1: \documentclass[]{emulateapj}
2:
3:
4: \newcounter{address}
5: \newcommand{\target}{target}
6: \newcommand{\control}{control}
7: \newcommand{\Ftarget}{\ensuremath{f_{\mathrm{\target}}}}
8: \newcommand{\Fcontrol}{\ensuremath{f_{\mathrm{\control}}}}
9: \newcommand{\Ntarget}{\ensuremath{\mathcal{N}_{\mathrm{\target}}}}
10: \newcommand{\Ncontrol}{\ensuremath{\mathcal{N}_{\mathrm{\control}}}}
11: \newcommand{\Rtarget}{\ensuremath{\mathcal{R}_{\mathrm{\target}}}}
12: \newcommand{\Rcontrol}{\ensuremath{\mathcal{R}_{\mathrm{\control}}}}
13: \newcommand{\Ha}{\ensuremath{\mathrm{H}\alpha}}
14: \newcommand{\composite}{composite}
15: \newcommand{\EBV}{\ensuremath{E(B-V)}}
16: \newcommand{\Q}{\ensuremath{Q}}
17: \newcommand{\sigv}{\ensuremath{\sigma_v}}
18: \newcommand{\mingals}{\ensuremath{\mathcal{N}_{\mathrm{gal}}^{\mathrm{min}}}}
19: \newcommand{\tauv}{\ensuremath{\tau_V}}
20: \newcommand{\ICM}{ICM}
21: \newcommand{\invvar}{invvar}
22: \newcommand{\template}{template}
23: \newcommand{\average}{average}
24: \newcommand{\averaging}{averaging}
25: \newcommand{\averaged}{averaged}
26: \newcommand{\SNR}{SNR}
27:
28: \renewcommand{\cite}[1]{\citeauthor{#1} \citeyearpar{#1}}
29:
30: \begin{document}
31:
32: \slugcomment{Astrophys.~J.~in press}
33:
34:
35: \title{The Transparency of Galaxy Clusters}
36: \shorttitle{Transparency of Galaxy Clusters}
37:
38: \author{
39: Jo~Bovy\altaffilmark{1},
40: David~W.~Hogg\altaffilmark{1,2}, and
41: John~Moustakas\altaffilmark{1}}
42: \shortauthors{Bovy et al.}
43:
44: %\setcounter{address}{1}
45: \altaffiltext{1}{Center for Cosmology and Particle Physics, Department of Physics, New York
46: University, 4 Washington Place, New York, NY 10003}
47: \altaffiltext{2}{To whom correspondence should be addressed: \texttt{david.hogg@nyu.edu}}
48:
49:
50: \begin{abstract}
51:
52:
53: If galaxy clusters contain intracluster dust, the spectra of galaxies
54: lying behind clusters should show attenuation by dust absorption. We
55: compare the optical (3500 - 7200 \AA) spectra of 60,267 luminous,
56: early-type galaxies selected from the Sloan Digital Sky Survey to
57: search for the signatures of intracluster dust in $z\sim\!0.05$ clusters.
58: We select massive, quiescent (i.e., non-star-forming) galaxies using an
59: EW(\Ha) $\leq 2$ \AA \ cut and consider galaxies in three bins of velocity
60: dispersion, ranging from 150 to 300 km s$^{-1}$. The uniformity of early-type
61: galaxy spectra in the optical allows us to construct
62: inverse-variance-weighted \composite\ spectra with high signal-to-noise
63: ratio (ranging from $10^2-10^3$). We compare the \composite\ spectra of
64: galaxies that lie behind and adjacent to galaxy clusters and find no
65: convincing evidence of dust attenuation on scales $\sim\! 0.15-2$ Mpc;
66: we derive a generic limit of $\EBV < 3 \times 10^{-3}$ mag on
67: scales $\sim\! 1-2$ Mpc at the 99\% confidence level, using conservative
68: jackknife error bars, corresponding to a dust mass
69: $\lesssim 10^8$ $M_{\odot}$. On scales smaller than 1 Mpc this limit is
70: slightly weaker, $\EBV < 8 \times 10^{-3}$ mag.
71:
72:
73:
74: \end{abstract}
75:
76:
77: \keywords{dust, extinction --- galaxies: clusters: general ---
78: intergalactic medium --- methods: statistical}
79:
80:
81: \section{Introduction}
82:
83: Galaxy clusters are known to contain galaxies and hot gas, and they may
84: contain extragalactic but intracluster dust, or be accreting intergalactic
85: dust from their neighborhoods. Indeed, there must be some intracluster dust
86: created by winds from the intracluster stars, which make up a significant
87: fraction of the total stellar mass in the cluster \citep{Ferguson98}.
88: Alternatively, dust could be introduced into the intracluster medium (ICM)
89: through such processes as cooling-flows \citep{Fabian94}, galaxy or cluster
90: mergers and collisions \citep{Popescu:2000qs}, supernovae-driven galactic
91: winds \citep{Okazaki93}, ram pressure stripping of galaxies as they travel
92: through the intracluster medium \citep{Gunn72} and accretion of primordial
93: dust \citep{Popescu:2000qs}. Many of these processes have associated
94: timescales of order 10$^8$-10$^9$ yr. A crucial question is then whether
95: the dust thus injected into the intracluster medium can survive thermal
96: sputtering in the hot gas. Typical dust grain sputtering timescales are
97: $\tau_{\mbox{sp}} \sim\! 10^6-10^9$ yr \citep{Draine79}, similar to the
98: timescales of the dust-producing processes. These timescales imply that
99: only the most recently injected dust is still surviving at any given moment
100: in time, from which we conclude that the amount of dust in the \ICM\ should
101: be small and non-uniformly distributed.
102:
103: Measurements of intracluster dust have a long and rich
104: history. The presence of dust was first hypothesized to explain the
105: discrepancy between counts of galaxies located behind and adjacent to the
106: Coma cluster \citep{Zwicky57}. A first estimate of 0.4 mag for the
107: magnitude of the $B$-band extinction was suggested
108: \citep[][1962]{Zwicky61}, although infrared emission of dust in the Coma
109: Cluster was not detected \citep{Dwek90}. Zwicky's method was improved over
110: the years as catalogs of galaxy clusters became available and $B$-band
111: extinctions of order 0.2 mag were reported based on various procedures:
112: using essentially the same approach as Zwicky, but
113: using a larger sample of 15 galaxy clusters \citep{Karachentsev};
114: considering color residuals to arrive at the amount of absorption
115: within the Local Supercluster \citep{Vaucouleurs}; looking at angular
116: correlations among clusters and quasars, leading to an extinction of
117: about 0.12 mag over radii of several Mpc \citep{Bogart}; considering
118: correlations of high-redshift quasars with low-redshift galaxies, which
119: gave evidence for dust in clusters at redshift z $\sim\! 0.15$ at a
120: characteristic linear radius of $500 h^{-1}$ kpc, corresponding to a
121: dust sphere of mass 10$^{10}$ $M_\odot$ \citep{Boyle}. However, galaxy
122: number counts are subject to a variety of biases and the dearth of
123: galaxies behind clusters could have other causes than
124: dust \citep{Nollenberg:2003cd}.
125:
126: Correlations of quasars with nearby clusters were reconsidered and a
127: $B$-band extinction of 0.15 mag was found \citep{Romani:1991nd}. A similar
128: result was found to explain an excess of higher redshift galaxies in nearby
129: small galaxy groups \citep{Girardi1992}. However, comparing the color
130: distribution of quasars behind a cluster with those in the vicinity of
131: the cluster limited the relative reddening of the two samples to
132: $\EBV \lesssim 0.05$ mag \citep{Maoz:1995yk}. Similar limits were
133: obtained comparing color distributions of galaxies behind and removed
134: from APM clusters \citep[on 1.3 Mpc scales]{Nollenberg:2003cd} and
135: using large, elliptical galaxies \citep{Ferguson93}.
136:
137: There have been some, but contradictory, reports on dust in the central
138: regions of clusters. A study of {\it IRAS} images of 56 clusters found
139: two clusters with far-infrared color excesses that could be due to
140: 10$^9$ $M_{\odot}$ of dust \citep{Wise93}. An average excess reddening in
141: 10 cooling-flow galaxy clusters of $\EBV \sim\! 0.19$ mag was reported for
142: lines-of-sight to the center of these clusters \citep{Hu92}. However, a
143: later report found no convincing evidence of submillimeter dust emission
144: in 11 cooling-flow clusters and set an upper limit of 10$^8$ $M_\odot$ on
145: the total mass of the dust \citep{Annis93}.
146:
147: More recently, observations of six Abell clusters found a rough estimate of
148: a dust mass of 10$^7$ $M_\odot$ in the Coma cluster, but no evidence of
149: dust in the other five observed clusters \citep{Stickel:2001nt}; and no
150: significant amount of infrared emission from intracluster dust in Abell
151: 2029 was found \citep{Bai:2007kp}. \cite{Chelouche:2007rm} reported
152: reddening in a $0.1 < z < 0.3 $ sample of $\sim\! 10^4$ galaxy clusters by
153: correlating the Sloan Digital Sky Survey cluster and quasar catalog and by
154: comparing photometric and spectroscopic properties of quasars behind
155: the clusters to those in the field. They found mean $\EBV$ values of a
156: few $\times 10^{-3}$ mag for sight lines passing $\sim\!$ Mpc from the
157: clusters' centers. However, a recent study found no evidence of dust in
158: $0.2 < z < 0.5$ clusters from a photometric study of color excesses in
159: several bands and, assuming a Galactic extinction law, derived an average
160: visual extinction of $\langle A_V \rangle = 0.004 \pm 0.010$ mag
161: \citep{Muller:2008hg}.
162:
163: In this paper we study the dust content of galaxy clusters by
164: comparing the spectra of galaxies behind clusters of galaxies with
165: those of galaxies not behind clusters. More specifically, we use the
166: optical (3500 - 7200 \AA) spectra of luminous, early-type
167: galaxies because they are known to dominate the stellar mass density
168: of the Universe \citep{Fukugita:1997bi,Hogg:2002ci} and show great
169: regularities in their properties \citep[e.g.][2003b, 2003c,
170: 2003d]{Oke68,Faber73,Visvanathan77,Djorgovski:1987vx,
171: Dressler:1987ny,Kormendy:1989dg,Bower:1992jx,Roberts:1994gp,
172: Bernardi03a}. Their spectra show a remarkable similarity with any
173: variation that does exist explained by the environment and luminosity
174: \citep{Eisenstein:2002ta}. Any dust-attenuation-like difference that
175: can be found between the composite spectra of galaxies behind galaxy
176: clusters and galaxies in the field can be attributed reliably to
177: interactions of the galaxies' light with the intracluster medium.
178:
179:
180: The difference between our study and the other precise studies cited above,
181: is that (1) we have more control over our galaxy population, (2) we are
182: selecting on properties that only weakly involve color and are
183: therefore less likely to be biased, and (3) we have well-calibrated spectrophotometry
184: of all objects. On the other hand, these considerations limit the size
185: of our sample, so what we gain in per-object precision we lose in
186: number of objects, in some sense.
187: Despite our relatively small sample, we obtain among the
188: most stringent upper limits ever.
189:
190:
191: In what follows, AB magnitudes are used throughout, a cosmological world model with
192: $(\Omega_{\mbox{m}},\Omega_\Lambda) = (0.3,0.7)$ is adopted, and the
193: Hubble constant is $H_0 = 70$ km s$^{-1}$ Mpc$^{-1}$ \citep{Komatsu}, for the purposes
194: of calculating distances \citep[e.g.][]{Hogg:1999ad}.
195:
196:
197:
198:
199:
200:
201: \section{Data}
202:
203: The Sloan Digital Sky Survey (SDSS) is obtaining \emph{u,g,r,i} and \emph{z}
204: CCD imaging of 10$^4$ deg$^2$ of the northern Galactic sky, and from that
205: imaging, selecting roughly 10$^6$ targets for spectroscopy, most of them
206: galaxies with $r < 17.77$ mag
207: \citep[][2004, 2005]{Gunn:1998vh,York:2000gk,Stoughton:2002ae,Abazajian:2003jy}.
208:
209: All the data processing, including astrometry \citep{Pier:2002iq},
210: source identification, deblending and photometry
211: \citep{Lupton:2001zb}, calibration \citep{Fukugita:1996qt,
212: Smith:2002pca, Ivezic:2004bf}, spectroscopic target selection
213: \citep{Eisenstein:2001cq, Strauss:2002dj, Richards:2002bb},
214: spectroscopic fiber placement \citep{Blanton:2001yk}, spectral data
215: reduction and analysis (Schlegel \& Burles 2006, in preparation;
216: Schlegel 2006, in preparation) are performed with automated SDSS
217: software.
218:
219: We use the spectroscopic and photometric catalog from the NYU Value Added
220: Galaxy Catalog \citep[NYU-VAGC;][]{Blanton2005} compiled from the SDSS Data
221: Release Four \citep[DR4;][]{AdelmanMcCarthy:2005se}.
222:
223: As we will be \averaging\ the spectra of galaxies coming from different
224: spectral ``plates'', we depend on the calibration of these fluxes. The
225: calibration procedure is as follows
226: \citep[D.~J. Schlegel, in preparation;][]{Stoughton:2002ae}: Every spectral
227: ``plate'' of fiber positions includes several faint (15.5-18.5 mag) F8
228: subdwarf stars. The spectrum of each standard star is spectrally typed by
229: comparing with a grid of theoretical spectra generated from Kurucz model
230: atmospheres \citep{Kurucz1992} using the spectral synthesis code SPECTRUM
231: \citep{Gray1994,Gray2001}. The spectra are calibrated with these F star spectra; i.e.
232: they are multiplied by the function of wavelength that makes the F star
233: spectra match the F star spectrophotometry (after correcting for Galactic reddening).
234: This calibration procedure produces consistent calibration at the
235: 5\% level\protect\footnote{$\sim\!4.4$\% in $g$-$r$ and $\sim\!2.8$\% in $r$-$i$ for galaxies, see
236: \url{http://www.sdss.org/dr4/products/spectra/spectrophotometry.html}}.
237: In addition to this, the SDSS does not use an atmospheric refraction corrector,
238: so the effective fiber position on the sky shifts slightly as a function of
239: wavelength. In the presence of brightness gradients, this creates a fluxing
240: error.
241:
242: Redshifts are measured on the reduced spectra by an automated system, which
243: models each galaxy spectrum as a linear combination of stellar eigenspectra
244: (D.~J. Schlegel, in preparation). The central velocity dispersion \sigv\ is
245: determined by fitting the detailed spectral shape as a velocity-smoothed sum
246: of stellar spectra (D.~J. Schlegel \& D.~P. Finkbeiner, in preparation).
247:
248: The measurements of the equivalent width (EW) of the \Ha\ line is measured
249: exactly as described in \cite{Quintero:2003we}. Briefly, a linear fit of the
250: spectral section to a linear combination of the mean SDSS old galaxy spectrum
251: and the mean SDSS A-star spectrum with the locations of possible emission
252: lines marked out is performed; this best fit model is then scaled down to
253: have the same flux continuum as the data in the vicinity of the \Ha\ emission
254: line and subtracted to leave a continuum-subtracted line spectrum; the \Ha\
255: line flux is then measured in a 20 \AA \ width interval centered on the line
256: and converted to a rest-frame equivalent width with a continuum found by
257: taking the inverse-variance-weighted average of two sections of the spectrum
258: about 150 \AA \ in size and on either side of the emission line. This method
259: fairly accurately models the absorption trough in the continuum, although in
260: detail it leaves small negative residuals.
261:
262: The galaxy clusters used here, are $0.015 < z < 0.067$ member clusters
263: taken from a friends-of-friends cluster catalog constructed from the
264: SDSS DR3 main sample galaxies with absolute magnitudes $M_{0.1r}$ $<$
265: -19.9 mag \citep{Berlind:2006bx}. We first consider a \mingals = 10
266: member minimum, corresponding to a total absolute r-band magnitude
267: M$_r \leq -21.9$; for reference, the Virgo cluster contains $13$
268: galaxies brighter than this limit \citep{Trentham:2002mq}. However,
269: we also vary the minimum number of galaxies in the cluster between
270: \mingals = 5 (M$_r \leq -20.8$ mag) and \mingals = 20 (M$_r \leq -23.1$ mag).
271:
272:
273: \section{Analysis}
274:
275:
276:
277:
278:
279: \subsection{Sample Construction}
280:
281: Our technique is analogous to the classic ``foreground screen'' test
282: of using stars to measure the Galactic extinction curve
283: \citep[e.g.,][and references therein]{Calzetti01}. Here, we compare
284: the spectra of galaxies that lie behind a cluster of galaxies,
285: i.e. galaxies whose light has had to traverse a galaxy cluster on its
286: way to observational astronomers, against the spectra of galaxies that
287: do not lie behind a cluster. Galaxies that lie behind a galaxy
288: cluster will constitute the \emph{\target}\ galaxies and galaxies that
289: are not behind any cluster will form the \emph{\control}\ sample. The
290: optical spectra of these objects must be intrinsically similar so that
291: we can ascribe any measured differences to dust attenuation as it
292: passes through a cluster of galaxies; massive, early-type galaxies are
293: known to have exactly this property \citep{Eisenstein:2002ta}.
294:
295:
296: %\clearpage
297: \begin{figure}%
298: \begin{center}
299: \includegraphics[width=0.25\textwidth,clip=]{f1a.eps}%
300: \includegraphics[width=0.25\textwidth,clip=]{f1b.eps}\\
301: \includegraphics[width=0.25\textwidth,clip=]{f1c.eps}%
302: \includegraphics[width=0.25\textwidth,clip=]{f1d.eps}%
303: \caption{Properties of galaxies in the SDSS sample: Plots of the quantities relevant for the sample selection. From left to right, top to bottom: equivalent width of the \Ha\ line (EW(\Ha)) vs. velocity dispersion \sigv; \sigv\ vs. redshift; $g$-$r$ color vs. EW(\Ha); $r$-band absolute magnitude vs. \sigv.}%
304: \label{galprop}
305: \end{center}
306: \end{figure}
307:
308:
309: %\clearpage
310: \begin{deluxetable*}{cccccc}
311: %\begin{deluxetable}{ccccrrc}
312: \tablecaption{Properties of the various samples\label{tablesamples}}
313: \tablecolumns{6}
314: \tablewidth{0pt}
315: \tablehead{
316: \colhead{Sample$^a$} & \colhead{EW(\Ha)} & \colhead{\sigv} & \colhead{\mingals$\,^b$} & \colhead{\Rtarget$^c$} & \colhead{\Rcontrol$^d$}\\
317: \colhead{} & \colhead{(\AA)} & \colhead{(km s$^{-1}$)} & \colhead{} & \colhead{(Mpc)} & \colhead{(Mpc)}}
318: \startdata
319: Primary & $\leq$ 2 & $200 \leq \sigv \leq 250$ & 10 & 0.50 & 2 \\
320: $150 \leq \sigv \leq 200$ & $\leq$ 2 & $150 \leq \sigv \leq 200$ & 10 & 0.50 & 2\\
321: $250 \leq \sigv \leq 300$ & $\leq$ 2 & $250 \leq \sigv \leq 300$ & 10 & 0.50 & 2\\
322: \mingals = 5 & $\leq$ 2 & $200 \leq \sigv \leq 250$ & 5 & 0.50 & 2\\
323: \mingals = 20 & $\leq$ 2 & $200 \leq \sigv \leq 250$ & 20 & 0.50 & 2\\
324: \Rtarget = 0.25 Mpc & $\leq$ 2 & $200 \leq \sigv \leq 250$ & 10 & 0.25 & 2\\
325: \Rtarget = 1 Mpc & $\leq$ 2 & $200 \leq \sigv \leq 250$ & 10 & 1.00 & 2
326: \enddata
327: \tablenotetext{a}{The redshift range for all these samples is $0.1 \leq z \leq 0.2$.}
328: \tablenotetext{b}{Minimum number of members to define a galaxy cluster.}
329: \tablenotetext{c}{Galaxies within a transverse distance \Rtarget\ of a cluster (and are at a larger redshift) are considered behind that cluster and make up the target subsample.}
330: \tablenotetext{d}{Galaxies more than \Rcontrol\ from every cluster make up the control subsample.}
331: \end{deluxetable*}
332:
333:
334:
335:
336: Luminous, early-type galaxies are part of the red sequence of galaxies
337: and can be identified in many different ways. Here, we select galaxies
338: from the SDSS sample based on two properties derived from their
339: optical spectra: the equivalent width of the \Ha\ emission line,
340: EW(\Ha), and the stellar velocity dispersion, \sigv. We select
341: quiescent (i.e., non-star-forming) galaxies using EW(\Ha)$<2$~\AA. In
342: addition, we restrict the sample to the redshift range $0.1 \leq z
343: \leq 0.2$. Figure \ref{galprop} shows plots of EW(\Ha), \sigv, color,
344: and absolute magnitude of relevant galaxies in the SDSS sample.
345:
346: The galaxy clusters used to define our \target\ and
347: \control\ subsamples were taken from a friends-of-friends cluster
348: catalog \citep{Berlind:2006bx}. We emphasize that this catalog is not
349: complete and that we only use part of the catalog to select our
350: \target\ and \control\ galaxies. In addition to this, our galaxy spectra
351: come from SDSS DR4, which has a larger coverage than SDSS DR3, out of which
352: the cluster catalog was constructed. We expect the number of ``false
353: negatives'', i.e. galaxies that might be catalogued as
354: \control\ galaxies that are actually behind a cluster, to be small.
355: In addition, because of the large number of galaxies in the
356: \control\ sample, the effect of any contaminating or misclassified
357: galaxies will be diluted in the stacking procedure. Finally, we will
358: see that the principal source of errors will be due to the
359: \target\ galaxies, for which this effect is not important.
360:
361: We can now specify when a galaxy is part of the \target\ subsample of
362: galaxies and when it belongs to the \control\ subsample. For each
363: galaxy cluster, every galaxy at a redshift exceeding that of the
364: cluster that is found to be within 0.5 Mpc transverse distance of it,
365: is considered to be behind the cluster. We define \Rtarget = 0.5 Mpc,
366: which we will vary to investigate the radial dependence of any
367: measured effect. Galaxies that are more than \Rcontrol\ = 2 Mpc
368: removed from every galaxy cluster, are classified as
369: \control\ galaxies. Any galaxy between these bounds is excluded from
370: the analysis.
371:
372:
373: %\clearpage
374: \begin{figure}%
375: \centering
376: \includegraphics[width=0.25\textwidth]{f2a.eps}%
377: \includegraphics[width=0.25\textwidth]{f2b.eps}\\
378: \includegraphics[width=0.25\textwidth]{f2c.eps}%
379: \includegraphics[width=0.25\textwidth]{f2d.eps}%
380: \caption{Galaxy properties for the primary sample: Plots of the quantities relevant for the sample selection for the primary sample (see Table \ref{tablesamples} for the definition of the primary sample). The \target\ subsample is represented by diamonds $\diamond$, the \control\ subsample by dots. From left to right, top to bottom: velocity dispersion vs. EW(\Ha); redshift vs. EW(\Ha); velocity dispersion vs. redshift; angular distribution of the \target\ and \control\ subsamples: RA vs. sin(DEC).}%
381: \label{mainprop}
382: \end{figure}
383:
384:
385:
386: Based on these criteria a \emph{primary} sample with \target\ and
387: \control\ subsamples was constructed as follows: galaxies with EW(\Ha)
388: $\leq 2$ \AA, $200 \leq \sigv \leq 250$ km s$^{-1}$ and $0.1 \leq z
389: \leq 0.2$ were classified as \target\ or \control\ based on the values
390: \Rtarget\ = 0.5 Mpc and \Rcontrol\ = 2 Mpc, using groups from the
391: cluster catalog with a minimum of 10 members. Figure \ref{mainprop}
392: shows plots of several of these properties for the primary sample as
393: well as their distribution on the celestial sphere. This figure
394: illustrates that the distribution of \target\ and \control\ galaxies
395: is approximately uniform in these properties. Several secondary
396: samples were also considered, based on variations of some of these
397: properties, i.e. \sigv, \mingals\ and \Rtarget\ (see
398: Table~\ref{tablesamples} for an overview of the properties of the
399: various samples used).
400:
401: Added together, the total number of galaxy spectra
402: that we consider in this paper is 60,267. This number includes the number
403: of galaxies in the primary and secondary samples, broken down in
404: Table~\ref{tableresults} by subsample (note, however, that there is a
405: significant overlap of used spectra between certain samples), as well as the
406: sizes of the \target\ subsamples used in the determination of the radial
407: dependency of dust-attenuation (see below).
408:
409:
410:
411: \newpage
412: \subsection{Stacking Procedure and Comparison}
413:
414:
415: Since we are creating \composite\ spectra from individual spectra that
416: are intrinsically very similar, we will weight the pixels of each
417: spectrum with their inverse variance, as this will
418: give us the highest signal-to-noise \composite\ spectrum. However,
419: the overall normalization of each spectrum is determined by its
420: distance from us and therefore the spectra are only alike when properly
421: normalized. This can be achieved by comparing each spectrum to a common
422: \template. This comparison consists
423: of a simple, one-parameter fit to determine the scale factor between
424: the \template\ and the spectrum to be added, which is used to rescale
425: the spectrum before adding it to the weighted \average.
426:
427:
428: %\clearpage
429: \begin{figure*}%
430: \centering
431: \includegraphics[width=0.3\textwidth]{f3a.eps}%
432: \includegraphics[width=0.3\textwidth]{f3b.eps}%
433: \includegraphics[width=0.3\textwidth]{f3c.eps}%
434: \caption{Convergence of the \average\ of the \control\ spectra for the different velocity dispersion bins: From left to right: $200 \leq \sigv \leq 250$, $150 \leq \sigv \leq 200$, and $250 \leq \sigv \leq 300$.}%
435: \label{mainconvergence}
436: \end{figure*}
437:
438:
439:
440: When stacking the spectra in the \control\ subsample, such a \template\
441: is not available a priori. Therefore, we have designed a procedure to
442: self-consistently create this \template\ from the constituent spectra,
443: using a simple, iterative algorithm. To start off, the spectra are
444: averaged as given, using inverse-variance-weighting, with no individual rescaling
445: applied. In the next step, this first \average\ features as the \template\
446: which is used to scale each spectrum before adding it. The resultant \average\
447: becomes the new \template\ in the next iteration. This continues until
448: the difference between successive iterations, defined as the maximum
449: fractional difference over the spectrum, is less than $\sim\!1$ part in
450: 10$^{-6}$. Convergence is achieved after about 5 iterations and can be
451: quite dramatic (see Fig.~\ref{mainconvergence}, which shows this
452: convergence for the primary sample and the other velocity
453: dispersion bins). The upshot of this iterative procedure is that
454: the resulting \average\ is essentially the same as the \template\ used to
455: compute it, i.e.~the \composite\ \control\ spectrum is obtained using itself
456: as template. Finally, the \composite\ spectrum is
457: smoothed using a Gaussian filter with a velocity width equal to the
458: maximum velocity dispersion of the sample.
459:
460: We use this \composite\ \control\ spectrum as the template when
461: obtaining the \composite\ \target\ spectrum, since the spectra in these
462: subsamples are intrinsically very similar. Indeed, the \composite\
463: \control\ spectrum is the best \template\ to use in the sum of the \target\
464: spectra, as it is a very high signal-to-noise representation of each
465: \target\ spectrum, and setting the overall normalization of each \target\
466: spectrum equal to that of the \composite\ \control\ spectrum will ensure that
467: small wavelength-dependent, attenuation-like differences will be the primary
468: difference between the \composite\ \target\ spectrum and the \composite\
469: \control\ spectrum, i.e.~the overall normalization of the \composite\
470: spectrum of the \target\ spectra will be very close to that of the \composite\
471: \control\ spectrum. The \composite\ \target\ spectrum obtained in this way
472: is smoothed using the same Gaussian filter as used on the \composite\
473: \control\ spectrum.
474:
475:
476: Finally, we fit to
477:
478: \begin{equation}\label{eqcomp}
479: \Ftarget(\lambda) = \Q \Fcontrol(\lambda) \ \mathrm{e}^{-\tau(\lambda)} ,
480: \end{equation}
481:
482: \noindent where \Ftarget\ and \Fcontrol\ is the \composite\ spectrum of the
483: \target\ and \control\ subsamples, respectively, and \Q\ is an arbitrary
484: scale factor. Following \cite{Charlot00}, we parameterize the dust
485: attenuation law as
486:
487: \begin{equation}
488: \tau(\lambda) = \tauv \Big(\frac{\lambda}{5500 \mbox{\AA}} \Big)^{-\alpha} \ ,
489: \end{equation}
490:
491: \noindent where \tauv\ is the $V$-band optical depth. As in
492: \cite{Charlot00}, we adopt $\alpha = 0.7$, which is a reasonable
493: approximation to the shape of the Milky Way optical extinction curve.
494: Note that our parameterization ignores the possibility that the
495: emission of the cluster itself is influencing our results; however,
496: since we expect the cluster light to be similar to the light of early-type
497: galaxies, this should not bias our conclusions.
498:
499:
500:
501:
502:
503:
504: \subsection{Error Estimation}
505:
506: The error on the individual pixels of the \average\ spectra
507: follows immediately from the stacking procedure. The
508: inverse variance of an inverse-variance-weighted \average\ is given by
509: the sum of the individual weights. The \composite\
510: \target\ spectra obtained have a median signal-to-noise ratio (\SNR) of
511: $\sim\!200$, while the larger \control\ subsamples lead to
512: \composite\ spectra with a median \SNR\ of $\sim\!2500$. Therefore, it is
513: clear that the main source of error is due to the \composite\ \target\
514: spectrum.
515:
516: The error $\sigma_{\tauv}$ on the value of \tauv\ is obtained by a jackknife
517: procedure. In general, a jackknife estimate of the variance of a statistic is
518: obtained by dividing the sample into a number of (equal size) subsamples and
519: obtaining the relevant statistic for each of these subsamples. The estimate
520: of the variance of the statistic is then approximately equal to the variance
521: of the values obtained for the subsamples \citep[with a proportionality constant
522: that depends on the number of subsamples, which rapidly converges to unity as
523: the number of subsamples increases; ][]{Efron}. Theoretically, this estimate is obtained
524: from subsamples created by leaving out one of the ``datapoints'' (in our
525: case, a ``datapoint'' is the spectrum of a galaxy, consisting of many
526: individual points); however, due to computational constraints, this
527: calculation is not always feasible. A possible way of dealing with this
528: limitation is by dividing the sample into a number of subsets, based on a
529: property that is unrelated to the relevant statistic, and creating jackknife
530: subsamples as unions of all but one of these subsets \citep{Shao}.
531:
532: Whenever a full-fledged jackknife estimate was deemed
533: too computationally intensive, an equal number of quantiles in
534: declination were chosen to subdivide the sample, and a minimum of 200
535: jackknife subsamples was used to calculate errors in all these cases. Since
536: the error is mostly due to inaccuracies in the \composite\ \target\ spectrum,
537: jackknife subsamples were only created using the \target\ subsample; i.e.
538: for the purpose of the jackknife procedure, the \composite\ \control\ spectrum
539: was supposed to be known exactly.
540:
541:
542: As a check on the validity of our error estimates, we performed a careful
543: examination of the error estimates for the primary sample. Jackknife error
544: estimates were obtained using different numbers of subsamples, ranging from
545: ten to the maximum, 110, the number of galaxies in the
546: primary \target\ subsample, see below. Additionally, we implemented a
547: bootstrap procedure, which works much in the same way as the jackknife
548: procedure, but creates bootstrap samples from randomly picking
549: ``datapoints'' with replacement from the set of, in this case, \target\ spectra. Values up
550: to 300 for the number of bootstrap samples were used. All jackknife and
551: bootstrap estimates agreed on the first two significant figures of the error.
552:
553:
554:
555: \subsection{Algorithm tests}\label{secmock}
556:
557: In order to examine the consistency of the stacking procedure and the
558: estimation of the error on \tauv, we have designed two algorithm-tests,
559: which also tell us about the precision with which we can perform our measurement.
560: First, we tested the accuracy and precision with which our stacking algorithm
561: could recover a known value of the reddening, \tauv. We selected a random
562: subset of 200 objects from the \control\ subsample of the primary sample
563: and reddened them with a value of \tauv\ = 0.025 with Gaussian noise of
564: standard deviation 0.027 (i.~e., mimicking the result for the primary sample,
565: see below). We found that our stacking algorithm retrieved a value
566: of \tauv\ = 0.024, with a 1$\sigma$-error of 0.038. Similar results were found
567: for different input-reddening values, however, the errors computed using the
568: jackknife procedure were consistently larger, but of the same magnitude as the
569: variation that went in.
570:
571: To better simulate the actual parameter estimation for the primary sample,
572: a larger random sample of 8000 spectra from the \control\ subsample was chosen
573: to be a mock-\control-subsample and a randomly chosen subsample of this
574: of one hundred spectra was artificially reddened to provide the
575: mock-\target-subsample. Several orders of magnitude of reddening were
576: tried and we found that for values of \tauv\ of $\sim\!10^{-1}$ and
577: Gaussian noise of the same magnitude, our algorithm returns the exact
578: amount of reddening and error, which means that our algorithm should be
579: able to detect reddening of this magnitude in the real samples.
580: Values of \tauv\ of order 10$^{-2}$ were recovered by our algorithm as
581: well, however, the jackknife estimate of the error is consistently larger
582: than the variation that was put in. We were unable to recover values of
583: \tauv\ of order 10$^{-3}$ and noise of the same magnitude. The computed
584: error in this case is still of the order 10$^{-2}$, which indicates that
585: this is a lower bound set by measurement errors. Any errors obtained that
586: are larger than this, are likely due to an intrinsic variation in dust
587: absorption among clusters, while errors of this size and smaller are
588: consistent with being true measurement errors.
589:
590:
591:
592: \section{Results}
593:
594: %\clearpage
595: \begin{figure}
596: \centering
597: \plotone{f4.eps}
598: \caption{Result for the primary sample: \average\ and comparison for the subsamples of the primary sample: the top panel shows the inverse-variance-weighted \average\ spectrum of the galaxies in the \control\ subsample, multiplied by the wavelength; the middle panel shows the same for the galaxies in the \target\ subsample; the bottom panel shows a comparison of these two \composite\ spectra (which should be exactly equal to zero if there were no dust absorption in galaxy clusters and the \averaged\ spectra were exact), and a fit to a standard dust law (see the text for more details on this procedure). The bottom panel shows error bars (the top curve is an upper bound and the bottom curve a lower bound). Errors on the \composite\ spectra in both the top panel and the middle panel are of the order of the line thickness.}%
599: \label{mainresult}
600: \end{figure}
601:
602:
603:
604:
605:
606: %\clearpage
607: \begin{deluxetable}{ccccccc}[b]
608: %\begin{deluxetable}{crrrrrrrr}
609: \tablecaption{Main results for the various samples\label{tableresults}}
610: \tablecolumns{5}
611: \tablewidth{0pt}
612: \tablehead{\colhead{Sample} & \colhead{\Ntarget} & \colhead{\Ncontrol} & \colhead{\tauv} & \colhead{$\sigma_{\tauv}$}}
613: \startdata
614: Primary & 110 & 21468 & 0.025 & 0.027\\
615: $150 \leq \sigv \leq 200$ & 122 & 26531 & 0.045 & 0.029\\
616: $250 \leq \sigv \leq 300$ & 32 & 7239 & -0.047 & 0.053\\
617: \mingals = 5 & 425 & 17993 & 0.027 & 0.014\\
618: \mingals = 20 & 29 & 22562 & -0.012 & 0.059\\
619: \Rtarget = 0.25 Mpc & 31 & 21468 & 0.037 & 0.039\\
620: \Rtarget = 1 Mpc & 392 & 21468 & -0.001 & 0.013
621: \enddata
622: \tablecomments{Results of the comparison of the \average\ of the \target\ galaxies and the \average\ of the \control\ galaxies for the primary sample and the secondary samples. \Ntarget\ gives the number of galaxies in the \target\ subsample; \Ncontrol\ gives the number of galaxies in the \control\ subsample; \tauv\ is the best fit parameter to the dust law (eq. \ref{eqcomp}); 1$\sigma$-errors on \tauv\ are given in the $\sigma_{\tauv}$ column.}
623: \end{deluxetable}
624:
625:
626: Figure \ref{mainresult} shows the result of \averaging\ the spectra in the
627: \control\ subsample (top panel) and of \averaging\ the spectra in the
628: \target\ subsample (middle panel) for the primary sample. The lower
629: panel shows a difference plot of these two quantities, with the fractional difference
630: defined as
631:
632: \begin{equation}
633: \frac{\Ftarget - \Q \Fcontrol}{\Q \Fcontrol} \ ,
634: \end{equation}
635:
636: \noindent which for the dust attenuation in equation (\ref{eqcomp})
637: equals e$^{-\tau(\lambda)} - 1$. A fit to equation (\ref{eqcomp}) gives
638: the value of \tauv, which together with the sizes of the samples
639: is given in Table \ref{tableresults}. Only the calculated error on the
640: difference in the lower panel is shown here. This error is substantial and
641: the error in \tauv\ is, likewise, not negligible (see Table
642: \ref{tableresults} for the error on \tauv).
643:
644:
645:
646: %\clearpage
647: \begin{figure*}%
648: \centering
649: \includegraphics[height=0.3\textwidth]{f5a.eps}%
650: \includegraphics[height=0.3\textwidth]{f5b.eps}%
651: \includegraphics[height=0.3\textwidth]{f5c.eps}\\
652: \includegraphics[height=0.3\textwidth]{f5d.eps}%
653: \includegraphics[height=0.3\textwidth]{f5e.eps}%
654: \includegraphics[height=0.3\textwidth]{f5f.eps}%
655: \caption{Results for the secondary samples: Each of these figures shows for the various secondary samples what was shown in Figure \ref{mainresult} for the primary sample. From left to right, top to bottom: $150 \leq \sigv \leq 200$; $250 \leq \sigv \leq 300$; \mingals\ = 5; \mingals\ = 20; \Rtarget\ = 0.25 Mpc; \Rtarget\ = 1 Mpc.}%
656: \label{sampleresults}
657: \end{figure*}
658:
659:
660:
661: %\clearpage
662: \begin{figure*}%
663: \centering
664: \includegraphics[height=0.3\textwidth]{f6a.eps}%
665: \includegraphics[height=0.3\textwidth]{f6b.eps}\\
666: \includegraphics[height=0.3\textwidth]{f6c.eps}%
667: \includegraphics[height=0.3\textwidth]{f6d.eps}%
668: \caption{Radial dependence of the dust attenuation: radial dependence of the parameter \tauv\ in the dust law (eq. \ref{eqcomp}) for three velocity dispersion bins and their combined signal: top-left: $200 \leq \sigv \leq 250$; top-right: $150 \leq \sigv \leq 200$; bottom-left: $250 \leq \sigv \leq 300$; bottom-right: average of the three velocity dispersion bins. $R$ sets \Rtarget; \mingals\ for each of these bins is 10, other parameters as for the primary sample. Note that the error bars in each plot are all covariant, since each datapoint is the cumulative signal from all galaxies inside \Rtarget.}%
669: \label{radialresults}
670: \end{figure*}
671:
672:
673: Figure \ref{sampleresults} gives the same analysis as Figure
674: \ref{mainresult} for the secondary samples (see also Table
675: \ref{tableresults}). Inspection of this figure and the results in the
676: table shows that for smaller \target\ subsamples, the errors are
677: significantly larger.
678:
679: We computed the radial dependence of \tauv\ for the
680: various velocity dispersion bins by varying the value of \Rtarget\ for
681: these samples, while keeping the value of \Rcontrol\ and other parameters
682: fixed. The result is shown in Figure \ref{radialresults}. The plots shown are
683: cumulative in the sense that each value of \Rtarget\ gives dust attenuation
684: for dust within a volume of radius \Rtarget. The range between 0.15 and 1 Mpc
685: was examined more carefully by increasing \Rtarget\ in steps of 0.05 Mpc,
686: whereas between 1 and 2 Mpc, steps of 0.1 Mpc were used. For smaller values
687: of \Rtarget\ the \target\ subsample consists of only a few galaxies, for which
688: the \composite\ spectrum cannot be reliably obtained.
689:
690:
691:
692:
693: \section{Discussion}
694:
695:
696: The lower panel of Figure \ref{mainresult} shows a value of the
697: difference that is nearly flat over the whole wavelength range, and the
698: 1-sigma error estimate on the value of \tauv\ in Table \ref{tableresults}
699: confirms that a null value is within the uncertainties. The other
700: samples confirm this result: Most of the samples give results that are
701: well within 1$\sigma$ of the null hypothesis, with a few of the samples
702: giving a formally negative value for \tauv, i.e. a negative
703: absorption. None of the values of \tauv\ are statistically significant
704: indicators of positive or negative absorption. Regarding the
705: negative absorptions, it must be remarked that the two most negative
706: values are obtained from relatively small \target\ subsample sizes,
707: about 30 spectra in the \target\ subsample (they both have the largest
708: error values as well), which could account for an estimate that is
709: significantly off. The other negative value is essentially zero and occurs
710: for the \Rtarget\ = 1 Mpc sample, which could
711: simply indicate that, generically, at this radial distance there is no
712: dust in galaxy clusters (see below for a discussion of the radial
713: dependence).
714:
715: The dust law used in equation (\ref{eqcomp}) is related to the extinction by
716: \begin{equation}
717: A(\lambda) = \tauv \times \frac{2.5}{\ln 10} \Big(\frac{\lambda}{5500 \mbox{\AA}} \Big)^{-0.7} \mbox{mag} \ ,
718: \end{equation}
719: which gives values of \EBV\ for the \tauv\ values obtained of the
720: order of 10$^{-3}$ mag. Our estimate of the typically error on
721: $\tauv$, $\sim0.03$, sets an upper bound on \EBV\ of $\sim\! 5 \times$
722: 10$^{-3}$ mag for values of \Rtarget\ = 0.5 Mpc. Comparing these errors to
723: what we found for our mock-samples in Section \ref{secmock}, we see that the
724: obtained errors are consistent with being true measurement errors. The most
725: positive value of \tauv\ is found for the $150 \leq \sigv \leq 200$ sample,
726: which gives a reddening of \EBV\ = ($8 \pm 5$) $\times$ 10$^{-3}$ mag. The most
727: statistically significant value of \tauv\ is found for the \mingals = 5
728: sample, with a value of \tauv\ that is 2$\sigma$
729: from the null result, corresponding to \EBV\ = ($5.0 \pm 2.5$) $\times$
730: 10$^{-3}$ mag. These are still not very significant, but they are
731: remarkable in that they are obtained for samples that have both a large
732: \target\ and a large \control\ subsample.
733:
734:
735: The radial dependence plots are the most instructive of the resulting
736: plots as they might reveal the radial location of the dust content of
737: galaxy clusters. If there were dust in a certain radial distance
738: range, we would expect the \tauv\ vs. \Rtarget\ plot to be essentially
739: zero up to the dust range, after which a sharp increase would occur
740: over the range in which the dust occurs, followed by a gradual decline
741: for larger values of \Rtarget\ as more and more unattenuated spectra
742: are added to the \target\ subsample. For example, to confirm the
743: results of \cite{Chelouche:2007rm}, who found evidence of dust around
744: 1 Mpc with no dust at smaller radii, we would expect to see a peak
745: around 1 Mpc.
746:
747:
748: In the range $< 1$ Mpc we do not find a consistent result in the three
749: velocity dispersion bins we considered. The errors in this range are large
750: because of the small number of galaxies in the \target\ subsamples of these
751: bins ($\sim\!100$ for \Rtarget$\sim\! 0.5$ Mpc). The value of \tauv\ in the
752: range $150 \leq \sigv \leq 250$ is positive; however, \tauv\ is in the
753: negative range in the $250 \leq \sigv \leq 300$ bin. The amount of
754: absorption rises in the interval between \Rtarget\ = 0.3 Mpc and
755: \Rtarget\ = 0.7 Mpc for the $150 \leq \sigv \leq 200$ bin, but the opposite
756: happens for the $200 \leq \sigv \leq 250$ bin. In the bottom-right panel of
757: Figure \ref{radialresults}, we see that the combined result of the three
758: bins is essentially flat within the error range.
759:
760: The results are more consistent in the range between 1 and 2 Mpc. All of the
761: bins show a small amount of dust absorption, but all the values
762: contain the null value in their error range. The overall significance of the
763: result is only slightly smaller than the significance of an individual point,
764: because of strong correlations between different \Rtarget\ values (since this
765: is a cumulative plot, the overlap between the different samples
766: is significant). Combining the results in the 1-2 Mpc range, we find an
767: average extinction \EBV\ = 0.002 mag with a significance of 1.5$\sigma$. At
768: the 99\% confidence level, we conclude that \EBV\ $< 3 \times 10^{-3}$ mag.
769:
770: A similar analysis for \Rtarget\ $\sim\!0.5$ Mpc gives an average extinction
771: \EBV\ = 0.004 mag, with a significance of 1.2$\sigma$. Therefore, in this
772: range we can derive a limit \EBV\ $< 8 \times 10^{-3}$ mag. Both of these
773: upper bounds are more stringent than the ones previously found.
774:
775:
776: We can translate an upper bound on the extinction into an upper bound on the
777: dust mass using \citep{Krugel03}
778:
779: \begin{equation}
780: M_{\mathrm{dust}} = 1.5 \times 10^8 \frac{\EBV}{3 \times 10^{-3}
781: \mathrm{ \ mag}}\Big( \frac{\mathcal{R}}{1 \mathrm{\ Mpc}} \Big)^2
782: M_\odot \ ,
783: \end{equation}
784:
785: \noindent which gives an upper bound $M_{\mathrm{dust}} \lesssim 10^8
786: M_\odot$ for $\sim\!$ Mpc-scales.
787:
788: The bottom-right panel of Figure \ref{radialresults} summarizes our results:
789: over the whole range we considered, \Rtarget\ $\sim\!0.15 - 2$ Mpc, we find a
790: signal that is essentially flat and consistent with zero. Our upper limit on
791: \EBV\ for distances between 1 and 2 Mpc from the center of a cluster is a
792: few 10$^{-3}$ mag, which is consistent with the amount of dust extinction
793: observed in $0.1 < z < 0.3$ clusters in \cite{Chelouche:2007rm}. Future work
794: that could significantly increase the size of the \target\ subsample could
795: lower the upper bound found here, or confirm the existence of dust on the
796: outskirts of galaxy clusters.
797:
798: \acknowledgements It is a pleasure to thank Mike Blanton for helpful
799: discussions and the anonymous referee for valuable comments. This
800: project made use of the NASA Astrophysics Data System, and the
801: idlutils codebase maintained by David Schlegel and others. Financial
802: support for this project was provided by the US National Aeronautics
803: and Space Administration (LTSA grant NAG5-11669, ADP grant
804: 07-ADP07-0099, GALEX grant 06-GALEX06-0030, and Spitzer grant
805: G05-AR-50443). During part of the period in which this research was
806: performed, DWH was a research fellow of the Alexander von Humboldt
807: Foundation of Germany.
808:
809: Funding for the SDSS and SDSS-II has been provided by the Alfred P. Sloan
810: Foundation, the Participating Institutions, the National Science Foundation,
811: the U.S. Department of Energy, the National Aeronautics and Space
812: Administration, the Japanese Monbukagakusho, the Max Planck Society, and the
813: Higher Education Funding Council for England. The SDSS Web Site is
814: \url{http://www.sdss.org/}.
815:
816: The SDSS is managed by the Astrophysical Research Consortium for the
817: Participating Institutions. The Participating Institutions are the American
818: Museum of Natural History, Astrophysical Institute Potsdam, University of
819: Basel, University of Cambridge, Case Western Reserve University, University
820: of Chicago, Drexel University, Fermilab, the Institute for Advanced Study,
821: the Japan Participation Group, Johns Hopkins University, the Joint Institute
822: for Nuclear Astrophysics, the Kavli Institute for Particle Astrophysics and
823: Cosmology, the Korean Scientist Group, the Chinese Academy of Sciences
824: (LAMOST), Los Alamos National Laboratory, the Max-Planck-Institute for
825: Astronomy (MPIA), the Max-Planck-Institute for Astrophysics (MPA), New
826: Mexico State University, Ohio State University, University of Pittsburgh,
827: University of Portsmouth, Princeton University, the United States Naval
828: Observatory, and the University of Washington.
829:
830:
831:
832: \begin{thebibliography}{}
833:
834:
835: %\cite{Abazajian:2003jy}
836: \bibitem[Abazajian et~al.~2003]{Abazajian:2003jy}
837: Abazajian, K.~et~al.~2003,
838: %``The First Data Release of the Sloan Digital Sky Survey,''
839: AJ, 126, 2081
840: [arXiv:astro-ph/0305492]
841: %%CITATION = ANJOA,126,2081;%%
842:
843:
844: %\cite{Abazajian:2004aja}
845: \bibitem[{Abazajian et al.~} {2004}]{Abazajian:2004aja}
846: Abazajian, K.~et~al.~2004,
847: %``The Second Data Release of the Sloan Digital Sky Survey,''
848: AJ, 128, 502
849: [arXiv:astro-ph/0403325]
850: %%CITATION = ANJOA,128,502;%%
851:
852: %\cite{Abazajian:2004it}
853: \bibitem[{Abazajian et al.~} {2005}]{Abazajian:2004it}
854: Abazajian, K.~et al. 2005,
855: %``The Third Data Release of the Sloan Digital Sky Survey,''
856: AJ, 129, 1755
857: [arXiv:astro-ph/0410239]
858: %%CITATION = ANJOA,129,1755;%%
859:
860:
861: %\cite{AdelmanMcCarthy:2005se}
862: \bibitem[Adelman-McCarthy et~al.~2006]{AdelmanMcCarthy:2005se}
863: Adelman-McCarthy, J.~K.~et~al.~2006,
864: %``The Fourth Data Release of the Sloan Digital Sky Survey,''
865: ApJS, 162, 38
866: [arXiv:astro-ph/0507711]
867: %%CITATION = APJSA,162,38;%%
868:
869:
870: %\cite{Annis93}
871: \bibitem[Annis \& Jewitt 1993]{Annis93}
872: Annis, J.~\& Jewitt, D.~1993,
873: %A Search for Cold Dust in Clusters of Galaxies with Cooling Flows}
874: MNRAS, 264, 593
875:
876:
877: %\cite{Bai:2007kp}
878: \bibitem[Bai et al.~2007]{Bai:2007kp}
879: Bai, L., Rieke, G.~H., \& Rieke, M.~J.~2007,
880: %``Search for IR Emission from Intracluster Dust in A2029,''
881: AJ, 668, L5
882: [arXiv:0708.3406]
883: %%CITATION = ARXIV:0708.3406;%%
884:
885:
886: %\cite{Berlind:2006bx}
887: \bibitem[{Berlind et al.} {2006}]{Berlind:2006bx}
888: Berlind, A.~A. et~al. 2006,
889: %``Percolation Galaxy Groups and Clusters in the SDSS Redshift Survey:
890: %Identification, Catalogs, and the Multiplicity Function,''
891: ApJS, 167, 1
892: [arXiv:astro-ph/0601346]
893: %%CITATION = APJSA,167,1;%%
894:
895:
896: %\cite{Bernardi03a}
897: \bibitem[Bernardi et al. 2003a]{Bernardi03a} Bernardi, M. et al.\
898: 2003a, AJ, 125, 1817
899:
900: %\cite{Bernardi03b}
901: \bibitem[Bernardi et al. 2003b]{Bernardi03b} Bernardi, M. et al.\
902: 2003b, AJ, 125, 1849
903:
904: %\cite{Bernardi03c}
905: \bibitem[Bernardi et al. 2003c]{Bernardi03c} Bernardi, M. et al.\
906: 2003c, AJ, 125, 1866
907:
908: %\cite{Bernardi03d}
909: \bibitem[Bernardi et al. 2003d]{Bernardi03d} Bernardi, M. et al.\
910: 2003d, AJ, 125, 1882
911:
912:
913: %\cite{Blanton:2001yk}
914: \bibitem[{Blanton et~ al.} {2003a}]{Blanton:2001yk}
915: Blanton, M.~R., Lupton, R.~H., Miller Malley, F., Young, N., Zehavi, I., \& Loveday, J. 2003a,
916: %``An Efficient algorithm for positioning tiles in the Sloan Digital Sky
917: %Survey,''
918: AJ, 125, 2276
919: [arXiv:astro-ph/0105535]
920: %%CITATION = ANJOA,125,2276;%%
921:
922:
923: %\cite{Blanton:2002wv}
924: \bibitem[Blanton et al.(2003b)]{Blanton:2002wv}
925: Blanton, M.~R. et al. 2003b,
926: %``The Galaxy Luminosity Function and Luminosity Density at Redshift z=0.1,''
927: ApJ, 592, 819
928: [arXiv:astro-ph/0210215]
929: %%CITATION = ASJOA,592,819;%%
930:
931:
932: %\cite{Blanton2005}
933: \bibitem[{Blanton et~al.} {2005}]{Blanton2005}
934: Blanton, M.~R. et~al. 2005,
935: % title = ``{New York University Value-Added Galaxy Catalog: A Galaxy Catalog Based on New Public Surveys}'',
936: AJ, 129, 2562
937: [arXiv:astro-ph/0410166]
938:
939:
940:
941: %\cite{Bogart}
942: \bibitem[Bogart et al. 1973]{Bogart}
943: Bogart, R.~S. \& Wagoner, R.~V. 1973,
944: %Clustering Effects among Clusters of Galaxies and Quasi-Stellar Sources
945: AJ, 181, 609
946:
947:
948: %\cite{Boyle}
949: %\harvarditem[Boyle et~al.]{Boyle, Fong, Shanks}{1990}{Boyle}
950: \bibitem[Boyle et~al. 1988]{Boyle}
951: Boyle, B.~J., Fong, R., \& Shanks, T. 1988,
952: %On the correlation of UVX QSOs with galaxies
953: MNRAS, 231, 897
954:
955:
956:
957: %\cite{Bower:1992jx}
958: \bibitem[Bower et~al. 1992]{Bower:1992jx}
959: Bower, R.~G., Lucey, J.~R., \& Ellis, R.~S. 1992,
960: %``Precision Photometry of Early Type Galaxies in the Coma and Virgo Clusters:
961: %A Test of the Universality of the Colour / Magnitude Relation. 2. Analysis,''
962: MNRAS, 254, 601.
963: %%CITATION = MNRAA,254,601;%%
964:
965:
966:
967:
968: %\cite{Calzetti01}
969: \bibitem[Calzetti 2001]{Calzetti01}
970: Calzetti, D. 2001,
971: %title = ``{The Dust Opacity of Star-forming Galaxies}'',
972: \pasp, 113, 1449 [arXiv:astro-ph/0109035]
973:
974:
975: %\cite{Charlot00}
976: \bibitem[Charlot \& Fall(2000)]{Charlot00}
977: Charlot, S. \& Fall, S.~M. 2000,
978: %title = ``{A Simple Model for the Absorption of Starlight by Dust in Galaxies}'',
979: \apj, 539, 718 [arXiv:astro-ph/0003128]
980:
981:
982: %\cite{Chelouche:2007rm}
983: \bibitem[Chelouche et al.(2007)]{Chelouche:2007rm}
984: Chelouche, D., Koester, B.~P., \& Bowen, D.V. 2007
985: % ``The Dust Content of Galaxy Clusters,''
986: ApJ, 671, L97
987: [arXiv:0711.1167]
988: %%CITATION = ARXIV:0711.1167;%%
989:
990:
991:
992: %\cite{Djorgovski:1987vx}
993: \bibitem[Djorgovski \& Davis 1987]{Djorgovski:1987vx}
994: Djorgovski, S. \& Davis, M. 1987
995: %``Fundamental Properties Of Elliptical Galaxies,''
996: ApJ, 313, 59
997: %%CITATION = ASJOA,313,59;%%
998:
999:
1000:
1001: %\cite{Draine79}
1002: \bibitem[Draine \& Salpeter 1979]{Draine79}
1003: Draine, B.~T. \& Salpeter, E.~E. 1979,
1004: %On the physics of dust grains in hot gas
1005: ApJ, 231, 77
1006:
1007:
1008: %\cite{Dressler:1987ny}
1009: \bibitem[Dressler et~al. 1987]{Dressler:1987ny}
1010: Dressler, A., Lynden-Bell, D., Burstein, D., Davies, R.~L., Faber, S.~M., Terlevich, R., \& and Wegner G. 1987,
1011: %``Spectroscopy and photometry of elliptical galaxies. 1. A New distance
1012: %estimator,''
1013: ApJ, 313, 42
1014: %%CITATION = ASJOA,313,42;%%
1015:
1016:
1017:
1018:
1019: %\cite{Dwek90}
1020: \bibitem[Dwek et~al. 1990]{Dwek90}
1021: Dwek, E., Rephaeli, Y., \& Mather, J.~C. 1990,
1022: %Infrared emission from dust in the Coma cluster of galaxies
1023: ApJ, 350, 104
1024:
1025:
1026: %\cite{Efron}
1027: \bibitem[Efron \& Tibshirani 1993]{Efron}
1028: Efron, B.~\& Tibshirani, R.~1993, An Introduction to the Bootstrap (Boca Raton: CRC Press)
1029:
1030:
1031:
1032: %\cite{Eisenstein:2001cq}
1033: \bibitem[Eisenstein et~al. 2001]{Eisenstein:2001cq}
1034: Eisenstein, D.~J. et al. 2001,
1035: %``Spectroscopic Target Selection for the Sloan Digital Sky Survey: The
1036: %Luminous Red Galaxy Sample,''
1037: AJ, 122, 2267
1038: [arXiv:astro-ph/0108153]
1039: %%CITATION = ANJOA,122,2267;%%
1040:
1041:
1042: %\cite{Eisenstein:2002ta}
1043: \bibitem[Eisenstein et~al. 2003]{Eisenstein:2002ta}
1044: Eisenstein, D.~J. et al. 2003,
1045: %``Average spectra of massive galaxies in the SDSS,''
1046: ApJ, 585, 694
1047: [arXiv:astro-ph/0212087]
1048: %%CITATION = ASJOA,585,694;%%
1049:
1050:
1051: %\cite{Faber73}
1052: \bibitem[Faber 1973]{Faber73}
1053: Faber, S.~M. 1973,
1054: %Variations in Spectral-Energy Distributions and Absorption-Line Strengths among Elliptical Galaxies,
1055: ApJ, 179, 731
1056:
1057: %\cite{Fabian94}
1058: \bibitem[Fabian et~al. 1994]{Fabian94}
1059: Fabian, A.~C., Johnstone, R.~M., \& Daines, S.~J. 1994,
1060: %title = ``{The Effects of Dust in Cold Clouds Embedded in Cooling Flows}'',
1061: \mnras, 271, 737
1062:
1063: %\cite{Ferguson93}
1064: \bibitem[Ferguson 1993]{Ferguson93}
1065: Ferguson, H.~C. 1993,
1066: %A Test for Dust in Clusters of Galaxies
1067: MNRAS, 263, 343
1068:
1069: %\cite{Ferguson98}
1070: \bibitem[Ferguson et~al. 1998]{Ferguson98}
1071: Ferguson, H.~C., Tanvir, N.~R., \& von Hippel, T. 1998,
1072: % title = ``{Detection of intergalactic red-giant-branch stars in the Virgo cluster}'',
1073: \nat, 391, 461 [arXiv:astro-ph/9801228]
1074:
1075:
1076: %\cite{Fukugita:1996qt}
1077: \bibitem[Fukugita et al. 1996]{Fukugita:1996qt}
1078: Fukugita, M., Ichikawa, T., Gunn, J.~E., Doi, M., Shimasaku, K., \& Schneider, D.~P. 1996,
1079: %``The Sloan digital sky survey photometric system,''
1080: AJ, 111, 1748
1081: %%CITATION = ANJOA,111,1748;%%
1082:
1083: %\cite{Fukugita:1997bi}
1084: \bibitem[Fukugita et~al. 1998]{Fukugita:1997bi}
1085: Fukugita, M., Hogan, C.~J., \& Peebles, P.~J.~E. 1998,
1086: %``The Cosmic Baryon Budget,''
1087: ApJ, 503, 518
1088: [arXiv:astro-ph/9712020].
1089: %%CITATION = ASJOA,503,518;%%
1090:
1091:
1092:
1093: %\cite{Girardi1992}
1094: \bibitem[{Girardi et~al.} {1992}]{Girardi1992}
1095: Girardi, M., Mezzetti, M., Giuricin, G., \& Mardirossian, F. 1992,
1096: % title = ``{Redshift asymmetry and color-velocity correlation in nearby
1097: %galaxy groups - Evidence of dust?}'',
1098: \apj, 394, 442
1099:
1100: %\cite{Gray1994}
1101: \bibitem[Gray \& Corbally 1994]{Gray1994}
1102: Gray, R.~O.~\& Corbally, C.~J.~1994, AJ, 107, 742
1103:
1104: %\cite{Gray2001}
1105: \bibitem[Gray et~al.~2001]{Gray2001}
1106: Gray, R.~O., Graham, P.~W., \& Hoyt, S.~R.~2001, AJ, 121, 2159
1107:
1108:
1109: %\cite{Gunn72}
1110: \bibitem[Gunn \& Gott 1972]{Gunn72}
1111: Gunn, J.~E. \& Gott, J.~R.~I. 1972,
1112: %title = ``{On the Infall of Matter Into Clusters of Galaxies and Some Effects on Their Evolution}'',
1113: \apj, 176, 1
1114:
1115:
1116: %\cite{Gunn:1998vh}
1117: \bibitem[Gunn et al. 1998]{Gunn:1998vh}
1118: Gunn, J.~E. et al. 1998,
1119: %``The Sloan digital sky survey photometric camera,''
1120: AJ, 116, 3040
1121: [arXiv:astro-ph/9809085]
1122: %%CITATION = ANJOA,116,3040;%%
1123:
1124:
1125: %\cite{Hogg:1999ad}
1126: \bibitem[Hogg 1999]{Hogg:1999ad}
1127: Hogg, D.~W. 1999
1128: %``Distance measures in cosmology,''
1129: [arXiv:astro-ph/9905116]
1130: %%CITATION = ASTRO-PH/9905116;%%
1131:
1132:
1133:
1134: %\cite{Hogg:2002ci}
1135: \bibitem[Hogg et~al. 2002]{Hogg:2002ci}
1136: Hogg, D.~W. et al. 2002,
1137: %``The luminosity density of red galaxies,''
1138: AJ, 124, 646
1139: [arXiv:astro-ph/0204436]
1140: %%CITATION = ANJOA,124,646;%%
1141:
1142:
1143: %\cite{Hu92}
1144: \bibitem[Hu 1992]{Hu92}
1145: Hu, E.~M. 1992,
1146: %Ly-alpha emission from cooling flows and measures of the dust content of rich clusters of galaxies}
1147: ApJ, 391, 608
1148:
1149:
1150: %\cite{Ivezic:2004bf}
1151: \bibitem[Ivezic et al. 2004]{Ivezic:2004bf}
1152: Ivezic, Z. et al. 2004,
1153: %``SDSS Data Management and Photometric Quality Assessment,''
1154: AN, 325, 583
1155: [arXiv:astro-ph/0410195]
1156: %%CITATION = ASNAA,325,583;%%
1157:
1158:
1159: %\cite{Karachentsev}
1160: \bibitem[Karachentsev \& Lipovetskii 1969]{Karachentsev}
1161: Karachentsev, I.~D. \& Lipovetskii, V.~A. 1969,
1162: %Absorbing Material in Clusters of Galaxies.
1163: SvA, 12, 909
1164:
1165:
1166: %\cite{Komatsu}
1167: \bibitem[Komatsu et~al. 2008]{Komatsu}
1168: Komatsu, E.~et~al.~2008,
1169: %Five-Year Wilkinson Microwave Anisotropy Probe (WMAP) Observations: Cosmological Interpretation,
1170: ApJS, in press
1171: [arXiv:0803.0547]
1172:
1173:
1174:
1175: %\cite{Kormendy:1989dg}
1176: \bibitem[Kormendy \& Djorgovski 1989]{Kormendy:1989dg}
1177: Kormendy, J. \& Djorgovski, S. 1989,
1178: %``Surface photometry and the structure of elliptical galaxies,''
1179: ARA\&A, 27, 235
1180: %%CITATION = ARAAA,27,235;%%
1181:
1182: %\cite{Krugel03}
1183: \bibitem[Kr\"ugel 2003]{Krugel03}
1184: Kr\"ugel, E. 2003,
1185: The Physics of Interstellar Dust (Bristol: IoP)
1186:
1187:
1188: %\cite{Kurucz1992}
1189: \bibitem[Kurucz 1992]{Kurucz1992}
1190: Kurucz, R.~L.~1992, in IAU Symp.~149, The Stellar Populations of Galaxies, ed.~B.~Barbuy \& A.~Renzini (Dordrecht: Kluwer), 225
1191:
1192:
1193: %\cite{Lupton:2001zb}
1194: \bibitem[Lupton et al. 2001]{Lupton:2001zb}
1195: Lupton, R., Gunn, J.~E., Ivezic, Z., Knapp, G.~R., Kent, S.~M., \& Yasuda, N. 2001,
1196: %``The SDSS Imaging Pipelines,''
1197: ASPC, 238, 269
1198: [arXiv:astro-ph/0101420]
1199: %%CITATION = ASPSF,238,269;%%
1200:
1201:
1202:
1203: %\cite{Maoz:1995yk}
1204: \bibitem[Maoz 1995]{Maoz:1995yk}
1205: Maoz, D.~1995,
1206: %``Limits on Dust in Rich Clusters of Galaxies from the Color of Background
1207: %Quasars,''
1208: ApJ, 455, L115
1209: [arXiv:astro-ph/9508093]
1210: %%CITATION = ASTRO-PH/9508093;%%
1211:
1212:
1213:
1214: %\cite{Muller:2008hg}
1215: \bibitem[{Muller et~al.} {2008}]{Muller:2008hg}
1216: Muller, S., Wu, S.~Y., Hsieh, B.~C., Gonzalez, R., Loinard, L., Yee, H., \& Gladders, M. 2008,
1217: %``Searching for dust in the intracluster medium from reddening of background
1218: %galaxies,''
1219: ApJ, 680, 975
1220: [arXiv:0801.2613]
1221: %%CITATION = ARXIV:0801.2613;%%
1222:
1223:
1224: %\cite{Nollenberg:2003cd}
1225: \bibitem[Nollenberg et~al. 2003]{Nollenberg:2003cd}
1226: Nollenberg, J.~G., Williams, L.~L.~R., \& Maddox, S.~J. 2003,
1227: %``Determination of Reddening and Extinction Due to Dust in APM Galaxy
1228: %Clusters,''
1229: AJ, 125, 2927
1230: [arXiv:astro-ph/0302537]
1231: %%CITATION = ASTRO-PH/0302537;%%
1232:
1233:
1234:
1235: %\cite{Okazaki93}
1236: \bibitem[Okazaki et~al. 1993]{Okazaki93}
1237: Okazaki, T., Chiba, M., Kumai, Y., \& Fujimoto, M. 1993,
1238: %title = ``{The distributions of hot gas and iron in the Virgo cluster}'',
1239: \pasj, 45, 669
1240:
1241: %\cite{Oke68}
1242: \bibitem[Oke \& Sandage 1968]{Oke68}
1243: Oke, J.~B. \& Sandage, A. 1968,
1244: %Energy Distributions, K Corrections, and the Stebbins-Whitford Effect for Giant Elliptical Galaxies,
1245: ApJ, 154, 21
1246:
1247:
1248:
1249: %\cite{Pier:2002iq}
1250: \bibitem[Pier et al. 2003]{Pier:2002iq}
1251: Pier, J.~R., Munn, J.~A., Hindsley, R.~B., Hennessy, G.~S., Kent, S.~M., Lupton, R.~H., \& Ivezic, Z. 2003
1252: %``Astrometric Calibration of the Sloan Digital Sky Survey,''
1253: AJ, 125, 1559
1254: [arXiv:astro-ph/0211375]
1255: %%CITATION = ANJOA,125,1559;%%
1256:
1257: %\cite{Popescu:2000qs}
1258: \bibitem[Popescu et~al. 2000]{Popescu:2000qs}
1259: Popescu, C.~C., Tuffs, R.~J., Fischera, J., \& Volk, H. 2000
1260: %``On the FIR emission from intracluster dust,''
1261: A\&A, 354, 480
1262: [arXiv:astro-ph/0001053]
1263: %%CITATION = ASTRO-PH/0001053;%%
1264:
1265:
1266: %\cite{Prugniel:2001xp}
1267: \bibitem[Prugniel \& Soubiran 2001]{Prugniel:2001xp}
1268: Prugniel, P. \& Soubiran, C. 2001,
1269: %``A database of high and medium-resolution stellar spectra,''
1270: A\&A, 369, 1048
1271: [arXiv:astro-ph/0101378]
1272: %%CITATION = ASTRO-PH/0101378;%%
1273:
1274:
1275:
1276: %\cite{Quintero:2003we}
1277: \bibitem[Quintero et al.(2004)]{Quintero:2003we}
1278: Quintero, A.~D. et al. 2004,
1279: %``Selection and photometric properties of K+A galaxies,''
1280: ApJ, 602, 190
1281: [arXiv:astro-ph/0307074]
1282: %%CITATION = ASJOA,602,190;%%
1283:
1284:
1285:
1286: %\cite{Richards:2002bb}
1287: \bibitem[Richards et al. 2002]{Richards:2002bb}
1288: Richards, G.~T. et al. 2002,
1289: %``Spectroscopic Target Selection in the Sloan Digital Sky Survey: The Quasar
1290: %Sample,''
1291: AJ, 123, 2945
1292: [arXiv:astro-ph/0202251]
1293: %%CITATION = ANJOA,123,2945;%%
1294:
1295: %\cite{Roberts:1994gp}
1296: \bibitem[Roberts \& Haynes 1994]{Roberts:1994gp}
1297: Roberts, M.~S. \& Haynes, M.~P. 1994
1298: %``Physical parameters along the hubble sequence,''
1299: ARA\&A, 32, 115
1300: %%CITATION = ARAAA,32,115;%%
1301:
1302:
1303:
1304: %\cite{Romani:1991nd}
1305: \bibitem[Romani \& Maoz 1992]{Romani:1991nd}
1306: Romani, R.~W. \& Maoz, D. 1992,
1307: %``The Angular correlation of quasars and Abell clusters,''
1308: %%CITATION = IASSNS-AST-91-21;%%
1309: ApJ, 386, 36
1310:
1311:
1312:
1313: %\cite{Shao}
1314: \bibitem[Shao \& Tu 1995]{Shao}
1315: Shao, J.~\& Tu, D.~1995, The Jackknife and Bootstrap (Berlin: Springer)
1316:
1317: %\cite{Smith:2002pca}
1318: \bibitem[{Smith et~al.} {2002}]{Smith:2002pca}
1319: Smith, J.~A. et al. 2002,
1320: %``The u'g'r'i'z' Standard star system,''
1321: AJ, 123, 2121
1322: [arXiv:astro-ph/0201143]
1323: %%CITATION = ANJOA,123,2121;%%
1324:
1325:
1326:
1327: %\cite{Stickel:2001nt}
1328: \bibitem[Stickel et~al. 2002]{Stickel:2001nt}
1329: Stickel, M., Klaas, U., Lemke, D., \& Mattila, K. 2002,
1330: %``Far-Infrared Emission from Intracluster Dust in Abell Clusters,''
1331: A\&A, 383, 367
1332: [arXiv:astro-ph/0112063]
1333: %%CITATION = ASTRO-PH/0112063;%%
1334:
1335:
1336: %\cite{Stoughton:2002ae}
1337: \bibitem[Stoughton et al. 2002]{Stoughton:2002ae}
1338: Stoughton, C. et al. 2002,
1339: %``The Sloan Digital Sky Survey: Early data release,''
1340: AJ, 123, 485
1341: %%CITATION = ANJOA,123,485;%%
1342:
1343:
1344:
1345: %\cite{Strauss:2002dj}
1346: \bibitem[Strauss et al. 2002]{Strauss:2002dj}
1347: Strauss, M.~A. et al. 2002,
1348: %``Spectroscopic Target Selection in the Sloan Digital Sky Survey: The Main
1349: %Galaxy Sample,''
1350: AJ, 124, 1810
1351: [arXiv:astro-ph/0206225]
1352: %%CITATION = ANJOA,124,1810;%%
1353:
1354: %\cite{Szalay89}
1355: \bibitem[Szalay et~al. 1989]{Szalay89}
1356: Szalay, A.~S., Hollosi, J. \& {Toth}, G. 1989,
1357: %Angular cross-relations of Abell clusters in different distance classes}
1358: ApJ, 339, L5
1359:
1360:
1361: %\cite{Trentham:2002mq}
1362: \bibitem[Trentham \& Hodgkin 2002]{Trentham:2002mq}
1363: Trentham, N. \& Hodgkin, S. 2002,
1364: %``The luminosity function of the Virgo Cluster from M_B=-22 to M_B=-11,''
1365: MNRAS 333, 423
1366: [arXiv:astro-ph/0202437]
1367: %%CITATION = MNRAA,333,423;%%
1368:
1369: %\cite{Vaucouleurs}
1370: \bibitem[de Vaucouleurs et~al. 1972]{Vaucouleurs}
1371: Vaucouleurs, G. de, Vaucouleurs, A. de \& Corwin, H.G., Jr. 1972,
1372: AJ, 77, 285
1373:
1374:
1375:
1376: %\cite{Visvanathan77}
1377: \bibitem[Visvanathan \& Sandage 1977]{Visvanathan77}
1378: Visvanathan, N. \& Sandage, A. 1977,
1379: %The color-absolute magnitude relation for E and S0 galaxies. I - Calibration and tests for universality using Virgo and eight other nearby clusters,
1380: ApJ, 216, 214
1381:
1382:
1383: %\cite{Wise93}
1384: \bibitem[Wise et~al. 1993]{Wise93}
1385: Wise, M.~W., O'Connell, R.~W., Bregman, J.~N., \& Roberts, M.~S. 1993,
1386: %Far-infrared emission from the intracluster medium
1387: ApJ, 405, 94
1388:
1389:
1390:
1391: %\cite{York:2000gk}
1392: \bibitem[York et al. 2000]{York:2000gk}
1393: York, D.~G. et al. 2000,
1394: %``The Sloan Digital Sky Survey: technical summary,''
1395: AJ, 120, 1579
1396: [arXiv:astro-ph/0006396]
1397: %%CITATION = ANJOA,120,1579;%%
1398:
1399:
1400: %\cite{Zwicky57}
1401: \bibitem[Zwicky 1957]{Zwicky57}
1402: Zwicky, F. 1957, Morphological Astronomy (Berlin: Springer)
1403:
1404:
1405: %\cite{Zwicky61}
1406: \bibitem[Zwicky 1961]{Zwicky61}
1407: Zwicky, F. 1961, \pasp, 63, 17
1408:
1409: %\cite{Zwicky62}
1410: \bibitem[Zwicky 1962]{Zwicky62}
1411: Zwicky, F. 1962, Problems of Extragalactic Research, ed. G.~C. McVittie (New York: Macmillan), 149
1412:
1413:
1414:
1415:
1416:
1417:
1418:
1419: \end{thebibliography}
1420:
1421: \end{document}
1422: