Movatterモバイル変換

• Top
• Abstract
• 1. Introduction
• 2. The VVV survey data
• 3. The globular cluster UKS 1
• 4. The globular cluster candidate VVV CL001
• 5. Conclusions
• Acknowledgments
• References
• List of tables
• List of figures

© ESO, 2011

1. Introduction

The inner regions of the Milky Way have been mapped extensively at all wavelengths. Nethertheless, it is unknown whether more distant globular clusters await discovery, hidden beyond the bulge, because of the high density of stellar sources and the large and inhomogeneous interstellar extinction. Near-IR surveys provide an invaluable resource in searching for these clusters. The 2MASS, for instance, discovered two new globular clusters (Hurt et al. 2000). But its limiting magnitude ofK_S = 14.3 (for 10σ-detections,Skrutskie et al. 2006) may prevent the discovery of fainter objects, especially if they are located in highly reddened regions.

Fig. 1 Discovery image (left panel,JHKS colour composite) showing VVV CL001 on the left and UKS 1 on the right. The cluster UKS 1 was discovered byMalkan et al. (1980), to have a total magnitudeI < 16 mag, and was invisible in Palomar Survey plates (up to a few years ago, UKS 1 was known as “the faintest globular cluster of the Milky Way”). The right panel shows a zoom into the region centered on VVV CL001, corresponding to the white box ().

The asymmetry of the spatial distribution of known globular clusters around the Galactic centre indicates that previous observations may have overlooked some additional globular clusters.Ivanov et al. (2005) estimated that there may be about ten clusters missing towards the inner Milky Way. Discoveries in the last ten years include both faint (low mass) halo clusters, as well as reddened globular clusters projected toward the bulge, e.g. 2MASS GC01 and 2MASS GC02 byHurt et al. (2000) (see alsoIvanov et al. 2000), ESO 280 SC06 byOrtolani et al. (2000), GLIMPSE C01 by Kobulnicky et al. (2005; but see also Ivanov et al. 2005; Davies et al. 2010), GLIMPSE C02 byKurtev et al. (2008), AL-3 byOrtolani et al. (2006), FSR 1735 byFroebrich et al. (2007), Koposov 1 and Koposov 2 byKoposov et al. (2007), FSR 1767 byBonatto et al. (2007), Whiting 1 (Carraro 2005), and Pfleiderer 2 byOrtolani et al. (2009).

The VISTA Variables in the Via Lactea (VVV) public survey has started mapping the inner disk and bulge of our Galaxy with the VISTA 4m telescope (Visible and Infrared Survey Telescope for Astronomy) in the near-IR (Minniti et al. 2010;Saito et al. 2010). One of the main scientific goals of the VVV survey is to study the bulge globular clusters and to search for new clusters. Here we present VVV CL001, the first globular cluster candidate discovered by the VVV survey.

2. The VVV survey data

The VVV survey data are acquired with the VISTA 4 m telescope at ESO Paranal Observatory (Emerson & Sutherland 2010). The VVV field b351 was observed in theJHK_S bands under subarcsec seeing conditions (0.8 arcsec inK_S). TheYZ-band observations are still pending. Each one of the VVV fields (tiles) covers 1.636deg² in total (1.475° inl by 1.109° inb). The bulge field b351 observed here is centered atα = 17:50:05.42,δ =  −23:43:16.7,l = 4.987°,b = 1.838°.

We use here the images processed by the CASU VIRCAM pipeline v1.0 (e.g.Irwin et al. 2004). The photometry was obtained with DoPhot (Schechter et al. 1993). In addition, the photometry is uniformly calibrated against the 2MASS catalog (Skrutskie et al. 2006). The limiting magnitude of the single epoch VVV images isKs = 18.1 in the bulge fields (for details of the observing strategy, see Minniti et al. 2010). The distance probed along the line of sight depends on the reddening of the fields. For example, in zero-reddening disk fields we would see horizontal-branch red clump beyond 50 kpc. Thus, we can search for distant galactic globular clusters and measure their physical parameters.

Visual inspection of the images of the field b351 led to the serendipitous discovery of a star cluster candidate that we name VVV CL001. This object is located in the vicinity of the known globular cluster UKS 1 (Fig. 1). On the basis of near-infrared stellar density maps (see Fig. 2), we conclude that this is not a statistical fluctuation of the background, and that the cluster VVV CL001 is centered atα = 17:54:42.5,δ =  −24:00:53 (l = 5.27°,b = 0.78°). Applying the procedure developed byKoposov et al. (2008) andBelokurov et al. (2009), the statistical significance of the overdensity at the coordinates given above isS = 9.31. There is no source identified at this location based on radio, infrared, and visible data from the SIMBAD, 2MASS and NED databases, or any other star cluster catalog. The second obvious density peak atα = 17:54:38,δ =  −24:00:49 is due to a saturated star.

The proximity of VVV CL001 and UKS 1 on the sky (Fig. 1) made it appropriate to use the latter cluster as a comparison when measuring the reddening and distance of the new globular cluster candidate. We therefore concentrate first on measuring the parameters of UKS 1 using VVV Survey data.

3. The globular cluster UKS 1

Fig. 2 J-band image of the VVV CL001 region (left) and density of all objects. The contours in the left panel refer to the intensity level of the image. In contrast, to derive the significance of the overdensity (right) only detections with a Dophot classification parameter 1 or 7 (Schechter et al. 1993) were selected. The sources were binned into 60 × 60 pixel bins (20 by 20 arcsec). We note that even with PSF photometry we are unable to resolve the inner region of VVV CL001 (or UKS 1), which certainly affects the numerical results of our analysis (density, significance of the detected over-density). Both plots show the same 272 by 272 arcsec section of theJ-band image.

Previous infrared photometry and spectroscopy of UKS 1 giants revealed that it is a very distant and reddened cluster, with distance modulus  ≈ 15, andE(B − V) ≈ 3 (Minniti et al. 1995), and moderately metal-rich, with [Fe/H ]  =  −0.78 (Origlia et al. 2005).

To determine the distance to UKS 1, we use the horizontal-branch red clump, which is very conspicuous for this cluster. The red clump stars with known parallaxes measured by the Hipparcos satellite are well-calibrated standard candles (Paczynski & Stanek 1998;Alves 2000).Alves (2000) obtained aK-band calibration of the horizontal-branch red clump luminosity, and applied the calibration to the red clump of the Galactic bulge. For example, this calibration has also been applied to the red clump of the Large Magellanic Cloud (e.g.,Alves et al. 2002;Pietrzyński et al. 2003a;Borissova et al. 2009). The uncertainties in the reddening should in all cases be larger than the uncertainties due to the unknown metallicity. OurK_S magnitudes are in the 2MASS magnitude system, hence we transform theK-band magnitudes of the red clump stars, studied byAlves (2000) toK_S magnitudes usingK = K_S + 0.044 (fromGrocholski & Sarajedini 2002). The zero-point differences should be smaller than 0.02 mag (Alves et al. 2002). From the previously cited works, we adopt the following mean values for the red clump stars:M_Ks =  −1.65 ± 0.03, and (J − K_s)₀ = 0.71 ± 0.03 (López-Corredoira et al. 2002). The distance modulus to the horizontal-branch red clump in the globular cluster would be:$\mathit{\mu }\mathrm{=}{\mathit{m}}_{{\mathit{K}}_{\mathrm{s}}}\mathrm{-}{\mathit{A}}_{{\mathit{K}}_{\mathrm{s}}}\mathit{/}\mathrm{(}{\mathit{A}}_{\mathit{J}}\mathrm{-}{\mathit{A}}_{{\mathit{K}}_{\mathrm{s}}}\mathrm{)}\mathrm{\left[}\mathrm{\right(}\mathit{J}\mathrm{-}{\mathit{K}}_{\mathrm{s}}\mathrm{)}\mathrm{-}\mathrm{(}\mathit{J}\mathrm{-}{\mathit{K}}_{\mathrm{s}}{\mathrm{)}}_{\mathrm{0}}\mathrm{]}\mathrm{-}{\mathit{M}}_{{\mathit{K}}_{\mathrm{s}}}\mathit{.}$(1)Adopting the mean red-clump magnitudes and colours discussed above, and the reddening coefficients fromCardelli et al. (1989),

this simplifies to$\mathit{\mu }\mathrm{=}\mathrm{5}\log \mathit{d}\mathrm{\left(}\mathrm{pc}\mathrm{\right)}\mathrm{-}\mathrm{5}\mathrm{=}{\mathit{m}}_{{\mathit{K}}_{\mathrm{s}}}\mathrm{-}\mathrm{0.73}\mathrm{(}\mathit{J}\mathrm{-}{\mathit{K}}_{\mathrm{s}}\mathrm{)}\mathrm{+}\mathrm{2.17.}$(2)Using this equation, we compute the mean distance modulus (and distance in kpc) for the horizontal-branch red clump in UKS 1, which has mean values ofK_S = 15.3 andJ − K_S = 2.0 (Fig. 4). This yields a meanE(J − K) = 1.30, equivalent toE(B − V) = 2.5, and a distance modulus (m − M)₀ = 16.01, equivalent to 15.9 kpc. The uncertainty in the red-clump position was estimated to be Δm_KS = 0.012 mag and Δ(J − K_S) = 0.08 mag, based on analysis of the relevant region of the CMD. This results in an error in the distance modulus and corresponding distance of Δμ = 0.078 and ΔD = 0.6 kpc. The distance to UKS 1 was also measured byOrtolani et al. (2007) using deep CST photometry. They found the mean magnitude of the horizontal branch to beJ = 17.69, and the mean colourJ − H = 1.39   mag or 1.71   mag depending on the calibration, and estimated a distance of betweenD = 9.3−14.3 kpc, placing it beyond the Galactic centre. They inferred a total reddening ofE(B − V) = 2.94−3.59   mag also depending on the calibration. Our values are consistent with the lower reddening values ofOrtolani et al. (2007). However, our distance measurement favours the largest distance measurement ofOrtolani et al. (2007), and is consistent with UKS 1 being located well beyond the Galactic centre, but not as far as the Sgr dwarf galaxy discovered byIbata et al. (1994).

4. The globular cluster candidate VVV CL001

Fig. 3 Discovery near-infrared colour − magnitude diagram of VVV CL001 for a 15 arcsec radius area centered on the cluster (left) compared with a field of the same area located east (middle), and a field of UKS 1 of the same size that is offset from the cluster centre by () (right). The small offset for UKS 1 was applied, since the centre of this cluster is barely resolved. The horizontal line atKs = 14.3 shows the limit of 2MASS photometry. Comparing UKS 1 and VVV CL001, we find that the RGB of UKS 1 is much more populated, hence VVV CL001 is far less massive then UKS 1.

Fig. 4 CMD of VVV CL001 (diamonds) compared with UKS 1 (dots). The horizontal-branch red clump of UKS 1 is seen atKS = 15.3,J − KS = 2.0. This diagram shows that the RGB of VVV CL001 is slightly bluer than that of UKS 1, from which one can conclude that VVV CL001 is less reddened or more metal-poor than UKS 1, and that the RGB of UKS 1 is much more populated, indicative of the low mass of VVV CL001. The legend gives the radius of the selected cluster region, whereas the arrows indicate the reddening derived for UKS 1.

Based on the CMD and assuming similar reddenings for both clusters, we can conclude that VVV CL001 has a similar metallicity or is slightly more metal-poor than UKS 1. It would otherwise be hard to reconcile the clearly different colours of the RGBs of these two clusters (see Fig. 3), particularly since a large difference in metallicity is impossible. The integrated near-infrared luminosity for Milky Way globular clusters are not very well known (Cohen et al. 2007, and references therein), nevertheless by comparing the RGB population in VVV CL001 and UKS 1 we also estimate that the new cluster candidate is  ≈ 3.7 mag fainter inK_S than UKS 1.

Thus, even though we cannot determine its distance accurately, the comparison with UKS 1 allows us to conclude that VVV CL001 is located beyond the Galactic centre, on the opposite side of the Milky Way. Therefore, this does not appear to be a bulge globular cluster according to the definitions ofMinniti et al. (1995) andBarbuy et al. (1998).

For consistency, we can also determine the reddening from the mean location of the VVV CL001 red giants in theJ − H versusH − K_S colour-colour diagram. The meanJ − H = 1.4 andH − K_S = 0.4 yieldA_V = 6   mag andE(B − V) = 2.0, which are slightly less reddened than UKS 1. On the other hand, the maps ofSchlegel et al. (1998) give a mean reddening ofE(B − V) ≈ 3.3 for this region of VVV field b351, which is larger than our measurements, but the reddening towards the bulge is known to be non-uniform and patchy.

The data are not good enough to measure the distance, but we have an accurate reddening and location, and deeper follow-up observations would allow us to define the other structural parameters of VVV CL001. On the basis of the number of giants, we crudely estimate that this cluster has a low mass, being about 50 − 100 times less massive than UKS 1. In that case, VVV CL001 would join the list of very low-luminosity globular clusters in the Milky Way (Table 1, see alsoKoposov et al. 2007). Given the uncertainty in the cluster distance, the possibility of VVV CL001 being an old, compact open cluster is at this point not strictly ruled out, although the high stellar concentration and the well populated RGB favour the globular cluster interpretation.

Alternatively, this may be just a chance alignment, and VVV CL001 could be a globular cluster or, although less likely, an old open cluster (see Sect. 4), either more distant or closer to the Sun than UKS 1. Deeper photometric observations are also clearly needed for this object.

5. Conclusions

The VVV survey is searching for missing globular clusters in the inner regions of the Milky Way galaxy. We have reported the discovery of VVV CL001 (Fig. 1), a low-mass globular cluster candidate atα = 17:54:42.5,δ =  −24:00:53. This is located only 8 arcmin away from the Galactic globular cluster UKS 1 in the sky. We have taken advantage of this spatial proximity to use UKS 1 as reference in order to estimate the parameters of this newly discovered cluster. We have measured the distance and reddening of UKS 1, findingE(B − V)_UKS1 = 2.5 andD_UKS1 = 15.9 kpc.

We have presented the first colour − magnitude diagrams of the new cluster VVV CL001 (Fig. 3), estimatingE(B − V) = 2.0. We have been unable to define the meanK_S-band magnitude of the horizontal-branch red clump for this cluster, hence and its distance remains uncertain. A comparison with the RGB of UKS 1 has provided a very crude estimate that it is at a similar distance, placing VVV CL001 well beyond the bulge of the Milky Way. Observations in theK_S-band that will be acquired in the future seasons of the VVV survey would allow us to improve this CMD of Fig. 3 and define the HB as well as the MSTO, which would allow a photometric age determination. In addition, we estimate that the RR Lyrae of VVV CL001 would be within the limit of detection of our VVV Survey.

We have been unable to definitely conclude that the proximity of UKS 1 and VVV CL001 on the sky imply that they represent a binary cluster because the distance to VVV CL001 is too uncertain. However, that they are a binary remains an exciting possibility that needs to be confirmed not only by means of a more accurate distance determination, but also by measuring their respective radial velocities from spectroscopic measurements. Finally, the present results are very encouraging and we conclude that the VVV survey can potentially provide the largest and most homogeneous census of globular clusters in the survey area, out to well beyond the Galactic centre, even in regions of large extinction.

Acknowledgments

We thank the Cambridge Astronomical Survey Unit (CASU) for processing the VISTA raw data. We acknowledge support by the FONDAP Center for Astrophysics 15010003, BASAL Center for Astrophysics and Associated Technologies PFB-06, MILENIO Milky Way Millennium Nucleus P07-021-F from MIDEPLAN, FONDECYT 1090213 from CONICYT, and the European Southern Observatory. We use data products from the Two Micron All Sky Survey, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center/California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. We also thank the referee, whose comments helped to improve the paper significantly.

References

All Tables

All Figures

Fig. 1 Discovery image (left panel,JHKS colour composite) showing VVV CL001 on the left and UKS 1 on the right. The cluster UKS 1 was discovered byMalkan et al. (1980), to have a total magnitudeI < 16 mag, and was invisible in Palomar Survey plates (up to a few years ago, UKS 1 was known as “the faintest globular cluster of the Milky Way”). The right panel shows a zoom into the region centered on VVV CL001, corresponding to the white box ().

In the text

Fig. 2 J-band image of the VVV CL001 region (left) and density of all objects. The contours in the left panel refer to the intensity level of the image. In contrast, to derive the significance of the overdensity (right) only detections with a Dophot classification parameter 1 or 7 (Schechter et al. 1993) were selected. The sources were binned into 60 × 60 pixel bins (20 by 20 arcsec). We note that even with PSF photometry we are unable to resolve the inner region of VVV CL001 (or UKS 1), which certainly affects the numerical results of our analysis (density, significance of the detected over-density). Both plots show the same 272 by 272 arcsec section of theJ-band image.

In the text

Fig. 3 Discovery near-infrared colour − magnitude diagram of VVV CL001 for a 15 arcsec radius area centered on the cluster (left) compared with a field of the same area located east (middle), and a field of UKS 1 of the same size that is offset from the cluster centre by () (right). The small offset for UKS 1 was applied, since the centre of this cluster is barely resolved. The horizontal line atKs = 14.3 shows the limit of 2MASS photometry. Comparing UKS 1 and VVV CL001, we find that the RGB of UKS 1 is much more populated, hence VVV CL001 is far less massive then UKS 1.

In the text

Fig. 4 CMD of VVV CL001 (diamonds) compared with UKS 1 (dots). The horizontal-branch red clump of UKS 1 is seen atKS = 15.3,J − KS = 2.0. This diagram shows that the RGB of VVV CL001 is slightly bluer than that of UKS 1, from which one can conclude that VVV CL001 is less reddened or more metal-poor than UKS 1, and that the RGB of UKS 1 is much more populated, indicative of the low mass of VVV CL001. The legend gives the radius of the selected cluster region, whereas the arrows indicate the reddening derived for UKS 1.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.