1: % AA vers. 6.1, LaTeX class for Astronomy & Astrophysics
2: %\documentclass[referee]{aa} % for a referee version
3: %\documentclass[onecolumn]{aa} % for a paper on 1 column
4: %\documentclass[longauth]{aa} % for the long lists of affiliations
5: %\documentclass[rnote]{aa} % for the research notes
6: %\documentclass[letter]{aa} % for the letters
7: \documentclass{aa}
8: %\documentclass[referee]{aa}
9: \topmargin -1.5cm
10: \usepackage{graphicx}
11: \usepackage{txfonts}
12: \newcommand{\be}{\begin{equation}}
13: \newcommand{\ee}{\end{equation}}
14: \newcommand{\lb}[1]{\label{#1}}
15: \newcommand{\sty}{\scriptstyle}
16: \newcommand{\ssty}{\scriptscriptstyle}
17: \newcommand{\apg}{\:^{>}_{\sim}\:}
18: \newcommand{\apl}{\:^{<}_{\sim}\:}
19: \newcommand{\etal}{{\it et al.\ }}
20: \newcommand{\dl}{d_{\ssty L}}
21: \newcommand{\da}{d_{\ssty A}}
22: \newcommand{\dg}{d_{\ssty G}}
23: \newcommand{\dz}{d_{\ssty Z}}
24: %\newcommand{\an}{\langle n \rangle}
25: \newcommand{\an}{\left[ n \right]}
26: \newcommand{\arho}{\langle \rho \rangle}
27: \newcommand{\g}{\gamma}
28: \newcommand{\gest}{\gamma^\ast}
29: \newcommand{\oito}{\! \! \! \! \! \! \! \!}
30: \newcommand{\onze}{\! \! \! \! \! \! \! \! \! \! \!}
31: \newcommand{\dezessete}{\! \! \! \! \! \! \! \! \! \! \! \! \! \! \! \! \!}
32: \newcommand{\ud}{^{^{\onze \ssty D}}}
33: \newcommand{\exud} {^{^{\dezessete \ssty D}}}
34: \newcommand{\Nid}{\; N_i\ud}
35: \newcommand{\nid}[1]{\; N_{\ssty {#1}}^{^{\onze \ssty D}}}
36: \newcommand{\anid}{\; \an_i\exud}
37: \newcommand{\gammaid}{\; \: \gamma_i^{{\oito \ssty D}}}
38: \newcommand{\gestid}{\; \: {\gamma_i^\ast}^{{\onze \ssty D}}\; \; \: \:}
39: \newcommand{\gamid}[2]{\; \: \gamma_{\ssty #1}^{{\oito \ssty #2}}}
40: \newcommand{\gamsid}[2]{\;\:{\gamma_{\ssty #1}^\ast}^{{\onze\ssty #2}}\;\;\:\:}
41: \newcommand{\NID}[2]{\; N_{\ssty {#1}}^{^{\onze \ssty #2}}}
42: \begin{document}
43: \title{Spatial and observational homogeneities of the
44: galaxy distribution in standard cosmologies}
45: \titlerunning{Types of homogeneities in standard cosmologies}
46: % \subtitle{}
47: \author{L. J. Rangel Lemos\thanks{\emph{Present address:} %ICRA,
48: International Center for Relativistic Astrophysics, Dept.\
49: of Physics, University of Rome ``La Sapienza'',
50: Rome, Italy
51: }
52: \inst{1}
53: \and
54: M. B. Ribeiro\inst{2}
55: }
56: \offprints{M. B. Ribeiro}
57: \institute{Department of Astronomy, Valongo Observatory, University
58: of Brazil -- UFRJ, Rio de Janeiro, Brazil\\
59: \and
60: Physics Institute, University of Brazil -- UFRJ,
61: CxP 68532, Rio de Janeiro, RJ 21941-972, Brazil\\
62: \email{mbr@if.ufrj.br}
63: }
64: \date{Received ; accepted }
65: % \abstract{}{}{}{}{}
66: % 5 {} token are mandatory
67: \abstract
68: % context heading (optional)
69: % {} leave it empty if necessary
70: % {Empirical verification of the geometrical concept of homogeneity
71: % of the standard relativistic cosmology considering its various
72: % definitions of distance and emphasizing that astronomical
73: % observations occur along the past light cone.}
74: {An important aim of standard relativistic cosmology is the empirical
75: verification of its geometrical concept of homogeneity by
76: considering various definitions of distance and astronomical
77: observations occurring along the past light cone.}
78: % aims heading (mandatory)
79: {We analyze the physical consequences of distinguishing between
80: \emph{spatial homogeneity} (SH), defined by the Cosmological
81: Principle, and \emph{observational homogeneity} (OH). We argue
82: that OH is falsifiable by means of astronomical observations,
83: whereas SH can be verified only indirectly.}
84: % methods heading (mandatory)
85: {We simulate observational counts of cosmological sources, such as
86: galaxies, by means of a generalized number-distance expression
87: that can be specialized to produce either the counts of the Einstein-de
88: Sitter (EdS) cosmology, which has SH by construction, or other types
89: of counts, which do, or do not, have OH by construction. Expressions
90: for observational volumes are derived using the various
91: cosmological-distance definitions in the EdS cosmological model. The
92: observational volumes and simulated counts are then used to derive
93: differential densities. We present the behavior of
94: these densities for increasing redshift values.}
95: % results heading (mandatory)
96: {Simulated counts that have OH by construction do not always exhibit
97: SH features. The reverse situation is also true. In addition,
98: simulated counts with no OH features at low redshift begin to show OH
99: characteristics at high redshift. The comoving distance appears to be
100: the only distance definition for which both SH and OH are applicable
101: simultaneously, even though with limitations.}
102: % conclusions heading (optional), leave it empty if necessary
103: {We demonstrate that observations indicative of a possible absence
104: of OH do \emph{not} necessarily falsify the standard Friedmannian
105: cosmology, which implies that this cosmology does not always
106: produce observable homogeneous densities. We conclude that using
107: different cosmological distances in the characterization of the
108: galaxy distribution can produce significant ambiguities in reaching
109: conclusions about the large-scale galaxy distribution in the Universe.}
110: \keywords{cosmology: theory --
111: large-scale structure of the universe --
112: galaxy distribution --
113: relativity
114: }
115:
116: \maketitle
117: %________________________________________________________________
118:
119: \section{Introduction}
120:
121: The determination of whether or not the large-scale distribution
122: of matter in the Universe reaches a homogeneous distribution on
123: some redshift scale has been a disputed topic in observational
124: cosmology for many decades. With the availability of increasingly
125: larger galaxy redshift survey databases stemming from increasingly
126: more complete and deeper galaxy samples, there have been renewed
127: efforts to solve this problem by means of statistical techniques
128: of growing sophistication. However, this issue has still not been
129: settled because contradictory results have been reported by various
130: authors who support opposite claims (Ribeiro \& Miguelote 1998;
131: Sylos Labini et al.\ 1998; Joyce et al.\ 1999, 2000, 2005; Gabrielli
132: et al.\ 2005; Hogg et al.\ 2005; Jones et al.\ 2005; Yadav et al.\
133: 2005).
134:
135: The conventional wisdom on this topic claims that the solution to
136: the problem lies in our ability to acquire more and better data, that
137: is, as more complete galaxy samples at both lower and higher redshift
138: ranges become available the statistical techniques currently applied
139: to the analysis of these data should suffice to settle the controversy.
140: In other words, this view assumes implicitly or explicitly that more
141: complete galaxy redshift survey samples will eventually clarify this
142: point once and for all. The difficulty with the conventional wisdom is
143: that we have witnessed an enormous improvement in the technological
144: methods for astronomical data acquisition and, as a direct result of
145: those technological advances, have indeed obtained higher quality
146: and more complete galaxy redshift survey data sets. Despite these
147: technical advances, the controversy, however, continues to resurface
148: (see Ribeiro 1994 for a brief historical account of this debate in the
149: past century). Ribeiro (1992) proposed that the source of the
150: controversy was not the observations themselves, but the conceptual
151: tools used to analyse the observations. The initial ideas were
152: gradually refined and an alternative perspective developed (Ribeiro
153: 2001b, 2005; Albani et al.\ 2007). This perspective suggests that
154: improvements in observational techniques and the acquisition of larger
155: and larger galaxy data samples is not the path that will shed light
156: and clarify this debate.
157:
158: This alternative conceptual framework grew out of various
159: works (Ribeiro 1992ab, 1993, 1994, 1995, 2001a; Abdalla et al.\
160: 2001) and was comprehensively analyzed in Ribeiro (2001b; hereafter
161: R01b). Albani et al.\ (2007; hereafter A07) further developed this
162: alternative perspective in a somewhat condensed version. First
163: principles were used to point out that General Relativity allows us
164: to define two different concepts of homogeneity perfectly applicable
165: to cosmological models: \emph{spatial homogeneity} (SH) and
166: \emph{observational homogeneity} (OH). Furthermore, it was argued
167: that the Cosmological Principle is based on the concept of SH,
168: whereas the astronomical search for the possible homogeneity
169: of the Universe occurs mostly in the context of OH: this is
170: because astronomical observations are completed where OH is defined,
171: that is, along the backward null cone. The Cosmological Principle
172: implies that SH is not directly observable on space-like surfaces of
173: constant time defined in the Friedmann-Lema\^{\i}tre-Robertson-Walker
174: (FLRW) cosmologies. On the other hand, the geometrical locus of OH is
175: not along those space-like surfaces of constant time, but along the
176: past null cone, and so OH will only occur if a density measured
177: directly from observations, usually galaxy number counts, remains
178: constant along these null surfaces. These two concepts of homogeneity
179: do overlap, but they are \emph{not} the same.\footnote{In earlier works
180: (Ribeiro 1992ab, 1993, 1994, 1995), the term ``apparent homogeneity'' was
181: used instead of the currently adopted term ``observational
182: homogeneity.''} Reports of the searches for the homogenization of the
183: matter distribution in the Universe have not, for the most part,
184: acknowledged this important difference since the use of the generic, but
185: ambiguous, term `homogeneity' has become commonplace (see Section
186: \ref{rel} below). As discussed in R01b and A07, such a distinction arises
187: only if one takes an entirely relativistic perspective of this problem.
188:
189: Nevertheless, acknowledging the existence of these two types of
190: homogeneities is not enough to clarify the controversial points
191: outlined above. One has to go a step further because the only
192: way to discriminate SH from OH is by building observational
193: densities using different distance measures with the same number
194: count data. Bearing in mind the conceptual framework summarized
195: above, R01b and Ribeiro (2005; hereafter R05) showed that
196: although the Einstein-de Sitter (EdS) cosmology has SH by
197: construction, it may, or may not exhibit OH, since the possible
198: presence of OH depends on the distance measure chosen
199: in the statistical analysis of its EdS theoretically derived
200: number count expression as a function of the redshift. Such a
201: result was confirmed by A07 using observations, for instance, by
202: means of number counts extracted from the Canadian
203: Network Observational Cosmology 2 (CNOC2) galaxy redshift survey
204: and applied to two types of standard cosmologies, EdS and FLRW
205: with $\Omega_{m_0} =0.3$, $\Omega_{\Lambda_0}=0.7$.
206:
207: When these two types of homogeneities are discussed, a question
208: which also arises is how deep the observations must be to be able
209: to distinguish SH from OH. In other words, we must determine
210: whether or not the redshift ranges of current galaxy surveys
211: are sufficiently deep to be able to detect this difference. It is
212: important to point out that previous work
213: (Ribeiro 1995) indicated that due to the high nonlinearity of
214: General Relativity the distinction between SH and OH effects
215: may occur, in theory, at redshifts as low as $z \apl 0.1$,
216: depending on the chosen relativistic cosmological model and the
217: observational quantity under study.
218:
219: The aim of this paper is to analyze further the issues discussed
220: above. Our goal here is to extend the studies presented in R01b,
221: R05, and A07. These papers started with SH by construction and
222: sought to show whether or not OH was also featured in the models.
223: R01b and R05 initiated their analysis from the theoretical EdS
224: number counts, whereas A07 used observed number counts
225: extracted from the luminosity function. We aim here to investigate
226: the opposite situation, i.e., start with OH by construction and then
227: investigate whether or not the models show SH. Instead of using
228: actual observations, our intention here is to
229: \textit{simulate observations} by means of a generalized
230: number-distance relation which may, or may not, produce OH from the
231: start and then search for possible SH in the resulting model.
232: {In addition, we shall analyze our results in two redshift
233: ranges, namely $0.001<z<0.1$ and $z>0.1$, since most direct
234: measurements of galaxy correlations have been limited to
235: $z \sim 0.1$. This aims to try to answer the question posed above
236: about the redshift depth which these effects begin to manifest
237: themselves, as well as probing which quantities could possibly
238: offer observational results from which we can attempt to obtain
239: observational evidence allowing us to discriminate between SH and
240: OH.}\footnote{It should be mentioned here that if one uses
241: the galaxy luminosity function data derived from various
242: surveys with a relativistic cosmology number count theory (Ribeiro
243: \& Stoeger 2003), one is able to indirectly obtain measurements of
244: galaxy number counts at $z \approx 1$, or at far higher
245: redshifts, where the distinction between SH and OH is easily
246: detected. See A07 and Iribarrem, Ribeiro \& Stoeger (2008, in
247: preparation).}
248:
249: Our results show that a model with no OH by construction does not
250: always remain that way at higher values of redshift. In fact,
251: with a specific distance definition we may have a model with OH
252: at low redshifts, but no OH at higher redshifts. We also found
253: that if we start a model with OH, it may or may not become SH.
254: These results imply that the use of different distance measures to
255: calculate cosmological densities produces significant ambiguities
256: in reaching conclusions about the behavior of the large-scale
257: distribution of galaxies in the Universe due to the impossibility
258: of uniquely characterizing densities from galaxy distribution
259: data. Conclusions on this matter reached by means of the use of just
260: one cosmological distance, usually comoving, should therefore be
261: seen as applicable to this distance measure only and are most likely
262: not valid generally. Therefore, the proposal of R05 that
263: \textit{observers should utilize all possible distance measures in
264: their data analysis} is reinforced here. It is the view of these authors
265: that \textit{this is the path towards clarifying the controversy
266: discussed above}.
267:
268: The plan of the paper is as follows. In Sect., \ref{sec2} we derive
269: the basic equations and definitions of observational distances,
270: areas, volumes, number counts, and densities in cosmology.
271: Section \ref{eds} reviews the results of R01b, R05, and A07,
272: where those definitions are applied to FLRW cosmological models,
273: this shows that although the EdS cosmology is spatially homogeneous
274: by construction, it has observational homogeneity only when the
275: comoving distance is adopted in calculating observational
276: densities. In Sect., \ref{sec4} we simulate models that do, or do
277: not, have observationally-homogeneous features, concluding
278: that even a model without OH at low redshifts may become
279: observationally homogeneous at higher values of $z$. Section
280: \ref{DSH} studies the asymptotic behavior of the fractal
281: dimension $D$ of the EdS model at $z\rightarrow 0$ and at the
282: Big Bang singularity hypersurface where $z \rightarrow \infty$,
283: showing that the former case leads to $D=3$ for all distance
284: measures. In contrast, the latter case implies that $D=0$ for all
285: distances, except the comoving distance where the value
286: $D=3$ remains unchanged for all $z$. { Section \ref{mags} discusses the
287: relationship between number counts $N$ and magnitudes by means
288: of the distance modulus $\mu$. We show that the same ambiguous
289: results obtained in Sect.\ \ref{sec4} for the $N \times z$
290: functions, constructed with the various distance definitions,
291: are also present in the $N \times \mu$ functions.} Section
292: \ref{rel} provides a conceptual discussion about the caveats of
293: the use of the generic term `homogeneity' in cosmological models,
294: arguing that observational homogeneity is a relative concept
295: entailed by the relativity of time intervals. Finally, Sect.\
296: \ref{conclu} summarizes the results obtained in this paper.
297:
298: %__________________________________________________________________
299:
300: \section{Distances, volumes, densities and number counts}\lb{sec2}
301:
302: As has been extensively argued elsewhere, measuring distances in
303: cosmology depends on circumstances, that is, on the method of
304: measurement (McVittie 1974, Sandage 1988, R01b, R05 and references
305: therein). This does not imply that distances cannot be
306: compared with each another. These are true, physical distances to
307: an object and they can indeed be compared simply because they are
308: distances to cosmological sources, mostly galaxies, for which
309: intrinsic physical characteristics can be determined
310: (intrinsic measurement, intrinsic luminosity, etc)
311: \emph{independently} of a cosmological model. In addition, the
312: reciprocity theorem (Ellis 2007) relates the various
313: distances to each another and allows conversions between them.
314:
315: That is the theory. In practice, however, due to technological
316: limitations and our incomplete knowledge of the physical processes
317: occurring in the evolution of galaxies, we are presently unable
318: to find those intrinsic measurements, that is, standard candles and
319: standard rods, for \emph{every} galaxy in a redshift survey.
320: Therefore, we are left to measure their redshifts only and, by
321: using a cosmological model, to relate those redshifts to some
322: distance. Textbooks and reviews discussing cosmological distance
323: definitions offer a plethora of names for the distance measures,
324: a fact which only adds confusion to, not infrequently, poorly
325: understood concepts about what is a distance to a cosmological
326: object, the definition that should be chosen, and the context for
327: choosing a certain definition and not another one. As soon as one
328: delves into this problem and reads how it is dealt with in the
329: literature, it becomes clear that familiar Newtonian concepts
330: slipped into a subject that can only be understood properly by
331: means of relativistic ideas. Thus, to avoid those Newtonian
332: concepts of absolute and unique definitions, which are not applicable
333: to the relativistic discussion proposed here, we must follow along
334: the wise footsteps of others and accept that \textit{there is no
335: such a thing as an unique cosmological distance: all are correct,
336: and all can be compared with each another.} To argue otherwise is to
337: allow Newtonian ideas to slip into a subject that is entirely
338: relativistic.
339:
340: The difficulties described above have, of course, been previously
341: perceived by others, such as observational cosmologists, who resorted
342: to the convenient convention of using, for
343: the most part, only one distance measure, the \emph{comoving
344: distance}. There are, however, three caveats to this practice.
345: Firstly, not all practitioners follow this convention and this
346: means that there are still those who are misled into treating
347: different distance definitions as if they were the same, when
348: they are not, and, worst of all, results derived from those
349: different distance measures are then improperly compared with each
350: other. This can obviously add even more confusion to an
351: already-confused subject. As we show below, the second caveat is
352: that the comoving distance implies that the task of distinguishing
353: spatial from observational homogeneity becomes very difficult.
354: Thirdly, by adopting just one distance as convention some may still
355: unconsciously fall into the familiar, but in this context very
356: misleading, \textit{Newtonian trap of believing that in cosmology
357: the distance to an object could be uniquely defined, when General
358: Relativity does not allow such a conclusion.}
359:
360: \subsection{Cosmological distances}
361:
362: Our proposed solution to these difficulties is therefore to use in
363: our analysis below all observational distances, the \emph{luminosity
364: distance} $\dl$, the \emph{area distance} $\da$, the \emph{galaxy
365: area distance} $\dg$, and the \emph{redshift distance}
366: $\dz$.\footnote{$\da$ is also known as `angular diameter distance',
367: `corrected luminosity distance' and `observer area distance'.
368: $\dg$ is also known as `effective distance', `angular size
369: distance', `transverse comoving distance' and `proper motion
370: distance' (see details and references in R01b and R05). It is
371: also possible to define another distance, the \textit{parallax
372: distance} $d_{\ssty P}$ due to galaxy parallaxes (Ellis 1971).
373: This distance is not often mentioned since galaxy parallaxes
374: cannot yet be measured.} These are quantities that can in principle,
375: be directly measured (R05; Ellis 2007). However, since we need to
376: assume a cosmological model to start with, other distances like the
377: comoving distance, proper distance, etc, can all be written in
378: terms of these four above (see below). The advantage of starting
379: with these observational four, which to simplify the notation
380: from now on shall be referred as $d_i \; (i={\sty A}, {\sty G},
381: {\sty L}, {\sty Z} \,)$, is to know beforehand that they are
382: defined along the past light cone, since they are observational
383: distances. Relating them to, for example, comoving distance, implies
384: that the solution of the null geodesic equation has therefore to be
385: included in the expression of the comoving distance. This is
386: standard practice, rarely mentioned, but, for the purposes of
387: this work, it must be stated explicitly to avoid confusion.
388:
389: These observational distances are related to each other by an
390: important result called the \textit{ reciprocity theorem},
391: or \textit{Etherington reciprocity law}, proven long ago by
392: Etherington (1933), which reads as follows,\footnote{
393: See Ellis (2007) for an appraisal of how fundamental this
394: theorem is in every aspect of modern cosmology.}
395: \be {(1+z)}^2 \da=(1+z) \: \dg = \dl.
396: \lb{eth}
397: \ee
398: This theorem is valid for \textit{all} cosmologies (Ellis 1971,
399: 2007; Pleba\'{n}ski and Krasi\'{n}ski 2006). In addition, R01b
400: and R05 also defined a distance by redshift as follows,
401: \be \dz=\frac{cz}{H_0},
402: \lb{dz}
403: \ee
404: where $c$ is the light speed and $H_0$ is the Hubble constant.
405: This is not a distance in the sense of the other distance measures:
406: since it is often used in observational cosmology, it is, however,
407: useful to adopt it here and assume Eq.\ (\ref{dz}) to be the definition
408: of redshift distance $\dz$ for all $z$.
409:
410: \subsection{Observational areas and volumes}
411:
412: We now define observational areas and volumes. Following R05,
413: the area of the observed spherical shell of radius $d_i$ may be
414: written as,
415: \be S_i=4 \pi {(d_i)}^2, \lb{area} \ee
416: and the observational volume at this same radius can be
417: straightforwardly written as,
418: \be V_i=\frac{4}{3} \pi \, {(d_i)}^3.
419: \lb{vol}
420: \ee
421: {These equations imply that the observed volume element
422: is given by,
423: \be dV_i=S_i \; d(d_i).
424: \lb{thick}
425: \ee
426: Here $d(d_i)$ is the elementary shell thickness and, since
427: cosmological distances are function of the redshift, the shell
428: thickness can be written approximately as $\Delta d_i = \left[
429: d(d_i)/dz \right] \Delta z$, for a certain observed redshift
430: interval $\Delta z$.}
431:
432: These expressions are completely general and Eq.\ (\ref{vol})
433: agrees with the usual volume definitions. We note that these areas
434: and volumes are observational, that is, they are defined along
435: the past light cone.
436:
437: %______________________________________________________________
438:
439: \subsection{Generalized number-distance relation}
440:
441: We aim to simulate number counts as if they were actual observations. To
442: do so in a general manner, it is
443: convenient to adopt the following expression for the \emph{number
444: of observed cosmological sources} $N$ at a certain observational
445: distance $d_i$,
446: \be \Nid = { \left( B \, d_i \right) }^D.
447: \lb{nd}
448: \ee
449: Here $B$ is an as yet unspecified constant and $D$ is the fractal
450: dimension (Pietronero 1987; Ribeiro and Miguelote 1998; Sylos
451: Labini et al.\ 1998). This expression provides a general
452: way of simulating galaxy counts, since, for $D=3$ we have an
453: \textit{observationally}-homogeneous distribution (see below),
454: whereas for $D<3$ we have an \textit{observationally-in}homogeneous
455: galaxy counting. The subscript index $i$ and superscript index $D$
456: therefore label respectively the choice of distance and how far the
457: simulated counting differs from OH. We note that it is \textit{not}
458: our intention to either prove or disprove the possible fractality of
459: the galaxy distribution. The adoption of the equation above for $N$
460: is just a matter of convenience due to its generality.
461:
462: de Vaucouleurs (1970) and Wertz (1970, 1971) long ago proposed similar
463: relations to Eq. (\ref{nd}), which were written instead in terms of
464: density. They appear to be the first authors to use it in the context
465: of galaxy distributions. Pietronero (1987) independently advanced an
466: expression virtually identical to the above. As discussed in Ribeiro
467: and Miguelote (1998; see also Ribeiro 1994), the models of Wertz
468: (1970, 1971) and Pietronero (1987) shared more similarities than
469: differences, both conceptually and analytically. Wertz did not use
470: the word `fractal', although self-similar ideas can be found in his
471: discussion. Pietronero (1987) named his expression the ``generalized
472: mass-length relation''; here we instead choose to refer to Eq.\
473: (\ref{nd}) as the \emph{generalized Pietronero-Wertz number-distance
474: relation}, or simply \emph{generalized number-distance relation}, since
475: we believe the emphasis on number-distance, rather than mass-length,
476: is more appropriate to observational cosmology.
477:
478: %______________________________________________________________
479:
480: \subsection{Observational densities}
481:
482: As discussed in R05 and A07, the \textit{differential density}
483: $\, \! \! \gammaid$ at a certain distance $d_i$, and with a specific
484: choice of the dimension $D$, is defined by the following expression,
485: \be \gammaid = \frac{1}{S_i} \frac{d \left( \Nid \right)}{d \left(
486: d_i \right) }.
487: \lb{diff}
488: \ee
489: This equation provides a measure of the rate of growth in the number
490: density as one moves down the past light cone along the observable
491: distance $d_i$. Obviously, the behavior of the differential density
492: depends heavily on the distance employed in Eq.\
493: (\ref{diff}). The \textit{integral differential density} $\gestid$
494: is defined to be the integration of $\gammaid$ over the observational
495: volume $V_i$, corresponding to,
496: \be \gestid= \frac{1}{V_i} \int_{V_i} \gammaid \: dV_i.
497: \lb{gest}
498: \ee
499: If we now define $\! \anid$ to be the \textit{radial number
500: density} for a given distance measure $d_i$, the following
501: result clearly holds, once we consider Eqs.\ (\ref{area}),
502: (\ref{vol}), (\ref{diff}), and (\ref{gest}),
503: \be \anid = \frac{\Nid}{V_i} = \gestid.
504: \lb{densi}
505: \ee
506:
507: {At this point some important remarks are necessary.
508: The radial number density $\! \anid$ considers the number
509: counts of objects as a function of distance from a single point.
510: In the context of FLRW cosmology, this single point can be any
511: point in a 4-dimensional Riemannian manifold, since this spacetime
512: assumes a maximal spatial isotropy (see \S \ref{rel} below).
513: Therefore, this is not an average quantity, or ensemble average,
514: obtained as an average made over many realizations of a stationary
515: stochastic process, where the ensemble average can be replaced by a
516: volume average. Confusion arises when Eq.\ (\ref{nd}) is viewed
517: in the framework of fractal geometry, where the number count of
518: this equation is interpreted as an average quantity. Here Eq.\
519: (\ref{nd}) is simply the radial matter distribution which
520: \textit{could} be given by a perfect fluid approximation of
521: cosmological solutions of Einstein's field equations. As
522: discussed in R05, the relationship between these two quantities,
523: the radial number density and the average number density, remains an
524: open question although it appears to be
525: reasonable to assume that they are related. The present paper
526: is not concerned with proving or disproving the fractal hypothesis
527: for the galaxy distribution, but aims to investigate the limitations
528: of the standard concept of ``homogeneity'', in particular
529: its application to interpreting real astronomical observations.}
530:
531: It is useful to write the differential densities
532: in terms of the redshift $z$. The results are as follows,
533: \be \gammaid (z) = \left[ \frac{d}{dz} \left( \Nid \right) \right]
534: {\left[ S_i \frac{d}{dz} \left( d_i \right)
535: \right]}^{-1},
536: \lb{gz}
537: \ee
538: \be \gestid (z) = \frac{1}{V_i} \int\limits_0^z \gammaid \left(
539: \frac{dV_i}{dz'} \right) dz' = \frac{\Nid(z)}{V_i(z)}.
540: \lb{gestz}
541: \ee
542: In both of the above equations one can clearly identify two distinct parts.
543: The first term is the geometrical term, determined by the spacetime
544: geometry of the chosen metric; this is the case of the functions
545: $d_i(z)$, $S_i(z)$ and $V_i(z)$ and their derivatives. The second
546: term is given by the number count $\! \Nid(z)$ and the differential
547: number count $d \left[ \Nid(z) \right] \big/ dz$ and are determined
548: either by theory or observationally. Various tests of cosmological
549: models rely, in one way or another, on the comparison of number
550: counting, determined observationally, with its theoretical prediction.
551: We note that, in Eq.\ (\ref{gz}), this division between the geometrical
552: and theoretical/observational parts are clearly visible, since they
553: are represented by each term inside the brackets on the right hand
554: side.
555:
556: As mentioned above, the geometrical part was determined entirely by
557: assuming special cases of the FLRW metric, EdS cosmology in R01b, R05,
558: and A07 and the FLRW open model in A07. The number count, however,
559: was either theoretically determined from the cosmological model, that
560: is, by taking the expression for $N(z)$ as given by the matter
561: distribution in these cosmologies, which meant assuming a
562: spatially-homogeneous matter distribution
563: from the start (R01b, R05), or by using differential counts
564: $dN/dz$ obtained from the luminosity function of galaxy surveys,
565: as demonstrated by A07.\footnote{It must be noted that when comparing
566: densities constructed from observationally-derived number
567: counts with the theoretical predictions of a FLRW model with
568: $\Omega_{m_0} =0.3$, $\Omega_{\Lambda_0}=0.7$, that is, having
569: SH by construction, A07 found deviations from pure SH. However,
570: it was not clear if these deviations were due to possible
571: incompleteness of the sample, the use of an inappropriate
572: evolution function when deriving the luminosity function
573: parameters, or if they were true deviations from SH. See details
574: in the caption Figs.\ 7 and 8 of A07.} We
575: determine the geometrical part by choosing a spacetime metric,
576: since this is unavoidable; instead of restricting ourselves however
577: to obtaining the number count solely from the chosen cosmology,
578: we use Eq.\ (\ref{nd}) to simulate number counts, \textit{as if} they
579: were real observations. This methodology attempts to make possible
580: the investigation of the model behavior, when OH is assumed from the
581: outset; it configures an approach that is opposite to that investigated
582: in R01b, R05, and A07. The advantage of this methodology is to depart
583: from pure spatially-homogeneous cosmologies.
584:
585: %______________________________________________________________
586:
587: \section{Observational \underline{in}homogeneity of the Einstein-de
588: Sitter spacetime}\lb{eds}
589:
590: We shall assume this spacetime metric for analytical simplicity
591: only, but this choice does not affect our main results (see below).
592: The methodology that we present can be extended easily to other
593: FLRW spacetime metrics. The EdS expressions shown next were derived
594: in R05 (see also R01b).
595:
596: In the EdS model, the four observational distances discussed above are
597: given by the following set of equations,\footnote{In EdS cosmology,
598: the comoving distance is equal to the galaxy area distance $\dg$
599: multiplied by a constant factor (see R05).}
600: \be \da(z)=\frac{2c}{H_0} \left[ \frac{1+z- \sqrt{1+z}}{{ \left(
601: 1+z \right) }^2} \right],
602: \lb{da}
603: \ee
604: \be \dg(z) =\frac{2c}{H_0} \left( \frac{1+z- \sqrt{1+z}}{1+z}
605: \right),
606: \lb{dg}
607: \ee
608: \be \dl(z) =\frac{2c}{H_0} \left( 1+z- \sqrt{1+z} \right),
609: \lb{dl}
610: \ee
611: \be \dz(z)= \frac{2c}{H_0} \left( \frac{z}{2} \right).
612: \lb{dz1}
613: \ee
614: {Equations (\ref{da}), (\ref{dg}), and (\ref{dl}) have
615: Taylor-series expansions in terms of redshift that are given by,
616: \be \da(z)=\frac{c}{H_0} \left(z-\frac{7}{4}z^2+\frac{19}{8}
617: z^3+ \ldots \right),
618: \lb{daz}
619: \ee
620: \be \dg(z)=\frac{c}{H_0} \left(z-\frac{3}{4}z^2+\frac{5}{8}
621: z^3+ \ldots \right),
622: \lb{dgz}
623: \ee
624: \be \dl(z)=\frac{c}{H_0} \left(z+\frac{1}{4}z^2-\frac{1}{8}
625: z^3+ \ldots \right),
626: \lb{dlz}
627: \ee
628: which shows that all four distances above reduce to the same expression
629: to first order.}
630:
631: The number count in this cosmology is well known, producing the
632: following expression,
633: \be N^{\mathrm{\ssty EdS}}= \alpha\;
634: { \left( \frac{1+z-\sqrt{1+z}}{1+z} \right) }^3,
635: \lb{neds}
636: \ee
637: where the dimensionless constant $\alpha$ is defined to be,
638: \be \alpha=\frac{4c^3}{H_0 M_g G}.
639: \lb{alpha}
640: \ee
641: Here $M_g$ is the average galactic rest mass $( \sim 10^{11}
642: M_\odot )$ and $G$ is the gravitational constant. The differential
643: densities in this cosmology are given by the following equations,
644: \be \gamma_{\ssty A}^{{\mathrm{\ssty EdS}}} = \mu_0
645: \left[ { \frac{{(1+z)}^3}{(3-2 \sqrt{1+z} ) } } \right] ,
646: \lb{ga}
647: \ee
648: \be \gamma_{\ssty G}^{{\mathrm{\ssty EdS}}} = \mu_0,
649: \lb{gg}
650: \ee
651: \be \gamma_{\ssty L}^{{\mathrm{\ssty EdS}}} = \mu_0
652: \left[ \frac{1}{ (2 \sqrt{1+z} - 1) { (1+z)}^3} \right],
653: \lb{gl}
654: \ee
655: \be \gamma_{\ssty Z}^{{\mathrm{\ssty EdS}}} = \mu_0 \left[ \frac{4 \,
656: {(1+z- \sqrt{1+z}) }^2}{z^{\, 2} \, { \left( 1+z \right)
657: }^{7/2}} \right],
658: \lb{gz1}
659: \ee
660: and the integral differential densities yield,
661: \be \gamma_{\ssty A}^{\ast^{\mathrm{EdS}}} = \mu_0
662: { \left(1+z \right) }^{3},
663: \lb{gamasa}
664: \ee
665: \be \gamma_{\ssty G}^{\ast^{\mathrm{EdS}}} = \mu_0,
666: \lb{gamasg}
667: \ee
668: \be \gamma_{\ssty L}^{\ast^{\mathrm{EdS}}} = \mu_0
669: { \left( 1+z \right) }^{-3},
670: \lb{gamasl}
671: \ee
672: \be \gamma_{\ssty Z}^{\ast^{\mathrm{EdS}}} = \mu_0 { \left[
673: \frac{ 2(1+z-\sqrt{1+z}) }{z { \left( 1+z \right)
674: }} \right] }^3,
675: \lb{gamasz}
676: \ee
677: where the constant $\mu_0$ is given as below,
678: \be \mu_0= \frac{3 {H_0}^2}{8\pi M_g G}.
679: \lb{mu0}
680: \ee
681:
682: {Taylor series expansions for the differential densities
683: above are given as follows,
684: \be \gamma_{\ssty A}^{{\mathrm{\ssty EdS}}} = \mu_0
685: \left(1+4z+\frac{27}{4}z^2+\frac{55}{8}z^3+\ldots \right)
686: \lb{gasz}
687: \ee
688: \be \gamma_{\ssty L}^{{\mathrm{\ssty EdS}}} = \mu_0
689: \left(1-4z+\frac{41}{4}z^2-\frac{171}{8}z^3+\ldots \right)
690: \lb{glsz}
691: \ee
692: \be \gamma_{\ssty Z}^{{\mathrm{\ssty EdS}}} = \mu_0
693: \left(1-3z+\frac{95}{16}z^2-\frac{39}{4}z^3+\ldots \right)
694: \lb{gz1sz}
695: \ee
696: \be \gamma_{\ssty A}^{\ast^{\mathrm{EdS}}} = \mu_0
697: \left(1+3z+3z^2+z^3 \right)
698: \lb{gamasasz}
699: \ee
700: \be \gamma_{\ssty L}^{\ast^{\mathrm{EdS}}} = \mu_0
701: \left(1-3z+6z^2-10z^3+\ldots \right)
702: \lb{gamaslsz}
703: \ee
704: \be \gamma_{\ssty Z}^{\ast^{\mathrm{EdS}}} = \mu_0
705: \left(1-\frac{9}{4}z+\frac{57}{16}z^2-\frac{39}{8}z^3+\ldots \right)
706: \lb{gamaszsz}
707: \ee
708: We note that all densities above have a non-vanishing zeroth order
709: term, whereas the series expansions for the cosmological distances,
710: given by Eqs.\ (\ref{daz}), (\ref{dgz}), and (\ref{dlz}), do not contain
711: this term. This implies that, according to Eqs.\
712: (\ref{gasz})--(\ref{gamaszsz}), deviations from a constant density
713: value, that is, from OH (see below), occur in the first-order
714: terms of the series, whereas Eqs.\ (\ref{daz})--(\ref{dlz})
715: imply that deviations from the Hubble law occur in the second-order
716: terms. In other words, Hubble-law deviations occur in higher
717: redshift ranges than deviations from OH. This was first noticed
718: in Ribeiro (1995), discussed again in R01b and explored further in
719: Abdalla et al.\ (2001) by means of a simple perturbed model.}
720:
721: As discussed at length in R01b, OH corresponds to a constant value
722: of observational average density, that is, when this average
723: density is calculated along the chosen spacetime's past null cone.
724: This requirement was operationally defined in R05 and A07 to
725: mean the following condition,
726: \be \gamma_i^\ast = \mathrm{constant}, \; \; \; \; \; \; \; \;
727: \mbox{\rm (OH definition)}.
728: \lb{oh}
729: \ee
730: It is clear from a simple inspection of Eqs.\ (\ref{gamasa}),
731: (\ref{gamasg}), (\ref{gamasl}), and (\ref{gamasz}) that although EdS
732: cosmology is SH by construction for \textit{all} distance measures,
733: according to the definition provided by Eq.\ (\ref{oh}) the conditions
734: of OH are met only when one builds an average density using the galaxy
735: area distance $\dg$ (see Eq.\ \ref{gamasg}). The three other distance
736: definitions, $\da$, $\dl$, and $\dz$ are apparently unsuitable
737: as tools for searching the possible OH in the galaxy redshift data,
738: if one adopts the EdS cosmology. This was the main conclusion reached
739: by R05, which was also found to be valid for open FLRW cosmology
740: in A07. This conclusion reproduced here to compare with
741: the analysis below is graphically summarized in Figs.\ \ref{fig1}
742: and \ref{fig12}.
743: \begin{figure}
744: \centering
745: \vspace{-5.cm}
746: \hspace{-2.cm} \includegraphics[width=12.5cm]{f1.ps}
747: %\input{f1.tex}
748: \vspace{-8.5cm}
749: \caption{Plot similar to the one appearing in R05 showing the
750: normalized radial number densities ${\left[ n \right]}_i$,
751: or integral differential densities $\gamma_i^\ast$ (see
752: eq.\ \ref{densi}), along the past light cone versus the
753: redshift $z$ in the EdS cosmology. Note that although all
754: cases are spatially-homogeneous by construction, only
755: ${\left[ n \right]}_{ \ssty G}^{\mathrm{\ssty EdS}}$ is
756: observationally homogeneous as well. The other three
757: radial number densities are spatially homogeneous, but
758: observationally {\it in}homogeneous. The asymptotic
759: limits for these densities are also different (see R01b),
760: yielding, $\lim\limits_{z \rightarrow \infty}
761: {\left[ n \right]}_{\ssty L}^{\mathrm{\ssty EdS}} = 0$,
762: $\lim\limits_{z \rightarrow \infty}
763: {\left[ n \right]}_{\ssty A}^{\mathrm{\ssty EdS}} =\infty$,
764: $\lim\limits_{z \rightarrow \infty}
765: {\left[ n \right]}_{\ssty G}^{\mathrm{\ssty EdS}} = \mu_0$,
766: $\lim\limits_{z \rightarrow \infty}
767: {\left[ n \right]}_{z}^{\mathrm{\ssty EdS}} = 0$.
768: }
769: \lb{fig1}
770: \end{figure}
771: \begin{figure}
772: \centering
773: \vspace{-5.cm}
774: \hspace{-2.cm} \includegraphics[width=12.5cm]{f2.ps}
775: %\input{f2.tex}
776: \vspace{-8.5cm}
777: \caption{This figure shows the normalized radial number
778: densities ${\left[ n \right]}_i$ of the previous graph,
779: but at a redshift range of up to $z=0.2$. As noted in Ribeiro
780: (1995), the distinction between the densities built with
781: different distance measures can be seen before the
782: redshift reaches $z=0.1$.}
783: \lb{fig12}
784: \end{figure}
785:
786: \section{Models with or without observational homogeneity}\lb{sec4}
787:
788: We now adopt the generalized number-distance relation provided by Eq.\
789: (\ref{nd}) to obtain more general models, whose main properties have or
790: do not have OH. As discussed above, we shall use Eq.\ (\ref{nd}) to
791: simulate possible matter distributions and test whether or not an assumed
792: OH number count distribution produces SH, or if a distribution with no
793: OH produces, or not, models with no SH.
794:
795: Considering the definitions given by Eqs.\ (\ref{area}) and
796: (\ref{vol}), it is straightforward to conclude that the
797: generalized number-distance relation provided by Eq.\ (\ref{nd})
798: corresponds to the following expressions for the differential density
799: given by Eq.\ (\ref{diff}) and the integral differential density
800: given by Eq.\ (\ref{gest}),
801: \be \gammaid \, (d_i) = \frac{DB^D}{4\pi} d_i^{D-3},
802: \lb{diffPW}
803: \ee
804: \be \gestid (d_i)= \frac{3B^D}{4\pi} d_i^{D-3}.
805: \lb{gestPW}
806: \ee
807: From these results, it becomes obvious that these
808: two densities are related by the following expression,
809: \be \gestid = \frac{3}{D}\, \gammaid \, .
810: \lb{doisgs}
811: \ee
812:
813: To proceed with our analysis, it is unavoidable at this stage to
814: choose a cosmological model (see Sect.\ \ref{sec2} above). Our
815: choice then is to continue using an EdS cosmology, due to its
816: simplicity, for everything besides number counts. However, our
817: results can be extended to other FLRW spacetimes, or even
818: to non FLRW cosmologies.
819:
820: We substitute the four EdS distance definitions, given by
821: Eqs.\ (\ref{da}), (\ref{dg}), (\ref{dl}), and (\ref{dz1}) into
822: Eq.\ (\ref{nd}). The results may be written as follows,
823: \be {\; \nid{A}}(z)={\left(\frac{2cB}{H_0}\right)}^D \;
824: {\left[\frac{1+z-\sqrt{1+z}}{{(1+z)}^2} \, \right]}^D,
825: \lb{Nad}
826: \ee
827: \be {\; \nid{G}}(z)={\left(\frac{2cB}{H_0}\right)}^D \;
828: {\left(\frac{1+z-\sqrt{1+z}}{1+z} \, \right)}^D,
829: \lb{Ngd}
830: \ee
831: \be {\; \nid{L}}(z)={\left(\frac{2cB}{H_0}\right)}^D \;
832: {\left({1+z-\sqrt{1+z}} \, \right)}^D,
833: \lb{Nld}
834: \ee
835: \be {\; \nid{Z}}(z)={\left(\frac{2cB}{H_0}\right)}^D \;
836: {\left(\frac{z}{2}\right)}^D.
837: \lb{Nzd}
838: \ee
839: If we now substitute these EdS distance measures into Eq.\
840: (\ref{diffPW}), we may write the expressions for the differential
841: densities as shown below,
842: \be \gamid{A}{D}(z) = \left(\frac{D{H_0}^3}{32\pi c^3}\right)
843: {\left(\frac{2cB}{H_0}\right)}^D \;
844: {\left[\frac{1+z-\sqrt{1+z}}{{(1+z)}^2} \, \right]}^{D-3},
845: \lb{gamidAD}
846: \ee
847: \be \gamid{G}{D}(z) = \left(\frac{D{H_0}^3}{32\pi c^3}\right)
848: {\left(\frac{2cB}{H_0}\right)}^D \;
849: {\left(\frac{1+z-\sqrt{1+z}}{1+z} \, \right)}^{D-3},
850: \lb{gamidGD}
851: \ee
852: \be \gamid{L}{D}(z) = \left(\frac{D{H_0}^3}{32\pi c^3}\right)
853: {\left(\frac{2cB}{H_0}\right)}^D \;
854: {\left({1+z-\sqrt{1+z}} \, \right)}^{D-3},
855: \lb{gamidLD}
856: \ee
857: \be \gamid{Z}{D}(z) = \left(\frac{D{H_0}^3}{32\pi c^3}\right)
858: {\left(\frac{2cB}{H_0}\right)}^D \;
859: {\left(\frac{z}{2}\right)}^{D-3}.
860: \lb{gamidZD}
861: \ee
862: For the integral differential density given by Eq.\
863: (\ref{gestPW}), the expression in Eq.\ (\ref{doisgs}) allows
864: us to write the results as below,
865: \be \gamsid{A}{D}(z) = \left(\frac{3{H_0}^3}{32\pi c^3}\right)
866: {\left(\frac{2cB}{H_0}\right)}^D \;
867: {\left[\frac{1+z-\sqrt{1+z}}{{(1+z)}^2} \, \right]}^{D-3},
868: \lb{gamsidAD}
869: \ee
870: \be \gamsid{G}{D}(z) = \left(\frac{3{H_0}^3}{32\pi c^3}\right)
871: {\left(\frac{2cB}{H_0}\right)}^D \;
872: {\left(\frac{1+z-\sqrt{1+z}}{1+z} \, \right)}^{D-3},
873: \lb{gamsidGD}
874: \ee
875: \be \gamsid{L}{D}(z) = \left(\frac{3{H_0}^3}{32\pi c^3}\right)
876: {\left(\frac{2cB}{H_0}\right)}^D \;
877: {\left({1+z-\sqrt{1+z}} \, \right)}^{D-3},
878: \lb{gamsidLD}
879: \ee
880: \be \gamsid{Z}{D}(z) = \left(\frac{3{H_0}^3}{32\pi c^3}\right)
881: {\left(\frac{2cB}{H_0}\right)}^D \;
882: {\left(\frac{z}{2}\right)}^{D-3}.
883: \lb{gamsidZD}
884: \ee
885:
886: When $D=3$, the definition (\ref{oh}) is fulfilled in view of
887: Eqs.\ (\ref{diffPW}) and (\ref{gestPW}) above and, therefore,
888: this choice of dimension clearly corresponds to OH. From this
889: follows an interesting result. We saw in Sect.\ \ref{eds} that
890: densities constructed using the galaxy area distance $\dg$ in an
891: EdS cosmology produces a model having both SH and OH. This
892: property allows us to find the constant $B$. For $D=3$, we can
893: equate Eq.\ (\ref{gamasg}) to Eq.\ (\ref{gamsidGD}) and,
894: considering Eq.\ (\ref{mu0}), we obtain the following result,
895: \be B={\left(\frac{{H_0}^2}{2M_gG}\right)}^{1/3}.
896: \lb{B}
897: \ee
898: This is valid as long as the geometrical part of the model is
899: given by EdS spacetime. Thus, each adopted metric corresponds to
900: a different value of the constant $B$, whose dimension is
901: ${[\,\mbox{length unit}\,]}^{-1}$.
902:
903: \subsection{Case of $D=3$}\lb{d3}
904:
905: As seen above, this is the condition for the existence of OH and
906: holds for \textit{all} distance definitions (see Eq.\ \ref{gestPW}).
907: This case reduces Eq.\ (\ref{doisgs}) to the following
908: simple expression,
909: \be \gamsid{\sty i}{3} = \gamid{\sty i}{3} =
910: \frac{3 {H_0}^2}{8\pi M_g G}.
911: \lb{special}
912: \ee
913: Expressions for the number counts $\NID{A}{3}(z)$,
914: $\NID{G}{3}(z)$, $\NID{L}{3}(z)$, and $\NID{Z}{3}(z)$ are
915: respectively obtained from Eqs.\ (\ref{Nad}),
916: (\ref{Ngd}), (\ref{Nld}), and (\ref{Nzd}). Recalling Eq.\
917: (\ref{neds}), we verify that $\NID{G}{3}(z)=N^{\mathrm{\ssty
918: EdS}}(z)$. These functions are plotted in Figs.\ \ref{fig2}
919: and \ref{fig34}, where the caption of the former Figure discusses
920: that only $\NID{G}{3}$ shows both SH and OH, although all
921: number-counting functions are observationally-homogeneous by
922: construction.
923: %\newpage
924: \begin{figure}
925: \centering
926: \vspace{-5.cm}
927: \hspace{-2.cm} \includegraphics[width=12.5cm]{f3.ps}
928: %\vspace{0.6cm}
929: %\input{f3.tex}
930: \vspace{-8.5cm}
931: \caption{Number counts for the case $D=3$ in Eqs.\ (\ref{Nad}),
932: (\ref{Ngd}), (\ref{Nld}), and (\ref{Nzd}). By definition these
933: counts are \textit{all} observationally-homogeneous in
934: \textit{all} values of $z$, as can be seen from Eq.\
935: (\ref{gestPW}). The four expressions start similarly, that
936: is, close to the spatially-homogeneous case $\NID{G}{3}=
937: N^\mathrm{\ssty EdS}$, but as $z$ increases deviations start
938: to occur. At $z=0.5$ these deviations are significant.
939: The only expression that exhibits both OH and SH is the one
940: obtained from the galaxy area distance $\dg$. Since $\NID{G}
941: {3}(z)$ and $\NID{L}{3}(z)$ have higher counts than
942: $N^\mathrm{\ssty EdS}$ they should be spatially
943: \textit{in}homogeneous. The same is also true for $\NID{A}{3}
944: (z)$, although in this case the counts are less than the EdS
945: one with SH. One can also easily verify that in the asymptotic
946: limit of the big bang singularity hypersurface the following
947: results hold,
948: $\lim\limits_{z \rightarrow \infty} \NID{A}{3}=0$,
949: $\lim\limits_{z \rightarrow \infty} \NID{G}{3}=\alpha$,
950: $\lim\limits_{z \rightarrow \infty} \NID{L}{3}=\infty$,
951: $\lim\limits_{z \rightarrow \infty} \NID{Z}{3}=\infty$.
952: }
953: \lb{fig2}
954: \end{figure}
955: \begin{figure}
956: \centering
957: \vspace{-5.cm}
958: \hspace{-2.cm} \includegraphics[width=12.5cm]{f4.ps}
959: %\vspace{0.9cm}
960: %\input{f4.tex}
961: \vspace{-8.5cm}
962: \caption{Graph also showing the number counts for the case $D=3$ in
963: Eqs.\ (\ref{Nad}), (\ref{Ngd}), (\ref{Nld}), and (\ref{Nzd}),
964: but at small redshifts compared to the previous figure.
965: Clearly only at $z \approx 0.1$, the differences between the
966: number counts constructed with the various distance definitions
967: can be seen.}
968: \lb{fig34}
969: \end{figure}
970:
971: \subsection{Case of $D=2$}\lb{d2}
972:
973: We have observational \textit{in}homogeneity by construction for
974: any $D<3$ since Sylos Labini et al.\ (1998) reported a value
975: close to $D=2$ for the fractal dimension of the distribution of
976: galaxies, this however provides a suitable choice for our toy
977: model. {By definition, this choice implies no SH and, therefore,
978: it might be argued that the Friedmann models, as well as their
979: distance definitions, are not valid for any $D<3$. We emphasize
980: that our aim is not to validate completely the standard cosmology,
981: but to determine an unambiguous answer to the following question.
982: \textit{If} the number counting produced by assuming $D=2$ in
983: Eq.\ (\ref{nd}) \textit{corresponded to true observations}, can
984: we conclude, by applying a FLRW framework, that the galaxy distribution
985: does not follow the standard cosmology? In other words, using
986: galaxy counts simulated by Eq.\ (\ref{nd}) with $D=2$, is it
987: possible by employing procedures based on the standard
988: cosmological model to conclude with certainty that this number count
989: could not possibly be an observationally-homogeneous galaxy
990: distribution? This is a very important question as this approach was
991: taken by many studies found in the literature, that is, the framework
992: given by the standard cosmology was used to determine if the observed
993: data was consistent with this model. Although most who carry out these
994: studies implicitly assume that this is possible, as we show below, our
995: results indicate that ambiguities in interpreting observations within
996: the standard model framework still remain. As noted by Joyce et al.\
997: (2000), \textit{straightforward interpretations of FLRW standard cosmologies
998: are problematic; they include the interpretation that the isotropy of
999: microwave background radiation implies that observationally-inhomogeneous
1000: matter distributions are impossible.}
1001:
1002: We proceed and assume that $D=2$ in Eqs.\ (\ref{Nad}) to
1003: (\ref{Nzd}). Considering the definitions provided in Eqs.\ (\ref{B})
1004: and (\ref{alpha}), we obtain the following results,
1005: \be {\; \NID{A}{2}}(z)={\alpha}^{2/3} \;
1006: {\left[\frac{1+z-\sqrt{1+z}}{{(1+z)}^2} \, \right]}^2,
1007: \lb{Nad2}
1008: \ee
1009: \be {\; \NID{G}{2}}(z)={\alpha}^{2/3} \;
1010: {\left(\frac{1+z-\sqrt{1+z}}{1+z} \, \right)}^2,
1011: \lb{Ngd2}
1012: \ee
1013: \be {\; \NID{L}{2}}(z)={\alpha}^{2/3} \;
1014: {\left({1+z-\sqrt{1+z}} \, \right)}^2,
1015: \lb{Nld2}
1016: \ee
1017: \be {\; \NID{Z}{2}}(z)={\alpha}^{2/3} \;
1018: {\left(\frac{z}{2}\right)}^2.
1019: \lb{Nzd2}
1020: \ee
1021: As can be seen in Fig.\ \ref{fig3}, the curves are similar
1022: to those in Fig.\ \ref{fig2}, apart from the scale. Although
1023: $\NID{G}{3}(z) \not = \NID{G}{2}(z)$, both functions converge to
1024: a constant value in their asymptotic limits. As we show below,
1025: this result has an interesting consequence.
1026: \begin{figure}
1027: \centering
1028: \vspace{-5.cm}
1029: \hspace{-2.cm} \includegraphics[width=12.5cm]{f5.ps}
1030: %\input{f5.tex}
1031: \vspace{-8.5cm}
1032: \caption{Number counts for the case of $D=2$. The functions show a
1033: behavior similar to those shown in figure \ref{fig2}. We note
1034: that $\NID{G}{2}$ tends to a constant value at its asymptotic
1035: limit. Indeed, the limits of the four functions are as follows,
1036: $\lim\limits_{z \rightarrow \infty} \NID{A}{2}=0$,
1037: $\lim\limits_{z \rightarrow \infty} \NID{G}{2}=\alpha^{2/3}$,
1038: $\lim\limits_{z \rightarrow \infty} \NID{L}{2}=\infty$, and
1039: $\lim\limits_{z \rightarrow \infty} \NID{Z}{2}=\infty$.
1040: }
1041: \lb{fig3}
1042: \end{figure}
1043: \begin{figure}
1044: \centering
1045: \vspace{-5.cm}
1046: \hspace{-2.cm} \includegraphics[width=12.5cm]{f6.ps}
1047: %\input{f6.tex}
1048: \vspace{-8.5cm}
1049: \caption{Number counts for the case where $D=2$ at very small
1050: redshifts. In a similar way to the graph shown in Fig.\
1051: \ref{fig34}, only at $z \approx 0.1$ the various
1052: distance measures begin to affect the number
1053: counts. This range is, nevertheless, dependent on the
1054: cosmological model, which means that if another cosmology
1055: is adopted, deviations caused by the use of different
1056: distance measures could possibly occur at redshifts
1057: smaller than $z=0.1$.}
1058:
1059: \lb{fig56}
1060: \end{figure}
1061:
1062: The integral differential densities are calculated using Eqs.\
1063: (\ref{gamsidAD}) to (\ref{B}), to be,
1064: \be \gamsid{A}{2}(z) = \mu_1 \;
1065: {\left[\frac{1+z-\sqrt{1+z}}{{(1+z)}^2} \, \right]}^{-1},
1066: \lb{gamsidAD2}
1067: \ee
1068: \be \gamsid{G}{2}(z) = \mu_1 \;
1069: {\left(\frac{1+z-\sqrt{1+z}}{1+z} \, \right)}^{-1},
1070: \lb{gamsidGD2}
1071: \ee
1072: \be \gamsid{L}{2}(z) = \mu_1 \;
1073: {\left({1+z-\sqrt{1+z}} \, \right)}^{-1},
1074: \lb{gamsidLD2}
1075: \ee
1076: \be \gamsid{Z}{2}(z) = \mu_1 \;
1077: {\left(\frac{z}{2}\right)}^{-1}.
1078: \lb{gamsidZD2}
1079: \ee
1080: where,
1081: \be \mu_1=\frac{3{H_0}^{7/3}}{8\pi c {(2M_g G)}^{2/3}}.
1082: \lb{mu1}
1083: \ee
1084: {We obtain the following power-series expansions of these
1085: expressions,
1086: \be \gamsid{A}{2}(z) = \mu_1 \left(\frac{2}{z}+\frac{7}{2}+
1087: \frac{11}{8}z-\frac{1}{16}z^2 + \ldots \right),
1088: \lb{gamsidAD2s}
1089: \ee
1090: \be \gamsid{G}{2}(z) = \mu_1 \left( \frac{2}{z}+\frac{3}{2}-
1091: \frac{1}{8}z+\frac{1}{16}z^2- \ldots \right),
1092: \lb{gamsidGD2s}
1093: \ee
1094: \be \gamsid{L}{2}(z) = \mu_1 \left(\frac{2}{z}-\frac{1}{2}+
1095: \frac{3}{8}z-\frac{5}{16}z^2+ \ldots \right).
1096: \lb{gamsidLD2s}
1097: \ee
1098: We note that for small redshifts the first terms of the
1099: series above dominate. Considering the approximation provided
1100: by Eq.\ (\ref{gamsidZD2}), we therefore have that
1101: $\gamsid{A}{2}$, $\gamsid{G}{2}$ and $\gamsid{L}{2}$
1102: become equal to $\gamsid{Z}{2}$.}
1103:
1104: The functions above are plotted versus the redshift in Figs.\
1105: \ref{fig4} and \ref{fig78}. We can see clearly in Fig.\ \ref{fig4}
1106: that, although we started with an observationally-\textit{in}homogeneous
1107: model, if we use the galaxy area distance $\dg$, the density, as a
1108: function of redshift, is constant for $z>10$. In other words,
1109: \textit{even a model that is observationally-inhomogeneous by
1110: construction appears to become observationally-homogeneous at higher
1111: $z$, if the density is built using the appropriate distance measure,
1112: in this case $\dg$}. Both $\dl$ and $\dz$ produce observational
1113: inhomogeneity for any $z$, which is reproduced as a power law decay.
1114: The density constructed with the area distance $\da$ has an odd
1115: behavior while starting to decay as a power law at small redshifts,
1116: but at $z \approx 1$ this decay turns into an increase.
1117:
1118: It is clear from these results that developing an
1119: observationally-inhomogeneous density that decreases as $z$ increases
1120: is no guarantee that it will remain so for all $z$. Clearly the use of
1121: the various distance measures creates too many ambiguities, which prevent
1122: definitive conclusions being made about the behavior of the large-scale
1123: galaxy distribution in the Universe.\footnote{We note that a further
1124: source of ambiguity in defining OH is that here it is defined
1125: in terms of a density that remains unchanged when the generic
1126: \textit{distance} $d_i$ changes, whereas we might possibly conceive a
1127: definition of OH in terms of densities that remain unchanged for
1128: different values of the \textit{redshift}.}
1129:
1130: \begin{figure}
1131: \centering
1132: \vspace{-5.cm}
1133: \hspace{-2.cm} \includegraphics[width=12.5cm]{f7.ps}
1134: %\input{f7.tex}
1135: \vspace{-8.5cm}
1136: \caption{Graph of the integral differential densities versus the
1137: redshift for the case of $D=2$. It is easy to show that
1138: the following asymptotic limits hold,
1139: $\lim\limits_{z \rightarrow \infty} \gamsid{A}{2}=\infty$,
1140: $\lim\limits_{z \rightarrow \infty} \gamsid{G}{2}=\mu_1$,
1141: $\lim\limits_{z \rightarrow \infty} \gamsid{L}{2}=0$, and
1142: $\lim\limits_{z \rightarrow \infty} \gamsid{Z}{2}=0$.
1143: All densities have a power-law decay at small redshifts.
1144: However, at $z \approx 1$ the densities constructed with the
1145: redshift and luminosity distances continue to decay, whereas
1146: $\gamsid{A}{2}$ begins to change from a decay to an increasing
1147: behavior. More interestingly, the density constructed with the
1148: galaxy area distance $\gamsid{G}{2}$ begins to change from a
1149: power-law decay to a constant value at $z \approx 2.5$. This
1150: is a consequence of the fact that the number counting
1151: constructed with $\dg$ becomes constant at higher values of
1152: $z$ (see figure \ref{fig3}). Since $\NID{G}{3}$ tends to a
1153: constant value as the redshift increases, and since
1154: $\NID{G}{3}=N^\mathrm{\ssty EdS}$ (see figure \ref{fig2}),
1155: this suggests that $\gamsid{G}{2} \;$ becomes spatially
1156: \textit{homogeneous} for $z > 10$. This occurs despite the
1157: fact that this density is observationally \textit{in}homogeneous
1158: by construction. This result therefore suggests that for
1159: $D=2$ the integral differential density constructed with
1160: the galaxy area distance $\dg$ is \textit{not} observationally
1161: and spatially homogeneous for $z<10$, but it seems to turn
1162: into both for $z>10$. This is a simple and clear example
1163: showing that the use of different distance measures in the
1164: characterization of cosmological densities may lead to
1165: significant ambiguities in reaching conclusions about the
1166: behavior of the large-scale galaxy distribution.
1167: }
1168: \lb{fig4}
1169: \end{figure}
1170: \begin{figure}
1171: \centering
1172: \vspace{-5.cm}
1173: \hspace{-2.cm} \includegraphics[width=12.5cm]{f8.ps}
1174: %\input{f8.tex}
1175: \vspace{-8.5cm}
1176: \caption{Graph of the integral differential densities versus the
1177: redshift for the case of $D=2$ at very small redshifts.
1178: Only when $z \approx 0.1$, this density starts to be affected
1179: by the different cosmological distance definitions.}
1180: \lb{fig78}
1181: \end{figure}
1182:
1183: \section{Behavior of the dimension D in a spatially
1184: homogeneous case}\lb{DSH}
1185:
1186: We have so far assumed constant values for the dimension $D$. However,
1187: it is interesting to study the behavior of this dimension as one
1188: approaches the Big Bang singularity hypersurface. Inasmuch as it is
1189: well known that at those very early times curvature effects become
1190: negligible, studying the behavior of $D$ as $z \rightarrow \infty$ in
1191: the EdS model should shed some light on its general behavior. To do
1192: so, we proceed as follows.
1193:
1194: In the EdS model, the generalized number-distance relation given by Eq.\
1195: (\ref{nd}) may be written as below if we consider the number counting
1196: given by Eq.\ (\ref{neds}),
1197: \be { \left( B \, d_i \right) }^D = \alpha \;
1198: { \left( \frac{1+z-\sqrt{1+z}}{1+z} \right) }^3.
1199: \lb{dsh}
1200: \ee
1201: These are in fact four equations, one for each distance definition
1202: indicated by the index $i={\sty A}, {\sty G}, {\sty L}, {\sty Z}$.
1203: We define the EdS distance measures as follows,
1204: \be d_i(z)=\frac{2c}{H_0}f_i(z),
1205: \lb{fi}
1206: \ee
1207: where $f_i(z)$ is a function given by each distance definition provided
1208: in Eqs.\ (\ref{da}), (\ref{dg}), (\ref{dl}), and (\ref{dz1}).
1209: Considering the definitions provided by Eqs.\ (\ref{alpha}) and (\ref{B}),
1210: the four equations given by the expression (\ref{dsh}) may be rewritten
1211: in terms of the dimension $D$, as can be seen below,
1212: \be D_i=3\left( \frac{\ln \alpha^{1/3} + \ln f_{\ssty G}}{\ln
1213: \alpha^{1/3} + \ln f_i} \right).
1214: \lb{Di}
1215: \ee
1216: The following result comes directly from this Eq.\ (\ref{Di}),
1217: \be D_{\ssty G}=3.
1218: \lb{Dg}
1219: \ee
1220: This value for the fractal dimension should not come as a surprise
1221: since the density defined with the galaxy area distance $\dg$
1222: remains constant, that is, observationally homogeneous for all $z$
1223: in the EdS cosmology (see figure \ref{fig1}). The other three
1224: expressions for the dimension $D$ in each distance measure yield,
1225: \be D_{\ssty A}=3\left\{ \frac{\ln \alpha^{1/3} +
1226: \ln \left( \displaystyle \frac{1+z-\sqrt{1+z}}{1+z} \right)}{\ln
1227: \alpha^{1/3} + \ln \left[ \displaystyle
1228: \frac{1+z-\sqrt{1+z}}{(1+z\,)^{\,2}} \right]} \right\},
1229: \lb{DA}
1230: \ee
1231: \be D_{\ssty L}=3\left[ \frac{\ln \alpha^{1/3} +
1232: \ln \left( \displaystyle \frac{1+z-\sqrt{1+z}}{1+z} \right)}{\ln
1233: \alpha^{1/3} + \ln \left( 1+z-\sqrt{1+z} \right) } \right],
1234: \lb{DL}
1235: \ee
1236: \be D_{\ssty Z}=3\left[ \frac{\ln \alpha^{1/3} +
1237: \ln \left( \displaystyle \frac{1+z-\sqrt{1+z}}{1+z} \right)}{\ln
1238: \alpha^{1/3} + \ln \displaystyle \left( \frac{z}{2} \right)} \right].
1239: \lb{DZ}
1240: \ee
1241: {Series expansions for the expressions above may be written
1242: as follows,
1243: \be D_{\ssty A}=\left[ 3+\left(\frac{3z}{\ln \alpha^{1/3} - \ln 2 +
1244: \ln z}\right)+ \ldots \right],
1245: \lb{DAs}
1246: \ee
1247: \be D_{\ssty L}=\left[ 3-\left( \frac{3z}{\ln \alpha^{1/3} - \ln 2 +
1248: \ln z} \right) + \ldots \right],
1249: \lb{DLs}
1250: \ee
1251: \be D_{\ssty Z}=\left[ 3- \left(\frac{3}{4}\right) \left( \frac{3z}
1252: { \ln \alpha^{1/3} - \ln 2 + \ln z } \right) + \ldots \right].
1253: \lb{DZs}
1254: \ee
1255: Clearly these results are valid for small, but nonzero, values of the
1256: redshift. However,} we show that functions given by the Eqs.\ (\ref{DA}),
1257: (\ref{DL}), and (\ref{DZ}) converge as the redshift vanishes. In fact,
1258: we have the following results,
1259: \be \lim\limits_{z \rightarrow 0} D_{\ssty A}=3,
1260: \lb{limDa}
1261: \ee
1262: \be \lim\limits_{z \rightarrow 0} D_{\ssty L}=3,
1263: \lb{limDl}
1264: \ee
1265: \be \lim\limits_{z \rightarrow 0} D_{\ssty Z}=3.
1266: \lb{limDz}
1267: \ee
1268: Again, these results are not surprising since, according to
1269: Fig.\ \ref{fig1}, all densities tend to OH at very small
1270: redshifts. The asymptotic limits of these functions at the
1271: Big Bang are also found easily, yielding,
1272: \be \lim\limits_{z \rightarrow \infty} D_{\ssty A}=0,
1273: \lb{limDa1}
1274: \ee
1275: \be \lim\limits_{z \rightarrow \infty} D_{\ssty L}=0,
1276: \lb{limDl1}
1277: \ee
1278: \be \lim\limits_{z \rightarrow \infty} D_{\ssty Z}=0.
1279: \lb{limDz1}
1280: \ee
1281: We compare these results with Eq.\ (\ref{Dg}) and conclude
1282: that the dimension $D$ constructed with the area distance $\da$,
1283: luminosity distance $\dl$, and redshift distance $\dz$
1284: corresponds to a vanishing fractal dimension at the Big Bang,
1285: whereas that constructed with the galaxy area distance $\dg$
1286: (or comoving distance) produce a finite nonzero dimension at
1287: the Big Bang singularity hypersurface. That indicates that the
1288: ambiguities arise when one considers all distance measures in
1289: cosmology, in view of the fact that the value of the fractal
1290: dimension at the Big Bang depends on the chosen cosmological
1291: distance definition. These results are graphically shown in
1292: Figs.\ \ref{fig5} and \ref{fig910}.
1293: \begin{figure}
1294: \centering
1295: \vspace{-5.cm}
1296: \hspace{-2.cm} \includegraphics[width=12.5cm]{f9.ps}
1297: %\input{f9.tex}
1298: \vspace{-8.5cm}
1299: \caption{Graph of the dimension $D$ defined in terms of the four
1300: distance measures $\da$, $\dg$, $\dl$, and $\dz$ and plotted
1301: against the redshift in the spatially-homogeneous EdS
1302: cosmological model. Assuming $H_0=70$ km s$^{-1}$ Mpc$^{-1}$
1303: and $M_g=10^{11} M_\odot$, we have $\ln\alpha^{1/3}=9.6353$.
1304: For $z \ll 1$, all dimensions are equal to 3 (see Fig.\
1305: \ref{fig910} and Eqs.\ [\ref{Dg}], [\ref{limDa}],
1306: [\ref{limDl}], [\ref{limDz}]). For higher values of $z$,
1307: both $D_{\ssty L}$ and $D_{\ssty Z}$ decrease steadily and
1308: vanish as $z \rightarrow \infty$. The dimension $D_{\ssty
1309: A}$ constructed with the area distance $\da$ shows an odd
1310: behavior, initially increasing very rapidly well above 3
1311: for $z>0.1$. However, according to Eq.\ (\ref{limDa1}),
1312: it eventually vanishes at the Big Bang singularity
1313: hypersurface which implies that this function must experience
1314: dramatic changes. Indeed, it is discontinuous at
1315: $z \approx 15170$, changing to negative values that increase
1316: towards zero. $D_{\ssty G}$ remains constant for all
1317: redshifts. This plot is a different way of presenting the
1318: results in Fig.\ \ref{fig1}. It is clear from this graph that
1319: although the EdS cosmology is spatially homogeneous, it may or
1320: may not be observationally homogeneous depending on the distance
1321: measure adopted for analyzing the behavior of the density
1322: in this model.
1323: }
1324: \lb{fig5}
1325: \end{figure}
1326: \begin{figure}
1327: \centering
1328: \vspace{-5.cm}
1329: \hspace{-2.cm} \includegraphics[width=12.5cm]{f10.ps}
1330: %\input{f10.tex}
1331: \vspace{-8.5cm}
1332: \caption{Graph of the dimension $D$ versus the redshift at very small
1333: values of $z$. Note that only when $z \approx 0.1$ that
1334: deviations from the constant values $D=3$ start to appear
1335: due to the definition of $D(z)$ in terms of the four
1336: distance measures.}
1337: \lb{fig910}
1338: \end{figure}
1339:
1340: \section{Number counts and magnitudes}\lb{mags}
1341:
1342: {The results discussed in the previous sections show that
1343: if we use the framework of the EdS cosmology, only when
1344: we have $z>0.1$ it does become theoretically possible to detect
1345: the distinction between SH and OH. If we depart from EdS
1346: cosmology and use, for instance, an open FLRW model the
1347: redshift ranges where this distinction becomes detectable
1348: could also change. Indeed, by means of differential
1349: densities calculated from the CNOC2 galaxy redshift survey
1350: data in the range $0.05 \le z \le 1$ in a FLRW model with
1351: $\Omega_{m_0}=0.3$ and $\Omega_{\Lambda_0}=0.7$, we are able
1352: to detect such a distinction for $z \approx 0.1$ (see A07)
1353: but, to do so we need to obtain indirectly the
1354: number counts by employing a method capable of calculating
1355: them from the galaxy luminosity function (LF). Presently
1356: the LF is evaluated from detailed observations of the
1357: apparent magnitude of galaxies and such magnitude
1358: measurements can go to redshifts much higher than the
1359: unity. To extract number counts from the LF, we
1360: require a relativistic theory connecting the theoretical
1361: aspects of a cosmological model with the usual procedures
1362: carried out by astronomers when quantifying galaxy
1363: catalogues. Ribeiro \& Stoeger (2003) developed such a
1364: method, A07 applied to the CNOC2 galaxy survey, and
1365: Iribarrem, Ribeiro \& Stoeger (2008, in preparation) applied
1366: to the FORS Deep Field redshift survey. These articles dealt
1367: with issues such as source evolution and K-correction (see
1368: also Ribeiro 2002) and the reader interested in those topics
1369: is referred to these articles since it is beyond the scope of
1370: this work to present a detailed discussion of these issues.
1371:
1372: Despite this, a simpler discussion about the relationship
1373: between number counts and magnitudes can be presented in the
1374: context of this paper. We define the bolometric \textit{apparent
1375: magnitude} $m$ to be given by the following Eq.,
1376: \be m=-2.5 \log \left( \frac{L}{4 \pi {\dl}^2} \right)
1377: + \mathrm{const.},
1378: \lb{m}
1379: \ee
1380: where $L$ is the intrinsic bolometric luminosity of a
1381: cosmological source, assumed point like (Ribeiro 2002),
1382: and the constant is due to the calibration of the magnitude
1383: system. Since by definition the bolometric \textit{absolute
1384: magnitude} $M$ is defined to be the apparent magnitude of a
1385: source located at a distance of 10 pc, the \textit{distance
1386: modulus} is defined as follows,
1387: \be \mu \equiv m-M = 5 \log \dl + 25.
1388: \lb{dm}
1389: \ee
1390: In this equation, the luminosity distance is measured in Mpc.
1391: Thus, if astronomical observations provide measurements of
1392: the pair $(m;z)$ for cosmological sources, which is common for
1393: sources of redshifts higher than unity, then, by assuming a
1394: cosmological model, we have the function $\dl=\dl(z)$ from which
1395: we can calculate $L$ using Eq.\ (\ref{m}) and then derive $M$ and
1396: $\mu$ for these sources. Therefore, the distance modulus is
1397: another way of representing observed magnitudes and
1398: redshifts of cosmological sources.
1399:
1400: Our aim is to relate the number counts given for each
1401: distance adopted in this paper to the distance modulus. This
1402: can be done if we use the reciprocity theorem provided by Eq.\
1403: (\ref{eth}) to write $\nid{A}$, $\nid{G}$, and $\nid{L}$, as given
1404: by Eq.\ (\ref{nd}), in terms of the luminosity distance
1405: and then use Eq.\ (\ref{dm}) to write the final
1406: expressions in terms of the distance modulus. The results may
1407: be written as follows,
1408: \be \nid{A} = { \left[ B \frac{10^{0.2(\mu-25)}}{{(1+z)}^2}
1409: \right] }^D,
1410: \lb{dnamu}
1411: \ee
1412: \be \nid{G} = { \left[ B \frac{10^{0.2(\mu-25)}}{(1+z)}
1413: \right] }^D,
1414: \lb{dngmu}
1415: \ee
1416: \be \nid{L} = { \left[ B \, 10^{0.2(\mu-25)} \right] }^D,
1417: \lb{dnlmu}
1418: \ee
1419: \be \nid{Z} = { \left[ \frac{cB}{H_0}z \right] }^D,
1420: \lb{dnzmu}
1421: \ee
1422: where the last Eq.\ is the result of simply substituting
1423: the definition given by Eq.\ (\ref{dz}) for the redshift
1424: distance, directly into Eq.\ (\ref{nd}) when taking $i=z$.
1425:
1426: To express the number counts above only in terms
1427: of the distance modulus, we require the function $z=z(\mu)$
1428: and to derive it we need to adopt a cosmological model as we
1429: did in the previous sections. In terms of observations, a
1430: cosmological model is required from the beginning of this approach,
1431: otherwise it would be impossible to find absolute magnitudes
1432: and, therefore, the distance moduli. As in previous sections, we
1433: adopt the EdS cosmology, which corresponds to using
1434: the luminosity distance as given in Eq.\ (\ref{dl}).
1435: Recalling that $\dl=0$ when $z=0$ Eq.\ (\ref{dl}) can
1436: be inverted to produce $z=z(\dl)$ and, after considering
1437: Eq.\ (\ref{dm}), we finally obtain the function $z(\mu)$
1438: given by,
1439: \be 1+z= \frac{1}{2}+ \frac{H_0}{2c}10^{0.2(\mu-25)}+
1440: \sqrt{ \frac{H_0}{2c}10^{0.2(\mu-25)}+ \frac{1}{4}}
1441: \; \; .
1442: \lb{zmu}
1443: \ee
1444: We note that Eqs.\ (\ref{dl}) and (\ref{dm}) imply that the
1445: small redshifts interval $0.001 \le z \le 0.1$ corresponds
1446: to $4.3 \, \mathrm{Mpc} \le \dl \le 439 \, \mathrm{Mpc}$ and
1447: $28.2 \le \mu \le 38.2$, if we assume that
1448: $H_0=70$~km~s$^{-1}$~Mpc$^{-1}$.
1449:
1450: For number counts to correspond to OH, we must choose $D=3$, which
1451: implies that the EdS cosmology number counts distribution is
1452: derived as in Sect.\ \ref{d3}, that is, is given by
1453: $N^\mathrm{\ssty EdS}(\mu)= \NID{G}{3} (\mu)$. Thus, only
1454: the expression $\NID{G}{3} (\mu)$ produces a number count
1455: distribution that is both OH and SH whereas the remaining
1456: functions $\NID{A}{3}(\mu)$, $\NID{L}{3}(\mu)$, and $\NID{Z}{3}
1457: (\mu)$ are OH, but are not SH. Figures \ref{fig-11} and
1458: \ref{fig-12} show graphs of these functions where it is
1459: clear that they are similar to those plotted in Figs.\
1460: \ref{fig2} and \ref{fig34}. Therefore, although these
1461: four expressions are built with a geometrical part
1462: consistent with EdS cosmology, this does not imply that
1463: the counts will be SH and OH, showing again the ambiguous
1464: nature of these expressions as far as ``homogeneity'' is
1465: concerned.
1466: \begin{figure}
1467: \centering
1468: \vspace{-5.cm}
1469: \hspace{-2.cm} \includegraphics[width=12.5cm]{f11.ps}
1470: %\input{f11.tex}
1471: \vspace{-8.5cm}
1472: \caption{This graph shows number counts versus distance modulus
1473: for the case $D=3$. The constant $B$ is given by Eq.\
1474: (\ref{nd}) which, for the EdS cosmology, turns out to be
1475: equal to Eq.\ (\ref{B}). The results are very similar
1476: to the ones shown in figure \ref{fig2}, meaning that the
1477: same conclusions reached there apply here.}
1478: \lb{fig-11}
1479: \end{figure}
1480: \begin{figure}
1481: \centering
1482: \vspace{-5.cm}
1483: \hspace{-2.cm} \includegraphics[width=12.5cm]{f12.ps}
1484: %\input{f12.tex}
1485: \vspace{-8.5cm}
1486: \caption{This is the same plot shown in figure \ref{fig-11}, but
1487: with a distance modulus range equivalent to small redshifts
1488: interval.}
1489: \lb{fig-12}
1490: \end{figure}
1491:
1492: }
1493:
1494: \section{The relativity of observational homogeneity}\lb{rel}
1495:
1496: We have seen how the concept of homogeneity applied to cosmological
1497: models is prone to ambiguities. Attempting to distinguish between
1498: spatial and observational homogeneities is a means of diminishing the
1499: ambiguities that, as seen above, have not been eliminated; this is
1500: because using various distance measures to calculate cosmological
1501: densities is still a source of ambiguity. At this point we recall
1502: some well known concepts that may help to clarify the physical
1503: interpretation of the effects discussed above.
1504:
1505: According to the reciprocity theorem given by Eq.\ (\ref{eth}),
1506: all distance
1507: definitions discussed above become equal at $z=0$. This means that
1508: if a signal such as pulses emitted at unit time intervals are
1509: emitted at the rest frame of the source and an observer measures
1510: the rate of change of the same signal, these rates of change are,
1511: by definition, the redshift $z$. In particular, the observed
1512: frequencies $\nu$ of light or radio waves are related to $z$ as
1513: $1+z={\nu_{_{\mbox \tiny \rm emi}}}/{\nu_{_{\mbox \tiny \rm obs}}}.$
1514: We can think of this as a {\it time dilation} effect (Ellis 1971).
1515: So, if a proper time interval $dt$ is observed to elapse between
1516: particular signals, then
1517: \begin{equation}
1518: \frac{dt_{\mbox \tiny \rm emitted}}{dt_{\mbox \tiny \rm observed}}
1519: = \frac{\nu_{\mbox \tiny \rm emitted}}{\nu_{\mbox \tiny \rm
1520: observed}}=1+z.
1521: \label{red}
1522: \end{equation}
1523: This relationship is true regardless of the separation of emitter and
1524: observer and implies that the difference between two distance
1525: measures in cosmology can be thought to be, in effect, a result of
1526: the time dilation between emitter and observer located in different
1527: reference frames in relative motion with one another. So, we can
1528: consider Eq.\ (\ref{eth}) to be produced by the relativity of
1529: time intervals. Inasmuch as the observable distances discussed in
1530: the previous sections are used to build observational densities,
1531: then we conclude that the concept of OH as defined in Eq.\
1532: (\ref{oh}) must also be relative. This means that similarly to the
1533: concept of a cosmological distance, we can talk about the relativity
1534: of OH and, therefore, we must abandon the notion of a ``true'', or
1535: unique, homogeneity of the observable Universe. In a similar way to
1536: the statement of McVittie (1974) about cosmological distances, we
1537: conclude that measuring the possible homogeneity of the large-scale
1538: structure in the Universe depends on circumstances, that is, on the
1539: method of measurement.
1540:
1541: The reasoning presented above can then help us to understand the
1542: limitations of the generic concept of homogeneity widely used in
1543: cosmology. It has its origins in the assumption of the maximal
1544: spatial isotropy in the Riemannian spacetime manifold, which then
1545: follows, as a mathematical result, that a perfect fluid cosmology
1546: metric ends up with its fluid variables (density and pressure)
1547: being time dependent only (Stephani et al.\ 2003, pp.\ 173, 210-212;
1548: Weinberg 1972, pp.\ 403, 412-415). This means that the local
1549: density $\rho$ that appears in the right hand side of Einstein's
1550: field equations becomes a function of the time coordinate only.
1551: Hence, a spatially-isotropic spacetime is, by mathematical
1552: requirement, spatially homogeneous as well. Due to this widely
1553: known result, it is usual to refer to the standard FLRW family of
1554: cosmological models as being characterized by isotropy and
1555: homogeneity. The adjective ``spatial'' is often dropped from
1556: appearing in front of the term homogeneity when the most basic
1557: features of the standard cosmology are described (e.g., see
1558: Peacock 1999, p.\ 65).
1559:
1560: Such an economy of language could, perhaps, have been thought
1561: harmless, but as a side effect it has in practice created a
1562: simplistic, but wrong, impression that all types of densities
1563: that can be derived in these cosmologies must also be
1564: homogeneous, that is, eventually become a constant value.
1565: As discussed in the previous sections, in contrast to this
1566: simplistic view it is possible to define densities in standard
1567: cosmologies that have different types of physical
1568: homogeneity.\footnote{Note that in this paper the term homogeneity
1569: was used with a physical meaning related to the average density,
1570: which can, in principle, be empirically determined, directly, or
1571: indirectly, by means of astronomical observations. Therefore,
1572: homogeneity has a wider meaning than the strict mathematical
1573: sense of spacetimes admitting isometries due to groups of motions
1574: (Stephani et al.\ 2003, pp.\ 157, 171).} The concept of OH is
1575: therefore fundamentally different from the concept of SH and,
1576: therefore, it is a misleading use of language to call the
1577: standard FLRW family of cosmological models simply isotropic and
1578: homogeneous. They are in fact isotropic and spatially homogeneous
1579: and either can or cannot be observationally homogeneous as well.
1580:
1581: %______________________________________________________________
1582:
1583: \section{Conclusion}\lb{conclu}
1584:
1585: In this paper we have presented an analysis of the physical
1586: consequences of the distinction between the usual concept of
1587: spatial homogeneity, as defined by the Cosmological Principle,
1588: and the concept of observational homogeneity. This distinction
1589: is based on calculating observational areas, volumes and
1590: densities with four cosmological distance measures
1591: $d_i \; (i={\sty A}, {\sty G}, {\sty L}, {\sty Z} \,)$,
1592: namely the area distance $\da$, the galaxy area distance $\dg$,
1593: the luminosity distance $\dl$, and the redshift distance $\dz$.
1594: Our aim was to simulate number counts as if they were actual
1595: observations. To do so in a general way, we have
1596: adopted the generalized number-distance relation $\Nid =
1597: { \left( B \, d_i \right) }^D$ to obtain the
1598: differential density $\gammaid$ and the integral differential
1599: density $\gestid$, where the latter is the observed radial
1600: number density $\anid$. In this way, these densities become a
1601: function of the fractal dimension $D$. We then reviewed the
1602: results of Ribeiro (2001b, 2005) and Albani et al.\ (2007),
1603: where those equations were applied to the Einstein-de Sitter
1604: cosmological model and the open FLRW cosmology with
1605: $\Omega_{m_0} =0.3$, $\Omega_{\Lambda_0}=0.7$, concluding
1606: that a spatially-homogeneous cosmology does not necessarily
1607: possess observational homogeneity. These features are only
1608: present if the galaxy area distance $\dg$, which in EdS
1609: cosmology is equivalent to the comoving distance apart from a
1610: constant, is used to calculate both differential densities.
1611:
1612: Models with and without observational homogeneity by construction
1613: were studied by means of setting $D=3$ and $D=2$ respectively in
1614: the generalized number-distance relation. It was found that
1615: models with $D=3$ do not seem to remain spatially homogeneous as
1616: well. The only exception appears to be when one adopts the galaxy
1617: area distance $\dg$. Models with $D=2$ were developed to be
1618: observationally inhomogeneous, although the integral differential
1619: density was constructed with the galaxy area distance $\dg$, which
1620: when plotted versus redshift shows a power-law decay for $z<2.5$
1621: that, after a transition, turns into a constant value for $z>10$.
1622: We have also studied the behavior of the dimension $D$ for
1623: the spatially-homogeneous EdS cosmology, showing that it tends
1624: to $D=3$ as $z \rightarrow 0$ for all distance measures, tends to
1625: $D=0$ as $z \rightarrow \infty$ for $\da$, $\dg$, and $\dl$,
1626: but remains $D=3$ for $\dg$ at the Big Bang singularity
1627: hypersurface. {Finally, we have also studied functions of
1628: number counts versus distance modulus with the various distance
1629: definitions and reached conclusions similar to models with
1630: $D=3$ and functions of number counts versus redshift.} The paper
1631: finishes with a conceptual discussion arguing that due to the
1632: relativity of time intervals for pulses emitted and observed at
1633: different reference frames, and in view of the reciprocity
1634: theorem linking various cosmological distances by means of
1635: $(1+z)$ factors, we can conclude that the concept of observational
1636: homogeneity should also be relative.
1637:
1638: To end this paper, it is important to emphasize that the conceptual
1639: distinction discussed above between different types of homogeneity
1640: in the standard cosmological model is fundamental and has important
1641: consequences for observational cosmology. In view of the fact that
1642: such a distinction is not generally recognized in the literature of
1643: observational cosmology, it is our opinion that it should be
1644: considered by all those who empirically probe the possible
1645: observational homogeneity of the large-scale distribution of galaxies
1646: in the Universe.
1647:
1648: %______________________________________________________________
1649:
1650: \begin{acknowledgements}
1651: We thank M. Montuori for discussions that motivated this work.
1652: {We are also grateful to a referee for interesting remarks,
1653: which improved the paper.} One of us (LJRL) acknowledges the financial
1654: support from CAPES Foundation.
1655: \end{acknowledgements}
1656:
1657: \begin{thebibliography}{}
1658:
1659: \bibitem[2001]{amr01} Abdalla, E., Mohayaee, R., \& Ribeiro, M.B.\ 2001,
1660: Fractals 9, 451, arXiv:astro-ph/9910003v4
1661: \bibitem[2007]{airs07} Albani, V.V.L., Iribarrem, A.S., Ribeiro, M.B.,
1662: \& Stoeger, W.R. 2007, Astrophys.\ J.\ 657, 760,
1663: arXiv:astro-ph/0611032v1 \textbf{(A07)}
1664: \bibitem[1970]{dV70} de Vaucouleurs, G.\ 1970, Science 167, 1203
1665: \bibitem[1971]{e71} Ellis, G.F.R.\ 1971, General Relativity and
1666: Cosmology, Proc.\ Int.\ School of Phys.\ ``Enrico Fermi'',
1667: ed.\ R.K.\ Sachs (Academic Press, New York), 104
1668: \bibitem[2007]{e07} {Ellis, G.F.R.\ 2007, Gen.\ Rel.\ Grav.\
1669: 39, 1047}
1670: \bibitem[1933]{eth33} Etherington, I.M.H.\ 1933, Phil.\ Mag.\ 15, 761;
1671: {reprinted in Gen.\ Rel.\ Grav.\ 39 (2007) 1055-1067}
1672: \bibitem[2005]{gsjp05} Gabrielli, A., Sylos Labini, F., Joyce, M., \&
1673: Pietronero, L.\ 2005, Statistical Physics for Cosmic Structures
1674: (Springer, Berlin)
1675: \bibitem[2005]{hebbbgs05}Hogg, D.W., Eisenstein, D.J., Blanton, M.R.,
1676: Bahcall, N.A., Brinkmann, J., Gunn, J.E., \& Schneider, D.P.
1677: 2005, Astrophys.\ J.\ 624, 54, arXiv:astro-ph/0411197v1
1678: \bibitem[2005]{jmst05} Jones, B.J.T., Mart\'{\i}nez, V.J., Saar, E.,
1679: \& Trimble, V.\ 2005, Rev.\ Mod.\ Phys.\ 76, 1211,
1680: arXiv:astro-ph/0406086v1
1681: \bibitem[1999]{jmsp99} Joyce, M., Montuori, M., Sylos Labini, F., \&
1682: Pietronero, L.\ 1999, A\&A 344, 387, arXiv:astro-ph/9805126v3
1683: \bibitem[2000]{jamps00} {Joyce, M., Anderson, P.\ W., Montuori, M.,
1684: Pietronero, L., \& Sylos Labini, F.\ 2000, Europhys.\ Lett.\
1685: 50, 416, arXiv:astro-ph/0002504v1}
1686: \bibitem[2005]{jsgmp05} Joyce, M., Sylos Labini, F., Gabrielli, A.,
1687: Montuori, M., \& Pietronero, L.\ 2005, A\&A 443, 11,
1688: arXiv:astro-ph/0501583v2
1689: \bibitem[1974]{m74} McVittie, G.C.\ 1974, Quart.\ J.\ Royal Astr.\
1690: Soc.\ 15, 246
1691: \bibitem[1999]{p99} Peacock, J.A.\ 1999, Cosmological Physics (Cambridge
1692: University Press)
1693: \bibitem[1987]{p87} Pietronero, L.\ 1987, Physica A 144, 257
1694: \bibitem[2006]{pk06} Pleba\'{n}ski, J., \& Krasi\'{n}ski, A.\ 2006, An
1695: Introduction to General Relativity and Cosmology (Cambridge
1696: University Press)
1697: \bibitem[1992]{rib92} Ribeiro, M.B.\ 1992, On Modelling a Relativistic
1698: Hierarchical (Fractal) Cosmology by Tolman's Spacetime, Ph.D.\
1699: Thesis, Queen Mary and Westfield College, University of London
1700: \bibitem[1992a]{rib92a} Ribeiro, M.B.\ 1992a, Astrophys.\ J.\ 388, 1
1701: \bibitem[1992b]{rib92b} Ribeiro, M.B.\ 1992b, Astrophys.\ J.\ 395, 29
1702: \bibitem[1993]{rib93} Ribeiro, M.B.\ 1993, Astrophys.\ J.\ 415, 469
1703: \bibitem[1994]{rib94} Ribeiro, M.B.\ 1994, in Deterministic Chaos in
1704: General Relativity, eds.\ D.W.\ Hobill, A.\ Burd, \& A.\
1705: Coley (Plenum, New York) 269
1706: \bibitem[1995]{rib95} Ribeiro, M.B.\ 1995, Astrophys.\ J.\ 441, 477,
1707: arXiv:astro-ph/9910145v1
1708: \bibitem[2001a]{rib01a} Ribeiro, M.B.\ 2001a, Fractals 9, 237,
1709: arXiv:gr-qc/9909093v2
1710: \bibitem[2001b]{rib01b} Ribeiro, M.B.\ 2001b, Gen.\ Rel.\ Grav.\ 33,
1711: 1699, arXiv:astro-ph/0104181v1 \textbf{(R01b)}
1712: \bibitem[2002]{rib02} {Ribeiro, M.B.\ 2002, Observatory 122,
1713: 201, arXiv:gr-qc/9910014v2}
1714: \bibitem[2005]{rib05} Ribeiro, M.B.\ 2005, A\&A 429, 65,
1715: arXiv:astro-ph/0408316v2 \textbf{(R05)}
1716: \bibitem[1998]{rm98} Ribeiro, M.B., \& Miguelote, A.Y.\ 1998, Braz.\
1717: J.\ Phys.\ 28, 132, arXiv:astro-ph/9803218v1
1718: \bibitem[2003]{rs03} {Ribeiro, M.B., \& Stoeger, W.R.\ 2003,
1719: Astrophys.\ J.\ 592, 1, arXiv:astro-ph/0304094v1}
1720: \bibitem[1988]{s88} Sandage, A.\ 1988, Ann.\ Rev.\ Astron.\
1721: Astrophys.\ 26, 561
1722: \bibitem[2003]{skmhh03} Stephani, H., Kramer, D., MacCallum, M.A.H.,
1723: Hoenselaers, C, \& Herlt, E.\ 2003, Exact Solutions of Einstein's
1724: Field Equations, 2nd ed. (Cambridge University Press)
1725: \bibitem[1998]{smp98} Sylos Labini, F., Montuori, M., \& Pietronero,
1726: L.\ 1998 Phys.\ Rep.\ 293, 61, arXiv:astro-ph/9711073v1
1727: \bibitem[1972]{w72} Weinberg, S.\ 1972, Gravitation and Cosmology
1728: (Wiley, New York)
1729: \bibitem[1970]{w70} Wertz, J.R.\ 1970, Newtonian Hierarchical Cosmology,
1730: Ph.D.\ Thesis, University of Texas at Austin
1731: \bibitem[1971]{w71} Wertz, J.R.\ 1971, Astrophys.\ J.\ 164, 227
1732: \bibitem[2005]{ybps05} Yadav, J., Bharadwaj, S., Pandey, B., \&
1733: Seshadri, T.R.\ 2005, Mon.\ Not.\ Roy.\ Astron.\ Soc.\ 364, 601,
1734: {arXiv:astro-ph/0504315v1}
1735: \end{thebibliography}
1736:
1737: \end{document}
1738: