1: %\documentclass[12pt,preprint]{aastex}
2:
3: %% manuscript produces a one-column, double-spaced document:
4: % \documentclass[manuscript]{aastex}
5:
6: %% preprint2 produces a double-column, single-spaced document:
7: % \documentclass[preprint2]{aastex}
8:
9: \documentstyle[emulateapj,apjfonts,psfig]{article}
10:
11: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
12: % Edit history
13: % July 2003 - Knebe first draft
14: % 29/07/2003 - Lewis first edit + comments
15: % 30/07/2003 - the man himself (implementing Brad's comments, too)
16: % 28/08/2003 - the man himself (first revision)
17: % 30/10/2003 - implementing referee report 02-02
18: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
19: \newcommand{\vdag}{(v)^\dagger}
20: \newcommand{\myemail}{skywalker@galaxy.far.far.away}
21: \newcommand{\apm}{AP$^3$M}
22: \newcommand{\LCDM}{$\Lambda$CDM}
23: \newcommand{\LWDM}{$\Lambda$WDM}
24: \newcommand{\hkpc}{$h^{-1}{\ }{\rm kpc}$}
25: \newcommand{\hMpc}{$h^{-1}{\ }{\rm Mpc}$}
26: \newcommand{\hMsun}{$h^{-1}{\ }{\rm M_{\odot}}$}
27: \newcommand{\kms}{${\rm{\ }km{\ }s^{-1}}$}
28: \newcommand{\nbody}{$N$-body}
29: \newcommand{\Rvir}{$R_{\rm vir}$}
30: \newcommand{\Mvir}{$M_{\rm vir}$}
31: \newcommand{\rtrunc}{$r_{\rm trunc}$}
32: \newcommand{\rtidal}{$r_{\rm tidal}$}
33: \newcommand{\zform}{$z_{\rm form}$}
34: \newcommand{\Lsat}{$\vec{L}_{\rm sat}$}
35: \newcommand{\Lhost}{$\vec{L}_{\rm host}$}
36: \newcommand{\Ehost}{$\vec{E}_{1, \rm host}$}
37: \newcommand{\Rapo}{$\vec{R}_{\rm sat}^{\rm apo}$}
38: \newcommand{\Lapo}{$\vec{L}_{\rm sat}^{\rm apo}$}
39: \newcommand{\vecx}{\vec{x}}
40: \newcommand{\vecp}{\vec{p}}
41: \newcommand{\vecr}{\vec{r}}
42: \newcommand{\vecF}{\vec{F}}
43: \newcommand{\vecg}{\vec{g}}
44: \newcommand{\ddx}{\ddot{\vec{x}}}
45: \newcommand{\ddr}{\ddot{{r}}}
46: \newcommand{\dx}{\dot{\vec{x}}}
47: \newcommand{\vecnabla}{\vec{\nabla}}
48: \newcommand{\dvecx}{\dot{\vec{x}}}
49: \newcommand{\ddvecx}{\ddot{\vec{x}}}
50: \newcommand{\ddvecr}{\ddot{\vec{r}}}
51: \newcommand{\REF}{\textbf{REFS}}
52: \newcommand{\Table}[1]{Table~\ref{#1}}
53: \newcommand{\Sec}[1]{Section~\ref{#1}}
54: \newcommand{\Eq}[1]{Eq.~(\ref{#1})}
55: \newcommand{\Fig}[1]{Fig.~\ref{#1}}
56: \newcommand{\mlapm}{\texttt{MLAPM}}
57: \newcommand{\mhf}{\texttt{MHF}}
58: \newcommand{\mht}{\texttt{MHT}}
59: \newcommand{\ea}{et~al.~} % \ea = et al.
60: \newcommand{\AN}[3] {\mbox{AN~\textbf{#1},~#2~(#3)}}
61: \newcommand{\AAA}[3] {\mbox{A\&A~\textbf{#1},~#2~(#3)}}
62: \newcommand{\AAR}[3] {\mbox{A\&A~Rev.~\textbf{#1},~#2~(#3)}}
63: \newcommand{\AAS}[3] {\mbox{A\&A~Suppl.~\textbf{#1},~#2~(#3)}}
64: \newcommand{\ApJ}[3] {\mbox{ApJ~\textbf{#1},~#2~(#3)}}
65: \newcommand{\ApJS}[3] {\mbox{ApJ~Suppl.~\textbf{#1},~#2~(#3)}}
66: \newcommand{\ApJL}[3] {\mbox{ApJ~Lett.~\textbf{#1},~#2~(#3)}}
67: \newcommand{\ARAA}[3] {\mbox{Ann.~Rev.~A~\&~A~\textbf{#1},~#2~(#3)}}
68: \newcommand{\AJ}[3] {\mbox{Astron.~J.~\textbf{#1},~#2~(#3)}}
69: \newcommand{\MNRAS}[3] {\mbox{MNRAS~\textbf{#1},~#2~(#3)}}
70: \newcommand{\Nature}[3] {\mbox{Nature~\textbf{#1},~#2~(#3)}}
71: \newcommand{\NewA}[3] {\mbox{NewA~\textbf{#1},~#2~(#3)}}
72: \newcommand{\Science}[3]{\mbox{Science~\textbf{#1},~#2~(#3)}}
73: \newcommand{\PhRevL}[3] {\mbox{Phys.~Rev.~Lett.~\textbf{#1},~#2~(#3)}}
74: \newcommand{\PhRevD}[3] {\mbox{Phys.~Rev.~\textbf{D#1},~#2~(#3)}}
75: \newcommand{\astroph}[1]{\mbox{\texttt{astro-ph/#1}}}
76:
77: \def\Rod#1{\noindent{\bf[$\heartsuit$ #1]}}
78: \def\GFL#1{\noindent{\bf[$\flat$ #1]}}
79:
80: \begin{document}
81:
82: %\shorttitle{Anisotropy in the Distribution of Satellite Galaxy Orbits}
83: %\shortauthors{Knebe et al.}
84:
85: \title{Anisotropy in the Distribution of Satellite Galaxy Orbits}
86:
87: \author{Alexander Knebe\altaffilmark{1},
88: Stuart P.D. Gill\altaffilmark{1},
89: Brad K. Gibson\altaffilmark{1},
90: Geraint F. Lewis\altaffilmark{2},
91: Rodrigo A. Ibata\altaffilmark{3}, and
92: Michael A. Dopita\altaffilmark{4}}
93:
94: \altaffiltext{1}{Centre for Astrophysics~\& Supercomputing,
95: Swinburne University,Mail\#31, PO Box 218,
96: Hawthorn, VIC 3122, Australia.}
97: \altaffiltext{2}{Institute of Astronomy, School of Physics, A29,
98: University of Sydney, NSW 2006, Australia}
99: \altaffiltext{3}{Observatoire de Strasbourg, 11, Rue de l'Universite,
100: F-6700, Strasbourg, France}
101: \altaffiltext{4}{Research School of Astronomy and Astrophysics,
102: Australian National University, Weston Creek Post Office,
103: ACT 2611, Australia}
104:
105:
106: \begin{abstract}
107: Nearby clusters such as Virgo and Coma possess galaxy distributions
108: which tend to be aligned with the principal axis of the cluster
109: itself. This has also been confirmed by a recent statistical analysis
110: of some 300 Abell clusters where the effect has been linked to the
111: dynamical state of the cluster. Moreover, the orbits of satellite
112: galaxies in galactic systems like our own Milky Way also demonstrate a
113: high degree of anisotropy -- the so-called Holmberg effect, the origin
114: of which has been the subject of debate for more than 30 years. This
115: study presents the analysis of cosmological simulations focusing on
116: the orbits of satellite galaxies within dark matter halos. The
117: apocentres of the orbits of these satellites are preferentially found
118: within a cone of opening angle $\sim$40$^\circ$ around the major axis
119: of the host halo, in accordance with the observed anisotropy found in
120: galaxy clusters. We do, however, note that a link to the dynamical
121: age of the cluster is not well established as both our oldest dark
122: matter halos do show a clear anisotropy signal. Further analysis
123: connects this distribution to the infall pattern of satellites along
124: the filaments: the orbits are determined rather by the environment of
125: the host halo than some "dynamical selection" during their life within
126: the host's virial radius.
127: \end{abstract}
128:
129: \keywords{galaxies: formation --- cosmology: theory --- methods: numerical}
130:
131: \section{Introduction} \label{observations}
132: %%%%%%%%%%%%%%%%%%%%%%%%%%
133: Observations of the distribution of bright elliptical galaxies within
134: the Virgo galaxy cluster show a remarkable collinear arrangement
135: (West~\& Blakeslee 2000). Moreover, this axis also appears to be part
136: of a filamentary bridge connecting Virgo and the rich cluster Abell
137: 1367. This phenomenon has already been recognized (Arp 1968;
138: Bingegeli, Tammann~\& Sandage 1987) but only with accurate
139: measurements of distance did it become possible to discriminate
140: between this being a genuine three-dimensional structure or merely a
141: chance alignment of galaxies. West~\& Blakeslee (2000) based their
142: distances upon the surface brightness fluctuations method
143: (cf. Blakeslee, Ajhar~\& Tonry 1999 for a recent review), concluding
144: that not only were the brightest cluster members distributed
145: anisotropically over the cluster, but the distribution of dwarf
146: ellipticals was elongated in the same direction (Bingelli 1999).
147: Plionis~\ea (2003) investigated this substructure-cluster correlation
148: statistically using 303 Abell clusters. They affirm that there is
149: indeed such a signal and also that this signal is related to the
150: dynamical state (and the environment) of the cluster.
151:
152: Moreover, even on smaller scales -- in galactic systems -- there are
153: clear observational indications that the distribution of the orbits of
154: satellite galaxies is biased towards the galactic pole (Holmberg 1969)
155: giving an anisotropic distribution. Zaritsky~\ea (1997) found that
156: satellites of (isolated) disk galaxies are scattered asymmetrically
157: about the parent galaxy and aligned preferentially with the disk minor
158: axis. At first this signal was only prominent for distances out to
159: $\sim$50~kpc from the host (Holmberg 1969), but an analysis based on a
160: much larger sample (cf. Zaritsky~\ea 1993) of satellites extends it
161: to 200~kpc (Zaritsky~\ea 1997). This result is confirmed by a study
162: of the satellites orbiting M31 (Hartwick 2000; Grebel, Kolatt~\&
163: Brandner 1999). In addition, for the one galaxy where individual
164: satellite orbits are known, the Milky Way, there is also evidence that
165: that the orbits are preferentially polar (Zaritsky~\& Gonzalez 1999).
166: This is derived from information based upon the alignment of
167: satellites on the sky (Kunkel~\& Demers 1976; Lynden-Bell 1982), the
168: orientation of the Magellanic Stream (Mathewson, Clearly~\& Murray
169: 1974), the three-dimensional distribution of satellites (Majewski
170: 1994; Hartwick 1996), and their actual velocities (Scholz~\& Irwin
171: 1994).
172:
173: All these observations clearly indicate that on scales spanning from
174: galaxy clusters down to galaxies there is a signal indicating a
175: correlation between the alignment of substructure and the shape of the
176: gravitational potential this substructure moves in. However, the
177: source of this alignment of satellites has been debated for a number
178: of years, with two potential solutions suggested to explain this
179: puzzling arrangement. If the distribution of infalling satellites is
180: initially spherical, then dynamical selection may preferentially
181: destroy or suppress the star-formation in those on more equatorial
182: orbits\footnote{With \textit{equatorial} we mean "perpendicular to the
183: major axis"}. For instance, Pe\~narubbia, Kroupa \& Boily (2002) have
184: argued that in flattened dark matter halos, the timescale for
185: dynamical friction is substantially longer for satellites that are on
186: polar orbits. Alternatively, the present non-isotropic distribution
187: of satellite systems may reflect the fact that it has always been
188: non-isotropic, with the satellites being accreted along preferred
189: directions. These hypotheses can be tested in numerical simulations
190: of structure formation. However, in previous attempts to decipher
191: this signal from cosmological \nbody\ simulations there still remains
192: a certain amount of uncertainty (Zaritsky~\ea 1997). The first fully
193: self-consistent simulations targeting the subject were performed by
194: Tormen (1997). They though lacked resolution in time, space and mass
195: to perform a detailed analysis of the satellite population for
196: differing environments, and hence more refined simulations are
197: required. Tormen (1997) could not follow the satellite distribution
198: within the host's virial radius but rather tracked all progenitors
199: prior to accretion. This allowed him to analyze the infall pattern
200: rather than the orbital evolution of the satellites. The aim of this
201: study is to investigate this subject with the latest state-of-the-art
202: high-resolution \nbody-simulations. We focus upon a detailed analysis
203: of the temporal and spatial properties of satellite galaxies residing
204: within host galaxy clusters that formed fully self-consistently within
205: a cosmological framework.
206:
207:
208: \begin{deluxetable}{ccccccccc}
209: \tablecaption{Summary of the eight host dark matter halos. \label{DMhost}}
210: \tablewidth{0pt}
211: \tablehead{
212: \colhead{Halo} &
213: \colhead{\Rvir [\hkpc]} &
214: \colhead{$v_{\rm circ}^{\rm max}$[\kms]} &
215: \colhead{\Mvir [$10^{14}$\hMsun]} &
216: \colhead{\zform} &
217: \colhead{age [Gyr]} &
218: \colhead{$N_{\rm sat}(<\!r_{\rm vir})$} &
219: \colhead{$T$}
220: }
221: \startdata
222: \# 1 & 1349 & 1125 & 2.87 & 1.16 & 8.30 & 158 & 0.67 \\
223: \# 2 & 1069 & 894 & 1.42 & 0.96 & 7.55 & 63 & 0.87 \\
224: \# 3 & 1081 & 875 & 1.48 & 0.87 & 7.16 & 87 & 0.83 \\
225: \# 4 & 980 & 805 & 1.10 & 0.85 & 7.07 & 57 & 0.77 \\
226: \# 5 & 1356 & 1119 & 2.91 & 0.65 & 6.01 & 175 & 0.65 \\
227: \# 6 & 1055 & 833 & 1.37 & 0.65 & 6.01 & 85 & 0.92 \\
228: \# 7 & 1014 & 800 & 1.21 & 0.43 & 4.52 & 59 & 0.89 \\
229: \# 8 & 1384 & 1041 & 3.08 & 0.30 & 3.42 & 251 & 0.90 \\
230: \enddata
231:
232: \end{deluxetable}
233:
234: \section{The Simulations} \label{simulations}
235: %%%%%%%%%%%%%%%%%%%%%%%%%
236: There is mounting evidence that, despite its problems, the Cold Dark
237: Matter structure formation scenario provides the most accurate
238: description of our Universe. Observations point towards the standard
239: \LCDM\ according to which the universe is comprised of about 28\% dark
240: matter, 68\% dark energy, and luminous baryonic matter (i.e. galaxies,
241: stars, gas, and dust) at a mere 4\% level (cf. Spergel~\ea 2003). This
242: so-called "concordance model" induces hierarchical structure formation
243: in which small objects form first and subsequently merge to form
244: larger objects.
245:
246: The analysis presented in this study is based upon a suite of eight
247: high-resolution \nbody\ simulations within the \LCDM\ concordance
248: cosmology. They were carried out using the publicly available
249: adaptive mesh refinement code \mlapm\ (Knebe, Green~\& Binney 2001).
250: \mlapm\ reaches high force resolution by refining all high-density
251: regions with an automated refinement algorithm. The refinements are
252: recursive: the refined regions can also be refined, each subsequent
253: refinement having cells that are half the size of the cells in the
254: previous level. This creates an hierarchy of refinement meshes of
255: different resolutions covering regions of interest. The refinement is
256: done cell-by-cell (individual cells can be refined or de-refined) and
257: meshes are not constrained to have a rectangular (or any other)
258: shape. The criterion for (de-)refining a cell is simply the number of
259: particles within that cell and a detailed study of the appropriate
260: choice for this number as well as more details about the particulars
261: of the code can be found elsewhere (Knebe, Green~\& Binney, 2001).
262:
263: Each run focuses on the formation and evolution of one particular dark
264: matter halo containing of order greater one million particles. The
265: force resolution reached by \mlapm\ in the high-density regions is
266: 2\hkpc\ corresponding to roughly 0.25\% of the host's virial radii
267: \Rvir. Following Lacey~\& Cole (1994) we define the formation redshift
268: \zform\ as the time when the halo contains half of its mass today.
269: The host halos were carefully chosen to sample a variety of different
270: triaxialities and merger histories as summarized in
271: \Table{DMhost}. The virial radii given in that Table correspond to the
272: point where the density of the host (measured in terms of the
273: cosmological background density $\rho_b$) drops below $\Delta_{\rm
274: vir}=340$ with \Mvir\ being the mass enclosed by that sphere. All
275: satellite galaxies orbiting in and around the host halo are identified
276: using the \mlapm-based halo finder described in a companion paper
277: (Gill, Knebe~\& Gibson 2004). This new halo finder is based upon the
278: \mlapm\ grid structure and works at the same resolution level as the
279: \nbody\ code itself thus assuring not to miss or include any objects
280: not resolved in the actual simulation. The satellites are then
281: individually traced from \zform\ onwards until redshift $z=0$. The
282: mass spectrum of those satellites accounted for in the following
283: analysis can be described by a declining power-law $dn/dM\propto
284: M^{-\alpha}$ with $\alpha \approx 1.7-1.9$ in the range from
285: $2\times10^{10}$\hMsun\ (applied mass-cut corresponding to 100
286: particles and explaining the rather 'low' number for $N_{\rm
287: sat}(<\!r_{\rm vir})$ in \Table{DMhost}) up to
288: $\sim10^{13}$\hMsun. The total fraction of mass locked up in
289: satellites never exceeds $\leq$15\% though.
290:
291: The high temporal sampling of the outputs (i.e. $\Delta t \approx 0.2$
292: Gyrs) allows us to accurately measure the orbital parameters of the
293: satellite galaxies and a detailed analysis of their disruption and
294: survival history will be presented elsewhere (Gill, Knebe~\& Gibson
295: 2004). Gill~\ea show that the distribution of orbital eccentricities
296: for the satellite population peaks about the value $\epsilon \approx
297: 0.6$ where $\epsilon = 1-p/a$ is defined using the last pericenter $p$
298: and apocenter $a$ of its orbit. The pericenter distribution itself
299: peaks about 35\% of \Rvir\ for all host halos and most of the
300: satellites (i.e. more than 75\%) had had one full orbit with the
301: maximum number of orbits found to be 3--5 depending on the host
302: halo. The angular momentum vector of the satellite \Lapo=\Rapo$\times
303: \vec{V}_{\rm sat}^{\rm apo}$ was calculated at its last apocenter
304: passage \Rapo.
305:
306: In order to probe the alignment with the shape and orientation of the
307: host halo we calculated the eigenvectors $\vec{E}_{\rm 1,2,3}$ (with
308: $\vec{E}_{\rm 1}$ being the major axis) of its inertia tensor
309: using only the "core" region as defined by the $6^{\rm th}$
310: refinement level in \mlapm, i.e. the boundary of this refinement level
311: is an isodensity contour. According to the refinement criterion
312: adopted in the simulations the $6^{\rm th}$ level surrounds material
313: about 3000 times denser than $\rho_b$ or in other words 9 times denser
314: than the material at the virial radius. The host halos are well
315: described by the density profiles advocated by Navarro, Frenk~\& White
316: (1997) with concentration in the range $c = 5-7$. Therefore,
317: a density of roughly $9\times \rho(r_{\rm vir})$ corresponds to about
318: the half-mass radius of the host. The eigenvectors now define the
319: orientation of the host halo and the coordinate system used to measure
320: the orbits of the satellites, respectively. Moreover, its eigenvalues
321: $a>b>c$ can be used to construct the triaxiality parameter
322: $T=(a^2-b^2)/(a^2-c^2)$ (Franx, Illingworth~\& Zeeuw 1991) presented
323: in \Table{DMhost}, too.
324:
325: The left panel of \Fig{NdotE} presents the (normalized) distribution
326: of angles between \Rapo\ and \Ehost. This graph shows that there is a
327: clear trend in at least six of the eight halos for the two vectors to
328: be aligned (the distribution peaks at $0^\circ$ and $180^\circ$,
329: respectively) meaning that the orbits of the satellites are
330: preferentially found along the major axis of the host. The right
331: panel of \Fig{NdotE} shows the (normalized) distribution of angles
332: between the angular momentum vector \Lapo\ of the satellite at its
333: last apocenter and \Ehost, clearly revealing evidence for these two
334: vectors to be parallel (distribution peaks at $90^\circ$). We like to
335: stress that only satellite galaxies that at least had one or more
336: complete orbits were taken into account in \Fig{NdotE}; the figure is
337: \textit{not} based upon the infall pattern of satellites as
338: investigated by, for instance, Tormen (1997). We also need to stress
339: that the distributions presented in that Figure are normalized,
340: i.e. they are corrected for the bias introduced by plotting them as a
341: function of the angle $\theta$ rather than $\cos\theta$. The data is
342: binned equally spaced in angle ranging from 0$^\circ$ to 180$^\circ$,
343: and therefore the area probed on the sphere varies with
344: $\theta$. Therefore the distribution needed to be normalized by the
345: the respective area $A$ specified by the actual range of angles
346: [$\theta-\Delta\theta/2,\theta+\Delta\theta/2$]. This area is
347: proportional to $A \propto \cos(\theta-\Delta\theta/2)-\cos(\theta+\Delta\theta/2)$.
348:
349:
350:
351: \section{Conclusions and Discussion} \label{conclusions}
352: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
353: There are two possible scenarios that explain the substructure-cluster
354: alignment found in observations of galaxy clusters (Plionis~\ea 2003;
355: West~\& Blakeslee 2000) and the analysis of \nbody-simulations
356: presented in this study. Either some dynamical process can be held
357: responsible, meaning that initially the orbits were randomly
358: distributed and only some satellites survive to give rise to the
359: observed correlation in \Fig{NdotE}, or alternatively, their orbital
360: parameters are imprinted upon them at the time they enter the host
361: halo as already pointed out by Tormen (1997).
362:
363: ``Dynamical destruction'', whereby satellites on more
364: equatorial orbits suffer more tidal disruption, can be rejected because a
365: comparable (although not as pronounced) alignment signal is found when
366: restricting the analysis to those satellites that are in fact
367: disrupted: We placed mass-less tracer particles at the last centre of
368: the satellites before they were classified "disrupted". These
369: ``disrupted systems'' are presented as the thin lines in \Fig{NdotE}.
370:
371: We therefore conclude that it must be the initial distribution of the
372: satellite systems that is responsible for the present day alignment.
373: For example, galaxies are "funneling along the filaments"
374: (Kitzbichler~\& Saurer 2003) which has also been confirmed by X-ray
375: observations of the spatial distribution of substructure in galaxy
376: clusters (West, Jones~\& Forman 1995) and biases the types of possible
377: orbits. It would now be reassuring to confirm a link between the host
378: halo shape and the surrounding environment, i.e. are the filaments
379: that feed the halo with material preferentially angled with respect to
380: the orientation of the host halo? When projecting the positions of
381: the satellites onto a sphere at the time they enter the virial radius
382: of the host it is clear that they are not randomly distributed; they
383: cluster in directions linked to the filamentary structure surrounding
384: the host halo, as found before observationally (e.g. West~\&
385: Blakeslee 2003; Plionis~\& Basilikos 2002, and references therein) and
386: in \nbody-simulations (e.g. Faltenbacher 2002; Hatton~\& Ninin 2001;
387: Onuora~\& Thomas 2000; Colberg~\ea 1999; Splinter~\ea 1997; Tormen
388: 1997). Tormen (1997) already pointed out that there is a strong
389: alignment between the distribution of infalling satellite galaxies and
390: the shape of the dark matter host. However, Tormen's simulations did
391: not have the spatial and mass resolution to investigate the orbits of
392: satellites \textit{within} the host halo. We are, for the first time,
393: tracing extremely well-resolved dynamics of the satellites within the
394: dark matter hosts confirming that the alignment for both, disrupted
395: and surviving, satellites is maintained for several orbital
396: periods. \Fig{SatelliteOrbits} now presents a visual impression of the
397: trajectories of all surviving satellites identified at formation time
398: of host halo~\#3 until redshift $z=0$, qualitatively supporting the
399: scenario that these orbits are linked to the filamentary structure
400: which will leave its imprint in observed anisotropy.
401:
402: Plionis~\ea (2003) also established a link between the alignment and
403: the dynamical age of the clusters. Our simulations though show that
404: this finding can not be generalized: both our oldest halos, in
405: which the satellites had as many as 4--5 orbits, do show the
406: correlation signal. We do not observe a trend for a "randomization"
407: of the orbits in older halos. The satellites actually preserve the
408: alignment with the host they had when they first fell into the
409: cluster.
410:
411: In \Sec{observations} we showed that the (observational)
412: substructure-alignment signal spans from galaxy clusters down to
413: galactic scales. Our results are mostly applicable to the former, and
414: any extrapolation to smaller scales has to be handled with care. In
415: the case for galaxies, for instance, the interaction between the
416: galactic disk and the incoming satellite could enhance destruction in
417: the galaxy plane, i.e. satellite on prograde orbits decay faster than
418: the ones on retrograde (or polar) orbits due to orbital resonances
419: between the disk and the satellites (Pe\~narubbia, Kroupa \& Boily
420: 2002). Based on theoretical and numerical studies (cf. Lacey~\& Cole
421: 1993 and Moore~\ea 1999) it should be possible to re-scale our
422: simulations to a Milky Way sized object by requiring that the maximum
423: of the circular velocity curves of our halos equals 220~\kms. The
424: scaling factor lies in the range of about 4--5 (cf.~$v^{\rm max}_{\rm
425: circ}$ values in \Table{DMhost}) and hence our ``rescaled dark matter
426: halos'' would correspond to the ones in the observational data with
427: virial radii in the range 260--295\hkpc. This scaling factor in length
428: entails a scaling in mass of 64--125 (simply the length scale to the
429: power of three). This brings the mass of our galaxy clusters down to
430: $\sim 10^{12}$\hMsun\ which agrees with the dark matter mass inferred
431: for our Milky Way (Freeman 1996). More problematic, however, are the
432: ages of our systems: our halos are only $\leq$8.3~Gyrs old opposed to
433: 12~Gyrs for the Milky Way. Satellites in the Milky Way had the chance
434: to complete nearly twice as many orbits leaving more space for an
435: explanation based upon the dynamical destruction scenario.
436:
437: When re-scaling our data, putting aside the age issue, and trying to
438: explain the Holmberg effect, another uncertainty comes into play: the
439: orientation of the stellar disk. In a triaxial potential there are in
440: general stable closed orbits about both the major and the minor axis
441: (e.g., Binney \& Tremaine 1987). So, in principal the disk plane could
442: either be perpendicular to the major or perpendicular to the minor
443: axis. Under the assumption it lies perpendicular to the major axis of
444: the dark matter halo the satellites in our simulations will be on
445: polar orbits. There are indications that this configuration results in
446: the most stable disk configuration within a triaxial halo (Hayashi~\ea
447: 2003). Moreover, even though there are clear indications that the
448: angular momentum of the dark matter is well aligned with the minor
449: axis of the halo (e.g. Warren~\ea 1992), van den Bosch~\ea (2002)
450: showed that the angular momentum of the baryonic component (i.e. gas)
451: not necessarily follows that of the dark matter distribution. They
452: found an average misalignment between $\vec{L}_{\rm gas}$ and
453: $\vec{L}_{\rm DM}$ of the order of 40$^\circ$ in their numerical
454: simulations. Hence the orientation of the galactic disk with respects
455: to the dark matter halo is not well determined. Turning to
456: observations does not resolve this question, with studies of polar
457: rings indicating strongly oblate halos (i.e. Iodice~\ea 2003), while
458: others suggest dark matter halos are more spherical, or even oblate
459: (Olling \& Merrifield 2000; Ibata~\ea 2001). The situation is not at
460: all clear, but the assumption that disks are perpendicular to the
461: major axis (which is in agreement with results presented by
462: Hayashi~\ea 2003) would provide an explanation for the Holmberg
463: effect, even though this is a very speculative interpretation.
464:
465:
466: \acknowledgments
467: %%%%%%%%%%%%%%%%
468: The simulations presented in this paper were carried out on the
469: Beowulf cluster at the Centre for Astrophysics~\& Supercomputing,
470: Swinburne University. The support of the Australian Research Council
471: and the Swinburne Research Development Grants Scheme is gratefully
472: acknowledged. GFL thanks Suede for ``Introducing the band''.
473:
474:
475: \begin{thebibliography}{}
476: %%%%%%%%%%%%%%%%%%%%%%%%%
477:
478: \bibitem[binney87]{binney87}
479: {Binney J., Tremaine S., 1987, {\em Galactic Dynamics}
480: (Princeton: Princeton University Press)}
481:
482: \bibitem[Bullock et al.(2000)]{bullock}
483: Bullock, J., Kravtsov, A., Weinberg, D. 2000, {\apj} 539, 517
484:
485: \bibitem[colberg99]{colberg99}
486: {Colberg J.M., White S.D.M., Jenkins A., Pearce F.R., 1999,
487: \mnras, 308, 593}
488:
489: \bibitem[cote01]{cote01}
490: {Cote P.~\ea, 2001, \apj, 559, 828}
491:
492: \bibitem[dalal02]{dalal02}
493: {Dalal N., Kochanek C.S., 2002, \apj, 572, 25}
494:
495: \bibitem[evans03]{evans03}
496: {Evans N.W., Witt H.J., 2003, {\it astro-ph/0212013}}
497:
498: \bibitem[franx91]{franx91}
499: {Franx M., Illingworth G., de Zeeuw T., 1991, \apj, 383,112}
500:
501: \bibitem[freeman96]{freeman96}
502: {Freeman K.C., 1996, in ASP Conf. Proc. 92,
503: Formation of the Galactic Halo - Inside and Out,
504: ed. H. Morrison \& A. Sarajedini (San Francisco: ASP), p.3}
505:
506: \bibitem[gill04]{gill04}
507: {Gill S., Knebe A., Gibson B.K., 2004, in preparation}
508:
509: \bibitem[grebel99]{grebel99}
510: {Grebel E.K., Kolatt T., Brandner W., 1999,
511: in \textit{The stellar content of the Local Group}, IAU 192,
512: eds. P. Whitelock \& R. Cannon, San Francisco: ASP Conf. Series}
513:
514: \bibitem[hayashi03]{hayashi03}
515: {Hayashi E.~\ea, 2003, in preparation}
516:
517: \bibitem[holmber69]{holmberg69}
518: {Holmberg E., 1969, Ark. Astron., 5, 305}
519:
520: \bibitem[hatton01]{hatton01}
521: {Hatton S., Ninin S., 2001, \mnras, 322, 576}
522:
523: \bibitem[Ibata et al.(2001)]{2001ApJ...551..294I}
524: {Ibata R., Lewis G.~F., Irwin M., Totten E., \& Quinn T.,
525: 2001, \apj, 551, 294}
526:
527: \bibitem[Iodice et al.(2003)]{2003ApJ...585..730I}
528: {Iodice E., Arnaboldi M., Bournaud F., Combes F., Sparke L.~S.,
529: van Driel W., \& Capaccioli M.\ 2003, \apj, 585, 730}
530:
531: \bibitem[kitz03]{kitz03}
532: {Kitzbichler M.G., Saurer W., 2003, \apj, 590, L9}
533:
534: \bibitem[klypin99]{klypin99}
535: {Klypin A.A., Kravtsov A.V., Valenzuela O., Prada F.,
536: 1999, \apj, 522, 82}
537:
538: \bibitem[knebe01]{knebe01}
539: {Knebe A., Green A., Binney J.J., 2001, \mnras, 325, 845}
540:
541: \bibitem[kunkel76]{kunkel76}
542: {Kunkel W.E., Demers S., 1976, Roy. Green. Obs. Bull., 182, 241}
543:
544: \bibitem[lacey93]{lacey93}
545: {Lacey C., Cole S., 1993, \mnras, 262, 627}
546:
547: \bibitem[lacey94]{lacey94}
548: {Lacey C., Cole S., 1994, \mnras, 271, 676}
549:
550: \bibitem[lynden82]{lynden82}
551: {Lynden-Bell D., 1982, Observatory, 102, 202}
552:
553: \bibitem[majewski94]{majewski94}
554: {Majewski S.R., 1994, \apjl, 431, L17}
555:
556: \bibitem[matthewson74]{matthewson74}
557: {Matthewson D.S., Clearly M.N., Murray J.D., 1974, \apj, 190, 291}
558:
559: \bibitem[moore99]{moore99}
560: {Moore B., Ghigna S., Governato F., Lake G., Quinn T., Stadel J.,
561: Tozzi P., 1999, \apjl, 524, L19}
562:
563: \bibitem[Olling \& Merrifield(2000)]{2000MNRAS.311..361O}
564: {Olling, R.~P.~\& Merrifield, M.~R.\ 2000, \mnras, 311, 361}
565:
566: \bibitem[Pe\~narubbia, Kroupa \& Boily(2002)]{penarubbia}
567: {Pe\~narubbia, J., Kroupa, P., Boily, C. 2002, \mnras, 333, 779}
568:
569: \bibitem[plionis02]{plionis02}
570: {Plionis M., Basilikos S., 2002, \mnras, 329, L47}
571:
572: \bibitem[scholz94]{scholz94}
573: {Scholz R.-D., Irwin M.J., 1994, in IAU 161, \textit{Wide Field Imaging},
574: eds. H.T. MacGillivray}
575:
576: \bibitem[Schechter \& Wambsganss(2002)]{2002ApJ...580..685S}
577: {Schechter P.~L.~\& Wambsganss J.\ 2002, \apj, 580, 685}
578:
579: \bibitem[spergel03]{spergel}
580: {Spergel D.N., Steinhardt P.J., 2000, Phys. Rev. Lett., {\bf 84}, 3760}
581:
582: \bibitem[Somerville(2002)]{somerville}
583: {Somerville, R. 2002, \apj, 572, 23}
584:
585: \bibitem[tormen97]{tormen97}
586: {Tormen G., 1997, \mnras, 290, 411}
587:
588: \bibitem[Tully et al.(2002)]{tully}
589: {Tully, B., Somerville, R., Trentham, N., Verheijen, M. 2002,
590: \apj, 569, 573}
591:
592: \bibitem[warren92]{warren92}
593: {Warren M.S., Quinn P.J., Salmon J.K., Zurek W.H., 1992, \apj, 399, 405}
594:
595: \bibitem[west95]{west95}
596: {West M.J., Jones C., Forman W., \apj, 451, L5}
597:
598: \bibitem[west00]{west00}
599: {West M.J., Blakeslee J.P., 2000, \apj, 543, L27}
600:
601: \bibitem[zaritsky93]{zaritsky93}
602: {Zaritsky D., Smith R., Frenk C.S., White S.D.M., 1993, \apj, 405, 464}
603:
604: \bibitem[zaritsky97]{zaritsky97}
605: {Zaritsky D., Smith R., Frenk C.S., White S.D.M., 1997, \apjl, 478, L53}
606:
607: \bibitem[zaritsky99]{zaritsky99}
608: {Zaritsky D., Gonzalez A.H., 1999, PASP, 111, 1508}
609:
610: \end{thebibliography}
611:
612:
613: \begin{figure}
614: \plottwo{f1a.eps}{f1b.eps}
615: \caption{The left panel shows the (normalized) distribution of angles between
616: position vector \Rapo\ of the satellites measured at the last
617: apocentre and the major axis \Ehost\ of the halo. $\theta$
618: values of 0$^\circ$ and 180$^\circ$ indicate alignment with
619: the the principal axis of the host. The right panel shows
620: the (normalized) distribution of angles between angular momentum vector
621: \Lapo\ of satellites measured at the last apocentre and the
622: major axis \Ehost\ of the halo. A $\theta$ value of
623: 90$^\circ$ means that the orbital plane of the satellite
624: contains the major axis of the host where $\theta$ of
625: 0$^\circ$ and 180$^\circ$ would require for the plane to be
626: perpendicular to the host's principal axis. The thin
627: histograms are for disrupted satellites. In both cases only
628: satellites more massive than $2\times 10^{10}$\hMsun\
629: (100 particles) that had one full orbit were taken into account.
630: \label{NdotE}}
631: \end{figure}
632:
633: \begin{figure}
634: \plotone{f2.eps}
635: \caption{Orbits of all satellite galaxies identified at formation time (dark)
636: to redshift $z=0$ (light). The left panel shows how the host halo
637: fits into the surrounding large-scale structure presented at formation
638: time. The right panel zooms into the region marked in the left panel,
639: this time not showing low-resolution particles. The shape (and position)
640: of the underlying host halo at redshift $z=0$ is indicated by the
641: best-fit ellipse to the grid used to calculate its triaxiality.
642: The vertical line extending to top edge of panel indicates the
643: principal axis of the host. With the assumptions presented in this
644: paper, the pole of the host galaxy within the dark matter halo is
645: aligned with this principle axis.
646: \label{SatelliteOrbits}}
647: \end{figure}
648:
649: %\clearpage
650:
651:
652: %% The following command ends your manuscript. LaTeX will ignore any text
653: %% that appears after it.
654:
655: \end{document}
656:
657: %%
658: %% End of file `sample.tex'.
659: