1: \documentclass[]{emulateapj}
2: %\documentclass[12pt,preprint]{aastex}
3: \usepackage{apjfonts,epsfig}
4: %\usepackage{epsfig}
5: %\newcounter{fred}
6:
7: \def\fsub{\mathrel{f_{\rm sub}}}
8: \def\tninety{\mathrel{t_{90}}}
9: \def\tfifty{\mathrel{t_{50}}}
10: \def\Ndm{\mathrel{N_{\rm DM}}}
11: \def\lx{\mathrel{L_X}}
12: \def\ls{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}}
13: \def\gs{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}
14: \def\Mtot{\mathrel{M_{\rm tot}}}
15: \def\Mcen{\mathrel{M_{\rm cen}}}
16: \def\ergs{\mathrel{\rm erg\,s^{-1}}}
17: \def\kms{\mathrel{\rm km\,s^{-1}}}
18: \def\Msol{\mathrel{\rm M_{\odot}}}
19: \def\hkpc{\mathrel{h^{-1}{\rm kpc}}}
20:
21: \lefthead{Smith \& Taylor}
22: \righthead{Cluster Substructure and Assembly Histories}
23:
24: \slugcomment{Received 2008 April 25; Accepted 2008 June 24}
25:
26: \begin{document}
27: %\input{definitions}
28:
29: \title{Connecting Substructure in Galaxy Cluster Cores at z=0.2 with
30: Cluster Assembly Histories}
31:
32: \author{Graham P.\ Smith\altaffilmark{1,3} and
33: James E.\ Taylor\altaffilmark{2,3}}
34:
35: \altaffiltext{1}{School of Physics and Astronomy, University of
36: Birmingham, Edgbaston, Birmingham, B15 2TT, England;
37: gps@star.sr.bham.ac.uk}
38: \altaffiltext{2}{Department of Physics and Astronomy, University of
39: Waterloo, Waterloo, Ontario, N2L 3G1, Canada;
40: taylor@sciborg.uwaterloo.ca}
41: \altaffiltext{3}{California Institute of Technology, Mail Code 105-24,
42: 1200 E.\ California Blvd., Pasadena, CA 91125}
43:
44: %\setcounter{footnote}{3}
45:
46: \begin{abstract}
47: We use semi-analytic models of structure formation to interpret
48: gravitational lensing measurements of substructure in galaxy cluster
49: cores ($R{\le}250{\hkpc}$) at $z=0.2$. The dynamic range of the
50: lensing-based substructure fraction measurements is well matched to
51: the theoretical predictions, both spanning $\fsub\sim0.05-0.65$. The
52: structure formation model predicts that $\fsub$ is correlated with
53: cluster assembly history. We use simple fitting formulae to
54: parameterize the predicted correlations:
55: $\Delta_{90}=\tau_{90}+\alpha_{90}\log(\fsub)$ and
56: $\Delta_{50}=\tau_{50}+\alpha_{50}\log(\fsub)$, where $\Delta_{90}$
57: and $\Delta_{50}$ are the predicted lookback times from $z=0.2$ to
58: when each theoretical cluster had acquired 90\% and 50\% respectively
59: of the mass it had at $z=0.2$. The best-fit parameter values are:
60: $\alpha_{90}=(-1.34\pm0.79)\,{\rm Gyr}$,
61: $\tau_{90}=(0.31\pm0.56)\,{\rm Gyr}$ and
62: $\alpha_{50}=(-2.77\pm1.66)\,{\rm Gyr}$,
63: $\tau_{50}=(0.99\pm1.18)\,{\rm Gyr}$. Therefore (i) observed clusters
64: with $\fsub\ls0.1$ (e.g.\ A\,383, A\,1835) are interpreted, on
65: average, to have formed at $z\gs0.8$ and to have suffered $\le10\%$
66: mass growth since $z\simeq0.4$, (ii) observed clusters with
67: $\fsub\gs0.4$ (e.g.\ A\,68, A\,773) are interpreted as, on average,
68: forming since $z\simeq0.4$ and suffering $>10\%$ mass growth in the
69: $\sim500\,{\rm Myr}$ preceding $z=0.2$, i.e.\ since $z=0.25$. In
70: summary, observational measurements of $\fsub$ can be combined with
71: structure formation models to estimate the age and assembly history of
72: observed clusters. The ability to ``age-date'' approximately clusters
73: in this way has numerous applications to the large clusters samples
74: that are becoming available.
75: \end{abstract}
76:
77:
78: \keywords{gravitational lensing --- cosmology:dark matter ---
79: galaxies:clusters:individual (A\,68, A\,209, A\,267, A\,383, A\,773,
80: A\,963, A\,1763, A\,1835, A\,2218, A\,2219)}
81:
82: \section{Introduction}\label{sec:intro}
83:
84: The mass growth of clusters is sensitive to the dark energy equation
85: of state parameter $w$, the matter density of the universe $\Omega_M$
86: and the normalization of the matter power spectrum $\sigma_8$ (e.g.\
87: \citealt{Evrard93,Smith03,Mantz07}). Clusters are inferred to grow
88: hierarchically via the ingestion of smaller dark matter halos (that
89: host galaxies) into the more massive parent halo (cluster). The
90: structure of galaxy clusters, specifically the internal substructure
91: of clusters, therefore contains a wealth of cosmological information,
92: including possible clues about the physics of the dark matter particle
93: itself (e.g.\ \citealt{Natarajan02b}). From an astrophysical point of
94: view, the mass growth of clusters brings new (generally gas rich)
95: galaxy populations into clusters (e.g.\ \citeauthor{Moran07}
96: \citeyear{Moran07}) and may lead to shock-heating of the intracluster
97: medium and/or disruption of cooling in cluster cores (e.g.\
98: \citealt{Poole08}). Reliable measurement and intepretation of cluster
99: substructure is therefore of broad interest.
100:
101: The most direct way to detect substructure within clusters is via
102: gravitational lensing. Group-scale substructures within individual
103: clusters were detected in early ground-based strong-lensing studies of
104: individual clusters \citep{Pello91,Kneib93,Kneib95} and subsequently
105: measured to high precision using \emph{Hubble Space Telescope (HST)}
106: data \citep{Kneib96}. \citeauthor{Smith05a} (\citeyear{Smith05a} --
107: hereafter Sm05 -- see \S\ref{sec:obs}) then measured the structure of
108: a sample of 10 clusters at $z{\simeq}0.2$. In this article we use
109: \citeauthor{Taylor04}'s (\citeyear{Taylor04} -- hereafter TB04)
110: semi-analytical models of structure formation to interpret Sm05's
111: cluster substructure measurements, as a means of exploring
112: lensing-based substructure measurements as a quantitative probe of
113: cluster age and assembly history.
114:
115: We summarize Sm05 and TB04 in \S\ref{sec:obs} and \S\ref{sec:theory}
116: respectively, and then synthesize observations and theory in
117: \S\ref{sec:results}. We discuss caveats in \S\ref{sec:discuss} and
118: summarize our conclusions and discuss future prospects in
119: \S\ref{sec:conc}. We assume $H_0{=}70{\rm km\,s^{-1}Mpc^{-1}}$,
120: ${\Omega}_{\rm M}{=}0.3$, ${\Omega}_{\rm \Lambda}{=}0.7$ and
121: $\sigma_8=0.9$ throughout. The lookback time from $z=0$ to $z=0.2$ is
122: $t_{z=0.2}=2.44\,{\rm Gyr}$ in this cosmology.
123:
124: \section{Summary of Observational Results}\label{sec:obs}
125:
126: \begin{table*}
127: \footnotesize
128: \caption{
129: Observational Substructure Measurements\label{tab:sample}
130: }
131: \begin{center}
132: \begin{tabular}{lccccllcc}
133: \hline
134: \hline
135: \noalign{\vspace{1mm}}
136: Cluster & Redshift & ~~~~~~~~$\Mtot$ & ${\Ndm}$ & ${\fsub}$ & Strong Lensing? & Sm05 Classification & ~~~~~~~~${\Delta_{50}}$ & ${\Delta_{90}}$ \cr
137: & & ~~~~~~~~($10^{14}M_\odot$) & & & & & ~~~~~~~~$({\rm Gyr})$ & $({\rm Gyr})$ \cr
138: \noalign{\vspace{1mm}}
139: \hline
140: \noalign{\vspace{1mm}}
141: A\,383 & 0.188 & ~~~~~~~~$2.6\pm0.1$ & 1 & $0.06{\pm}0.01$ & Confirmed & Undisturbed uni-modal & ~~~~~~~~$4.3{\pm}2.4$ & $2.0{\pm}1.1$ \cr
142: A\,963 & 0.206 & ~~~~~~~~$2.4\pm0.2$ & 1 & $0.06{\pm}0.01$ & Confirmed & Undisturbed uni-modal & ~~~~~~~~$4.3{\pm}2.4$ & $2.0{\pm}1.1$ \cr
143: A\,1835 & 0.253 & ~~~~~~~~$4.1\pm1.1$ & 1 & $0.06{\pm}0.01$ & Unconfirmed & Undisturbed uni-modal & ~~~~~~~~$4.3{\pm}2.4$ & $2.0{\pm}1.1$ \cr
144: A\,267 & 0.230 & ~~~~~~~~$1.9\pm0.4$ & 1 & $0.08{\pm}0.01$ & Unconfirmed & Disturbed uni-modal & ~~~~~~~~$4.0{\pm}2.2$ & $1.8{\pm}1.0$ \cr
145: A\,1763 & 0.288 & ~~~~~~~~$1.5\pm0.8$ & 1 & $0.15{\pm}0.05$ & No? & Disturbed uni-modal & ~~~~~~~~$3.2{\pm}1.8$ & $1.4{\pm}0.9$ \cr
146: A\,209 & 0.209 & ~~~~~~~~$1.1\pm0.5$ & 1 & $0.18{\pm}0.06$ & No? & Disturbed uni-modal & ~~~~~~~~$3.0{\pm}1.7$ & $1.3{\pm}0.8$ \cr
147: A\,2219 & 0.228 & ~~~~~~~~$2.4\pm0.1$ & 3 & $0.20{\pm}0.01$ & Confirmed & Disturbed multi-modal & ~~~~~~~~$2.9{\pm}1.7$ & $1.3{\pm}0.8$ \cr
148: A\,2218 & 0.171 & ~~~~~~~~$4.0\pm0.1$ & 2 & $0.27{\pm}0.01$ & Confirmed & Disturbed multi-modal & ~~~~~~~~$2.5{\pm}1.5$ & $1.1{\pm}0.7$ \cr
149: A\,68 & 0.255 & ~~~~~~~~$3.1\pm0.1$ & 2 & $0.36{\pm}0.01$ & Confirmed & Disturbed multi-modal & ~~~~~~~~$2.2{\pm}1.4$ & $0.9{\pm}0.7$ \cr
150: A\,773 & 0.217 & ~~~~~~~~$3.6\pm1.2$ & 3 & $0.61{\pm}0.20$ & Unconfirmed & Disturbed multi-modal & ~~~~~~~~$1.6{\pm}1.3$ & $0.6{\pm}0.6$ \cr
151: \hline
152: \end{tabular}
153: \end{center}
154: \end{table*}
155:
156: Sm05 investigated the projected mass and structure of ten X-ray
157: luminous ($L_X\ge4\times10^{44}\,{\rm erg\,s^{-1}}$, $0.1-2.4\,{\rm
158: keV}$) cluster cores at $0.17\le z\le0.25$
159: (Table~\ref{tab:sample}). \emph{Hubble Space Telescope (HST)}/WFPC2
160: imaging and ground-based spectroscopy of gravitational arcs
161: \citep{Smith01a,Smith02b,Sand05,Richard07}, were used to characterize
162: the strong and weak gravitational lensing signal of each cluster core.
163: The lensing signals were then used to constrain a detailed
164: parametrized model of the projected mass distribution in each cluster
165: core following \citeauthor{Kneib93Th} (\citeyear{Kneib93Th} -- see
166: also \citealt{Kneib96} and \citealt{Smith02Th}). Each lens model
167: includes mass components that account for both the underlying dark
168: matter distribution in the cluster (cluster/group-scale mass
169: components) and the cluster galaxies down to
170: $L_K{\ge}0.1L^\star_K$. For clarity, we refer to the main central
171: cluster dark matter halo as the cluster-scale mass component, and all
172: other massive substructures associated with infallen clusters and/or
173: groups as group-scale mass components. The typical number of
174: galaxy-scale mass components in each lens model was 30.
175:
176: The \emph{HST} data probe out to a typical projected cluster-centric
177: radius of $R=250\hkpc$ at the cluster redshifts. Sm05 measured the
178: cluster substructure within this radius. Here we define $\fsub$ as the
179: fraction of mass associated with substructures, $\fsub{\equiv}M_{\rm
180: sub}/M_{\rm tot}=1-M_{\rm cen}/M_{\rm tot}$, where $M_{\rm cen}/M_{\rm
181: tot}$ is the central mass fraction used by Sm05. For this purpose,
182: ``substructures'' include group-scale masses and galaxy-scale masses, with
183: the exception of the brightest cluster galaxy (BCG). This is because BCG
184: mass is degenerate with cluster-scale mass in the lens models, and, in any
185: event, Sm05 found no evidence for BCGs being offset from the center of the
186: cluster-scale mass components. The clusters are listed in order of
187: increasing $\fsub$ in Table~\ref{tab:sample}, together with $\Ndm$, the
188: number of cluster/group-scale mass components in each lens model. As
189: expected, $\Ndm$ and $\fsub$ are correlated, however there is a factor of
190: 3 spread in $\fsub$ for sub-samples of clusters with $\Ndm=1$ and
191: $\Ndm>1$. Note, that to aide comparison of Sm05's lens modeling results
192: with predictions, the values of $\fsub$ in Table~\ref{tab:sample} have
193: been adjusted upwards to take account of the fact that the ``L''-shaped
194: observed WFPC2 field of view ($\sim5\,{\rm arcmin}^2$) covers just
195: $\sim50\%$ of a circle of radius $250\hkpc$ ($\sim10\,{\rm arcmin}^2$).
196: The WFPC2 observations were originally designed to include the likely
197: group-scale substructures in the cluster cores. The adjustments to
198: $\fsub$ therefore account statistically for the galaxy-scale masses not
199: included in Sm05's analysis, leaving $\Ndm$ unchanged.
200:
201: Cluster-to-cluster differences in $\fsub$ arise for two reasons: (i)
202: group-scale mass components in the lens models that are associated
203: with massive, likely in-falling structures such as groups of galaxies,
204: and (ii) cluster galaxies that are associated both with the central
205: cluster dark matter halo (and are presumably virialized) and with the
206: in-falling structures. Sm05 broadly interpreted measurements of
207: $\fsub$ as indicating the merger history of clusters, however
208: quantitative conclusions on cluster assembly and age were impossible
209: without theoretical models.
210:
211: \section{Summary of Theoretical Models}\label{sec:theory}
212:
213: TB04's semi-analytic model of halo evolution provides a fast way of
214: generating a large number of model halos for comparison with
215: observations, and agrees well in its preditions with self-consistent
216: n-body simulations \citep{Taylor05b}. The model includes two main
217: components, a merger-tree code for determining the assembly history of
218: an individual cluster, and an analytic description of how the main
219: halo and the merging subcomponent evolve after each merger. We will
220: summarize these components here, and refer the reader to TB04 for full
221: details.
222:
223: The merger tree describing the assembly of a single cluster is
224: generated randomly using the algorithm of
225: \cite{Somerville99}. Starting from a halo of specified mass at $z=0$,
226: the algorithm chooses an interval to step back in redshift and picks
227: progenitors at this redshift following extended Press-Schechter merger
228: statistics \citep{LaceyCole}. Iterating produces a complete history of
229: the mergers through which the final object assembled, down to some
230: limiting mass resolution and back to some redshift. In this article we
231: use a set of 1000 merger trees whose final masses at $z=0$ are
232: randomly drawn from the massive end ($>5\times10^{14}M_\odot$) of a
233: halo mass function. For the range of concentration parameters derived
234: for our halos (as explained below), this produces projected masses
235: within a cluster-centric radius of $R=250\hkpc$ (integrating out to
236: the virial radius along the line of sight) in the range
237: $0.4-8\times10^{14}M_{\odot}$, similar to those in the observed sample
238: (Table~\ref{tab:sample}). The merger trees for these systems are
239: followed back to a redshift of $50$, or until they drop below the
240: resolution limit of $10^{-4}\times$ the final mass.
241:
242: Given a merger tree, the model of TB04 selects the most massive
243: progenitor in the most recent merger and traces its history back,
244: selecting the most massive progenitor at each time step. This object
245: is considered the ``main'' system, and is modelled as a spherical halo
246: with a radial density profile similar to the ``universal'' profile
247: found in simulations \citep{Navarro97}. To account for some of the
248: centrally concentrated (baryonic) mass in the cluster, we use the
249: fitting formula proposed by \cite{Moore98}, which has a steeper
250: $r^{-1.5}$ central cusp. In an update to the model presented in TB04,
251: in this work we use the whole mass assembly history of the system,
252: rather than just its instantaneous mass, to derive the value of its
253: concentration parameter at each redshift step, following
254: \cite{Wechsler02}. Each merging subhalo is added to this main system
255: as a smaller spherical object with a Moore profile, and its subsequent
256: orbit and mass-loss history are determined using the analytic
257: description of dynamical friction, tidal stripping and tidal heating
258: in \cite{Taylor01}. Subhalos are tracked until they are disrupted
259: either by repeated tidal stripping or by passing within $0.01R_{vir}$
260: of the centre of the main potential. The model also includes simple
261: treatments of sub-substructure (subhalos within subhalos), correlated
262: orbits in infalling groups of objects, and collisions and encounters
263: between objects. See TB04 for a detailed explanation of these
264: elements.
265:
266: For each merger tree, we use the semi-analytic model to determine how
267: much substructure exists at $z=0.2$, and calculate $\fsub$ as the
268: fraction of the
269: mass projected within $250\hkpc$ of the halo centre that is contained
270: in the 30 most massive subhalos in the system (to match the
271: observations -- see \S\ref{sec:obs}). We exclude from the final
272: analysis any substructure which ends up within 5\% of $R_{vir}$ of the
273: halo centre, since in a real cluster most such objects would interact
274: strongly and merge rapidly with the dominant central galaxy. We
275: also record several indicators of the mass assembly history of the main
276: system in each merger tree, including $\tninety$ and $\tfifty$, the
277: lookback times by which it had assembled 90\% and 50\% respectively of
278: the mass it has at redshift $0.2$. These thresholds are chosen to
279: probe the recent infall history and cluster age respectively. We show
280: the resulting $\fsub-\tninety$ and $\fsub-\tfifty$ distributions in
281: Fig.~\ref{fig:epoch}. The observed clusters span a dynamic range of
282: $0.06{\le}{\fsub}{\le}0.62$ (Table~\ref{tab:sample}), which is
283: comparable with that of the TB04 theoretical models, and consistent
284: with other theoretical estimates of cluster substructure (e.g.\
285: \citealt{Diemand04,Gao04,Nagai05,vandenbosch05}).
286:
287:
288: \section{Results}\label{sec:results}
289:
290: \begin{figure}
291: \centerline{
292: \psfig{file=f1.ps,width=50mm,angle=-90}
293: }
294: \caption{Predicted distribution of substructure fraction versus
295: $\tninety$ (left) and $\tfifty$ (right). Note that the lookback
296: time from the $z=0$ to $z=0.2$ is $2.4\,{\rm Gyr}$ (dashed line).
297: The grey open points show clusters from TB04's model -- they span a
298: similar dynamic range to the observed clusters (see
299: Table~\ref{tab:sample}). The big black points show the mean
300: prediction in equally spaced logarithmic bins; the error bars show
301: the $1\sigma$ scatter in each bin. The solid black line shows the
302: best fit relations described in the text (\S\ref{sec:results})
303: \label{fig:epoch} }
304: \end{figure}
305:
306: Fig.~\ref{fig:epoch} reveals that the TB04 model predicts that $\fsub$
307: and $\tninety$ are correlated, as are $\fsub$ and $\tfifty$. However,
308: at $\fsub<0.2$ the distributions of ${\tninety}$ and $\tfifty$ are
309: bi-modal -- some theoretical clusters exhibiting both a low
310: substructure fraction and recent significant infall. The unphysical
311: location of these clusters in Fig.~1 is due to a timing difference
312: between the axes -- $\tninety$ is sensitive to mass growth within the
313: virialized region of the cluster (approximately a sphere of diameter
314: $\sim3\,{\rm Mpc}$); $\fsub$ is sensitive to substructures in a
315: cylinder of diameter $500\hkpc$ through the center of the cluster. If
316: an infalling structure has crossed the virial radius but not yet
317: entered the cylinder, then it may appear in the ``spike''of clusters
318: at $\fsub<0.2$ and $\tninety<3\,{\rm Gyr}$ (or $\tfifty<3\,{\rm
319: Gyr}$). We exclude these clusters from the calculations described
320: below.
321:
322: We have therefore confirmed and considerably extended Sm05's
323: qualitative interpretation of the lensing results: clusters with high
324: substructure fractions have (i) suffered more pronounced recent infall
325: than clusters with low substructure fractions, and (ii) formed at
326: later times than clusters with lower substructure fractions. To
327: quantify this, we fit the following simple formulae to the theoretical
328: data:
329: $\Delta_{90}=\tninety-t_{z=0.2}=\tau_{90}+\alpha_{90}\log(\fsub)$ and
330: $\Delta_{50}=\tfifty-t_{z=0.2}=\tau_{50}+\alpha_{50}\log(\fsub)$,
331: where $\alpha_{90}$ and $\alpha_{50}$ parameterize the dependence of
332: $\Delta_{90}$ and $\Delta_{50}$ respectively on $\fsub$; $\tau_{90}$
333: and $\tau_{50}$ are the intercepts at $\fsub=1$. We first bin up the
334: individual theoretical data points in bins of width
335: $\Delta\log(\fsub)=0.2$, and then weight each bin in the fit by the
336: reciprocal of the sample variance in the respective bin. The best fit
337: parameter values obtained in this way are:
338: $\alpha_{90}=(-1.34\pm0.79)\,{\rm Gyr}$,
339: $\tau_{90}=(0.31\pm0.56)\,{\rm Gyr}$ and
340: $\alpha_{50}=(-2.77\pm1.66)\,{\rm Gyr}$,
341: $\tau_{50}=(0.99\pm1.18)\,{\rm Gyr}$. The $\tau_{90}$ and $\tau_{50}$
342: parameters are measurements of the timing difference between the axes
343: in Fig.~\ref{fig:epoch} discussed above, and are well-matched to the
344: infall timescale of $\sim0.5-1\,{\rm Gyr}$.
345:
346: The best-fit models were then used to interpret quantitatively the
347: observed substructure fractions from Sm05 listed in
348: Table~\ref{tab:sample} -- i.e.\ to estimate the age ($\Delta_{50}$)
349: and recent infall history ($\Delta_{90}$) of each cluster -- see
350: Table~\ref{tab:sample}. The uncertainties on $\Delta_{90}$ and
351: $\Delta_{50}$ quoted in Table~\ref{tab:sample} incorporate errors on
352: $\fsub$ and on the best-fit parameter values derived above, and are
353: dominated by the scatter on $\fsub-\tfifty$ and $\fsub-\tninety$. The
354: typical uncertainty on cluster age is $\sim1.7\,{\rm Gyr}$, with ages
355: spanning $\sim1-4\,{\rm Gyr}$ . Cluster with the largest substructure
356: fractions, i.e.\ $\fsub\gs0.4$ formed within the $\sim2$Gyr preceding
357: $z=0.2$, i.e.\ since $z\ls0.4$, and had not assembled 90\% of the mass
358: they had at $z=0.2$ until $z\simeq0.25$, i.e.\ just $\sim500$Myr
359: before $z=0.2$. In contrast, clusters with the lowest subtructure
360: fractions, i.e.\ $\fsub<0.1$ formed $\gs4\,{\rm Gyr}$ before $z=0.2$,
361: i.e.\ at $z\gs0.8$. and then went on to assemble $>90\%$ of their
362: $z=0.2$ mass within the ensuing $\sim2\,{\rm Gyr}$, and suffered
363: negligible mass growth in the $\sim2\,{\rm Gyr}$ prior to $z=0.2$,
364: i.e.\ since $z\simeq0.4$. Therefore we interpret clusters with the
365: lowest $\fsub$ as being, on average, almost fully ($\sim90\%$)
366: assembled by the time that the clusters with highest $\fsub$ had
367: barely formed (i.e.\ assembled half of their mass).
368:
369: \section{Caveats}\label{sec:discuss}
370:
371: TB04 make some simplifying assumptions that may affect our results:
372: (i) dark matter halos are assumed to be spherical, whereas real dark
373: matter halos are likely triaxial; (ii) all matter is treated as
374: collisonless, thus ignoring baryonic physics including adiabatic
375: contraction due to gas cooling and the dynamical effects of galaxies;
376: (iii) matter outside the cluster virial radius is ignored; in
377: contrast, lensing is sensitive to all the matter along the line of
378: sight, including chance foreground and background projections. We will
379: consider these complications in more detail in future work.
380:
381: The main systematic uncertainty on the observed substructure fractions
382: is the completeness of Sm05's lens models as a function of sub-halo
383: mass. A detailed study of this issue will be published in the future
384: (Hamilton-Morris et al., in prep.). Here we identify how many of
385: Sm05's models may suffer from incompleteness. Incompleteness is most
386: likely to arise from features in the dark matter distribution not
387: identified in the luminous properties of the clusters within
388: $R<250\hkpc$, including group-scale dark matter halos in which the
389: cluster galaxies may be embedded. Strong lensing has been detected in
390: eight of the ten observed clusters (Table~\ref{tab:sample});
391: group-scale dark matter halos are detected in three of the eight by
392: the measurable way they alter the appearance of strongly-lensed arcs
393: compared to simpler mass distributions. Clusters in which
394: strong-lensing has not been detected are therefore of greatest
395: concern, as pointed out in Sm05's discussion of the filament feeding
396: A\,1763; see also \citeauthor{Mercurio03}'s (\citeyear{Mercurio03})
397: discussion of the complex dynamical structure of A\,209. Gross
398: substructure incompleteness may therefore be a concern for $\sim20\%$
399: of the observed sample.
400:
401: \begin{figure}
402: \centerline{\psfig{file=f2.ps,width=40mm,angle=-90}}
403: \caption{ Distribution of substructure fractions from the observed sample
404: (black solid line) and from the theoretical model (red dashed).
405: The observed clusters show a possible excess of very low and very high
406: substructure systems -- this is discussed in \S\ref{sec:discuss}.
407: \label{fig:fsub}}
408: \end{figure}
409:
410: We also consider whether there is any evidence in the respective
411: $\fsub$ distributions for systematic differences between Sm05's X-ray
412: selected sample and TB04's mass-selected sample. To address this we
413: compare the cumulative and differential $\fsub$ distributions in
414: Fig.~\ref{fig:fsub}. A KS test obtains a probability of 20\% that the
415: observed and theoretical distributions are drawn from the same parent
416: distribution, i.e.\ formally evidence of systematic differences
417: between the samples at $1.3\sigma$ significance. The small observed
418: sample size of just 10 clusters is clearly a limiting factor here.
419: The right panel of Fig.~\ref{fig:fsub} helps somewhat, as it
420: highlights more clearly the \emph{possible} bias of the observed
421: sample to very high and very low values of $\fsub$. Such a bias would
422: be plausible because cluster-cluster mergers and cool cores both
423: suffer an excess of X-ray flux at a fixed mass over a non-merging,
424: non-cool core cluster. A much larger observational sample is required
425: to investigate this issue further.
426:
427: \section{Conclusions and Future Prospects}\label{sec:conc}
428:
429: We have combined theoretical models of structure formation (TB04) with
430: gravitational lens models of galaxy clusters (Sm05) to explore how
431: measurements of cluster substructure from lensing observations can be
432: interpreted in the context of the age and assembly history of
433: clusters. The main result is that $\fsub$, the fraction of cluster
434: mass within a projected cluster-centric radius of $R=250\hkpc$
435: associated with substructure (galaxies and group-scale halos), as can
436: be measured from lensing data, is predicted to be strongly correlated
437: with the age and recent mass growth of galaxy clusters. We fitted the
438: the following simple formulae to the theoretical data to quantify the
439: predicted behavior in a convenient form:
440: $\Delta_{90}=\tninety-t_{z=0.2}=\tau_{90}+\alpha_{90}\log(\fsub)$ and
441: $\Delta_{50}=\tninety-t_{z=0.2}=\tau_{50}+\alpha_{50}\log(\fsub)$,
442: where $\tninety$ and $\tfifty$ are the lookback times at which a
443: cluster had acquired 90\% and 50\% of its mass at $z=0.2$. The best
444: fit parameter values are: $\alpha_{90}=(-1.34\pm0.79)\,{\rm Gyr}$,
445: $\tau_{90}=(0.31\pm0.56)\,{\rm Gyr}$ and
446: $\alpha_{50}=(-2.77\pm1.66)\,{\rm Gyr}$,
447: $\tau_{50}=(0.99\pm1.18)\,{\rm Gyr}$. Low-$\fsub$ clusters
448: ($\fsub\ls0.1$; e.g.\ A\,383, A\,1835) are therefore interpreted as,
449: on average, having formed at $z{\gs}0.8$ and having suffered
450: ${\le}10\%$ mass growth in the 2\,Gyr preceding $z=0.2$, i.e.\ since
451: $z\simeq0.4$. In contrast, high-$\fsub$ clusters ($\fsub\gs0.4$;
452: e.g.\ A\,68, A\,773) are interpreted, on average, to have formed just
453: $\sim2\,{\rm Gyr}$ before $z=0.2$, i.e.\ since $z\simeq0.4$, and
454: suffered $10\%$ mass growth in the $\sim0.5\,{\rm Gyr}$ preceding
455: $z=0.2$, i.e.\ since $z{\simeq}0.25$.
456:
457: Our synthesis therefore demonstrate that lensing-based measurements of
458: $\fsub$ can be combined with semi-analytic structure formation models
459: to estimate the average age and assembly history of observed clusters.
460: This suggests numerous avenues for further exploration, including: (i)
461: expansion of the observed samples by at least an order of magnitude,
462: (ii) calibration of the completeness of the lensing-based mass
463: function of sub-halos in clusters, (iii) investigation of how
464: lensing-based cluster age and assembly history estimates might allow
465: new cosmological constraints to be derived, for example, on the dark
466: energy equation of state parameter $w$, (iv) analysis of cluster
467: galaxy populations and cluster scaling relations as a function of
468: cluster age.
469:
470: \acknowledgements
471: We thank Arif Babul, Habib Khosroshahi, Jean-Paul Kneib, Ian Smail and
472: Risa Wechsler for helpful discussions. GPS acknowledges support from
473: NASA (HST-GO-10420.04-A), Caltech, and a Royal Society University
474: Research Fellowship. JET acknowledges financial support from the US
475: NSF (grant AST-0307859) and DoE (contract DE-FG02-04ER41316) and from
476: NSERC Canada.
477:
478: \begin{thebibliography}{31}
479: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
480:
481: \bibitem[{{Diemand} {et~al.}(2004){Diemand}, {Moore}, \& {Stadel}}]{Diemand04}
482: {Diemand}, J., {Moore}, B., \& {Stadel}, J. 2004, \mnras, 352, 535
483:
484: \bibitem[{{Evrard} {et~al.}(1993){Evrard}, {Mohr}, {Fabricant}, \&
485: {Geller}}]{Evrard93}
486: {Evrard}, A.~E., {Mohr}, J.~J., {Fabricant}, D.~G., \& {Geller}, M.~J. 1993,
487: \apjl, 419, L9+
488:
489: \bibitem[{{Gao} {et~al.}(2004){Gao}, {White}, {Jenkins}, {Stoehr}, \&
490: {Springel}}]{Gao04}
491: {Gao}, L., {White}, S.~D.~M., {Jenkins}, A., {Stoehr}, F., \& {Springel}, V.
492: 2004, \mnras, 355, 819
493:
494: \bibitem[{{Kneib}(1993)}]{Kneib93Th}
495: {Kneib}, J.-P. 1993, Ph.D.~Thesis, Universit\'e Paul Sabatier, Toulouse, France
496:
497: \bibitem[{{Kneib} {et~al.}(1996){Kneib}, {Ellis}, {Smail}, {Couch}, \&
498: {Sharples}}]{Kneib96}
499: {Kneib}, J.-P., {Ellis}, R.~S., {Smail}, I., {Couch}, W.~J., \& {Sharples},
500: R.~M. 1996, \apj, 471, 643
501:
502: \bibitem[{{Kneib} {et~al.}(1993){Kneib}, {Mellier}, {Fort}, \&
503: {Mathez}}]{Kneib93}
504: {Kneib}, J.~P., {Mellier}, Y., {Fort}, B., \& {Mathez}, G. 1993, \aap, 273, 367
505:
506: \bibitem[{{Kneib} {et~al.}(1995){Kneib}, {Mellier}, {Pello}, {Miralda-Escude},
507: {Le Borgne}, {Boehringer}, \& {Picat}}]{Kneib95}
508: {Kneib}, J.~P., {Mellier}, Y., {Pello}, R., {Miralda-Escude}, J., {Le Borgne},
509: J.-F., {Boehringer}, H., \& {Picat}, J.-P. 1995, \aap, 303, 27
510:
511: \bibitem[{{Lacey} \& {Cole}(1993)}]{LaceyCole}
512: {Lacey}, C. \& {Cole}, S. 1993, \mnras, 262, 627
513:
514: \bibitem[{{Mantz} {et~al.}(2007){Mantz}, {Allen}, {Ebeling}, \&
515: {Rapetti}}]{Mantz07}
516: {Mantz}, A., {Allen}, S.~W., {Ebeling}, H., \& {Rapetti}, D. 2007, ArXiv
517: e-prints, 709
518:
519: \bibitem[{{Mercurio} {et~al.}(2003){Mercurio}, {Girardi}, {Boschin},
520: {Merluzzi}, \& {Busarello}}]{Mercurio03}
521: {Mercurio}, A., {Girardi}, M., {Boschin}, W., {Merluzzi}, P., \& {Busarello},
522: G. 2003, \aap, 397, 431
523:
524: \bibitem[{{Moore} {et~al.}(1998){Moore}, {Governato}, {Quinn}, {Stadel}, \&
525: {Lake}}]{Moore98}
526: {Moore}, B., {Governato}, F., {Quinn}, T., {Stadel}, J., \& {Lake}, G. 1998,
527: \apjl, 499, L5+
528:
529: \bibitem[{{Moran} {et~al.}(2007){Moran}, {Ellis}, {Treu}, {Smith}, {Rich}, \&
530: {Smail}}]{Moran07}
531: {Moran}, S.~M., {Ellis}, R.~S., {Treu}, T., {Smith}, G.~P., {Rich}, R.~M., \&
532: {Smail}, I. 2007, \apj, 671, 1503
533:
534: \bibitem[{{Nagai} \& {Kravtsov}(2005)}]{Nagai05}
535: {Nagai}, D. \& {Kravtsov}, A.~V. 2005, \apj, 618, 557
536:
537: \bibitem[{{Natarajan} {et~al.}(2002){Natarajan}, {Loeb}, {Kneib}, \&
538: {Smail}}]{Natarajan02b}
539: {Natarajan}, P., {Loeb}, A., {Kneib}, J.-P., \& {Smail}, I. 2002, \apjl, 580,
540: L17
541:
542: \bibitem[{{Navarro} {et~al.}(1997){Navarro}, {Frenk}, \& {White}}]{Navarro97}
543: {Navarro}, J.~F., {Frenk}, C.~S., \& {White}, S.~D.~M. 1997, \apj, 490, 493
544:
545: \bibitem[{{Pello} {et~al.}(1991){Pello}, {Sanahuja}, {Le Borgne}, {Soucail}, \&
546: {Mellier}}]{Pello91}
547: {Pello}, R., {Sanahuja}, B., {Le Borgne}, J., {Soucail}, G., \& {Mellier}, Y.
548: 1991, \apj, 366, 405
549:
550: \bibitem[{{Poole} {et~al.}(2008){Poole}, {Babul}, {McCarthy}, {Sanderson}, \&
551: {Fardal}}]{Poole08}
552: {Poole}, G.~B., {Babul}, A., {McCarthy}, I.~G., {Sanderson}, A.~J.~R., \&
553: {Fardal}, M.~A. 2008, ArXiv e-prints, 804
554:
555: \bibitem[{{Richard} {et~al.}(2007){Richard}, {Kneib}, {Jullo}, {Covone},
556: {Limousin}, {Ellis}, {Stark}, {Bundy}, {Czoske}, {Ebeling}, \&
557: {Soucail}}]{Richard07}
558: {Richard}, J., {Kneib}, J.-P., {Jullo}, E., {Covone}, G., {Limousin}, M.,
559: {Ellis}, R., {Stark}, D., {Bundy}, K., {Czoske}, O., {Ebeling}, H., \&
560: {Soucail}, G. 2007, \apj, 662, 781
561:
562: \bibitem[{{Sand} {et~al.}(2005){Sand}, {Treu}, {Ellis}, \& {Smith}}]{Sand05}
563: {Sand}, D.~J., {Treu}, T., {Ellis}, R.~S., \& {Smith}, G.~P. 2005, \apj, 627,
564: 32
565:
566: \bibitem[{{Smith}(2002)}]{Smith02Th}
567: {Smith}, G.~P. 2002, Ph.D.~Thesis, University of Durham, UK
568:
569: \bibitem[{{Smith} {et~al.}(2003){Smith}, {Edge}, {Eke}, {Nichol}, {Smail}, \&
570: {Kneib}}]{Smith03}
571: {Smith}, G.~P., {Edge}, A.~C., {Eke}, V.~R., {Nichol}, R.~C., {Smail}, I., \&
572: {Kneib}, J. 2003, \apjl, 590, L79
573:
574: \bibitem[{{Smith} {et~al.}(2001){Smith}, {Kneib}, {Ebeling}, {Czoske}, \&
575: {Smail}}]{Smith01a}
576: {Smith}, G.~P., {Kneib}, J., {Ebeling}, H., {Czoske}, O., \& {Smail}, I. 2001,
577: \apj, 552, 493
578:
579: \bibitem[{{Smith} {et~al.}(2005){Smith}, {Kneib}, {Smail}, {Mazzotta},
580: {Ebeling}, \& {Czoske}}]{Smith05a}
581: {Smith}, G.~P., {Kneib}, J.-P., {Smail}, I., {Mazzotta}, P., {Ebeling}, H., \&
582: {Czoske}, O. 2005, \mnras, 359, 417
583:
584: \bibitem[{{Smith} {et~al.}(2002){Smith}, {Smail}, {Kneib}, {Davis}, {Takamiya},
585: {Ebeling}, \& {Czoske}}]{Smith02b}
586: {Smith}, G.~P., {Smail}, I., {Kneib}, J.-P., {Davis}, C.~J., {Takamiya}, M.,
587: {Ebeling}, H., \& {Czoske}, O. 2002, \mnras, 333, L16
588:
589: \bibitem[{{Somerville} \& {Kolatt}(1999)}]{Somerville99}
590: {Somerville}, R.~S. \& {Kolatt}, T.~S. 1999, \mnras, 305, 1
591:
592: \bibitem[{{Tasitsiomi} {et~al.}(2004){Tasitsiomi}, {Kravtsov}, {Gottl{\"o}ber},
593: \& {Klypin}}]{Tasitsiomi04}
594: {Tasitsiomi}, A., {Kravtsov}, A.~V., {Gottl{\"o}ber}, S., \& {Klypin}, A.~A.
595: 2004, \apj, 607, 125
596:
597: \bibitem[{{Taylor} \& {Babul}(2001)}]{Taylor01}
598: {Taylor}, J.~E. \& {Babul}, A. 2001, \apj, 559, 716
599:
600: \bibitem[{{Taylor} \& {Babul}(2004)}]{Taylor04}
601: ---. 2004, \mnras, 348, 811
602:
603: \bibitem[{{Taylor} \& {Babul}(2005)}]{Taylor05b}
604: ---. 2005, \mnras, 364, 535
605:
606: \bibitem[{{van den Bosch} {et~al.}(2005){van den Bosch}, {Tormen}, \&
607: {Giocoli}}]{vandenbosch05}
608: {van den Bosch}, F.~C., {Tormen}, G., \& {Giocoli}, C. 2005, \mnras, 359, 1029
609:
610: \bibitem[{{Wechsler} {et~al.}(2002){Wechsler}, {Bullock}, {Primack},
611: {Kravtsov}, \& {Dekel}}]{Wechsler02}
612: {Wechsler}, R.~H., {Bullock}, J.~S., {Primack}, J.~R., {Kravtsov}, A.~V., \&
613: {Dekel}, A. 2002, \apj, 568, 52
614:
615: \end{thebibliography}
616:
617:
618: \end{document}
619:
620: