC 2360 and NGC 2818. If confirmed, this would support the contention of Cordoni et al. (2018)that the feature is not unusual among Galactic open clusters and the UV may be more sensitive tothe effect than the optical or IR.Considering that the eMSTO involves the brightest stars in any specific population, this is an effectthat may need to be accounted for in spectral synthesis models. While the precise effect of stellarrotation on integrated light is beyond the scope of this paper, the indications that stellar rotationcan increase both mass loss rates and main sequence lifetime makes their potential ubiquity in youngstellar populations both critical to confirm and critical to incorporate into the models.
VOT Stars: Open Clusters CONCLUSIONSWe have measured photometry of stars in 103 open clusters using the Swift/UVOT telescope. Wehave analyzed 49 of these clusters, using GAIA DR2 and spectroscopic studies to separate membersstars from the field to study precise CMDS and compared the CMDs to theoretical isochrones. Ourmain results are: • The theoretical isochrones reproduce the features of the CMDs very well. The only consistentdiscrepancy is at faint magnitudes where cool late-type stars are bluer in uvw − uvw • Using these isochrones, we measure age, reddening and distance for 49 clusters with well-definedcolor-magnitude sequences. For those that have been studied in detail before, we generally find agree-ment with literature values. However, there are a number of clusters for which we find significantlydifferent values than the literature, particularly in the third Galactic quadrant where some previousstudies have mistaken the disk sequence for the cluster. We catalog substantially revised parametersfor the clusters NGC 2304, NGC 2343, NGC 2360, NGC 2396, NGC 2428, NGC 2509, NGC 2533,NGC 2571, NGC 2818, Collinder 220 and NGC 6939. • We confirm the presence of an extended MSTO in two previously studied clusters – NGC 2360and NGC 2818. We also identify broad MSTO features in at least five other clusters that couldwarrant further spectroscopic investigation. • Most of our clusters have minimal reddening and are thus unsuited to probe the UV propertiesof the foreground dust. However, one cluster – Collinder 220 – shows significant improvement in theisochrone fits if an “LMC-like” reddening law – one with a smaller red bump at 2175 ˚A and a steeperextinction curve – is used.Our investigation has only scratched the surface of what this remarkable data set can yield. Mea-suring the integrated light would allow direct comparison between isochrones and synthetic spectralmodels to test the validity of the latter for unresolved stellar populations. Combining this databasewith extra ground-based optical photometry or the broad-band photometry of GAIA DR2 would al-low the detection of unresolved white dwarf-main sequence binaries, providing additional insight intothe role of binarism in spectral synthesis models and the integrated luminosity of stellar population(see, e.g., Buzzoni et al. 2012, Hern´andez-P´erez & Bruzual 2013). A more thorough examination ofthe MSTOs – especially in combination with spectroscopic study – would allow a much more detailedexploration of eMSTOs and their connection to cluster age.The authors acknowledge sponsorship at PSU by NASA contract NAS5-00136. This research wasalso supported by the NASA ADAP through grants NNX13AI39G and NNX12AE28G. The authorsthank L. Girardi for useful discussions about the NUV isochrones. The Institute for Gravitation andthe Cosmos is supported by the Eberly College of Science and the Office of the Senior Vice Presidentfor Research at the Pennsylvania State University.REFERENCES
Ahumada, J. A. 2003, RMdAA, 39, 41 Ahumada, J. A. 2005, Astronomische Nachrichten,326, 3 Siegel et al.
Ahumada, J. A., & Lapasset, E. 2007, A&A, 463,789Alter, G., Balazs, B., Ruprecht, J., & Vanysek, J.1970, in Catalogue of star clusters andassociations 2nd edn., ed. G. Alter, B. Balazs,& J. Ruprecht (Budapest: Akademiai Kiado)Andreuzzi, G., Bragaglia, A., Tosi, M., &Marconi, G. 2004, MNRAS, 348, 297Ann, H. B., Lee, S. H., Sung, H., et al. 2002, AJ,123, 905Anthony-Twarog, B. J., & Twarog, B. A. 2004,AJ, 127, 1000Anthony-Twarog, B. J., Tanner, D., Cracraft, M.,& Twarog, B. A. 2006, AJ, 131, 461Anthony-Twarog, B. J., Deliyannis, C. P., &Twarog, B. A. 2016, AJ, 152, 192Anthony-Twarog, B. J., Lee-Brown, D. B.,Deliyannis, C. P., & Twarog, B. A. 2018, AJ,155, 138Balaguer-N´u˜nez, L., Jordi, C., & Galad´ı-Enr´ıquez,D. 2005, A&A, 437, 457Barker, H., & Paust, N. E. Q. 2018, PASP, 130,034204Bastian, N., Kamann, S., Cabrera-Ziri, I., et al.2018, MNRAS, 480, 3739Bica, E., Bonatto, C., & Blumberg, R. 2006,A&A, 460, 83B¨ocek Topcu, G., Af¸sar, M., Schaeuble, M., &Sneden, C. 2015, MNRAS, 446, 3562Bostancı, Z. F., Ak, T., Yontan, T., et al. 2015,MNRAS, 453, 1095Bragaglia, A., Gratton, R. G., Carretta, E., et al.2012, A&A, 548, A122Breeveld, A. A., et al. 2011, AIP Conf. Proc.1358, 373Brown, T. M., Sweigart, A. V., Lanz, T., et al.2010, ApJ, 718, 1332Bruzual A., G. 2009, A&A, 500, 521Burrows, D. N., et al. 2005, Space ScienceReviews, 120, 165Buzzoni, A., Bertone, E., Carraro, G., & Buson,L. 2012, ApJ, 749, 35Calzetti, D., Lee, J. C., Sabbi, E., et al. 2015, AJ,149, 51Cantat-Gaudin, T., Jordi, C., Vallenari, A., et al.2018, A&A, 618, A93 [CG18]Carraro, G., & Chiosi, C. 1994, A&A, 288, 751Carraro, G., Chiosi, C., Bressan, A., & Bertelli, G.1994, A&AS, 103, 375 Carraro, G., Girardi, L., & Chiosi, C. 1999,MNRAS, 309, 430Carraro, G., & Costa, E. 2007, A&A, 464, 573Carraro, G., V´azquez, R. A., Costa, E., Ahumada,J. A., & Giorgi, E. E. 2015, AJ, 149, 12Casamiquela, L., Carrera, R., Jordi, C., et al.2016, MNRAS, 458, 3150Chavez, M., Bertone, E., Buzzoni, A., et al. 2007,ApJ, 657, 1046Claria, J. J. 1985, A&AS, 59, 195Clari´a, J. J., Piatti, A. E., Lapasset, E., & Parisi,M. C. 2005, Baltic Astronomy, 14, 301Clari´a, J. J., Piatti, A. E., Mermilliod, J.-C., &Palma, T. 2008, Astronomische Nachrichten,329, 609Conrad, C., Scholz, R.-D., Kharchenko, N. V., etal. 2014, A&A, 562, A54Cordoni, G., Milone, A. P., Marino, A. F., et al.2018, ApJ, 869, 139Daniel, S. A., Latham, D. W., Mathieu, R. D., &Twarog, B. A. 1994, PASP, 106, 281Davidge, T. J. 2013, PASP, 125, 115De Martino, C., Bianchi, L., Pagano, I., Herald,J., & Thilker, D. 2008,Mem. Soc. Astron. Italiana, 79, 704Dias, W. S., Alessi, B. S., Moitinho, A., & L´epine,J. R. D. 2002, A&A, 389, 871Dobbie, P. D., Day-Jones, A., Williams, K. A., etal. 2012, MNRAS, 423, 2815Donati, P., Bragaglia, A., Carretta, E., et al. 2015,MNRAS, 453, 4185Friel, E. D. 1995, ARA&A, 33, 381Frinchaboy, P. M., Mu˜noz, R. R., Phelps, R. L.,Majewski, S. R., & Kunkel, W. E. 2006, AJ,131, 922Frinchaboy, P. M., & Majewski, S. R. 2008, AJ,136, 118Frogel, J. A., & Twarog, B. A. 1983, ApJ, 274, 270Gaia Collaboration, Prusti, T., de Bruijne,J. H. J., et al. 2016, A&A, 595, A1Gaia Collaboration, Brown, A. G. A., Vallenari,A., et al. 2018, A&A, 616, A1Gehrels, N., et al. 2004, ApJ, 611, 1005Geller, A. M., Latham, D. W., & Mathieu, R. D.2015, AJ, 150, 97Georgy, C., Charbonnel, C., Amard, L., et al.2019, A&A, 622, A66Giorgi, E. E., V`azquez, R. A., Baume, G.,Seggewiss, W., & Will, J.-M. 2002, A&A, 381,884
VOT Stars: Open Clusters Glaspey, J. W. 1987, PASP, 99, 1089Gordon, K. D., Clayton, G. C., Misselt, K. A.,Landolt, A. U., & Wolff, M. J. 2003, ApJ, 594,279Goudfrooij, P., Puzia, T. H., Kozhurina-Platais,V., & Chandar, R. 2011, ApJ, 737, 3Hagen, L. M. Z., Siegel, M. H., Hoversten, E. A.,et al. 2017, MNRAS, 466, 4540Hagen, L. M. Z., Siegel, M. H. & Gronwall, C. A.,2019, MNRAS, submitted