Spectroscopy of Triangulum II

Triangulum II: a very metal-poor and dynamically hot stellar system

Nicolas F. Martin, Rodrigo A. Ibata, Michelle L. M. Collins, R. Michael Rich, Eric F. Bell, Annette M. N. Ferguson, Benjamin P. M. Laevens, Hans-Walter Rix, Scott C. Chapman, Andreas Koch nicolas.martin@astro.unistra.fr

We present a study of the recently discovered compact stellar system Triangulum II. From observations conducted with the DEIMOS spectrograph on Keck II, we obtained spectra for 13 member stars that follow the CMD features of this very faint stellar system and include two bright red giant branch stars. Tri II has a very negative radial velocity () that translates to and confirms it is a Milky Way satellite. We show that, despite the small data set, there is evidence that Tri II has complex internal kinematics. Its radial velocity dispersion increases from in the central to outwards. The velocity dispersion of the full sample is inferred to be . From the two bright RGB member stars we measure an average metallicity , placing Tri II among the most metal-poor Milky Way dwarf galaxies. In addition, the spectra of the fainter member stars exhibit differences in their line-widths that could be the indication of a metallicity dispersion in the system. All these properties paint a complex picture for Tri II, whose nature and current state are largely speculative. The inferred metallicity properties of the system however lead us to favor a scenario in which Tri II is a dwarf galaxy that is either disrupting or embedded in a stellar stream.

Subject headings:
Local Group — galaxies: individual: Tri II — galaxies: kinematics and dynamics
11affiliationtext: Observatoire astronomique de Strasbourg, Université de Strasbourg, CNRS, UMR 7550, 11 rue de l’Université, F-67000 Strasbourg, France22affiliationtext: Max-Planck-Institut für Astronomie, Königstuhl 17, D-69117 Heidelberg, Germany33affiliationtext: Astronomy Department, Yale University, New Haven, CT 0652044affiliationtext: Department of Physics, University of Surrey, Guildford, GU2 7XH, Surrey, UK55affiliationtext: Hubble Fellow66affiliationtext: Department of Physics and Astronomy, University of California, Los Angeles, PAB, 430 Portola Plaza, Los Angeles, CA 90095-1547, USA77affiliationtext: Department of Astronomy, University of Michigan, 500 Church St., Ann Arbor, MI 48109, USA88affiliationtext: Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, UK99affiliationtext: Department of Physics and Atmospheric Science, Dalhousie University, Coburg Road, Halifax, NS B3H 1A6, Canada1010affiliationtext: Landessternwarte, Zentrum für Astronomie der Universität Heidelberg, Königstuhl 12, 69117 Heidelberg, Germany

1. Introduction

A large number of faint and small stellar systems have been uncovered over the last decade thanks to wide photometric surveys. The harvest of such objects, which started with Willman 1 (Willman et al., 2005) and then Segue 1 (Belokurov et al., 2007), blossomed through systematic searches of the Sloan Digital Sky Survey (SDSS; Belokurov et al., 2009) and, more recently, of the Dark Energy Survey (DES; Bechtol et al., 2015; Drlica-Wagner et al., 2015; Koposov et al., 2015) and the Panoramic Telescope and Rapid Response System 1 (Pan-STARRS1; Laevens et al., 2015b, a). The photometric properties of many of these systems ambiguously locate them in a region of parameter space where dwarf galaxies appear to mix with globular clusters (Gilmore et al., 2007). Spectroscopic studies of their stars are therefore unavoidable to show that they are either dynamically cold and display no metallicity dispersion (e.g., Laevens 1; Kirby et al., 2015), as expected for globular clusters, or that they are dynamically hot (e.g., Segue 1; Geha et al., 2009), have a metallicity dispersion (e.g., Segue 2; Kirby et al., 2013a), and/or lie on the luminosity–metallicity relation followed by dwarf galaxies (e.g., Hydra II and Draco II; Kirby et al., 2015; Martin et al., 2015).

Although they are expected by the dozen in simulations (Tollerud et al., 2008; Bullock et al., 2010), only a handful of these faint and small stellar systems have so far been confirmed as dwarf galaxies. Any new addition to the list is particularly valuable as these objects are among the most promising for the indirect detection of the elusive dark matter particle (e.g.,  Geringer-Sameth et al., 2015). Their small baryonic component () makes them hard to find and study but, at the same time, gives powerful insight into the interplay of physical processes that drive galaxy formation at low masses and in shallow potential wells. The characterization of these systems is, however, made difficult by the potential presence of binary stars that can significantly inflate the intrinsic velocity dispersion of a system with a velocity dispersion of only a few (McConnachie & Côté, 2010). In addition, the usual assumption of dynamical equilibrium can be inappropriate for systems that are often found within of the Galactic center, further impeding their study (e.g., Willman 1; Willman et al., 2011). Finally, the difficulty to disentangle member stars from foreground contamination can sometimes further compound the analysis of these faint objects (e.g.,  Bonnivard et al., 2015).

In this paper, we report a spectroscopic study of the Triangulum II (Tri II) stellar system discovered by Laevens et al. (2015b) in Pan-STARRS1 and confirmed with deep Large Binocular Camera (LBC) photometry. Tri II is very faint (), fairly compact (), and located at from the Sun, or from the Galactic center. So far as one can infer from the photometry, it appears to contain only old and metal-poor stars.

We present the Keck II/DEIMOS data used for the analysis in section 2 of this paper, the results of the spectroscopic study in section 3, while section 4 is devoted to a discussion on our findings.

2. Observations and data

Two masks targeting Tri II potential member stars were observed during the night of September 17, 2015, with the DEep Imaging Multi-Object Spectrograph (DEIMOS) on Keck II (Faber et al., 2003) under reasonably good conditions (0.7–1.0 seeing and % humidity; PI: Rich, program ID: 2015B_U064D). The LBC photometry used by Laevens et al. (2015b) to confirm the discovery of the satellite were used to place slits on stars selected in the color-magnitude diagram (CMD) to follow the system’s main sequence turn off (MSTO), sub-giant branch (SGB), and red giant branch (RGB). The selection was purposefully tight around the sharp MSTO and loose around the RGB region that cannot be selected out of the MW foreground contamination from the photometry alone.

Each mask was integrated for 3600s, split into 3 sub-exposures for cosmic-ray removal. The spectrograph was set up with the 1200 lines/mm grating, which translates to  per pixel in the Calcium ii triplet (CaT) region we focus on. The full spectra cover the range 6600–9400. Raw frames are processed through our own pipeline, which we developed over the years to specifically reduce DEIMOS spectra. We refer the reader to Ibata et al. (2011) for an overview of the details of the processing, and the results of using the pipeline on high quality data. Briefly, the reduction method calibrates each pixel of the original spectroscopic frame in both wavelength and spatial position on the sky. In this way the data retain their original pixel binning, and one avoids introducing the correlated noise that occurs when spectra are extracted and co-added. The wavelength solution is given by a fit to arc-lamp frames taken immediately after the science frames. We also allow for a wavelength re-calibration using the Fraunhofer A band in the range 7595–7630 in order to perform small telluric corrections when the signal to noise of the spectra is sufficient. A two-dimensional sky spectrum model is built for each slitlet following a procedure inspired by the method of Kelson (2003). Finally, the radial velocity (and corresponding uncertainty) of the target stars is measured by fitting a simple Gaussian model of the Ca II triplet lines to the pixel data minus the sky spectrum. Only stars with large enough signal-to-noise ( per pixel) and velocity uncertainties lower than are kept for the analysis, leaving a total sample of 50 stars.

Velocity uncertainties measured by the pipeline are known not to fully account for low level systematics. Following Ibata et al. (2011), we add an uncertainty floor of in quadrature to the velocity uncertainties measured directly from the spectra. Finally, we measure the equivalent widths of the CaT lines and their uncertainties by independently fitting to the three lines Moffat functions shifted to the velocity of a given star (Ibata et al., 2005).

Unless specified otherwise, radial velocities reported in this paper are heliocentric velocities, corrected from the motion of the Earth around the Sun, but not corrected for the Solar motion. The known and derived properties of Tri II are summarized in Table 1 and the properties of the spectroscopic sample stars are listed below in Table 2.

RA111from Laevens et al. (2015b) (ICRS) 02:13:17.4
Dec (ICRS) +36:10:42.4
Heliocentric distance ( kpc)
Galactocentric distance ( kpc)
Global kinematics
Inner kinematics ()
Outer kinematics ()
Table 1Properties of Tri II

3. Analysis

3.1. Velocities

Figure 1.— Left: LBC CMD of stars within of Tri II’s centroid. Stars with spectra that passed our quality cuts are shown color-coded by velocity whereas stars without spectroscopic information are represented by gray dots. The 13 Tri II member stars appear as dark blue points with large negative velocities and follow the Parsec isochrone (Bressan et al., 2012) shown in blue, favored by Laevens et al. (2015b) to reproduce the CMD features of the stellar system ( and ). Right: Distribution of the LBC stars selected to follow the Tri II CMD features. The color-coding is the same as in the left-hand panel. The blue ellipse represents the region within the half-light radius of Tri II.

Stars with good quality spectra are displayed in Figure 1 over the CMD of Tri II in the left-hand panel and over the spatial distribution of possible Tri II stars in the right-hand panel. A group of stars with highly negative velocities, shown in dark-blue, is almost perfectly aligned with the favored old and metal-poor isochrone of Laevens et al. (2015b, and ). Most of these stars are MSTO or SGB stars but the sample also contains 2 RGB stars that shall prove valuable to derive the metallicity of Tri II. A large fraction of the stars with very negative velocities is also concentrated within the half-light radius of Tri II represented by the blue ellipse, even though some member stars are also located beyond and throughout the region covered by the DEIMOS masks.

Figure 2.— Top: Heliocentric velocity distribution of the spectroscopic sample. The velocity peak of Tri II stars is visible at , separated from the MW contamination with . The orange line displays the best fit to the velocity distribution of Tri II stars, convolved by the median velocity uncertainty. Bottom: Pdfs of the systemic velocity (left) and velocity dispersion (right) of the full Tri II sample (black histograms). The blue and red histograms correspond to the pdfs for the inner and outer half of the sample, respectively. The vertical lines indicate the modes of the distributions. Note the discrepant velocity dispersion pdfs for the inner and outer samples.

The velocity distribution of the sample stars shown in the top panel of Figure 2 reveals the velocity peak produced by Tri II stars. Located around , it appears completely isolated from the MW foreground contamination and confirms that all the stars shown in dark blue in Figure 1 are member stars. The velocity peak is also surprisingly wide. Fitting a Gaussian distribution to the 13 member stars corroborates this first impression as we infer a velocity dispersion of around a systemic velocity of (see the bottom panels of Figure 2 for the parameters’ probability distribution functions or pdfs). Such a value is at odds with velocity dispersion measurements usually obtained in similarly faint and compact MW systems. Recent studies consistently infer values of, at most, a few (Martin et al., 2007a; Simon & Geha, 2007; Geha et al., 2009; Willman et al., 2011; Kirby et al., 2013b, 2015; Martin et al., 2015).

Figure 3.— Velocities of Tri II member stars as a function of their distance from the system’s centroid, showing an apparent flaring of the radial velocity distribution with distance. The hashed regions correspond to the velocities within for the full sample (gray), the inner sample (blue), and the outer sample (red).

Figure 3 shows, however, that the radial velocities of member stars appear to flare up with distance from the center of the system. The central half of the sample is much more closely aligned in velocity than its outer half. Fitting the velocity distribution for the 6 member stars within yields and , whereas the outermost 7 stars yield and . While the systemic velocities of the inner and outer samples are compatible, this is hardly the case for the velocity dispersion measurements (see pdfs in the bottom panels of Figure 2). Although we cannot completely rule out the compatibility of the two measurements inferred from a small number of stars in both samples, we find that the velocity dispersions are nevertheless discrepant at the level, confirming the visual impression from Figure 3.

The large velocity dispersion of the outer sample is robust to the velocity uncertainties as it remains present, even if we increase the uncertainty floor to the unlikely value of . We also checked for the presence of a velocity gradient as a function of position that could artificially give the impression of the flaring from stars with significantly different velocities on opposite sides of the system. No such gradient was found in the data.

3.2. Metallicities

Figure 4.— Smoothed spectra of two Tri II RGB (top) and three lower-RGB/SGB member stars in the region around the CaT. The spectra are smoothed with a 3-pixel boxcar filter. The properties of the stars are listed in each panel. Note the weak lines of the two RGB stars and the varying lines widths of the three lower-RGB/SGB stars that share similar photometric properties, indicating a potential metallicity dispersion in the system.

Two of the observed member stars (stars 40 & 46 in Table 2) are RGB stars bright enough to allow for a determination of their metallicity via a measure of the equivalent widths of the CaT lines. Starkenburg et al. (2010) has calibrated the relation between the equivalent widths of the second and third Caii lines, , down to very low metallicities (). We first convert the magnitudes of the two stars to magnitudes with the Tonry et al. (2012) color equation and, using the Starkenburg et al. (2010) relation, we calculate for both RGB stars. These very low metallicity values are confirmed by the inspection of the spectra (top two panels of Figure 4) that both exhibit very weak CaT lines. Assuming a Gaussian metallicity distribution function, we infer a mean metallicity for Tri II, which is therefore among the most metal-poor MW satellites.

It is harder to provide a definite conclusion on the presence or absence of a metallicity dispersion in Tri II since two stars alone cannot rule out the presence of a dispersion, even if they are measured to have the same metallicity. Moreover, the other observed member stars are located far below the horizontal branch of the system and the horizontal branch marks the faint limit to which the Starkenburg et al. (2010) relation has been calibrated. Finally, directly extracting a measure of Fe line strengths from DEIMOS spectra with is fraught with peril.

We note however that the three Tri II stars above the SGB that have similar colors (), magnitudes (), and signal-to-noise (), have inconsistent equivalent widths with , , and  from the brighter to the fainter star (stars # 20, 21, and 31 from Table 2, respectively). Moreover, none of these stars show strong Nai doublet lines (8183 and 8192) that would indicate that they are foreground contaminants. These differences are directly visible on the spectra (lower three panels of Figure 4) and could be interpreted as evidence of a metallicity dispersion in Tri II. Blindly applying the Starkenburg et al. (2010) relation for these stars that are fainter than the magnitude range over which it was calibrated222However, see Leaman et al. (2013) and their study of the metal-poor globular cluster M15 for which the calibration is shown to hold magnitudes below this system’s horizontal branch. yields tentative metallicities of , , and , respectively. From the 5 brightest confirmed Tri II members, we infer and a large metallicity dispersion of dex. We nevertheless stress that direct measurements are needed from higher spectra to bolster this marginal conclusion.

4. Discussion

We obtained spectra for 13 member stars in the very faint MW satellite Tri II. These stars follow the CMD features of the stellar system and include mainly MSTO and SGB stars, as well as two bright RGB stars. With , Tri II has a very negative radial velocity that translates to . We have further shown that, as far as we can tell from only 13 member stars, the internal kinematics of Tri II appear complex with evidence for a radial velocity dispersion increase from in the central to outwards. The velocity dispersion of the full sample is inferred to be . Finally, the two bright RGB member stars are both measured to have and point to Tri II being among the most metal-poor MW satellites. The spectra of fainter member stars exhibit differences in their line-widths that could be due to a metallicity dispersion in the system.

At the distance of Tri II ( from the Galactic center), such a fast infalling velocity is not unexpected for a satellite bound to the MW. It does, however, rule out any association with the numerous stellar structures found nearby in the Milky Way halo. TriAnd (Rocha-Pinto et al., 2004), TriAnd 2 (Martin et al., 2007b), or the PAndAS MW stream (Martin et al., 2014) all have positive (Deason et al., 2014). Tri II is also unrelated to the Segue 2 satellite that is located only away but has a very different velocity (; Belokurov et al., 2009).

But what is the nature of Tri II? Taken at face value, the large global velocity dispersion, the very low-metallicity, and the potential metallicity dispersion seem to point towards Tri II being a dwarf galaxy rather than a globular cluster. However, the complex kinematics of the system question the assumption of dynamical equilibrium that is required to translate a large velocity dispersion into a large mass and mass-to-light ratio.

Figure 5.— Distribution of MW satellites in the mean metallicity versus luminosity plane. Large black points correspond to MW dwarf galaxies, as listed in Kirby et al. (2013b), supplemented by Norris et al. (2010), Kirby et al. (2015), and Simon et al. (2015). MW globular clusters are shown as small black dots (Harris, 1996). The red squares corresponds to the Tri II measurements, with the filled square representing the inference from the two robust individual stellar metallicities (stars 40 and 46) while the hollow circle corresponds to the inference from the 5 stars shown in Figure 4 (the two Tri II points have been slightly offset from each other along the luminosity axis so their error bars do not overlap). Both measurements compare well with those of other MW dwarf galaxies.

Can discriminate between globular cluster and dwarf galaxy? Irrespective of its velocity dispersion, Tri II is among the most metal-poor systems known. No globular cluster is known with a metallicity below (Harris, 1996) and only the Segue 1, Bootes II, and Reticulum II dwarf galaxies, with whom Tri II shares many similar properties (total luminosity, size, distance), are as metal-poor with (Norris et al., 2010) or (Simon et al., 2011) for Segue 1, (Koch & Rich, 2014) for Bootes II, and (Simon et al., 2015) or (Walker et al., 2015) for Reticulum II. Figure 5 shows that Tri II is in agreement with the dwarf galaxy metallicity–luminosity relation of Kirby et al. (2013b), even if we include the 3 stars with tentative metallicity measurements for the inference of the mean metallicity (hollow red square). By analogy, the metallicity of the system therefore appears to favor the dwarf galaxy hypothesis, which would be bolstered further by the marginal evidence of a metallicity dispersion.

Figure 6.— Distribution of MW dwarf galaxies in the plane, compared to the Local Group dwarf galaxy mass profiles. The black dashed and dot-dashed lines correspond to the favored Local Group mass profiles of Collins et al. (2014) for a cored or NFW model, respectively, and the gray bands represent the model dispersions determined by these authors. Black points correspond to Milky Way dwarf galaxies, as listed in McConnachie (2012), Kirby et al. (2015), Martin et al. (2015), and Simon et al. (2015). The Tri II data point is shown as a hollow square for the global kinematics determined in this paper, or as a filled square when using the kinematics of the inner sample. The former is a strong outlier whereas the latter fits well with other MW dwarf galaxies. The two Tri II points have been slightly offset from each other along the axis so their error bars do not overlap.

What is the dynamical mass of Tri II? It is hard to tell if one takes the increase in the velocity dispersion with radius as a sign that the system is out of equilibrium. On the other hand, if one assumes that the change in is an (unlikely ) statistical fluctuation and that the global velocity dispersion measured at is representative of the true properties of Tri II, one can easily note that it is a significant outlier among other dwarf galaxies of this size, as displayed in Figure 6. Equation (1) of Wolf et al. (2010), relates the mass within the 3-dimensional half-light radius, , to the half-light radius and velocity dispersion of a bound system in equilibrium and, in the case of Tri II, yields and in Solar units. It would mean that Tri II is almost an order of magnitude more massive than Segue 1 or Reticulum II and, by a wide margin, the most dark matter system known in the universe. Such a large mass seems very improbable.

Could MW contaminants pollute the velocity peak? This also appears unlikely as the inflated velocity dispersion in the outskirts of Tri II is not driven by any single outlier (see Figure 3). The very negative systemic velocity of Tri II also means it is improbable for the spectroscopic data set to contain more than a single contaminating halo star, if any.

Is Tri II disrupting and/or embedded in a stellar stream? After ruling out that Tri II is in equilibrium or contaminated by MW halo stars, the most likely hypothesis is that the observed increase in the velocity dispersion with radius is genuine. The large velocity dispersion beyond would then be produced by stars that are not bound to the body of the satellite. This seems at odds with the current measure of the half-light radius of Tri II (; Laevens et al. 2015b), but this measure could be systematically biased by the comparatively small LBC field of view (Muñoz et al., 2012). In fact, the spatial distribution of possible satellite member stars exhibits what could be a more compact central core within surrounded by a more diffuse component (see the right-hand panel of Figure 1 from Laevens et al. 2015b). Upcoming wider and deeper photometric data will allow us to robustly investigate this morphology.

If this core really is the true extent of the main body of Tri II, the velocity dispersion we measure from the inner half of the spectroscopic sample () would be more representative of the satellite’s intrinsic properties and would yield a consistent picture with the very faint dwarf galaxies Segue 1 (Simon et al., 2011), Reticulum II (Simon et al., 2015; Walker et al., 2015), or Draco II (Martin et al., 2015), as can be seen in Figure 6. The only conundrum would then be whether Tri II is disrupted by tidal interactions with the Milky Way, or if it is in equilibrium but embedded in a stellar stream. Such a stellar stream could, for instance, be produced by a potentially more massive dwarf galaxy it would have been a satellite of in the past (e.g.,  Wheeler et al., 2015). The large negative velocity of the system however likely rules out a system that has just now been tidally disrupted after a pericentric passage. The fact that the star with the most discrepant spectrum among those shown in Figure 4 (star 31) is also the only one beyond could point towards the latter hypothesis but this is hardly conclusive.

At the moment, the puzzling properties of Tri II mean that its nature and current state are largely speculative. We favor the scenario in which Tri II is a dwarf galaxy that is either disrupting or embedded in a stellar stream but cannot completely rule out that it could be a disrupting globular cluster. Whatever the true nature of the satellite, it exhibits unexpected properties that make it very exciting and call for more observations to understand its complexity.

# RA Dec R333Distance from Tri II’s centroid () S/N Member? EW
(ICRS) (ICRS) () () () (per pixel) ()
1 33.1946678 36.1328316 6.8 19.955 19.347 2.8 23.2 N
2 33.2111244 36.1401672 5.9 20.006 19.410 2.6 13.2 N
3 33.4113731 36.2505836 6.1 20.026 19.375 2.6 15.5 N
4 33.3050842 36.1931953 1.2 20.092 19.578 2.9 13.4 N
5 33.3774185 36.2628899 5.7 20.344 19.702 2.8 12.0 N
6 33.3839149 36.2578049 5.6 20.658 20.168 3.8 7.3 N
7 33.3341675 36.2451935 4.0 20.799 20.253 3.5 7.1 N
8 33.2591667 36.2090836 3.6 21.165 20.737 7.7 4.9 Y
9 33.3639183 36.2251930 3.4 21.499 21.158 5.1 3.8 Y
10 33.4349174 36.3003044 9.1 20.180 19.403 4.1 7.0 N
11 33.3940010 36.2913055 7.6 21.106 20.682 4.6 4.8 N
11 33.3940010 36.2913055 7.6 21.101 20.714 4.6 4.8 N
12 33.2499161 36.1958885 3.7 18.036 17.258 2.3 48.1 N
13 33.2220421 36.1532211 5.1 19.294 18.586 2.4 26.5 N
14 33.2612076 36.1757774 3.0 19.427 18.863 2.5 19.4 N
15 33.3087502 36.2062492 1.8 17.416 16.429 2.3 54.4 N
16 33.3540840 36.2714462 5.8 17.152 16.330 2.3 67.3 N
17 33.3997078 36.2773628 7.0 18.224 17.460 2.3 45.1 N
18 33.4496231 36.2708626 8.3 19.047 18.278 2.6 17.8 N
19 33.3174171 36.1878624 0.6 19.387 18.731 2.4 22.2 N
20 33.3305016 36.1926117 0.9 20.565 19.907 2.9 11.7 Y 444These stars are fainter that the magnitude range over which the Starkenburg et al. (2010) relation was calibrated; their measurements should therefore be taken with cautious.
21 33.3165016 36.1710815 0.5 20.667 20.059 3.1 10.8 Y
22 33.3029175 36.1470566 2.1 20.957 20.458 3.8 7.6 Y
23 33.3359184 36.1629181 1.1 21.109 20.562 3.6 7.1 Y
24 33.3416672 36.1738892 1.0 21.552 21.222 4.9 5.3 Y
25 33.3214149 36.1205826 3.5 21.568 21.196 5.6 4.5 Y
26 33.3535004 36.1727219 1.5 21.799 21.398 8.2 3.5 N
27 33.3389587 36.1414452 2.4 21.814 21.458 6.6 4.2 Y
28 33.3247490 36.1699982 0.5 21.991 21.616 6.9 3.5 N
29 33.3789597 36.1988907 3.0 22.326 21.877 7.8 3.1 Y
30 33.4619598 36.2307205 7.4 20.690 20.180 2.9 11.2 N
31 33.4694176 36.2234154 7.6 20.845 20.223 3.1 9.8 Y
32 33.4769173 36.2040291 7.6 21.132 20.700 3.5 6.8 N
33 33.2707901 36.0966949 5.5 17.659 16.706 2.4 53.3 N
34 33.2869987 36.0956383 5.3 18.630 18.031 2.4 31.0 N
35 33.2285843 36.1292763 5.4 17.306 16.496 2.3 60.3 N
36 33.2309570 36.1342239 5.2 19.463 18.691 2.4 21.5 N
37 33.2211685 36.1413040 5.4 17.928 17.216 2.3 46.5 N
38 33.4219170 36.1992760 5.0 17.195 16.231 2.3 78.9 N
39 33.4609985 36.1992493 6.8 16.906 16.181 2.4 57.1 N
40 33.3189583 36.1793900 0.2 17.585 16.692 2.3 73.3 Y
41 33.3302078 36.1485825 1.8 17.699 16.858 2.3 54.9 N
42 33.3511658 36.1708603 1.5 17.561 16.772 2.3 57.2 N
43 33.4364586 36.2213593 6.1 18.337 17.426 2.3 39.8 N
44 33.3328743 36.2034149 1.6 18.925 18.122 4.8 30.0 N
45 33.3890419 36.1674995 3.3 19.115 18.539 2.4 25.2 N
46 33.3397484 36.1659431 1.1 19.286 18.540 2.4 26.4 Y
47 33.3222084 36.1384430 2.4 19.019 18.452 2.4 25.7 N
48 33.3827095 36.2173882 3.7 17.149 16.205 2.3 70.2 N
49 33.4824982 36.2344437 8.4 17.344 16.462 2.3 61.6 N
50 33.4132080 36.1776123 4.4 19.096 18.472 2.4 26.9 N
Table 2Properties of observed stars meeting the quality criteria
B.P.M.L. acknowledges funding through a 2012 Strasbourg IDEX (Initiative d’Excellence) grant, awarded by the University of Strasbourg. N.F.M. and B.P.M.L. gratefully acknowledge the CNRS for support through PICS project PICS06183. H.-W.R. acknowledges support by the DFG through the SFB 881 (A3). M.L.M.C. acknowledges financial support from the European Research Council (ERC-StG-335936) and from NASA through Hubble Fellowship grant #51337 awarded by the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Inc., for NASA, under contract NAS 5-26555. The data presented herein were obtained at the W.M. Keck Observatory, which is operated as a scientific partnership among the California Institute of Technology, the University of California and the National Aeronautics and Space Administration. The Observatory was made possible by the generous financial support of the W.M. Keck Foundation. The authors wish to recognize and acknowledge the very significant cultural role and reverence that the summit of Mauna Kea has always had within the indigenous Hawaiian community. We are most fortunate to have the opportunity to conduct observations from this mountain.


  • Bechtol et al. (2015) Bechtol, K., Drlica-Wagner, A., Balbinot, E., Pieres, A., Simon, J. D., Yanny, B., Santiago, B., Wechsler, R. H., Frieman, J., Walker, A. R., Williams, P., Rozo, E., Rykoff, E. S., Queiroz, A., Luque, E., Benoit-Levy, A., Bernstein, R. A., Tucker, D., Sevilla, I., Gruendl, R. A., da Costa, L. N., Fausti Neto, A., Maia, M. A. G., Abbott, T., Allam, S., Armstrong, R., Bauer, A. H., Bernstein, G. M., Bertin, E., Brooks, D., Buckley-Geer, E., Burke, D. L., Carnero Rosell, A., Castander, F. J., D’Andrea, C. B., DePoy, D. L., Desai, S., Diehl, H. T., Eifler, T. F., Estrada, J., Evrard, A. E., Fernandez, E., Finley, D. A., Flaugher, B., Gaztanaga, E., Gerdes, D., Girardi, L., Gladders, M., Gruen, D., Gutierrez, G., Hao, J., Honscheid, K., Jain, B., James, D., Kent, S., Kron, R., Kuehn, K., Kuropatkin, N., Lahav, O., Li, T. S., Lin, H., Makler, M., March, M., Marshall, J., Martini, P., Merritt, K. W., Miller, C., Miquel, R., Mohr, J., Neilsen, E., Nichol, R., Nord, B., Ogando, R., Peoples, J., Petravick, D., Plazas, A. A., Romer, A. K., Roodman, A., Sako, M., Sanchez, E., Scarpine, V., Schubnell, M., Smith, R. C., Soares-Santos, M., Sobreira, F., Suchyta, E., Swanson, M. E. C., Tarle, G., Thaler, J., Thomas, D., Wester, W., & Zuntz, J. 2015, ArXiv e-prints, arXiv:1503.02584
  • Belokurov et al. (2009) Belokurov, V., Walker, M. G., Evans, N. W., Gilmore, G., Irwin, M. J., Mateo, M., Mayer, L., Olszewski, E., Bechtold, J., & Pickering, T. 2009, MNRAS, 397, 1748
  • Belokurov et al. (2007) Belokurov, V., Zucker, D. B., Evans, N. W., Kleyna, J. T., Koposov, S., Hodgkin, S. T., Irwin, M. J., Gilmore, G., Wilkinson, M. I., Fellhauer, M., Bramich, D. M., Hewett, P. C., Vidrih, S., De Jong, J. T. A., Smith, J. A., Rix, H.-W., Bell, E. F., Wyse, R. F. G., Newberg, H. J., Mayeur, P. A., Yanny, B., Rockosi, C. M., Gnedin, O. Y., Schneider, D. P., Beers, T. C., Barentine, J. C., Brewington, H., Brinkmann, J., Harvanek, M., Kleinman, S. J., Krzesinski, J., Long, D., Nitta, A., & Snedden, S. A. 2007, ApJ, 654, 897
  • Bonnivard et al. (2015) Bonnivard, V., Combet, C., Daniel, M., Funk, S., Geringer-Sameth, A., Hinton, J. A., Maurin, D., Read, J. I., Sarkar, S., Walker, M. G., & Wilkinson, M. I. 2015, MNRAS, 453, 849
  • Bressan et al. (2012) Bressan, A., Marigo, P., Girardi, L., Salasnich, B., Dal Cero, C., Rubele, S., & Nanni, A. 2012, MNRAS, 427, 127
  • Bullock et al. (2010) Bullock, J. S., Stewart, K. R., Kaplinghat, M., Tollerud, E. J., & Wolf, J. 2010, ApJ, 717, 1043
  • Collins et al. (2014) Collins, M. L. M., Chapman, S. C., Rich, R. M., Ibata, R. A., Martin, N. F., Irwin, M. J., Bate, N. F., Lewis, G. F., Peñarrubia, J., Arimoto, N., Casey, C. M., Ferguson, A. M. N., Koch, A., McConnachie, A. W., & Tanvir, N. 2014, ApJ, 783, 7
  • Deason et al. (2014) Deason, A. J., Belokurov, V., Hamren, K. M., Koposov, S. E., Gilbert, K. M., Beaton, R. L., Dorman, C. E., Guhathakurta, P., Majewski, S. R., & Cunningham, E. C. 2014, MNRAS, 444, 3975
  • Drlica-Wagner et al. (2015) Drlica-Wagner, A., Bechtol, K., Rykoff, E. S., Luque, E., Queiroz, A., Mao, Y.-Y., Wechsler, R. H., Simon, J. D., Santiago, B., Yanny, B., Balbinot, E., Dodelson, S., Fausti Neto, A., James, D. J., Li, T. S., Maia, M. A. G., Marshall, J. L., Pieres, A., Stringer, K., Walker, A. R., Abbott, T. M. C., Abdalla, F. B., Allam, S., Benoit-Levy, A., Bernstein, G. M., Bertin, E., Brooks, D., Buckley-Geer, E., Burke, D. L., Carnero Rosell, A., Carrasco Kind, M., Carretero, J., Crocce, M., da Costa, L. N., Desai, S., Diehl, H. T., Dietrich, J. P., Doel, P., Eifler, T. F., Evrard, A. E., Finley, D. A., Fosalba, P., Frieman, J., Gaztanaga, E., Gerdes, D. W., Gruen, D., Gruendl, R. A., Gutierrez, G., Honscheid, K., Kuehn, K., Kuropatkin, N., Lahav, O., Martini, P., Miquel, R., Nord, B., Ogando, R., Plazas, A. A., Reil, K., Roodman, A., Sako, M., Sanchez, E., Scarpine, V., Schubnell, M., Sevilla-Noarbe, I., Smith, R. C., Soares-Santos, M., Sobreira, F., Suchyta, E., Swanson, M. E. C., Tarle, G., Tucker, D., Vikram, V., Wester, W., Zhang, Y., & Zuntz, J. 2015, ArXiv e-prints
  • Faber et al. (2003) Faber, S. M., Phillips, A. C., Kibrick, R. I., Alcott, B., Allen, S. L., Burrous, J., Cantrall, T., Clarke, D., Coil, A. L., Cowley, D. J., Davis, M., Deich, W. T. S., Dietsch, K., Gilmore, D. K., Harper, C. A., Hilyard, D. F., Lewis, J. P., McVeigh, M., Newman, J., Osborne, J., Schiavon, R., Stover, R. J., Tucker, D., Wallace, V., Wei, M., Wirth, G., & Wright, C. A. 2003, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 4841, Instrument Design and Performance for Optical/Infrared Ground-based Telescopes, ed. M. Iye & A. F. M. Moorwood, 1657–1669
  • Geha et al. (2009) Geha, M., Willman, B., Simon, J. D., Strigari, L. E., Kirby, E. N., Law, D. R., & Strader, J. 2009, ApJ, 692, 1464
  • Geringer-Sameth et al. (2015) Geringer-Sameth, A., Koushiappas, S. M., & Walker, M. G. 2015, Phys. Rev. D, 91, 083535
  • Gilmore et al. (2007) Gilmore, G., Wilkinson, M. I., Wyse, R. F. G., Kleyna, J. T., Koch, A., Evans, N. W., & Grebel, E. K. 2007, ApJ, 663, 948
  • Harris (1996) Harris, W. E. 1996, AJ, 112, 1487
  • Ibata et al. (2005) Ibata, R., Chapman, S., Ferguson, A. M. N., Lewis, G., Irwin, M., & Tanvir, N. 2005, ApJ, 634, 287
  • Ibata et al. (2011) Ibata, R., Sollima, A., Nipoti, C., Bellazzini, M., Chapman, S. C., & Dalessandro, E. 2011, ApJ, 738, 186
  • Kelson (2003) Kelson, D. D. 2003, PASP, 115, 688
  • Kirby et al. (2013a) Kirby, E. N., Boylan-Kolchin, M., Cohen, J. G., Geha, M., Bullock, J. S., & Kaplinghat, M. 2013a, ApJ, 770, 16
  • Kirby et al. (2013b) Kirby, E. N., Cohen, J. G., Guhathakurta, P., Cheng, L., Bullock, J. S., & Gallazzi, A. 2013b, ApJ, 779, 102
  • Kirby et al. (2015) Kirby, E. N., Simon, J. D., & Cohen, J. G. 2015, ApJ, 810, 56
  • Koch & Rich (2014) Koch, A., & Rich, R. M. 2014, ApJ, 794, 89
  • Koposov et al. (2015) Koposov, S. E., Belokurov, V., Torrealba, G., & Wyn Evans, N. 2015, ArXiv e-prints, arXiv:1503.02079
  • Laevens et al. (2015a) Laevens, B. P. M., Martin, N. F., Bernard, E. J., Schlafly, E. F., Sesar, B., Rix, H.-W., Bell, E. F., Ferguson, A. M. N., Slater, C. T., Sweeney, W. E., Wyse, R. F. G., Huxor, A. P., Burgett, W. S., Chambers, K. C., Draper, P. W., Magnier, E. A., Metcalfe, N., Tonry, J. L., Wainscoat, R. J., & Waters, C. 2015a, ArXiv e-prints
  • Laevens et al. (2015b) Laevens, B. P. M., Martin, N. F., Ibata, R. A., Rix, H.-W., Bernard, E. J., Bell, E. F., Sesar, B., Ferguson, A. M. N., Schlafly, E. F., Slater, C. T., Burgett, W. S., Chambers, K. C., Flewelling, H., Hodapp, K. A., Kaiser, N., Kudritzki, R.-P., Lupton, R. H., Magnier, E. A., Metcalfe, N., Morgan, J. S., Price, P. A., Tonry, J. L., Wainscoat, R. J., & Waters, C. 2015b, ApJ, 802, L18
  • Leaman et al. (2013) Leaman, R., Venn, K. A., Brooks, A. M., Battaglia, G., Cole, A. A., Ibata, R. A., Irwin, M. J., McConnachie, A. W., Mendel, J. T., Starkenburg, E., & Tolstoy, E. 2013, ApJ, 767, 131
  • Martin et al. (2015) Martin, N. F., Geha, M., Ibata, R. A., Collins, M. L. M., Laevens, B. P. M., Bell, E. F., Rix, H.-W., Ferguson, A. M. N., Chambers, K. C., Wainscoat, R. J., & Waters, C. 2015, ArXiv e-prints
  • Martin et al. (2007a) Martin, N. F., Ibata, R. A., Chapman, S. C., Irwin, M., & Lewis, G. F. 2007a, MNRAS, 380, 281
  • Martin et al. (2007b) Martin, N. F., Ibata, R. A., & Irwin, M. 2007b, ApJ, 668, L123
  • Martin et al. (2014) Martin, N. F., Ibata, R. A., Rich, R. M., Collins, M. L. M., Fardal, M. A., Irwin, M. J., Lewis, G. F., McConnachie, A. W., Babul, A., Bate, N. F., Chapman, S. C., Conn, A. R., Crnojević, D., Ferguson, A. M. N., Mackey, A. D., Navarro, J. F., Peñarrubia, J., Tanvir, N. T., & Valls-Gabaud, D. 2014, ApJ, 787, 19
  • McConnachie (2012) McConnachie, A. W. 2012, AJ, 144, 4
  • McConnachie & Côté (2010) McConnachie, A. W., & Côté, P. 2010, ApJ, 722, L209
  • Muñoz et al. (2012) Muñoz, R. R., Padmanabhan, N., & Geha, M. 2012, ApJ, 745, 127
  • Norris et al. (2010) Norris, J. E., Wyse, R. F. G., Gilmore, G., Yong, D., Frebel, A., Wilkinson, M. I., Belokurov, V., & Zucker, D. B. 2010, ApJ, 723, 1632
  • Rocha-Pinto et al. (2004) Rocha-Pinto, H. J., Majewski, S. R., Skrutskie, M. F., Crane, J. D., & Patterson, R. J. 2004, ApJ, 615, 732
  • Simon et al. (2015) Simon, J. D., Drlica-Wagner, A., Li, T. S., Nord, B., Geha, M., Bechtol, K., Balbinot, E., Buckley-Geer, E., Lin, H., Marshall, J., Santiago, B., Strigari, L., Wang, M., Wechsler, R. H., Yanny, B., Abbott, T., Bauer, A. H., Bernstein, G. M., Bertin, E., Brooks, D., Burke, D. L., Capozzi, D., Carnero Rosell, A., Carrasco Kind, M., D’Andrea, C. B., da Costa, L. N., DePoy, D. L., Desai, S., Diehl, H. T., Dodelson, S., Cunha, C. E., Estrada, J., Evrard, A. E., Fausti Neto, A., Fernandez, E., Finley, D. A., Flaugher, B., Frieman, J., Gaztanaga, E., Gerdes, D., Gruen, D., Gruendl, R. A., Honscheid, K., James, D., Kent, S., Kuehn, K., Kuropatkin, N., Lahav, O., Maia, M. A. G., March, M., Martini, P., Miller, C. J., Miquel, R., Ogando, R., Romer, A. K., Roodman, A., Rykoff, E. S., Sako, M., Sanchez, E., Schubnell, M., Sevilla, I., Smith, R. C., Soares-Santos, M., Sobreira, F., Suchyta, E., Swanson, M. E. C., Tarle, G., Thaler, J., Tucker, D., Vikram, V., Walker, A. R., Wester, W., & The DES Collaboration. 2015, ApJ, 808, 95
  • Simon & Geha (2007) Simon, J. D., & Geha, M. 2007, ApJ, 670, 313
  • Simon et al. (2011) Simon, J. D., Geha, M., Minor, Q. E., Martinez, G. D., Kirby, E. N., Bullock, J. S., Kaplinghat, M., Strigari, L. E., Willman, B., Choi, P. I., Tollerud, E. J., & Wolf, J. 2011, ApJ, 733, 46
  • Starkenburg et al. (2010) Starkenburg, E., Hill, V., Tolstoy, E., González Hernández, J. I., Irwin, M., Helmi, A., Battaglia, G., Jablonka, P., Tafelmeyer, M., Shetrone, M., Venn, K., & de Boer, T. 2010, A&A, 513, A34
  • Tollerud et al. (2008) Tollerud, E. J., Bullock, J. S., Strigari, L. E., & Willman, B. 2008, ApJ, 688, 277
  • Tonry et al. (2012) Tonry, J. L., Stubbs, C. W., Lykke, K. R., Doherty, P., Shivvers, I. S., Burgett, W. S., Chambers, K. C., Hodapp, K. W., Kaiser, N., Kudritzki, R.-P., Magnier, E. A., Morgan, J. S., Price, P. A., & Wainscoat, R. J. 2012, ApJ, 750, 99
  • Walker et al. (2015) Walker, M. G., Mateo, M., Olszewski, E. W., Bailey, III, J. I., Koposov, S. E., Belokurov, V., & Evans, N. W. 2015, ApJ, 808, 108
  • Wheeler et al. (2015) Wheeler, C., Oñorbe, J., Bullock, J. S., Boylan-Kolchin, M., Elbert, O. D., Garrison-Kimmel, S., Hopkins, P. F., & Kereš, D. 2015, MNRAS, 453, 1305
  • Willman et al. (2005) Willman, B., Blanton, M. R., West, A. A., Dalcanton, J. J., Hogg, D. W., Schneider, D. P., Wherry, N., Yanny, B., & Brinkmann, J. 2005, AJ, 129, 2692
  • Willman et al. (2011) Willman, B., Geha, M., Strader, J., Strigari, L. E., Simon, J. D., Kirby, E., Ho, N., & Warres, A. 2011, AJ, 142, 128
  • Wolf et al. (2010) Wolf, J., Martinez, G. D., Bullock, J. S., Kaplinghat, M., Geha, M., Muñoz, R. R., Simon, J. D., & Avedo, F. F. 2010, MNRAS, 406, 1220
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description