1: \documentclass[useAMS,usenatbib]{mn2e}
2: \bibliographystyle{mn2e}
3: %\def\baselinestretch{1}
4: %\usepackage[dvips]{graphicx}
5: \usepackage{graphicx}
6: \usepackage{times}
7: \usepackage{amssymb}
8: \usepackage{natbib}
9: %%%%%%%%%% user-defined commands %%%%%%%%%%%%%%%%%
10: \def\gsim { \lower .75ex \hbox{$\sim$} \llap{\raise .27ex \hbox{$>$}} }
11: \def\lsim { \lower .75ex \hbox{$\sim$} \llap{\raise .27ex \hbox{$<$}} }
12: %%%%%%%%%%%%%%%%%%%%%%%%%%%
13:
14: \newcommand{\apj}{ApJ}
15: \newcommand{\apjl}{ApJL}
16: \newcommand{\apjs}{ApJS}
17: \newcommand{\aj}{AJ}
18: \newcommand{\mnras}{MNRAS}
19: \newcommand{\nat}{Nature}
20: \newcommand{\pasj}{PASJ}
21: \newcommand{\araa}{ARA\&A}
22: \newcommand{\aapr}{A\&ARv}
23: \newcommand{\aap}{A\&A}
24: \newcommand{\prd}{PhRvD}
25: \newcommand{\physrep}{PhysRep}
26:
27: %%%%%%%%%% end user-defined commands %%%%%%%%%%%%%%%%%
28:
29: \begin{document}
30:
31: \title[Satellites on Extreme Orbits]
32: {Cosmic M\'enage \`a Trois: The Origin of Satellite Galaxies \\ on Extreme Orbits}
33:
34: \author[Sales, Navarro, Abadi \& Steinmetz]{Laura V. Sales$^{1,2}$, Julio F. Navarro,$^{3,4}$\thanks{Fellow of the Canadian Institute for Advanced Research.} Mario G. Abadi $^{1,2,3}$
35: and Matthias Steinmetz$^{5}$
36: \\
37: $^{1}$ Observatorio Astron\'omico, Universidad Nacional de C\'ordoba, Laprida
38: 854, 5000 C\'ordoba, Argentina.
39: \\
40: $^{2}$ Instituto de Astronom\'{\i}a Te\'orica y Experimental, Conicet, Argentina.
41: \\
42: $^{3}$Department of Physics and Astronomy, University of Victoria, Victoria, BC V8P 5C2,
43: Canada\\
44: %
45: $^{4}$Max-Planck Institut f\"ur Astrophysik, Karl-Schwarzschild Strasse 1,
46: Garching, D-85741, Germany\\
47: %
48: $^{5}$Astrophysikalisches Institut Potsdam, An der Sternwarte 16, Potsdam 14482, Germany\\
49: %
50: %
51: }
52:
53: \maketitle
54:
55: \begin{abstract}
56: We examine the orbits of satellite galaxies identified in a suite of
57: N-body/gasdynamical simulations of the formation of $L_*$ galaxies in
58: a $\Lambda$CDM universe. The numerical resolution of the simulations
59: allows us to track in detail the orbits of the $\sim $ ten brightest
60: satellites around each primary. Most satellites follow conventional
61: orbits; after turning around, they accrete into their host halo and
62: settle on orbits whose apocentric radii are steadily eroded by
63: dynamical friction. As a result, satellites associated with the
64: primary are typically found within its virial radius, $r_{\rm vir}$,
65: and have velocities consistent with a Gaussian distribution with mild
66: radial anisotropy. However, a number of outliers are also present. We
67: find that a surprising number (about one-third) of satellites
68: identified at $z=0$ are on unorthodox orbits, with apocenters that
69: exceed their turnaround radii. These include a number of objects with
70: extreme velocities and apocentric radii at times exceeding $\sim 3.5\,
71: r_{\rm vir}$ (or, e.g., $\gsim \, 1$ Mpc when scaled to the Milky
72: Way). This population of satellites on extreme orbits consists
73: typically of the faint member of a satellite pair whose kinship is
74: severed by the tidal field of the primary during first approach. Under
75: the right circumstances, the heavier member of the pair remains bound
76: to the primary, whilst the lighter companion is ejected onto a
77: highly-energetic orbit. Since the concurrent accretion of multiple
78: satellite systems is a defining feature of hierarchical models of
79: galaxy formation, a fairly robust prediction of this scenario is that
80: at least some of these extreme objects should be present in the Local
81: Group. We speculate that this three-body ejection mechanism may be the
82: origin of (i) some of the newly discovered high-speed satellites
83: around M31 (such as Andromeda XIV); (ii) some of the distant
84: fast-receding Local Group members, such as Leo I; and (iii) the oddly
85: isolated dwarf spheroidals Cetus and Tucana in the outskirts of the
86: Local Group. Our results suggest that care must be exercised when
87: using the orbits of the most weakly bound satellites to place
88: constraints on the total mass of the Local Group.
89: \end{abstract}
90:
91: \begin{keywords}
92: galaxies: haloes - galaxies: formation -
93: galaxies: evolution - galaxies: kinematics and dynamics.
94: \end{keywords}
95:
96: \section{Introduction}
97: \label{sec:intro}
98:
99: The study of Local Group satellite galaxies has been revolutionized by
100: digital imaging surveys of large areas of the sky. More than a dozen
101: new satellites have been discovered in the past couple of years
102: \citep{zucker04,zucker06,willman05b,martin06,belokurov06,
103: belokurov07,irwin07,majewski07},
104: due in large part to the completion of the Sloan Digital Sky
105: Survey \citep{york00,strauss02} and to concerted campaigns designed to image in detail
106: the Andromeda galaxy and its immediate surroundings (\citealt{ibata01,ferguson02,
107: reitzel02,mcconnachie03,rich04,guhathakurta06,gilbert06,chapman06},
108: Ibata et al. 2007 submitted). The
109: newly discovered satellites have extended the faint-end of the galaxy
110: luminosity function down to roughly $\sim 10^3 \,
111: L_{\odot}$, and are likely to provide important constraints regarding
112: the mechanisms responsible for ``lighting up'' the baryons in low-mass
113: halos. These, in turn, will serve to validate (or falsify) the various
114: theoretical models attempting to reconcile the wealth of
115: ``substructure'' predicted in cold dark matter (CDM) halos with the
116: scarcity of luminous satellites in the Local Group
117: \citep[see, e.g.][]{klypin99b,bullock00, benson02,stoehr02,
118: kazantzidis04,kravtsov04,penarrubia07}.
119:
120: At the same time, once velocities and distances are secured for the
121: newly-discovered satellites, dynamical studies of the total mass and
122: spatial extent of the Local Group will gain new impetus. These studies
123: have a long history \citep{littleandtremaine87,zaritsky89,kochanek96,wilkinson99,
124: evans00,battaglia05}, but their results have traditionally
125: been regarded as tentative rather than conclusive, particularly
126: because of the small number of objects involved, as well as the
127: sensitivity of the results to the inclusion (or omission) of one or
128: two objects with large velocities and/or distances \citep{zaritsky89,
129: kochanek96,sakamoto03}. An enlarged satellite sample will likely
130: make the conclusions of satellite dynamical studies more
131: compelling and robust.
132:
133: To this end, most theoretical work typically assumes that satellites are in
134: equilibrium, and use crafty techniques to overcome the limitations of
135: small-N statistics when applying Jeans' equations to estimate masses
136: \citep[see, e.g.,][]{littleandtremaine87,wilkinson99,evans00}. With increased
137: sample size, however, follow enhanced
138: opportunities to discover satellites on unlikely orbits; i.e.,
139: dynamical ``outliers'' that may challenge the expectations of
140: simple-minded models of satellite formation and evolution. It is
141: important to clarify the origin of such systems, given their
142: disproportionate weight in mass estimates.
143:
144: One issue to consider is that the assumption of equilibrium must break
145: down when considering outliers in phase space. This is because the
146: finite age of the Universe places an upper limit to the orbital period
147: of satellites observed in the Local Group; high-speed satellites have
148: typically large apocenters and long orbital periods, implying that
149: they cannot be dynamically well-mixed and casting doubts on the
150: applicability of Jeans' theorem-inspired analysis tools.
151:
152: To make progress, one possibility is to explore variants of the
153: standard secondary infall model \citep{gunnandgott72, gott75,
154: gunn77,fillmore84}, where satellites are assumed to recede initially with
155: the universal expansion, before turning around and collapsing onto the
156: primary due to its gravitational pull. This is the approach adopted by
157: \citet{zaritskyandwhite94} in order to interpret statistically the
158: kinematics of observed satellite samples without assuming well-mixed
159: orbits and taking into account the proper timing and phase of the
160: accretion process.
161:
162:
163:
164: %%%%%%%%%%%%%%%%%%%%%%%%%%
165: \begin{center}
166: \begin{figure*}
167: \includegraphics[width=100mm]{figs/letfig1.ps}
168: \caption{Star particles in one of our simulations, shown at
169: $z=0$. Particles are colored according to the age of the star; blue
170: means a star is younger than $\simeq 1$ Gyr, red that it is older than
171: $\simeq 10$ Gyr. The large box is $2\, r_{\rm vir}$ ($632$ kpc) on a
172: side and centered on the primary galaxy. More than $85\%$ of all stars
173: are in the inner regions of the primary, within about $\sim 20$
174: kpc from the center \citep[for more details see][]{abadi06}.
175: surround the satellites ``associated'' with the primary galaxy; i.e.,
176: satellites that have been within $r_{\rm vir}$ in the past. Note that
177: a few ``associated'' satellites lie well beyond the virial boundary of
178: the system. Two of these satellites are highlighted for analysis in
179: Figures~\ref{fig:orbesc} and \ref{fig:orbxyesc}.}
180: \label{fig:xyzsat}
181: \end{figure*}
182: \end{center}
183: %%%%%%%%%%%%%%%%%%%%%%%%%%
184:
185:
186:
187: In the secondary infall accretion sequence, satellites initially
188: farther away accrete later, after turning around from larger
189: turnaround radii. The turn-around radius grows with time, at a rate
190: the depends on the mass of the primary and its environment, as well as
191: on the cosmological model. Three distinct regions surround a system
192: formed by spherical secondary infall \citep[see, e.g.,][]{bertschinger85,
193: navarroandwhite93}: (i) an outer region beyond the current
194: turnaround radius where satellites are still expanding away, (ii) an
195: intermediate region containing satellites that are approaching the
196: primary for the first time, and (iii) an inner, ``virialized'' region
197: containing all satellites that have turned around at earlier times and
198: are still orbiting around the primary. To good approximation, the
199: latter region is delineated roughly by the conventional virial radius
200: of a system{\footnote{We define the {\it virial} radius, $r_{\rm
201: vir}$, of a system as the radius of a sphere of mean density $\simeq
202: \Delta_{\rm vir}(z)$ times the critical density for closure. This
203: definition defines implicitly the virial mass, $M_{\rm vir}$, as that
204: enclosed within $r_{\rm vir}$, and the virial velocity, $V_{\rm vir}$,
205: as the circular velocity measured at $r_{\rm vir}$. We compute
206: $\Delta_{\rm vir}(z)$ using $\Delta_{vir}(z)=18\pi^2+82f(z)-39f(z)^2$, where
207: $f(z)=[\Omega_0(1+z)^3/(\Omega_0(1+z)^3+\Omega_\Lambda))]-1$ and
208: $\Omega_0=\Omega_{\rm CDM}+\Omega_{\rm
209: bar}$ \citep{bryanandnorman98}, and is $\sim 100$ at $z=0$}}, $r_{\rm vir}$;
210: the turnaround radius is of order $r_{\rm ta} \sim 3 \,r_{\rm vir}$
211: \citep[see, e.g.][]{white93}.
212:
213: We note a few consequences of this model. (a) Satellites outside the
214: virial radius are on their first approach to the system and thus have
215: not yet been inside $r_{\rm vir}$. (b) Satellites inside the virial
216: radius have apocentric radii that typically do not exceed $r_{\rm
217: vir}$. (c) The farther the turnaround radius the longer it takes for a
218: satellite to turn around and accrete and the higher its orbital
219: energy. (d) Satellites with extreme velocities will, in general, be
220: those completing their first orbit around the primary. Velocities will
221: be maximal near the center, where satellites may reach speeds as high
222: as $\sim 3 \,V_{\rm vir}$. (e) Since all satellites associated with
223: the primary are bound (otherwise they would not have turned around and
224: collapsed under the gravitational pull of the primary), the velocity
225: of the highest-speed satellite may be used to estimate a lower limit
226: to the escape velocity at its location and, thus, a lower bound to the
227: total mass of the system.
228:
229: Hierarchical galaxy formation models, such as the current $\Lambda$CDM
230: paradigm, suggest further complexity in this picture. Firstly,
231: although numerical simulations show that the sequence of expansion,
232: turnaround and accretion of satellites described above is more or less
233: preserved in hierarchical models, the evolution is far from
234: spherically symmetric
235: \citep{navarro94,ghigna98,jing02,bailin05,knebe06}. Much of the mass
236: (as well as many of the satellites) is accreted through filaments of
237: matter embedded within sheets of matter formation \citep[see,
238: e.g.,][]{navarro04}. The anisotropic collapse pattern onto a primary
239: implies that the turnaround ``surface'' won't be spherical and that
240: the virial radius may not contain {\it all} satellites that have
241: completed at least one orbit around the primary \citep[see,
242: e.g.,][]{balogh00,diemand07}.
243:
244: More importantly for the purposes of this paper, in hierarchical
245: models galaxy systems are assembled by collecting smaller systems
246: which themselves, in turn, were assembled out of smaller units. This
247: implies that satellites will in general not be accreted in isolation,
248: but frequently as part of larger structures containing multiple
249: systems. This allows for complex many-body interactions to take place
250: during approach to the primary that may result in substantial
251: modification to the orbits of accreted satellites.
252:
253: We address this issue in this contribution using N-body/gasdynamical
254: simulations of galaxy formation in the current $\Lambda$CDM paradigm.
255: We introduce briefly the simulations in \S~\ref{sec:numexp}, and
256: analyze and discuss them in \S~\ref{sec:analysis}. We speculate on
257: possible applications to the Local Group satellite population in
258: \S\ref{sec:LG} and conclude with a brief summary in \S~\ref{sec:conc}.
259:
260:
261: \section{The Numerical Simulations}
262: \label{sec:numexp}
263:
264: We identify satellite galaxies in a suite of eight simulations of the
265: formation of $L_*$ galaxies in the $\Lambda$CDM scenario. This series
266: has been presented by Abadi, Navarro \& Steinmetz (2006), and follow
267: the same numerical scheme originally introduced by
268: \citet{steinmetzandnavarro02}. The ``primary'' galaxies in
269: these simulations have been analyzed in detail in several recent
270: papers, which the interested reader may wish to consult for
271: details \citep{abadi03a,abadi03b,meza03,meza05,navarro04}.
272: We give a brief outline below for completeness.
273:
274: Each simulation follows the evolution of a small region of the
275: universe chosen so as to encompass the mass of an $L_{*}$ galaxy
276: system. This region is chosen from a large periodic box and
277: resimulated at higher resolution preserving the tidal fields from the
278: whole box. The simulation includes the gravitational effects of dark
279: matter, gas and stars, and follows the hydrodynamical evolution of the
280: gaseous component using the Smooth Particle Hydrodynamics (SPH)
281: technique \citep{steinmetz96}. We adopt the following cosmological
282: parameters for the $\Lambda$CDM scenario: $H_0=65$ km/s/Mpc,
283: $\sigma_8=0.9$, $\Omega_{\Lambda}=0.7$, $\Omega_{\rm CDM}=0.255$,
284: $\Omega_{\rm b}=0.045$, with no tilt in the primordial power
285: spectrum.
286:
287: All re-simulations start at redshift $z_{\rm init}=50$, have force
288: resolution of order $1$ kpc, and the mass resolution is chosen so that
289: each galaxy is represented on average, at $z=0$, with $\sim 50,000$
290: dark matter/gas particles. Gas is turned into stars at rates
291: consistent with the empirical Schmidt-like law of \citet{kennicutt98}.
292: Because of this, star formation proceeds efficiently only in
293: high-density regions at the center of dark halos, and the stellar
294: components of primary and satellite galaxies are strongly segregated
295: spatially from the dark matter.
296:
297: Each re-simulation follows a single $\sim L_*$ galaxy in detail, and
298: resolves as well a number of smaller, self-bound systems of stars,
299: gas, and dark matter we shall call generically ``satellites''. We
300: shall hereafter refer to the main galaxy indistinctly as ``primary''
301: or ``host''. The resolved satellites span a range of luminosities,
302: down to about six or seven magnitudes fainter than the primary. Each
303: primary has on average $\sim 10$ satellites within the virial radius.
304:
305: Figure~\ref{fig:xyzsat} illustrates the $z=0$ spatial configuration of
306: star particles in one of the simulations of our series. Only star
307: particles are shown here, and are colored according to their age:
308: stars younger than $\simeq 1$ Gyr are shown in blue; those older than
309: $\simeq 10$ Gyr in red. The large box is centered on the primary and
310: is $2\, r_{\rm vir}$ ($632$ kpc) on a side. The ``primary'' is
311: situated at the center of the large box and contains most of the
312: stars. Indeed, although not immediately apparent in this rendition,
313: more than $85\%$ of all stars are within $\sim 20$ kpc from the
314: center. Outside that radius most of the stars are old and belong to
315: the stellar halo, except for a plume of younger stars stripped from a
316: satellite that has recently merged with the primary. Satellites
317: ``associated'' with the primary (see \S~\ref{ssec:convorb} for a
318: definition) are indicated with small boxes. Note that a few of them
319: lie well beyond the virial radius of the primary.
320:
321: A preliminary analysis of the properties of the simulated satellite
322: population and its relation to the stellar halo and the primary galaxy
323: has been presented in Abadi, Navarro \& Steinmetz (2006) and Sales et
324: al (2007, submitted), where the interested reader may find further details.
325: %%%%%%%%%%%%%%%%%%%%%%%%%%
326: \begin{center}
327: \begin{figure}
328: \includegraphics[width=84mm]{figs/letfig2.ps}
329: \caption{ Distance to the primary as a function of
330: time for four satellites selected in one of our simulations. The four
331: satellites are accreted into the primary in two pairs of unequal
332: mass. The heavier satellite of the pair, shown by solid lines, follows
333: a ``conventional'' orbit: after turning around from the universal
334: expansion, it accretes into the primary on a fairly eccentric orbit
335: which becomes progressively more bound by the effects of dynamical
336: friction. Note that, once accreted, these satellites on
337: ``conventional'' orbits do not leave the virial radius of the primary,
338: which is shown by a dotted line. The light member of the pair, on
339: the other hand, is ejected from the system as a result of a three-body
340: interaction between the pair and the primary during first
341: approach. One of the ejected satellites shown here is the ``escaping''
342: satellite identified in Figure~\ref{fig:rvMW}; the other is the most
343: distant ``associated'' satellite in that Figure. The latter is still
344: moving toward apocenter at $z=0$, which we estimate to be as far as
345: $\sim 3.5\, r_{\rm vir}$.}
346: \label{fig:orbesc}
347: \end{figure}
348: \end{center}
349: %%%%%%%%%%%%%%%%%%%%%%%%%%
350:
351: \section{Results and Discussion}
352: \label{sec:analysis}
353:
354: \subsection{Satellites on conventional orbits}
355: \label{ssec:convorb}
356:
357: The evolution of satellites in our simulations follows roughly the
358: various stages anticipated by our discussion of the secondary infall
359: model; after initially receding with the universal expansion, satellites turn around and are
360: accreted into the primary. Satellites massive enough to be well
361: resolved in our simulations form stars actively before accretion and,
362: by the time they cross the virial radius of the primary, much of their
363: baryonic component is in a tightly bound collection of stars at the
364: center of their own dark matter halos.
365:
366: The stellar component of a satellite is thus quite resilient to the
367: effect of tides and can survive as a self-bound entity for several
368: orbits. This is illustrated by the {\it solid lines} in
369: Figure~\ref{fig:orbesc}, which show, for one of our simulations, the
370: evolution of the distance to the primary of two satellites that turn
371: around and are accreted into the primary at different times. As
372: expected from the secondary infall model, satellites that are
373: initially farther away turn around later; do so from larger radii; and
374: are on more energetic orbits. After accretion (defined as the time
375: when a satellite crosses the virial radius of the primary), their
376: orbital energy and eccentricity are eroded by dynamical friction, and
377: these two satellites do not leave the virial radius of the primary,
378: shown by the dotted line in Figure~\ref{fig:orbesc}. Depending on
379: their mass and orbital parameters, some of these satellites merge with
380: the primary shortly after accretion, while others survive as
381: self-bound entities until $z=0$. For short, we shall refer to
382: satellites that, by $z=0$, have crossed the virial radius boundary at
383: least once as satellites ``associated'' with the primary.
384:
385: The ensemble of surviving satellites at $z=0$ have kinematics
386: consistent with the evolution described above. This is illustrated in
387: Figure~\ref{fig:rvMW}, where we show the radial velocities of all
388: satellites as a function of their distance to the primary, scaled to
389: virial units. Note that the majority of ``associated'' satellites
390: (shown as circles in this figure) are confined within $r_{\rm vir}$,
391: and that their velocity distribution is reasonably symmetric and
392: consistent with a Gaussian (Sales et al 2007). The most recently
393: accreted satellites tend to have higher-than-average speed at all
394: radii, as shown by the ``crossed'' circles, which identify all
395: satellites accreted within the last $3$ Gyr.
396:
397: Crosses (without circles) in this figure correspond to satellites that
398: have not yet been accreted into the primary. These show a clear infall
399: pattern outside $r_{\rm vir}$, where the mean infall velocity
400: decreases with radius and approaches zero at the current turnaround
401: radius, located at about $3 \, r_{\rm vir}$. All of these properties
402: agree well with the expectations of the secondary infall model
403: discussed above.
404:
405: \subsection{Three-body interactions and satellites on unorthodox orbits}
406:
407: Closer examination, however, shows a few surprises. To begin with, a
408: number of ``associated'' satellites are found outside $r_{\rm
409: vir}$. As reported in previous work \citep[see, e.g.,][]{balogh00,moore04,
410: gill05,diemand07}, these are a minority ($\sim 15\%$ in
411: our simulation series), and have been traditionally linked to
412: departures from spherical symmetry during the accretion
413: process. Indeed, anisotropies in the mass distribution during
414: expansion and recollapse may endow some objects with a slight excess
415: acceleration or, at times, may push satellites onto rather tangential
416: orbits that ``miss'' the inner regions of the primary, where
417: satellites are typically decelerated into orbits confined within the
418: virial radius.
419:
420: These effects may account for some of the associated satellites found
421: outside $r_{\rm vir}$ at $z=0$, but cannot explain why $\sim 33\%$ of
422: all associated satellites are today on orbits whose apocenters exceed
423: their turnaround radius. This is illustrated in
424: Figure~\ref{fig:rhist}, where we show a histogram of the ratio between
425: apocentric radius (measured at $z=0$; $r_{\rm apo}$) and turnaround
426: radius ($r_{\rm ta}$). The histogram highlights the presence of two
427: distinct populations: satellites on ``conventional'' orbits with
428: $r_{\rm apo}/r_{\rm ta} <1$, and satellites on orbital paths that lead
429: them well beyond their original turnaround radius.
430:
431: Intriguingly, a small but significant fraction ($\sim 6\%$) of
432: satellites have extremely large apocentric radius, exceeding their
433: turnaround radius by $50\%$ or more. These systems have clearly been
434: affected by some mechanism that propelled them onto orbits
435: substantially more energetic than the ones they had followed until
436: turnaround. This mechanism seems to operate preferentially on low-mass
437: satellites, as shown by the dashed histogram in
438: Figure~\ref{fig:rhist}, which corresponds to satellites with stellar
439: masses less than $\sim 3\%$ that of the primary.
440:
441: %%%%%%%%%%%%%%%%%%%%%%%%%%
442: %\begin{center}
443: \begin{figure}
444: \includegraphics[width=84mm]{figs/letfig3.ps}
445: \caption{Radial velocity of satellites versus distance to the
446: primary. Velocities are scaled to the virial velocity of the system,
447: distances to the virial radius. Circles denote ``associated''
448: satellites; i.e., those that have been {\it inside} the virial radius
449: of the primary at some earlier time. Crosses indicate satellites that
450: are on their first approach, and have never been inside $r_{\rm
451: vir}$. Filled circles indicate associated satellites whose apocentric
452: radii exceed their turnaround radius by at least $25\%$, indicating
453: that their orbital energies have been substantially altered during
454: their evolution. ``Crossed'' circles correspond to associated
455: satellites that have entered $r_{\rm vir}$ during the last $3$
456: Gyrs. The curves delineating the top and bottom boundaries of the
457: distribution show the escape velocity of an NFW halo with
458: concentration $c=10$ and $c=20$, respectively.
459: Note that there is one satellite ``escaping'' the system with positive
460: radial velocity. Solid lines show the trajectories in the $r-V_r$ plane
461: of the two ''ejected'' satellites shown in figure \ref{fig:orbesc}.
462: Filled squares correspond to the fourteen brightest Milky
463: Way satellites, taken from \citet{vandenbergh99} (complemented with NED data for
464: the Phoenix, Tucana and NGC6822), and plotted assuming that $V_{\rm vir}^{\rm
465: MW} \sim 109$ km/s and $r_{\rm vir}^{\rm MW}=237$ kpc (see Sales et al
466: 2007). Arrows indicate how the positions of MW satellites in this plot
467: would be altered if our estimate of $V_{\rm vir}^{\rm MW}$ (and,
468: consequently, $r_{\rm vir}^{\rm MW}$) is allowed to vary by $\pm 20\%$.}
469: \label{fig:rvMW}
470: \end{figure}
471: %\end{center}
472:
473: We highlight some of these objects in Figure~\ref{fig:rvMW}, using
474: ``filled'' circles to denote ``associated'' satellites whose
475: apocenters at $z=0$ exceed their turnaround radii by at least
476: $25\%$. Two such objects are worth noting in this figure: one of them
477: is the farthest ``associated'' satellite, found at more than $\sim 2.5
478: \, r_{\rm vir}$ from the primary; the second is an outward-moving
479: satellite just outside the virial radius but with radial velocity
480: approaching $\sim 2\, V_{\rm vir}$. The latter, in particular, is an
481: extraordinary object, since its radial velocity alone exceeds the
482: nominal escape velocity{\footnote{The notion of binding energy and
483: escape velocity is ill-defined in cosmology; note, for example, that
484: the {\it whole universe} may be considered formally bound to any
485: positive overdensity in an otherwise unperturbed Eistein-de Sitter
486: universe. We use here the nominal escape velocity of an NFW model
487: \citep{nfw96,nfw97} to guide the
488: interpretation. This profile fits reasonably well the mass
489: distribution of the primaries inside the virial radius, and has a
490: finite escape velocity despite its infinite mass. Certainly satellites
491: with velocities exceeding the NFW escape velocity are likely to move
492: far enough from the primary to be considered true {\it escapers}.}} at
493: that radius. This satellite is on a trajectory which, for all
494: practical purposes, will remove it from the vicinity of the primary
495: and leave it wandering through intergalactic space.
496:
497: The origin of these unusual objects becomes clear when inspecting
498: Figure~\ref{fig:orbesc}. The two satellites in question are shown with
499: {\it dashed lines} in this figure; each is a member of a bound {\it
500: pair} of satellites (the other member of the pair is shown with solid
501: lines of the same color). During first pericentric approach, the pair
502: is disrupted by the tidal field of the primary and, while one member
503: of the pair remains bound and follows the kind of ``conventional''
504: orbit described in \S~\ref{ssec:convorb}, the other one is ejected
505: from the system on an extreme orbit. The trajectories of these two
506: ``ejected'' satellites in the $r$-$V_r$ plane are shown by the wiggly
507: lines in Figure~\ref{fig:rvMW}.
508:
509: These three-body interactions typically involve the first pericentric
510: approach of a bound pair of accreted satellites and tend to eject the
511: lighter member of the pair: in the example of Figure~\ref{fig:orbesc},
512: the ``ejected'' member makes up, respectively, only $3\%$ and $6\%$ of
513: the total mass of the pair at the time of accretion. Other interaction
514: configurations leading to ejection are possible, such as an unrelated
515: satellite that approaches the system during the late stages of a
516: merger event, but they are rare, at least in our simulation series. We
517: emphasize that not all satellites that have gained energy during
518: accretion leave the system; most are just put on orbits of unusually
519: large apocenter but remain bound to the primary. This is shown by the
520: filled circles in Figure~\ref{fig:rvMW}; many affected satellites are
521: today completing their second or, for some, third orbit around the
522: primary.
523:
524: The ejection mechanism is perhaps best appreciated by inspecting the
525: orbital paths of the satellite pairs. These are shown in
526: Figure~\ref{fig:orbxyesc}, where the top (bottom) panels correspond to
527: the satellite pair accreted later (earlier) into the primary in
528: Figure~\ref{fig:orbesc}. Note that in both cases, as the pair
529: approaches pericenter, the lighter member (dashed lines) is also in
530: the process of approaching the pericenter of its own orbit around the
531: heavier member of the pair. This coincidence in orbital phase
532: combines the gravitational attraction of the two more massive members
533: of the trio of galaxies, leading to a substantial gain in orbital
534: energy by the lightest satellite, effectively ejecting it from the
535: system on an approximately radial orbit. The heavier member of the
536: infalling pair, on the other hand, decays onto a much more tightly
537: bound orbit.
538:
539: Figure~\ref{fig:orbxyesc} also illustrates the complexity of orbital
540: configurations that are possible during these three-body
541: interactions. Although the pair depicted in the top panels approaches
542: the primary as a cohesive unit, at pericenter each satellite circles
543: about the primary in opposite directions: in the $y$-$z$ projection
544: the heavier member circles the primary {\it clockwise} whereas the
545: ejected companion goes about it {\it counterclockwise}. After
546: pericenter, not only do the orbits of each satellite have different
547: period and energy, but they differ even in the {\it sign} of their
548: orbital angular momentum. In this case it would clearly be very
549: difficult to link the two satellites to a previously bound pair on the
550: basis of observations of their orbits after pericenter.
551:
552: Although not all ejections are as complex as the one illustrated in
553: the top panels of Figure~\ref{fig:orbxyesc}, it should be clear from
554: this figure that reconstructing the orbits of satellites that have
555: been through pericenter is extremely difficult, both for satellites
556: that are ejected as well as for those that remain bound. For example,
557: the massive member of the late-accreting pair in
558: Figure~\ref{fig:orbesc} sees its apocenter reduced by more than a
559: factor of $\sim 5$ from its turnaround value in a single pericentric
560: passage. Such dramatic variations in orbital energy are difficult to
561: reproduce with simple analytic treatments inspired on Chandrasekhar's
562: dynamical friction formula (Pe\~narrubia 2007, private communication).
563:
564: %%%%%%%%%%%%%%%%%%%%%%%%%%
565: \begin{center}
566: \begin{figure}
567: \includegraphics[width=84mm]{figs/letfig4.ps}
568: \caption{Distribution of the ratio between the apocentric radius of
569: satellites (measured at $z=0$) and their turnaround radius, defined as
570: the maximum distance to the primary before accretion. Note the
571: presence of two groups. Satellites on ``conventional'' orbits have
572: $r_{\rm apo}/r_{\rm ta}<1$, the rest have been catapulted into
573: high-energy orbits by three-body interactions during first
574: approach. The satellite marked with a rightward arrow is the
575: ``escaping'' satellite identified by a dot-centered circle in
576: Figure~\ref{fig:rvMW}; this system has nominally infinite $r_{\rm
577: apo}$. The dashed histogram highlights the population of low-mass
578: satellites; i.e., those with stellar masses at accretion time not
579: exceeding $2.6\%$ of the primary's final $M_{str}$. The satellite
580: marked with an arrow is a formal ``escaper'' for which $r_{\rm apo}$
581: cannot be computed.}
582: \label{fig:rhist}
583: \end{figure}
584: \end{center}
585: %%%%%%%%%%%%%%%%%%%%%%%%%%
586:
587: \section{Application to the Local Group}
588: \label{sec:LG}
589:
590: We may apply these results to the interpretation of kinematical
591: outliers within the satellite population around the Milky Way (MW) and
592: M31, the giant spirals in the Local Group. Although part of the
593: discussion that follows is slightly speculative due to lack of
594: suitable data on the three-dimensional orbits of nearby satellites, we
595: feel that it is important to highlight the role that the concomitant
596: accretion of multiple satellites may have played in shaping the
597: dynamics of the dwarf members in the Local Group.
598:
599: \subsection{Milky Way satellites}
600: \label{ssec:MW}
601:
602: The filled squares in Figure~\ref{fig:rvMW} show the galactocentric
603: radial velocity of thirteen bright satellites around the Milky Way and
604: compare them with the simulated satellite population. This comparison
605: requires a choice for the virial radius and virial velocity of the
606: Milky Way, which are observationally poorly constrained.
607:
608: We follow here the approach of Sales et al (2007), and use the
609: kinematics of the satellite population itself to set the parameters of
610: the Milky Way halo. These authors find that simulated satellites are
611: only mildly biased in velocity relative to the dominant dark matter
612: component: $\sigma_{\rm r} \sim 0.9 (\pm 0.2) V_{\rm vir}$, where
613: $\sigma_{\rm r}$ is the radial velocity dispersion of the satellite
614: population within $r_{vir}$. Using this, we find
615: $V_{\rm vir}^{\rm MW}=109 \pm 22$ km/s
616: and $r_{\rm vir}^{\rm MW}=237\pm 50$ kpc from the observed radial
617: velocity dispersion of $\sim 98$ km/s. This corresponds to $M_{\rm
618: vir}^{\rm MW}=7 \times 10^{11} M_{\odot}$, in reasonable agreement
619: with the $1$-$2 \times 10^{12} M_{\odot}$ estimate of \citet{klypin02}
620: and with the recent findings of \citet{smith06} based on
621: estimates of the escape velocity in the solar neighbourhood.
622:
623: Since Leo I dwarf has the largest radial velocity of the
624: Milky Way satellites, we have recomputed the radial velocity
625: dispersion excluding it from the sample. We have found that
626: $\sigma_r$ drops from 98 to 82 km/s when Leo I is not taken
627: into account changing our estimation of $V_{\rm vir}^{\rm MW}$ from
628: 109 to 91 km/s, still within the errors of the value previously found.
629: Given the recent rapid growth in the number of known Milky Way
630: satellite one would suspect that the velocity dispersion
631: will significantly increase if more Leo I-like
632: satellites are detected. However, we notice that given their
633: high velocities they are not expected to remain inside the
634: virial radius for a long time period hence not contributing
635: to the $\sigma_r$ computation.
636:
637: Figure~\ref{fig:rvMW} shows that, considering $V_{\rm vir}^{\rm MW}=109$ km/s,
638: the velocities and positions of all MW satellites are reasonably
639: consistent with the
640: simulated satellite population, with the possible exception of Leo I,
641: which is located near the virial radius and is moving outward with a
642: velocity clearly exceeding $V_{\rm vir}$. Indeed, for $V_{\rm
643: vir}^{\rm MW}=109$ km/s, Leo I lies right on the escape velocity curve
644: of an NFW profile with concentration parameter similar to those
645: measured in the simulations. This is clearly a kinematical outlier
646: reminiscent of the satellite expelled by three-body interactions
647: discussed in the previous subsection and identified by a dot-centered
648: circle in Figure~\ref{fig:rvMW}. This is the {\it only} ``associated''
649: satellite in our simulations with radial velocity exceeding $V_{\rm
650: vir}$ and located outside $r_{\rm vir}$.
651:
652: Could Leo I be a satellite that has been propelled into a
653: highly-energetic orbit through a three-body interaction? If so, there
654: are a number of generic predictions that might be possible to verify
655: observationally. One is that its orbit must be now basically radial in
656: the rest frame of the Galaxy, although it might be some time before
657: proper motion studies are able to falsify this prediction. A second
658: possibility is to try and identify the second member of the pair to
659: which it belonged. An outward moving satellite on a radial orbit takes
660: only $\sim 2$-$3$ Gyr to reach $r_{\rm vir}$ with escape
661: velocity. Coincidentally, this is about the time that the Magellanic
662: Clouds pair were last at pericenter, according to the traditional
663: orbital evolution of the Clouds \citep[see, e.g.,][]{gardiner96,vandermarel02}.
664:
665: Could Leo I have been a Magellanic Cloud satellite ejected from the
666: Galaxy a few Gyrs ago? Since most satellites that are ejected do so
667: during {\it first} pericentric approach, this would imply that the
668: Clouds were accreted only recently into the Galaxy, so that they
669: reached their first pericentric approach just a few Gyr ago. This is
670: certainly in the spirit of the re-analysis of the orbit of the Clouds
671: presented recently by \citealt{besla07} and based on new proper
672: motion measurements recently reported by \citet{kallivayalil06}. In
673: this regard, the orbit of the Clouds might resemble the orbit of the
674: companion of the ``escaping'' satellite located next to Leo I in
675: Figure~\ref{fig:rvMW}. The companion is fairly massive and, despite a
676: turnaround radius of almost $\sim 600$ kpc and a rather late accretion
677: time ($t_{\rm acc}=10.5$ Gyr, see Figure~\ref{fig:orbesc}), it is left
678: after pericenter on a tightly bound, short-period orbit resembling
679: that of the Clouds today \citep{gardiner96,vandermarel02}. To
680: compound the resemblance, this satellite has, at accretion time, a
681: total luminosity of order $\sim 10\%$ of that of the primary, again on
682: a par with the Clouds.
683:
684: We also note that an ejected satellite is likely to have picked up its
685: extra orbital energy through a rather close pericentric passage and
686: that this may have led to substantial tidal damage. This, indeed, has
687: been argued recently by \citealt{sohn06} on the basis of asymmetries
688: in the spatial and velocity distribution of Leo I giants (but see
689: \citealt{koch07} for a radically different interpretation).
690:
691:
692: On a final note, one should not forget to mention another (less
693: exciting!) explanation for Leo I: that our estimate of $V_{\rm
694: vir}^{\rm MW}$ is a substantial underestimate of the true virial
695: velocity of the Milky Way. The arrows in Figure~\ref{fig:rvMW}
696: indicate how the position of the MW satellites in this plane would
697: change if our estimate of $V_{\rm vir}^{\rm MW}$ is varied by $\pm
698: 20\%$. Increasing $V_{\rm vir}^{\rm MW}$ by $\sim 20\%$ or more would
699: make Leo I's kinematics less extreme, and closer to what would be
700: expected for a high-speed satellite completing its first orbit. This
701: rather more prosaic scenario certainly cannot be discounted on the
702: basis of available data (see, e.g., Zaritsky et al 1989, Kochanek
703: 1996, Wilkinson \& Evans 1999)
704:
705: %%%%%%%%%%%%%%%%%%%%%%%%%%
706: \begin{center}
707: \begin{figure}
708: \includegraphics[width=84mm]{figs/letfig5.ps}
709: \caption{As Figure~\ref{fig:rvMW} but for {\it line-of-sight}
710: velocities and {\it projected} distances. Three random orthogonal
711: projections have been chosen for each simulated satellite
712: system. Signs for $V_{\rm los}$ have been chosen so that it is
713: positive if the satellite is receding away from the primary in
714: projection, negative otherwise. The ``escaping'' satellite from
715: Figure~\ref{fig:rvMW} is shown by a starred symbol. Filled squares
716: correspond to the M31 satellites taken from \citealt{mcconnachieandirwin06},
717: plus And XIV \citep{majewski07} and And XII (Chapman et al 2007, submitted)
718: and assuming that $V_{\rm vir}^{\rm M31}\sim 138$ km/s and $r_{\rm
719: vir}^{\rm M31}=300$ kpc. Arrows indicate how the positions of M31
720: satellites in this plot would be altered if our estimate of $V_{\rm
721: vir}^{\rm M31}$ (and, consequently, $r_{\rm vir}^{\rm M31}$) is allowed
722: to vary by $20\%$.}
723: \label{fig:rvM31}
724: \end{figure}
725: \end{center}
726: %%%%%%%%%%%%%%%%%%%%%%%%%%
727: \begin{center}
728: \begin{figure}
729: \includegraphics[width=84mm]{figs/letfig6.ps}
730: \caption{Orbital paths for both pair of satellites shown in
731: Figure~\ref{fig:orbesc}. Upper (bottom) panel shows the pair that
732: accretes later (earlier) in that figure and shows the orbits in the
733: rest frame
734: of the primary. The coordinate system is chosen so that the angular
735: momentum of the primary is aligned with the $z$ axis. A solid curve
736: tracks the path of the heavier satellite; a dashed line follows the
737: satellite that is propelled into a highly energetic orbit after.}
738: \label{fig:orbxyesc}
739: \end{figure}
740: \end{center}
741: %%%%%%%%%%%%%%%%%%%%%%%%%%
742:
743:
744: \subsection{M31 satellites}
745: \label{ssec:M31}
746:
747: A similar analysis may be applied to M31 by using the projected
748: distances and line-of-sight velocities of simulated satellites, shown
749: in Figure~\ref{fig:rvM31}. Three orthogonal projections of the
750: simulated satellites are overlapped in this figure, with symbols as
751: defined in Figure~\ref{fig:rvMW}. Following the same approach as in
752: \S~\ref{ssec:MW}, we use the fact that the line-of-sight satellite
753: velocity dispersion is $\sigma_{\rm los} \sim 0.8 (\pm 0.2) \, V_{\rm
754: vir}$ in our simulations to guide our choice of virial velocity and
755: radius for M31; $V_{\rm vir}^{\rm M31}=138 \pm 35$ km/s and $r_{\rm
756: vir}^{\rm M31}=300 \pm 76$ kpc. (We obtain $\sigma_{\rm los}=111$ km/s
757: for all $17$ satellites within $300$ kpc of M31.) This compares
758: favourably with the $V_{\rm vir}^{\rm M31}\sim 120$ km/s estimate
759: recently obtained by \citealt{seigar06} under rather different
760: assumptions.
761:
762: With this choice, we show the $19$ satellites around M31 compiled by
763: McConnachie \& Irwin (2006), plus two recently-discovered satellites
764: for which positions and radial velocities have become available (And
765: XII, Chapman et al 2007, and And XIV, Majewski et al. 2007).
766: As in Figure~\ref{fig:rvMW}, arrows
767: indicate how the position of M31 satellites would change in this
768: figure if $V_{\rm vir}^{\rm M31}$ were allowed to vary by $\pm
769: 20\%$. We notice that the exclusion of And XII and And XIV (the
770: highest velocity satellites within 300 kpc from Andromeda) in the
771: $V_{\rm vir}^{\rm M31}$ estimation gives $\sim 100$ km/s, consistent
772: with the $V_{\rm vir}^{\rm M31}=138 \pm 35$ km/s previously found
773: considering all satellites.
774: Projected distances are as if viewed from infinity
775: along the direction joining the Milky Way with M31 and that the
776: {\it sign} of the line-of-sight velocity in Figure~\ref{fig:rvM31} is
777: chosen to be positive if the satellite is receding from the primary
778: (in projection) and negative otherwise.
779:
780: There are a few possible outliers in the distribution of M31 satellite
781: velocities: And XIV (Majewski et al 2007), the Pegasus dwarf irregular
782: (UGC 12613, \citealt{gallagher98}), And XII (Chapman et al 2007), and UGCA 092
783: (labelled U092 in Figure~\ref{fig:rvM31}, \citealt{mcconnachieandirwin06}).
784: And XIV and PegDIG
785: seem likely candidates for the three-body ``ejection'' mechanism
786: discussed above: they have large velocities for their position, and,
787: most importantly, they are receding from M31; a {\it requirement} for
788: an escaping satellite. Note, for example, that And XIV lies very close
789: to the ``escaping'' satellite (dot-centered symbol in
790: Figure~\ref{fig:rvM31}) paired to Leo I in the previous
791: subsection. Escapers should move radially away from the primary, and
792: they would be much harder to detect in projection as extreme velocity
793: objects, unless they are moving preferentially along the line of
794: sight. It is difficult to make this statement more conclusive without
795: further knowledge of the orbital paths of these satellites. Here, we
796: just note, in agreement with Majewski et al (2007), that whether And
797: XIV and PegDIG are dynamical ``rogues'' depends not only on the
798: (unknown) transverse velocity of these galaxies, but also on what is
799: assumed for M31's virial velocity. With our assumed $V_{\rm vir}^{\rm
800: M31}=138$ km/s, neither And XIV nor PegDIG look completely out of
801: place in Figure~\ref{fig:rvM31}; had we assumed the lower value of
802: $120$ km/s advocated by Seigar et al (2006) And XIV would be almost
803: on the NFW escape velocity curve, and would certainly be a true
804: outlier.
805:
806: High-velocity satellites {\it approaching} M31 in projection are
807: unlikely to be escapers, but rather satellites on their first
808: approach. This interpretation is probably the most appropriate for
809: And XII and UGCA 092. As discussed by Chapman et al (2007), And XII is
810: almost certainly {\it farther} than M31 but is approaching us at much
811: higher speed ($\sim 281$ km/s faster) than M31. This implies that And
812: XII is actually getting closer in projection to M31 (hence the
813: negative sign assigned to its $V_{\rm los}$ in
814: Figure~\ref{fig:rvM31}), making the interpretation of this satellite
815: as an escaping system rather unlikely.
816:
817: Note, again, that although And XII (and UGCA 092) are just
818: outside the loci delineated by simulated satellites in
819: Figure~\ref{fig:rvM31}, revising our assumption for $V_{\rm vir}^{\rm
820: M31}$ upward by $20\%$ or more would render the velocity of this
821: satellite rather less extreme, and would make it consistent with that
822: of a satellite on its first approach to M31. As was the case for Leo
823: I, this more prosaic interpretation of the data is certainly
824: consistent with available data.
825:
826: \section{SUMMARY and Conclusions}
827: \label{sec:conc}
828:
829: We examine the orbits of satellite galaxies in a series of
830: Nbody/gasdynamical simulations of the formation of $L_*$ galaxies in a
831: $\Lambda$CDM universe. Most satellites follow orbits roughly in accord
832: with the expectations of secondary infall-motivated models. Satellites
833: initially follow the universal expansion before being decelerated by
834: the gravitational pull of the main galaxy, turning around and
835: accreting onto the main galaxy. Their apocentric radii decrease
836: steadily afterwards as a result of the mixing associated with the
837: virialization process as well as of dynamical friction. At $z=0$ most
838: satellites associated with the primary are found within its virial
839: radius, and show little spatial or kinematic bias relative to the dark
840: matter component (see also Sales et al 2007).
841:
842: A number of satellites, however, are on rather unorthodox orbits, with
843: present apocentric radii exceeding their turnaround radii, at times
844: by a large factor. The apocenters of these satellites are typically
845: beyond the virial radius of the primary; one satellite is formally
846: ``unbound'', whereas another is on an extreme orbit and is found today
847: more than $2.5\, r_{\rm vir}$ away, or $\gsim \, 600$ Mpc when scaling
848: this result to the Milky Way.
849:
850: These satellites owe their extreme orbits to three-body interactions
851: during first approach: they are typically the lighter member of a pair
852: of satellites that is disrupted during their first encounter with the
853: primary. This process has affected a significant fraction of
854: satellites: a full one-third of the simulated satellite population
855: identified at $z=0$ have apocentric radii exceeding their turnaround
856: radii. These satellites make up the majority ($63\%$) of systems on
857: orbits that venture outside the virial radius.
858:
859: We speculate that some of the kinematical outliers in the Local Group
860: may have been affected by such process. In particular, Leo I might
861: have been ejected $2$-$3$ Gyr ago, perhaps as a result of interactions
862: with the Milky Way and the Magellanic Clouds. Other satellites on
863: extreme orbits in the Local Group may have originated from such
864: mechanism. Cetus \citep{lewis07} and Tucana \citep{oosterloo96}
865: ---two dwarf spheroidals in the periphery
866: of the Local Group---may owe their odd location (most dSphs are found
867: much closer to either M31 or the Galaxy) to such ejection mechanism.
868:
869: If this is correct, the most obvious culprits for such ejection events
870: are likely to be the largest satellites in the Local Group (M33 and
871: the LMC/SMC), implying that their possible role in shaping the
872: kinematics of the Local Group satellite population should be
873: recognized and properly assessed. In this regard, the presence of
874: kinematical oddities in the population of M31 satellites, such as the
875: fact that the majority of them lie on ``one side'' of M31 and seem to
876: be receding away from it (McConnachie \& Irwin 2006), suggest the
877: possibility that at least some of the satellites normally associated
878: with M31 might have actually been brought into the Local Group fairly
879: recently by M33. Note, for example, that two of the dynamical outliers
880: singled out in our discussion above (And XII and And XIV) are close to
881: each other in projection; have rather similar line-of-sight velocities
882: (in the heliocentric frame And XII is approaching us at $556$ km/s, And
883: XIV at $478$ km/s); and belong to a small subsystem of satellites
884: located fairly close to M33.
885:
886: The same mechanism might explain why the spatial distribution of at
887: least some satellites, both around M31 and the Milky Way, seem to
888: align themselves on a ``planar'' configuration \citep{majewski94,libeskind05,
889: kochandgrebel06}, as this may
890: just reflect the orbital accretion plane of a multiple system of
891: satellites accreted simultaneously in the recent past \citep{kroupa05,
892: metz07}.
893:
894: From the point of view of hierarchical galaxy formation models, it
895: would be rather unlikely for a galaxy as bright as M33 to form in
896: isolation and to accrete as a single entity onto M31. Therefore, the
897: task of finding out {\it which} satellites (rather than {\it whether}) have
898: been contributed by the lesser members of the Local Group, as well as
899: what dynamical consequences this may entail, should be undertaken
900: seriously, especially now, as new surveys begin to bridge our incomplete
901: knowledge of the faint satellites orbiting our own backyard.
902:
903: \section*{Acknowledgements}
904: \label{acknowledgements}
905:
906:
907: LVS and MGA are grateful for the hospitality of the Max-Planck
908: Institute for Astrophysics in Garching, Germany, where much of the
909: work reported here was carried out. LVS thanks financial support from
910: the Exchange of Astronomers Programme of the IAU and to the ALFA-LENAC
911: network. JFN acknowledges support from Canada's NSERC, from the
912: Leverhulme Trust, and from the Alexander von Humboldt Foundation, as
913: well as useful discussions with Simon White, Alan McConnachie, and
914: Jorge Pe\~narrubia. MS acknowledges support by the German Science
915: foundation (DFG) under Grant STE 710/4-1. We thank Scott Chapman and
916: collaborators for sharing their results on Andromeda XII in advance of
917: publication. We also acknowledge a very useful report from an anonymous
918: referee that helped to improve the first version.
919:
920: \bibliography{references}
921:
922: \end{document}
923: