1: \documentclass[11pt,preprint]{emulateapj}
2:
3: %\documentclass[12pt,preprint2]{aastex}
4:
5:
6: \usepackage{natbib}
7:
8:
9: \shorttitle{Searching for ICM dust}
10: \shortauthors{Muller et al. 2008}
11:
12:
13:
14: \begin{document}
15:
16: \title{Searching for dust in the intracluster medium from reddening of background galaxies}
17:
18:
19:
20: \author{Muller, S.\altaffilmark{1}, Wu, S.-Y.\altaffilmark{1,2}, Hsieh, B.-C.\altaffilmark{1},
21: Gonz\'alez, R. A.\altaffilmark{3},
22: Loinard, L.\altaffilmark{3},
23: Yee, H. K. C.\altaffilmark{4}
24: \& Gladders, M. D.\altaffilmark{5}}
25:
26: \affil{\altaffilmark{1} Academia Sinica Institute of Astronomy and Astrophysics, P.O. Box 23-141, Taipei, 106 Taiwan}
27: \affil{\altaffilmark{2} Current address: Department of Astronomy, University of Massachusetts, Amherst, MA 01106, USA}
28: \affil{\altaffilmark{3} Centro de Radioastronom\'ia y Astrof\'isica,
29: %Instituto de Astronom\'ia, Universidad Nacional Aut\'onoma de M\'exico,
30: 58190 Morelia, Michoac\'an, Mexico}
31: \affil{\altaffilmark{4} Department of Astronomy and Astrophysics, University of Toronto, ON, M5S 3H4, Canada}
32: \affil{\altaffilmark{5} Department of Astronomy and Astrophysics, University of Chicago, Chicago, IL, USA}
33:
34:
35:
36: \begin{abstract}
37:
38: We report a search for the presence of dust in the intra-cluster medium based on the study of statistical
39: reddening of background galaxies. Armed with the Red Sequence Cluster survey data, from which we extracted
40: ({\em i}) a catalog of 458 clusters with $z_{clust} < 0.5$ and
41: ({\em ii}) a catalog of $\sim$ 90,000 galaxies with photometric redshift $0.5 < z_{ph} < 0.8$ and photometric
42: redshift uncertainty $\delta z_{ph} / (1+z_{ph}) < 0.06$, we have constructed several samples of galaxies
43: according to their projected distances to the cluster centers.
44: No significant color differences [$<$E($B-R_c$)$>$ = $0.005 \pm 0.008$, and $<$E($V-z'$)$>$ = $0.000 \pm 0.008$]
45: were found for galaxies background to the clusters, compared to the references.
46: Assuming a Galactic extinction law, we derive an average visual extinction of $<$A$_V$$>$ = $0.004 \pm 0.010$
47: towards the inner $1\times R_{200}$ of clusters.
48:
49: \end{abstract}
50:
51: \keywords{dust, extinction --- intergalactic medium --- galaxies: clusters: general}
52:
53:
54:
55: \section{Introduction}
56:
57: \cite{zwi51} was the first one to observe a weak and diffuse intra-cluster light in the Coma cluster, suggesting
58: the presence of material in the intra-cluster medium (ICM). The jump in sensitivity of the CCD detectors later
59: allowed intra-cluster light to be studied more quantitatively (see e.g., \citealt{ber95,fel02,zib05}). The
60: existence of an intra-cluster stellar population has further been clearly established with the detection of
61: planetary nebulae \citep{arn96,fel98}, red giant stars \citep{fer98} and supernovae \citep{gal03} in the ICM.
62: In parallel, X-ray spectroscopic observations revealed the presence of heavy metals in the ICM \citep{mit76,mus96}.
63: This ICM matter is commonly believed to originate from stripping and/or disruption of cluster galaxies, and
64: one can then naturally wonder about the existence of dust in the ICM.
65:
66: It is true that the ICM is a harsh medium, where strong X-ray radiation and gas temperatures of $\sim$ 10$^7$ K
67: may impose severe restrictions on the presence of dust. However, the survival timescale for dust grains
68: in this environment is still of the order of few 10$^8$ yr \citep{dra79}. Besides, many mechanisms, such as tidal
69: or ram pressure stripping, galactic winds, mass loss from ICM red giants and supergiants, or possible accretion
70: of primordial dust, may contribute to the release or replenishment of dust in the ICM [see, e.g.,\cite{pop00},
71: \cite{sch05}, \cite{dom06} for predictions on the efficiency of the different mechanisms].
72:
73: While there have been many attempts, using various methods, to detect the existence of dust in the ICM (see
74: Table~\ref{prevsearch}), the debate is still open. The methods are mostly twofold: either based on a direct
75: search for dust thermal emission, typically peaking in the far infrared to submillimeter range, or on an
76: indirect search for extinction and reddening of background sources.
77:
78: The direct detection of extended thermal emission in the intra-cluster medium would constitute the best evidence
79: for the presence of ICM dust. Theoretical predictions of the IR emission
80: have been made (see e.g., \citealt{dwe90}, \citealt{yam05}); taking into account dust injection by galaxies,
81: sputtering, and heating by the hot ICM plasma, they yield a mean dust temperature of $\sim 20 - 30$ K, and
82: dust-to-gas ratio $\sim 10^{-6}$. \cite{ann93} could not detect significant emission from a submillimeter search
83: towards 11 clusters with cooling flows, up to a limit of $\sim 10^8$ M$_\odot$. \cite{wis93} observed with IRAS a
84: larger sample of 56 clusters, from which only two may show evidence for extended diffuse emission. Further ISO
85: observations revealed FIR excess emission towards Coma \citep{sti98}, possibly due to a dust mass of $\sim
86: 10^7 - 10^9$ M$_\odot$ in the inner 0.4 Mpc region, but none towards five other Abell clusters \citep{sti02}.
87: More recently, \cite{mon05} adopted a statistical approach and co-added IRAS maps of a large number ($> 11,000$)
88: of clusters and detected a significant emission in all four IRAS band (12, 25, 60 and 100 $\mu$m). The origin of
89: the emission, tentatively attributed to ICM dust, is, however, questionable. Indeed, in addition to the
90: challenging detection of low surface brightness and extended emission, IR observations of ICM dust might be
91: affected by contamination from Galactic cirrus (e.g., \citealt{wis93,bai07}), as well as from cluster members
92: \citep{qui99}, dusty star forming galaxies in clusters (\citealt{gea06}), dust associated with the central
93: dominant galaxies or the presence of a cooling flow \citep{bre90,gra90,cox95,edg99}, or even from background
94: (possibly magnified) galaxies.
95:
96: The first claim for ICM dust was, nevertheless, historically made by \cite{zwi62}, based on the extinction of
97: light from distant clusters by nearby ones. He estimated A$_V \sim 0.4$ for Coma. \cite{kar69} derived a
98: similar value of A$_V \sim 0.3$ mag for Coma, and found a mean cluster extinction of A$_V \sim 0.20 \pm 0.05$
99: mag by averaging over 15 clusters. \cite{bog73} measured that distant Abell clusters appear to be inversely
100: correlated on the sky with nearby ones. To account for this effect, they argued for an extinction corresponding
101: to A$_{V} \sim 0.4$ mag and extending to $\sim 2.5$ times the optical radii of the nearby clusters. However,
102: these first results were probably severly affected by the difficulties of identifying background clusters.
103:
104: Obvious background sources to search for dust extinction and reddening are quasars. \cite{boy88}
105: measured a value of A$_B = 0.2$ mag (A$_ V \sim 0.15$ mag) from a deficiency of ultraviolet-excess-selected
106: objects behind clusters of galaxies. Similarly, \cite{rom92} concluded that an area-averaged extinction of
107: A$_B \sim 0.5$ mag was required to explain the anti-correlation between quasars and Abell clusters. On the
108: other hand, \cite{mao95} could not measure a significant difference [E($B-V$)$\leq 0.05$] between the color
109: distribution of radio-selected quasars located within 1$\degr$ of a foreground Abell cluster, and quasars
110: with no cluster in their line of sight. He reached the conclusion that the apparent deficiency of optically
111: discovered quasars in the background of clusters could be attributed to selection effects, reflecting the
112: difficulty to identify quasars in crowded fields.
113:
114: More recently, \cite{nol03} proposed, to our knowledge, the first and only search so far of ICM dust based
115: on extinction and reddening of background galaxies. They used a catalog of $\sim 140$ nearby ($z < 0.08$)
116: clusters with $B$ and $R$ photometric data, complete to $B \sim 20.5$ mag and $R \sim 19.5$ mag. For each
117: cluster, they constructed two samples. The ``cluster'' sample includes all galaxies lying in a circular
118: region centered on the cluster, and with an angular radius equivalent to the physical radius of 1.3 Mpc at
119: the distance of the cluster. The second sample, namely ``control sample'', contains galaxies projected
120: within a ring of inner radius 1.3 Mpc and of outer radius such that the area covered by the cluster and
121: control samples are identical. The cluster and control samples for all clusters were then respectively
122: merged to increase the number of galaxies and improve the statistics. The cluster sample inevitably suffers
123: from contamination of cluster members and foreground galaxies, although \cite{nol03} estimate that these
124: latest amount to less than 10\% contamination. The cluster and control group galaxies were then distributed
125: in two color-magnitude diagrams (CMD), and any extinction and reddening of the cluster group was searched
126: by shifting the cluster CMD both in magnitude and color with respect to the control CMD, until the two CMD
127: matched. The pixel resolution (in magnitude and color) of the CMD is constrained by the number of galaxies
128: in each pixel, which should be high enough. Given the number of galaxies in each of their samples ($<10^5$),
129: \cite{nol03} fixed the smallest reasonable pixel size for $\Delta$($B-R$) and $\Delta R$ to 0.025. They
130: could not detect a significant difference between the cluster and control samples, thus yielding upper
131: limits of A$_R < 0.025$ and E($B-R$)$< 0.025$ mag (A$_V < 0.044$) for ICM dust.
132:
133: We note briefly that dust extinction and reddening of background galaxies have been successfully detected
134: in galactic disks (see, e.g., \citealt{gon98}, \citealt{hol05}), galactic halos \citep{zar94}, and, recently,
135: in the intergalactic medium towards the M81 group \citep{xil06}. This simple method should become more and
136: more popular with the advent of wide field cameras on board of large telescopes.
137:
138: Finally, in a very recent paper, \cite{che07} claim the detection of reddening of background quasars behind
139: $0.1 < z < 0.3$ galaxy clusters using the {\em Sloan Digital Sky Survey} (SDSS) photometric and
140: spectroscopic data. They report E($g-i$) of $0.003 \pm 0.002$ from photometry analysis, on one hand, and
141: $0.008 \pm 0.003$ from spectroscopic extinction curve, on the other hand, towards the central $\sim 1$ Mpc
142: of clusters. Assuming a Galactic extinction curve (\citealt{sch98}), this converts to A$_V \sim 0.005 \pm 0.003$
143: and $0.013 \pm 0.004$, respectively. The reference quasars were taken at large projected distances ($> 7$ Mpc)
144: from the cluster centers. Their measurement of average reddening at different radial annuli seems
145: to indicate both a large covering factor and a large scale distribution of dust, up to $5-6$ Mpc.
146:
147: In the present paper, we propose to address the question of the presence of dust in the ICM by studying
148: the statistical reddening of galaxies located in the background of clusters, as compared to a reference
149: sample of galaxies located away from the line of sight to clusters. We have constructed the samples based
150: on the cluster and galaxy catalogs from the Red-sequence Cluster Survey (RCS, \citealt{gla05}), and using
151: galaxy photometric redshifts. In \S2, we briefly describe the RCS data and present our galaxy sample
152: selection and analysis. Our results are discussed in \S3, and a summary is given in \S4.
153:
154:
155: \section{Data and Data Analysis}
156:
157:
158: \subsection{The Red-Sequence Cluster Catalog} \label{RCS}
159:
160: The Red-Sequence Cluster Survey (RCS, \citealt{gla05}) made use of both the Canada-France-Hawaii Telescope
161: (CFHT) and the Cerro Tololo Inter-American Observatory to cover a total sky area of $\sim$90 deg$^{2}$ in
162: the $R_c$ (6500\AA) and $z'$ (9100\AA) bands. We hereby consider only the CFHT RCS data overlapping with
163: the CFHT follow-up observations in the $B$ and $V$ bands presented in \cite{hsi05}, briefly described below
164: in \S\ref{photoz}. The intersection of the $B, V, R_c$ and $z'$ data results in a total of 33.6 deg$^{2}$
165: over 10 different fields (see Table~\ref{RCSsubset}), located at high Galactic latitude ($>$ 40$\degr$)
166: to minimize Galactic extinction.
167:
168: The RCS was originally designed to obtain a large sample of galaxy clusters up to redshift $z \sim 1.4$,
169: using the cluster red-sequence method \citep{gla00}. In this method, clusters are identified as density
170: enhancements in the four-dimensional space of position, color and magnitude. Red-sequence models provide
171: cluster photometric redshifts for which a comparison with spectroscopic data shows that the accuracy is
172: typically better than 0.05 over the redshift range $0.2 < z < 1.0$.
173:
174: In total, the RCS cluster catalog includes $\sim$ 4,000 identified clusters with redshift $0.2 < z < 1.4$,
175: each listed with position, redshift, size and richness [see \cite{gla05}]. The cluster size is defined by
176: the $R_{200}$ radius, within which the average galaxy density is 200 times the critical density. The cluster
177: richness is estimated through the amplitude of the cluster galaxy correlation function Bgc. This parameter
178: basically traces the excess galaxy counts around a reference point (e.g., the center of a cluster), given a
179: luminosity function and a spatial distribution for galaxies, and is known to be a robust quantitative
180: measurement of the cluster richness [see \cite{yee99} for a detailed discussion on the Bgc parameter]. For
181: this study, we extracted only clusters with $z_{clust}$ $< 0.5$ and Bgc $> 200$ (in unit of
182: Mpc$^{1.8}$~$h_{50}^{-1.8}$, where $h_{50} = H_0 /50$~km~s$^{-1}$~Mpc$^{-1}$), yielding a subset catalog
183: of 458 clusters. The typical uncertainty in Bgc for RCS clusters is 150 to 200, and the average Bgg (i.e.,
184: the galaxy-galaxy correlation amplitude computed by picking any random galaxy) is about 60.
185:
186: The RCS cluster catalog is not complete for nearby clusters (i.e., $z < 0.2$), due to the choice of the $R_c$
187: and $z'$ filters, optimized to find clusters at higher redshift. We have used the MaxBCG cluster catalog issued
188: from the SDSS, recently published by \cite{koe07}, to check the apparent surface covered by SDSS nearby clusters
189: ($0.05 < z < 0.35$), with respect to RCS data. 7 out of the 10 RCS patches overlap with the SDSS and we found a
190: total of 41 SDSS clusters within these limits. By fixing a typical size of 5$\arcmin$ (in radius) for each cluster,
191: we estimate that the coverage of these nearby clusters represents a surface area less than 4\% of the corresponding
192: RCS data. We therefore neglect the contamination by nearby ($z < 0.2$) clusters in our study.
193:
194:
195: \subsection{Photometric Redshifts} \label{photoz}
196:
197: Additional photometric data were obtained in the $V$ and $B$ bands with the CFHT for some part of the RCS fields.
198: The intersection of $B, V, R_c$ and $z'$ data covers a total $\sim 33$ deg$^2$, with limiting magnitudes of 23.9
199: in $z'$ (AB), 25.0 in $R_c$ (Vega), 24.5 in $V$ (Vega), and 25.0 in $B$ (Vega). The four-band photometry data,
200: corrected for Galactic extinction, were used by \cite{hsi05} to derive photometric redshifts for $\sim$ 1.2 million
201: galaxies, using an empirical quadratic polynomial fitting technique with spectroscopic redshifts for 4,924 galaxies
202: in the training set. However, since then more spectroscopic data became available for the RCS fields, which can be
203: added to the training set. In particular, the DEEP2 DR2 \footnote{http://deep.berkeley.edu/DR2/} provides 4,297
204: matched spectroscopic redshifts in the range $0.7 < z < 1.5$. The new training set not only contains almost twice
205: as many objects compared to the original one but also provides a much larger sample for $z > 0.7$, i.e., better
206: photometric redshift solution for high redshift objects. Besides having a larger training set, we also used an
207: empirical third-order polynomial fitting technique with 16 Kd-Tree cells for the photometric redshift estimation,
208: allowing a higher redshift accuracy and giving a rms scatter $< 0.05$ for $\delta z_{ph}$ within the redshift
209: range $0.2 < z < 0.5$, and $< 0.09$ for galaxies at $0.0 < z < 1.2$. The refined calibration procedure and
210: photometric redshift method will be described in detail elsewhere (Hsieh et al., in prep.)
211:
212:
213: \subsection{Sample Selection and Analysis} \label{sampleselection}
214:
215: The goal of our study is to measure any statistical color difference between a sample of galaxies seen in the
216: background of clusters and a reference sample of galaxies, presumably not affected by intervening dust. We first
217: extracted only the galaxies with $0.5 < z_{ph}$ $< 0.8$ (i.e, at redshift larger than that of selected clusters)
218: and $\delta z_{ph}/(1+z_{ph}) < 0.06$ (i.e., with good photometric redshift accuracy). A total of $\sim$ 90,000
219: galaxies satisfied these criteria. Each galaxy of our RCS subset is defined by its photometric redshift $z_{ph}$
220: and $B, V, R_c, z'$ magnitudes (with the associated uncertainties $\delta z_{ph}$, $\delta B, \delta V, \delta R_c$
221: and $\delta z'$).
222:
223: Our samples were built by using a Monte Carlo method, in order to take into account the redshift and photometric
224: uncertainties. We repeated the process explained hereafter 100 times. Before checking its line of sight, we
225: replaced any given galaxy by a 'new' galaxy with:
226: \begin{equation}
227: z_{ph} \rightarrow z_{ph} + G(\delta z_{ph}),
228: \end{equation}
229: and
230: \begin{equation}
231: i-j \rightarrow (i + G(\delta i)) - (j + G(\delta j)),
232: \end{equation}
233: where i and j are any two filters taken from ($B, V, R_c, z'$), and G(x) is a random value following a Gaussian
234: distribution with standard deviation x. Two independent colors, $B-R_c$ and $V-z'$, were derived.
235:
236: The line of sight to each galaxy was investigated by checking against the positions of all clusters in our
237: RCS cluster subset. We constructed five samples, containing all galaxies with redshifts such that
238: $z_{ph} - z_{clust} > 0.3$ and located at projected distances $l < 1\times R_{200}$ (sample $a$),
239: $l \in [1-2]\times R_{200}$ (sample $b$), $l \in [2-3]\times R_{200}$ (sample $c$),
240: $l \in [3-4]\times R_{200}$ (sample $d$), and $l \in [6-7]\times R_{200}$ (sample $e$),
241: respectively, from a cluster center.
242: A bootstrap technique (\citealt{efr79}) was applied to rebuild each sample, by randomly picking out N times
243: a galaxy amongst the N galaxies of the sample. Each 'new' sample was further binned in each color with a bin
244: resolution of $\delta$=0.05 magnitude. Fig.\ref{color} illustrates the redshift and color distributions obtained
245: for one given Monte-Carlo+bootstrap realization. We finally checked any color difference between two samples by
246: progressively shifting the color distribution of one relatively to the other by a bin offset $\Delta$
247: ($\Delta$ being an integer), and calculating the value:
248: \begin{equation}
249: \chi (\Delta) = \sum_i ({\rm sample}1(i-\Delta)-{\rm sample}2(i))^2.
250: \end{equation}
251: The color difference $\Delta_{COL}$ between the two samples was then assumed to be the minimum value for
252: $\chi (\Delta)$ (i.e., that for which the two samples are most similar), calculated as:
253: \begin{equation}
254: \Delta_{COL} = \sum_\Delta \frac{\Delta \delta}{\chi^2(\Delta)} / \sum_\Delta \frac{1}{\chi^2(\Delta)}.
255: \end{equation}
256: In practice, we limited the shifts $\Delta$ between $-$10 to +10 bins, thus exploring a color difference
257: range $-$0.5 to +0.5 magnitude. For each Monte Carlo realization, we performed a total of 100 bootstrap runs,
258: which allowed us to derive a statistical dispersion associated with $\Delta_{COL}$.
259:
260: The combination of Monte Carlo + bootstrap techniques allowed us (1) to take into account the redshift and
261: photometric uncertainties, and (2) to derive the average color differences between the samples, with their
262: associated dispersion. To give some numbers, we counted $\sim 8,000$ galaxies in sample $a$, $\sim 23,000$
263: in sample $b$, $\sim 36,000$ in sample $c$, $\sim 47,000$ in sample $d$ and $\sim 73,000$ in sample $e$.
264: The final results for the color differences between the various samples are presented in Table~\ref{results}.
265:
266:
267:
268: \section{Discussion}
269:
270: Our results indicate that there is statistically no color difference between the samples of background galaxies seen
271: behind the inner part of clusters ($<1\times R_{200}$) and in the periphery up to $[6-7]\times R_{200}$. We
272: emphasize that while the color differences are derived from independent filter pairs with independent calibration,
273: the results in columns 2 and 3 of Table~\ref{results} are all consistent with no reddening, given the uncertainties.
274: Assuming a Galactic extinction law, we convert the E($B-R_c$) and E($V-z'$) values into visual extinction, A$_V$,
275: from the following relations: A$_V = 1.9 \times$E($B-R_c$) and A$_V = 1.9 \times$ E($V-z'$) (taking the coefficients
276: from \citealt{sch98}), and calculate the weighted average of the two independent values (column 4 in Table~\ref{results}).
277: We end up with a visual extinction of $<$A$_V$$>$ = $0.004 \pm 0.010$ when comparing the sample of galaxies projected
278: within $1\times R_{200}$ and the reference sample of galaxies projected at $[3-4]\times R_{200}$ away from the cluster
279: centers.
280:
281: The relation between the total mass of intracluster dust, M$_d$, and average visual extinction, A$_V$, within a radius
282: R$_{clust}$ can be expressed as (\citealt{whi92}):
283: \begin{equation}
284: M_d = \frac{16 \pi}{9} \frac{A_V}{1.086} \frac{r_g\rho_g}{Q_{ext}} R_{clust}^2,
285: \end{equation}
286: where the dust grains are characterized by their radius r$_g$ ($\sim 0.1 \mu$m), mass density $\rho_g$
287: ($\sim 1$ g~cm$^{-3}$), and extinction efficiency factor Q$_{ext}$ ($\sim 2$). The values given in parenthesis are
288: characteristic for standard silicate grains (\citealt{hil83}). There is, however, no guarantee that ICM dust grains
289: have similar properties. Should there be dust with such properties and uniformly distributed within $R_{clust}
290: \sim 1.5$ Mpc (roughly corresponding in average to $1\times R_{200}$) from the cluster center, the $3\sigma$ upper
291: limit on A$_V$ would yield a dust mass of M$_d \sim 8 \times 10^9$ M$_\odot$.
292:
293: \cite{che07} have recently claimed the detection of ICM dust by studying both the photometric and spectroscopic
294: properties of background quasars at different projected distances from galaxy clusters, using SDSS data. They
295: report reddening of few $10^{-3}$ magnitudes over large scale up to $[6-7]\times R_{200}$. The slight difference
296: with respect to our results could be attributed either to systematic errors or to the fact that they used
297: relatively closer clusters $0.1 < z < 0.3$. Despite a larger number of background sources in our data, our error
298: bars are about $2-3$ times larger. This could be explained by the fact that the color distribution of quasars is
299: narrower than that of galaxies.
300:
301: Indeed, under Gaussian probability distribution, the uncertainty on the mean can
302: be expressed as $\sigma / \sqrt{N}$, where $\sigma$ is the dispersion of the distribution, and N, the number of
303: elements. The dispersion in color is $\sim 0.5$ for SDSS quasars (see, e.g., \citealt{ric01}), and $\sim 2$ for
304: our galaxies samples (Fig.\ref{color}). Therefore, with N = 8,000 background galaxies in our sample $a$, and
305: N = 3,000 background quasars in \cite{che07} study, we naturally end up with uncertainties $\sim 2.5 \times$
306: larger in our case. The one way to improve our results would then be to significantly increase the number of
307: background galaxies, for example with a larger survey. We note that, within a radius of $1 \times R_{200}$, we
308: get about 20 background galaxies per cluster, whereas \cite{che07} count a total of about 3,000 quasars for their
309: sample of $10^4$ clusters, i.e., in average, less than one background quasar per cluster.
310:
311: Given the uncertainties associated with those measurements, it is still difficult to reach a definite conclusion about
312: the presence of dust in the intracluster medium. All the studies based on the measurement of reddening of background
313: sources (cf. \citealt{mao95, nol03, che07}; and this work) agree with null or very small average reddening, in contrast
314: to previous counts of background clusters or quasars (Table \ref{prevsearch}), which were probably severely affected by
315: selection effects. At infrared or submillimeter wavelengths, on the other hand, only the combination of high angular
316: resolution and sensitivity might help to clarify the claimed detection by \cite{sti98} and \cite{mon05}. Still, the
317: task will not be eased by the large angular size of galaxy clusters, as, e.g., emission from diffuse and extended ICM
318: dust might be filtered out by interferometers.
319:
320:
321: %The associated intensity $I_\nu$ of the dust thermal emission can be calculated as:
322: %\begin{equation}
323: %I_\nu = \mathcal{N}_d B_\nu(T_d) \pi r_g^2 Q_{ext}
324: %\end{equation}
325: %where $\mathcal{N}_d$ is the dust column density and B$_\nu$(T$_d$) the Planck function for the average dust temperature T$_d$.
326: %With T$_d = 20$ K (\citealt{dwe90,yam05}), we derive an upper limit for the 100 $\mu$m intensity of $\sim 8~10^5$ Jy/s
327: %%towards the inner $\sim 1$ Mpc of a cluster,
328: %again assuming that the dust in uniformly distributed within $1 \times R_{200}$.
329: %This value is about one order of magnitude higher than that obtained by \cite{mon05} towards cluster center from the
330: %co-addition of IRAS maps.
331:
332:
333: \section{Summary and Conclusions}
334:
335: In this paper, we report a search for the presence of dust in the intracluster medium. Our method relies on the
336: statistical measurement of the color differences between galaxies located in the background of a cluster and
337: reference galaxies. We take advantage of data acquired with the CFHT in the frame of the Red Sequence Cluster Survey,
338: providing us with deep and homogeneous catalogs of galaxies and galaxy clusters. For our purpose, a major asset of
339: these RCS data is the availability of photometric redshifts for more than a million galaxies. We extracted 458
340: clusters up to a redshift of 0.5, and a total of $\sim 90,000$ galaxies, selected from their photometric redshift
341: $0.5 < z_{ph} < 0.8$ and with photometric redshift uncertainty $\delta z_{ph}/(1+z_{ph}) < 0.06$. In particular, our
342: selection criteria ensure low contamination by foreground galaxies and cluster members. Special attention was given
343: to the propagation of photometry and photometric redshift uncertainties, which were taken into account by a combination
344: of Monte-Carlo and bootstrap methods.
345:
346: A sample of about 8,000 galaxies projected within $1\times R_{200}$ of a cluster is compared to different reference
347: samples composed of galaxies projected at distances [1-2], [2-3], [3-4] and [6-7]$\times R_{200}$ from a cluster center.
348: We do not detect significant color differences between the first sample and the different references. Unless intracluster
349: dust extends over scales larger than $\sim 10$ Mpc in radius, we therefore conclude that the dust extinction is
350: statistically low, with visual extinction $<$A$_V$$>$ = $0.004 \pm 0.010$ for the 458 RCS clusters with $z < 0.5$,
351: corresponding to an upper dust mass limit of $8 \times 10^9$ M$_\odot$ within $1 \times R_{200}$.
352: Given the uncertainties, our results are consistent with the recent measurements of reddening of background quasars
353: by galaxy clusters from \cite{che07}.
354:
355: The on going RCS2 survey, extending the RCS coverage to $\sim 1,000$ deg$^2$ in four photometric bands with the CFHT
356: MegaCam will soon be available. The observed sky surface will be multiplied by a factor of 30, providing a much larger
357: number of background galaxies. These new data could allow us to improve our measurements and compare
358: the dust reddening towards clusters with different properties, such as richness, X-ray brightness or merging activity.
359:
360:
361: \acknowledgements
362: Based on observations obtained at the Canada-France-Hawaii Telescope (CFHT) which is operated by the
363: National Research Council of Canada, the Institut National des Sciences de l'Univers of the Centre
364: National de la Recherche Scientifique of France, and the University of Hawaii.
365:
366:
367: \begin{thebibliography}{}
368:
369: \bibitem[Annis \& Jewitt(1993)]{ann93}{Annis, J. \& Jewitt, D., 1993, \mnras, 264, 593}
370: \bibitem[Arnaboldi et al.(1996)]{arn96}{Arnaboldi, M., Freeman, K. C., Mendez, R. H., et al., 1996, \apj, 472, 145}
371: %\bibitem[Bai et al.(2006)]{bai06}{Bai, L., Rieke, G. H., Rieke, M. J., et al., 2006, \apj, 639, 827}
372: \bibitem[Bai et al.(2007)]{bai07}{Bai, L., Rieke, G. H. \& Rieke, M. J., 2007, \apj, 668, L5}
373: \bibitem[Bernstein et al.(1995)]{ber95}{Bernstein, G. M., Nichol, R. C., Tyson, J. A., et al., 1995, \aj, 110, 1507}
374: \bibitem[Bogart \& Wagoner(1973)]{bog73}{Bogart, R. S., \& Wagoner, R. V., 1973, \apj, 181, 609}
375: \bibitem[Boyle et al.(1988)]{boy88}{Boyle, B. J., Fong, R., \& Shanks, T., 1988, \mnras, 231, 897}
376: \bibitem[Bregman et al.(1990)]{bre90}{Bregman, J. N., McNamara, B. R. \& O'Connell, R. W., 1990, \apj, 351, 406}
377: \bibitem[Chelouche et al.(2007)]{che07}{Chelouche, D., Koester, B. P. \& Bowen D. V., 2007, \apjl, 671, L97}
378: \bibitem[Cox et al.(1995)]{cox95}{Cox, C. V., Bregman, J. N. \& Schombert, J. M., 1995, \apjs, 99, 405}
379: \bibitem[Domainko et al.(2006)]{dom06}{Domainko, W., Mair, M., Kapferer, W., et al., 2006, \aap, 452, 795}
380: \bibitem[Draine \& Salpeter(1979)]{dra79}{Draine, B., \& Salpeter, E. E. 1979, \apj, 231, 77}
381: \bibitem[Dwek et al.(1990)]{dwe90}{Dwek, E., Rephaeli, Y., \& Mather, J. C., 1990, \apj, 350, 104}
382: \bibitem[Edge et al.(1999)]{edg99}{Edge, A. C., Ivison, R. J., Smail, I., Blain, A. W., \& Kneib, J.-P., 1999, \mnras, 306, 599}
383: \bibitem[Efron(1979)]{efr79}{Efron, B., 1979, The Annals of Statistics, Vol. 7, No. 1, Page 1-26}
384: \bibitem[Feldmeier et al.(1998)]{fel98}{Feldmeier, J. J., Ciardullo, R. \& Jacoby, G. H., 1998, \apj, 503, 109}
385: \bibitem[Feldmeier et al.(2002)]{fel02}{Feldmeier, J. J., Mihos, J. C., Morrison, H. L., et al., 2002, \apj, 575, 779}
386: \bibitem[Ferguson et al.(1998)]{fer98}{Ferguson, H. C., Tanvir, N. R. \& von Hippel, T., 1998, \nat, 391, 461}
387: \bibitem[Ferguson(1993)]{fer93}{Ferguson, H. C., 1993, \mnras, 263, 343}
388: \bibitem[Gal-Yam et al.(2003)]{gal03}{Gal-Yam, A., Maoz, D., Guhathakurta, P. \& Filippenko, A. V., 2003, \aj, 125, 1087}
389: \bibitem[Geach et al.(2006)]{gea06}{Geach, J. E., Smail, I., Ellis, R. S., et al., 2006, \apj, 649, 661}
390: \bibitem[Girardi et al.(1992)]{gir92}{Girardi, M., Mezzetti, M., Giuricin, G. \& Mardirossian, F., 1992, \apj, 394, 442}
391: \bibitem[Gladders \& Yee(2000)]{gla00}{Gladders, M. D. \& Yee, H. K. C., 2000, \aj, 120, 2148}
392: \bibitem[Gladders \& Yee(2005)]{gla05}{Gladders, M. D. \& Yee, H. K. C., 2005, \apjs, 157, 1}
393: \bibitem[Gonz\'alez et al.(1998)]{gon98}{Gonz\'alez, R. A., Allen, R. J.; Dirsch, B., et al., 1998, \apj, 506, 152}
394: \bibitem[Grabelsky \& Ulmer(1990)]{gra90}{Grabelsky, D. A. \& Ulmer, M. P., 1990, \apj, 355, 401}
395: \bibitem[Hildebrand(1983)]{hil83}{Hildebrand, R.H., 1983, QJRAS 24, 267}
396: \bibitem[Holwerda et al.(2005)]{hol05}{Holwerda, B. W., Gonz\'alez, R. A., Allen, R. J. \& van der Kruit, P. C., 2005, \aj, 129, 1396}
397: \bibitem[Hsieh et al.(2005)]{hsi05}{Hsieh, B. C., Yee, H. K. C., Lin, H. \& Gladders, M. D., 2005, \apjs, 158, 161}
398: \bibitem[Hu(1992)]{hu92}{Hu, E. M., 1992, \apj, 391, 608}
399: \bibitem[Karachentsev \& Lipovetskii(1969)]{kar69}{Karachentsev, I. D., \& Lipovetskii, V. A., 1969, Soveit Phys., 12, 909}
400: \bibitem[Koester et al.(2007)]{koe07}{Koester, B. P., McKay, T. A., Annis, J., et al., 2007, \apj, 660, 239}
401: \bibitem[Maoz(1995)]{mao95}{Maoz, D., 1995, \apj, 455, L115}
402: \bibitem[Mitchell et al.(1976)]{mit76}{Mitchell, R. J., Culhane, J. L., Davison, P. J. N. \& Ives, J. C., 1976, \mnras, 175, 29}
403: \bibitem[Montier \& Giard(2005)]{mon05}{Montier, L. A. \& Giard, M., 2005, \aap, 439, 35}
404: \bibitem[Mushotzky et al.(1996)]{mus96}{Mushotzky, R., Loewenstein, M., Arnaud, K. A., et al., 1996, \apj, 466, 686}
405: \bibitem[Nollenberg et al.(2003)]{nol03}{Nollenberg, J. G., Williams, L. L. R. \& Maddox, S., 2003, \aj, 125, 2927}
406: \bibitem[Popescu et al.(2000)]{pop00}{Popescu, C. C., Tuffs, R. J., Fischera, J \& V\"{o}lk, H., 2000, \aap, 354, 480}
407: \bibitem[Quillen et al.(1999)]{qui99}{Quillen, A. C., Rieke, G. H., Rieke, M. J., Caldwell, N. \& Engelbracht, C. W., 1999, \apj, 518, 632}
408: %\bibitem[Rieke \& Lebofsky(1985)]{rie85}{Rieke, G. H. \& Lebofsky, M. J., 1985, \apj, 288, 618}
409: \bibitem[Richards et al.(2001)]{ric01}{Richards, G. T., Fan, X., Schneider, D. P., et al., 2001, \aj, 121, 2308}
410: \bibitem[Romani \& Maoz(1992)]{rom92}{Romani, R. W. \& Maoz, D., 1992, \apj, 386, 36}
411: \bibitem[Schindler et al.(2005)]{sch05}{Schindler, S., Kapferer, W., Domainko, W., et al., 2005, \aap, 435, L25}
412: \bibitem[Schlegel et al.(1998)]{sch98}{Schlegel, D. J., Finkbeiner, D. , P. \& Davis, M., 1998, \apj, 500, 525}
413: \bibitem[Stickel et al.(1998)]{sti98}{Stickel, M., Lemke, D., Mattila, K., Haikala, L. K. \& Haas, M., 1998, \aap, 329, 55}
414: \bibitem[Stickel et al.(2002)]{sti02}{Stickel, M., Klaas, U., Lemke, D. \& Mattila, K., 2002, \aap, 383, 367}
415: \bibitem[White et al.(1991)]{whi91}{White, D. A., et al., 1991, \mnras, 252, 72}
416: \bibitem[Whittek(1992)]{whi92}{Whittek, D.C.B, 1992, {\em Dust in the Galactic Environment}, Institute of Physics Publishing, Bristol, 1992}
417: \bibitem[Wise et al.(1993)]{wis93}{Wise, M. W., O'Connell, R. W., Bregman, J. N. \& Roberts, M. S., 1993, \apj, 405, 94}
418: \bibitem[Xilouris et al.(2006)]{xil06}{Xilouris, E., Alton, P., Alikakos, J., et al., 2006, \apj, 651, L107}
419: \bibitem[Yamada \& Kitayama(2005)]{yam05}{Yamada, K. \& Kitayama, T., 2005, \pasj, 57, 611}
420: \bibitem[Yee \& Lopez-Cruz(1999)]{yee99}{Yee, H. K. C., \& Lopez-Cruz, O., 1999, \aj, 117, 1985}
421: \bibitem[Zaritsky(1994)]{zar94}{Zaritsky, D., 1994, \aj, 108, 1619}
422: \bibitem[Zibetti et al.(2005)]{zib05}{Zibetti, S.; White, S. D. M., Schneider, D. P. \& Brinkmann, J., 2005, \mnras, 358, 949}
423: \bibitem[Zwicky(1951)]{zwi51}{Zwicky, F., 1951, \pasp, 63, 61}
424: \bibitem[Zwicky(1962)]{zwi62}{Zwicky, F., 1962, in Problems of Extragalactic Research, er. G. C. McVittie (New York: Macmillan), p. 149}
425:
426:
427: \end{thebibliography}
428:
429:
430:
431: \input{tab1.tex}
432:
433: \input{tab2.tex}
434:
435: \input{tab3.tex}
436:
437:
438:
439: \begin{figure}
440: \includegraphics[width=8cm]{f1.eps}
441: \caption{Comparison of the redshift and color distributions of galaxies in sample $a$ ($< 1 \times$R$_{200}$;
442: {\em thick grey curve}) and sample $d$ ($[3-4]\times$R$_{200}$; {\em black curve}) for a given Monte Carlo
443: realization. All distributions were normalized to the total number of galaxies in each sample.}
444: \label{color}
445: \end{figure}
446:
447:
448: \end{document}
449:
450: