1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{natbib}
3: \usepackage{ifthen}
4: \newcounter{address}
5: \newcommand{\latin}[1]{\textit{#1}}
6: \newcommand{\ie}{\latin{i.e.}}
7: \newcommand{\eg}{\latin{e.g.}}
8: \newcommand{\cf}{\latin{cf.}}
9: \newcommand{\etc}{\latin{etc.}}
10: \newcommand{\etal}{\latin{et~al.}}
11: \newcommand{\lcdm}{$\Lambda$CDM}
12:
13: \begin{document}
14: \title{
15: Cosmic homogeneity demonstrated with luminous red galaxies}
16: \author{
17: David~W.~Hogg\altaffilmark{\ref{NYU},\ref{email}},
18: Daniel~J.~Eisenstein\altaffilmark{\ref{Steward}},
19: Michael~R.~Blanton\altaffilmark{\ref{NYU}},
20: Neta~A.~Bahcall\altaffilmark{\ref{Princeton}},
21: J.~Brinkmann\altaffilmark{\ref{APO}},
22: James~E.~Gunn\altaffilmark{\ref{Princeton}}, and
23: Donald~P.~Schneider\altaffilmark{\ref{PennState}}
24: }
25: \setcounter{address}{1}
26: \altaffiltext{\theaddress}{\stepcounter{address}\label{NYU}
27: Center for Cosmology and Particle Physics, Department of Physics,
28: New York University, 4 Washington Pl, New York, NY 10003}
29: \altaffiltext{\theaddress}{\stepcounter{address}\label{email}
30: \texttt{david.hogg@nyu.edu}}
31: \altaffiltext{\theaddress}{\stepcounter{address}\label{Steward}
32: Steward Observatory, 933 N Cherry Ave, Tucson, AZ 85721}
33: \altaffiltext{\theaddress}{\stepcounter{address}\label{Princeton}
34: Princeton University Observatory, Princeton, NJ 08544}
35: \altaffiltext{\theaddress}{\stepcounter{address}\label{APO}
36: Apache Point Observatory, P.O. Box 59, Sunspot, NM 88349}
37: \altaffiltext{\theaddress}{\stepcounter{address}\label{PennState}
38: Department of Astronomy and Astrophysics,
39: Pennsylvania State University, University Park, PA 16802}
40:
41: \begin{abstract}
42: We test the homogeneity of the Universe at $z\sim 0.3$ with the
43: Luminous Red Galaxy (LRG) spectroscopic sample of the Sloan Digital
44: Sky Survey. First, the mean number $N(R)$ of LRGs within completely
45: surveyed LRG-centered spheres of comoving radius $R$ is shown to be
46: proportional to $R^3$ at radii greater than $R\sim
47: 70\,h^{-1}~\mathrm{Mpc}$. The test has the virtue that it does not
48: rely on the assumption that the LRG sample has a finite mean density;
49: its results show, however, that there \emph{is} such a mean density.
50: Secondly, the survey sky area is divided into 10 disjoint solid
51: angular regions and the fractional rms density variations of the LRG
52: sample in the redshift range $0.2<z<0.35$ among these ($\sim
53: 2\times10^7\,h^{-3}~\mathrm{Mpc^3}$) regions is found to be 7~percent
54: of the mean density. This variance is consistent with typical biased
55: \lcdm\ models and puts very strong constraints on the quality of SDSS
56: photometric calibration.
57: \end{abstract}
58:
59: \keywords{
60: cosmology: observations
61: ---
62: cosmological parameters
63: ---
64: galaxies: statistics
65: ---
66: large-scale structure of universe
67: ---
68: methods: statistical
69: ---
70: surveys
71: }
72:
73: \section{Introduction}
74:
75: One of the principal assumptions of successful physical cosmological
76: models is \emph{homogeneity;} \ie, the assumption that sufficiently
77: large independent volumes of the Universe will contain similar mean
78: densities of matter (and everything else). In detail, any test of
79: this assumption becomes a quantitative one: Do the observed variations
80: of density agree with the predictions of the leading physical
81: theories? Inasmuch as astronomical observations are used to
82: \emph{rule out} physical theories, the homogeneity of the Universe
83: cannot be demonstrated definitively beyond this. This is not because
84: the observations suggest an inhomogeneous Universe, but rather because
85: there are no \emph{physical} inhomogeneous models to rule out!
86:
87: In fact the observations do strongly suggest homogeneity, and the
88: great success of the \lcdm\ model in explaining the statistical
89: fluctuations in the isotropic cosmic microwave background \citep[CMB;
90: \eg,][]{bennett03a}, the growth of present-day large-scale structure
91: \citep[\eg,][]{tegmark04a}, and the tight redshift-distance relation
92: inferred for type Ia supernovae \citep{schmidt98a, perlmutter99a}
93: should be taken as very strong evidence that the Universe is
94: homogeneous on large scales. It does not make sense to postulate
95: inhomogeneous matter distributions without providing a physical model
96: in which such distributions are consistent with modern observations.
97:
98: Nevertheless, in what follows we use the enormous Luminous Red Galaxy
99: (LRG) sample \citep{eisenstein01a} of the Sloan Digital Sky Survey
100: \citep[SDSS;][]{york00a} to test the homogeneity of the Universe with
101: the most conservative statistical tests we know. We may not have
102: physical inhomogeneous models to test, but such models \emph{may}
103: exist in the future, and homogeneity on large scales \emph{is} an
104: extremely strong prediction of \lcdm\ and its variants. Even if
105: homogeneity of the matter distribution is taken for granted,
106: homogeneity in the galaxy distribution is not guaranteed if there are,
107: \eg, factors contributing to galaxy formation that act over large
108: distances. For these reasons, cosmic homogeneity is worthy of study.
109:
110: Certainly the prevailing view in modern cosmology is that homogeneity
111: has been very well established by the observed isotropy of the CMB
112: \citep{partridge67a,wilson67a,smoot92,bennett03a}, x-ray background
113: \citep[\eg,][]{scharf00a}, and the isotropies of various source
114: populations, \eg, radio galaxies \citep{peebles93a}. In the context
115: of physical models in which the CMB is emitted during recombination at
116: early times, its isotropy is indeed a very strong argument for
117: homogeneity. However it is certainly possible to imagine
118: inhomogeneous distributions of finite-sized (redshift-dependent)
119: blackbody sources such that every line of sight happens to terminate
120: on the surface of one of them. In other words, isotropy does not by
121: itself guarantee homogeneity. The fundamental issue is that
122: homogeneity is a property of distributions in three-dimensional space
123: and isotropy is a property of distributions on a two-dimensional sky.
124:
125: Isotropy of a population of discrete sources merits additional
126: discussion. As with the CMB, the isotropy of the point distribution
127: itself does not guarantee homogeneity, as there can in principle be
128: inhomogeneous three-dimensional distributions that project to
129: isotropic distributions on the sky at high probability for typical
130: observers \citep{durrer97a}. However, we do expect that isotropy of
131: such discrete populations provides strong evidence for homogeneity
132: when combined with the observation that flux and redshift
133: distributions are similar in different directions. Isotropy itself is
134: a two-dimensional test and therefore not sufficient, but isotropy as a
135: function of flux or redshift is indeed a three-dimensional test and
136: probably does establish homogeneity, although at lower precision than
137: the tests we perform below. Of course these statements (or their
138: contraries) are hard to maintain with great confidence in the absence
139: of any inhomogeneous physical model.
140:
141: By far the cleanest tests of homogeneity involve simply counting
142: sources in three dimensional regions. Homogeneity is established---in
143: principle---when the three-space correlation function of galaxies can
144: be shown to vanish over a range of scales at large scales.
145: Unfortunately, most investigations of the correlation function use the
146: largest scales available in the survey under analysis to determine the
147: mean density (sometimes with an integral constraint correction). The
148: estimation of clustering statistics involves subtraction of the mean
149: density, so it is in some sense required by the methodology that the
150: clustering tends to zero at the largest available scales. Indeed, it
151: has sometimes been the case that clustering amplitudes have been
152: underestimated in small surveys. For these reasons the vanishing of
153: the correlation function only establishes homogeneity if it is shown
154: to vanish over a substantial \emph{range} of scales, \eg, as it has
155: been shown to do for optically selected QSOs on scales of 100 to
156: $2000\,h^{-1}~\mathrm{Mpc}$ \citep{croom04a}.
157:
158: The Sloan Digital Sky Survey Luminous Red Galaxy sample is ideal for
159: performing these tests in an extremely conservative manner. The
160: sample used here fills a huge volume ($\sim
161: 0.6\,h^{-3}~\mathrm{Gpc^3}$) with high spectroscopic completeness. It
162: contains only galaxies of a limited range of luminosities and it is
163: close to volume-limited \citep{eisenstein01a}. The sample contains
164: large numbers; there are 55,000 used below. The SDSS footprint also
165: has a good angular shape for this project; we show below that the LRG
166: sample contains many independent filled spheres of comoving radius
167: $R>100\,h^{-1}~\mathrm{Mpc}$. Previous work on cosmic homogeneity,
168: some of which has concluded that there is evidence for a fractal-like
169: galaxy distribution \citep{syloslabini98a,joyce99a}, have not had
170: samples that have measured such large scales so cleanly.
171:
172: In what follows, a cosmological world model with
173: $(\Omega_\mathrm{M},\Omega_\mathrm{\Lambda})=(0.3,0.7)$ is adopted,
174: and the Hubble constant is parameterized
175: $H_0=100\,h~\mathrm{km\,s^{-1}\,Mpc^{-1}}$, for the purposes of
176: calculating distances\citep[\eg,][]{hogg99cosm}.
177:
178: \section{The LRG sample}
179:
180: The SDSS \citep{york00a,stoughton02a,abazajian03a,abazajian04a} is
181: conducting an imaging survey of $10^4$ square degrees in 5 bandpasses
182: $u$, $g$, $r$, $i$, and $z$ \citep{fukugita96a,gunn98a}. Photometric
183: monitoring \citep{hogg01a}, image processing
184: \citep{lupton01a,stoughton02a,pier03a}, and good photometric
185: calibration \citep{smith02a} allow one to select galaxies
186: \citep{strauss02a,eisenstein01a}, quasars \citep{richards02a}, and
187: stars for follow-up spectroscopy with twin fiber-fed
188: double-spectrographs. The spectra cover 3800\AA\ to 9200\AA\ with a
189: resolution of 1800. Targets are assigned to plug plates with a tiling
190: algorithm that ensures nearly complete samples \citep{blanton03a}.
191:
192: We focus here on the luminous red galaxy spectroscopic sample
193: \citep{eisenstein01a}. This uses color-magnitude cuts in $g$, $r$,
194: and $i$ to select galaxies that are likely to be luminous early-type
195: galaxies at redshifts between 0.15 and 0.5. The selection is highly
196: efficient and the redshift success rate is excellent. The sample is
197: constructed so as to be close to volume-limited up to $z=0.36$, with a
198: dropoff in density towards $z=0.5$. The comoving number density of
199: the sample is close to that required to maximize the signal-to-noise
200: ratio on the large-scale power spectrum.
201:
202: The sample we use here is drawn from NYU LSS {\tt sample14}
203: \citep{blanton04a} and covers 3,836 square degrees containing 55,000
204: LRGs between redshift of 0.16 and 0.47. We use an absolute $g$-band
205: magnitude cut of $-21.2$, including $k$-corrections and evolution to
206: $z=0.3$. The details of the radial and angular selection functions
207: are described in \citet{zehavi04a}. The radial modeling of the
208: expected number of galaxies as a function of redshift is based closely
209: on the observed distribution. The exact survey geometry is expressed
210: in terms of spherical polygons. We exclude the few regions that have
211: less than 48~percent spectroscopy coverage. We weight all of our
212: counts by the inverse of the angular selection function, including an
213: explicit correction for fiber collisions, to bring the weighted LRG
214: catalog to a uniform mean angular density within the chosen sky
215: region.
216:
217: We create large catalogs of randomly distributed points based on these
218: angular and radial models. These catalogs match the distribution of
219: the LRGs in redshift and are isotropic within the survey region.
220: These catalogs allow us to check the survey completeness of any given
221: volume and provide a homogeneous baseline (\eg, expected numbers) for
222: the tests that follow.
223:
224: The LRG sample depends considerably on the photometric uniformity of
225: the SDSS. Photometric calibration is involved at two different stages
226: in this work: The first stage is at initial target selection, because
227: the LRG spectroscopic targets are selected on the imaging photometry,
228: calibrated via the Photometric Telescope \citep{smith02a}. An 0.01 mag
229: shift in $(r-i)$ color or a 0.03 mag shift in $r$ magnitude would
230: modulate the LRG target density by up to 10~percent at most redshifts
231: \citep{eisenstein01a}. The second stage comes when the spectroscopic
232: LRG sample is ``cut down'' to the volume limited sample used here by
233: an additional absolute magnitude cut. This cut is based on improved
234: calibration (using survey scan stripe overlap regions). The sample
235: density is very sensitive to small shifts in calibration at this stage
236: also.
237:
238: The homogeneity that we find below is a testament to the calibration
239: of the survey. It appears that while the survey may have small
240: regions or even scan stripes that are miscalibrated (at $\sim
241: 2$~percent rms), the large-scale photometric homogeneity is better
242: than 1~percent in $(r-i)$ color, \ie, the errors average down on large
243: scales. At no point is the uniformity of the LRG sample assumed or
244: used in setting calibration or calibration parameters.
245:
246: \section{Conditioned density scaling}
247:
248: The fundamental test of the homogeneity of a point set, which makes no
249: assumption about even the existence of a mean density, is the
250: measurement of the scaling of the average number of neighboring points
251: to any given point, as a function of maximum separation
252: \citep{pietronero87a,gabrielli04a}. In a homogenous distribution, at
253: large enough distances, this number $N$ scales with the maximum
254: distance $R$ as $R^D$, where $D$ is the dimensionality of the space.
255:
256: Because our goal is to be extremely conservative, for this test, we
257: considered neighbors not of all LRGs, but only of those ``target''
258: LRGs in the redshift range $0.20<z<0.40$ that have enormous spherical
259: comoving regions around them that have been completely surveyed by the
260: SDSS spectroscopic survey. In detail, we required that \emph{both}
261: the $50$ \emph{and} the $100\,h^{-1}$~Mpc radius comoving sphere
262: around each target galaxy to be at least 95~percent complete (in the
263: sense of 95~percent coverage by the spectroscopic survey, which itself
264: is complete at the 94~percent level for LRGs). The sky distribution
265: of the 3658 LRGs that satisfy this highly restrictive selection are
266: shown in Figure~\ref{fig:sky}.
267:
268: In Figure~\ref{fig:density} the scaling of the average number $N$ of
269: neighbors of the target galaxies as a function of radius $R$ is shown,
270: with an ``expected number'' divided out. The expected number is
271: estimated by counting the number of ``random'' points in the spherical
272: volume, where the random points are from a homogeneous random catalog
273: described above with the same footprint, angular variation of
274: completeness, and radial selection function as the LRGs, but 100 times
275: the total number. Since the sample being used is highly complete,
276: this is essentially equivalent to dividing all the counts by a
277: constant times $R^3$. Figure~\ref{fig:density} shows that the mean
278: number $N$ scales as $R^3$ at large distances (and $R^2$ at small
279: distances).
280:
281: Also shown in Figure~\ref{fig:density} is the average number for each
282: of five samples, split by RA by the RA lines shown in
283: Figure~\ref{fig:sky}. Each of the five sky regions shows the $R^3$
284: scaling individually. The expected number densities from the random
285: catalogs has been kept constant across all regions. This test is
286: particularly conservative, because the five sky regions contain
287: (mostly) independent spheres and are expected to have different
288: observing and calibration properties.
289:
290: \section{Density variation}
291:
292: In an inhomogeneous universe (absent a physical inhomogeneous model),
293: the natural expectation is for order unity differences in population
294: densities on all scales. For this reason we performed a test in which
295: the sky is split into disjoint regions and the number density of LRGs
296: is tested in a comoving volume in each of the disjoint regions.
297:
298: Figure~\ref{fig:squeeze} shows the redshift $0.20<z<0.35$ LRG sample
299: split into 10 disjoint sky regions, each of which is of comparable
300: total solid angle and therefore (for this redshift range) comoving
301: volume. Each region corresponds to a comoving volume of roughly
302: $2.2\times 10^7\,h^{-3}~\mathrm{Mpc^3}$. Figure~\ref{fig:counts}
303: shows the relative densities of these regions. The rms scatter
304: between the 10 regions is 7.3~percent. Subtracting the expected
305: Poisson variation among these regions yields a relative rms scatter of
306: 7.0~percent.
307:
308: In a $\Lambda$CDM model with $\sigma_8=0.85$, the mass density
309: fluctuations among this set of sky regions in the redshift range
310: $0.20<z<0.35$ is expected to be 3.1~percent. The LRG real-space
311: autocorrelation has been measured to have an amplitude $\sigma_8=1.80$
312: \citep{zehavi04a}, so that the galaxies have a bias of around 2, and
313: one would predict a variation of roughly 6.5~percent. Hence, the
314: 7.0~percent variance in the LRG sample does not leave much room for
315: calibration errors, which ought to come in in quadrature. For the LRG
316: sample, 1~percent calibration variations in the $r-i$ color or
317: 3~percent variations in the $r$ flux would produce $\sim 10$~percent
318: variations in the LRG number density \citep{eisenstein01a}, so the
319: consistency of this measurement with the biased CDM description of the
320: LRG population puts extremely strong constraints on the quality of
321: SDSS photometric calibration. Note that the LRG target selection is
322: never used to tune the photometric calibration, so there is no sense
323: in which this uniformity is enforced directly.
324:
325: \section{Discussion}
326:
327: The extremely large (in both volume and number) and complete LRG
328: sample from the SDSS was used to test the three-dimensional
329: homogeneity (\ie, not just isotropy) of the Universe at $z\sim 0.3$ by
330: the most conservative possible method. We find that the Universe has
331: a well-defined mean density and that it is homogeneous. Furthermore,
332: the variations we see in the density of LRGs on large scales is
333: consistent with the predictions of a biased \lcdm\ cosmogonic model.
334:
335: The number $N(R)$ of LRGs within LRG-centered spheres of comoving
336: radius $R$ is shown to be proportional to $R^3$ at radii greater than
337: $R\sim 70\,h^{-1}~\mathrm{Mpc}$. For this test we used not all LRGs
338: as central LRGs, but only those in the centers of highly complete
339: spheres, so there is no ``interpolation'' or ``extrapolation'' over
340: unobserved regions. In the terminology of fractals, this test shows
341: that the fractal dimension of the LRG distribution is very close to
342: $D=3$. The test has the virtue that it does not rely on the
343: assumption that the LRG sample has a finite mean density; its results
344: show, however, that there \emph{is} such a mean density.
345:
346: The $D=3$ result presented here is in qualitative disagreement with
347: some previous studies. There are some similar kinds of analyses of
348: redshift surveys that show $D\approx 2$ \citep{syloslabini98a,
349: joyce99a}, but there is no quantitative disagreement, because the most
350: robust measurements of $D$ are generally made at scales much smaller
351: than those measured by the LRGs here. Indeed, Figure~2 shows that
352: $D=2$ is a remarkably good fit out to roughly
353: $20\,h^{-1}~\mathrm{Mpc}$. Analyses of the ESO Slice Project, in
354: which the scaling of the number of galaxies in the sample is measured
355: as a function of the redshift depth to which they are counted
356: \citep{scaramella98a,joyce99b}, depend crucially on $K$ corrections,
357: evolution, and world model.
358:
359: The survey sky area was divided into 10 disjoint solid angular regions
360: and the fractional rms density variations of the LRG sample in the
361: redshift range $0.2<z<0.35$ among these ($\sim
362: 2\times10^7\,h^{-3}~\mathrm{Mpc^3}$) regions was found to be 7~percent
363: of the mean density. This variance is consistent with typical biased
364: \lcdm\ models, with a bias of $\approx 2$, as is found for the LRG
365: sample at smaller scales \citep{zehavi04a}. This result confirms
366: homogeneity and supports the biased \lcdm\ model on the largest
367: observable scales.
368:
369: Finally, it is worthy of note that because the LRG sample selection is
370: so sensitive to photometric calibration, these results demonstrate
371: that the calibration of the SDSS in the $g$, $r$, and $i$ bands is
372: consistent at the sub-percent level when averaged on angular scales of
373: tens of degrees and larger. This is a tremendous technical
374: achievement and recommends the LRG sample and the SDSS for making many
375: extremely precise measurements in the future.
376:
377: \acknowledgments It is a pleasure to thank Michael Joyce and Francesco
378: Sylos Labini for the discussions which led to this project and the
379: Laboratoire de Physique Th\'eorique, Universit\'e de Paris XI, for
380: hospitality during that time. Additional thanks are due to Jim
381: Peebles and Michael Strauss for useful discussions. DWH and MRB are
382: partially supported by NASA (grant NAG5-11669) and NSF (grant
383: PHY-0101738). DJE is supported by NSF (grant AST-0098577) and an
384: Alfred P.~Sloan Research Fellowship.
385:
386: Funding for the creation and distribution of the SDSS Archive has been
387: provided by the Alfred P. Sloan Foundation, the Participating
388: Institutions, the National Aeronautics and Space Administration, the
389: National Science Foundation, the U.S. Department of Energy, the
390: Japanese Monbukagakusho, and the Max Planck Society. The SDSS Web site
391: is http://www.sdss.org/.
392:
393: The SDSS is managed by the Astrophysical Research Consortium (ARC) for
394: the Participating Institutions. The Participating Institutions are the
395: University of Chicago, Fermilab, the Institute for Advanced Study, the
396: Japan Participation Group, the Johns Hopkins University, the Korean
397: Scientist Group, Los Alamos National Laboratory, the
398: Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute
399: for Astrophysics (MPA), New Mexico State University, the University of
400: Pittsburgh, Princeton University, the United States Naval Observatory,
401: and the University of Washington.
402:
403: \begin{thebibliography}{28}
404: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
405:
406: \bibitem[{Abazajian \etal(2003)}]{abazajian03a}
407: Abazajian, K. \etal\ 2003, \aj, 126, 2081
408:
409: \bibitem[{{Abazajian} \etal(2004)}]{abazajian04a}
410: {Abazajian}, K. \etal\ 2004, \aj, 128, 502
411:
412: \bibitem[{{Bennett} \etal(2003)}]{bennett03a}
413: {Bennett}, C.~L. \etal\ 2003, \apjs, 148, 1
414:
415: \bibitem[{{Blanton} \etal(2001)}]{blanton01a}
416: {Blanton}, M.~R. \etal\ 2001, \aj, 121, 2358
417:
418: \bibitem[{{Blanton} \etal(2003){Blanton}, {Lin}, {Lupton}, {Maley}, {Young},
419: {Zehavi}, \& {Loveday}}]{blanton03a}
420: {Blanton}, M.~R., {Lin}, H., {Lupton}, R.~H., {Maley}, F.~M., {Young}, N.,
421: {Zehavi}, I., \& {Loveday}, J. 2003, \aj, 125, 2276
422:
423: \bibitem[{Blanton \etal(2004)}]{blanton04a}
424: Blanton, M.~R. \etal\ 2004, \aj, submitted
425:
426: \bibitem[{Croom \etal(2004)}]{croom04a}
427: Croom, S.~M., Boyle, B.~J., Shanks, T., Smith, R.~J., Miller, L.,
428: Outram, P.~J., Loaring, N.~S., Hoyle, F., \& da~Angela, J. 2004,
429: \mnras, in press
430:
431: \bibitem[{Durrer \etal(1997)}]{durrer97a}
432: Durrer, R., Eckmann, J.-P.,
433: {Sylos~Labini}, F., Montuori, M., \& Pietronero, L. 1997,
434: Europhys.\ Lett., 40, 491
435:
436: \bibitem[{Eisenstein \etal(2001)}]{eisenstein01a}
437: Eisenstein, D.~J. \etal\ 2001, \aj, 122, 2267
438:
439: \bibitem[{Fukugita \etal(1996)}]{fukugita96a}
440: Fukugita, M., Ichikawa, T., Gunn, J.~E., Doi, M., Shimasaku, K., \& Schneider,
441: D.~P. 1996, \aj, 111, 1748
442:
443: \bibitem[{Gabrielli \etal(2004)}]{gabrielli04a}
444: Gabrielli, A., Sylos Labini, F., Joyce, M., \& Pietronero, L., 2004,
445: Statistical Physics for Cosmic Structure (Springer Verlag)
446:
447: \bibitem[{Gunn \etal(1998)}]{gunn98a}
448: Gunn, J.~E., Carr, M.~A., Rockosi, C.~M., Sekiguchi, M., \etal\ 1998, \aj,
449: 116, 3040
450:
451: \bibitem[{Hogg(1999)}]{hogg99cosm}
452: Hogg, D.~W. 1999, astro-ph/9905116
453:
454: \bibitem[{{Hogg} \etal(2001){Hogg}, {Finkbeiner}, {Schlegel}, \&
455: {Gunn}}]{hogg01a}
456: {Hogg}, D.~W., {Finkbeiner}, D.~P., {Schlegel}, D.~J., \& {Gunn}, J.~E. 2001,
457: \aj, 122, 2129
458:
459: \bibitem[{{Joyce} \etal(1999{\natexlab{a}}){Joyce}, {Montuori}, \&
460: {Sylos~Labini}}]{joyce99a}
461: {Joyce}, M., {Montuori}, M., \& {Sylos~Labini}, F. 1999{\natexlab{a}}, \apjl,
462: 514, L5
463:
464: \bibitem[{{Joyce} \etal(1999{\natexlab{b}}){Joyce}, {Montuori},
465: {Sylos~Labini}, \& {Pietronero}}]{joyce99b}
466: {Joyce}, M., {Montuori}, M., {Sylos~Labini}, F., \& {Pietronero}, L.
467: 1999{\natexlab{b}}, \aap, 344, 387
468:
469: \bibitem[{Lupton \etal(2001)Lupton, Gunn, {Ivezi{\'c}}, Knapp, Kent, \&
470: Yasuda}]{lupton01a}
471: Lupton, R.~H., Gunn, J.~E., {Ivezi{\'c}}, Z., Knapp, G.~R., Kent, S., \&
472: Yasuda, N. 2001, in ASP Conf.\ Ser.\ 238:\ Astronomical Data Analysis
473: Software and Systems X, Vol.~10, 269
474:
475: \bibitem[{Partridge \& Wilkinson(1967)}]{partridge67a}
476: Partridge, R.~B. \& Wilkinson, D.~T. 1967, PRL, 18, 557
477:
478: \bibitem[{{Peebles}(1993)}]{peebles93a}
479: {Peebles}, P.~J.~E. 1993, Principles of physical cosmology (Princeton
480: University Press)
481:
482: \bibitem[{{Perlmutter} \etal(1999)}]{perlmutter99a}
483: {Perlmutter}, S. \etal\ 1999, \apj, 517, 565
484:
485: \bibitem[{Pier \etal(2003)Pier, Munn, Hindsley, Hennessy, Kent, Lupton, \&
486: {Ivezi{\' c}}}]{pier03a}
487: Pier, J.~R., Munn, J.~A., Hindsley, R.~B., Hennessy, G.~S., Kent, S.~M.,
488: Lupton, R.~H., \& {Ivezi{\' c}}, {\v Z}. 2003, \aj, 125, 1559
489:
490: \bibitem[{Pietronero \etal(1987)}]{pietronero87a}
491: Pietronero, L. 1987, Physica, 144, 257
492:
493: \bibitem[{Richards \etal(2002)}]{richards02a}
494: Richards, G. \etal\ 2002, \aj, 123, 2945
495:
496: \bibitem[{Scaramella \etal(1998)}]{scaramella98a}
497: Scaramella, R. \etal\ 1998, \aap, 334, 404
498:
499: \bibitem[{Scharf \etal(2000)}]{scharf00a}
500: Scharf, C.~A., Jahoda, K., Treyer, M., Lahav, O., Boldt, E., \&
501: Piran, T. 2000, \apj, 544, 49
502:
503: \bibitem[{{Schmidt} \etal(1998)}]{schmidt98a}
504: {Schmidt}, B.~P. \etal\ 1998, \apj, 507, 46
505:
506: \bibitem[{Smith, Tucker \etal(2002)}]{smith02a}
507: Smith, J.~A., Tucker, D.~L., \etal\ 2002, \aj, 123, 2121
508:
509: \bibitem[{Smoot \etal(1992)}]{smoot92}
510: Smoot, G.~F., et al, 1992, ApJ, 396, L1
511:
512: \bibitem[{Stoughton \etal(2002)}]{stoughton02a}
513: Stoughton, C. \etal\ 2002, \aj, 123, 485
514:
515: \bibitem[{Strauss \etal(2002)}]{strauss02a}
516: Strauss, M.~A. \etal\ 2002, \aj, 124, 1810
517:
518: \bibitem[{{Sylos~Labini} \etal(1998){Sylos~Labini}, Montuori, \&
519: Pietronero}]{syloslabini98a}
520: {Sylos~Labini}, F., Montuori, M., \& Pietronero, L. 1998, \physrep, 293, 61
521:
522: \bibitem[{{Tegmark} \etal(2004)}]{tegmark04a}
523: {Tegmark}, M. \etal\ 2004, \apj, 606, 702
524:
525: \bibitem[{{Wilson} \& {Penzias}(1967)}]{wilson67a}
526: {Wilson}, R.~W. \& {Penzias}, A.~A. 1967, Science, 156, 1100
527:
528: \bibitem[{York \etal(2000)}]{york00a}
529: York, D. \etal\ 2000, \aj, 120, 1579
530:
531: \bibitem[{Zehavi \etal(2004)}]{zehavi04a}
532: Zehavi, I. \etal\ 2004, \apj, submitted
533:
534: \end{thebibliography}
535:
536: \clearpage
537: \begin{figure}
538: \plotone{f1.eps}
539: \caption{The entire LRG sample used in this study is shown in grey.
540: The 3658 LRGs in the redshift range $0.2<z<0.4$ with $>95$~percent
541: complete volumetric coverage inside the surrounding comoving spheres
542: of radii $R=50$ and $100\,h^{-1}~\mathrm{Mpc}$ are shown in black.
543: Also shown are grey lines that split the 3658 points into five
544: quantiles by RA.\label{fig:sky}}
545: \end{figure}
546:
547: \clearpage
548: \begin{figure}
549: \plotone{f2.eps}
550: \caption{The average comoving number density (\ie, number counted
551: divided by expected number from a homogeneous random catalog) of LRGs
552: inside comoving spheres centered on the 3658 LRGs shown in
553: Figure~\ref{fig:sky}, as a function of comoving sphere radius $R$.
554: The average over all 3658 spheres is shown in with open squares, and
555: the averages of each of the five RA quantiles is shown as a separate
556: grey line. At small scales, the number density drops with radius
557: because the LRGs are clustered; at large scales the number density
558: approaches a constant because the sample is
559: homogeneous.\label{fig:density}}
560: \end{figure}
561:
562: \clearpage
563: \begin{figure}
564: \plotone{f3.eps}
565: \caption{The LRGs in the redshift range $0.20<z<0.35$, divided into 10
566: disjoint sky regions (where LRGs are labeled and marked by dots and
567: squares), with inter-region gaps (where LRGs are marked by dots
568: only).\label{fig:squeeze}}
569: \end{figure}
570:
571: \clearpage
572: \begin{figure}
573: \plotone{f4.eps}
574: \caption{The relative comoving number densities of LRGs in the
575: redshift range $0.20<z<0.35$ in the 10 disjoint sky areas shown in
576: Figure~\ref{fig:squeeze}. The quoted ``$\sigma$'' is the rms scatter
577: of the points, with the expected Poisson rms subtracted out in
578: quadrature.\label{fig:counts}}
579: \end{figure}
580:
581: \end{document}
582: