1: %
2: % Cosmos w(theta) paper
3: %
4: %% AASTeX v5.x LaTeX 2e macros.
5:
6: %% The first piece of markup in an AASTeX v5.x document
7: %% is the \documentclass command. LaTeX will ignore
8: %% any data that comes before this command.
9:
10: %% The command below calls the preprint style
11: %% which will produce a one-column, single-spaced document.
12: %% Examples of commands for other substyles follow. Use
13: %% whichever is most appropriate for your purposes.
14: %%
15:
16: %\documentclass[12pt,preprint]{aastex}
17: \documentclass{emulateapj}
18: \slugcomment{}
19: %% manuscript produces a one-column, double-spaced document:
20:
21: %\documentclass[manuscript]{aastex}
22:
23: %% preprint2 produces a double-column, single-spaced document:
24:
25: %\documentclass[preprint2]{aastex}
26:
27: %% Sometimes a paper's abstract is too long to fit on the
28: %% title page in preprint2 mode. When that is the case,
29: %% use the longabstract style option.
30:
31: %% \documentclass[preprint2,longabstract]{aastex}
32:
33: \usepackage{graphicx}
34: \usepackage{natbib}
35: %\usepackage{hyperref}
36: \newcommand{\vdag}{(v)^\dagger}
37:
38: \newcommand{\etal}{\mbox{et~al.}}
39: \newcommand{\Msol}{\mbox{$M_{\odot}$}}
40: \newcommand{\msun}{\mbox{$M_{\odot}$}}
41: \newcommand{\Msun}{\mbox{${\bf M_{\odot} }$}}
42: \newcommand{\Lsol}{\mbox{$L_{\odot}$}}
43: \newcommand{\lsun}{\mbox{$L_{\odot}$}}
44: \newcommand{\ergsec}{\mbox{erg s$^{-1}$}}
45: \newcommand{\beam}{\mbox{beam$^{-1}$}}
46: \newcommand{\Mdyn}{\mbox{$M_{\rm dyn}$}}
47: \newcommand{\Mgas}{\mbox{$M_{\rm gas}$}}
48: \newcommand{\HH}{\mbox{H$_2$}}
49: \newcommand{\Mgasfive}{\mbox{$M_{\rm gas}(r<500{\rm pc})$ }}
50: \newcommand{\Halpha}{\mbox{H$\alpha$} }
51: \newcommand{\tm}[1]{\tablenotemark{#1}}
52: \def\deg {{\ifmmode^\circ\else$^\circ$\fi}} %%% Overwrites TeX \deg
53: \def\Ho {{$H_{0}$} }
54: \def\qo {{$q_{0}$} }
55: \def\kmsMpc {{\ km\ s$^{-1}$\ Mpc$^{-1}$} }
56:
57: %\def\revised#1{\textbf{#1}}
58: \def\revised#1{{#1}}
59:
60: %% You can insert a short comment on the title page using the command below.
61:
62: % \slugcomment{Not to appear in Nonlearned J., 45.}
63:
64: %% If you wish, you may supply running head information, although
65: %% this information may be modified by the editorial offices.
66: %% The left head contains a list of authors,
67: %% usually a maximum of three (otherwise use et al.). The right
68: %% head is a modified title of up to roughly 44 characters.
69: %% Running heads will not print in the manuscript style.
70:
71: \shorttitle{Galaxy clustering in the COSMOS field}
72: \shortauthors{McCracken et al.}
73:
74: %% This is the end of the preamble. Indicate the beginning of the
75: %% paper itself with \begin{document}.
76:
77: \begin{document}
78:
79: %% LaTeX will automatically break titles if they run longer than
80: %% one line. However, you may use \\ to force a line break if
81: %% you desire.
82:
83: \title{The angular correlations of galaxies in the COSMOS field}
84:
85: %% Use \author, \affil, and the \and command to format
86: %% author and affiliation information.
87: %% Note that \email has replaced the old \authoremail command
88: %% from AASTeX v4.0. You can use \email to mark an email address
89: %% anywhere in the paper, not just in the front matter.
90: %% As in the title, use \\ to force line breaks.
91:
92: % this list of authors and their addresses in ap j format is
93: % based on the web site team member list + some new additions.
94: % PLEASE be cautious using it since there are undoubtedly some people
95: % missing and perhaps a few errors.
96: %
97: % AUTHORS SHOULD NOT USE THIS LIST IN ENTIRETY BUT DELETE
98: % AND EDIT IT
99: %
100: % COMPILED BY NZS
101: % used 2 initials when known, did not use spelled out names
102: %
103:
104: \author{H. J. McCracken\altaffilmark{1,2},
105: J. A. Peacock\altaffilmark{3},
106: L. Guzzo\altaffilmark{4,5,6},
107: P. Capak\altaffilmark{7,8}
108: C. Porciani\altaffilmark{9},
109: N. Scoville\altaffilmark{7,8},
110: H. Aussel\altaffilmark{10},
111: A. Finoguenov\altaffilmark{11},
112: J. B. James\altaffilmark{3},
113: M. G. Kitzbichler\altaffilmark{5},
114: A. Koekemoer\altaffilmark{12},
115: A. Leauthaud\altaffilmark{13},
116: O. Le F\`{e}vre\altaffilmark{13},
117: R. Massey\altaffilmark{7},
118: Y. Mellier\altaffilmark{1,2},
119: B. Mobasher\altaffilmark{12},
120: P. Norberg\altaffilmark{3},
121: J. Rhodes\altaffilmark{7,14},
122: D. B. Sanders\altaffilmark{15},
123: S. S. Sasaki\altaffilmark{16,17},
124: Y. Taniguchi\altaffilmark{17},
125: D. J. Thompson\altaffilmark{18,19},
126: S. D. M. White\altaffilmark{5},
127: A. El-Zant\altaffilmark{20}
128: }
129:
130: %% Notice that each of these authors has alternate affiliations, which
131: %% are identified by the \altaffilmark after each name. Specify alternate
132: %% affiliation information with \altaffiltext, with one command per each
133: %% affiliation.
134:
135: \altaffiltext{$\star$}{Based on observations with the NASA/ESA {\em
136: Hubble Space Telescope},obtained at the Space Telescope Science
137: Institute, which is operated by AURA Inc, under NASA contract NAS
138: 5-26555; also based on data collected at : the Subaru Telescope,
139: which is operated by the National Astronomical Observatory of Japan;
140: the XMM-Newton, an ESA science mission with instruments and
141: contributions directly funded by ESA Member States and NASA; the
142: European Southern Observatory under Large Program 175.A-0839, Chile;
143: Kitt Peak National Observatory, Cerro Tololo Inter-American
144: Observatory, and the National Optical Astronomy Observatory, which
145: are operated by the Association of Universities for Research in
146: Astronomy, Inc. (AURA) under cooperative agreement with the National
147: Science Foundation; the National Radio Astronomy Observatory which is
148: a facility of the National Science Foundation operated under
149: cooperative agreement by Associated Universities, Inc; and and the
150: Canada-France-Hawaii Telescope with MegaPrime/MegaCam operated as a
151: joint project by the CFHT Corporation, CEA/DAPNIA, the National
152: Research Council of Canada, the Canadian Astronomy Data Centre, the
153: Centre National de la Recherche Scientifique de France, TERAPIX and
154: the University of Hawaii.}
155:
156: \altaffiltext{1}{Institut d'Astrophysique de Paris, UMR7095 CNRS, Universit\`{e} Pierre et Marie Curie, 98 bis Boulevard Arago, 75014 Paris, France}
157: \altaffiltext{2}{Observatoire de Paris, LERMA, 61 Avenue de l'Observatoire, 75014 Paris, France}
158: \altaffiltext{3}{Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ, U.K.}
159: \altaffiltext{4}{INAF-Osservatorio Astronomico di Brera, via Bianchi 46, I-23807 Merate (LC), Italy}
160: \altaffiltext{5}{Max-Planck-Institut f\"{u}r Astrophysik, D-85748 Garching bei M\"{u}nchen, Germany}
161: \altaffiltext{6}{Visiting Scientist, European Southern Observatory,Karl-Schwarzschild-Str. 2, D-85748 Garching, Germany}
162: %
163: \altaffiltext{7}{California Institute of Technology, MC 105-24, 1200 East California Boulevard, Pasadena, CA 91125}
164: %
165: \altaffiltext{8}{Visiting Astronomer, Univ. Hawaii, 2680 Woodlawn Dr., Honolulu, HI, 96822}
166: %
167: \altaffiltext{9}{Department of Physics, ETH Zurich, CH-8093 Zurich, Switzerland}
168: %
169: \altaffiltext{10}{Service d'Astrophysique, CEA/Saclay, 91191 Gif-sur-Yvette, France}
170: %
171: \altaffiltext{11}{Max-Planck-Institut f\"ur Extraterrestrische Physik, Giessenbachstra\ss e, 85748 Garching, Germany}
172:
173: \altaffiltext{12}{Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218}
174: %
175: \altaffiltext{13}{Laboratoire d'Astrophysique de Marseille, BP 8, Traverse du Siphon, 13376 Marseille Cedex 12, France}
176:
177: \altaffiltext{14}{Jet Propulsion Laboratory, Pasadena, CA 91109}
178:
179: \altaffiltext{15}{Institute for Astronomy, 2680 Woodlawn Dr.,University of Hawaii, Honolulu, Hawaii, 96822}
180:
181: \altaffiltext{16}{Astronomical Institute, Graduate School of Science, Tohoku University, Aramaki, Aoba, Sendai 980-8578, Japan}
182:
183: \altaffiltext{17}{Physics Department, Graduate School of Science, Ehime University, 2-5 Bunkyou, Matuyama, 790-8577, Japan}
184:
185: \altaffiltext{18}{Large Binocular Telescope Observatory, University of Arizona, 933 N. Cherry Ave.Tucson, AZ 85721-0065, USA}
186: \altaffiltext{19}{Caltech Optical Observatories, MS 320-47, California Institute of Technology, Pasadena, CA 91125}
187: \altaffiltext{20}{Canadian Institute for Theoretical Astrophysics, Mclennan Labs, University of Toronto, 60 St. George St, Room 1403, Toronto, ON M5S 3H8, Canada}
188:
189: \begin{abstract}
190:
191: We present measurements of the two-point galaxy angular correlation
192: function $w(\theta)$ in the COSMOS field. Independent determinations
193: of $w(\theta)$ as a function of magnitude limit are presented for
194: both the HST ACS catalog and also for the ground-based data from
195: Subaru and the CFHT. Despite having significantly different masks,
196: these three determinations agree well. At bright magnitudes ($I_{\rm
197: AB}<22$), our data generally match very well with existing
198: measurements and with mock catalogs based on semi-analytic galaxy
199: formation calculations of \citet{2006astro.ph..9636K} from the
200: Millennium Simulation. The exception is that our result is at the
201: upper end of the expected cosmic variance scatter for $\theta > 10$
202: arcmin, which we attribute to a particularly rich structure known to
203: exist at $z\simeq 0.8$. For fainter samples, however, the level of
204: clustering is somewhat higher than reported by some previous
205: studies: \revised {in all three catalogues} we find
206: $w(\theta=1')\simeq 0.014$ at a median $I_{\rm AB}$ magnitude of
207: 24. At these very faintest magnitudes, our measurements agree well
208: with the latest determinations from the Canada-France Legacy Survey.
209: This level of clustering is approximately double what is predicted
210: by the semi-analytic catalogs (at all angles). The semi-analytic
211: results allow an estimate of cosmic variance, which is too small to
212: account for the discrepancy. We therefore conclude that the mean
213: amplitude of clustering at this level is higher than previously
214: estimated.
215:
216: \end{abstract}
217:
218: \keywords{cosmology: observations --- cosmology: large scale structure of universe
219: --- cosmology: dark matter --- galaxies: formation --- galaxies: evolution --- surveys}
220: \section{Introduction}
221: The COSMOS field \citep{scoville_cosmos_overview} is the largest
222: contiguous multi-wavelength probe of the high-redshift galaxy
223: distribution, and a major task for the survey will be to extract
224: improved measurements of galaxy clustering at these early times. In
225: this initial paper, we will be concerned with the simplest of these
226: measures: the angular two-point correlation function, $w(\theta)$.
227: Demonstrating a robust measurement of this quantity is a minimum
228: requirement for verifying that the survey completeness is understood,
229: as a basis for future more elaborate analysis. The main aim of this
230: paper is therefore to present measurements of the two-point galaxy
231: clustering statistic on the COSMOS field using three independently
232: generated catalogs, and to compare the results with existing data. We
233: also compare the amplitudes we measure to those found in a
234: semi-analytic model of galaxy formation.
235:
236: The key feature of the COSMOS field is that it is completely covered
237: by the largest existing mosaic of image tiles from the Advanced Camera
238: for Surveys (ACS) on the Hubble Space Telescope
239: \citep{scoville_cosmos_hst_observations}. With respect to ground-based
240: surveys, in addition to the exceptional image quality, the great
241: advantage of this dataset is the superior photometric accuracy and
242: stability over the entire field of view of the survey, which in turn
243: makes it possible to measure the clustering of galaxies on large
244: scales and at faint magnitudes where the amplitude of the galaxy
245: correlation function $w(\theta)$ is very small. \revised{The COSMOS
246: field is currently the survey that probes the largest comoving
247: scales at redshifts of around one. The total area covered, 2
248: deg$^2$, does not exceed existing studies, in particular the UH8k
249: study \citep[1.5 deg$^2$,][]{2003ApJ...585..191W} and the shallower
250: DEEP2 measurements \citep[5.0 deg$^2$,][]{2004ApJ...617..765C}.
251: However, COSMOS offers the unique combination of large contiguous
252: area and depth, and also has the virtue of several independent and
253: quite different imaging datasets in the same field. }
254:
255: We use this rich dataset to investigate the clustering properties of
256: the field galaxy population on degree scales. In future papers, we
257: will present a more detailed study of galaxy clustering using
258: photometric redshifts \revised {which can be used, for example, to
259: divide our galaxy catalogues by type and apparent magnitude}. Our
260: objective here is simply to present the global properties of the field
261: in terms of simple two-point statistics for catalogs selected by
262: apparent magnitude. \revised{As our aim is to demonstrate the
263: robustness of the results, we restrict ourselves to the $i$-band
264: data, which is available for all three datasets considered here.}
265:
266: The COSMOS field has been imaged by many ground-based facilities,
267: amongst them the Subaru telescope \citep{taniguchi_cosmos_subaru}, and
268: we compare the COSMOS ACS catalog (described in
269: \citealp{lea_cosmos_catalog}) with two ground-based catalogs: the
270: Subaru optical catalog described by \citet{capak_cosmos_catalog} and
271: the CFHTLS-T03 catalog used by McCracken et al. (2007, in preparation).
272: Each of these have significantly different masks, especially with
273: regard to ghosting around bright stars. As will be demonstrated, the
274: results of these independent determinations are in good agreement, and
275: display a consistently higher amplitude at faint magnitudes than has
276: been suggested in previous work.
277:
278: \section{Catalogs and methods}
279:
280: All of the three catalogs were prepared independently. For full
281: details of the ACS catalog, see \citet{lea_cosmos_catalog}; the
282: Subaru catalogs are described in \citet{capak_cosmos_catalog}. A
283: full description of the CFHTLS catalogs, based on the images
284: corresponding to the release CFHTLS-T03, can be found in McCracken
285: et al 2007.
286:
287: The ACS data are constructed from a mosaic of 575 image tiles taken
288: over 588 orbits of the HST. The 50\% completeness limit of the
289: catalog is 26.6 F814W magnitudes.
290:
291: SuprimeCam consists of 10 Lincoln-Labs 8k $\times$ 4k CCDs with a plate
292: scale of $0.2''$ per pixel. However, a plate scale of $0.15''$ per
293: pixel was used for the final image to ensure that the images with good
294: seeing were not undersampled. A special dither pattern including
295: camera rotations (Subaru is an alt-azimuth telescope) was used to
296: ensure every portion of the field was imaged by at least four different
297: CCDs. Further details can be found in \citet{taniguchi_cosmos_subaru}.
298: The Subaru catalog was based on a mosaic of 115 Subaru images taken
299: with the Suprime camera on the Subaru 8m telescope. The image has a
300: median seeing of $0.6''$ and a $50\%$ completeness of $i=27.4$.
301:
302: \revised{In order to produce the best possible catalogue for
303: correlation function measurements, we re-extracted a catalogue
304: ourselves from the Subaru tiles produced by Capak et al. We
305: downloaded each tile and assembled them into a single large mosaic
306: using the TERAPIX software tool \texttt{swarp}; we carried out the
307: same procedure for the RMS maps. Following this, we used
308: \texttt{sextractor} to extract a catalogue. Star-galaxy separation
309: to $i=21.5$ was performed using the \texttt{flux\_radius} compactness
310: parameter which measures the radius which encloses $50\%$ of an
311: object's flux. Bright stars and defects were masked on the images. }
312:
313:
314: The CFHTLS catalog is derived from the TERAPIX CFHTLS-T03 release.
315: The CFHTLS stacks were taken using the Megacam camera on the 3.6m CFHT
316: telescope. Megacam covers 1 deg$^2$ with $0.205''$ pixels using 36
317: separate $2048\times4096$ Rockwell CCDs. Note that, unlike the Subaru
318: and ACS data, the CFHTLS-T03 image consists of a single Megacam
319: pointing. The
320: CFHTLS-T03 D2-i dataset comprises 153 images and has a median seeing
321: of $0.9''$. The 50\% completeness of this dataset is
322: $I_{\rm AB}\simeq 25.7$ magnitudes\footnote{For details see
323: {\tt \url {http://terapix.iap.fr/cplt/tab\_t03ym.html}}}.
324:
325: In all three catalogs, the star-galaxy separation is carried out
326: using a morphological classifier. We emphasize that the classifiers
327: were determined independently for each catalog; we did not, for
328: example, use the ACS morphologies to perform star-galaxy separation
329: on the CFHTLS or Subaru images, or use the Subaru images as
330: detection images for the ACS data. The Subaru images cover the full
331: 2~deg$^2$ of the COSMOS field while the ACS tiling covers a total of
332: 1.7~deg$^2$; the CFHTLS-T03 images cover just the central 1~deg$^2$
333: of the field.
334:
335: In each catalog, regions around bright stars and near the edges of the
336: field were masked. For the ACS catalogs, we use the same set of masks
337: that were used for weak lensing measurements
338: \citep{massey_cosmos_cosmicshear}. These masks also remove many
339: blended objects.
340: \revised{After masking, in the Subaru catalogue there are 134,397
341: galaxies in the magnitude range $20 < i < 25$; in the ACS catalogue
342: and there are there are 124,665 galaxies in the same interval in
343: magnitude. For the CFHTLS, there are 52,521 galaxies in the
344: magnitude range $20 < i < 24$. The effective areas (total available
345: area after masking) of the three surveys (Subaru, ACS and CFHT) is
346: $1.6$, $1.5$, and $0.7$ deg$^2$ respectively
347: (i.e. completeness of 80\%, 88\% and 70\%). We have experimented with
348: varying the degree of masking by `growing' the mask to eliminate
349: pixels adjacent to masked pixels. The results are robust even
350: when $>50\%$ of the area is masked.}
351:
352:
353:
354: \revised{Figure~\ref{fig:counts} shows the galaxy number counts extracted from the
355: three catalogues. The dotted lines indicate the magnitude limits
356: adopted in this paper. The slight `break' in the counts at $i*\simeq 21$ is
357: an artefact caused by our morphologically based star-galaxy
358: separation. The ACS count are slightly lower at $i\simeq 22$ as a
359: consequence of the improved star-galaxy separation in this
360: dataset. }
361:
362:
363:
364: \section{Clustering measurements}
365:
366: We selected galaxies in progressively fainter slices of apparent $i$
367: magnitude. For the purposes of the paper we assume that the
368: instrumental AB total magnitudes measured in each catalog are
369: equivalent; this is approximately true. \citet{capak_cosmos_catalog}
370: present a detailed comparison of galaxy photometry between the three
371: catalogs described here. Figure 8 of their paper demonstrates that
372: total instrumental magnitudes in each catalog agree well, to
373: within 0.05 magnitudes.
374:
375: \revised{For each slice, we measure $w$ for at range of angular separations
376: $\theta$ to $\theta+\delta \theta$ in a series of logarithmically
377: separated bins using the standard \citet{1993ApJ...412...64L} estimator,
378: %
379: \begin{equation}
380: w ( \theta) ={\mbox{DD} - 2\mbox{DR} + \mbox{RR}\over \mbox{RR}}
381: \label{eq:1.ls}
382: \end{equation}
383: %
384: with the $DD$, $DR$ and $RR$ terms referring to the number of
385: data--data, data--random and random--random galaxy pairs between
386: $\theta$ and $\theta+\delta\theta$. }
387: \revised{The fitted amplitudes quoted in this paper assume a power-law
388: slope for the galaxy correlation function,
389: $w(\theta)=A_w(\theta/{\rm deg})^{-\delta}$; however this amplitude must be
390: adjusted for the `integral constraint' correction, arising from
391: the need to estimate the mean galaxy density from the sample
392: itself. This can be estimated as \citep[e.g.][]{2005ApJ...619..697A},
393: %
394: \begin{equation}
395: C = {1 \over {\Omega^2}} \int\!\!\! \int w(\theta)\, d\Omega_1\, d\Omega_2,
396: \label{eq:5}
397: \end{equation}
398: %
399: where $\Omega$ is the area subtended by each of our survey fields. For
400: the COSMOS field, We find $C\sim 1A_w$ by numerical integration of
401: Equation~\ref{eq:5} over our field geometry and assuming that galaxies
402: closer than $1''$ cannot be distinguished.}
403:
404:
405:
406: We used a sorted linked list in order to reduce computing time given
407: the very large number of objects in each slice. These results are
408: compared in Figure~\ref{fig:wtheta-comp}. The solid line shows
409: measurements from the ACS; the triangles and stars correspond to
410: measurements from Subaru and CFHT. At each angular bin for each
411: survey, the error bars plotted are simple bootstrap errors. Although
412: these are not in general a perfect substitute for a full estimate of
413: cosmic variance (e.g. using an ensemble of simulations), they should
414: give the correct magnitude of the uncertainty
415: \citep{1992ApJ...392..452M}. \revised{In particular, these authors
416: show that bootstrap errors yield sensible uncertainties on power-law
417: fits to correlation-function data when the points are treated as
418: independent.}
419:
420:
421: \begin{figure}
422: \epsscale{1.2}
423: \plotone{f1.eps}
424: \caption{Galaxy counts for the three catalogues presented in this
425: paper: ACS (solid line); CFHTLS-T03 (open squares) and Subaru (red
426: filled triangles). For reference, we also show galaxy counts extracted
427: from the HDF-N by \cite{2001MNRAS.323..795M}}
428: \label{fig:counts}
429: \end{figure}
430:
431:
432:
433:
434: \begin{figure}
435: \epsscale{1.2}
436: \plotone{f2.eps}
437: \caption{The angular correlation function, $w(\theta)$ as a function of
438: angular separation, $\theta$ in four slices of apparent
439: magnitude. In each panel we show three different measurements from
440: three different catalogs: ACS (connected lines); Subaru (triangles);
441: CFHTLS/Megacam (stars).}
442: \label{fig:wtheta-comp}
443: \end{figure}
444:
445:
446: \begin{figure}
447: \epsscale{1.2}
448: \plotone{f3.eps}
449: \caption{The amplitude of the angular correlation function $w$, as a
450: function of angular separation, $\theta$, for the ACS catalog.
451: Measurements for galaxies selected in the magnitude ranges $21<i<22$
452: and $24<i<25$ and are presented. The dotted lines show the
453: best-fitting power-law correlations (with slopes $-0.59$ and
454: $-0.47$, respectively) and the
455: integral correction for the COSMOS field included. The dashed line
456: is plotted at $1'$. There seems little evidence here for any deviation
457: from power-law correlations, although the brighter bin does hint at an
458: inflection around 1 arcmin. The correlation function for the fainter bin
459: is clearly flatter.
460: }
461: \label{fig:wtheta-acs}
462: \end{figure}
463:
464: For each of the four slices in apparent magnitude, the amplitude of
465: $w(\theta)$ measured in the Subaru data agrees well with the
466: measurements in the ACS. At very small angular separations the ACS data
467: are higher than the ground-based results. For the two faintest bins
468: ($23 < i < 24$ and $24 < i < 25$) the agreement between the ACS
469: measurements and the Subaru measurements is excellent.
470:
471: We now examine more closely the galaxy correlation function measured
472: from the ACS catalogs. As we have already seen, thanks to the
473: excellent resolution of the ACS images, we are able to measure
474: clustering amplitudes to small separations, on the order of $1''$.
475: Conversely, as a consequence of the large areal coverage of the ACS
476: COSMOS field, we can also measure amplitudes to large angular
477: separations. In Figure~\ref{fig:wtheta-acs} we show the angular
478: correlation function $w(\theta)$ as a function of angular separation
479: for bright and faint samples with $21<i<22$ and $24<i<25$
480: respectively.
481: \revised{The dotted line shows the best-fitting lines
482: with an integral constraint correction applied. For the bright bin,
483: we find a best fitting slope of $-0.59\pm0.05$; for the fainter bin,
484: $-0.47\pm0.02.$ We find unambiguously that in this magnitude limited
485: sample, the galaxy correlation function becomes flatter towards
486: fainter magnitude bins, in agreement with previous works
487: \citep{2001A&A...376..756M}.}
488:
489:
490: \section{Comparisons with simulations}
491: \label{sec:repr-cosm-field}
492:
493: In the previous Sections we established that measurements between the
494: different datasets are consistent. In this Section, we compare our
495: measurements to those made on catalogs produced by
496: \cite{2006astro.ph..9636K}. These catalogs were created using a
497: semianalytic model to simulate galaxy formation within the evolving
498: halo population of the extremely large Millennium Simulation. These
499: simulated Universes were then `observed' to produce light-cones, which
500: can be then used to produce observations of identical geometry to those
501: of real catalogs. \citep{2005MNRAS.360..159B,2006astro.ph..9636K}.
502: Each COSMOS light-cone covers 2~deg$^2$.
503: Figure~\ref{fig:wtheta-millennium} shows, as before, the amplitude of
504: $w(\theta)$ as a function of angular separation in four magnitude
505: slices. Points with error bars show measurements from the ACS data. The
506: solid line shows the average of measurements made from twenty
507: light-cones extracted from the millennium simulation; the dotted lines
508: show the amplitude of the $\pm1 \sigma$ error bars.
509:
510: \begin{figure}
511: %\resizebox{\hsize}{!}{\includegraphics{./eps/cosmos_millennium.ps}}
512: \epsscale{1.2}
513: \plotone{f4.eps}
514: \caption{The amplitude of the angular correlation function $w$, as a
515: function of angular separation, $\theta$, for the ACS catalog. The
516: lines show measurements made using mock catalogs extracted from the
517: millennium simulation. The dotted lines show the one sigma scatter of
518: the results for individual mock catalogs around their mean computed
519: from the variance over twenty mock catalogs. Note that these mock
520: results automatically incorporate integral corrections due to the
521: finite field size. }
522: \label{fig:wtheta-millennium}
523: \end{figure}
524:
525: For the two brighter slices, $21<i<22$ and $22<i<23$, the agreement
526: between the simulations and observations at intermediate and small
527: scales is remarkably good. At fainter magnitudes, and at larger scales,
528: the amplitudes measured in the COSMOS catalogs are consistently
529: higher than the prediction of the simulations. Since the simulations
530: allow us to assess the amplitude of cosmic variance directly, we
531: can easily conclude that this discrepancy cannot plausibly be attributed
532: to the COSMOS field having above-average clustering. It therefore
533: appears that the model predictions for $w(\theta)$ at $i\simeq 24$
534: are too low; we discuss this further below.
535:
536: \section{Comparisons with literature measurements}
537:
538: From our measurements it is clear that the form of $w(\theta)$ in the
539: COSMOS field does not correspond to a simple power law with a slope
540: that is independent of median magnitude of the sample. In the past,
541: determinations of galaxy clustering amplitude were usually given at a
542: fixed angular scale over a small range of angular separations. From
543: our data, it is clear that fitting $w(\theta)$ only at small angular
544: separations will result in different fitted amplitudes as compared to
545: a fit over the entire range. Therefore, in order to compare with
546: results presented in the literature, we choose to carry out fits over
547: \revised{a similar range of angular separations}.
548:
549: \begin{figure}
550: \epsscale{1.2}
551: \plotone{f5.eps}
552: \caption{Comparisons of the COSMOS results for the dependence of
553: the amplitude of $w(\theta)$ on depth, compared with existing
554: measurements and with the mock results. We choose to specify
555: the amplitude at 1 arcmin, based on a $\delta=-0.8$ power-law
556: fit around this point. The results are highly insensitive to this choice
557: of slope.}
558: \label{fig:wtheta-comparisons}
559: \end{figure}
560: \nocite{2003ApJ...585..191W}
561:
562: In Figure~\ref{fig:wtheta-comparisons} we show the fitted amplitudes of
563: $w(\theta)$ as a function of the median magnitude of each slice. In
564: comparing with literature measurements, we see that at bright
565: magnitudes $(i\simeq 20)$ our measurements are approximately in
566: agreement with the values presented in the literature. However, by
567: $i\simeq 23$, the COSMOS field measurements are significantly higher
568: than, for example the CFDF measurements of \citet{2001A&A...376..756M}.
569: This is true at all angles and for all measurements, whether they are
570: from the ground based or space-based catalogs.
571:
572: Photometric redshifts are available for all objects in the COSMOS field
573: \citep{mobasher_cosmos_photz}. The slice with $22<i<23$ has a median
574: redshift of $z\simeq0.8$; the next faintest slice is at $z\simeq0.9$.
575: If we examine Figure 2 of \citet{scoville_cosmos_lss} we can see there
576: is a significant over-density in the redshift range $0.7<z<0.9$. It is
577: possible that this structure could contribute to the enhanced signal on
578: large scales seen in our data in these slices.
579:
580: Some independent evidence exists that this structure raises the
581: COSMOS clustering amplitude above the ensemble average. In
582: Figure~\ref{fig:wtheta-comparisons} we have plotted the average
583: amplitude of $w$ as a function of apparent median magnitude for the
584: four independent deep fields of the Canada-France Legacy Survey (one
585: of which is the D2 field described above), totalling an effective
586: area of 3.2 deg$^2$. Error bars correspond to the variance over all
587: four fields. Interestingly, at bright magnitudes, the CFHTLS
588: magnitudes agree with the COSMOS measurements, and other literature
589: values; at intermediate magnitudes ($22<i<23$) they are between the
590: COSMOS values and those from other surveys, whereas at fainter
591: magnitudes they agree perfectly with the COSMOS measurements,
592: presumably because the median redshift probed by both surveys at
593: these magnitude limits is beyond that of the rich structure in the
594: COSMOS field. The size of the cosmic variance error bars on the
595: CFHTLS measurements are also largest at intermediate magnitudes.
596: These measurements are discussed in greater detail in a forthcoming
597: paper (McCracken et al., in preparation).
598:
599: \section {Discussion and conclusions}
600:
601: This paper has presented measurements of the angular two-point
602: correlation function in the COSMOS field, $w(\theta)$, and how it
603: depends on $i$-band magnitude depth. We have shown that consistent
604: results are obtained using three independent datasets: HST ACS; Subaru
605: SuprimeCam; and CFHT MegaCam. The results agree well at bright
606: magnitudes ($i\simeq 22$) with previous measurements and with the
607: predictions of semi-analytic mock catalogs constructed from the
608: Millennium Simulation (MS). The only slight caveat here is that the
609: results at $\theta > 10$ arcmin are at the high end of the MS
610: predictions, which may reflect a single rich $z\simeq 0.8$ structure
611: in the field.
612:
613: At fainter magnitudes, however, a different picture emerges. By the
614: time we reach the $24<i<25$ bin, the COSMOS measurements are
615: consistently a factor 2 higher than the MS predictions at all angles.
616: Moreover, the COSMOS measurements are consistent with the four-field
617: average of measurements from the CFHTLS survey. This discrepancy is
618: well beyond the compass of cosmic variance from limited numbers of rich
619: structures, as measured via the ensemble of simulations. Thus, barring
620: some undetected systematic that is consistent between all the datasets
621: we have used, the conclusion must be that the MS predictions are too
622: low at these magnitude levels. This could arise in a number of ways:
623: the predicted degree of bias at high redshifts might be too low;
624: the luminosity function might be incorrect, resulting in too high a predicted
625: mean redshift at these depths; or alternatively the MS may miss
626: foreground pairs of intrinsically faint galaxies because of the
627: resolution limit of the simulation. The first possibility is
628: particularly interesting given the current debate over the
629: normalization of the primordial power spectrum, $\sigma_8$. The MS used
630: $\sigma_8=0.9$, whereas WMAP favors a smaller result -- perhaps as low
631: as $\sigma_8=0.7$ (Spergel et al. 2006). Since high-redshift galaxies
632: are strongly biased already, a reduced $\sigma_8$ will in fact boost
633: the predicted galaxy clustering (for a given galaxy mass). These issues
634: will be explored further in future papers, where we make direct use of
635: the photometric redshift data in the COSMOS field.
636:
637:
638: \acknowledgements
639: This work is based in part on data products
640: produced at TERAPIX located at the Institut d'Astrophysique de
641: Paris. H. J. McCracken wishes to acknowledge the use of TERAPIX
642: computing facilities.
643:
644: The HST COSMOS Treasury program was supported through NASA grant
645: HST-GO-09822. We wish to thank Tony Roman, Denise Taylor, and David
646: Soderblom for their assistance in planning and scheduling of the
647: extensive COSMOS observations. We gratefully acknowledge the
648: contributions of the entire COSMOS collaboration consisting of more
649: than 70 scientists. More information on the COSMOS survey is available
650: at \url{http://cosmos.astro.caltech.edu}. It is a
651: pleasure to acknowledge the excellent services provided by the NASA
652: IPAC/IRSA staff (Anastasia Laity, Anastasia Alexov, Bruce Berriman and
653: John Good) in providing online archive and server capabilities for the
654: COSMOS datasets. Data for the Millennium Simulation are publically
655: available at \url{http://www.mpa-garching.mpg.de/millennium}.
656:
657: %% To help institutions obtain information on the effectiveness of their
658: %% telescopes, the AAS Journals has created a group of keywords for telescope
659: %% facilities. A common set of keywords will make these types of searches
660: %% significantly easier and more accurate. In addition, they will also be
661: %% useful in linking papers together which utilize the same telescopes
662: %% within the framework of the National Virtual Observatory.
663: %% See the AASTeX Web site at http://www.journals.uchicago.edu/AAS/AASTeX
664: %% for information on obtaining the facility keywords.
665:
666: %% After the acknowledgements section, use the following syntax and the
667: %% \facility{} macro to list the keywords of facilities used in the research
668: %% for the paper. Each keyword will be checked against the master list during
669: %% copy editing. Individual instruments or configurations can be provided
670: %% in parentheses, after the keyword, but they will not be verified.
671:
672: % {\it Facilities:} \facility{HST (ACS)}, \facility{HST (NICMOS)}, \facility{HST (WFPC2)}.
673:
674: %% Appendix material should be preceded with a single \appendix command.
675: %% We have used macros to produce journal name abbreviations.
676: %% AASTeX provides a number of these for the more frequently-cited journals.
677: %% See the Author Guide for a list of them.
678:
679: %% Note that the style of the \bibitem labels (in []) is slightly
680: %% different from previous examples. The natbib system solves a host
681: %% of citation expression problems, but it is necessary to clearly
682: %% delimit the year from the author name used in the citation.
683: %% See the natbib documentation for more details and options.
684:
685: %\bibliographystyle{apj}
686: %\bibliography{thesis,cosmos_plus}
687: \begin{thebibliography}{}
688:
689: \bibitem[\protect\citeauthoryear{{Adelberger} et~al.}{{Adelberger}
690: et~al.}{2005}]{2005ApJ...619..697A}
691: {Adelberger}, K.~L., {Steidel}, C.~C., {Pettini}, M., {Shapley}, A.~E.,
692: {Reddy}, N.~A., \& {Erb}, D.~K. 2005, \apj, 619, 697
693:
694: \bibitem[\protect\citeauthoryear{{Blaizot} et~al.}{{Blaizot}
695: et~al.}{2005}]{2005MNRAS.360..159B}
696: {Blaizot}, J., {Wadadekar}, Y., {Guiderdoni}, B., {Colombi}, S.~T., {Bertin},
697: E., {Bouchet}, F.~R., {Devriendt}, J.~E.~G., \& {Hatton}, S. 2005, \mnras,
698: 360, 159
699:
700: \bibitem[\protect\citeauthoryear{{Capak} et~al.}{{Capak}
701: et~al.}{2007}]{capak_cosmos_catalog}
702: {Capak}, P.~L., et~al. 2007, \apjs, 999, 999
703:
704: \bibitem[\protect\citeauthoryear{{Coil} et~al.}{{Coil}
705: et~al.}{2004}]{2004ApJ...617..765C}
706: {Coil}, A.~L., {Newman}, J.~A., {Kaiser}, N., {Davis}, M., {Ma}, C.-P.,
707: {Kocevski}, D.~D., \& {Koo}, D.~C. 2004, \apj, 617, 765
708:
709: \bibitem[\protect\citeauthoryear{{Kitzbichler} \& {White}}{{Kitzbichler} \&
710: {White}}{2006}]{2006astro.ph..9636K}
711: {Kitzbichler}, M.~G., \& {White}, S.~D.~M. 2006, ArXiv Astrophysics e-prints
712:
713: \bibitem[\protect\citeauthoryear{{Landy} \& {Szalay}}{{Landy} \&
714: {Szalay}}{1993}]{1993ApJ...412...64L}
715: {Landy}, S.~D., \& {Szalay}, A.~S. 1993, \apj, 412, 64
716:
717: \bibitem[\protect\citeauthoryear{{Leauthaud} et~al.}{{Leauthaud}
718: et~al.}{2007}]{lea_cosmos_catalog}
719: {Leauthaud}, A., et~al. 2007, \apjs, 999-999
720:
721: \bibitem[\protect\citeauthoryear{{Massey} et~al.}{{Massey}
722: et~al.}{2007}]{massey_cosmos_cosmicshear}
723: {Massey}, R., et~al. 2007, \apjs, 999-999
724:
725: \bibitem[\protect\citeauthoryear{{McCracken} et~al.}{{McCracken}
726: et~al.}{2001}]{2001A&A...376..756M}
727: {McCracken}, H.~J., {Le F{\` e}vre}, O., {Brodwin}, M., {Foucaud}, S., {Lilly},
728: S.~J., {Crampton}, D., \& {Mellier}, Y. 2001, \aap, 376, 756
729:
730: \bibitem[\protect\citeauthoryear{{Metcalfe} et~al.}{{Metcalfe}
731: et~al.}{2001}]{2001MNRAS.323..795M}
732: {Metcalfe}, N., {Shanks}, T., {Campos}, A., {McCracken}, H.~J., \& {Fong}, R.,
733: 323, 795
734:
735: \bibitem[\protect\citeauthoryear{{Mo}, {Jing}, \& {Boerner}}{{Mo}
736: et~al.}{1992}]{1992ApJ...392..452M}
737: {Mo}, H.~J., {Jing}, Y.~P., \& {Boerner}, G. 1992, \apj, 392, 452
738:
739: \bibitem[\protect\citeauthoryear{{Mobasher} et~al.}{{Mobasher}
740: et~al.}{2007}]{mobasher_cosmos_photz}
741: {Mobasher}, B., et~al. 2007, \apjs
742:
743: \bibitem[\protect\citeauthoryear{{Scoville} et~al.}{{Scoville}
744: et~al.}{2007a}]{scoville_cosmos_hst_observations}
745: {Scoville}, N.~Z., et~al. 2007a, \apjs, 999-999
746:
747: \bibitem[\protect\citeauthoryear{{Scoville} et~al.}{{Scoville}
748: et~al.}{2007b}]{scoville_cosmos_lss}
749: {Scoville}, N.~Z., et~al. 2007b, \apjs, 999-999
750:
751: \bibitem[\protect\citeauthoryear{{Scoville} et~al.}{{Scoville}
752: et~al.}{2007c}]{scoville_cosmos_overview}
753: {Scoville}, N.~Z., et~al. 2007c, \apjs, 999-999
754:
755: \bibitem[\protect\citeauthoryear{{Taniguchi} et~al.}{{Taniguchi}
756: et~al.}{2007}]{taniguchi_cosmos_subaru}
757: {Taniguchi}, N.~Z., Y.~{Scoville}, et~al. 2007, \apjs, 999-999
758:
759: \bibitem[\protect\citeauthoryear{{Wilson}}{{Wilson}}{2003}]{2003ApJ...585..191%
760: W}
761: {Wilson}, G. 2003, \apj, 585, 191
762:
763: \end{thebibliography}
764:
765:
766: \clearpage
767:
768: %% Use the figure environment and \plotone or \plottwo to include
769: %% figures and captions in your electronic submission.
770: %% To embed the sample graphics in
771: %% the file, uncomment the \plotone, \plottwo, and
772: %% \includegraphics commands
773: %%
774: %% If you need a layout that cannot be achieved with \plotone or
775: %% \plottwo, you can invoke the graphicx package directly with the
776: %% \includegraphics command or use \plotfiddle. For more information,
777: %% please see the tutorial on "Using Electronic Art with AASTeX" in the
778: %% documentation section at the AASTeX Web site,
779: %% http://www.journals.uchicago.edu/AAS/AASTeX.
780:
781: \end{document}
782: