1: \documentclass{emulateapj}
2: \usepackage{apjfonts,graphics}
3:
4: %\documentclass[12pt,preprint]{aastex}
5:
6: \newcommand{\simgt}{\lower.5ex\hbox{$\; \buildrel > \over \sim \;$}}
7: \newcommand{\simlt}{\lower.5ex\hbox{$\; \buildrel < \over \sim \;$}}
8: %\newcommand{\citet}[1] {\cite{#1}}
9: %\newcommand{\citep}[1] {(\cite{#1})}
10: \newcommand{\bm}[1]{\mbox{\boldmath$#1$}}
11: \newcommand{\kaco}[1]{\left\langle{#1}\right\rangle}
12: \newcommand{\skaco}[1]{\langle{#1}\rangle}
13: \newcommand{\ereion}{\eta_{\rm reion}}
14: \newcommand{\solam}{M_\odot}
15: \newcommand{\tildel}{\tilde{\kappa}}
16: \newcommand{\kdd}{l^{\prime}}
17: %\newcommand{\arcsec}{$' '$}
18: \newcommand{\vkdd}{\bm{l}^{\prime}}
19: \newcommand{\rvir}{r_{\rm vir}}
20: \newcommand{\rcor}{r_{\rm c}}
21: \newcommand{\baredth}{\;\overline{\raise1.0pt\hbox{$'$}\hskip-6pt \partial}\;}
22: \newcommand{\edth}{\;\raise1.0pt\hbox{$'$}\hskip-6pt\partial\;}
23:
24: \usepackage{epsfig}
25: \usepackage{subfigure}
26: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
27:
28: \begin{document}
29:
30: \title{Comparison of Cluster Lensing Profiles with
31: $\Lambda$CDM Predictions\altaffilmark{1}}
32:
33:
34: %\title{Comparison of $\Lambda$CDM with Lensing Profiles of
35: %Massive Galaxy Clusters\altaffilmark{1}}
36:
37: \author{Tom Broadhurst\altaffilmark{2}, Keiichi Umetsu\altaffilmark{3,4},
38: Elinor Medezinski\altaffilmark{2}, Masamune Oguri\altaffilmark{5},
39: Yoel Rephaeli\altaffilmark{2,6}}
40:
41: \altaffiltext{1}{Based on data collected at Subaru Telescope, which is
42: operated by the National Astronomical Observatory of Japan.}
43: \altaffiltext{2}{School of Physics and Astronomy Tel Aviv University, Israel}
44: \altaffiltext{3}{Institute of Astronomy and Astrophysics, Academia Sinica, P.~O. Box
45: 23-141, Taipei 10617, Taiwan}
46: \altaffiltext{4}{Leung Center for Cosmology and Particle Astrophysics,
47: National Taiwan University, Taipei 10617, Taiwan}
48: \altaffiltext{5}{Kavli Institute for Particle Astrophysics and
49: Cosmology, Stanford University, 2575 Sand Hill Road, Menlo Park, CA
50: 94025, USA}
51: \altaffiltext{6}{Center for Astrophysics and Space Sciences, UC San Diego,
52: California, USA}
53:
54: \begin{abstract}
55:
56: We derive lens distortion and magnification profiles of four well
57: known clusters observed with Subaru. Each cluster is very well fitted
58: by the general form predicted for Cold Dark Matter (CDM) dominated
59: halos, with good consistency found between the independent distortion
60: and magnification measurements. The inferred level of mass
61: concentration is surprisingly high, $8<c_{\rm vir}<15$ ($\langle
62: c_{\rm vir}\rangle = 10.39 \pm 0.91$), compared to the relatively
63: shallow profiles predicted by the $\Lambda$CDM model, $c_{\rm
64: vir}=5.06 \pm 1.10$ (for $\langle M_{\rm vir}\rangle =1.25\times
65: 10^{15}M_{\odot}/h)$. This represents a $4\sigma$ discrepancy, and
66: includes the relatively modest effects of projection bias and profile
67: evolution derived from N-body simulations, which oppose each other
68: with little residual effect. In the context of CDM based cosmologies,
69: this discrepancy implies clusters collapse earlier ($z\geq 1$) than
70: predicted ($z<0.5$), when the Universe was correspondingly denser.
71:
72: \end{abstract}
73:
74: \keywords{cosmology: observations -- gravitational lensing
75: -- large-scale structure of universe
76: -- galaxies: clusters: individual (A1689, A1703, A370, RXJ1347-11)}
77:
78:
79: \section{Introduction}\label{section1}
80:
81: Dark matter (DM) is understood to comprise $\simeq 85\%$ of the
82: material Universe (Fukugita \& Peebles 2004), in the form of massive
83: halos around galaxies and clusters of galaxies. The density profile of
84: a dark halo depends on the unknown nature of the DM and the way
85: structure develops over cosmic time. Galaxy halos may be modified
86: significantly when gas associated with the disk of a galaxy cools and
87: condenses during its formation. In contrast, clusters are so massive
88: that the virial temperature of the gas (typically, $3-15$ keV) is too
89: high for efficient cooling and hence the cluster potential reflects
90: the dominant DM.
91:
92: Models for the development of structure now rest on accurate
93: measurements of the power spectrum of mass fluctuations, the
94: cosmological density of DM, and the contribution of a cosmological
95: constant, $\Lambda$ (Spergel et al. 2007; Komatsu et al. 2008). This
96: framework is the standard ``$\Lambda$CDM'' cosmological model, with
97: the added simple assumptions that DM is collisionless, reacts only
98: gravitationally, was never relativistic ('Cold'), and with initially
99: Gaussian-distributed density perturbations. In this context, detailed
100: $N$-body simulations have established a clear prediction that
101: CDM-dominated cluster halos should have relatively shallow,
102: low-concentration mass profiles, where the logarithmic gradient
103: flattens continuously toward the center with a central slope tending
104: towards $r^{-1}$, interior to a characteristic radius, $r_s\lesssim
105: 100-200 {\rm kpc}\, h^{-1}$ (Navarro, Frenk, \& White 1997, NFW, hereafter;
106: Bullock et al. 2001; Maccio et al 2006; Hennawi et al. 2007; Neto et
107: al. 2007; Duffy et al. 2008). This ``NFW'' profile is characterized
108: by the (total) mass, $M_{\rm vir}$, within the virial radius, $r_{\rm
109: vir}$, and by the concentration parameter, $c_{\rm vir}\equiv r_{\rm
110: vir}/r_s$.
111:
112: Interaction between clusters indicates DM is collisionless, in
113: particular the ``bullet cluster''(Markevitch et al. 2002) - where
114: shocked gas lies between two substantial galaxy clusters, implying
115: these clusters passed through each other recently (Markevitch et
116: al. 2002). Here the weak lensing signal follows the bimodal
117: distribution of member galaxies reflecting substantial DM associated
118: with each cluster component (Clowe et al. 2006) and that this DM is
119: relatively collisionless like the galaxies. These observations
120: disfavour the class of alternative gravity theories for which the
121: lensing signal is expected to trace the dominant baryonic contribution
122: of the gas (Clowe et al. 2006). Other cases of interaction show that in
123: general displacement of the gas relative to the DM is typically
124: related to interaction (Jee et al. 2006; Okabe \& Umetsu 2008).
125:
126: For many clusters no obvious evidence of recent interaction is seen as
127: the gas and member galaxies follow a symmetric, structureless
128: distribution. Measurements of the mass profiles of these relaxed
129: clusters may help in understanding the nature of DM, preferably
130: relying on gravitational lensing signals where model-dependent
131: assumptions are not required. For A1689, over 100 multiply-lensed
132: images have been used to derive the inner mass profile (Broadhurst et
133: al. 2005a), with the outer profile determined from weak lensing
134: (Broadhurst et al. 2005b). Together, the full profile has the predicted
135: NFW form, but with a surprisingly high concentration when compared to
136: the shallow profiles of the standard $\Lambda$CDM model (Broadhurst et
137: al. 2005b; Umetsu \& Broadhurst 2008). Furthermore, good consistency is
138: also found between the lensing based mass profile and the X-ray and
139: dynamical structure of this cluster in model independent analyses
140: (Lemze et al. 2007; Lemze et al. 2008, in preparation). A similar
141: discrepancy is also indicated by lensing observations of other
142: clusters e.g., MS2137-23 (Gavazzi et al. 2003) and CL0024+16 (Kneib et
143: al. 2003).
144:
145: Here we examine a sample of relaxed, high-mass clusters, to test the
146: distinctive prediction of $\Lambda$CDM, and to settle empirically
147: whether the mass profile of A1689 is unusual. In section \S1 we
148: describe the data reduction in \S3 we present the weak lensing
149: analysis and the magnification profiles derived from background number
150: counts and make comparison with the prediction of the $\Lambda$CDM;
151: in \S4 we discuss our results and conclusions.
152:
153: \section{Data Reduction}
154:
155: \begin{deluxetable*}{ccccccccc}
156: \tablecaption{ the Subaru distortion measurements
157: combined with the Einstein-radius constraint.}
158: \tabletypesize{\scriptsize}
159: \tablewidth{\textwidth}
160: \tablehead{
161: \colhead{Cluster} &
162: \colhead{$z$} &
163: \colhead{Filters} &
164: \colhead{Einstein Radius} &
165: \colhead{$\langle D_{ds}/D_s\rangle$} &
166: \colhead{$\frac{d\log N(<m)}{dm}$} &
167: \colhead{$M_{\rm vir}$} &
168: \colhead{ $c_{\rm vir}$} &
169: \colhead{$\chi^2/{\rm dof}$} \\
170: \colhead{} &
171: \colhead{} &
172: \colhead{} &
173: \colhead{($\arcsec$)} &
174: \colhead{} &
175: \colhead{} &
176: \colhead{($10^{15}M_\odot/h_{70}$)} &
177: \colhead{} &
178: \colhead{}
179: }
180: \startdata
181: %%%%%%%%%%%%%%%%%%%%
182: A1689 & 0.183 & $V_{\rm J}i'$ & 52 ($z_s=3.05$) & $0.704$ & 0.150 &
183: $1.59^{+0.24}_{-0.22}$ & $15.69^{+3.96}_{-2.88}$ & $4.94/9$ \\%[.7ex]
184: A1703 & 0.258 & $g'r'i'$ & 33 ($z_s=2.8$) & $0.722$ & 0.062 &
185: $1.30^{+0.24}_{-0.20}$ & $9.92^{+2.39}_{-1.63}$ & $2.69/5$ \\%[.7ex]
186: A370 & 0.375 & $BR_{\rm C}z'$ & 43 ($z_s=1.5$) & $0.606$ & 0.088 &
187: $2.93^{+0.36}_{-0.32}$ & $7.75^{+1.12}_{-0.92}$ & $5.54/8$ \\%[.7ex]
188: RX J1347-11 & 0.451 & $V_{\rm J}R_{\rm C}z'$ & 35 ($z_s=1.8$) & $0.553$ & 0.066 &
189: $1.47^{+0.26}_{-0.23}$ & $10.42^{+3.25}_{-2.13}$ & $6.25/7$ \\%[1.5ex]
190: %%%%%%%%%%%%%%%%%%%%%
191: \enddata
192:
193: \end{deluxetable*}
194:
195: We analyse deep images of A1689, A1703, A370, RXJ1347-11 taken by
196: the wide-field camera Suprime-Cam ($34\prime\times 27\prime$, Miyazaki
197: et al. 2002) in Subaru (8.3m) which are observed deeply in several
198: optical passbands, listed in Table~1. These clusters are of interest
199: due to the exceptional quality of the data, with exposures in the
200: range 2000s-10000s per pass band, with seeing ranging from 0\arcsec.5
201: to 0\arcsec.75. We use either the R or I bands for our weak lensing
202: measurements (described below \S3) for which the instrumental
203: response, sky background and seeing conspire to provide the best
204: images. All the available bands are used to define colours with which
205: we separate cluster members from the background in order to minimise
206: dilution of the weak lensing signal by unlensed objects.
207:
208:
209: %Archival images are retrieved from the Subaru archive system;
210: %SMOKA\footnote[1]{http://smoka.nao.ac.jp.} and togehter with
211: %proprietry data the image reduction follows the proceedures outlined
212: %in Medezinski et al, based on the standard Subaru pipeline (Miyazaki et
213: %al. 2002).
214:
215: The standard pipeline reduction software for Suprime-Cam (Yagi et
216: al. 2002) is applied for flat-fielding, instrumental distortion
217: correction, differential refraction, PSF matching, sky subtraction
218: and stacking. Photometric catalogs are constructed from stacked and
219: matched images using Sextractor (Bertin \& Arnaut 1996). We select
220: red galaxies with colours redder than the colour-magnitude sequence
221: of cluster E/SO galaxies. The sequence forms a well defined line due
222: to the richness and relatively low redshifts of our clusters. These
223: red galaxies are expected to lie in the background by virtue of
224: k-corrections which are greater than for the red cluster sequence
225: galaxies and convincingly demonstrated spectroscopically by Rines \&
226: Geller (2008). We also include very blue galaxies falling far from
227: the cluster sequence,$> 1.^{m}5-2.^{m}5$, depending on the passbands
228: available for each cluster (listed in Table 1), to minimise cluster
229: contamination (see Medzinski et al. 2007). Typically the proportion
230: of blue galaxies used is around 50\% of the red background. In
231: addition, we adopt a conservative magnitude limit of $m<25.5-26.0$,
232: depending on the depth of the data for each cluster, to avoid
233: incompleteness. Imposing these strict limits is a necessary
234: precaution against the diluting effect that unlensed cluster and
235: foreground galaxies will otherwise produce, leading to spuriously
236: shallow central mass profiles, as shown in Broadhurst et al. 2005b and
237: Medezinski et al. 2008.
238:
239:
240:
241: \section{Measurements of Lensing Effects}
242:
243: These data permit high quality weak lensing measurements, following
244: established methods required to deal with instrumental and atmospheric
245: effects following the formalism outlined in Kaiser, Squires, \&
246: Broadhurst (1995), with modifications described in Erben et
247: al. (2001). The mean residual stellar ellipticity after PSF correction
248: is less than, $\sim 10^{-4}$, in all fields, averaged over 400-600
249: stars and consistent with the standard error on this
250: measurement, $\simeq 10^{-4}$. Note also the lack of any systematic
251: deviation from zero in the B-mode, $g_\times$, shown in Figure
252: 2a. Full details of the methods are presented in Umetsu \& Broadhurst
253: (2008).
254:
255: Figure 1 shows the gravitational shear field by locally averaging
256: the corrected distortions of colour-selected background galaxies of
257: each cluster. This is compared to the distribution of color-selected
258: cluster sequence galaxies. In each case, one large symmetric cluster
259: is visible around which the lensing distortion pattern is clearly
260: tangential, with little significant substructure. The derived radial
261: profiles of the lensing distortion are seen to be very similar in form
262: (left panel of Figure 2), with differing amplitudes reflecting a range
263: of mass.
264:
265: The NFW profile fits well each cluster (Figure 2; Table 1), but
266: surprisingly the derived concentrations lie well in excess of the
267: standard $\Lambda$CDM model (Figure 3). For the derived masses of our
268: clusters, the predicted mean concentration is $c_{\rm vir}\sim 5.06\pm
269: 1.1$ based on Duffy et al. (2007) using the improved WMAP5 (Komatsu et
270: al. 2008). Our best-fitting values instead range over $8<c_{\rm
271: vir}<15$, with a mean value of ($\langle c_{\rm vir}\rangle = 10.39
272: \pm 0.91$) representing a $4.8\sigma$ discrepancy. Our measurements
273: also imply A370 is the most massive cluster now known, $M_{\rm
274: vir}=2.93 \times 10^{15}M_{\odot}/h_{70}$, with a virial mass twice
275: that of RXJ1347-11, the most X-ray luminous cluster known. Our
276: estimate of the virial mass of RXJ1347-11 is in very good agreement
277: with independent X-ray and lensing analyses (Miranda et al. 2007,
278: Kling et al. 2005). For A1689, a somewhat lower concentration parameter
279: is derived by Limousin et al. (2007), $c_{\rm vir}=9.6\pm 2.0$, from
280: independent weak lensing measurements, which is consistent with our
281: findings at the $1.5\sigma$ level, whereas the observed Einstein
282: radius of approximately 52\arcsec is under-predicted by the
283: NFW parameters obtained by Limousin et al. (2007), $24\arcsec
284: \pm 11\arcsec$ (see Umetsu \& Broadhurst 2008).
285:
286: Note, when model fitting an estimate of the background depth is
287: required. The mean depth is sufficient for our purposes as the
288: variation of the lens distance ratio, $D_{ds}/D_s$, is slow for our
289: sample because the clusters are at relatively low redshift compared to
290: the redshift range of the background galaxies. The estimated mean
291: depth of the combined red+blue background galaxies is listed in column
292: 5, which we obtain by applying our colour-magnitude selection to
293: Subaru imaging of the HDF region (Capak et al. 2004), where photometric
294: redshifts are reliable for red galaxies, with a mean redshift close to
295: $z=1.2$ depending on the details of the colour magnitude
296: selection. For the blue galaxies we may rely on the zCOSMOS deep
297: redshift survey (Lilly et al. 2008) where our mean depth is typically
298: $z\simeq 2.1$.
299:
300: Lensing measures projected mass and so a statistical bias arises from
301: the triaxiality of clusters in cases where the major axis lies along
302: the line-of-sight. This leads to an $\sim 18\%$ increase in the mean
303: value of lensing derived concentrations based on $\Lambda$CDM (Oguri
304: et al. 2005; Hennawi et al. 2007), and an overall discrepancy in $c_{\rm
305: vir}$ of $4.0\sigma$ with respect to the predictions. A larger bias
306: is inferred for clusters selected by the presence of large arcs, $\sim
307: 34\%$, representing the most triaxial cases (Hennawi et al. 2007). Our
308: sample is defined by the quality of available imaging and includes
309: clusters observed for reasons other than lensing, hence it is unlikely
310: that these clusters are all particularly triaxial with the long axes
311: pointing to the observer. Even so, applying the maximum estimated
312: bias ($\sim50\%$, Oguri \& Blandford 2008) cannot account
313: for our measurements (Figure 3).
314:
315: Multiply-lensed images are visible in all our clusters, from which the
316: inner mass distribution may be determined (Broadhurst et al. 2005a;
317: Halkola et al. 2008; Hennawi et al. 2008) and an equivalent Einstein
318: radius derived by averaging azimuthally (Figure 2) which for most
319: cases is close to the observed radius. For A1703 and A370, the
320: tangential critical curves are significantly elongated, for
321: $r<1\arcmin$ (Limousin et al. 2008; Kneib et al. 1993) although in
322: general the mass distribution is always less elliptical than the
323: critical curve. X-ray observations of A370 show the inner region has
324: some substructure (eg. Shu etal 2008), but beyond $r>1\arcmin$ the
325: weak lensing pattern and the distribution of galaxies is
326: symmetrical, as seen in Figure 1. For each cluster the NFW parameters
327: derived with or without the Einstein radius constraint are found to be
328: closely consistent and in combination they are more precisely
329: determined - see Table 1.
330:
331: We also examine the magnification profile, $\mu(r)$, via the surface
332: number density of background galaxies (Broadhurst, Taylor, \& Peacock
333: 1995). At faint fluxes where the counts follow a power-law slope,
334: $s=d\log N(<m)/dm$, lensing modifies the true density above the
335: magnitude limit , $N_0(m<m_{\rm lim})$, by, $N(<m_{\rm
336: lim})=N_0(<m_{\rm lim})\mu(r)^{2.5s-1}$, implying competition between
337: the magnified sky, which reduces the surface density, and an increase
338: of galaxies magnified above the flux limit. Here we use only galaxies
339: lying redward of the cluster sequence because for these the intrinsic
340: count-slope of faint red galaxies is relatively flat, $s \sim 0.1$, so
341: a net count depletion results, which we readily detect for each
342: cluster, increasing towards the center (Figure 2). For each cluster we
343: calculate the depletion of the counts expected for the the best
344: fitting NFW profile derived from the corresponding distortion
345: measurements above (Figure 2, right panel, with same colour code),
346: finding clear consistency, which considerably strengthens our
347: conclusions, and also establishes the utility of the
348: background red galaxies for measuring magnification. The variation in
349: the total number of background counts seen in Figure 2 simply reflects
350: the relative depth of the imaging data and also to some extent the
351: redshift of the cluster, which lies redder in colour space with
352: increasing cluster redshift, thereby reducing the numbers of
353: background galaxies which can be selected to lie redward of the
354: sequence for our purposes. Note, absolute counts are not used here,
355: the models are instead normalised to the observed density when
356: comparing with the model to avoid reliance on control fields.
357:
358:
359: %very similar to each. It may
360: %not be surprising that A370 is also the most distant cluster in the
361: %Abell catalog ($z=0.37$), and so conceivably ranks as the most massive
362: %cluster in the Universe, providing a useful limit for comparison with
363: %theory. For $\Lambda$CDM, taking the normalization,
364: %$\sigma_8=0.8$(Komatsu et al 2008), the most massive cluster expected
365: %within $z<0.4$ exceeds $M_{\rm vir}=5\times 10^{15}
366: %M_{\odot}$(Broadhurst \& Barkana 2008), significantly more massive
367: %than A370. Equivalently, $\approx 20$ clusters are predicted with a
368: %mass exceeding A370, $M_{\rm vir} > 3\times 10^{15} M_{\odot}$, and
369: %these should lie within a redshift of $z<0.8$, which seems excessive
370: %compared with large area X-ray surveys.
371:
372: %For the distortion measurement, small faint galaxies cannot be used to
373: %make reliable shape measurements (Umetsu \& Broadhurst 2008). For the
374: %magnification measurements no size restriction is necessary, however,
375: %we do not make use of the blue galaxies as the count slope lies close
376: %to the magnification invariant slope of, $dlogN(<m)/dm=0.4$, where no
377: %lensing signal can be generated (Broadhurst, Taylor \& Peacock 1995).
378:
379:
380: % To estimate the lensing distortion
381: %from the observed ellipticities, $e_\alpha$, the PSF anisotropy needs
382: %to be correced using the star images as references: $e'_{\alpha} =
383: %e_{\alpha} - P^{sm}_{\alpha
384: %\beta} q_{\beta}$, where $P_{sm}$ is the {\it smear polarisability}
385: %tensor and $q^*_{\alpha} = (P^{*sm})^{-1}_{\alpha \beta}e^*_{\beta}$
386: %is the stellar anisotropy kernel. The quantity $q^*_{\alpha}$ is
387: %calibrated from stars in the field. The pre-seeing reduced shear $g_\alpha$
388: %can be estimated from $g_{\alpha} =(P_g^{-1})_{\alpha\beta}
389: %e'_{\beta}$ with the {\it pre-seeing shear polarizability} tensor
390: %$P^g_{\alpha\beta}$. To measure $P^g$ we adopt the scalar correction,
391: %$P^g_{\alpha\beta}=\frac{1}{2}{\rm tr}[P^g]\delta_{\alpha\beta}\equiv
392: %P^g_{\rm s}\delta_{\alpha\beta}$. Individual galaxies have noisy
393: %$P^g$, so thirty neighbors are identified in the size ($r_g$) and
394: %magnitude ($i'$) plane, and the median value of $P^g_{\rm s}$ is used
395: %as the smoothed $P^g_{\rm s}$ for the object: $\left< P^g_{\rm s}
396: %\right>$.
397:
398:
399: \section{Discussion}
400:
401: We have shown that the mass distribution of a sample of the most
402: massive clusters follow a very similar form in terms of the observed
403: lensing based profiles. The distortion profiles measured are amongst
404: the most accurate constructed to date and the great depth of the
405: Subaru imaging permits magnification profiles to be
406: established with unprecedented detail using the background red galaxy
407: counts. The magnification and distortion profiles are fitted well by
408: the general NFW profile for CDM dominated halos, which continuously
409: steepens with radius, but the profiles derived are more concentrated
410: than predicted by $\Lambda$CDM based simulations for which clusters
411: of $\sim 10^{15}M_{\odot}$, form late ($z<0.5$) with low
412: concentrations, reflecting the low mean cosmological mass density today.
413:
414: Tension with the standard $\Lambda$CDM model is also indicated by the
415: observed Einstein radii of very massive clusters, including those
416: studied here, which lie well beyond the predicted range derived from
417: massive halos generated in $\Lambda$CDM based simulations (Broadhurst
418: \& Barkana 2008). In addition, a significant discrepancy is also
419: claimed for X-ray derived cluster concentrations, with respect to the
420: $\Lambda$CDM based on WMAP5 parameters (Duffy et al. 2008). The
421: combined X-ray and lensing based concentrations compiled by Comerford
422: and Natarajan (2008) also exceeds somewhat the standard expectation,
423: despite including X-ray based concentrations that are found to be
424: systematically lower than for lensing (Comerford and Natarajan 2008).
425: The mass-concentration relation of Mandelbaum etal (2008) is claimed to
426: be more consistent with $\Lambda$CDM. Here allowance is made for the
427: smoothing of the central lensing signal when stacking many SDSS
428: clusters, relying on the brightest cluster galaxy to define the
429: center of mass. Cluster members misclassified as background
430: galaxies when estimating photometric redshifts will also act to
431: reduce the derived concentrations, diluting the central lensing
432: signal.
433:
434: Without resorting to radical proposals regarding the nature of DM, our
435: findings imply the central region of clusters collapsed earlier than
436: expected. Assuming the simple redshift relation $c_{\rm vir}(z)=c_0
437: (1+z)$ (e.g., Wechsler et al. 2006), relating central cluster densities
438: to the cosmological mean density, then the formation of clusters with
439: $c_{\rm vir}(1+z)=10-15$ corresponds to $z \geq 1$, significantly
440: earlier than in the standard $\Lambda$CDM, for which clusters form today
441: with $c_0 \sim 5.6$.
442:
443: The presence of massive clusters at high redshift ($z\sim 2$), and the
444: old ages of their member galaxies (Zirm et al. 2008; Blakeslee et al.
445: 2003), may also imply clusters collapsed at relatively early
446: times(Mathis, Diego, \& Silk 2004), for which accelerated growth factors
447: have been proposed, adopting a generalised equation of state (Sadeh \&
448: Rephaeli 2008). Alternatively, since clusters correspond to rare
449: density maxima, then any non-Gaussianity in the early fluctuation
450: spectrum may advance cluster formation (Matthis, Diego, \& Silk 2004;
451: Sadeh, Rephaeli, \& Silk 2007). Our results present a challenge to
452: some of these models, since earlier cluster formation must not also
453: enhance significantly the abundance of clusters.
454:
455:
456:
457: %\section*{Acknowledgements}
458:
459: \acknowledgements
460:
461: This work is in part supported by the Israeli Science
462: Foundation and the National Science Council of Taiwan under the grant
463: NSC95-2112-M-001-074-MY2 and by the US Department of Energy, contract
464: DE-AC02-76SF00515.
465:
466:
467: \begin{thebibliography}{}
468:
469: \bibitem[Bertin
470: \& Arnouts(1996)]{1996A&AS..117..393B}
471: Bertin, E., \& Arnouts, S.
472: 1996, \aaps, 117, 393
473:
474:
475: \bibitem{27}
476: Blakeslee, J.~P., et al.
477: 2003, ApJL, 596, L143
478:
479: \bibitem{22}
480: Broadhurst, T.~J., Taylor, A.~N., Peacock, J.~A.,
481: 1995, ApJ, 438, 49
482:
483: \bibitem{11}
484: Broadhurst, T., et al. 2005a, ApJ, 621, 53
485:
486: \bibitem{12}
487: Broadhurst, T., Takada, M., Umetsu, K. et al. 2005b,
488: ApJL, 619, L143
489:
490: \bibitem{24}
491: Broadhurst,T.J. \& Barkana R. 2008, MNRAS, submitted.
492:
493: \bibitem{18}
494: Capak, P., et al. 2004, AJ, 127, 180
495:
496: \bibitem{10}
497: Clowe, D., et al 2006, ApJL, 648, L109
498:
499: \bibitem[Comerford
500: \& Natarajan(2007)]{2007MNRAS.379..190C}
501: Comerford, J.~M., \& Natarajan, P.
502: 2007, \mnras, 379, 190
503:
504:
505: \bibitem{}
506: Bullock, J, et al, 2001, MNRAS, 321, 559
507:
508: \bibitem{5}
509: Duffy, A.R., Schaye, J., Kay, S.T., Della Vecchia, C.
510: 2008, MNRAS, submitted
511:
512:
513: \bibitem{17}
514: Erben, T., Van Waerbeke, L.,
515: Bertin, E., Mellier, Y., Schneider, P.,
516: 2001, A\&A, 366, 717
517:
518: \bibitem{}
519: Fukugita \& Peebles 2004, ApJ, 616, 643
520:
521: \bibitem{14}
522: Gavazzi, R., Fort, B., Mellier, Y., Pell{\'o}, R.,
523: \& Dantel-Fort, M.\ 2003, A\&A, 403, 11
524:
525: \bibitem{21}
526: Halkola, A., et al. 2008, A\&A, 481, 65
527:
528: \bibitem{4}
529: Hennawi, J.~F., Dalal, N., Bode, P.,
530: \& Ostriker, J.~P. 2007, ApJ, 654, 714
531:
532: \bibitem[Hennawi et al.(2008)]{2008AJ....135..664H}
533: Hennawi, J.~F., et al. 2008, \aj, 135, 664
534:
535:
536: \bibitem[Jee et al.(2005)]{2005ApJ...618...46J}
537: Jee, M.~J., White, R.~L.,
538: Ben{\'{\i}}tez, N., Ford, H.~C., Blakeslee, J.~P.,
539: Rosati, P., Demarco, R.,
540: \& Illingworth, G.~D. 2005, \apj, 618, 46
541:
542:
543: \bibitem{16}
544: Kaiser, N., Squires, G., Broadhurst, T., 1995, ApJ, 449, 460
545:
546: \bibitem[Kling et al.(2005)]{2005ApJ...625..643K}
547: Kling, T.~P., Dell'Antonio, I., Wittman, D., \& Tyson,
548: J.~A.
549: 2005, \apj, 625, 643
550:
551:
552: \bibitem[Kneib et
553: al.(1993)]{1993A&A...273..367K}
554: Kneib, J.~P., Mellier, Y., Fort, B., \& Mathez, G.
555: 1993, \aap, 273, 367
556:
557:
558: \bibitem{15}
559: Kneib, J.-P. {et~al.} 2003, ApJ, 598, 804
560:
561: \bibitem{23}
562: Komatsu, E., et al.\ 2008, ApJS, submitted.
563:
564: \bibitem[Lemze et al.(2008)]{2008MNRAS.386.1092L}
565: Lemze, D., Barkana, R.,
566: Broadhurst, T.~J., \& Rephaeli, Y.
567: 2008, \mnras, 386, 1092
568:
569:
570: \bibitem{19}
571: Lilly, S.~J., et al.\ 2007, ApJS, 172, 70
572:
573:
574: \bibitem[Limousin et al.(2007)]{2007ApJ...668..643L}
575: Limousin, M., et al. 2007, \apj, 668, 643
576:
577:
578: \bibitem[Limousin et al.(2008)]{2008arXiv0802.4292L}
579: Limousin, M., et al. 2008, preprint (arXiv:0802.4292)
580:
581: \bibitem[Macci{\`o} et al.(2007)]{2007MNRAS.378...55M}
582: Macci{\`o}, A.~V., Dutton, A.~A., van den Bosch, F.~C.,
583: Moore, B., Potter, D., \& Stadel, J. 2007, \mnras, 378, 55
584:
585:
586: \bibitem[Mandelbaum et al.(2006)]{2006MNRAS.372..758M}
587: Mandelbaum, R.,
588: Seljak, U., Cool, R.~J., Blanton, M., Hirata, C.~M.,
589: \& Brinkmann, J.\ 2006, \mnras, 372, 758
590:
591:
592: \bibitem{9}
593: Markevitch, M., et al, 2002, ApJL, 567, L27
594:
595:
596: \bibitem{28}
597: Mathis, H., Diego, J.~M., \& Silk, J.\ 2004, MNRAS, 353, 681
598:
599: \bibitem[Medezinski et al.(2007)]{2007ApJ...663..717M}
600: Medezinski, E., et al. 2007, \apj, 663, 717
601:
602: \bibitem[Miranda et al.(2008)]{2008MNRAS.385..511M}
603: Miranda, M., Sereno, M., de Filippis, E., \& Paolillo,
604: M. 2008, \mnras, 385, 511
605:
606:
607: \bibitem[Miyazaki et al. (2002)]{miya02}
608: Miyazaki, S. et al. 2002, PASJ, 54, 833
609:
610: \bibitem{7}
611: Navarro, J.~F., Frenk, C.~S., White, S.~D.~M.
612: 1997, ApJ, 490, 493
613:
614: \bibitem{}
615: Neto, A.F. et al. 2007, MNRAS, 381, 1450
616:
617: \bibitem{}
618: Oguri, M. \& Blandford, R.D. 2008,
619: preprint (arXiv:0808.0192)
620:
621: \bibitem{20}
622: Oguri, M., Takada, M., Umetsu, K., \& Broadhurst, T.
623: 2005, ApJ, 632, 841
624:
625: \bibitem{}
626: Okabe, N., Umetsu, K. 2008, PASJ, 60, 345 (arXiv:0702649)
627:
628:
629: \bibitem[Rines \& Geller(2008)]{2008AJ....135.1837R}
630: Rines, K., \& Geller, M.~J. 2008, \aj, 135, 1837
631:
632: \bibitem{29}
633: Sadeh, S., \& Rephaeli, Y. 2008, MNRAS, submitted.
634:
635: \bibitem{30}
636: Sadeh, S., Rephaeli, Y., \& Silk, J. 2007, MNRAS, 380, 637
637:
638: \bibitem[Shu et al.(2008)]{2008arXiv0805.1148S}
639: Shu, C., Zhou, B.,
640: Bartelmann, M., Comerford, J.~M., Huang, J.-S.,
641: \& Mellier, Y. 2008, ApJ in press (arXiv:0805.1148)
642:
643:
644: \bibitem{6}
645: Spergel, D.~N. et al. 2007, ApJS, 170, 377
646:
647: \bibitem{8}
648: Springel et al. 2005, Nature, 435, 629
649:
650:
651: \bibitem{13}
652: Umetsu, K. \& Broadhurst, T. 2008, ApJ, 684, 177
653: (arXiv:0712.3441)
654:
655:
656: \bibitem{25}
657: Wechsler, R., Zentner, A., Bullock, J.~S., Kravtsov, A.~V.,
658: \& Allgood, B. 2006, ApJ, 652, 71
659:
660:
661: \bibitem[Yagi et al. (2002)]{yagi02}
662: Yagi, M., Kashikawa, N., Sekiguchi, M., et al.
663: 2002, AJ, 123, 66
664:
665: \bibitem{26}
666: Zirm, A.,W., et al. 2008, ApJ, 679,2
667:
668:
669:
670: \end{thebibliography}
671:
672:
673:
674: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
675: %%
676: %% Tables
677: %%
678: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
679:
680:
681:
682:
683: \clearpage
684:
685:
686:
687: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
688:
689: \begin{figure}[t]
690: \begin{center}
691: \resizebox{16cm}{!}{\includegraphics{f1.eps}}
692: \end{center}
693: \caption{Maps of the surface number-density distribution of
694: color-selected cluster member galaxies, with the gravitational shear
695: of background galaxies overlayed; $10\%$ ellipticity is indicated top
696: right, and the resolution of the distortion map is shown bottom right.
697: In each case a single concentration of galaxies is visible, around
698: which a coherent tangential pattern is centered, with little
699: significant substructure.
700: }
701: \label{fig:shear}
702: \end{figure}
703: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
704:
705:
706: \begin{figure}[tbh]
707: \begin{center}
708: \hglue0.0cm{\includegraphics[angle=0,width=8.cm]{f2a.eps}}
709: \hglue0.5cm{\includegraphics[angle=0,width=8.cm]{f2b.eps}}
710: \caption {({\it Left}): Tangential distortion profiles of the 4
711: clusters based on the combined red and blue background samples
712: and compared with the best-fitting NFW model. The Einstein-radius
713: determined from multiply-lensed images is indicated, marking the
714: point of maximum distortion, $g_+=1$. Shown below is the rotated
715: shear, $g_x$, demonstrating no obvious non-tangential
716: distortion. ({\it Right}): Profiles of background red galaxy
717: counts, whose intrinsic slope is relatively shallow, so lens
718: magnification reduces the observed numbers towards the center.
719: Over-plotted are normalised NFW models derived from the
720: distortion profiles in the left panel, demonstrating consistency
721: between these two independent lensing observables; note A370 is
722: shifted upward 40\% for clarity. The count uncertainty is
723: obtained by sub-dividing each annulus into equal area cells, with
724: a tail of $>2\sigma$ cells excluded to remove inherent small
725: scale clustering of the background. Small areas around each
726: bright object ($m_R<21$) are excluded, including cluster members,
727: interior to 3 times the isophotal radius, where the detection of
728: faint galaxies is significantly compromised (Broadhurst et al.
729: 2005b; Umetsu \& Broadhurst 2008). }
730: \label{one}
731: \end{center}
732: \end{figure}
733:
734:
735:
736:
737: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
738: \begin{figure}[t]
739: \begin{center}
740: \resizebox{14cm}{!}{\includegraphics{f3.eps}}
741: \end{center}
742: \caption{ Comparison of observations with the $\Lambda$CDM model, based on
743: $N$-body simulations for the $c_{\rm vir}-M_{\rm vir}$ relation,
744: derived at $z=0$. The predictions of Duffy et al. (2008)
745: ($\sigma_8=0.8$, WMAP5) and Neto et al. (2007)
746: ($\sigma_8=0.9$, WMAP1) are shown as solid curves, with
747: $1\sigma$ uncertainty (from Neto et al. 2007) indicated by the hatched
748: area. Also shown is the level of selection and orientation bias for
749: projected masses based on $\Lambda$CDM (dashed curves, see text). The
750: data points are all derived from lensing alone, and are multiplied by
751: $(1+z)^{0.66}$ at the cluster redshift, for consistency with the
752: evolution of $c_{\rm vir}(M_{\rm vir})$ derived from $\Lambda$CDM
753: simulations by Duffy et al. 2008. Note, a significant correction for
754: substructure is required for Cl0024+16 (Umetsu et al., in
755: preparation). Clearly the data lie well above the predicted relation even
756: when possible sources of bias are considered.}
757: \label{fig:magbias}
758: \end{figure}
759: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
760:
761:
762:
763: \end{document}
764: