A Sequoia in the Garden: FSR 1758—Dwarf Galaxy or Giant Globular Cluster?^∗

More than two dozen of new low-luminosity globular cluster (GC) candidates have been discovered in the past year to the direction of the Galactic Bulge (Minniti et al. 2017a, 2017b; Bica et al. 2018; Camargo 2018; T. Palma et al. 2019, in preparation; Ryu & Lee 2018). These objects are very difficult to detect, due to the heavy extinction and high field stellar density lying well inside the Bulge, and if proved to be genuine clusters, most are expected to be of low mass.

Very recently, Cantat-Gaudin et al. (2018), on the basis of the Gaia optical color–magnitude diagram (CMD), proposed another object, [FSR2007] 1758 (hereafter FSR 1758), to be a new GC located toward the Bulge. This serendipitous discovery was made while studying 1229 open clusters and noticing the striking proper motion (PM) separation from the field stellar population. They estimated a Galactocentric distance R_G = 1600 pc and a height below the plane z = −470 pc, which places it inside the Bulge itself. This object was listed as a diffuse open cluster in the catalogs of Froebrich et al. (2007) and Kharchenko et al. (2013), although we note that the position and physical properties in these initial studies are very preliminary.

In this Letter, we use combined data from 2nd Gaia Data Release (GDR2; Gaia Collaboration et al. 2018a), the DECam Plane Survey (DECaPS; Schlafly et al. 2018), and the VISTA Variables in the Vía Láctea Extended (VVVX) Survey (Minniti 2018) to investigate the physical properties of this impressive object in much greater detail. We also use OGLE RR Lyrae stars to provide an external distance determination. Sections 2–4 describe the discovery, observations used, and derived CMDs, respectively. Then, based on our findings, we discuss its physical nature in Section 5 and summarize some conclusions in Section 6.

One of us (R.B.) made the independent discovery of FSR 1758 serendipitously by visual inspection of the images from the DECaPS Survey of Schlafly et al. (2018). It stands out as a bright, diffuse glow in an otherwise patchy and generally high extinction zone, next to an extended dark cloud (Figure 1). A zoom-in reveals that the diffuse object is really a plethora of stars. A quick inspection of the GDR2 PMs in the region revealed that the cluster motion is very different from the field stars, as also noted by Cantat-Gaudin et al. (2018).

Zoom In Zoom Out Reset image size

FSR 1758 appears to be quite large, perhaps rivaling in size the largest Galactic GCs, ω Cen and NGC 2419 (e.g., Harris 1996; Ripepi et al. 2007). The visible part seen in the DECaPS images (Figure 1) is probably just "the tip of the iceberg," being that much of its population is possibly hidden by field contamination and differential reddening. Part of the area has surprisingly low reddening, and it is this region that we use to obtain the best cluster CMDs in order to determine its parameters. Nevertheless, approximately one-third of the object is heavily obscured by one or more interstellar clouds toward the plane. Thus, evidence suggests that the cluster may be more extended and/or have tidal tails, as discussed below.

Because of the large non-uniform reddening in the bulge fields studied here (e.g., González et al. 2012; Alonso-García et al. 2015, 2018; Minniti et al. 2018; Schlafly et al. 2018), we also use the combination of optical GDR2 with near-IR VVVX data to help determine the parameters.

The GDR2 (Gaia Collaboration 2018) significantly improves the photometric and astrometric measurements of the first release and provides additional information on astrophysical parameters, variability, and median radial velocities for some sources. The GDR2 data have been processed by the Gaia Data Processing and Analysis Consortium (DPAC) and contain G-band magnitudes for over 1.6 × 10⁹ sources with G < 21 mag and broadband colors G_BP covering (330–680 nm) and G_RP covering (630–1050 nm) for 1.4 × 10⁹ sources. PM components in equatorial coordinates are available for 1.3 × 10⁹ sources, with an accuracy of 0.06 mas yr⁻¹, 0.2 mas yr⁻¹, and 1.2 mas yr⁻¹, for sources with G < 15 mag, G ∼ 17 mag, and G ∼ 20 mag, respectively (Gaia Collaboration 2018).

DECaPS (Schlafly et al. 2018) is an optical multiband survey of the Southern Galactic Plane performed with the DECAM camera attached to the Victor Blanco 4 m telescope at Cerro Tololo Inter-American Observatory (Chile).

The VVVX survey (Minniti 2018) maps the Galactic Bulge and southern disk in the near-IR with the VISTA InfraRed CAMera (VIRCAM) at the 4.1 m wide-field Visible and Infrared Survey Telescope for Astronomy (VISTA; Emerson & Sutherland 2010) at ESO Paranal Observatory (Chile). In the Galactic Bulge, the VVVX Survey covers about 600 deg², using the J (1.25 μm), H (1.64 μm), and Ks (2.14 μm) near-IR passbands. The VVVX Survey data reduction and the archival merging were carried out at the Cambridge Astronomical Survey Unit (CASU; Irwin et al. 2004) and VISTA Science Archive at the Wide-Field Astronomy Unit, within the VISTA Data Flow System (Cross et al. 2012). In order to deal with the high crowding in this VVVX, we follow Alonso-García et al. (2018), extracting the point-spread function (PSF) photometry and obtaining a highly complete near-IR catalog.

Figure 2 (top panels) show the optical CMDs using the DECaPS photometry from Schlafly et al. (2018). The photometry is very deep, and the cluster red giant branch (RGB) is clearly seen, although heavily contaminated by foreground RGB stars. However, the most striking indication of the cluster is the presence of an extended blue horizontal branch (BHB), which is absent in the surrounding fields.

Zoom In Zoom Out Reset image size

Figure 2 (bottom left panel) shows the near-IR CMD within 5' of the cluster center obtained from PSF photometry of the VVVX tile e682, following the procedure from Alonso-García et al. (2018). The cluster RGB and extended BHB are clearly seen, on top of the foreground disk and background Bulge stars. In particular, the cluster RGB is bluer than the Bulge RGB. A clump of stars located at Ks = 13.4, $J-{Ks}=0.87$ can be identified; this feature is likely the red giant branch bump (RGBb). This interesting GC CMD feature is related to the evolution of the RGB stars during the first dredge-up, and it is sensitive to metallicity, helium content, and mixing efficiency (see Fu et al. 2018). The locus in a CMD of the RGBb depends on both age and metallicity. Once the cluster metallicity has been estimated, the location of the RGBb in a CMD helps to determine the age, and vice versa (Alves & Sarajedini 1999). The optical and near-IR CMDs exhibit a well-populated steep RGB, with no clear red clump, an indication of a metal-poor GC, in agreement with the presence of a prominent BHB.

Schlegel et al. (1998) and Schlafly & Finkbeiner (2011) determined the extinction in the area to be in the ranges ${A}_{V}={\rm{3.35:2.89}}$ (Landolt filters), ${A}_{I}={\rm{2.11:1.79}}$ (SDSS filters), and A_K = 0.37:0.32 (UKIRT filters). By comparing the near-IR CMD with known metal-poor clusters, we obtained $E(J-{Ks})=0.20\pm 0.03$ mag, in agreement with the extinction values from Schlafly & Finkbeiner (2011)

Knowing the reddening, we can accurately measure the distance differentially with respect to the known Bulge GCs NGC 6642 and NGC 6266, using the VVVX near-IR CMDs. With respect to NGC 6642, we measured ${\rm{\Delta }}(J-{Ks})\,=0.10$ mag and ${\rm{\Delta }}{Ks}=0.90$ mag. This yields a fainter distance modulus (by about 0.85 ± 0.1 mag) than the GC NGC 6642, whose distance and metallicity are D = 8.1 kpc and $[\mathrm{Fe}/{\rm{H}}]\,=-1.26$ dex, respectively. Our first estimate of the distance to $\mathrm{FSR}\,1758$ is then D₁ = 12.0 ± 0.5 kpc. Similarly, with respect to NGC 6266, we measured ${\rm{\Delta }}(J-{Ks})=0.10$ mag and ${\rm{\Delta }}{Ks}=0.95$ mag. This gives a fainter distance modulus than the GC NGC 6266, which has $D=6.8\,\mathrm{kpc}$ and $[\mathrm{Fe}/{\rm{H}}]\,=-1.18$ dex. Then, our second estimate is ${D}_{2}=11.0\pm 0.5\,\mathrm{kpc}$. We can take the average of the former two values as the distance to FSR 1758, i.e., ${D}_{\mathrm{FSR}1758}=11.5\pm 1.0\,\mathrm{kpc}$. The error is estimated as the sum of the individual errors in order to include possible systematic differences. The alignment of the RGB and the BHB with these comparison GCs is remarkably good, which gives added confidence to this determination. We note that the GDR2 parallaxes in the distant and crowded cluster field would be unreliable.

We repeated the same procedure using the Gaia CMD and considering the clusters studied by Gaia Collaboration et al. (2018b) as references. Using this optical CMD we find a magnitude difference of 17.1 ± 0.1 mag and a reddening $E({BP}-{RP})=0.90\pm 0.05$ mag. Using the extinction ratio from Andrae et al. (2018) yields ${AG}=2.0\,E({BP}-{RP})\,=1.8$ mag. This gives an absolute distance modulus of ${(m-M)}_{0}\,=15.30\pm 0.10$ mag, equivalent to a distance $D=11.5\,\pm 0.5\,\mathrm{kpc}$. The Gaia CMD distance is then very consistent with the determination using the near-IR CMD. We note here that the fitting uncertainty of about 0.1 mag is quite optimistic because it does not consider the uncertainty in the reddening law. Therefore, as mentioned before, a distance error of the order of 1 kpc must be considered. As a conclusion, FSR 1758 is located on the far side beyond the Bulge. Its corresponding Galactocentric distance is R_G = 3.7 kpc, and its height below the plane is z = −760 pc.

Figure 3 (middle panel) shows the optical Gaia CMD, decontaminated using PM data. Again, the cluster RGB and extended BHB are clearly visible. The cluster RGBb is also well defined, located at G = 16.9, ${BP}-{RP}=2.0$. The RGB tip is located at Gaia magnitude G = 13.5, bright enough for future spectroscopic follow-up. Figure 3 also shows the cluster PM (left panel), which is strikingly different from nearby field stars. This characteristic led Cantat-Gaudin et al. (2018) to claim the discovery of a new GC. The longitudinal component of the PM is close to zero, and indeed this cluster has a PM indicating it is plunging into the disk.

Zoom In Zoom Out Reset image size

The wide wavelength coverage available allows us to constrain the cluster metallicity. Once the reddening is known, the metallicity can be readily estimated from the optical and near-IR CMDs. Compared to other well-known GCs (e.g., Gaia Collaboration et al. 2018b), the Gaia CMD of FSR 1758 resembles those of M3 and M13, thus suggesting a metallicity close to [Fe/H] = −1.5 dex. Using the RGB location interpolated in the optical CMDs from Gaia Collaboration et al. (2018b) and the near-IR CMDs from Valenti et al. (2004), we estimated a value of [Fe/H] = −1.5 ± 0.3 dex for the metallicity of this cluster.

A concentration of RR Lyrae variables is also evident in the cluster region. From the OGLE catalog by Soszyński et al. (2014), we found eight fundamental RR Lyrae pulsators (RRab), plus three first overtone RR Lyrae pulsators (RRc) within 015 (9') from the cluster center. The centroid of the RR Lyrae distribution is located to the west of the centroid of the BHB distribution, probably due to a differential reddening effect. Table 1 lists the RRLyr stars including their OGLE IDs, types, periods, GDR2 equatorial coordinates and PMs (in J2000), OGLE mean V- and I-band and GDR2 magnitudes, extinctions A_I, distance from the cluster center, and probable cluster membership. We select the most likely cluster members based on the sky positions, PMs, and mean magnitudes and positions in the CMDs. The mean period for the five RRab members is $\langle P\rangle$ = 0.684 day. As a consequence, we classify this cluster as an Oosterhoff type II, which is consistent with being metal-poor, like, for example, ω Cen (Clement & Rowe 2000).

RR Lyrae can be used to determine GCs distances even in deeply reddened cases (e.g., Alonso-García et al. 2015; Minniti et al. 2017b). We measured the distance to FSR 1758 differentially with respect to ω Cen following Braga et al. (2018), who relied on the period–luminosity-metallicity relations by Marconi et al. (2015). Considering all eight RRab listed in Table 1, we obtained a mean distance modulus of $\langle m-{M}_{0}\rangle =15.16\pm 0.3$ mag, equivalent to a distance $D\,=10.8\pm 1.0\,\mathrm{kpc}$. On the other hand, using only the five RRab that are most probably cluster members we obtained $\langle m-{M}_{0}\rangle \,=15.02\pm 0.3$ mag, equivalent to $D=10.1\pm 1.0\,\mathrm{kpc}$. Although the former estimate has a larger scatter due to the large individual reddening corrections, it is still in agreement with the distance determination based on the optical and near-IR CMDs.

Figure 4 shows the radial profile and spatial extension of FSR 1758, obtained combining DECaPS and GDR2 data. We isolated stars with PMs similar to the cluster, within 1.2 mas from μ_α = −2.85 mas yr⁻¹, and μ_δ = 2.55 mas yr⁻¹, and parallaxes smaller than 0.3 mas, in order to avoid most of the foreground stars. The DECaPS star counts in i-band (shown in the left panel) indicates that the cluster stellar density joins the field at large radii from the center (R > 15'). The differential extinction in the field produces a structured radial profile. The fitting of a King profile (King 1962, 1966) gives a core radius ${R}_{c}=0\buildrel{\circ}\over{.} 050\pm 0\buildrel{\circ}\over{.} 004$ (about 10 ± 1 pc) and a tidal radius ${R}_{t}=0\buildrel{\circ}\over{.} 78\pm 0\buildrel{\circ}\over{.} 22$ (about 150 ± 45 pc). The derived value of R_t = 150 pc should be taken with caution due to the presence of strong differential reddening in the area. Although the structure of FSR 1758 determined from the star counts yields a concentrated radial profile typical of a GC, with a concentration index $c=1.20{\pm }_{0.13}^{0.19}$, it suggests an extended nature of the cluster. Figure 4 shows that this object is indeed very extended, possibly even larger than ω Cen, the most massive GC in the Milky Way (e.g., Meylan 1987; Harris 1996, 2010; Ferraro et al. 2006), with a tidal radius of R_t = 45' (Trager et al. 1995) and a mass of 4 × 10⁶  M_⊙ (D'Souza & Rix 2013). It is also a significantly flattened GC (White & Shawl 1987; Chen & Chen 2010). The appearance of FSR 1758 is also very flattened, like ω Cen, although this issue needs further investigation since the dark cloud located to the northeast of the cluster center could be significantly affecting this result. The total cluster luminosity is very difficult to estimate in the presence of the high background and heavy differential reddening. A lower crude estimate was obtained using the DECaPS photometry in the i band. Coadding up all stars within 01 from the GC center (to avoid further field contamination), after accounting for the background taken in four different fields of similar area surrounding the cluster, assuming a distance of 11.5 ± 1.0 kpc, and an extinction of A_i = 1.79 mag from Schlafly & Finkbeiner (2011), we obtain a total i-band absolute magnitude brighter than ${M}_{i}\lt -8.6\pm 1.0$ mag. This is a very bright GC indeed, and we emphasize that this total magnitude is only a lower limit because of incompleteness and differential reddening. A detailed study of reddening is needed in order to improve the values of astrophysical parameters allowing for a more robust comparison with ω Cen.

Zoom In Zoom Out Reset image size

The large size of the cluster is confirmed using stellar tracers like BHB stars. Note that BHB stars are particularly sensitive to extinction, and therefore the empty patches just represent high extinction fields. Because there was no clear edge to the distribution of BHB stars, we decided to explore the spatial distribution of comoving stars using the GDR2 PMs. We then searched for associated streaming motions within two degrees of the cluster, using the same set of stars isolated from GDR2. Then, we selected the stars that lie in the main locus of the cluster RGB and BHB, with an additional constraint of G < 18.8. The result, shown in the right panel of Figure 4, reveals an asymmetric source distribution, with the potential comoving stars located preferentially to the southeast. This suggests that FSR 1758 may be the nucleus of a dwarf galaxy, which we tentatively name as Scorpius dwarf galaxy, joining in this category with the GCs ω Cen (Meylan et al. 2001), M54 (Monaco et al. 2005), and M31-G1 (Gregg et al. 2015). In particular, some of the BHB stars in the outskirts of the field studied may belong to the body of this putative dwarf galaxy, in analogy with the BHB stars found by Monaco et al. (2003) in the field surrounding the GC M54, which is now known to be the nucleus of the Sgr dwarf galaxy. It could also be one of the putative primordial bulge building blocks such as Terzan 5 (Ferraro et al. 2009).

Also interesting in this regard is the fact that FSR 1758 appears not to fit to the well-defined size–metallicity and size–Galactocentric radius relations for the Galactic GCs (e.g., Vanderbeke et al. 2015), but to lie far above the mean correlations observed between these parameters. A thorough comparison of its properties with those of dwarf galaxies is beyond the scope of this Letter. For example, in the absence of radial-velocity data, it is not possible to determine its mass-to-light ratio. Given its metallicity, if it was a dwarf galaxy, its stellar mass would be of the order of ${10}^{7}\,{M}_{\odot }$, according to the mass–metallicity relation for dwarf galaxies (Kirby et al. 2013). This relation could remain valid even if the object is tidally stripped.

Since FSR 1758 is close in projection to two other GCs belonging to the Bulge, Ton 2 and NGC 6380, we explored the possibility of a real association with any of them. We found that they all share similar PMs; all three clusters might be plunging onto the Galactic disk.

With an angular separation of only 08 and a probable physical separation of about 600 pc (perhaps even smaller considering the large errors in the distances), the association of FSR 1758 with NGC 6380 (Harris 1996, 2010) is strong. They might be a binary cluster or part of a larger structure like a dwarf galaxy. However, their compositions are different by 0.7 dex, with NGC 6380 being more metal-rich. The angular separation between Ton 2 and FSR 1758 is 14. The distance to the former cluster is believed to be D = 8.2 kpc, although it is not well established, and its metallicity was determined to be [Fe/H] = −0.70 dex. Given the different metallicities, a putative association is not strongly supported. In addition, the radial velocities of Ton 2 and NGC 6380 are very different, −182 km s⁻¹ and −4 km s⁻¹, respectively, so these clusters are definitely not associated with each other.

Spectroscopic data of a number of stellar members of FSR 1758 are essential, not only to compute the radial velocity, necessary to determine the orbital path of the object, but also to derive information about its origin and potential association with any of the known GCs and to obtain metallicity estimates. Combining the velocity dispersion and metallicities for various stars will allow us to definitively determine the nature of this object by estimating its dark matter content and whether or not it possesses a metallicity spread.

FSR 1758 is a new GC recently discovered serendipitously in the Milky Way Bulge. In this Letter, we used combined data from the Gaia, DECaPS, and VVVX surveys to determine the cluster physical parameters for the first time, including its position, distance, reddening, size, metallicity, and mean PM. The cluster is centered at equatorial coordinates α =17:31:12, δ = −39:48:30 (J2000), and Galactic coordinates l = 349217, b = −3292 degrees. A distance of D = 11.5 ±0.5 kpc was estimated using the optical and near-IR CMDs, and confirmed with OGLE RR Lyrae variable stars. From the CMDs, extinction values were also measured, resulting in $E(J-{Ks})=0.20\pm 0.03$ mag, and $E({BP}-{RP})=0.90\,\pm 0.05$ mag. The metallicity of FSR 1758 was estimated to be $[\mathrm{Fe}/{\rm{H}}]=-1.5\pm 0.3$ dex, based on the appearance of the RGB in the optical and near-IR CMDs. Finally, its derived mean PM, μ_α = −2.85 mas yr⁻¹ and μ_δ = 2.55 mas yr⁻¹, indicates that this cluster is plunging onto the Galactic plane. The acid test for this cluster will be to obtain spectra for a number of members. The measurement of radial velocities which, combined with the distance and PM information, will give us the orbital parameters for the individual stars, and the chemical information will allow us to search for any metallicity variation or the presence of multiple populations.

The stellar density distribution shows that FSR 1758 is a very large cluster, with a core radius R_c = 10 pc and a large tidal radius of about R_t = 150 pc, with a lower limit for the i-band luminosity of about M_i < −8.6. The inspection of a larger field of view of several degrees around the cluster led to the detection of extended streaming motions, i.e., a collection of stars with characteristics similar to the BHB and RGB stars of the cluster, with coherent PMs, and moving parallel to the cluster. This points to the hypothesis that this object is actually much larger, possibly the nucleus of a dwarf galaxy.

The census of Galactic GCs is not complete, and several low-luminosity GCs may still be missing in the Galactic Bulge (Bica et al. 2016; Minniti et al. 2017a; Camargo 2018; Ryu & Lee 2018). The discovery of FSR 1758 represents a paradigm shift, as it clearly shows that not only low-luminosity GCs may be missing in the Bulge, but also some quite luminous and massive ones. Future searches to be carried out with the Large Synoptic Survey Telescope (Ivezic et al. 2008) or with the Wide Field Infrared Survey Telescope (Spergel et al. 2015; Stauffer et al. 2018) might detect more GCs hidden in this region. Finally, the recently started multi-epoch observation campaign for the VVVX survey would significantly extend the areal coverage, allowing us to search for more RR Lyrae variable stars likely associated with this new large GC.

We thank the reviewer, who did a thorough reading, which helped to substantially improve the manuscript. We gratefully acknowledge data from the ESO Public Survey program ID179.B-2002 taken with the VISTA telescope, and products from the Cambridge Astronomical Survey Unit (CASU). We have made use of tools from Aladin/Simbad at the Centre des Donnees Stelaires (CDS) Strassbourg, and TopCat (Taylor 2005). R.H.B. thanks support from DIDULS project PR18143, and the "Big Data and Cosmography: the skyscape of large astrophysical surveys" project. D.M. and D.G. gratefully acknowledge support provided by the BASAL Center for Astrophysics and Associated Technologies (CATA) through grant AFB-170002. D.G. also acknowledges financial support from the Dirección de Investigación y Desarrollo de la Universidad de La Serena through the Programa de Incentivo a la Investigación de Académicos (PIA-DIDULS). D.M. acknowledges the Ministry for the Economy, Development and Tourism, Programa Iniciativa Científica Milenio grant IC120009, awarded to the Millennium Institute of Astrophysics (MAS), and from FONDECYT Regular grant 1170121. J.A-G. acknowledges support by FONDECYT Iniciation grant 11150916, and by the Ministry of Education through grant ANT-1656. A.M. acknowledges support from FONDECYT Regular grant 1181797. F.G. acknowledges support from FONDECYT Regular grant 1181264. F.G. and A.M. acknowledge funding from the Max Planck Society through a "Partner Group" grant. J.A. acknowledges support from DIDULS project PR16142.