1: \documentclass[aps,prl,superscriptaddress,twocolumn,floats,tightenlines,balancelastpage,floatfix]{revtex4}
2: \input epsf
3: \usepackage{amsmath,amssymb}
4: \usepackage{epsfig}
5: \usepackage{graphicx}
6:
7: %\addtolength{\voffset}{2cm}
8: %\addtolength{\hoffset}{1.5cm}
9: %\addtolength{\textwidth}{-3cm}
10:
11: \def\mathrelfun#1#2{\lower3.6pt\vbox{\baselineskip0pt\lineskip.9pt
12: \ialign{$\mathsurround=0pt#1\hfil##\hfil$\crcr#2\crcr\sim\crcr}}}
13: \def\simlt{\mathrel{\mathpalette\mathrelfun <}}
14: \def\simgt{\mathrel{\mathpalette\mathrelfun >}}
15: \def\simpropto{\mathrel{\mathpalette\mathrelfun \propto}}
16: \def\ln {{\rm ln}}
17: \def\hatbfn{{\hat{\bf n}}}
18: \def\hatbfc{{\hat{\bf c}}}
19: \def\bfx{{\bf x}}
20: \def\RV{{R_{\rm V}}}
21: \def\zV{{z_{\rm edge}}}
22: \def\zrei{{z_{\rm rei}}}
23: \def\zi{{z_{\rm i}}}
24: \def\ze{{z_{\rm e}}}
25: \def\DA{{D_{\rm A}}}
26: \def\Heo{{H_{{\rm e}0}}}
27: \def\rrei{{r_{\rm rei}}}
28:
29: \def\sqz{\vspace{-0.1cm}}
30:
31: \begin{document}
32:
33: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
34: \title{A Test of the Copernican Principle}
35:
36: \author{R.~R. Caldwell}
37: \affiliation{Department of Physics \& Astronomy, 6127 Wilder Lab, Dartmouth College,
38: Hanover, NH 03755}
39:
40: \author{A. Stebbins}
41: \affiliation{Theoretical Astrophysics Group, Fermi National Accelerator
42: Laboratory, P.O. Box 500, Batavia, IL 60510}
43:
44: \date{\today}
45:
46: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
47: \begin{abstract}
48:
49: The blackbody nature of the cosmic microwave background (CMB) radiation spectrum is
50: used in a modern test of the Copernican Principle. The reionized universe serves as
51: a mirror to reflect CMB photons, thereby permitting a view of ourselves and the
52: local gravitational potential. By comparing with measurements of the CMB spectrum, a
53: limit is placed on the possibility that we occupy a privileged location, residing at
54: the center of a large void. The Hubble diagram inferred from lines-of-sight
55: originating at the center of the void may be misinterpreted to indicate cosmic
56: acceleration. Current limits on spectral distortions are shown to exclude the
57: largest voids which mimic cosmic acceleration. More sensitive measurements of the
58: CMB spectrum could prove the existence of such a void or confirm the validity of the
59: Copernican Principle.
60:
61: \end{abstract}
62: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
63:
64: \maketitle
65:
66: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
67:
68: {\it Introduction:} The observed accelerating expansion of the universe
69: \cite{Riess:1998cb,Perlmutter:1998np} poses deep questions for cosmology. Is the
70: universe filled by some new, exotic dark energy, or is one of the basic tenets of
71: the standard model of cosmology invalid? One such tenet is the Cosmological
72: Principle, the assumption of approximate homogeneity and isotropy of matter and
73: radiation throughout the universe. The Cosmological Principle is known to be partly
74: satisfied. The universe is observed to be very nearly isotropic on our celestial
75: sphere, on the basis of the near-isotropy of the CMB temperature pattern
76: %\cite{Hajian:2006ud,Jaffe:2005pw}
77: \cite{CMBisotropy}. The universe is observed to be approximately
78: homogeneous across the distances probed by large-scale structures
79: \cite{Hogg:2004vw}. Yet, radial homogeneity on cosmic scales $\simgt1\,$Gpc remains
80: to be proven. If the assumption of radial homogeneity is relaxed, and if we observe
81: from a preferred vantage point, then it may be possible to explain the apparent
82: cosmic acceleration in terms of a peculiar distribution of matter centered upon our
83: location
84: %\cite{Celerier:1999hp, Barrett:1999fd, Tomita:2001a, Tomita:2001b,
85: %Iguchi:2001sq, Moffat:2005ii, Alnes:2005rw, Mansouri:2005rf, Chuang:2005yi,
86: %Vanderveld:2006rb, Garfinkle:2006sb, BiswasMansouriNotari2006, Chung:2006xh,
87: %Enqvist:2006cg, Alnes:2006, Celerier:2007}
88: \cite{RefGroupA}. In fact, models of the universe
89: consisting of a spherically-symmetric distribution of matter, mathematically
90: described by a Lemaitre-Tolman-Bondi spacetime \cite{Bondi:1947av}, have been shown
91: to produce a Hubble diagram which is consistent with observations. These models
92: require no cosmological constant or other form of dark energy, and locally resemble
93: a matter-dominated low-density universe or void. The observed near-isotropy
94: constrains us to occupy a very special location, at or near the center of the void,
95: in violation of the Copernican Principle. Although the Copernican Principle may be
96: widely accepted by {\it fiat}, it is imperative that such a foundational principle
97: be proven.
98:
99: We propose a test of the Copernican Principle, to verify radial homogeneity and
100: thereby constrain non-accelerating void cosmological models. The test relies on a
101: previously under-appreciated effect: the mixture of anisotropic CMB radiation
102: through scattering leads to distortions of the blackbody spectrum
103: \cite{Stebbins2007}. The CMB is initially thermal (blackbody), but small
104: inhomogeneities cause variations in the temperature at different locations and along
105: different lines-of-sight that preserve the blackbody spectrum. However, scattering
106: of this anisotropic radiation into our line-of-sight by ionized gas produces
107: observable spectral distortions. This allows us to indirectly detect large
108: anisotropies in other parts of the universe.
109:
110: Here we are interested in anisotropies caused by a large, local void. Such a
111: structure causes ionized gas to move outward, in motion relative to the CMB frame
112: which leads to a Doppler anisotropy in the gas frame. The gravitational potential of
113: such a structure also leads to a Sachs-Wolfe (SW) effect for photons which originate
114: inside of the void and scatter back toward us. The geometry of these effects is
115: illustrated in fig.~\ref{fig:bubble}. A large void, or any other non-Copernican
116: structure, will lead to large anisotropies in other places which will be reflected
117: back at us in the form of spectral distortions. Hence, deviations from a blackbody
118: spectrum can indicate a violation of the Copernican Principle. In essence, we use
119: the reionized universe as a mirror to look at ourselves in CMB light. If we see
120: ourselves in the the mirror it is because ours is a privileged location. If we see
121: nothing in the mirror, then the Copernican Principle is upheld.
122:
123: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
124: {\it Spectral Distortions:}
125: %Here we give the tools needed to implement our test of the Copernican Principle.
126: %\subsection{$u$-distortions}
127: %\sqz
128: The distortion of the CMB blackbody spectrum due to scattering by anisotropic CMB
129: radiation is \cite{Stebbins2007}
130: $u[\hatbfn]={3\over16\pi}\int_0^\infty
131: dz\,{d\tau\over dz}\, \int
132: d^2\hatbfn'\,\left(1+(\hatbfn\cdot\hatbfn')^2\right)$
133: $\times$
134: $\left( {\Delta T\over T}[\hatbfn ,\hatbfn,z] -{\Delta T\over
135: T}[\hatbfn',\hatbfn,z]\right)^2,
136: $
137: where $\Delta T/T[\hatbfn',\hatbfn,z]$ is the CMB temperature anisotropy in the
138: direction $\hatbfn'$, as observed at redshift $z$ in the direction $\hatbfn$ from
139: the central observer, and $\tau$ is the optical depth. For cosmic voids extending
140: out to redshifts $z\simlt1$, reflections back at us may occur up to $z\simlt3$
141: (see fig.~\ref{fig:bubble}). The optical depth to Thomson scattering is small,
142: so that it is appropriate to consider single scattering. Since the mean CMB
143: temperature is not known {\it a priori}, but rather is fit to the observations, $u$
144: is observationally degenerate with the Compton $y$-distortion parameter according to
145: the relation $u=2y$. (Compare Refs.~\cite{Stebbins2007,Chluba:2003} for details.)
146: Thus observational constraints on $2y$ can can be treated as constraints on $u$.
147:
148:
149: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
150: \begin{figure}[h]
151: \centerline{\epsfig{file=Figure1.eps,width=20pc,angle=0}}
152: \caption{Illustrated is a cross-section through a model universe with the observer
153: (O) at the center of a void, in violation of the Copernican Principle. CMB photons
154: traveling in any direction may Thomson scatter off reionized gas toward the observer.
155: The final spectrum of the observed light will be a mixture of blackbody spectra with
156: different (anisotropic) tempertures, producing a distorted blackbody. The
157: yellow lines represent: incoming beams of unscattered, primary CMB photons (dashed);
158: incoming beams of scattered photons (thin), and the observed beams (thick) for
159: representative scattering centers with last scattering surfaces represented by
160: the dark circles. A is in the {\it Doppler zone}: Beams 1-3 experience the same SW
161: temperature shift, introducing no anisotropy. However,
162: gradients in the void gravitational potential cause the gas to move with respect to
163: the CMB frame, so A sees a differential Doppler anisotropy, resulting in spectral
164: distortions. B is in the {\it reflection zone}: B is at rest with respect to
165: the CMB frame and sees no Doppler anisotropy. However, some of the incoming photons,
166: {\it e.g.} beam 4, originate inside the void so there will be an anisotropic SW
167: temperature shift, leading to spectral distortions.
168: %B is
169: %outside the void so it is not moving with respect to the CMB frame and sees no
170: %Doppler anisotropy; however some of the incoming photons, {\it e.g.} beam 4,
171: %originate inside the void so there will be an anisotropic SW temperature shift,
172: %leading to a spectral distortion. B is in the {\it reflection zone}. C is outside
173: %the void and all its incoming beams originate outside the void, so there is no
174: %temperature anisotropy and C does not contribute to the spectral distortion.
175: }
176: \label{fig:bubble}
177: \vspace{-0.3cm}
178: \end{figure}
179: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
180:
181: %\subsection{Temperature Anisotropies}
182:
183: %\sqz
184: We consider a low-amplitude void embedded in a flat, Einstein-deSitter (EdS or
185: $\Omega=1$) matter-dominated universe. The gravitational potential due to
186: the void, $\Phi[\bfx]$, is a function of comoving position, $\bfx$, with Earth near
187: $\bfx=0$. The temperature anisotropy can be divided into a Sachs-Wolfe and Doppler
188: term
189: ${\Delta T\over T}[\hatbfn',\hatbfn,z]=
190: {\Delta T\over T}|_{\rm SW}+{\Delta T\over T}|_{\rm Doppler}$
191: where
192: ${\Delta T\over T}|_{\rm SW}=
193: \frac{1}{3 c^2}(\Phi[\bfx_{\rm rec}]-\Phi[\bfx_{\rm scatter}])$
194: and
195: ${\Delta T\over T}|_{\rm Doppler}=\frac{2}{3}
196: {\hatbfn'\cdot{\bf \nabla}_\bfx\Phi[\bfx_{\rm scatter}]}/{c\,H_0\sqrt{1+z}}$,
197: where
198: $\bfx_{\rm scatter}= D_{\rm A}^{\rm co}[z]\,\hatbfn$,
199: $\bfx_{\rm rec}= \bfx_{\rm scatter}
200: +(D_{\rm A}^{\rm co}[z_{\rm rec}]-D_{\rm A}^{\rm co}[z])\hatbfn'$,
201: $D_{\rm A}^{\rm co}[z]=2{c\over\Heo}\,\left(1-{1\over\sqrt{1+z}}\right)$.
202: Here $D_{\rm A}^{\rm co}$ is the comoving angular diameter distance, and the
203: redshift of recombination, $z_{\rm rec}$, will be approximated by $\infty$ for
204: simplicity. The Hubble constant at the present time in the background cosmology,
205: outside the void, is $\Heo$, whereas $H_0$ is the larger, present-day Hubble
206: constant at the center of the void.
207:
208: We neglect the integrated Sachs-Wolfe (ISW) effect, meaning that a CMB photon does
209: not contribute to the $u$-distortion simply because it passes across the void. This
210: approximation is justified for a low-amplitude void in the EdS background where the
211: ISW is a second-order effect. As $\Omega$ deviates from unity and/or the void
212: amplitude becomes non-linear we expect a larger ISW contribution to the anisotropy
213: and thus to the spectral distortion, but we do not expect that the ISW will ever be
214: the dominant contributor to $u$ for the small voids needed to mimic an accelerating
215: universe.
216:
217: %\subsection{Optical Depth}
218:
219: %\sqz
220: The run of optical depth with redshift is taken from the unperturbed, background
221: cosmology. We assume a rapid reionization at $z=\zrei$ such that
222: ${d\tau\over dz}=\tau_{\rm e0}' \, \sqrt{1+z}\,\Theta[\zrei-z]$
223: $\tau_{\rm e0}' = {3\Heo\Omega_{\rm b0}\sigma_{\rm T} c\over8\pi G m_{\rm H}}\,
224: \left(1-{1\over2}Y_{\rm He}\right)$,
225: where $\Theta[x]$ is the Lorentz-Heaviside step function, $\sigma_{\rm T},\, m_{\rm
226: H},\, \Omega_{\rm b0}$, and $Y_{\rm He}$ are the Thomson cross-section, the hydrogen
227: mass, the current baryonic mass density in units of the critical density, and the
228: helium mass fraction, respectively. We use $\Omega_{\rm b0}h^2=0.022$
229: ($h\equiv\Heo/100$km/s/Mpc), $Y_{\rm He}=0.24$. For $H_0$ we use the
230: locally-measured expansion rate: $73$~km/s/Mpc ({\it e.g.}
231: Refs.~\cite{Freedman:2000cf,Riess:2005zi}). Where needed we use the WMAP3
232: \cite{SpergelBean2006} value, $\tau_{\rm obs}=0.9$, for the optical depth to the
233: surface of last-scattering which in our model gives $\zrei=11$. These numbers
234: specify the cosmic evolution of the density of scatterers.
235:
236: %\subsection{Spherical Void}
237:
238: %\sqz
239: We assume spherical symmetry for the local void. Consequently, the gravitational
240: potential is $\Phi[\bfx]=\Phi[R=|\bfx|]$, where $R$ is the comoving radial distance
241: from Earth. The temperature anisotropy $\Delta T/T$ depends on the
242: directions $\hatbfn$ and $\hatbfn'$ only through the combination
243: $\hatbfn\cdot\hatbfn'$, which leaves $u$ $\hatbfn$-independent. Thus the final
244: result is a single number, the $u$-distortion at Earth, which can be translated into
245: a limit on any local spherical inhomogeneity.
246:
247:
248: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
249: {\it Void Model:}
250: %\sqz
251: We cannot compute $u$ for every possible void profile, so we focus our attention on
252: a particularly simple, two parameter class of voids, sometimes known as a
253: {\it Hubble bubble}:
254: $\Phi[R]=\Phi_0\,\left(1-{R^2\over R_{\rm V}^2}\right)\,{\Theta}[\RV-R]$.
255: The parameters $\Phi_0,\,\RV$ give the void amplitude and comoving radius. The
256: reason it is called a Hubble bubble is that the Hubble parameter
257: is uniform inside and outside the void, but the values differ. Nonlinear growth
258: leads to the appearance of a shell of mass overdensity which compensates the
259: underdensity in the void at the boundary of the outer and inner region. This
260: compensating shell has a complicated density and velocity structure, which is safely
261: ignored in linear theory. Away from the compensating shell this model resembles an
262: open ($\Omega_0<1$) FRW cosmology embedded inside a flat EdS cosmology. Any smooth
263: spherical void which is asymptotically EdS at large $R$ and has finite density in
264: the center can be thought of in this way; what differs is the radial profile of the
265: transition between the two FRW spacetimes. The Hubble bubble is the limit of a
266: sharp transition between the interior and exterior regions.
267:
268:
269: %\subsection{Nonlinear Parameterizations}
270:
271: %\sqz
272: The Hubble bubble amplitude can be expressed in terms of the present-day density
273: parameter, $\Omega_0$, inside the void as
274: $\Phi_0={3\over20}\,(\Heo\RV)^2
275: \left[\frac{(1-\Omega_0)^{3/2}}{\Omega_0}\frac{H_0}{\Heo}\right]^{2/3}$.
276: Next, the void radius can be expressed as a function of the redshift at the edge of
277: the void, $\zV$. This relationship is complicated by finite peculiar velocities and
278: non-linear clustering of the compensating shell, but to first order is simply
279: $\RV=2(c/ H_0)(1-{1/\sqrt{1+\zV}})$. Finally, the exterior Hubble parameter, $\Heo$,
280: differs from the interior value, $H_0$. At the same ``time since bang" they are
281: related as
282: $\frac{H_0}{\Heo}={3\over2}{\sqrt{1-\Omega_0}-\Omega_0\,
283: {\rm sinh}^{-1}\left[\sqrt{1-\Omega_0\over\Omega_0}\right]
284: \over(1-\Omega_0)^{3/2}}$.
285: Note that a small jump in the Hubble parameter corresponds to a large jump
286: in the density parameter (see fig.~\ref{HBfig}).
287:
288: %\subsection{Zones and Void Sizes}
289: %\label{sizes}
290:
291: %\sqz
292: The gas at different redshift must satisfy several criteria in order to contribute
293: to the $u$-distortion. A patch of gas at redshift $z$ must be ionized, on our past
294: light cone, and see an anisotropic $\Delta T$ from the void. We refer to the region
295: $z>\zrei$ as the {\it neutral zone} because the gas is not ionized, producing no
296: contribution to $u$. Even if the gas is ionized, if $z> z_{\rm max} \equiv 3+4\,\zV$
297: then the gas is not in causal contact with the void so $\Delta T=0$. We refer to the
298: range $(2\sqrt{z_{\rm max}+1}-1)/(\sqrt{z_{\rm max}+1}-1)^2 \le z \le z_{\rm max}$
299: as the {\it reflection zone}; the last scattering surface of gas in this range intersects
300: the interior of the void so that, depending on the scattering angle, some CMB photons
301: will reflect back towards us with anisotropy $\Delta T_{\rm SW}$. Gas which is on
302: our past light cone and within the void will see $\Delta T_{\rm Doppler}$, which we
303: call the {\it Doppler zone}.
304:
305: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
306: \begin{figure}
307: \centerline{\epsfig{file=Figure2.eps,width=19pc,angle=0}}
308: \caption{The dependence of the spectral distortion, $u$, on the size of a Hubble
309: bubble parameterized by $\zV$, is shown in units of $u_{\rm scale}\equiv\tau_{\rm
310: e0}'\,(\Phi_0/(3c^2))^2(1+\zV)/(\sqrt{1+\zV}-1)$. The thick curves show the
311: various contributions to $u$. The dashed curves correspond to the case in
312: which $\zrei\to\infty$. For small voids the Doppler contribution
313: dominates, and the value of $\zrei$ is unimportant.}
314: \label{uofzv}
315: \vspace{-0.3cm}
316: \end{figure}
317: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
318:
319: Five classes of void sizes are identified depending on how the different zones
320: overlap (assuming $\zrei > 8$): {\it small} ($\zV\le{5\over4}$) whereby the neutral
321: zone, reflection zone, and Doppler zone are all disjoint; {\it big}
322: (${5\over4}<\zV\le{1\over4}(\zrei-3)$) whereby the Doppler zone and reflection zones
323: overlap, but neither overlap the neutral zone; {\it large}
324: (${1\over4}(\zrei-3)<\zV\le\zrei$) in which the Doppler and reflection zones
325: overlap, as do the reflection and neutral zones, but the neutral zone does not
326: overlap the Doppler zone; {\it huge} ($\zV>\zrei$) in which the neutral, reflection,
327: and Doppler zones all overlap; and {\it super-horizon} ($\RV>2c/H_0$) for which the
328: void encompasses the entire observable universe. This classification is not
329: restricted to the Hubble bubble void profile, but applies to any void profile with a
330: sharp edge at $z=\zV$. As we shall see it is only the small voids that can explain
331: the current SNe data.
332:
333: %\subsection{Expression for $u$}
334:
335: %\sqz
336: In the linear perturbation approximation for this void model the spectral distortion
337: $u$ is proportional to $({1\over3c^2}\Phi_0)^2\tau_{\rm e0}$ and may be decomposed
338: as
339: $u=\tau_{\rm e0}\,\left({\Phi_0\over3c^2}\right)^2\,
340: \left(U_{\rm D}+U_{\rm S}+U_{\rm DS}\right)$
341: where the three terms are, respectively, the contribution from gas where the
342: temperature anisotropies are Doppler only (subscript D), Sachs-Wolfe only
343: (subscript S), and a combination of the two (subscript DS). All of these can be
344: expressed analytically. For small and big voids $u$ does not depend on $\zrei$ but
345: only on the dimensionless size parameter $r\equiv{1\over2}{H_0\over
346: c}\RV=1-{1\over\sqrt{1+z_V}}$. The general expression for $u$ is long and we do not
347: give it here. For small voids, which are the most relevant, we find
348: $U_{\rm D}^{\rm small}={28\over5}{1\over r^3}
349: \left(1+{1\over1-r}+{2\over r}\ln[1-r]\right)$
350: and $U_{\rm S}^{\rm small}\ll U_{\rm D}^{\rm small}$ and $U_{\rm DS}^{\rm small}=0$.
351:
352:
353: %\subsection{Hubble Diagram for a Hubble Bubble}
354:
355: %\sqz
356: The angular-diameter distance $\DA$ is a solution of the Dyer-Roeder
357: \cite{DyerRoeder} equation, $\frac{d}{dz}\left((1+z)^2 H \frac{d}{dz}\DA\right) +
358: \frac{3}{2}\Omega H \DA=0$. In the interior open and exterior flat cosmologies the
359: respective solutions are
360: $\DA[z<\zi]={2c\over H_0}
361: ({2-\Omega_0(1-z)-(2-\Omega_0)\sqrt{1+z\Omega_0})/(1+z)^2 \Omega_0^2} $ and
362: $\DA[z>\ze]={2c\over H_0}(\frac{C_1}{(1+z)} + \frac{C_2}{(1+z)^{3/2}})$,
363: where the coefficients $C_1,\,C_2$ are set by the continuity $\DA^{\rm
364: int}[\zi]=\DA^{\rm ext}[\ze]$, and the jump in $d\DA/dz$ as determined by integrating
365: the Dyer-Roeder equation across the delta-function density spike at the void edge.
366: The radial velocity drop, $\Delta v$, at the void edge means a double-valued
367: $\DA[z]$ for $z\in[\ze,\zi]$ and ${1+\zi\over1+\ze}=\sqrt{c+\Delta v\over c-\Delta
368: v}$. This drop also gives the Doppler anisotropy at the edge. To get $\zi$ and
369: $\ze$ we use the approximations ${\Delta v\over c}={\Delta T\over T}_{\rm
370: Doppler}[\zV,\hatbfn,\hatbfn]$ and $\zV={1\over2}(\zi+\ze)$. The luminosity distance
371: versus redshift, a.k.a. the Hubble diagram, is $(1+z)^2\,D_{\rm A}[z]$.
372:
373: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
374: {\it Constraints:}
375: %\sqz
376: The $u$-distortion is evaluated according to the procedure described above. We are
377: primarily interested in small and big voids which extend out to $z\sim 1$. Hence our
378: constraints are independent of $\zrei$. The other cosmological parameters only
379: enter into the overall normalization of $u$ through $\tau_{\rm e0}' $. What remains
380: are the void, size and amplitude: ($\zV,\,\Omega_0$).
381:
382: The best current bound on $u$ is due to FIRAS
383: \cite{Mather:1994,BoggessMatherEtAl1992,FixsenCheng1996} which constrains
384: $y<15\times10^{-6}$ or $u<3\times10^{-5}$ at $95\%$ C.L.. The corresponding
385: constraint on Hubble bubble parameters are shown in fig.~\ref{HBfig}. Also shown
386: are constraints for projected bounds $y<10^{-6},\, 10^{-7}$. The limits are
387: expected to improve \cite{Fixsen:2002,Kogut:2006kf}, but a $y$-distortion from the
388: IGM would likely mask the signal discussed here if $u\simlt10^{-6}$
389: \cite{ZhangPenTrac}.
390:
391:
392: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
393: \begin{figure}
394: \centerline{\epsfig{file=Figure3.eps,width=20pc,angle=0}}
395: \caption{The test of the Copernican Principle, in terms of constraints on the
396: size and depth of a local, spherically-symmetric void, is shown. The blue shaded
397: regions show the range of parameters excluded by the $u$-distortion test, whereas
398: the red regions show the range of parameters compatible with the current SNe Hubble
399: diagram data.}
400: \label{HBfig}
401: \vspace{-0.3cm}
402: \end{figure}
403: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
404:
405: The results rule out large voids with large density contrasts -- the most egregious
406: violations of the Copernican Principle. The larger the void, the smaller the
407: density contrast must be in order to pass the test. Although not shown, the
408: constraints become weaker for huge (nearly super-horizon sized) voids. Since
409: observationally $\Omega_0 \lesssim 0.3$, only small bubbles with $\zV <0.9$ are
410: allowed. Improving the constraint to $y<10^{-6}$ would lower this bound to
411: $\zV\lesssim0.3$ or a radius of $1$~Gpc. These constraints are consistent with the
412: very small Hubble bubble proposed in Ref.~\cite{Jha:2006fm}, with $H_0-\Heo \sim0.1
413: H_0$ and $\zV\simgt0.025$.
414:
415: The observed SNe data can be compared with our model Hubble diagram to further
416: constrain void parameters. Using the SNe data \cite{Riess:2006fw,WoodVasey:2007jb}
417: compiled in Ref.~\cite{Davis:2007na}, we computed the likelihood of $\Omega_0$ and
418: $\zV$. The best-fit parameter combinations give $\chi^2=207$ for the $192$ SNe
419: magnitudes (within $3\sigma$ of the best-fit $\Lambda$CDM model based on a
420: $\Delta\chi^2$ test, for a family of models with a sufficient number of parameters to
421: encompass both $\Lambda$ and the void). Voids which explain the observed Hubble
422: diagram have low density and large size, $\zV\sim 1$ (radii $\sim 2.5$~Gpc). However,
423: as shown in fig.~\ref{HBfig}, combining the SNe data with current limits on $u$
424: ($\chi^2=225,\,250$ for $191$ degrees of
425: freedom), we find that nearly all such voids are ruled out. These specific
426: constraints only apply to the Hubble bubble class of models, which also suffers from
427: other flaws not mentioned here. This test will be applied to more general and more
428: realistic void profiles.
429:
430: An improvement in the bound on $u$ by an order of magnitude may confirm or refute a
431: wider variety of such voids as an explanation of the dark energy phenomena. Yet, the
432: $u$-distortion test presented here is more general than the question of dark energy.
433: Future pursuit of this test will help improve our view of the universe on the largest
434: scales.
435:
436:
437: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
438: \begin{acknowledgments}
439:
440: We thank the Galileo Galilei Institute for Theoretical Physics for the hospitality
441: and the INFN for partial support during the completion of this work. R.C. was
442: supported in part by NSF AST-0349213 at Dartmouth and A.S. by the DoE at
443: Fermilab.
444: \vspace{-0.5cm}
445: \end{acknowledgments}
446:
447:
448:
449: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
450: \begin{thebibliography}{99}
451:
452: \bibitem{Perlmutter:1998np}
453: S.~Perlmutter {\it et al.} [Supernova Cosmology Project Collaboration],
454: Astrophys.\ J.\ {\bf 517}, 565 (1999).
455:
456: \bibitem{Riess:1998cb}
457: A.~Riess {\it et al.} [Supernova Search Team Collaboration],
458: Astron.\ J.\ {\bf 116}, 1009 (1998).
459:
460: %\bibitem{Hajian:2006ud}
461: \bibitem{CMBisotropy}
462: A.~Hajian and T.~Souradeep,
463: %``Testing Global Isotropy of Three-Year WMAP Date: Temperature Analysis,''
464: Phys.\ Rev.\ D {\bf 74}, 123521 (2006);
465: %\bibitem{Jaffe:2005pw}
466: T.~R.~Jaffe, {\it et al.},
467: %``Evidence of vorticity and shear at large angular scales in the WMAP
468: % data: A violation of cosmological isotropy?''
469: Astrophys.\ J.\ {\bf 629}, L1 (2005).
470:
471: \bibitem{Hogg:2004vw}
472: D.~W.~Hogg, {\it et al.},
473: %``Cosmic homogeneity demonstrated with luminous red galaxies,''
474: Astrophys.\ J.\ {\bf 624}, 54 (2005).
475:
476: \bibitem{RefGroupA}
477: %\bibitem{Celerier:1999hp}
478: M.~N.~Celerier,
479: %``Do we really see a Cosmological Constant in the Supernovae data ?,''
480: Astron.\ Astrophys.\ {\bf 353}, 63 (2000);
481: %\bibitem{Barrett:1999fd}
482: R.~K.~Barrett and C.~A.~Clarkson,
483: %``Undermining the Cosmological Principle: Almost Isotropic Observations in
484: %Inhomogeneous Cosmologies,''
485: Class.\ Quant.\ Grav.\ {\bf 17}, 5047 (2000);
486: %\bibitem{Tomita:2001a}
487: K.~Tomita,
488: %``A local void and the accelerating Universe,''
489: Mon.\ Not.\ R.A.S. {\bf 326}, 287 (2001);
490: %\bibitem{Tomita:2001b}
491: K.~Tomita,
492: %``Analyses of Type Ia Supernova Data
493: % in Cosmological Models with a Local Void,''
494: Prog.\ Theor.\ Phys. {\bf 106}, 929 (2001);
495: %\bibitem{Iguchi:2001sq}
496: H.~Iguchi, T.~Nakamura and K.~I.~Nakao,
497: %``Is dark energy the only solution to the apparent acceleration of the
498: %present universe?,''
499: Prog.\ Theor.\ Phys.\ {\bf 108}, 809 (2002);
500: %\bibitem{Moffat:2005ii}
501: J.~W.~Moffat,
502: %``Late-time inhomogeneity and acceleration without dark energy,''
503: JCAP {\bf 0605}, 001 (2006);
504: %\bibitem{Alnes:2005rw}
505: H.~Alnes, M.~Amarzguioui and O.~Gron,
506: %``An inhomogeneous alternative to dark energy?,''
507: Phys.\ Rev.\ D {\bf 73}, 083519 (2006);
508: %\bibitem{Mansouri:2005rf}
509: R.~Mansouri,
510: %``Structured FRW universe leads to acceleration: A non-perturbative
511: %approach,''
512: arXiv:astro-ph/0512605;
513: %\bibitem{Chuang:2005yi}
514: C.~H.~Chuang, J.~A.~Gu and W.~Y.~Hwang,
515: %``Inhomogeneity-Induced Cosmic Acceleration in a Dust Universe,''
516: arXiv:astro-ph/0512651;
517: %\bibitem{Vanderveld:2006rb}
518: R.~A.~Vanderveld, E.~E.~Flanagan and I.~Wasserman,
519: %``Mimicking Dark Energy with Lemaitre-Tolman-Bondi Models: Weak Central
520: %Singularities and Critical Points,''
521: Phys.\ Rev.\ D {\bf 74}, 023506 (2006);
522: %\bibitem{Garfinkle:2006sb}
523: D.~Garfinkle,
524: %``Inhomogeneous space-times as a dark energy model,''
525: Class.\ Quant.\ Grav.\ {\bf 23}, 4811 (2006);
526: %\bibitem{BiswasMansouriNotari2006}
527: R.~Mansouri,
528: %``Nonlinear Structure Formation and ÒApparentÓ Acceleration: an Investigation''
529: arXiv:astro-ph/0606703;
530: %\bibitem{Chung:2006xh}
531: D.~J.~H.~Chung and A.~E.~Romano,
532: %``Mapping Luminosity-Redshift Relationship to LTB Cosmology,''
533: Phys.\ Rev.\ D {\bf 74}, 103507 (2006);
534: %\bibitem{Enqvist:2006cg}
535: K.~Enqvist and T.~Mattsson,
536: %``The effect of inhomogeneous expansion on the supernova observations,''
537: JCAP {\bf 0702}, 019 (2007);
538: %\bibitem{Alnes:2006}
539: H.~Alnes and M.~Amarzguioui,
540: %``The supernova Hubble diagram for off-center observers in a ...,''
541: Phys.\ Rev.\ D {\bf 75}, 023506 (2007);
542: %\bibitem{Celerier:2007}
543: M.~N.~Celerier,
544: %``The accelerated expansion of the Universe challenged by an effect of
545: % the inhomogeneities. A review,''
546: [arXiv:astro-ph/0702416].
547:
548: \bibitem{Bondi:1947av}
549: H.~Bondi,
550: %``Spherically symmetrical models in general relativity,''
551: Mon.\ Not.\ Roy.\ Astron.\ Soc.\ {\bf 107}, 410 (1947).
552:
553: \bibitem{Stebbins2007}
554: A.~Stebbins,
555: %``CMB Spectral Distortions from the Scattering of Temperature Anisotropies,''
556: arXiv:astro-ph/0703541.
557:
558: \bibitem{Chluba:2003}
559: J.~Chluba and R.~A.~Sunyaev,
560: Astron.\ Astrophys.\ {\bf 424}, 389 (2003).
561:
562: \bibitem{Freedman:2000cf}
563: W.~L.~Freedman {\it et al.},
564: %``Final Results from the Hubble Space Telescope Key Project to Measure the
565: %Hubble Constant,''
566: Astrophys.\ J.\ {\bf 553}, 47 (2001).
567:
568: \bibitem{Riess:2005zi}
569: A.~G.~Riess {\it et al.},
570: %``Cepheid Calibrations from the Hubble Space Telescope of the Luminosity of
571: %Two Recent Type Ia Supernovae and a Re-determination of the Hubble
572: %Constant,''
573: Astrophys.\ J.\ {\bf 627}, 579 (2005).
574:
575: \bibitem{SpergelBean2006}
576: D.N.~Spergel, R.~Bean, {\it et al.},
577: %``Wilkinson Microwave Anisotropy Probe (WMAP) Three Year Results:
578: % Implications for Cosmology,''
579: Astrophys.\ J.\ Suppl.\ {\bf 170}, 377 (2007).
580:
581: \bibitem{DyerRoeder}
582: C.~Dyer and R.~Roeder,
583: % "Distance-Redshift Relations for Universes with Some Intergalactic Medium"
584: Astrophys.~J. {\bf 180} 31 (1973).
585:
586: \bibitem{Mather:1994}
587: J.~C.~Mather, {\it et al.},
588: %``Measurement of the cosmic microwave background spectrum by the
589: % COBE FIRAS instrument,''
590: Astrophys.\ J.\ {\bf 420}, 439 (1994).
591:
592: \bibitem{BoggessMatherEtAl1992}
593: N.W.~Boggess, J.C.~Mather, R.~Weiss {\it et al.},
594: %``The COBE Mission: Its Design and Performance Two Years After Launch"
595: Astrophys.\ J.\ {\bf 397}, 420 (1992).
596:
597: \bibitem{FixsenCheng1996}
598: D.~Fixsen, E.~Cheng, {\it et al.},
599: %``The Cosmic Microwave Background Spectrum from the Full COBE
600: % FIRAS Data Set,''
601: Astrophys.\ J.\ {\bf 473}, 576 (1996).
602:
603: \bibitem{Fixsen:2002}
604: D.~J.~Fixsen and J.~C.~Mather,
605: Astrophys.\ J.\ {\bf 581}, 817 (2002).
606:
607: \bibitem{Kogut:2006kf}
608: A.~Kogut {\it et al.},
609: %``ARCADE: Absolute Radiometer for Cosmology, Astrophysics, and Diffuse
610: %Emission,''
611: New Astron.\ Rev.\ {\bf 50}, 925 (2006).
612:
613: \bibitem{ZhangPenTrac}
614: P.~Zhang, U.~Pen and H.~Trac,
615: % "The Temperature of the intergalactic medium and the Compton y parameter"
616: Mon.\ Not.\ Roy.\ Astron.\ Soc.\ {\bf 355} 451 (2004).
617:
618: \bibitem{Jha:2006fm}
619: S.~Jha, A.~G.~Riess and R.~P.~Kirshner,
620: %``Improved Distances to Type Ia Supernovae with Multicolor Light Curve
621: %Shapes: MLCS2k2,''
622: Astrophys.~J. {\bf 659}, 122 (2007).
623:
624: \bibitem{Riess:2006fw}
625: A.~G.~Riess {\it et al.},
626: %``New Hubble Space Telescope Discoveries of Type Ia Supernovae at $z>1$:
627: %Narrowing Constraints on the Early Behavior of Dark Energy,''
628: Astrophys.~J. {\bf 659} 98 (2007).
629:
630: \bibitem{WoodVasey:2007jb}
631: W.~M.~Wood-Vasey {\it et al.},
632: %``Observational Constraints on the Nature of the Dark Energy: First
633: %Cosmological Results from the ESSENCE Supernova Survey,''
634: Astrophys.~J. {\bf 666} 694 (2007).
635:
636: \bibitem{Davis:2007na}
637: T.~M.~Davis {\it et al.},
638: %``Scrutinizing exotic cosmological models using (ESSENCE) supernova data
639: %combined with other cosmological probes,''
640: Astrophys.~J. {\bf 666} 716 (2007).
641:
642:
643: \end{thebibliography}
644:
645: \vfill
646: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
647:
648:
649: \end{document}
650:
651:
652:
653: