1:
2: \documentclass[12pt,preprint]{aastex}
3:
4: \shorttitle{Dynamics and Metallicity of Bo\"{o}tes}
5: \shortauthors{Mu\~noz et al.}
6:
7: \begin{document}
8:
9: \title{Exploring Halo Substructure with Giant Stars: The Dynamics and Metallicity
10: of the Dwarf Spheroidal in Bo\"{o}tes}
11:
12: \author{
13: Ricardo R. Mu\~noz\altaffilmark{1},
14: Jeffrey L. Carlin\altaffilmark{1},
15: Peter M. Frinchaboy\altaffilmark{1},
16: David L. Nidever\altaffilmark{1},
17: Steven R. Majewski\altaffilmark{1} \&
18: Richard J. Patterson\altaffilmark{1}}
19:
20:
21: \altaffiltext{1}{Dept. of Astronomy, University of Virginia, P. O. Box
22: 400325, Charlottesville, VA 22904-4325 (rrm8f, jc4qn, pmf8b, dln5q, srm4n,
23: ricky@virginia.edu)}
24:
25: \begin{abstract}
26:
27: We report the results of a spectroscopic study of the Bo\"{o}tes (Boo)
28: dwarf spheroidal (dSph) galaxy carried out with the WIYN telescope and
29: the Hydra multifiber spectrograph. Radial velocities have been
30: measured for 58 Boo candidate stars selected to have magnitudes and
31: colors consistent with its red and asymptotic giant branches. Within
32: the 13\arcmin~half-light radius, seven members of Boo yield a systemic
33: velocity of $V_{r}=95.6\pm3.4$ km s$^{-1}$ and a velocity
34: dispersion of $\sigma_{\rm o}=6.6\pm2.3$ km s$^{-1}$. This implies a
35: mass on the order of $1\times10^7$ M$_{\sun}$, similar to the inferred
36: masses of other Galactic dSphs. Adopting a total Boo luminosity of
37: $L=1.8\times10^4$ L$_{\sun}$ to $8.6\times10^4$ L$_{\sun}$ implies
38: $M/L\sim610$ to 130, making Boo, the most distorted known Milky Way
39: dwarf galaxy, potentially also the darkest. From the
40: spectra of Boo member stars we estimate its metallicity to be [Fe/H]
41: $\sim-2.5$, which would make it the most metal poor dSph known to date.
42:
43: \end{abstract}
44:
45:
46: \keywords{ galaxies: individual (Bo\"{o}tes dwarf spheroidal) -- galaxies: kinematics and dynamics -- Local Group}
47:
48: \section{Introduction}
49:
50: Since the early work of \citet{Aaronson1983} on the Draco system, dwarf
51: spheroidal (dSph) galaxies in the Local Group have been suspected to
52: be heavily dark matter (DM) dominated. Analyses of dSph internal dynamics
53: under the assumption of virial equilibrium suggest mass contents far
54: exceeding those inferred from their luminosities, with central
55: mass-to-light ratios ($M/L$) ranging from a few to about a hundred (in
56: solar units). Such $M/L$ imply that dSphs have the largest
57: DM fraction of all galaxy types in the
58: universe. Yet, despite the wide range of inferred $M/L$ for dSphs,
59: their central velocity dispersions ($\sigma_{\rm o}$) and half-light
60: radii ($r_{\rm h}$) seem to be remarkably similar, $\sim7-10$ km
61: s$^{-1}$ and $\sim200$ pc respectively. Using these values, typical total
62: dSph masses of a few times $10^7$ M$_{\sun}$ are inferred (e.g.,
63: \citealt{Mateo1998})\footnote{Recent kinematical studies of dSphs have
64: attempted to derive their mass content by modeling the shape of the
65: velocity dispersion profile (e.g., \citealt{Kleyna2002};
66: \citealt{Walker2006a,Walker2006b}), but these calculations do not take
67: into account likely tidal effects in the outskirts of dSphs, as pointed
68: out by \citet{Lokas2005,M05,M06,Westfall2006,Sohn2006}.}. This mass
69: conspiracy seemingly extends to even the very low end of the galaxy
70: luminosity scale. An early kinematical survey of the dSph in Ursa
71: Major (UMa; \citealt{Kleyna2005}), the faintest and most diffuse of
72: the known Milky Way (MW) dwarf satellites (\citealt[][hereafter {W05}]{W05}),
73: found $\sigma_{\rm o}$ = 9.3$^{+11.7}_{-1.2}$ km s$^{-1}$, which,
74: coupled with UMa's $r_{\rm h}$ of 250 pc, also results in a mass of
75: $\sim10^7$ M$_{\sun}$.
76:
77: Recently, a new Galactic dSph candidate has been found in the
78: constellation of Bo\"{o}tes (\citealt[][hereafter {B06b}]{B06b})
79: during a search for halo substructure using Sloan
80: Digital Sky Survey (SDSS; \citealt{Abazajian2005}) data. Along with UMa,
81: Bo\"{o}tes (Boo) is one of the faintest MW satellites found,
82: having (at a 60 kpc distance) an absolute magnitude of $M_{V}=-5.8$
83: (but maybe brighter; see \S 3.2). In addition, the Boo dSph exhibits
84: the most irregular density contours of any Galactic dSph
85: (\citeauthor*{B06b}), which suggests that the satellite may be undergoing
86: tidal disruption.
87:
88: In this Letter, we present the results of a spectroscopic survey of
89: Boo candidate stars.
90: We have identified its kinematical signature, which allows us to
91: estimate its systemic velocity, mass and central density, as well as
92: its $M/L$. We also use the spectral data to make a rough measurement
93: of the metallicity of Boo giant stars.
94:
95: \section{Spectroscopic Observations and Data Reduction}
96:
97: Though \citeauthor*{B06b} detected Boo as an overdensity in SDSS Data
98: Release 5 (DR5), that dataset was
99: proprietary at the time we selected observing targets. However, Boo
100: appears near the northern limit of the SDSS Data Release 4 (DR4,
101: \citealt{sdss_dr4}) as well. Because the core of the Boo dSph is well
102: within the DR4 coverage area, we select Boo candidate stars in the
103: core and southward of declination $\sim$14.8\arcdeg~(Figure 1).
104:
105:
106: Spectroscopic follow-up of Boo red giant branch (RGB) and asymptotic giant branch
107: targets with $g$ $<$ 19.0, as shown in Figure 2,
108: was carried out with the WIYN 3.5-m
109: Hydra multi-fiber spectrograph on UT 2006 May 6--7.
110: Targets are distributed
111: throughout the $\sim$1-degree WIYN+Hydra field of view.
112: The 600@10.1
113: grating was used with the red fiber cable to yield a wavelength
114: coverage $\lambda$ = 4500--7200 \AA~with a spectral resolution of 2.80
115: \AA~per resolution element.
116: To achieve adequate $S/N$ for velocity determination, we
117: observed one Hydra configuration containing 72 Boo candidate stars for
118: $4 \times 30$ min. We also observed 9 radial velocity (RV) standards
119: covering spectral types F through M, each through multiple
120: fibers, yielding a total of 64 RV cross-correlation template spectra.
121:
122: Preliminary processing used the {\it IRAF} CCDRED package, with
123: spectral reduction following the standard DOHYDRA routine. RVs were
124: derived using the {\it IRAF} FXCOR task, with each
125: spectrum cross-correlated against all RV standard spectra. The regions
126: around the H$\alpha$, Mg triplet, and H$\beta$ lines were used for
127: cross-correlation. In the end, we measured reliable RVs for 58 of the
128: 72 observed Boo candidates.
129: RV standards (observed multiple times) showed deviations
130: from the published IAU values for $V_{r}$ of less than 1 km s$^{-1}$ on average.
131:
132: RV uncertainties were determined using the \citet{Vogt1995} method,
133: based on analysis of repeatedly observed standard-star
134: spectra. This technique takes advantage of the fact that the
135: Tonry-Davis Ratio (TDR, \citealt{TonryDavis1979}) scales with $S/N$,
136: so that individual RV uncertainties may be found via: $Error(V_{r})
137: = \alpha/(1+TDR)$, where $\alpha$ is determined from the standards.
138: For our set of 64 standards, we measure $\alpha$ = 107, which yields
139: an average velocity uncertainty of $\sim$ 4.0 km s$^{-1}$ (TDR $>$ 25,
140: with typical $S/N$ of 15-20).
141:
142:
143: \section{Spectroscopic Results}
144:
145: \subsection{Bootes Membership}
146:
147:
148: With our RV data alone, the signal of Boo stars is not clear. To
149: improve the contrast of Boo stars with respect to MW contaminants, we
150: make use of the gravity sensitive MgH absorption feature near 5150
151: \AA~ to remove foreground Galactic dwarfs, as we have done
152: photometrically in our previous dSph studies (e.g.,
153: \citealt{Majewski2005}; \citealt{M05,M06}; \citealt{Westfall2006}).
154: Because the MgH absorption is very strong in all but the most
155: metal-poor ([Fe/H] $\lesssim$ -2) K--dwarf stars, visual inspection
156: alone is sufficient to ``clean" our spectroscopic sample of the
157: majority of foreground dwarfs by simply identifying those with strong
158: MgH features, like the one shown in Fig.\ 4 (see also Fig.\ 1 in
159: \citealt{Majewski2000} for an illustration).
160:
161:
162: To further clean this sample, we make use of instrumental Lick
163: spectroscopic indices (\citealt{Worthey1994}). A proper determination
164: of stellar $T_{\rm eff}$, log $g$ and [Fe/H] values based on Lick
165: indices is beyond the purpose here. Instead, we
166: identify stars with similar metal line strengths using
167: the quantitative indices of the Lick system.
168: Figure 3a shows the Mg$_{1}$ + Mg$_{2}$ versus Mg
169: $b$ trend for all
170: stars visually classified as likely giants with
171: RV uncertainties less than 7.5 km s$^{-1}$. Figure 3b shows their RV
172: distribution as a function of radial distance from the nominal center
173: of Boo. We have marked those having relatively low and high Mg indices
174: with circles and squares respectively. We also mark with asterisks a
175: group of stars clumped in $V_{r}$ ($\sim50$ km s$^{-1}$ in Fig.\
176: 3b) and Mg strength. These stars have
177: velocities compatible with those expected for debris from the leading
178: arm of the Sagittarius (Sgr) dSph (\citealt{Law2004}); because Boo
179: lies in the background of the Sgr leading arm (see Fig. 1 in
180: \citealt{field_of_streams2006}), giant stars with these RVs are not
181: unexpected in our sample. Figure 3b shows that within the $r_{\rm h}$
182: of Boo the distribution of stars shows a distinct grouping of giant
183: stars with low Mg indices and $V_{r}\sim100$ km s$^{-1}$.
184: Such a velocity grouping is unexpected for a random halo population at
185: this Galactic position, and is clearly distinct from the distribution
186: of stars beyond $r_{\rm h}$, which is centered at 0 km s$^{-1}$ (the
187: nominal mean velocity for any MW stellar population at this Galactic
188: longitude). A two-sample Kolmogorov-Smirnov test shows that the
189: populations inside and outside 13\arcmin~ are drawn from different
190: RV distributions with a probability greater than 99.99\%. Thus, we
191: identify this group of RV-clumped stars within 13\arcmin~ with the Boo dSph.
192:
193: This ``cleaner" RV distribution inside $r_{\rm h}$ helps us define a
194: membership criterion to identify Boo stars. We first draw attention
195: to the broadening of the RV distribution beyond $\sim10$\arcmin. This
196: behavior might be expected if Boo is undergoing tidal stripping, as
197: its distorted morphology suggests. Models of tidally disrupting
198: satellites predict a rising velocity dispersion at large radii (e.g.,
199: \citealt{Kroupa1997}), a trend that is also observed in a number of
200: other Galactic dSphs: Ursa Minor (UMi, \citealt{M05}), Sculptor
201: (\citealt{Westfall2006}), Carina (\citealt{M06}), Fornax
202: (\citealt{Walker2006a}), Sextans (\citealt{Walker2006b}) and Leo I
203: (\citealt{Sohn2006}). Because of the broadening of the RV
204: distribution, we conservatively use the innermost seven stars to
205: define a 3$\sigma$ (see \S 3.2) RV criterion for membership of
206: $75<V_{r}<116$ km s$^{-1}$. Twelve stars lie within this velocity
207: range, all with low Mg line strengths, the most widely separated
208: (within our survey) at 27\arcmin~ to the south of the Boo center.
209:
210: Two likely giants at higher velocity than our 3$\sigma$ RV limit are
211: seen in Figure 3b. They lie right in the narrow RGB of Boo (see
212: Fig. 2 where they are marked with open circles) and the strength of
213: their Mg features makes them consistent with being Boo members; these
214: stars suggest a possibly dramatic increase of the Boo velocity dispersion
215: with radius.
216:
217: \subsection{Velocity Dispersion, Mass and $M/L$}
218:
219: Using the innermost seven stars to define the properties of the Boo
220: core RV distribution gives a systemic velocity of
221: $V_{r}=95.6\pm3.4$ km s$^{-1}$ and a velocity dispersion of
222: $\sigma_{\rm o}=6.6\pm2.3$ km s$^{-1}$, calculated using the Maximum
223: Likelihood Method (\citealt{Pryor1993}; \citealt{Hargreaves1994};
224: \citealt{Kleyna2002}). This $V_{r}$ translates into a
225: $V_{GSR}=102.9\pm2.8$ km s$^{-1}$, implying that Boo is likely not in
226: a circular orbit, but in a rather radial one. Including all twelve
227: potential $3\sigma$ members in the calculation yields
228: $V_{r}=98.4\pm2.9$ km s$^{-1}$ and $\sigma_{\rm o}=9.0\pm2.2$ km
229: s$^{-1}$; the latter corroborating the apparent increase of the
230: velocity dispersion with radius. If we include the two likely high
231: velocity members that lie above the 3$\sigma$ RV limit (\S 3.1), we
232: obtain $\sigma_{\rm o}=14.6\pm3.0$ km s$^{-1}$.
233:
234: As is customary for this type of study, one can use the observed
235: $\sigma_{\rm o}$ to estimate the mass content in the Boo dSph under
236: the assumption that it is in dynamical equilibrium. While this
237: assumption has been a matter of some debate, it has been shown (e.g.,
238: \citealt{PP95}; \citealt{OLA95}; \citealt{Kroupa1997}; Mu\~noz et al.,
239: in prep.) that unless the satellite is completely destroyed or near
240: complete destruction, the value of $\sigma_{\rm o}$ indeed reflects
241: the instantaneous mass content. From \citet{Illingworth76}, the total
242: mass of the dSph system can be approximated by: $M_{\rm tot} = 167
243: \beta R_{c,g} V^{2}_{s}$, where $\beta$ is a correction parameter
244: dependent on the concentration value, $R_{c,g}$ is the geometric-mean
245: King core radius in pc, and $V_{\rm s}$ is a parameter related to
246: $\sigma_{\rm o}$. \citet{Mateo1998} approximates $\beta=8.0$ and
247: $V_{s}\sim\sigma_{\rm o}$ for pressure-supported systems that follow
248: low concentration King models, as dSphs do. We can further
249: approximate $R_{c,g}$ to $r_{\rm h}$ derived by \citeauthor*{B06b}
250: (13\arcmin, corresponding to 225 pc at a distance of 60 kpc).
251: With $\sigma_{\rm o}=6.6\pm2.3$ km s$^{-1}$ we obtain
252: $M_{\rm tot} = 1.1^{+1.3}_{-0.5}\times10^7$ M$_{\sun}$.
253: We can also estimate the central mass density in Boo as
254: $\rho_{\rm o} \approx 166 \sigma^{2}_{\rm o} \ R^{2}_{c,g}$ \citep{Mateo1998};
255: this yields $\rho_{\rm o} = 0.14 M_{\sun}$ pc$^{-3}$.
256:
257: These results are remarkably similar to the values obtained for the
258: majority of the Galactic dSphs. \citet{Mateo1998} points out that
259: Local Group dSph systems seem to be embedded in DM halos of
260: $M_{\rm tot} \approx 10^7$ M$_{\sun}$ regardless of the contribution
261: of their luminous component. He further finds the following empirical
262: relation between the $M/L$ and the total luminosity of a dSph: $M/L =
263: 2.5+[10^{7}/(L/L_{\sun})]$. Adopting an absolute magnitude for Boo of
264: $M_{\rm V}=-5.8$ (from \citeauthor*{B06b}) yields a total luminosity
265: of $L_{\rm V} = 1.8\times10^{4}$ L$_{\sun}$, which, in turn, gives $M/L_{\rm
266: V}=610$ (M/L)$_{\sun}$. We note that this luminosity would make Boo
267: even fainter than UMa if we adopt for the latter the absolute magnitude
268: estimate of $M_{\rm V}=-6.75$ from \citeauthor*{W05}.
269: A visual comparison of the CMDs of Boo and UMa reveals that the Boo
270: RGB contains at least a factor of two more stars than that
271: of UMa (see also \citealt{Siegel2006}).
272: Since they have comparable half-light radii, this implies that Boo
273: must be at least twice as bright as UMa. Were we to adopt the
274: \citeauthor*{W05} estimate for the UMa $M_{\rm V}$, this yields a
275: total luminosity for Boo of $L_{\rm V}\sim8.6\times10^{4}$ L$_{\sun}$,
276: making $M/L_{\rm V}=130$. With either of these values, however, Boo
277: lies squarely on the \citet{Mateo1998} relation.
278:
279:
280: \subsection{Metallicity}
281:
282: The Mg I triplet+MgH absorption features near 5150 \AA~ can be
283: exploited to estimate the Boo [Fe/H]. We have already shown (Fig.\ 3)
284: that Boo stars exhibit by far the weakest Mg features of all likely giant stars in
285: our sample. We have
286: used the CTIO 4-m telescope + the Hydra multifiber spectrograph, with
287: a comparable instrument setup (yielding similar spectral resolution)
288: to that used for the Boo observations, for RV measurements of red
289: giant candidates in the globular clusters (GCs) NGC 288 and NGC 5634
290: (Moskowitz et al., in prep.). These clusters have [Fe/H] of $-1.24$
291: and $-1.88$ respectively (\citealt{Harris1996}) and we use them as
292: metallicity calibrators, selecting for this purpose giant stars that
293: are confirmed RV cluster members. In order to minimize the effect of
294: surface gravity and temperature in the measurements of equivalent
295: width (EW) of the Mg features, we pick and combine the spectra of stars
296: that lie in the upper part of their respective red giant branches,
297: which provides bright members with fairly similar colors.
298: Figure 4 shows the combined spectra obtained for the two clusters and for Boo.
299: We then add the EWs for the three Mg lines to compensate for the
300: weakness of the Mg features, and assume a linear
301: function between this sum and [Fe/H]\footnote{Figure 9 from \citet{Buzzoni1992}
302: suggests that a linear relationship between Mg$_{2}$ and [Fe/H] is not
303: unreasonable for [Fe/H] $< -1.3$.}.
304: Using this derived relationship and assuming
305: a similar [Mg/Fe] ratio between the GC and Boo stars\footnote{\citet{Shetrone2001} and
306: \citet{Shetrone2003} showed
307: that metal poor stars in Sculptor, Leo I and UMi
308: ([Fe/H] $<-2.0$) have comparable [$\alpha$/Fe] to Galactic GC stars.},
309: we translate the EWs measured for Boo into an extrapolated [Fe/H] of $-2.5$.
310: The uncertainties in the EW measurements translate into
311: uncertainties in [Fe/H] of about 0.2 dex. This does
312: not include, of course, the error introduced by assuming a linear
313: relationship between EWs and [Fe/H] in the first place, nor the uncertainties
314: due to surface gravity, temperature effects, or possible variations in
315: $\alpha$-element abundances between the GCs and Boo.
316:
317:
318: These calculations, although only intended to provide a very rough
319: estimation of the metallicity of Boo, are consistent
320: with the [Fe/H] = $-2.6$ derived by \citet{Siegel2006} using RR Lyrae variables,
321: and with the fact that the Boo RGB seems to be slightly bluer
322: than that of M92 ([Fe/H] = $-2.3$; see Fig. 2 of \citeauthor*{B06b}).
323: Boo is potentially the most metal poor of the Galactic dSphs known to date.
324:
325: \section{Discussion}
326:
327: We spectroscopically survey the Boo dSph and
328: derive both its systemic velocity
329: ($95.6\pm3.4$ km s$^{-1}$) and central velocity dispersion
330: ($6.6\pm2.3$ km s$^{-1}$), which
331: yields a mass of $M_{\rm tot} = 1.1^{+1.3}_{-0.5}\times10^7$ M$_{\sun}$.
332: This mass is
333: similar to that of the other dSph galaxies and puts Boo squarely on the
334: ``same mass-just different luminosities" trend identified by \citet{Mateo1998},
335: despite the fact that Boo is one of the faintest known Galactic satellites.
336:
337: The dynamical mass derived for Boo, taken at face value, implies that
338: this is possibly also the darkest dSph known to date. If other
339: systems of similar luminosity have the same mass, the current pace of
340: discovery of these systems (three in the past year;
341: \citeauthor*{W05}; \citealt{Zucker_CVn2006}; \citeauthor*{B06b})
342: --- and in only the approximately 20\% of the sky covered by the SDSS
343: DR5 --- will help alleviate the current order of magnitude or two
344: deficit of known Galactic satellites compared to that predicted by
345: $\Lambda$CDM simulations, albeit only for one part of the mass
346: spectrum exhibiting the apparent ``missing satellites" shortfall
347: (e.g., \citealt{Klypin1999}). While Boo seems
348: to be the most DM dominated dwarf, it is, at the same time,
349: the Galactic satellite with the most distorted known morphology
350: (\citeauthor*{B06b}) and possibly most dramatic increase in velocity
351: dispersion with radius. This implies that Boo may be among the most
352: disrupted Galactic dwarfs, to the extent that
353: it even lacks a proper core (\citeauthor*{B06b}). In fact a puzzling
354: correlation is now emerging between the DM fraction of a dSph
355: and its morphology, wherein the faintest and most distorted systems (the UMi, UMa and Boo dSphs)
356: seem also to present the largest central $M/L$'s. If the distorted contours
357: are a response to the influence of Galactic tides, and if all dSphs indeed
358: have a similar current total mass and density, then why is it that tides
359: seem to affect preferentially the least luminous systems?
360: Perhaps these faintest systems
361: represent the \citet{Kroupa1997} regime in which disruption has
362: proceeded to the point where the central velocity dispersions
363: are inflated by tides, artificially increasing the derived $M/L$. But
364: even if so, then it is curious that the current
365: $\sigma_{\rm o}$ for Boo would be inflated to just such a value that the derived
366: (but artificial) dynamical mass still participates in the ``same mass"
367: conspiracy of dSph galaxies.
368:
369: We appreciate useful discussions with Gregory Sivakoff, Jeffrey Crane and Allyson
370: Polak. We gratefully acknowledge support by NSF grant AST-0307851,
371: NASA/JPL contract 1228235, the Virginia Space Grant Consortium and
372: Frank Levinson through the Celerity Foundation.
373: D.L.N. is also supported by the ARCS Foundation and the Green Bank Telescope
374: Student Support Program. P.M.F. is supported by NASA GSRP and UVa dissertation fellowships.
375:
376:
377: \begin{thebibliography}{}
378:
379: \bibitem[Aaronson(1983)]{Aaronson1983} Aaronson, M. 1983, \apj, 266, L11
380:
381: \bibitem[Abazajian et al.(2005)]{Abazajian2005} Abazajian, K., et al. 2005, \aj, 129, 1755
382:
383: \bibitem[Adelman-McCarthy et al.(2006)]{sdss_dr4}
384: Adelman-McCarthy, J.~K., et al.\ 2006, \apjs, 162, 38
385:
386: \bibitem[Belokurov et al.(2006a)]{field_of_streams2006} Belokurov, V., et
387: al.\ 2006a, \apjl, 642, L137
388:
389: \bibitem[Belokurov et al.(2006b)B06b]{B06b} Belokurov, V., et
390: al.\ 2006b, ApJ, 647, L111 (B06b)
391:
392: \bibitem[Burkert(1997)]{Burkert1997} Burkert, A. 1997, \apjl, 474, L99
393:
394: \bibitem[Buzzoni, Gariboldi, \& Mantegazza(1992)]{Buzzoni1992} Buzzoni, A.,
395: Gariboldi, G., \& Mantegazza, L. 1992, \aj, 130, 1814
396:
397: \bibitem[Hargreaves et al.(1994)]{Hargreaves1994} Hargreaves, J. C., Gilmore, G.,
398: Irwin, M. J., \& Carter, D. 1994, \mnras, 269, 957
399:
400:
401: \bibitem[Harris(1996)]{Harris1996} Harris, W.~E.\ 1996, \aj, 112, 1487
402:
403:
404: \bibitem[Illingworth(1976)]{Illingworth76} Illingworth, G. 1976, \apj, 204, 73
405:
406:
407: \bibitem[Kleyna et al.(2002)]{Kleyna2002} Kleyna, J. T., Wilkinson, M.
408: I., Evans, N. W., Gilmore, G., \& Frayn, C. 2002, \mnras, 330, 792
409:
410: \bibitem[Kleyna et al.(2005)]{Kleyna2005} Kleyna, J. T., Wilkinson, M.
411: I., Evans, N. W., \& Gilmore, G. 2005, \apj, 630, L141
412:
413: \bibitem[Klypin et al.(1999)]{Klypin1999} Klypin, A., Kravtsov, A. V.,
414: Valenzuela, O., \& Prada, F. 1999, \apj, 522, 82
415:
416: \bibitem[Kroupa(1997)]{Kroupa1997} Kroupa, P. 1997, New Astronomy, 2, 139
417:
418:
419: \bibitem[{\L}okas, Mamon, \& Prada(2005)]{Lokas2005} {\L}okas, E. L., Mamon, G. A., \&
420: Prada, F. 2005, \mnras, 363, 918
421:
422: \bibitem[Law et al.(2004)]{Law2004} Law, D. R., Johnston, K. V.,
423: \& Majewski, S. R. 2004, ApJ, 619, 807
424:
425:
426: \bibitem[Majewski et al.(2000)]{Majewski2000} Majewski, S. R., Ostheimer,
427: J. C., Kunkel, W. E., \& Patterson, R. J. 2000, \aj, 120, 2550
428:
429:
430: \bibitem[Majewski et al.(2005)]{Majewski2005} Majewski, S. R., et al. 2005, \aj, 130, 2677
431:
432:
433: \bibitem[Mateo(1998)]{Mateo1998} Mateo, M. L. 1998, \araa, 36, 435
434:
435:
436: \bibitem[Mu\~noz et al.(2005)]{M05} Mu\~noz, R. R. et al. 2005 \apjl, 631, L137
437:
438: \bibitem[Mu\~noz et al.(2006)]{M06} Mu\~noz, R. R. et al. 2006 \apj, {\it in press} (astro-ph/0605098)
439:
440:
441: \bibitem[Oh, Lin, \& Aarseth(1995)]{OLA95} Oh, K.~S., Lin,
442: D.~N.~C., \& Aarseth, S.~J.\ 1995, \apj, 442, 142
443:
444: \bibitem[Piatek \& Pryor(1995)]{PP95} Piatek, S., \& Pryor, C.\ 1995, \aj, 109, 1071
445:
446: \bibitem[Pryor \& Meylan(1993)]{Pryor1993} Pryor, C., \& Meylan, G.\ 1993, in ASP Conf. Ser. 50,
447: Structure and Dynamics of Globular Clusters, ed. S. Djorgovski \& G. Meylan (San Francisco: ASP), 357
448:
449:
450: \bibitem[Shetrone, Cote, \& Sargent(2001)]{Shetrone2001} Shetrone, M. D.,
451: Cote, P., \& Sargent, W. L. W. 2001, \apj, 548, 592
452:
453: \bibitem[Shetrone et al.(2003)]{Shetrone2003} Shetrone, M. D., Venn, K. A., Tolstoy, E.,
454: Primas, F., Hill, V., \& Kaufer, A. 2003, \aj, 125, 684
455:
456: \bibitem[Siegel(2006)]{Siegel2006} Siegel, M.~H.\ 2006, ApJL, {\it submitted} (astro-ph/0607091)
457:
458: \bibitem[Sohn et al.(2006)]{Sohn2006} Sohn, S. et al. 2006, \apj, {\it submitted}
459:
460: \bibitem[Tonry \& Davis(1979)]{TonryDavis1979} Tonry, J., \& Davis,
461: M.\ 1979, \aj, 84, 1511
462:
463:
464: \bibitem[Vogt et al.(1995)]{Vogt1995} Vogt, S. S., Mateo, M., Olszewski,
465: E. W., \& Keane, M. J. 1995, \aj, 109, 151
466:
467: \bibitem[Walker et al.(2006a)]{Walker2006a} Walker, M. G., Mateo, M., Olszewski, E. W.,
468: Bernstein, R. A., Wang, X., \& Woodroofe, M. 2006a, \aj, 131, 2114
469:
470: \bibitem[Walker et al.(2006b)]{Walker2006b} Walker, M. G., Mateo, M., Olszewski, E. W.,
471: Pal, J. K., Sen, B., \& Woodroofe, M. 2006b, \apj, 642, L44
472:
473: \bibitem[Westfall et al.(2006)]{Westfall2006} Westfall, K. B., Ostheimer,
474: J. C., Frinchaboy, P. M., Patterson, R. J., Majewski, S. R., \&
475: Kunkel, W. E. 2006, \aj, 131, 375
476:
477: \bibitem[Willman et al.(2005)W05]{W05} Willman, B. et al. 2005, \apj, 626, L85 (W05)
478:
479: \bibitem[Worthey et al.(1994)]{Worthey1994} Worthey, G., Faber,
480: S.~M., Gonzalez, J.~J., \& Burstein, D.\ 1994, \apjs, 94, 687
481:
482: \bibitem[Zucker et al.(2006)]{Zucker_CVn2006} Zucker, D.~B., et al. 2006, \apj,
483: 643, L103
484:
485:
486: \end{thebibliography}
487:
488: \clearpage
489:
490: \begin{figure}
491: \plotone{f1.eps}
492: \caption{Spatial distribution of all 58 stars with measured RVs.
493: Filled triangles correspond to the final Boo sample (from Fig.\ 3 analysis),
494: while open triangles mark stars not considered members.
495: Two open circles represent the higher-velocity stars discussed in the text as
496: possible members based on the similarity of their Mg indices to those of other Boo stars.
497: The Boo $r_{\rm h}$ determined by \citeauthor*{B06b} is delineated by the dotted circle.}
498: \end{figure}
499:
500: \clearpage
501:
502: \begin{figure}
503: \plotone{f2.eps}
504: \caption{CMD of the Boo region showing the clear Boo RGB and
505: horizontal branch. The dots represent stars within the $r_{\rm h}$ of
506: Boo (dotted circle in Fig.\ 1) in order to highlight Boo's features,
507: while the other symbols show {\it all} 58 stars as in Fig.\ 1.}
508: \end{figure}
509:
510: \clearpage
511:
512: \begin{figure}
513: \plotone{f3.eps}
514: \caption{(a) Mg$_{1}$+Mg$_{2}$ versus Mg $b$ instrumental Lick
515: indices for all stars with velocity uncertainties $<$ 7.5 km s$^{-1}$
516: visually classified as likely giants. Circles mark stars most likely
517: to be metal poor, while squares show stars more likely to be metal
518: rich. We mark with asterisks a clump of stars with similar indices and
519: RVs, possibly from the Sgr dSph. (b) RVs of all stars in (a) as a
520: function of radial distance from the center of the Boo dSph. Symbols
521: as in panel (a). The dotted vertical line marks the Boo $r_{\rm h}$,
522: while the dotted horizontal lines delimit a 3$\sigma$ RV spread. The
523: two higher-velocity stars discussed in \S 3.1--3.2 (open circles in
524: Fig.\ 1) lie just above the upper 3$\sigma$ limit.}
525: \end{figure}
526:
527: \clearpage
528:
529: \begin{figure}
530: \plotone{f4.eps}
531: \caption{Sample of the spectra used to estimate the metallicity of
532: Boo. From top to bottom: Combined spectrum of the three brightest Boo
533: giant stars, combined spectrum for NGC 5634 giants and combined
534: spectrum for NGC 288 giant stars. The comparison spectrum of a dwarf star
535: demonstrates the broad MgH absorption feature between $\sim$4900-5250 \AA, which is an
536: obvious signature of late G or K dwarf stars.}
537: \end{figure}
538:
539:
540: \end{document}
541: