1: \chapter{Dark Energy} \label{deintro}
2:
3:
4: In this chapter we will introduce the Dark-Energy problem, at first its experimental evidences and then the possible explanations in which we are interested.
5: It will help us to introduce the Back-Reaction problem in the next chapter.
6:
7: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
8: \section{Evidences}
9: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
10:
11: The fact that matter energy density does not dominate the universe has been one of the most stunning discoveries in the last few years in cosmology.
12:
13:
14: Most, if not all, observations are consistent with the cosmic concordance model
15: according to which, today, one-fourth of the mass-energy of the universe is
16: clustered and dominated by cold dark matter. The remaining three-quarters is
17: uniform and dominated by a fluid with a negative pressure which we call dark energy.
18:
19: The three main experimental evidences for dark energy are summarized in Fig.~\ref{verde03}.\\
20: The blue area is about the results of the 2dF galaxy survey which gives $\Omega_{M} \simeq 0.3$.
21: This measure is independent of $\Omega_{DE}$ and so it is vertical\footnote{The blue area is not actually vertical, but slightly tilted. This because the 2dF galaxy survey is mainly at $z=0.14$. In order to evolve the data till $z=0$ a cosmological model (in particular a $\Lambda$CDM model) has been used and therefore a dependence on dark energy introduced.} in the plane $\Omega_{M}$-$\Omega_{DE}$.\\
22: The orange area represents the constraints from the WMAP observations of the CMB anisotropies.
23: They give $1 \simeq \Omega_{TOT}=\Omega_{M}+\Omega_{DE}$ and so the orange area is aligned to the line for $(0,1)$ and $(1,0)$.
24: \newpage
25: Finally, the green area shows the Supernovae Ia measurements about the deceleration parameter which, in the $\Lambda$CDM model, is given by $q= \Omega_{M}/2-\Omega_{DE}$. Therefore the green area is almost perpendicular to the orange one.
26:
27:
28: The fact that these constrains cross perpendicularly and consistently gave the name {\it cosmic concordance model} to a $\Lambda$CDM with:
29: %
30: \begin{eqnarray}
31: \Omega_{M} &\simeq& 0.25 \cr
32: \Omega_{DE} &\simeq& 0.75 \cr
33: w_{DE} &\simeq& -1
34: \end{eqnarray}
35: %
36: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
37: \begin{figure}[htbp]
38: \begin{center}
39: \includegraphics[width=14cm]{verde03.eps}
40: \end{center}
41: \caption{\small \slshape Experimental constraints on the plane $\Omega_{M}$-$\Omega_{DE}$ from the 2dF galaxy survey (blue), CMB (orange) and Supernovae Ia (green).}
42: \label{verde03}
43: \end{figure}
44: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
45:
46:
47: \clearpage
48: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
49: \section{Candidates}
50: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
51:
52: In this section we will take into account three possible explanations for dark energy: the cosmological constant, a cosmological scalar field and inhomogeneous universe models.
53: In this thesis we are really interested only in the last possibility.
54: We will describe the first two because, beside being interesting in themselves, they will help us understand the problem and will turn out to be useful to describe inhomogeneous models effectively.
55: They are indeed based on the cosmological principle of homogeneity and isotropy.
56:
57: For a more comprehensive examination of possible approaches and different points of view to dark energy see \cite{Nobbenhuis:2006yf}.
58:
59: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
60: \subsection{The Cosmological Constant}
61:
62: Trying to explain dark energy through the cosmological constant $\Lambda$, that is by using the concordance model, has the advantage of simplicity and agreement with experimental data.
63:
64: The cosmological constant appears as a free parameter in the Einstein's equations:
65: %
66: \begin{equation}
67: R_{\mu \nu}-{1 \over 2} \, g_{\mu \nu} \, R -\Lambda \, g_{\mu \nu}= 8 \pi \, T_{\mu \nu}
68: \end{equation}
69: %
70: where we are using the signature $(-,+,+,+)$ and geometric units, $c=1=G$.
71:
72: The freedom about the value of $\Lambda$ is cause of problems.
73: In spite of being treated as the single problem of the value of $\Lambda$, from a conceptual point of view they are actually three distinct problems\footnote{For a similar approach see \cite{Quartin:2008px}.}.
74: This distinction will help us understand which ones are more pertinent to the cosmological problem and which are actually the advantages in using alternative explanations for dark energy.
75:
76: For a wider overview of cosmology in the presence of the cosmological constant we refer to \cite{weinberg-lambda, carroll:lambda0, Carroll:2000fy, Peebles:2002gy, Padmanabhan:2002ji}.
77:
78:
79: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
80: \begin{figure}[htbp]
81: \begin{center}
82: \includegraphics[width=12cm]{cosmoconst.eps}
83: \end{center}
84: \caption{\small \slshape Evolution of the energy density of radiation (short-dashed line), matter (long-dashed line) and cosmological constant (solid line) with respect to the redshift. All energy densities are expressed in units of the present critical energy density $\rho_{C 0}$.}
85: \label{cosmoconst}
86: \end{figure}
87: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
88:
89: %%%%%%%%%%%%%%%%
90: \subsubsection{Initial conditions problem}
91: The first problem is about the initial conditions that we have to give to the concordance model: it would be expectable to have comparable values of the initial energy densities. However, as you can see from Fig.~\ref{cosmoconst}, matter and radiation initial energy-density values are much bigger than the cosmological constant one. The concordance model implies therefore a sizeable fine tuning of the initial conditions.
92:
93: %%%%%%%%%%%%%%%%
94: \subsubsection{Coincidence problem}
95: The second issue is about the values of the present-day energy density of matter and cosmological constant. As you can see from Fig.~\ref{cosmoconst}, they evolve differently and therefore generally, during the evolution of the universe, we will expect them to have different values. However their values happen to be of the same order at present time. This fact needs a sizeable fine tuning to happen.
96:
97: %%%%%%%%%%%%%%%%
98: \subsubsection{Origin problem}
99: Finally, problems show up if we ask where the cosmological constant comes from.\\
100: Classically, it is a free parameter given as an initial condition, $\rho_{\Lambda}$.\\
101: Quantum mechanically, we have the contribution from the ground state of the energy-momentum tensor which does not need to be zero:
102: %
103: \begin{equation}
104: \langle 0 |T_{\mu \nu}|0 \rangle \equiv T_{\mu \nu}^{GS} = \rho_{GS} \, g_{\mu \nu}
105: \end{equation}
106: %
107: $\rho_{GS}$ has the same equation of state of $\rho_{\Lambda}$ and therefore we can combine them into the vacuum energy density $\rho_{V}$:
108: %
109: \begin{equation} \label{vacu}
110: \rho_{V}=\rho_{\Lambda}+\rho_{GS}
111: \end{equation}
112: %
113: The problem is that any quantum-mechanic estimation of $\rho_{V}$ largely exceeds the experimental value of
114: $\rho_{DE} \sim (10^{-3} \mbox{ eV})^{4}$. Typical estimations are around $\rho_{V} \sim (10^{27} \mbox{ eV})^{4}$, that is, $30$ orders of magnitude bigger.
115:
116:
117: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
118: \subsection{Quintessence models} \label{quiquoqua}
119:
120: Quintessence or cosmological scalar-field models will be, as effective models, a useful tool in our analysis of a inhomogeneous universe.
121:
122: The idea is to let the cosmological constant be a dynamical quantity. The Lagrangian of a quintessence field is, therefore:
123: %
124: \begin{equation} \label{lphi}
125: {\cal L}_{\phi} = \frac{1}{2} \partial^{\mu}\phi \, \partial_{\mu}\phi - V(\phi)
126: \end{equation}
127: %
128: from which, assuming spatial homogeneity, we obtain:
129: %
130: \begin{equation} \label{rho-p}
131: \rho_{\phi}={1 \over 2} \dot{\phi}^{2} + V(\phi)
132: \quad \mbox{ e } \quad
133: p_{\phi}={1 \over 2} \dot{\phi}^{2} - V(\phi)
134: \end{equation}
135: %
136: The equation of state is:
137: %
138: \begin{equation} \label{wphi}
139: w_{\phi}={p_{\phi} \over \rho_{\phi}}={\dot{\phi}^{2}/2 - V(\phi) \over \dot{\phi}^{2} /2 + V(\phi)}
140: \end{equation}
141: %
142: Quintessence models are extensively studied (see for example \cite{Copeland:2006wr}) mainly because of their flexibility. In particular there exist classes of potentials which exhibit attractor solutions independent of initial conditions \cite{Steinhardt:1999nw}.
143:
144: For illustrative purposes, in Fig.~\ref{quinti2} we show the evolution of the energy density and of the equation of state for a scalar potential $V=M^{5} \phi^{-1}$ and initial conditions $\rho_{\phi}^{in}/\rho_{C 0}=10^{30}$ at $z=10^{10}$. $M$ gives the energy scale of the potential and has to be fine-tuned in order to have the right present-day quintessence energy density.
145:
146: As you can see from Fig.~\ref{quinti2}, at early times the kinetic energy is dominant over the potential energy: during this period, called kination and characterized by $w_{\phi}\simeq 1$, the scalar field rapidly rolls down along the potential. Then the friction performed by the Hubble term\footnote{The continuity equation for a quintessence field can be written as $\dot{\rho}_{\phi}=-3 H \dot{\phi}^{2}$.} slows down the run of the scalar till its freeze: during this period, called slow-roll and characterized by $w_{\phi}\simeq -1$, the potential energy dominates and the scalar mimics a cosmological constant. Finally the scalar reaches the attractor and follows a dynamics characterized by its features. In the illustrative case shown, the attractor is characterized by $w_{\phi}\simeq -0.67$. See Appendix~\ref{alphaevo} for more details.
147:
148: %%%%%%%%%%%%%%%%%%
149: \begin{figure}[!htbp]
150: \begin{flushright}
151: \includegraphics[width=16cm]{quinti.eps}
152: \caption{\small \slshape Evolution of the energy densities (left) and scalar equation of state (right) for a quintessence model with potential $V=M^{5} \phi^{-1}$ and initial conditions $\rho_{\phi}^{in}/\rho_{C 0}=10^{30}$ at $z=10^{10}$.
153: The dot-dashed line represents the energy density of radiation, the dotted line the energy density of matter, the green dashed line the energy density of quintessence and the red solid line the attractor. All of the energy densities are expressed in units of the present critical energy density $\rho_{C 0}$. From \cite{Marra:2005yt}.}
154: \label{quinti2}
155: \end{flushright}
156: \end{figure}
157: %%%%%%%%%%%%%%%%%%%
158:
159: %%%%%%%%%%%%%%%%
160: \subsubsection{Pros}
161: A first advantage of the example shown is that it solves the initial conditions problem: we can change of many order of magnitude the chosen value of $\rho_{\phi}^{in}/\rho_{C 0}=10^{30}$ at $z=10^{10}$ and still the late time dynamics will be the same: only the moment at which the scalar reaches the attractor will change. This is a real improvement in comparison with the cosmological constant.
162:
163: In the second place, a dynamical scalar field may have chances to solve the coincidence problem. Indeed, if it couples to the matter energy density, it could be possible to explain why, today, they are of the same order of magnitude.
164: However, quintessence models generally feature a sizeable fine-tuning about the mass parameter of the potential, similarly to what happens with the cosmological constant.
165:
166: %%%%%%%%%%%%%%%%
167: \subsubsection{Cons}
168: Beside the problem of the origin of quintessence, its first drawback is about its mass which has to be ultra-light:
169: %
170: \begin{equation} \label{mphi}
171: m_{\phi_{0}} \sim H_{0} \sim 10^{-33} \mbox{ eV}
172: \end{equation}
173: %
174: Quantum-mechanics direct or indirect corrections will easily bring $m_{\phi_{0}}$ toward much bigger values.
175:
176: Other problems come from the possible coupling of the quintessence field with the other terms in the matter-radiation Lagrangian and from the fact that such an ultra-light field has long-range interactions: $\lambda \sim m_{\phi_{0}}^{-1} \sim H_{0}^{-1}$. Variations of fundamental constants and violations of the equivalence principle are therefore expected. For more details see Appendix~\ref{alphaevo}.
177:
178: %%%%%%%%%%%%%%%%
179: \subsubsection{Effective theory?}
180: In the case quintessence models are used as effective models, none of the drawback examined above will be applicable.
181: They indeed are about the quantum mechanics implications a primary quintessence field brings to light.
182:
183: As for this thesis, the most appealing feature of quintessence is its dynamical nature which can be linked to the evolution of the universe. In particular, a coupling of an effective quintessence field to inhomogeneities could solve the coincidence problem.
184: We will discuss this point in the next chapter.
185:
186: \clearpage
187: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%
188: \subsection{Inhomogeneous models - a point of view} \label{poview}
189:
190: This overview of advantages and drawbacks of $\Lambda$CDM and quintessence models gives us some hints to better set the dark energy problem. We will present now the point of view that will characterize this thesis.
191: It is clarifying to write:
192: %
193: \begin{equation} \label{core}
194: \rho_{V} \equiv (\rho_{\Lambda} \; \, +) \; \, \rho_{GS} \ggg \rho_{DE} \sim \rho_{M}
195: \end{equation}
196: %
197: On the left-hand side, the two terms $\rho_{\Lambda}$ and $\rho_{GS}$ are really different and could be inappropriate to mix them. The latter comes from quantum mechanics, the former comes from general relativity. $\rho_{\Lambda}$ is likely connected to cosmology and therefore to $\rho_{DE}$ while for $\rho_{GS}$ it could not be the case. Indeed typical estimations for $\rho_{GS}$ are far away from the cosmological value.\\
198: Therefore it could be possible that we should be only concerned about $\rho_{\Lambda}$ the role of which has to be understood in order to have a satisfactory general-relativist cosmological model.
199:
200: On the right-hand side, the similarity in density values of the dark energy and matter components suggests a direct or indirect connection: the dark energy problem might be a cosmological one.
201:
202: If we accept this reasoning and remember that the homogeneous $\Lambda$CDM model is a good fit to the real inhomogeneous universe, we could think that the connection is by means of the smoothing of the inhomogeneities: {\it the back-reaction of inhomogeneities makes an inhomogeneous matter FRW model appear as a $\Lambda$CDM model.}\\
203: One immediate benefit is to turn the coincidence problem into a hint in favor of the back-reaction: only recently have the cosmic structures evolved enough to have a sufficiently strong back-reaction of inhomogeneities which could affect observables.
204:
205: The issue now is what we mean by back-reaction of inhomogeneities.
206: There are broadly speaking two distinct approaches as we will now discuss.
207: One is focused on theoretical aspects while the other on observations.\\
208: Both the approaches will, however, share the idea of smoothing out inhomogeneities.
209: The duality in the interpretation of this concept will characterize this thesis.
210:
211: \clearpage
212: %%%%%%%%%%%%%%%%
213: \subsubsection{Observational side - $\rho_{DE}$} \label{dake}
214:
215: The ``safe'' consequence of the success of the concordance model is that the isotropic and homogeneous $\Lambda$CDM model is a good {\it observational} fit to the real inhomogeneous universe.
216: And this is, in some sense, a verification of the cosmological principle:
217: the inhomogeneous universe can be described by means of an isotropic and
218: homogeneous solution.
219:
220: However this does not imply that a primary source of dark energy
221: exists, but only that it exists effectively as far as the
222: observational fit is concerned.
223: For example it is not straightforward that the universe, as Supernovae Ia measurements seem to tell us, is globally
224: accelerating.
225: If dark energy does not exist at a fundamental level,
226: its evidence coming from the
227: concordance model would tell us that the purely-matter inhomogeneous
228: model has been renormalized, from the observational point of view
229: (luminosity and redshift of photons), into a homogeneous $\Lambda$CDM model.
230: Moreover, the very homogeneous nature of dark energy seems a clue about its effective nature.
231:
232: There could indeed be the possibility that there are large effects on the {\it observed} expansion rate due to the back-reaction of inhomogeneities in the universe (see, e.g., Ref.~\cite{Kolb:2005da} and refs. therein).
233: The basic idea is that all evidence for dark energy comes from observational determination of the expansion history of the universe.
234: Anything that affects the observed expansion history of the universe alters the
235: determination of the parameters of dark energy; in the extreme it may remove
236: the need for dark energy.\\
237:
238: Summarizing, this approach is tied to our past light cone: it will focus on the effects
239: of large-scale non-linear inhomogeneities on observables such as the
240: luminosity-distance--redshift relation.
241: Even though there is no explicit averaging in here, the $d_{L}(z)$ is a about the luminosity and redshift of photons that travelled through inhomogeneities: their effects are therefore ``integrated out'', averaged in an implicit way.
242:
243: \clearpage
244: %%%%%%%%%%%%%%%%
245: \subsubsection{Theoretical side - $\rho_{\Lambda}$} \label{lambda}
246:
247: We said that $\rho_{\Lambda}$ is likely connected to cosmology and that, in any case, we have to understand its meaning in order to have a satisfactory general-relativistic cosmological model.
248:
249: It could be that $\rho_{\Lambda}$ is not directly related to what we measure as $\rho_{DE}$.
250: The latter comes from the fitting of the real inhomogeneous universe by means of a $\Lambda$CDM, while the former is about a free parameter of Einstein's equations.
251:
252: As we will discuss in the next chapter, $\rho_{\Lambda}$ could be the result of a smoothing process \cite{ellis-1984, Buchert:2007ik}. This is an important issue in general relativity because, generally, every measurement involves some form of smoothing or averaging.
253:
254: Following this point of view, $\rho_{\Lambda}$ could be a product of the back-reaction of inhomogeneities and have its freedom fixed by a measuring process. Inhomogeneities introduce a scale in the otherwise scale-free general relativity.
255: We think this is a crucial step in understanding how General Relativity effectively works in a lumpy universe.\\
256:
257: As we will see, the issue is to understand how to carry out this smoothing meaningfully.
258: In particular if it is possible to connect $\rho_{\Lambda}$ to $\rho_{DE}$.
259: Within the previous approach we were only concern about the luminosity-distance--relation, while here we want, in addition to that, a phenomenological model that fits observations, in other words, we want a description by means of a mean field.
260:
261:
262:
263:
264:
265:
266:
267:
268:
269:
270:
271:
272:
273:
274: