This action might not be possible to undo. Are you sure you want to continue?
r
X
i
v
:
h
e
p

p
h
/
0
0
0
4
1
8
8
v
1
1
9
A
p
r
2
0
0
0
ASTROPHYSICS AND COSMOLOGY
Juan Garc´ıaBellido
Theoretical Physics Group, Blackett Laboratory, Imperial College of Science,
Technology and Medicine, Prince Consort Road, London SW7 2BZ, U.K.
Abstract
These notes are intended as an introductory course for experimental particle
physicists interested in the recent developments in astrophysics and cosmo
logy. I will describe the standard Big Bang theory of the evolution of the
universe, with its successes and shortcomings, which will lead to inﬂation
ary cosmology as the paradigm for the origin of the global structure of the
universe as well as the origin of the spectrum of density perturbations respon
sible for structure in our local patch. I will present a review of the very rich
phenomenology that we have in cosmology today, as well as evidence for the
observational revolution that this ﬁeld is going through, which will provide us,
in the next few years, with an accurate determination of the parameters of our
standard cosmological model.
1 GENERAL INTRODUCTION
Cosmology (from the Greek: kosmos, universe, world, order, and logos, word, theory) is probably the
most ancient body of knowledge, dating from as far back as the predictions of seasons by early civiliza
tions. Yet, until recently, we could only answer to some of its more basic questions with an order of
magnitude estimate. This poor state of affairs has dramatically changed in the last few years, thanks to
(what else?) raw data, coming from precise measurements of a wide range of cosmological parameters.
Furthermore, we are entering a precision era in cosmology, and soon most of our observables will be
measured with a few percent accuracy. We are truly living in the Golden Age of Cosmology. It is a very
exciting time and I will try to communicate this enthusiasm to you.
Important results are coming out almost every month from a large set of experiments, which pro
vide crucial information about the universe origin and evolution; so rapidly that these notes will probably
be outdated before they are in print as a CERN report. In fact, some of the results I mentioned during
the Summer School have already been improved, specially in the area of the microwave background
anisotropies. Nevertheless, most of the new data can be interpreted within a coherent framework known
as the standard cosmological model, based on the Big Bang theory of the universe and the inﬂationary
paradigm, which is with us for two decades. I will try to make such a theoretical model accesible to
young experimental particle physicists with little or no previous knowledge about general relativity and
curved spacetime, but with some knowledge of quantum ﬁeld theory and the standard model of particle
physics.
2 INTRODUCTION TO BIG BANG COSMOLOGY
Our present understanding of the universe is based upon the successful hot Big Bang theory, which
explains its evolution from the ﬁrst fraction of a second to our present age, around 13 billion years
later. This theory rests upon four strong pillars, a theoretical framework based on general relativity,
as put forward by Albert Einstein [1] and Alexander A. Friedmann [2] in the 1920s, and three robust
observational facts: First, the expansion of the universe, discovered by Edwin P. Hubble [3] in the 1930s,
as a recession of galaxies at a speed proportional to their distance fromus. Second, the relative abundance
of light elements, explained by George Gamow [4] in the 1940s, mainly that of helium, deuterium and
lithium, which were cooked fromthe nuclear reactions that took place at around a second to a few minutes
after the Big Bang, when the universe was a few times hotter than the core of the sun. Third, the cosmic
microwave background (CMB), the afterglow of the Big Bang, discovered in 1965 by Arno A. Penzias
and Robert W. Wilson [5] as a very isotropic blackbody radiation at a temperature of about 3 degrees
Kelvin, emitted when the universe was cold enough to form neutral atoms, and photons decoupled from
matter, approximately 500,000 years after the Big Bang. Today, these observations are conﬁrmed to
within a few percent accuracy, and have helped establish the hot Big Bang as the preferred model of the
universe.
2.1 Friedmann–Robertson–Walker universes
Where are we in the universe? During our lectures, of course, we were in
ˇ
Casta Papierniˇ cka, in “the heart
of Europe”, on planet Earth, rotating (8 lightminutes away) around the Sun, an ordinary star 8.5 kpc
1
from the center of our galaxy, the Milky Way, which is part of the local group, within the Virgo cluster
of galaxies (of size a few Mpc), itself part of a supercluster (of size ∼ 100 Mpc), within the visible
universe (∼ few 1000 Mpc), most probably a tiny homogeneous patch of the inﬁnite global structure
of spacetime, much beyond our observable universe.
Cosmology studies the universe as we see it. Due to our inherent inability to experiment with it,
its origin and evolution has always been prone to wild speculation. However, cosmology was born as a
science with the advent of general relativity and the realization that the geometry of spacetime, and thus
the general attraction of matter, is determined by the energy content of the universe [6],
G
µν
≡ R
µν
−
1
2
g
µν
R = 8πGT
µν
+ Λg
µν
. (1)
These nonlinear equations are simply too difﬁcult to solve without some insight coming from the sym
metries of the problem at hand: the universe itself. At the time (19171922) the known (observed)
universe extended a few hundreds of parsecs away, to the galaxies in the local group, Andromeda and
the Large and Small Magellanic Clouds: The universe looked extremely anisotropic. Nevertheless, both
Einstein and Friedmann speculated that the most “reasonable” symmetry for the universe at large should
be homogeneity at all points, and thus isotropy. It was not until the detection, a few decades later, of the
microwave background by Penzias and Wilson that this important assumption was ﬁnally put onto ﬁrm
experimental ground. So, what is the most general metric satisfying homogeneity and isotropy at large
scales? The FriedmannRobertsonWalker (FRW) metric, written here in terms of the invariant geodesic
distance ds
2
= g
µν
dx
µ
dx
ν
in four dimensions, µ = 0, 1, 2, 3, see Ref. [6],
2
ds
2
= dt
2
−a
2
(t)
_
dr
2
1 − K r
2
+r
2
(dθ
2
+ sin
2
θ dφ
2
)
_
, (2)
characterized by just two quantities, a scale factor a(t), which determines the physical size of the uni
verse, and a constant K, which characterizes the spatial curvature of the universe,
(3)
R =
6K
a
2
(t)
.
_
¸
_
¸
_
K = −1 OPEN
K = 0 FLAT
K = +1 CLOSED
(3)
Spatially open, ﬂat and closed universes have different geometries. Light geodesics on these universes
behave differently, and thus could in principle be distinguished observationally, as we shall discuss later.
Apart from the threedimensional spatial curvature, we can also compute a fourdimensional spacetime
curvature,
(4)
R = 6
¨ a
a
+ 6
_
˙ a
a
_
2
+ 6
K
a
2
. (4)
1
One parallax second (1 pc), parsec for short, corresponds to a distance of about 3.26 lightyears or 3 × 10
18
cm.
2
I am using c = 1 everywhere, unless speciﬁed.
Depending on the dynamics (and thus on the matter/energy content) of the universe, we will have dif
ferent possible outcomes of its evolution. The universe may expand for ever, recollapse in the future or
approach an asymptotic state in between.
2.1.1 The expansion of the universe
In 1929, Edwin P. Hubble observed a redshift in the spectra of distant galaxies, which indicated that they
were receding from us at a velocity proportional to their distance to us [3]. This was correctly interpreted
as mainly due to the expansion of the universe, that is, to the fact that the scale factor today is larger
than when the photons were emitted by the observed galaxies. For simplicity, consider the metric of a
spatially ﬂat universe, ds
2
= dt
2
− a
2
(t) dx
2
(the generalization of the following argument to curved
space is straightforward). The scale factor a(t) gives physical size to the spatial coordinates x, and the
expansion is nothing but a change of scale (of spatial units) with time. Except for peculiar velocities, i.e.
motion due to the local attraction of matter, galaxies do not move in coordinate space, it is the spacetime
fabric which is stretching between galaxies. Due to this continuous stretching, the observed wavelength
of photons coming from distant objects is greater than when they were emitted by a factor precisely equal
to the ratio of scale factors,
λ
obs
λ
em
=
a
0
a
≡ 1 + z , (5)
where a
0
is the present value of the scale factor. Since the universe today is larger than in the past, the
observed wavelengths will be shifted towards the red, or redshifted, by an amount characterized by z, the
redshift parameter.
In the context of a FRW metric, the universe expansion is characterized by a quantity known as
the Hubble rate of expansion, H(t) = ˙ a(t)/a(t), whose value today is denoted by H
0
. As I shall deduce
later, it is possible to compute the relation between the physical distance d
L
and the present rate of
expansion, in terms of the redshift parameter,
3
H
0
d
L
= z +
1
2
(1 −q
0
) z
2
+O(z
3
) . (6)
At small distances from us, i.e. at z ¸1, we can safely keep only the linear term, and thus the recession
velocity becomes proportional to the distance from us, v = c z = H
0
d
L
, the proportionality constant
being the Hubble rate, H
0
. This expression constitutes the socalled Hubble law, and is spectacularly
conﬁrmed by a huge range of data, up to distances of hundreds of megaparsecs. In fact, only recently
measurements from very bright and distant supernovae, at z · 1, were obtained, and are beginning to
probe the secondorder term, proportional to the deceleration parameter q
0
, see Eq. (22). I will come
back to these measurements in Section 3.
One may be puzzled as to why do we see such a stretching of spacetime. Indeed, if all spatial
distances are scaled with a universal scale factor, our local measuring units (our rulers) should also be
stretched, and therefore we should not see the difference when comparing the two distances (e.g. the two
wavelengths) at different times. The reason we see the difference is because we live in a gravitationally
bound system, decoupled from the expansion of the universe: local spatial units in these systems are not
stretched by the expansion.
4
The wavelengths of photons are stretched along their geodesic path from
one galaxy to another. In this consistent world picture, galaxies are like point particles, moving as a ﬂuid
in an expanding universe.
2.1.2 The matter and energy content of the universe
So far I have only discussed the geometrical aspects of spacetime. Let us now consider the matter
and energy content of such a universe. The most general matter ﬂuid consistent with the assumption of
3
The subscript L refers to Luminosity, which characterizes the amount of light emitted by an object. See Eq. (61).
4
The local spacetime of a gravitationally bound system is described by the Schwarzschild metric, which is static [6].
homogeneity and isotropy is a perfect ﬂuid, one in which an observer comoving with the ﬂuid would see
the universe around it as isotropic. The energy momentum tensor associated with such a ﬂuid can be
written as [6]
T
µν
= p g
µν
+ (p +ρ) U
µ
U
ν
, (7)
where p(t) and ρ(t) are the pressure and energy density of the ﬂuid at a given time in the expansion, and
U
µ
is the comoving fourvelocity, satisfying U
µ
U
µ
= −1.
Let us now write the equations of motion of such a ﬂuid in an expanding universe. According to
general relativity, these equations can be deduced from the Einstein equations (1), where we substitute
the FRW metric (2) and the perfect ﬂuid tensor (7). The µ = ν = 0 component of the Einstein equations
constitutes the socalled Friedmann equation
H
2
=
_
˙ a
a
_
2
=
8πG
3
ρ +
Λ
3
−
K
a
2
, (8)
where I have treated the cosmological constant Λ as a different component from matter. In fact, it can
be associated with the vacuum energy of quantum ﬁeld theory, although we still do not understand why
should it have such a small value (120 orders of magnitude below that predicted by quantum theory),
if it is nonzero. This constitutes today one of the most fundamental problems of physics, let alone
cosmology.
The conservation of energy (T
µν
;ν
= 0), a direct consequence of the general covariance of the
theory (G
µν
;ν
= 0), can be written in terms of the FRW metric and the perfect ﬂuid tensor (7) as
d
dt
_
ρ a
3
_
+p
d
dt
_
a
3
_
= 0 , (9)
where the energy density and pressure can be split into its matter and radiation components, ρ = ρ
M
+ρ
R
,
p = p
M
+ p
R
, with corresponding equations of state, p
M
= 0, p
R
= ρ
R
/3. Together, the Friedmann
and the energyconservation equation give the evolution equation for the scale factor,
¨a
a
= −
4πG
3
(ρ + 3p) +
Λ
3
, (10)
I will now make a few useful deﬁnitions. We can write the Hubble parameter today H
0
in units of
100 kms
−1
Mpc
−1
, in terms of which one can estimate the order of magnitude for the present size and
age of the universe,
H
0
= 100 h kms
−1
Mpc
−1
, (11)
c H
−1
0
= 3000 h
−1
Mpc , (12)
H
−1
0
= 9.773 h
−1
Gyr . (13)
The parameter h has been measured to be in the range 0.4 < h < 1 for decades, and only in the last few
years has it been found to lie within 10% of h = 0.65. I will discuss those recent measurements in the
next Section.
One can also deﬁne a critical density ρ
c
, that which in the absence of a cosmological constant
would correspond to a ﬂat universe,
ρ
c
≡
3H
2
0
8πG
= 1.88 h
2
10
−29
g/cm
3
(14)
= 2.77 h
−1
10
11
M
/(h
−1
Mpc)
3
, (15)
where M
= 1.989 10
33
g is a solar mass unit. The critical density ρ
c
corresponds to approximately
4 protons per cubic meter, certainly a very dilute ﬂuid! In terms of the critical density it is possible to
deﬁne the ratios Ω
i
≡ ρ
i
/ρ
c
, for matter, radiation, cosmological constant and even curvature, today,
Ω
M
=
8πGρ
M
3H
2
0
Ω
R
=
8πGρ
R
3H
2
0
(16)
Ω
Λ
=
Λ
3H
2
0
Ω
K
= −
K
a
2
0
H
2
0
. (17)
We can evaluate today the radiation component Ω
R
, corresponding to relativistic particles, from
the density of microwave background photons, ρ
CMB
=
π
2
15
(kT
CMB
)
4
/(¯ hc)
3
= 4.510
−34
g/cm
3
, which
gives Ω
CMB
= 2.4 10
−5
h
−2
. Three massless neutrinos contribute an even smaller amount. Therefore,
we can safely neglect the contribution of relativistic particles to the total density of the universe today,
which is dominated either by nonrelativistic particles (baryons, dark matter or massive neutrinos) or by
a cosmological constant, and write the rate of expansion H
2
in terms of its value today,
H
2
(a) = H
2
0
_
Ω
R
a
4
0
a
4
+ Ω
M
a
3
0
a
3
+ Ω
Λ
+ Ω
K
a
2
0
a
2
_
. (18)
An interesting consequence of these redeﬁnitions is that I can now write the Friedmann equation today,
a = a
0
, as a cosmic sum rule,
1 = Ω
M
+ Ω
Λ
+ Ω
K
, (19)
where we have neglected Ω
R
today. That is, in the context of a FRW universe, the total fraction of
matter density, cosmological constant and spatial curvature today must add up to one. For instance, if
we measure one of the three components, say the spatial curvature, we can deduce the sum of the other
two. Making use of the cosmic sum rule today, we can write the matter and cosmological constant as a
function of the scale factor (a
0
≡ 1)
Ω
M
(a) =
8πGρ
M
3H
2
(a)
=
Ω
M
a + Ω
M
(1 −a) + Ω
Λ
(a
3
− a)
_
a→0
−→ 1
a→∞
−→ 0
, (20)
Ω
Λ
(a) =
Λ
3H
2
(a)
=
Ω
Λ
a
3
a + Ω
M
(1 − a) + Ω
Λ
(a
3
−a)
_
a→0
−→ 0
a→∞
−→ 1
. (21)
This implies that for sufﬁciently early times, a ¸ 1, all matterdominated FRW universes can be de
scribed by Einsteinde Sitter (EdS) models (Ω
K
= 0, Ω
Λ
= 0).
5
On the other hand, the vacuum energy
will always dominate in the future.
Another relationship which becomes very useful is that of the cosmological deceleration parameter
today, q
0
, in terms of the matter and cosmological constant components of the universe, see Eq. (10),
q
0
≡ −
¨a
aH
2
¸
¸
¸
¸
0
=
1
2
Ω
M
− Ω
Λ
, (22)
which is independent of the spatial curvature. Uniform expansion corresponds to q
0
= 0 and requires a
precise cancellation: Ω
M
= 2Ω
Λ
. It represents spatial sections that are expanding at a ﬁxed rate, its scale
factor growing by the same amount in equallyspaced time intervals. Accelerated expansion corresponds
to q
0
< 0 and comes about whenever Ω
M
< 2Ω
Λ
: spatial sections expand at an increasing rate, their
scale factor growing at a greater speed with each time interval. Decelerated expansion corresponds to
q
0
> 0 and occurs whenever Ω
M
> 2Ω
Λ
: spatial sections expand at a decreasing rate, their scale factor
growing at a smaller speed with each time interval.
5
Note that in the limit a →0 the radiation component starts dominating, see Eq. (18), but we still recover the EdS model.
2.1.3 Mechanical analogy
It is enlightening to work with a mechanical analogy of the Friedmann equation. Let us rewrite Eq. (8)
as
1
2
˙ a
2
−
GM
a
−
Λ
6
a
2
= −
K
2
= constant , (23)
where M ≡
4π
3
ρ a
3
is the equivalent of mass for the whole volume of the universe. Equation (23) can
be understood as the energy conservation law E = T + V for a test particle of unit mass in the central
potential
V (r) = −
GM
r
+
1
2
k r
2
, (24)
corresponding to a Newtonian potential plus a harmonic oscillator potential with a negative spring con
stant k ≡ −Λ/3. Note that, in the absence of a cosmological constant (Λ = 0), a critical universe,
deﬁned as the borderline between indeﬁnite expansion and recollapse, corresponds, through the Fried
mann equations of motion, precisely with a ﬂat universe (K = 0). In that case, and only in that case, a
spatially open universe (K = −1) corresponds to an eternally expanding universe, and a spatially closed
universe (K = +1) to a recollapsing universe in the future. Such a well known (textbook) correspon
dence is incorrect when Ω
Λ
,= 0: spatially open universes may recollapse while closed universes can
expand forever. One can see in Fig. 1 a range of possible evolutions of the scale factor, for various pairs
of values of (Ω
M
, Ω
Λ
).
One can show that, for Ω
Λ
,= 0, a critical universe (H =
˙
H = 0) corresponds to those points
x ≡ a
0
/a > 0, for which f(x) ≡ H
2
(a) and f
(x) vanish, while f
(x) > 0,
f(x) = x
3
Ω
M
+ x
2
Ω
K
+ Ω
Λ
= 0 , (25)
f
(x) = 3x
2
Ω
M
+ 2xΩ
K
= 0
_
x = 0
x = −2Ω
K
/3Ω
M
> 0
, (26)
f
(x) = 6xΩ
M
+ 2Ω
K
=
_
+2Ω
K
> 0 x = 0
−2Ω
K
> 0 x = 2[Ω
K
[/3Ω
M
. (27)
Using the cosmic sum rule (19), we can write the solutions as
Ω
Λ
=
_
_
_
0 Ω
M
≤ 1
4Ω
M
sin
3
_
1
3
arcsin(1 −Ω
−1
M
)
_
Ω
M
≥ 1
. (28)
The ﬁrst solution corresponds to the critical point x = 0 (a = ∞), and Ω
K
> 0, while the second one
to x = 2[Ω
K
[/3Ω
M
, and Ω
K
< 0. Expanding around Ω
M
= 1, we ﬁnd Ω
Λ
·
4
27
(Ω
M
− 1)
3
/Ω
2
M
, for
Ω
M
≥ 1. These critical solutions are asymptotic to the Einsteinde Sitter model (Ω
M
= 1, Ω
Λ
= 0), see
Fig. 2.
2.1.4 Thermodynamical analogy
It is also enlightening to ﬁnd an analogy between the energy conservation equation (9) and the second
law of Thermodynamics,
TdS = dU + pdV , (29)
where U = ρV is the total energy of the closed system and V = a
3
is its physical volume. Equation (9)
implies that the expansion of the universe is adiabatic or isoentropic (dS = 0), corresponding to a ﬂuid
in thermal equilibriumat a temperature T. For a barotropic ﬂuid, satisfying the equation of state p = ωρ,
we can write the energy density evolution as
d
dt
(ρa
3
) = −p
d
dt
(a
3
) = −3Hω (ρa
3
) . (30)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3 2 1 0 1 2 3
A B
C
D
No Λ
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3 2 1 0 1 2 3
A
E
F
Flat
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3 2 1 0 1 2 3
D
G
H
Closed
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3 2 1 0 1 2 3
B I
J
Open
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3 2 1 0 1 2 3
F
K
L
Loitering
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3 2 1 0 1 2 3
M
N
Bouncing
Fig. 1: Evolution of the scale parameter with respect to time for different values of matter density and cosmological parameter.
The horizontal axis represents τ = H0(t − t0), while the vertical axis is y = a/a0 in each case. The values of (ΩM, ΩΛ)
for different plots are: A=(1,0), B=(0.1,0), C=(1.5,0), D=(3,0), E=(0.1,0.9), F=(0,1), G=(3,.1), H=(3,1), I=(.1,.5), J=(.5,−1),
K=(1.1,2.707), L=(1,2.59), M=(0.1,1.5), N=(0.1,2.5). From Ref. [7].
For relativistic particles in thermal equilibrium, the trace of the energymomentum tensor vanishes (be
cause of conformal invariance) and thus p
R
= ρ
R
/3 ⇒ ω = 1/3. In that case, the energy density of
radiation in thermal equilibriumcan be written as [8]
ρ
R
=
π
2
30
g
∗
T
4
, (31)
g
∗
=
i=bosons
g
i
_
T
i
T
_
4
+
7
8
i=fermions
g
i
_
T
i
T
_
4
, (32)
where g
∗
is the number of relativistic degrees of freedom, coming from both bosons and fermions. Using
the equilibrium expressions for the pressure and density, we can write dp = (ρ +p)dT/T, and therefore
dS =
1
T
d[(ρ + p)V ] − (ρ + p)V
dT
T
2
= d
_
(ρ + p)V
T
+ const.
_
(33)
That is, up to an additive constant, the entropy per comoving volume is S = a
3
(ρ + p)V/T, which is
conserved. The entropy per comoving volume is dominated by the contribution of relativistic particles,
1.5
1.0
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Ω
Λ
Ω
M
A
c
c
e
le
r
a
t
in
g
D
e
c
e
le
r
a
t
in
g
C
l
o
s
e
d
O
p
e
n
B
o
u
n
c
e
Expansion
Recollapse
Fig. 2: Parameter space (ΩM, ΩΛ). The line ΩΛ = 1 − ΩM corresponds to a ﬂat universe, ΩK = 0, separating open from
closed universes. The line ΩΛ = ΩM/2 corresponds to uniform expansion, q0 = 0, separating accelerating from decelerating
universes. The dashed line corresponds to critical universes, separating eternal expansion from recollapse in the future. Finally,
the dotted line corresponds to t0H0 = ∞, beyond which the universe has a bounce.
so that, to very good approximation,
S =
2π
2
45
g
∗s
(aT)
3
= constant , (34)
g
∗s
=
i=bosons
g
i
_
T
i
T
_
3
+
7
8
i=fermions
g
i
_
T
i
T
_
3
. (35)
A consequence of Eq. (34) is that, during the adiabatic expansion of the universe, the scale factor grows
inversely proportional to the temperature of the universe, a ∝ 1/T. Therefore, the observational fact
that the universe is expanding today implies that in the past the universe must have been much hotter and
denser, and that in the future it will become much colder and dilute. Since the ratio of scale factors can
be described in terms of the redshift parameter z, see Eq. (5), we can ﬁnd the temperature of the universe
at an earlier epoch by
T = T
0
(1 +z) . (36)
Such a relation has been spectacularly conﬁrmed with observations of absorption spectra from quasars
at large distances, which showed that, indeed, the temperature of the radiation background scaled with
redshift in the way predicted by the hot Big Bang model.
2.2 Brief thermal history of the universe
In this Section, I will brieﬂy summarize the thermal history of the universe, from the Planck era to the
present. As we go back in time, the universe becomes hotter and hotter and thus the amount of energy
available for particle interactions increases. As a consequence, the nature of interactions goes from those
described at low energy by long range gravitational and electromagnetic physics, to atomic physics,
nuclear physics, all the way to high energy physics at the electroweak scale, gran uniﬁcation (perhaps),
and ﬁnally quantum gravity. The last two are still uncertain since we do not have any experimental
evidence for those ultra high energy phenomena, and perhaps Nature has followed a different path.
6
6
See the recent theoretical developments on large extra dimensions and quantum gravity at the TeV [9].
The way we know about the high energy interactions of matter is via particle accelerators, which
are unravelling the details of those fundamental interactions as we increase in energy. However, one
should bear in mind that the physical conditions that take place in our high energy colliders are very
different from those that occurred in the early universe. These machines could never reproduce the
conditions of density and pressure in the rapidly expanding thermal plasma of the early universe. Nev
ertheless, those experiments are crucial in understanding the nature and rate of the local fundamental
interactions available at those energies. What interests cosmologists is the statistical and thermal proper
ties that such a plasma should have, and the role that causal horizons play in the ﬁnal outcome of the early
universe expansion. For instance, of crucial importance is the time at which certain particles decoupled
from the plasma, i.e. when their interactions were not quick enough compared with the expansion of the
universe, and they were left out of equilibrium with the plasma.
One can trace the evolution of the universe from its origin till today. There is still some speculation
about the physics that took place in the universe above the energy scales probed by present colliders.
Nevertheless, the overall layout presented here is a plausible and hopefully testable proposal. According
to the best accepted view, the universe must have originated at the Planck era (10
19
GeV, 10
−43
s)
from a quantum gravity ﬂuctuation. Needless to say, we don’t have any experimental evidence for such
a statement: Quantum gravity phenomena are still in the realm of physical speculation. However, it
is plausible that a primordial era of cosmological inﬂation originated then. Its consequences will be
discussed below. Soon after, the universe may have reached the Grand Uniﬁed Theories (GUT) era (10
16
GeV, 10
−35
s). Quantum ﬂuctuations of the inﬂaton ﬁeld most probably left their imprint then as tiny
perturbations in an otherwise very homogenous patch of the universe. At the end of inﬂation, the huge
energy density of the inﬂaton ﬁeld was converted into particles, which soon thermalized and became the
origin of the hot Big Bang as we know it. Such a process is called reheating of the universe. Since
then, the universe became radiation dominated. It is probable (although by no means certain) that the
asymmetry between matter and antimatter originated at the same time as the rest of the energy of the
universe, from the decay of the inﬂaton. This process is known under the name of baryogenesis since
baryons (mostly quarks at that time) must have originated then, from the leftovers of their annihilation
with antibaryons. It is a matter of speculation whether baryogenesis could have occurred at energies
as low as the electroweak scale (100 GeV, 10
−10
s). Note that although particle physics experiments
have reached energies as high as 100 GeV, we still do not have observational evidence that the universe
actually went through the EW phase transition. If conﬁrmed, baryogenesis would constitute another
“window” into the early universe. As the universe cooled down, it may have gone through the quark
gluon phase transition (10
2
MeV, 10
−5
s), when baryons (mainly protons and neutrons) formed from
their constituent quarks.
The furthest window we have on the early universe at the moment is that of primordial nucleosyn
thesis (1 − 0.1 MeV, 1 s – 3 min), when protons and neutrons were cold enough that bound systems
could form, giving rise to the lightest elements, soon after neutrino decoupling: It is the realm of nuclear
physics. The observed relative abundances of light elements are in agreement with the predictions of the
hot Big Bang theory. Immediately afterwards, electronpositron annihilation occurs (0.5 MeV, 1 min)
and all their energy goes into photons. Much later, at about (1 eV, ∼ 10
5
yr), matter and radiation have
equal energy densities. Soon after, electrons become bound to nuclei to form atoms (0.3 eV, 3 10
5
yr), in a process known as recombination: It is the realm of atomic physics. Immediately after, photons
decouple from the plasma, travelling freely since then. Those are the photons we observe as the cosmic
microwave background. Much later (∼ 1−10 Gyr), the small inhomogeneities generated during inﬂation
have grown, via gravitational collapse, to become galaxies, clusters of galaxies, and superclusters, char
acterizing the epoch of structure formation. It is the realm of long range gravitational physics, perhaps
dominated by a vacuum energy in the form of a cosmological constant. Finally (3K, 13 Gyr), the Sun,
the Earth, and biological life originated from previous generations of stars, and from a primordial soup
of organic compounds, respectively.
I will now review some of the more robust features of the Hot Big Bang theory of which we have
precise observational evidence.
2.2.1 Primordial nucleosynthesis and light element abundance
In this subsection I will brieﬂy review Big Bang nucleosynthesis and give the present observational
constraints on the amount of baryons in the universe. In 1920 Eddington suggested that the sun might
derive its energy from the fusion of hydrogen into helium. The detailed reactions by which stars burn
hydrogen were ﬁrst laid out by Hans Bethe in 1939. Soon afterwards, in 1946, George Gamow realized
that similar processes might have occurred also in the hot and dense early universe and gave rise to the
ﬁrst light elements [4]. These processes could take place when the universe had a temperature of around
T
NS
∼ 1 − 0.1 MeV, which is about 100 times the temperature in the core of the Sun, while the density
is ρ
NS
=
π
2
30
g
∗
T
4
NS
∼ 82 g cm
−3
, about the same density as the core of the Sun. Note, however, that
although both processes are driven by identical thermonuclear reactions, the physical conditions in star
and Big Bang nucleosynthesis are very different. In the former, gravitational collapse heats up the core of
the star and reactions last for billions of years (except in supernova explosions, which last a few minutes
and creates all the heavier elements beyond iron), while in the latter the universe expansion cools the hot
and dense plasma in just a few minutes. Nevertheless, Gamow reasoned that, although the early period of
cosmic expansion was much shorter than the lifetime of a star, there was a large number of free neutrons
at that time, so that the lighter elements could be built up quickly by succesive neutron captures, starting
with the reaction n + p →D + γ. The abundances of the light elements would then be correlated with
their neutron capture cross sections, in rough agreement with observations [6, 10].
Fig. 3: The relative abundance of light elements to Hidrogen. Note the large range of scales involved. From Ref. [10].
Nowadays, Big Bang nucleosynthesis (BBN) codes compute a chain of around 30 coupled nuclear
reactions, to produce all the light elements up to beryllium7.
7
Only the ﬁrst four or ﬁve elements can
be computed with accuracy better than 1% and compared with cosmological observations. These light
elements are H,
4
He, D,
3
He,
7
Li, and perhaps also
6
Li. Their observed relative abundance to hydrogen
is [1 : 0.25 : 3 10
−5
: 2 10
−5
: 2 10
−10
] with various errors, mainly systematic. The BBN
codes calculate these abundances using the laboratory measured nuclear reaction rates, the decay rate of
the neutron, the number of light neutrinos and the homogeneous FRW expansion of the universe, as a
function of only one variable, the number density fraction of baryons to photons, η ≡ n
B
/n
γ
. In fact,
the present observations are only consistent, see Fig. 3 and Ref. [11, 10], with a very narrow range of
values of
η
10
≡ 10
10
η = 4.6 −5.9 . (37)
Such a small value of η indicates that there is about one baryon per 10
9
photons in the universe today.
Any acceptable theory of baryogenesis should account for such a small number. Furthermore, the present
baryon fraction of the critical density can be calculated from η
10
as [10]
Ω
B
h
2
= 3.6271 10
−3
η
10
= 0.0190 ± 0.0024 (95% c.l.) (38)
Clearly, this number is well below closure density, so baryons cannot account for all the matter in the
universe, as I shall discuss below.
2.2.2 Neutrino decoupling
Just before the nucleosynthesis of the lightest elements in the early universe, weak interactions were too
slow to keep neutrinos in thermal equilibrium with the plasma, so they decoupled. We can estimate the
temperature at which decoupling occurred from the weak interaction cross section, σ
w
· G
2
F
T
2
at ﬁnite
temperature T, where G
F
= 1.2 10
−5
GeV
−2
is the Fermi constant. The neutrino interaction rate, via
W boson exchange in n +ν ↔p + e
−
and p + ¯ ν ↔n + e
+
, can be written as [8]
Γ
ν
= n
ν
¸σ
w
[v[) · G
2
F
T
5
, (39)
while the rate of expansion of the universe at that time (g
∗
= 10.75) was H · 5.4 T
2
/M
P
, where
M
P
= 1.22 10
19
GeV is the Planck mass. Neutrinos decouple when their interaction rate is slower
than the universe expansion, Γ
ν
≤ H or, equivalently, at T
ν−dec
· 0.8 MeV. Below this temperature,
neutrinos are no longer in thermal equilibriumwith the rest of the plasma, and their temperature continues
to decay inversely proportional to the scale factor of the universe. Since neutrinos decoupled before
e
+
e
−
annihilation, the cosmic background of neutrinos has a temperature today lower than that of the
microwave background of photons. Let us compute the difference. At temperatures above the the mass
of the electron, T > m
e
= 0.511 MeV, and below 0.8 MeV, the only particle species contributing to
the entropy of the universe are the photons (g
∗
= 2) and the electronpositron pairs (g
∗
= 4
7
8
); total
number of degrees of freedom g
∗
=
11
2
. At temperatures T · m
e
, electrons and positrons annihilate into
photons, heating up the plasma (but not the neutrinos, which had decoupled already). At temperatures
T < m
e
, only photons contribute to the entropy of the universe, with g
∗
= 2 degrees of freedom.
Therefore, from the conservation of entropy, we ﬁnd that the ratio of T
γ
and T
ν
today must be
T
γ
T
ν
=
_
11
4
_
1/3
= 1.401 ⇒ T
ν
= 1.945 K, (40)
where I have used T
CMB
= 2.725 ± 0.002 K. We still have not measured such a relic background of
neutrinos, and probably will remain undetected for a long time, since they have an average energy of
order 10
−4
eV, much below that required for detection by present experiments (of order GeV), precisely
because of the relative weakness of the weak interactions. Nevertheless, it would be fascinating if, in the
future, ingenious experiments were devised to detect such a background, since it would conﬁrm one of
the most robust features of Big Bang cosmology.
7
The rest of nuclei, up to iron (Fe), are produced in heavy stars, and beyond Fe in novae and supernovae explosions.
2.2.3 Matterradiation equality
Relativistic species have energy densities proportional to the quartic power of temperature and therefore
scale as ρ
R
∝ a
−4
, while nonrelativistic particles have essentially zero pressure and scale as ρ
M
∝ a
−3
,
see Eq. (30). Therefore, there will be a time in the evolution of the universe in which both energy
densities are equal ρ
R
(t
eq
) = ρ
M
(t
eq
). Since then both decay differently, and thus
1 + z
eq
=
a
0
a
eq
=
Ω
M
Ω
R
= 3.1 10
4
Ω
M
h
2
, (41)
where I have used Ω
R
h
2
= Ω
CMB
h
2
+Ω
ν
h
2
= 3.24 10
−5
for three massless neutrinos at T = T
ν
. As
I will show later, the matter content of the universe today is below critical, Ω
M
· 0.3, while h · 0.65,
and therefore (1 +z
eq
) · 3900, or about t
eq
= 1.2 10
3
(Ω
M
h
2
)
−2
· 7 10
4
years after the origin of
the universe. Around the time of matterradiation equality, the rate of expansion (18) can be written as
(a
0
≡ 1)
H(a) = H
0
_
Ω
R
a
−4
+ Ω
M
a
−3
_
1/2
= H
0
Ω
1/2
M
a
−3/2
_
1 +
a
eq
a
_
1/2
. (42)
The horizon size is the coordinate distance travelled by a photon since the beginning of the universe,
d
H
∼ H
−1
, i.e. the size of causally connected regions in the universe. The comoving horizon size is
then given by
d
H
=
c
aH(a)
= c H
−1
0
Ω
−1/2
M
a
1/2
_
1 +
a
eq
a
_
−1/2
. (43)
Thus the horizon size at matterradiation equality (a = a
eq
) is
d
H
(a
eq
) =
c H
−1
0
√
2
Ω
−1/2
M
a
1/2
eq
· 12 (Ω
M
h)
−1
h
−1
Mpc . (44)
This scale plays a very important role in theories of structure formation.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1000 1100 1200 1300 1400 1500 1600
X
e
e
q
(1+z)
Fig. 4: The equilibrium ionization fraction X
eq
e
as a function of redshift. The two lines show the range of η10 = 4.6 −5.9.
2.2.4 Recombination and photon decoupling
As the temperature of the universe decreased, electrons could eventually become bound to protons to
form neutral hydrogen. Nevertheless, there is always a nonzero probability that a rare energetic photon
ionizes hydrogen and produces a free electron. The ionization fraction of electrons in equilibrium with
the plasma at a given temperature is given by [8]
1 − X
eq
e
X
eq
e
=
4
√
2ζ(3)
√
π
η
_
T
m
e
_
3/2
e
E
ion
/T
, (45)
where E
ion
= 13.6 eV is the ionization energy of hydrogen, and η is the baryontophoton ratio (37). If
we now use Eq. (36), we can compute the ionization fraction X
eq
e
as a function of redshift z, see Fig. 4.
Note that the huge number of photons with respect to electrons (in the ratio
4
He : H : γ · 1 : 4 : 10
10
)
implies that even at a very low temperature, the photon distribution will contain a sufﬁciently large num
ber of highenergy photons to ionize a signiﬁcant fraction of hydrogen. In fact, deﬁning recombination
as the time at which X
eq
e
≡ 0.1, one ﬁnds that the recombination temperature is T
rec
= 0.3 eV ¸E
ion
,
for η
10
· 5.2. Comparing with the present temperature of the microwave background, we deduce the
corresponding redshift at recombination, (1 +z
rec
) · 1270.
Photons remain in thermal equilibrium with the plasma of baryons and electrons through elastic
Thomson scattering, with cross section
σ
T
=
8πα
2
3m
2
e
= 6.65 10
−25
cm
2
= 0.665 barn , (46)
where α = 1/137.036 is the dimensionless electromagnetic coupling constant. The mean free path of
photons λ
γ
in such a plasma can be estimated from the photon interaction rate, λ
−1
γ
∼ Γ
γ
= n
e
σ
T
.
For temperatures above a few eV, the mean free path is much smaller that the causal horizon at that
time and photons suffer multiple scattering: the plasma is like a dense fog. Photons will decouple from
the plasma when their interaction rate cannot keep up with the expansion of the universe and the mean
free path becomes larger than the horizon size: the universe becomes transparent. We can estimate
this moment by evaluating Γ
γ
= H at photon decoupling. Using n
e
= X
e
η n
γ
, one can compute
the decoupling temperature as T
dec
= 0.26 eV, and the corresponding redshift as (1 + z
dec
) · 1100.
This redshift deﬁnes the so called last scattering surface, when photons last scattered off protons and
electrons and travelled freely ever since. This decoupling occurred when the universe was approximately
t
dec
= 1.8 10
5
(Ω
M
h
2
)
−1/2
· 5 10
5
years old.
Fig. 5: The Cosmic Microwave Background Spectrum seen by the FIRAS instrument on COBE. The left panel corresponds to
the monopole spectrum, T0 = 2.725 ±0.002 K, where the error bars are smaller than the line width. The right panel shows the
dipole spectrum, δT1 = 3.372 ±0.014 mK. From Ref. [12].
2.2.5 The microwave background
One of the most remarkable observations ever made my mankind is the detection of the relic background
of photons from the Big Bang. This background was predicted by George Gamow and collaborators
in the 1940s, based on the consistency of primordial nucleosynthesis with the observed helium abun
dance. They estimated a value of about 10 K, although a somewhat more detailed analysis by Alpher and
Herman in 1950 predicted T
γ
≈ 5 K. Unfortunately, they had doubts whether the radiation would have
survived until the present, and this remarkable prediction slipped into obscurity, until Dicke, Peebles,
Roll and Wilkinson [13] studied the problem again in 1965. Before they could measure the photon back
ground, they learned that Penzias and Wilson had observed a weak isotropic background signal at a radio
wavelength of 7.35 cm, corresponding to a blackbody temperature of T
γ
= 3.5 ± 1 K. They published
their two papers back to back, with that of Dicke et al. explaining the fundamental signiﬁcance of their
measurement [6].
Since then many different experiments have conﬁrmed the existence of the microwave background.
The most outstanding one has been the Cosmic Background Explorer (COBE) satellite, whose FIRAS
instrument measured the photon background with great accuracy over a wide range of frequencies (ν =
1 −97 cm
−1
), see Ref. [12], with a spectral resolution
∆ν
ν
= 0.0035. Nowadays, the photon spectrum is
conﬁrmed to be a blackbody spectrum with a temperature given by [12]
T
CMB
= 2.725 ± 0.002 K (systematic, 95% c.l.) ± 7 µK (1σ statistical) (47)
In fact, this is the best blackbody spectrum ever measured, see Fig. 5, with spectral distortions below the
level of 10 parts per million (ppm).
Fig. 6: The Cosmic Microwave Background Spectrum seen by the DMR instrument on COBE. The top ﬁgure corresponds to
the monopole, T0 = 2.725 ± 0.002 K. The middle ﬁgure shows the dipole, δT1 = 3.372 ± 0.014 mK, and the lower ﬁgure
shows the quadrupole and higher multipoles, δT2 = 18 ± 2 µK. The central region corresponds to foreground by the galaxy.
From Ref. [14].
Moreover, the differential microwave radiometer (DMR) instrument on COBE, with a resolution
of about 7
◦
in the sky, has also conﬁrmed that it is an extraordinarily isotropic background. The devia
tions from isotropy, i.e. differences in the temperature of the blackbody spectrum measured in different
directions in the sky, are of the order of 20 µK on large scales, or one part in 10
5
, see Ref. [14]. There
is, in fact, a dipole anisotropy of one part in 10
3
, δT
1
= 3.372 ± 0.007 mK (95% c.l.), in the direction
of the Virgo cluster, (l, b) = (264.14
◦
± 0.30, 48.26
◦
± 0.30) (95% c.l.). Under the assumption that a
Doppler effect is responsible for the entire CMB dipole, the velocity of the Sun with respect to the CMB
rest frame is v
= 371 ± 0.5 km/s, see Ref. [12].
8
When subtracted, we are left with a whole spectrum
of anisotropies in the higher multipoles (quadrupole, octupole, etc.), δT
2
= 18 ± 2 µK (95% c.l.), see
Ref. [14] and Fig. 6.
Soon after COBE, other groups quickly conﬁrmed the detection of temperature anisotropies at
around 30 µK and above, at higher multipole numbers or smaller angular scales. As I shall discuss below,
these anisotropies play a crucial role in the understanding of the origin of structure in the universe.
2.3 Largescale structure formation
Although the isotropic microwave background indicates that the universe in the past was extraordinarily
homogeneous, we know that the universe today is not exactly homogeneous: we observe galaxies, clus
ters and superclusters on large scales. These structures are expected to arise from very small primordial
inhomogeneities that grow in time via gravitational instability, and that may have originated from tiny
ripples in the metric, as matter fell into their troughs. Those ripples must have left some trace as temper
ature anisotropies in the microwave background, and indeed such anisotropies were ﬁnally discovered
by the COBE satellite in 1992. The reason why they took so long to be discovered was that they appear
as perturbations in temperature of only one part in 10
5
.
While the predicted anisotropies have ﬁnally been seen in the CMB, not all kinds of matter and/or
evolution of the universe can give rise to the structure we observe today. If we deﬁne the density contrast
as [15]
δ(x, a) ≡
ρ(x, a) − ¯ ρ(a)
¯ ρ(a)
=
_
d
3
k δ
k
(a) e
i
k·x
, (48)
where ¯ ρ(a) = ρ
0
a
−3
is the average cosmic density, we need a theory that will grow a density contrast
with amplitude δ ∼ 10
−5
at the last scattering surface (z = 1100) up to density contrasts of the order of
δ ∼ 10
2
for galaxies at redshifts z ¸ 1, i.e. today. This is a necessary requirement for any consistent
theory of structure formation [16].
Furthermore, the anisotropies observed by the COBE satellite correspond to a smallamplitude
scaleinvariant primordial power spectrum of inhomogeneities
P(k) = ¸[δ
k
[
2
) ∝ k
n
, with n = 1 , (49)
where the brackets ¸) represent integration over an ensemble of different universe realizations. These
inhomogeneities are like waves in the spacetime metric. When matter fell in the troughs of those waves,
it created density perturbations that collapsed gravitationally to form galaxies and clusters of galaxies,
with a spectrum that is also scale invariant. Such a type of spectrum was proposed in the early 1970s by
Edward R. Harrison, and independently by the Russian cosmologist Yakov B. Zel’dovich, see Ref. [17],
to explain the distribution of galaxies and clusters of galaxies on very large scales in our observable
universe.
Today various telescopes – like the Hubble Space Telescope, the twin Keck telescopes in Hawaii
and the European Southern Observatory telescopes in Chile – are exploring the most distant regions
of the universe and discovering the ﬁrst galaxies at large distances. The furthest galaxies observed so
far are at redshifts of z · 5, or 12 billion light years from the Earth, whose light was emitted when
the universe had only about 5% of its present age. Only a few galaxies are known at those redshifts,
but there are at present various catalogs like the CfA and APM galaxy catalogs, and more recently the
IRAS Point Source redshift Catalog, see Fig. 7, and Las Campanas redshift surveys, that study the spatial
distribution of hundreds of thousands of galaxies up to distances of a billion light years, or z < 0.1, that
recede from us at speeds of tens of thousands of kilometres per second. These catalogs are telling us
about the evolution of clusters of galaxies in the universe, and already put constraints on the theory of
8
COBE even determined the annual variation due to the Earth’s motion around the Sun – the ultimate proof of Copernicus’
hypothesis.
Fig. 7: The IRAS Point Source Catalog redshift survey contains some 15,000 galaxies, covering over 83% of the sky up to
redshifts of z ≤ 0.05. We show here the projection of the galaxy distribution in galactic coordinates. From Ref. [18].
structure formation. From these observations one can infer that most galaxies formed at redshifts of the
order of 2 − 6; clusters of galaxies formed at redshifts of order 1, and superclusters are forming now.
That is, cosmic structure formed from the bottom up: from galaxies to clusters to superclusters, and not
the other way around. This fundamental difference is an indication of the type of matter that gave rise
to structure. The observed power spectrum of the galaxy matter distribution from a selection of deep
redshift catalogs can be seen in Fig. 8.
We know from Big Bang nucleosynthesis that all the baryons in the universe cannot account for
the observed amount of matter, so there must be some extra matter (dark since we don’t see it) to account
for its gravitational pull. Whether it is relativistic (hot) or nonrelativistic (cold) could be inferred from
observations: relativistic particles tend to diffuse from one concentration of matter to another, thus trans
ferring energy among them and preventing the growth of structure on small scales. This is excluded by
observations, so we conclude that most of the matter responsible for structure formation must be cold.
How much there is is a matter of debate at the moment. Some recent analyses suggest that there is not
enough cold dark matter to reach the critical density required to make the universe ﬂat. If we want to
make sense of the present observations, we must conclude that some other form of energy permeates
the universe. In order to resolve this issue, even deeper galaxy redshift catalogs are underway, looking
at millions of galaxies, like the Sloan Digital Sky Survey (SDSS) and the AngloAustralian two degree
ﬁeld (2dF) Galaxy Redshift Survey, which are at this moment taking data, up to redshifts of z
∼
<
0.5,
over a large region of the sky. These important observations will help astronomers determine the nature
of the dark matter and test the validity of the models of structure formation.
Before COBE discovered the anisotropies of the microwave background there were serious doubts
whether gravity alone could be responsible for the formation of the structure we observe in the universe
today. It seemed that a new force was required to do the job. Fortunately, the anisotropies were found
with the right amplitude for structure to be accounted for by gravitational collapse of primordial inho
mogeneities under the attraction of a large component of nonrelativistic dark matter. Nowadays, the
standard theory of structure formation is a cold dark matter model with a non vanishing cosmological
constant in a spatially ﬂat universe. Gravitational collapse ampliﬁes the density contrast initially through
linear growth and later on via nonlinear collapse. In the process, overdense regions decouple from
the Hubble expansion to become bound systems, which start attracting eachother to form larger bound
structures. In fact, the largest structures, superclusters, have not yet gone nonlinear.
The primordial spectrum(49) is reprocessed by gravitational instability after the universe becomes
matter dominated and inhomogeneities can grow. Linear perturbation theory shows that the growing
Fig. 8: The left panel shows the matter power spectrum for clusters of galaxies, from three different cluster surveys. The right
panel shows a compilation of the most recent estimates of the power spectrum of galaxy clustering, from four of the largest
available redshift surveys of opticallyselected galaxies, compared to the deprojected spectrum of the 2D APM galaxy survey.
From Ref. [19].
mode
9
of small density contrasts go like [15, 16]
δ(a) ∝ a
1+3ω
=
_
a
2
, a < a
eq
a , a > a
eq
(50)
in the Einsteinde Sitter limit (ω = p/ρ = 1/3 and 0, for radiation and matter, respectively). There are
slight deviations for a ¸ a
eq
, if Ω
M
,= 1 or Ω
Λ
,= 0, but we will not be concerned with them here.
The important observation is that, since the density contrast at last scattering is of order δ ∼ 10
−5
, and
the scale factor has grown since then only a factor z
dec
∼ 10
3
, one would expect a density contrast
today of order δ
0
∼ 10
−2
. Instead, we observe structures like galaxies, where δ ∼ 10
2
. So how can
this be possible? The microwave background shows anisotropies due to ﬂuctuations in the baryonic
matter component only (to which photons couple, electromagnetically). If there is an additional matter
component that only couples through very weak interactions, ﬂuctuations in that component could grow
as soon as it decoupled from the plasma, well before photons decoupled from baryons. The reason why
baryonic inhomogeneities cannot grow is because of photon pressure: as baryons collapse towards denser
regions, radiation pressure eventually halts the contraction and sets up acoustic oscillations in the plasma
that prevent the growth of perturbations, until photon decoupling. On the other hand, a weakly interacting
cold dark matter component could start gravitational collapse much earlier, even before matterradiation
equality, and thus reach the density contrast amplitudes observed today. The resolution of this mismatch
is one of the strongest arguments for the existence of a weakly interacting cold dark matter component
of the universe.
How much dark matter there is in the universe can be deduced from the actual power spectrum(the
Fourier transform of the twopoint correlation function of density perturbations) of the observed large
scale structure. One can decompose the density contrast in Fourier components, see Eq. (48). This is
very convenient since in linear perturbation theory individual Fourier components evolve independently.
A comoving wavenumber k is said to “enter the horizon” when k = d
−1
H
(a) = aH(a). If a certain
perturbation, of wavelength λ = k
−1
< d
H
(a
eq
), enters the horizon before matterradiation equality, the
fast radiationdriven expansion prevents darkmatter perturbations from collapsing. Since light can only
cross regions that are smaller than the horizon, the suppression of growth due to radiation is restricted
to scales smaller than the horizon, while largescale perturbations remain unaffected. This is the reason
why the horizon size at equality, Eq. (44), sets an important scale for structure growth,
k
eq
= d
−1
H
(a
eq
) · 0.083 (Ω
M
h) h Mpc
−1
. (51)
9
The decaying modes go like δ(t) ∼ t
−1
, for all ω.
CDM
n = 1
HDM
n = 1
MDM
n = 1
0.001 0.01 0.1 1 10
k ( h Mpc )
1
5
P
(
k
)
(
h
M
p
c
)

3
3
10
4
10
1000
100
10
1
0.1
TCDM
n = .8
C
O
B
E
d ( h Mpc )
1
1000 100 10 1
Microwave Background Superclusters Clusters Galaxies
Fig. 9: The power spectrum for cold dark matter (CDM), tilted cold dark matter (TCDM), hot dark matter (HDM), and mixed
hot plus cold dark matter (MDM), normalized to COBE, for largescale structure formation. From Ref. [20].
The suppression factor can be easily computed from (50) as f
sup
= (a
enter
/a
eq
)
2
= (k
eq
/k)
2
. In other
words, the processed power spectrum P(k) will have the form:
P(k) ∝
_
k , k ¸k
eq
k
−3
, k ¸k
eq
(52)
This is precisely the shape that largescale galaxy catalogs are bound to test in the near future, see Fig. 9.
Furthermore, since relativistic Hot Dark Matter (HDM) transfer energy between clumps of matter, they
will wipe out small scale perturbations, and this should be seen as a distinctive signature in the matter
power spectra of future galaxy catalogs. On the other hand, nonrelativistic Cold Dark Matter (CDM)
allow structure to form on all scales via gravitational collapse. The dark matter will then pull in the
baryons, which will later shine and thus allow us to see the galaxies.
Naturally, when baryons start to collapse onto dark matter potential wells, they will convert a large
fraction of their potential energy into kinetic energy of protons and electrons, ionizing the medium. As a
consequence, we expect to see a large fraction of those baryons constitutinga hot ionized gas surrounding
large clusters of galaxies. This is indeed what is observed, and conﬁrms the general picture of structure
formation.
3 DETERMINATION OF COSMOLOGICAL PARAMETERS
In this Section, I will restrict myself to those recent measurements of the cosmological parameters by
means of standard cosmological techniques, together with a few instances of new results from recently
applied techniques. We will see that a large host of observations are determining the cosmological
parameters with some reliability of the order of 10%. However, the majority of these measurements are
dominated by large systematic errors. Most of the recent work in observational cosmology has been
the search for virtually systematicfree observables, like those obtained from the microwave background
anisotropies, and discussed in Section 4.4. I will devote, however, this Section to the more ‘classical’
measurements of the following cosmological parameters: The rate of expansion H
0
; the matter content
Ω
M
; the cosmological constant Ω
Λ
; the spatial curvature Ω
K
, and the age of the universe t
0
.
10
These ﬁve basic cosmological parameters are not mutually independent. Using the homogeneity
and isotropy on large scales observed by COBE, we can infer relationships between the different cosmo
10
We will take the baryon fraction as given by observations of light element abundances, in accordance with Big Bang
nucleosynthesis, see Eq. (38).
logical parameters through the EinsteinFriedmann equations. In particular, we can deduce the value of
the spatial curvature from the Cosmic Sum Rule,
1 = Ω
M
+ Ω
Λ
+ Ω
K
, (53)
or viceversa, if we determine that the universe is spatially ﬂat from observations of the microwave back
ground, we can be sure that the sum of the matter content plus the cosmological constant must be one.
Another relationshipbetween parameters appears for the age of the universe. In a FRWcosmology,
the cosmic expansion is determined by the Friedmann equation (8). Deﬁning a new time and normalized
scale factor,
y ≡
a
a
0
=
1
1 +z
, τ ≡ H
0
(t −t
0
) , (54)
we can write the Friedmann equation with the help of the Cosmic Sum Rule (19) as
y
(τ) =
_
1 + (y
−1
− 1)Ω
M
+ (y
2
−1)Ω
Λ
_
1/2
, (55)
with initial condition y(0) = 1, y
(0) = 1. Therefore, the present age t
0
is a function of the other
parameters, t
0
= f(H
0
, Ω
M
, Ω
Λ
), determined from
t
0
H
0
=
_
1
0
dy
_
1 + (y
−1
−1)Ω
M
+ (y
2
−1)Ω
Λ
_
−1/2
. (56)
We show in Fig. 10 the contour lines for constant t
0
H
0
in parameter space (Ω
M
, Ω
Λ
).
1.0
0.5
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Ω
Λ
Ω
M
Fig. 10: The contour lines correspond to equal t0H0 = 0.5 −1.0, 1.2, 1.5, 2.0 and 5.0, from bottom to top, in parameter space
(ΩM, ΩΛ). The line t0H0 = ∞would be indistinguishable from that of t0H0 = 5. From Ref. [7].
There are two speciﬁc limits of interest: an open universe with Ω
Λ
= 0, for which the age is given by
t
0
H
0
=
1
1 −Ω
M
−
Ω
M
(1 −Ω
M
)
3/2
ln
_
1 + (1 − Ω
M
)
1/2
Ω
1/2
M
_
= 2
∞
n=0
(1 − Ω
M
)
n
(2n + 1)(2n + 3)
, (57)
and a ﬂat universe with Ω
Λ
= 1 −Ω
M
, for which the age can also be expressed in compact form,
t
0
H
0
=
2
3(1 −Ω
M
)
1/2
ln
_
1 + (1 −Ω
M
)
1/2
Ω
1/2
M
_
=
2
3
∞
n=0
(1 − Ω
M
)
n
2n + 1
. (58)
We have plotted these functions in Fig. 11. It is clear that in both cases t
0
H
0
→ 2/3 as Ω
M
→ 1.
We can now use these relations as a consistency check between the cosmological observations of H
0
,
Ω
M
, Ω
Λ
and t
0
. Of course, we cannot measure the age of the universe directly, but only the age of its
constituents: stars, galaxies, globular clusters, etc. Thus we can only ﬁnd a lower bound on the age of
the universe, t
0 ∼
>
t
gal
+ 1.5 Gyr. As we will see, this is not a trivial bound and, in several occasions,
during the progress towards better determinations of the cosmological parameters, the universe seemed
to be younger than its constituents, a logical inconsistency, of course, only due to an incorrect assessment
of systematic errors [21].
Fig. 11: The age of the universe as a function of the matter content, for an open and a ﬂat universe. From Ref. [22].
In order to understand those recent measurements, one should also deﬁne what is known as the
luminosity distance to an object in the universe. Imagine a source that is emitting light at a distance d
L
from a detector of area dA. The absolute luminosity L of such a source is nothing but the energy emitted
per unit time. A standard candle is a luminous object that can be calibrated with some accuracy and
therefore whose absolute luminosity is known, within certain errors. For example, Cepheid variable stars
and type Ia supernovae are considered to be reasonable standard candles, i.e. their calibration errors are
within bounds. The energy ﬂux T received at the detector is the measured energy per unit time per unit
area of the detector coming from that source. The luminosity distance d
L
is then deﬁned as the radius
of the sphere centered on the source for which the absolute luminosity would give the observed ﬂux,
T ≡ L/4πd
2
L
. In a FriedmannRobertsonWalker universe, light travels along null geodesics, ds
2
= 0,
or, see Eq. (2),
dr
_
1 +a
2
0
H
2
0
r
2
Ω
K
=
1
a
2
0
H
2
0
dz
_
(1 + z)
2
(1 +zΩ
M
) − z(2 + z)Ω
Λ
, (59)
which determines the coordinate distance r = r(z, H
0
, Ω
M
, Ω
Λ
), as a function of redshift z and the other
cosmological parameters. Now let us consider the effect of the universe expansion on the observed ﬂux
coming from a source at a certain redshift z from us. First, the photon energy on its way here will be
redshifted, and thus the observed energy E
0
= E/(1 + z). Second, the rate of photon arrival will be
timedelayed with respect to that emitted by the source, dt
0
= (1 +z)dt. Finally, the fraction of the area
of the 2sphere centered on the source that is covered by the detector is dA/4πa
2
0
r
2
(z). Therefore, the
total ﬂux detected is
T =
L
4πa
2
0
r
2
(z)
≡
L
4πd
2
L
. (60)
The ﬁnal expression for the luminosity distance d
L
as a function of redshift is thus given by [8]
H
0
d
L
= (1 +z) [Ω
K
[
−1/2
sinn
_
[Ω
K
[
1/2
_
z
0
dz
_
(1 +z
)
2
(1 +z
Ω
M
) − z
(2 +z
)Ω
Λ
_
, (61)
where sinn(x) = x if K = 0; sin(x) if K = +1 and sinh(x) if K = −1. Expanding to second order
around z = 0, we obtain Eq. (6),
H
0
d
L
= z +
1
2
_
1 −
Ω
M
2
+ Ω
Λ
_
z
2
+O(z
3
) . (62)
This expression goes beyond the leading linear term, corresponding to the Hubble law, into the second
order term, which is sensitive to the cosmological parameters Ω
M
and Ω
Λ
. It is only recently that cos
mological observations have gone far enough back into the early universe that we can begin to probe the
second term, as I will discuss shortly. Higher order terms are not yet probed by cosmological observa
tions, but they would contribute as important consistency checks.
Let us now pursue the analysis of the recent determinations of the most important cosmological
parameters: the rate of expansion H
0
, the matter content Ω
M
, the cosmological constant Ω
Λ
, the spatial
curvature Ω
K
, and the age of the universe t
0
.
3.1 The rate of expansion H
0
Over most of last century the value of H
0
has been a constant source of disagreement [21]. Around
1929, Hubble measured the rate of expansion to be H
0
= 500 kms
−1
Mpc
−1
, which implied an age of
the universe of order t
0
∼ 2 Gyr, in clear conﬂict with geology. Hubble’s data was based on Cepheid
standard candles that were incorrectly calibrated with those in the Large Magellanic Cloud. Later on,
in 1954 Baade recalibrated the Cepheid distance and obtained a lower value, H
0
= 250 kms
−1
Mpc
−1
,
still in conﬂict with ratios of certain unstable isotopes. Finally, in 1958 Sandage realized that the bright
est stars in galaxies were ionized HII regions, and the Hubble rate dropped down to H
0
= 60 kms
−1
Mpc
−1
, still with large (factor of two) systematic errors. Fortunately, in the past 15 years there has
been signiﬁcant progress towards the determination of H
0
, with systematic errors approaching the 10%
level. These improvements come from two directions. First, technological, through the replacement of
photographic plates (almost exclusively the source of data from the 1920s to 1980s) with charged couple
devices (CCDs), i.e. solid state detectors with excellent ﬂux sensitivity per pixel, which were previously
used successfully in particle physics detectors. Second, by the reﬁnement of existing methods for mea
suring extragalactic distances (e.g. parallax, Cepheids, supernovae, etc.). Finally, with the development
of completely new methods to determine H
0
, which fall into totally independent and very broad cate
gories: a) Gravitational lensing; b) SunyaevZel’dovich effect; c) Extragalactic distance scale, mainly
Cepheid variability and type Ia Supernovae; d) Microwave background anisotropies. I will review here
the ﬁrst three, and leave the last method for Section 4.4, since it involves knowledge about the primordial
spectrum of inhomogeneities.
3.1.1 Gravitational lensing
Imagine a quasistellar object (QSO) at large redshift (z ¸ 1) whose light is lensed by an intervening
galaxy at redshift z ∼ 1 and arrives to an observer at z = 0. There will be at least two different
images of the same background variable point source. The arrival times of photons from two different
gravitationally lensed images of the quasar depend on the different path lengths and the gravitational
potential traversed. Therefore, a measurement of the time delay and the angular separation of the different
images of a variable quasar can be used to determine H
0
with great accuracy. This method, proposed in
1964 by Refsdael [23], offers tremendous potential because it can be applied at great distances and it is
based on very solid physical principles [24].
Unfortunately, there are very few systems with both a favourable geometry (i.e. a known mass
distribution of the intervening galaxy) and a variable background source with a measurable time delay.
That is the reason why it has taken so much time since the original proposal for the ﬁrst results to come
out. Fortunately, there are now very powerful telescopes that can be used for these purposes. The best
candidate todate is the QSO 0957 + 561, observed with the 10m Keck telescope, for which there is a
model of the lensing mass distributionthat is consistent with the measured velocity dispersion. Assuming
a ﬂat space with Ω
M
= 0.25, one can determine [25]
H
0
= 72 ±7 (1σ statistical) ± 15% (systematic) kms
−1
Mpc
−1
. (63)
The main source of systematic error is the degeneracy between the mass distribution of the lens and
the value of H
0
. Knowledge of the velocity dispersion within the lens as a function of position helps
constrain the mass distribution, but those measurements are very difﬁcult and, in the case of lensing by
a cluster of galaxies, the dark matter distribution in those systems is usually unknown, associated with a
complicated cluster potential. Nevertheless, the method is just starting to give promising results and, in
the near future, with the recent discovery of several systems with optimum properties, the prospects for
measuring H
0
and lowering its uncertainty with this technique are excellent.
3.1.2 SunyaevZel’dovich effect
As discussed in the previous Section, the gravitational collapse of baryons onto the potential wells gen
erated by dark matter gave rise to the reionization of the plasma, generating an Xray halo around rich
clusters of galaxies, see Fig. 12. The inverseCompton scattering of microwave background photons
off the hot electrons in the Xray gas results in a measurable distortion of the blackbody spectrum
of the microwave background, known as the SunyaevZel’dovich (SZ) effect. Since photons acquire
extra energy from the Xray electrons, we expect a shift towards higher frequencies of the spectrum,
(∆ν/ν) · (k
B
T
gas
/m
e
c
2
) ∼ 10
−2
. This corresponds to a decrement of the microwave background tem
perature at low frequencies (RayleighJeans region) and an increment at high frequencies, see Ref. [26].
Fig. 12: The Coma cluster of galaxies, seen here in an optical image (left) and an Xray image (right), taken by the recently
launched Chandra Xray Observatory. From Ref. [27].
Measuring the spatial distribution of the SZ effect (3 K spectrum), together with a high resolution
Xray map (10
8
K spectrum) of the cluster, one can determine the density and temperature distribution
of the hot gas. Since the Xray ﬂux is distancedependent (T = L/4πd
2
L
), while the SZ decrement is
not (because the energy of the CMB photons increases as we go back in redshift, ν = ν
0
(1 + z), and
exactly compensates the redshift in energy of the photons that reach us), one can determine from there
the distance to the cluster, and thus the Hubble rate H
0
.
The advantages of this method are that it can be applied to large distances and it is based on
clear physical principles. The main systematics come from possible clumpiness of the gas (which would
reduce H
0
), projection effects (if the clusters are prolate, H
0
could be larger), the assumption of hy
drostatic equilibrium of the Xray gas, details of models for the gas and electron densities, and possible
contaminations from point sources. Present measurements give the value [26]
H
0
= 60 ± 10 (1σ statistical) ±20% (systematic) kms
−1
Mpc
−1
, (64)
compatible with other determinations. A great advantage of this completely new and independent method
is that nowadays more and more clusters are observed in the Xray, and soon we will have highresolution
2D maps of the SZ decrement from several balloon ﬂights, as well as from future microwave background
satellites, together with precise Xray maps and spectra from the Chandra Xray observatory recently
launched by NASA, as well as from the European Xray satellite XMM launched a few months ago by
ESA, which will deliver orders of magnitude better resolution than the existing Einstein Xray satellite.
3.1.3 Cepheid variability
Cepheids are lowmass variable stars with a periodluminosity relation based on the helium ionization
cycles inside the star, as it contracts and expands. This time variability can be measured, and the star’s
absolute luminosity determined from the calibrated relationship. From the observed ﬂux one can then
deduce the luminosity distance, see Eq. (61), and thus the Hubble rate H
0
. The Hubble Space Telescope
(HST) was launched by NASA in 1990 (and repaired in 1993) with the speciﬁc project of calibrating the
extragalactic distance scale and thus determining the Hubble rate with 10% accuracy. The most recent
results from HST are the following [28]
H
0
= 71 ± 4 (random) ± 7 (systematic) kms
−1
Mpc
−1
. (65)
The main source of systematic error is the distance to the Large Magellanic Cloud, which provides the
ﬁducial comparison for Cepheids in more distant galaxies. Other systematic uncertainties that affect the
value of H
0
are the internal extinction correction method used, a possible metallicity dependence of the
Cepheid periodluminosity relation and cluster population incompleteness bias, for a set of 21 galaxies
within 25 Mpc, and 23 clusters within z
∼
<
0.03.
With better telescopes coming up soon, like the Very Large Telescope (VLT) interferometer of the
European Southern Observatory (ESO) in the Chilean Atacama desert, with 4 synchronized telescopes
by the year 2005, and the Next Generation Space Telescope (NGST) proposed by NASA for 2008, it is
expected that much better resolution and therefore accuracy can be obtained for the determination of H
0
.
3.2 The matter content Ω
M
In the 1920s Hubble realized that the so called nebulae were actually distant galaxies very similar to our
own. Soon afterwards, in 1933, Zwicky found dynamical evidence that there is possibly ten to a hundred
times more mass in the Coma cluster than contributed by the luminous matter in galaxies [29]. However,
it was not until the 1970s that the existence of dark matter began to be taken more seriously. At that time
there was evidence that rotation curves of galaxies did not fall off with radius and that the dynamical
mass was increasing with scale from that of individual galaxies up to clusters of galaxies. Since then,
new possible extra sources to the matter content of the universe have been accumulating:
Ω
M
= Ω
B, lum
(stars in galaxies) (66)
+ Ω
B, dark
(MACHOs?) (67)
+ Ω
CDM
(weakly interacting : axion, neutralino?) (68)
+ Ω
HDM
(massive neutrinos?) (69)
The empirical route to the determination of Ω
M
is nowadays one of the most diversiﬁed of all
cosmological parameters. The matter content of the universe can be deduced from the masstolight ratio
of various objects in the universe; from the rotation curves of galaxies; from microlensing and the direct
search of Massive Compact Halo Objects (MACHOs); from the cluster velocity dispersion with the use
of the Virial theorem; from the baryon fraction in the Xray gas of clusters; from weak gravitational
lensing; from the observed matter distribution of the universe via its power spectrum; from the cluster
abundance and its evolution; from direct detection of massive neutrinos at SuperKamiokande; fromdirect
detection of Weakly Interacting Massive Particles (WIMPs) at DAMA and UKDMC, and ﬁnally from
microwave background anisotropies. I will review here just a few of them.
3.2.1 Luminous matter
The most straight forward method of estimating Ω
M
is to measure the luminosity of stars in galaxies and
then estimate the masstolight ratio, deﬁned as the mass per luminosity density observed from an object,
Υ = //L. This ratio is usually expressed in solar units, /
/L
, so that for the sun Υ
= 1. The
luminosity of stars depends very sensitively on their mass and stage of evolution. The masstolight ratio
of stars in the solar neighbourhood is of order Υ ≈ 3. For globular clusters and spiral galaxies we can
determine their mass and luminosity independently and this gives Υ ≈ few. For our galaxy,
L
gal
= (1.0 ± 0.3) 10
8
hL
Mpc
−3
and Υ
gal
= 6 ± 3 . (70)
The contribution of galaxies to the luminosity density of the universe (in the visibleV spectral band,
centered at ∼ 5500
˚
A) is [30]
L
V
= (1.7 ± 0.6) 10
8
hL
Mpc
−3
, (71)
which can be translated into a mass density by multiplying by the observed Υ in that band,
Ω
M
h = (6.1 ± 2.2) 10
−4
Υ
V
. (72)
All the luminous matter in the universe, from galaxies, clusters of galaxies, etc., account for Υ ≈ 10,
and thus [31]
0.002 ≤ Ω
lum
h ≤ 0.006 . (73)
As a consequence, the luminous matter alone is far from the critical density. Moreover, comparing with
the amount of baryons from Big Bang nucleosynthesis (38), we conclude that Ω
lum
¸Ω
B
, so there must
be a large fraction of baryons that are dark, perhaps in the form of very dim stars.
3.2.2 Rotation curves of spiral galaxies
The ﬂat rotation curves of spiral galaxies provide the most direct evidence for the existence of large
amounts of dark matter. Spiral galaxies consist of a central bulge and a very thin disk, stabilized against
gravitational collapse by angular momentum conservation, and surrounded by an approximately spher
ical halo of dark matter. One can measure the orbital velocities of objects orbiting around the disk as a
function of radius from the Doppler shifts of their spectral lines. The rotation curve of the Andromeda
galaxy was ﬁrst measured by Babcock in 1938, from the stars in the disk. Later it became possible to
measure galactic rotation curves far out into the disk, and a trend was found [32]. The orbital velocity
rose linearly from the center outward until it reached a typical value of 200 km/s, and then remained ﬂat
out to the largest measured radii. This was completely unexpected since the observed surface luminosity
of the disk falls off exponentially with radius, I(r) = I
0
exp(−r/r
D
), see Ref. [32]. Therefore, one
would expect that most of the galactic mass is concentrated within a few disk lengths r
D
, such that the
rotation velocity is determined as in a Keplerian orbit, v
rot
= (GM/r)
1/2
∝ r
−1/2
. No such behaviour
is observed. In fact, the most convincing observations come from radio emission (from the 21 cm line) of
neutral hydrogen in the disk, which has been measured to much larger galactic radii than optical tracers.
A typical case is that of the spiral galaxy NGC 6503, where r
D
= 1.73 kpc, while the furthest measured
hydrogen line is at r = 22.22 kpc, about 13 disk lengths away. The measured rotation curve is shown in
Fig. 13 together with the relative components associated with the disk, the halo and the gas.
Fig. 13: The rotation curve of the spiral galaxy NGC 6503, determined by radio observations of hydrogen gas in the disk [33].
The dashed line shows the rotation curve expected from the disk material alone, the dotdashed line is from the dark matter halo
alone.
Nowadays, thousands of galactic rotation curves are known, and all suggest the existence of about
ten times more mass in the halos of spiral galaxies than in the stars of the disk. Recent numerical simula
tions of galaxy formation in a CDM cosmology [34] suggest that galaxies probably formed by the infall
of material in an overdense region of the universe that had decoupled from the overall expansion. The
dark matter is supposed to undergo violent relaxation and create a virialized system, i.e. in hydrostatic
equilibrium. This picture has led to a simple model of darkmatter halos as isothermal spheres, with
density proﬁle ρ(r) = ρ
c
/(r
2
c
+r
2
), where r
c
is a core radius and ρ
c
= v
2
∞
/4πG, with v
∞
equal to the
plateau value of the ﬂat rotation curve. This model is consistent with the universal rotation curve seen in
Fig. 13. At large radii the dark matter distribution leads to a ﬂat rotation curve. Adding up all the matter
in galactic halos up to maximum radii, one ﬁnds Υ
halo
≥ 30 h, and therefore
Ω
halo
≥ 0.03 − 0.05 . (74)
Of course, it would be extraordinary if we could conﬁrm, through direct detection, the existence of
dark matter in our own galaxy. For that purpose, one should measure its rotation curve, which is much
more difﬁcult because of obscuration by dust in the disk, as well as problems with the determination of
reliable galactocentric distances for the tracers. Nevertheless, the rotation curve of the Milky Way has
been measured and conforms to the usual picture, with a plateau value of the rotation velocity of 220
km/s, see Ref. [35]. For dark matter searches, the crucial quantity is the dark matter density in the solar
neighbourhood, which turns out to be (within a factor of two uncertainty depending on the halo model)
ρ
DM
= 0.3 GeV/cm
3
. We will come back to direct searched of dark matter in a later subsection.
3.2.3 Microlensing
The existence of large amounts of dark matter in the universe, and in our own galaxy in particular, is now
established beyond any reasonable doubt, but its nature remains a mystery. We have seen that baryons
cannot account for the whole matter content of the universe; however, since the contribution of the halo
(74) is comparable in magnitude to the baryon fraction of the universe (38), one may ask whether the
galactic halo could be made of purely baryonic material in some nonluminous form, and if so, how one
should search for it. In other words, are MACHOs the nonluminous baryons ﬁlling the gap between
Ω
lum
and Ω
B
? If not, what are they?
Let us start a systematic search for possibilities. They cannot be normal stars since they would
be luminous; neither hot gas since it would shine; nor cold gas since it would absorb light and reemit
in the infrared. Could they be burntout stellar remnants? This seems implausible since they would
arise from a population of normal stars of which there is no trace in the halo. Neutron stars or black
holes would typically arise from Supernova explosions and thus eject heavy elements into the galaxy,
while the overproduction of helium in the halo is strongly constrained. They could be white dwarfs, i.e.
stars not massive enough to reach supernova phase. Despite some recent arguments, a halo composed
by white dwarfs is not rigorously excluded. Are they stars too small to shine? Perhaps Mdwarfs, stars
with a mass M ≤ 0.1 M
which are intrinsically dim; however, very long exposure images of the
Hubble Space Telescope restrict the possible Mdwarf contribution to the galaxy to be below 6%. The
most plausible alternative is a halo composed of brown dwarfs with mass M ≤ 0.08 M
, which never
ignite hydrogen and thus shine only from the residual energy due to gravitational contraction.
11
In fact,
the extrapolation of the stellar mass function to small masses predicts a large number of brown dwarfs
within normal stellar populations. A ﬁnal possibility is primordial black holes (PBH), which could have
been created in the early universe from early phase transitions [36], even before baryons were formed,
and thus may be classiﬁed as nonbaryonic. They could make a large contribution towards the total Ω
M
,
and still be compatible with Big Bang nucleosynthesis.
Fig. 14: Geometry of the light deﬂection by a pointlike mass which gives two images of a source viewed by an observer. From
Ref. [22].
Whatever the arguments for or against baryonic objects as galactic dark matter, nothing would be
more convincing than a direct detection of the various candidates, or their exclusion, in a direct search
experiment. Fortunately, in 1986 Paczy´ nski proposed a method for detecting faint stars in the halo of our
galaxy [39]. The idea is based on the well known effect that a pointlike mass deﬂector placed between
an observer and a light source creates two different images, as shown in Fig. 14. When the source is
exactly aligned with the deﬂector of mass M
D
, the image would be an annulus, an Einstein ring, with
radius
r
2
E
= 4GM
D
d , where d =
d
1
d
2
d
1
+d
2
(75)
is the reduced distance to the source, see Fig. 14. If the two images cannot be separated because their
angular distance α is below the resolving power of the observer’s telescope, the only effect will be an
11
A sometimes discussed alternative, planetsize Jupiters, can be classiﬁed as lowmass brown dwarfs.
Fig. 15: The apparent lightcurve of a source if a pointlike MACHO passes through the line of sight with a transverse velocity v
and an impact parameter b. The ampliﬁcation factor A is shown in logarithmic scale to give the usual astronomical magnitude
of an object. From Ref. [22].
apparent brightening of the star, an effect known as gravitational microlensing. The ampliﬁcation factor
is [39]
A =
2 + u
2
u
√
4 + u
2
, where u ≡
r
r
E
, (76)
with r the distance from the line of sight to the deﬂector. Imagine an observer on Earth watching a distant
star in the Large Magellanic Cloud (LMC), 50 kpc away. If the galactic halo is ﬁlled with MACHOs,
one of them will occasionally pass near the line of sight and thus cause the image of the background
star to brighten. If the MACHO moves with velocity v transverse to the line of sight, and if its impact
parameter, i.e. the minimal distance to the line of sight, is b, then one expects an apparent lightcurve
as shown in Fig. 15 for different values of b/r
E
. The natural time unit is ∆t = r
E
/v, and the origin
corresponds to the time of closest approach to the line of sight.
The probability for a target star to be lensed is independent of the mass of the dark matter ob
ject [39, 22]. For stars in the LMC one ﬁnds a probability, i.e. an optical depth for microlensing of
the galactic halo, of approximately τ ∼ 10
−6
. Thus, if one looks simultaneously at several millions of
stars in the LMC during extended periods of time, one has a good chance of seeing at least a few of
them brightened by a dark halo object. In order to be sure one has seen a microlensing event one has to
monitor a large sample of stars long enough to identify the characteristic light curve shown in Fig. 15.
The unequivocal signatures of such an event are the following: it must be a) unique (nonrepetitive in
time); b) timesymmetric; and c) achromatic (because of general covariance). These signatures allow
one to discriminate against variable stars which constitute the background. The typical duration of the
light curve is the time it takes a MACHO to cross an Einstein radius, ∆t = r
E
/v. If the deﬂector mass
is 1 M
, the average microlensing time will be 3 months, for 10
−2
M
it is 9 days, for 10
−4
M
it is
1 day, and for 10
−6
M
it is 2 hours. A characteristic event, of duration 34 days, is shown in Fig. 16.
The ﬁrst microlensing events towards the LMC were reported by the MACHO and EROS collab
orations in 1993 [40, 41]. Nowadays, there are 12 candidates towards the LMC, 2 towards the SMC,
around 40 towards the bulge of our own galaxy, and about 2 towards Andromeda, seen by AGAPE [42],
with a slightly different technique based on pixel brightening rather than individual stars. Thus, mi
crolensing is a well established technique with a rather robust future. In particular, it has allowed the
MACHO and EROS collaboration to draw exclusion plots for various mass ranges in terms of their max
imum allowed halo fraction, see Fig. 17. The MACHO Collaboration conclude in their 5year analysis,
Fig. 16: The best candidate (LMC1) for microlensing from the MACHO Collaboration in the direction of the Large Mage
llanic Cloud. A recent reanalysis of this event suggested an ampliﬁcation factor Amax = 7.20 ± 0.09, with achromaticity
Ared/Ablue = 1.00 ± 0.05, and a duration of
ˆ
t = 34.8 ± 0.2. From Ref. [37].
see Ref. [38], that the spatial distribution of events is consistent with an extended lens distribution such
as Milky Way or LMC halo, consisting partially of compact objects. A maximum likelihood analysis
gives a MACHO halo fraction of 20% for a typical halo model with a 95% conﬁdence interval of 8% to
50%. A 100% MACHO halo is ruled out at 95% c.l. for all except their most extreme halo model. The
most likely MACHO mass is between 0.15 M
and 0.9 M
, depending on the halo model. The lower
mass is characteristic of white dwarfs, but a galactic halo composed primarily of white dwarfs is barely
compatible with a range of observational constraints. On the other hand, if one wanted to attribute the
observed events to brown dwarfs, one needs to appeal to a very nonstandard density and/or velocity dis
tribution of these objects. It is still unclear what sort of objects the microlensing experiments are seeing
towards the LMC and where the lenses are. Nevertheless, the ﬁeld is expanding, with several new exper
iments already underway, to search for clear signals of parallax, or binary systems, where the degeneracy
between mass and distance can be resolved. For a discussion of those new results, see Ref. [37].
3.2.4 Virial theorem and large scale motion
Clusters of galaxies are the largest gravitationally bound systems in the universe (superclusters are not
yet in equilibrium). We know today several thousand clusters; they have typical radii of 1 − 5 Mpc and
typical masses of 2−9 10
14
M
. Zwicky noted in 1933 that these systems appear to have large amounts
of dark matter [29]. He used the virial theorem (for a gravitationally bound system in equilibrium),
2¸E
kin
) = −¸E
grav
), where ¸E
kin
) =
1
2
m¸v
2
) is the average kinetic energy of one of the bound objects
(galaxies) of mass m and ¸E
grav
) = −m¸GM/r) is the average gravitational potential energy caused
by the attraction of the other galaxies. Measuring the velocity dispersion ¸v
2
) from the Doppler shifts of
the spectral lines and estimating the geometrical size of the systemgives an estimate of its total mass M.
Fig. 17: Likelihood contours for MACHO mass m (in units of solar mass) and halo fraction f for a typical size halo. The plus
sign shows the maximum likelihood estimate and the contours enclose regions of 68%, 90%, 95% and 99% probability. The
panels are labeled according to different sets of selection criteria (A or B), and whether or not an LMC halo with MACHO
fraction f is included. From Ref. [38].
As Zwicky noted, this virial mass of clusters far exceeds their luminous mass, typically leading to a
masstolight ratio Υ
cluster
= 200 ± 70. Assuming that the average cluster Υ is representative of the
entire universe
12
one ﬁnds for the cosmic matter density [44]
Ω
M
= 0.24 ±0.05 (1σ statistical) ±0.09 (systematic) . (77)
On scales larger than clusters the motion of galaxies is dominated by the overall cosmic expansion.
Nevertheless, galaxies exhibit peculiar velocities with respect to the global cosmic ﬂow. For example,
our Local Group of galaxies is moving with a speed of 627 ± 22 km/s relative to the cosmic microwave
background reference frame, towards the Great Attractor.
In the context of the standard gravitational instability theory of structure formation, the peculiar
motions of galaxies are attributed to the action of gravity during the universe evolution, caused by the
matter density inhomogeneities which give rise to the formation of structure. The observed largescale
velocity ﬁelds, together with the observed galaxy distributions, can then be translated into a measure for
the masstolight ratio required to explain the largescale ﬂows. An example of the reconstruction of the
matter density ﬁeld in our cosmological vicinity from the observed velocity ﬁeld is shown in Fig. 18.
The cosmic matter density inferred from such analyses is [43, 45]
Ω
M
> 0.3 95% c.l. (78)
Related methods that are more modeldependent give even larger estimates.
3.2.5 Baryon fraction in clusters
Since large clusters of galaxies form through gravitational collapse, they scoop up mass over a large
volume of space, and therefore the ratio of baryons over the total matter in the cluster should be rep
resentative of the entire universe, at least within a 20% systematic error. Since the 1960s, when Xray
telescopes became available, it is known that galaxy clusters are the most powerful Xray sources in the
sky [46]. The emission extends over the whole cluster and reveals the existence of a hot plasma with
temperature T ∼ 10
7
− 10
8
K, where Xrays are produced by electron bremsstrahlung. Assuming the
12
Recent observations indicate that Υ is independent of scale up to supercluster scales ∼ 100 h
−1
Mpc.
Fig. 18: The velocity and density ﬂuctuation ﬁelds in the Supergalactic Plane as recovered by the POTENT method from the
Mark III velocities of about 3,000 galaxies with 12 h
−1
smoothing. The vectors are projections of the 3D velocity ﬁeld in
the frame of the CMB. Coordinates are in units of 10 h
−1
Mpc. The marked structures are the Local Group (LG), the “Great
Attractor” (GA), the Coma cluster “Great Wall” (GW), the PerseusPisces (PP) region and the “Southern Wall” (SW). From
Ref. [43].
gas to be in hydrostatic equilibrium and applying the virial theorem one can estimate the total mass in
the cluster, giving general agreement (within a factor of 2) with the virial mass estimates. From these
estimates one can calculate the baryon fraction of clusters
f
B
h
3/2
= 0.03 −0.08 ⇒
Ω
B
Ω
M
≈ 0.15 , for h = 0.65 , (79)
which together with (73) indicates that clusters contain far more baryonic matter in the form of hot gas
than in the form of stars in galaxies. Assuming this fraction to be representative of the entire universe,
and using the Big Bang nucleosynthesis value of Ω
B
= 0.05 ±0.01, for h = 0.65, we ﬁnd
Ω
M
= 0.3 ±0.1 (statistical) ±20% (systematic) . (80)
This value is consistent with previous determinations of Ω
M
. If some baryons are ejected fromthe cluster
during gravitational collapse, or some are actually bound in nonluminous objects like planets, then the
actual value of Ω
M
is smaller than this estimate.
3.2.6 Weak gravitational lensing
Since the mid 1980s, deep surveys with powerful telescopes have observed huge arclike features in
galaxy clusters, see for instance Fig. 19. The spectroscopic analysis showed that the cluster and the
giant arcs were at very different redshifts. The usual interpretation is that the arc is the image of a
distant background galaxy which is in the same line of sight as the cluster so that it appears distorted
and magniﬁed by the gravitational lens effect: the giant arcs are essentially partial Einstein rings. From
a systematic study of the cluster mass distribution one can reconstruct the shear ﬁeld responsible for the
gravitational distortion, see Ref. [47].
Fig. 19: The most famous image of weak gravitational lensing around the Abell 2218 cluster, made by the Hubble Space
Telescope. From Ref. [48].
This analysis shows that there are large amounts of dark matter in the clusters, in rough agreement
with the virial mass estimates, although the lensing masses tend to be systematically larger. At present,
the estimates indicate Ω
M
= 0.2 −0.3 on scales
∼
<
6 h
−1
Mpc, while Ω
M
= 0.4 for the Corona Borealis
supercluster, on scales of order 20 Mpc.
3.2.7 Structure formation and the matter power spectrum
One the most important constraints on the amount of matter in the universe comes from the present distri
bution of galaxies. As we mentioned in the Section 2.3, gravitational instability increases the primordial
density contrast, seen at the last scattering surface as temperature anisotropies, into the present density
ﬁeld responsible for the large and the small scale structure.
Since the primordial spectrum is very approximately represented by a scaleinvariant Gaussian
random ﬁeld, the best way to present the results of structure formation is by working with the 2point
correlation function in Fourier space (the equivalent to the Green’s function in QFT), the socalled power
spectrum. If the reprocessed spectrum of inhomogeneities remains Gaussian, the power spectrum is all
we need to describe the galaxy distribution. NonGaussian effects are expected to arise from the non
linear gravitational collapse of structure, and may be important at small scales [15].
The power spectrum measures the degree of inhomogeneity in the mass distribution on different
scales. It depends upon a few basic ingredientes: a) the primordial spectrumof inhomogeneities, whether
they are Gaussian or nonGaussian, whether adiabatic (perturbations in the energy density) or isocur
vature (perturbations in the entropy density), whether the primordial spectrum has tilt (deviations from
scaleinvariance), etc.; b) the recent creation of inhomogeneities, whether cosmic strings or some other
topological defect from an early phase transition are responsible for the formation of structure today; and
c) the cosmic evolution of the inhomogeneity, whether the universe has been dominated by cold or hot
dark matter or by a cosmological constant since the beginning of structure formation, and also depending
on the rate of expansion of the universe.
The working tools used for the comparison between the observed power spectrum and the pre
dicted one are very precise Nbody numerical simulations and theoretical models that predict the shape
but not the amplitude of the present power spectrum. Even though a large amount of work has gone
into those analyses, we still have large uncertainties about the nature and amount of matter necessary for
structure formation. A model that has become a working paradigm is a ﬂat cold dark matter model with
a cosmological constant and Ω
M
= 0.3 − 0.4. This model will soon be confronted with very precise
measurements from SDSS, 2dF, and several other large redshift catalogs, that are already taking data, see
Section 4.5.
The observational constraints on the power spectrum have a huge lever arm of measurements
at very different scales, mainly from the observed cluster abundance, on 10 Mpc scales, to the CMB
ﬂuctuations, on 1000 Mpc scales, which determines the normalization of the spectrum. At present, deep
redshift surveys are probing scales between 100 and 1000 Mpc, which should begin to see the turnover
corresponding to the peak of the power spectrumat k
eq
, see Figs. 8 and 9. The standard CDM model with
Ω
M
= 1, normalized to the CMB ﬂuctuations on large scales, is inconsistent with the cluster abundance.
The power spectra of both a ﬂat model with a cosmological constant or an open universe with Ω
M
= 0.3
(deﬁned as ΛCDM and OCDM, respectively) can be normalized so that they agree with both the CMB
and cluster observations. In the near future, galaxy survey observations will greatly improve the power
spectrum constraints and will allow a measurement of Ω
M
from the shape of the spectrum. At present,
these measurements suggest a low value of Ω
M
, but with large uncertainties.
3.2.8 Cluster abundance and evolution
Rich clusters are the most recently formed gravitationally bound systems in the universe. Their number
density as a function of time (or redshift) helps determine the amount of dark matter. The observed
present (z ∼ 0) cluster abundance provides a strong constraint on the normalization of the power spec
trumof density perturbations on cluster scales. Both ΛCDM and OCDM are consistent with the observed
cluster abundance at z ∼ 0, see Fig. 20, while Standard CDM (EinsteinDe Sitter model, with Ω
M
= 1),
when normalized at COBE scales, produces too many clusters at all redshifts.
10
12
10
10
10
8
10
6
A
b
u
n
d
a
n
c
e
o
f
m
a
s
s
i
v
e
c
l
u
s
t
e
r
s
(
c
l
u
s
t
e
r
s
p
e
r
[
M
p
c
/
h
]
3
)
Redshift
0
0.5
1.0
ΛCDM
OCDM
SCDM
TCDM
Fig. 20: The evolution of the cluster abundance as a function of redshift, compared with observations from massive clusters.
The four models are normalized to COBE. From Ref. [49].
The evolution of the cluster abundance with redshift breaks the degeneracy among the models
at z ∼ 0. The lowmass models (Open and ΛCDM) predict a relatively small change in the number
density of rich clusters as a function of redshift because, due to the low density, hardly any structure
growth occurs since z ∼ 1. The highmass models (Tilted and Standard CDM) predict that structure has
grown steadily and rich clusters only formed recently: the number density of rich clusters at z ∼ 1 is
predicted to be exponentially smaller than today. The observation of a single massive cluster is enough to
rule out the Ω
M
= 1 model. In fact, three clusters have been seen, suggesting a low density universe [50],
Ω
M
= 0.25
+0.15
−0.10
(1σ statistical) ±20% (systematic) . (81)
But one should be cautious. There is the caveat that for this constraint it is assumed that the initial
spectrum of density perturbations is Gaussian, as predicted in the simplest models of inﬂation, but that
has not yet been conﬁrmed observationally on cluster scales.
Fig. 21: The observed cosmic matter components as functions of the Hubble expansion parameter. The luminous matter
component is given by Eq. (73); the galactic halo component is the horizontal band, Eq. (74), crossing the baryonic component
from BBN, Eq. (38); and the dynamical mass component from large scale structure analysis is given by Eq. (80). Note that in
the range H0 = 70 ± 7 km/s/Mpc, there are three dark matter problems, see the text.
3.2.9 Summary of the matter content
We can summarize the present situation with Fig. 21, for Ω
M
as a function of H
0
. There are four bands,
the luminous matter Ω
lum
; the baryon content Ω
B
, from BBN; the galactic halo component Ω
halo
, and
the dynamical mass from clusters, Ω
M
. From this ﬁgure it is clear that there are in fact three dark matter
problems: The ﬁrst one is where are 90% of the baryons. Between the fraction predicted by BBN and
that seen in stars and diffuse gas there is a huge fraction which is in the form of dark baryons. They could
be in small clumps of hydrogen that have not started thermonuclear reactions and perhaps constitute the
dark matter of spiral galaxies’ halos. Note that although Ω
B
and Ω
halo
coincide at H
0
· 70 km/s/Mpc,
this could be just a coincidence. The second problem is what constitutes 90% of matter, from BBN
baryons to the mass inferred from cluster dynamics. This is the standard dark matter problem and could
be solved by direct detection of a weakly interacting massive particle in the laboratory. And ﬁnally,
since we know from observations of the CMB, see Section 4.4, that the universe is ﬂat, what constitutes
around 60% of the energy density, from dynamical mass to critical density, Ω
0
= 1? One possibility
could be that the universe is dominated by a diffuse vacuum energy, i.e. a cosmological constant, which
only affects the very large scales. Alternatively, the theory of gravity (general relativity) may need to
be modiﬁed on large scales, e.g. due to quantum gravity effects. The need to introduce an effective
cosmological constant on large scales is nowadays the only reason why gravity may need to be modiﬁed
at the quantum level. Since we still do not have a quantum theory of gravity, such a proposal is still very
speculative, and most of the approaches simply consider the inclusion of a cosmological constant as a
phenomenological parameter.
3.2.10 Massive neutrinos
One of the ‘usual suspects’ when addressing the problem of dark matter are neutrinos. They are the only
candidates known to exist. If neutrinos have a mass, could they constitute the missing matter? We know
from the Big Bang theory, see Section 2.2.2, that there is a cosmic neutrino background at a temperature
of approximately 2K. This allows one to compute the present number density in the form of neutrinos,
which turns out to be, for massless neutrinos, n
ν
(T
ν
) =
3
11
n
γ
(T
γ
) = 112 cm
−3
, per species of neutrino.
If neutrinos have mass, as recent experiments seem to suggest, see Fig. 22, the cosmic energy density in
massive neutrinos would be ρ
ν
=
n
ν
m
ν
=
3
11
n
γ
m
ν
, and therefore its contribution today,
Ω
ν
h
2
=
m
ν
94 eV
. (82)
The discussion in the previous Sections suggest that Ω
M
≤ 0.4, and thus, for any of the three families
of neutrinos, m
ν
≤ 40 eV. Note that this limit improves by six orders of magnitude the present bound
on the tauneutrino mass [51]. Supposing that the missing mass in nonbaryonic cold dark matter arises
from a single particle dark matter (PDM) component, its contribution to the critical density is bounded
by 0.05 ≤ Ω
PDM
h
2
≤ 0.4, see Fig. 21.
10
4
10
3
10
2
10
1
10
0
sin
2
2θ
10
11
10
10
10
9
10
8
10
7
10
6
10
5
10
4
10
3
10
2
10
1
10
0
10
1
10
2
10
3
∆
m
2
(
e
V
2
)
Solar
Solar
ν
e
−ν
µ, τ
ν
e
−ν
µ
limit
ν
µ
−ν
τ
limit
BBN
Limit
ν
µ
−ν
s
ν
e
−ν
s
ν
e
 ν
µ,τ
LSND
ν
µ
 ν
e
Atmos
ν
µ
−ν
τ
Solar
ν
e
 ν
µ,τ,s
Cosmologically
Important
Cosmologically
Detectable
Cosmologically
Excluded
Fig. 22: The neutrino parameter space, mixing angle against ∆m
2
, including the results from the different solar and atmospheric
neutrino oscillation experiments. Note the threshold of cosmologically important masses, cosmologically detectable neutrinos
(by CMB and LSS observations), and cosmologically excluded range of masses. From Ref. [52].
I will nowgo through the various logical arguments that exclude neutrinos as the dominant compo
nent of the missing dark matter in the universe. Is it possible that neutrinos with a mass 4 eV ≤ m
ν
≤ 40
eV be the nonbaryonic PDM component? For instance, could massive neutrinos constitute the dark
matter halos of galaxies? For neutrinos to be gravitationally bound to galaxies it is necessary that their
velocity be less that the escape velocity v
esc
, and thus their maximum momentum is p
max
= m
ν
v
esc
.
How many neutrinos can be packed in the halo of a galaxy? Due to the Pauli exclusion principle,
the maximum number density is given by that of a completely degenerate Fermi gas with momen
tum p
F
= p
max
, i.e. n
max
= p
3
max
/3π
2
. Therefore, the maximum local density in dark matter
neutrinos is ρ
max
= n
max
m
ν
= m
4
ν
v
3
esc
/3π
2
, which must be greater than the typical halo density
ρ
halo
= 0.3 GeVcm
−3
. For a typical spiral galaxy, this constraint, known as the TremaineGunn limit,
gives m
ν
≥ 40 eV, see Ref. [53]. However, this mass, even for a single species, say the tauneutrino,
gives a value for Ω
ν
h
2
= 0.5, which is far too high for structure formation. Neutrinos of such a low mass
would constitute a relativistic hot dark matter component, which would washout structure below the su
percluster scale, against evidence from present observations, see Fig. 22. Furthermore, applying the same
phasespace argument to the neutrinos as dark matter in the halo of dwarf galaxies gives m
ν
≥ 100 eV,
beyond closure density (82). We must conclude that the simple idea that light neutrinos could constitute
the particle dark matter on all scales is ruled out. They could, however, still play a role as a subdominant
hot dark matter component in a ﬂat CDM model. In that case, a neutrino mass of order 1 eV is not
cosmological excluded, see Fig. 22.
Another possibility is that neutrinos have a large mass, of order a few GeV. In that case, their num
ber density at decoupling, see Section 2.2.2, is suppressed by a Boltzmann factor, ∼ exp(−m
ν
/T
dec
).
For masses m
ν
> T
dec
· 0.8 MeV, the present energy density has to be computed as a solution
of the corresponding Boltzmann equation. Apart from a logarithmic correction, one ﬁnds Ω
ν
h
2
·
0.1(10 GeV/m
ν
)
2
for Majorana neutrinos and slightly smaller for Dirac neutrinos. In either case, neu
trinos could be the dark matter only if their mass was a few GeV. Laboratory limits for ν
τ
of around 18
MeV [51], and much more stringent ones for ν
µ
and ν
e
, exclude the known light neutrinos. However,
there is always the possibility of a fourth unknown heavy and stable (perhaps sterile) neutrino. If it
couples to the Z boson and has a mass below 45 GeV for Dirac neutrinos (39.5 GeV for Majorana neu
trinos), then it is ruled out by measurements at LEP of the invisible width of the Z. There are two logical
alternatives, either it is a sterile neutrino (it does not couple to the Z), or it does couple but has a larger
mass. In the case of a Majorana neutrino (its own antiparticle), their abundance, for this mass range,
is too small for being cosmologically relevant, Ω
ν
h
2
≤ 0.005. If it were a Dirac neutrino there could
be a lepton asymmetry, which may provide a higher abundance (similar to the case of baryogenesis).
However, neutrinos scatter on nucleons via the weak axialvector current (spindependent) interaction.
For the small momentum transfers imparted by galactic WIMPs, such collisions are essentially coherent
over an entire nucleus, leading to an enhancement of the effective cross section. The relatively large
detection rate in this case allowes one to exclude fourthgeneration Dirac neutrinos for the galactic dark
matter [54]. Anyway, it would be very implausible to have such a massive neutrino today, since it would
have to be stable, with a lifetime greater than the age of the universe, and there is no theoretical reason
to expect a massive sterile neutrino that does not oscillate into the other neutrinos.
Of course, the deﬁnitive test to the possible contribution of neutrinos to the overall density of
the universe would be to measure directly their mass in laboratory experiments.
13
There are at present
two types of experiments: neutrino oscillation experiments, which measure only differences in squared
masses, and direct masssearches experiments, like the tritiumβspectrum and the neutrinoless doubleβ
decay experiments, which measure directly the mass of the electron neutrino and give a bound m
νe ∼
<
2
eV. Neutrinos with such a mass could very well constitute the HDM component of the universe, Ω
HDM ∼
<
0.15. The oscillation experiments give a variety of possibilities for ∆m
2
ν
= 0.3 − 3 eV
2
from LSND
(not yet conﬁrmed), to the atmospheric neutrino oscillations from SuperKamiokande (∆m
2
ν
· 3
10
−3
eV
2
) and the solar neutrino oscillations (∆m
2
ν
· 10
−5
eV
2
). Only the ﬁrst two possibilities would
be cosmologically relevant, see Fig. 22.
13
For a review of Neutrinos, see Bilenky’s contribution to these Proceedings [55].
3.2.11 Weakly Interacting Massive Particles
Unless we drastically change the theory of gravity on large scales, baryons cannot make up the bulk
of the dark matter. Massive neutrinos are the only alternative among the known particles, but they are
essentially ruled out as a universal dark matter candidate, even if they may play a subdominant role as
a hot dark matter component. There remains the mystery of what is the physical nature of the dominant
cold dark matter component.
Something like a heavy stable neutrino, a generic Weakly Interacting Massive Particle (WIMP),
could be a reasonable candidate because its present abundance could fall within the expected range,
Ω
PDM
h
2
∼
G
3/2
T
3
0
h
2
H
2
0
¸σ
ann
v
rel
)
=
3 10
−27
cm
3
s
−1
¸σ
ann
v
rel
)
. (83)
Here v
rel
is the relative velocity of the two incoming dark matter particles and the brackets ¸. . .) denote
a thermal average at the freezeout temperature, T
f
· m
PDM
/20, when the dark matter particles go out
of equilibrium with radiation. The value of ¸σ
ann
v
rel
) needed for Ω
PDM
≈ 1 is remarkably close to
what one would expect for a WIMP with a mass m
PDM
= 100 GeV, ¸σ
ann
v
rel
) ∼ α
2
/8π m
PDM
∼
3 10
−27
cm
3
s
−1
. We still do not know whether this is just a coincidence or an important hint on the
nature of dark matter.
Fig. 23: The maximumlikelihood region from the annualmodulation signal consistent with a neutralino of mass mχ = 59
+17
−14
GeVand a proton cross section of ξσp = 7.0
+0.4
−1.2
×10
−6
pb, see the text. The scatter plot represents the theoretical predictions
of a generic MSSM. From Ref. [56].
There are a few theoretical candidates for WIMPs, like the neutralino, coming from supersymme
tric extensions of the standard model of particle physics,
14
but at present there is no empirical evidence
that such extensions are indeed realized in nature. In fact, the nonobservation of supersymmetric par
ticles at current accelerators places stringent limits on the neutralino mass and interaction cross sec
tion [57].
If WIMPs constitute the dominant component of the halo of our galaxy, it is expected that some
may cross the Earth at a reasonable rate to be detected. The direct experimental search for them rely
14
For a review of Supersymmetry (SUSY), see Carena’s contribution to these Proceedings.
on elastic WIMP collisions with the nuclei of a suitable target. Dark matter WIMPs move at a typical
galactic virial velocity of around 200 − 300 km/s, depending on the model. If their mass is in the
range 10 − 100 GeV, the recoil energy of the nuclei in the elastic collision would be of order 10 keV.
Therefore, one should be able to identify such energy depositions in a macroscopic sample of the target.
There are at present three different methods: First, one could search for scintillation light in NaI crystals
or in liquid xenon; second, search for an ionization signal in a semiconductor, typically a very pure
germanium crystal; and third, use a cryogenic detector at 10 mK and search for a measurable temperature
increase of the sample. The main problem with such a type of experiment is the low expected signal rate,
with a typical number below 1 event/kg/day. To reduce natural radioactive contamination one must
use extremely pure substances, and to reduce the background caused by cosmic rays requires that these
experiments be located deeply underground.
DAMA/
NaI1
DAMA/
NaI2
DAMA/
NaI3
DAMA/
NaI4
Fig. 24: The DAMA experiment sees an annual variation, of order 7%, in the WIMP ﬂux due to the Earth’s motion around the
Sun. The model independent residual rate in the lowest (2 − 6 keV) cumulative energy interval (in counts per day/kg/keV) is
shown as a function of time since 1 January of the ﬁrst year of data taking. The expected behaviour of a WIMP signal is a
cosine function with a minimum (maximum) roughly at the dashed (dotted) vertical lines. From Ref. [56].
The best limits on WIMP scattering cross sections come from some germanium experiments [58],
as well as from the NaI scintillation detectors of the UK dark matter collaboration (UKDMC) in the
Boulby salt mine in England [59], and the DAMA experiment in the Gran Sasso laboratory in Italy [56].
Current experiments already touch the parameter space expected from supersymmetric particles, see
Fig. 23, and therefore there is a chance that they actually discover the nature of the missing dark matter.
The problem, of course, is to attribute a tentative signal unambiguously to galactic WIMPs rather than to
some unidentiﬁed radioactive background.
One speciﬁc signature is the annual modulation which arises as the Earth moves around the Sun.
15
Therefore, the net speed of the Earth relative to the galactic dark matter halo varies, causing a modulation
of the expected counting rate. The DAMA/NaI experiment has actually reported such a modulation sig
nal, see Fig. 24, fromthe combined analysis of their 4year data [56], which provides a conﬁdence level of
99.6% for a neutralino mass of m
χ
= 52
+10
−8
GeV and a proton cross section of ξσ
p
= 7.2
+0.4
−0.9
10
−6
pb, where ξ = ρ
χ
/0.3 GeVcm
−3
is the local neutralino energy density in units of the galactic halo
density. There has been no conﬁrmation yet of this result from other dark matter search groups, but
hopefully in the near future we will have much better sensitivity at low masses from the Cryogenic Rare
Event Search with Superconducting Thermometers (CRESST) experiment at Gran Sasso as well as at
weaker cross sections from the CDMS experiment at Stanford and the Soudan mine, see Fig. 25. The
CRESST experiment [60] uses sapphire crystals as targets and a new method to simultaneously measure
15
The time scale of the Sun’s orbit around the center of the galaxy is too large to be relevant in the analysis.
Fig. 25: Exclusion range for the spinindependent WIMP scattering cross section per nucleon from the NaI experiments and
the Ge detectors. Also shown is the range of expected counting rates for neutralinos in the MSSM. The search goals for the
upcoming largescale cryogenic detectors CRESST and CDMS are also shown. From Ref. [22].
the phonons and the scintillating light from particle interactions inside the crystal, which allows excel
lent background discrimination. Very recently there has been the interesting proposal of a completely
new method based on a Superheated Droplet Detector (SDD), which claims to have already a similar
sensitivity as the more standard methods described above, see Ref. [61].
There exist other indirect methods to search for galactic WIMPs [62]. Such particles could self
annihilate at a certain rate in the galactic halo, producing a potentially detectable background of high
energy photons or antiprotons. The absence of such a background in both gamma ray satellites and the
Alpha Matter Spectrometer [63] imposes bounds on their density in the halo. Alternatively, WIMPs
traversing the solar system may interact with the matter that makes up the Earth or the Sun so that a
small fraction of them will lose energy and be trapped in their cores, building up over the age of the
universe. Their annihilation in the core would thus produce high energy neutrinos from the center of the
Earth or from the Sun which are detectable by neutrino telescopes. In fact, SuperKamiokande already
covers a large part of SUSY parameter space. In other words, neutrino telescopes are already competitive
with direct search experiments. In particular, the AMANDA experiment at the South Pole [64], which
is expected to have 10
3
Cherenkov detectors 2.3 km deep in very clear ice, over a volume ∼ 1 km
3
, is
competitive with the best direct searches proposed. The advantages of AMANDA are also directional,
since the arrays of Cherenkov detectors will allow one to reconstruct the neutrino trajectory and thus its
source, whether it comes from the Earth or the Sun.
3.3 The cosmological constant Ω
Λ
A cosmological constant is a term in the Einstein equations, see Eq. (1), that corresponds to the energy
density of the vacuum of quantum ﬁeld theories, Λ ≡ 8πGρ
v
, see Ref. [65]. These theories predict
a value of order ρ
v
∼ M
4
P
· 5 10
93
g/cm
3
, which is about 123 orders of magnitude larger than
the critical density (14). Such a discrepancy is one of the biggest problems of theoretical physics [66].
It has always been assumed that quantum gravity effects, via some as yet unknown symmetry, would
exactly cancel the cosmological constant, but this remains a downright speculation. Moreover, one of the
difﬁculties with a nonzero value for Λ is that it appears coincidental that we are now living at a special
epoch when the cosmological constant starts to dominate the dynamics of the universe, and that it will do
so forever after, see Section 2.1.2 and Eq. (20). Nevertheless, ever since Einstein introduced it in 1917,
this ethereal constant has been invoked several times in history to explain a number of apparent crises,
always to disappear under further scrutiny [21].
Fig. 26: The Type Ia supernovae observed nearby show a relationship between their absolute luminosity and the timescale of
their light curve: the brighter supernovae are slower and the fainter ones are faster. Asimple linear relation between the absolute
magnitude and a “stretch factor” multiplying the light curve timescale ﬁts the data quite well. From Ref. [68].
In spite of the theoretical prejudice towards Λ = 0, there are new observational arguments for a
nonzero value. The most compelling ones are recent evidence that we live in a ﬂat universe, from obser
vations of CMB anisotropies, together with strong indications of a low mass density universe (Ω
M
< 1),
from the large scale distribution of galaxies, clusters and voids, that indicate that some kind of dark
energy must make up the rest of the energy density up to critical, i.e. Ω
Λ
= 1 − Ω
M
. In addition, the
discrepancy between the ages of globular clusters and the expansion age of the universe may be cleanly
resolved with Λ ,= 0. Finally, there is growing evidence for an accelerating universe from observations
of distant supernovae. I will now discuss the different arguments one by one.
The only known way to reconcile a low mass density with a ﬂat universe is if an additional “dark”
energy dominates the universe today. It would have to resist gravitational collapse, otherwise it would
have been detected already as part of the energy in the halos of galaxies. However, if most of the energy
of the universe resists gravitational collapse, it is impossible for structure in the universe to grow. This
dilemma can be resolved if the hypothetical dark energy was negligible in the past and only recently
became the dominant component. According to general relativity, this requires that the dark energy have
negative pressure, since the ratio of dark energy to matter density goes like a(t)
−3p/ρ
. This argument [67]
would rule out almost all of the usual suspects, such as cold dark matter, neutrinos, radiation, and kinetic
energy, since they all have zero or positive pressure. Thus, we expect something like a cosmological
constant, with negative pressure, p ≈ −ρ, to account for the missing energy.
This negative pressure would help accelerate the universe and reconcile the expansion age of the
universe with the ages of stars in globular clusters, see Fig. 11, where t
0
H
0
is shown as a function
of Ω
M
, in a ﬂat universe, Ω
Λ
= 1 − Ω
M
, and an open one, Ω
Λ
= 0. For the present age of the
universe of t
0
= 13 ± 1 Gyr, and the measured rate of expansion, H
0
= 70 ± 7 km/s/Mpc, one ﬁnds
t
0
H
0
= 0.93 ± 0.12 (adding errors in quadrature), which corresponds to Ω
M
= 0.05
+0.24
−0.10
for an open
universe, see Fig. 11, marginally consistent with observations of large scale structure. On the other hand,
for a ﬂat universe with a cosmological constant, t
0
H
0
= 0.93 ± 0.12 corresponds to Ω
M
= 0.34
+0.20
−0.12
,
which is perfectly compatible with recent observations. These suggest that we probably live in a ﬂat
universe that is accelerating, dominated today by a vacuum energy density.
Calan/Tololo
(Hamuy et al,
A.J. 1996)
Supernova
Cosmology
Project
e
f
f
e
c
t
i
v
e
m
B
(0.5,0.5)
(0, 0)
( 1, 0 ) (1, 0)
(1.5,–0.5) (2, 0)
(Ω
Μ,
Ω
Λ
) =
( 0, 1 )
F
l
a
t
Λ
=
0
redshift z
14
16
18
20
22
24
26
0.02 0.05 0.1 0.2 0.5 1.0 0.02 0.05 0.1 0.2 0.5 1.0
Fig. 27: Hubble diagram for the high redshift supernovae found by the SN Cosmology Project. From Ref. [68]. A similar
diagram is found by the High Redshift Supernova Project [69]. Both groups conclude that distant supernovae are fainter than
expected, and this could be due to an accelerating universe.
This conclusions have been supported by growingly robust observational evidence from distant
supernovae. In their quest for the cosmological parameters, astronomers look for distant astrophysical
objects that can serve as standard candles to determine the distance to the object from their observed
apparent luminosity. A candidate that has recently been exploited with great success is a certain type of
supernova explosions at large redshifts, called SN of type Ia. These are white dwarf stars at the end of
their life cycle that accrete matter from a companion until they become unstable and violently explode
in a natural thermonuclear explosion that outshines their progenitor galaxy. The intensity of the distant
ﬂash varies in time, it takes about three weeks to reach its maximum brightness and then it declines over
a period of months. Although the maximum luminosity varies from one supernova to another, depending
on their original mass, their environment, etc., there is a pattern: brighter explosions last longer than
fainter ones. By studying the characteristic light curves, see Fig. 26, of a reasonably large statistical
sample, cosmologists from two competing groups, the Supernova Cosmology Project [68] and the High
redshift Supernova Project [69], are conﬁdent that they can use this type of supernova as a standard
candle. Since the light coming from some of these rare explosions has travelled for a large fraction of
the size of the universe, one expects to be able to infer from their distribution the spatial curvature and
the rate of expansion of the universe.
One of the surprises revealed by these observations is that high redshift type Ia supernovae appear
fainter than expected for either an open (Ω
M
< 1) or a ﬂat (Ω
M
= 1) universe, see Fig. 27. In fact, the
universe appears to be accelerating instead of decelerating, as was expected from the general attraction
Ω
Μ
No Big Bang
1 2 0 1 2 3
expands forever
Ω
Λ
Flat
Λ = 0
Universe 1
0
1
2
3
2
3
c
l
o
s
e
d
o
p
e
n
90%
68%
99%
95
%
recollapses eventually
f
l
a
t
Ω
Μ
1 2 0 1 2 3
Ω
Λ
1
0
1
2
3
2
3
18.4 Gyr
13.9 Gyr
a
c
c
e
le
r
a
tin
g
d
e
c
e
le
r
a
tin
g
11.5 Gyr
9.2 Gyr
7.4 Gyr
H
0
t
0
65 km s
1
Mpc
1
=
Fig. 28: The left ﬁgure shows the bestﬁt conﬁdence regions (68% – 99% c.l.) in the (ΩM, ΩΛ) plane, for the high redshift
supernovae results. The systematic uncertainty is not shown, and would shift the ellipses vertically. Present observations
disfavour the Eisnteinde Sitter model (circle) by several standarddeviations. The upperleft shadedregion represents “bouncing
universe” cosmologies with no Big Bang in the past. The lowerright shaded region corresponds to a universe that is younger
than the oldest heavy elements, for any value of h ≥ 0.5. The right ﬁgure shows the isochrones of constant H0t0, the age of
the universe in units of the Hubble time, H
−1
0
, with the bestﬁt 68% and 90% conﬁdence regions in the (ΩM, ΩΛ) plane. From
Ref. [68].
of matter, see Eq. (22); something seems to be acting as a repulsive force on very large scales. The
most natural explanation for this is the presence of a cosmological constant, a diffuse vacuum energy
that permeates all space and, as explained above, gives the universe an acceleration that tends to separate
gravitationally bound systems from each other. The bestﬁt results from the Supernova Cosmology
Project give a linear combination 0.8Ω
M
−0.6Ω
Λ
= −0.2±0.1 (1σ), and, for a ﬂat universe (Ω
M
+Ω
Λ
=
1), the bestﬁt values for the combined analysis of both groups [68, 69], are
Ω
ﬂat
M
= 0.28
+0.09
−0.08
(1σ statistical)
+0.05
−0.04
(identiﬁed systematics) , (84)
Ω
ﬂat
Λ
= 0.72
+0.08
−0.09
(1σ statistical)
+0.04
−0.05
(identiﬁed systematics) . (85)
However, one may think that it is still premature to conclude that the universe is indeed accel
erating, because of possibly large systematic errors inherent to most cosmological measurements, and
in particular to observations of supernovae at large redshifts. There has been attempts to ﬁnd crucial
systematic effects like evolution, chemical composition dependence, reddening by dust, etc. in the su
pernovae observations that would invalidate the claims, but none of them are now considered as a serious
threat. Perhaps the most critical one today seems to be sampling effects, since the luminosities of the
highredshift supernovae (z ∼ 0.5 − 1.0) are all measured relative to the same set of local supernovae
(z < 0.3). Hence, absolute calibrations, completeness levels, and any other systematic effects related
to both data sets are critical. For instance, the intense efforts to search for highredshift objects have
led to the peculiar situation where the nearby sample, which is used for calibration, is now smaller than
the distant one. Further searches, already underway, for increasing the nearby supernovae sample will
provide an important check.
Moreover, there are bounds on a cosmological constant that come from the statistics of gravita
tional lensing, with two different methods. Gravitational lensing can be due to various accumulations of
matter along the line of sight to the distant light sources. The ﬁrst method uses the abundance of multiply
imaged sources like quasars, lensed by intervening galaxies. The probability of ﬁnding a lensed image
is directly proportional to the number of galaxies (lenses) along the path and thus to the distance to the
source. This distance, for ﬁxed H
0
, increases dramatically for a large value of the cosmological constant:
the age of the universe and the distance to the galaxy become large for Ω
Λ
,= 0 because the universe has
been expanding for a longer time; therefore, more lenses are predicted for Ω
Λ
> 0. Using this method,
an upper limit of
Ω
Λ
< 0.75 (95% c.l.) (86)
has recently been obtained [70], marginally consistent with the supernovae results, but there are caveats
to this powerful method due to uncertainties in the number density and lensing cross section of the lensing
galaxies as well as the distant quasars. A second method is lensing by massive clusters of galaxies, which
produces widely separated lensed images of quasars and distorted images of background galaxies. The
observed statistics, when compared with numerical simulations, rule out the Ω
M
= 1 models and set an
upper bound on the cosmological constant, Ω
Λ
< 0.7, see Ref. [47]. However, this limit is very sensitive
to the resolution of the numerical simulations, which are currently improving.
3.4 The spatial curvature Ω
K
As we will discuss in detail in Section 4.4, observations of the twopoint correlation function of tem
perature anisotropies in the microwave background provide a crucial test for the spatial curvature of the
universe. From those observations one can tell whether the photons that left the last scattering surface,
at redshift z = 1100, have travelled in straight lines, like in a ﬂat universe, or in curved paths, like in
an open one. Very recent observations made by the balloon experiment BOOMERANG suggest that the
universe is indeed spatially ﬂat (Ω
K
= 0) with about 10% accuracy [71],
Ω
0
= Ω
M
+ Ω
Λ
= 1.0 ±0.1 (95% c.l.) (87)
These measuremnts are bound to be improved in the near future, by both balloon experiments and by the
Microwave Anisotropy Probe (MAP) satellite, to be launched by NASA at the end of year 2000 [72].
Furthermore, with the launch in 2007 of Planck satellite [73] we will be able to determine Ω
0
with 1%
accuracy.
3.5 The age of the universe t
0
The universe must be older than the oldest objects it contains. Those are believed to be the stars in the
oldest clusters in the Milky Way, globular clusters. The most reliable ages come from the application
of theoretical models of stellar evolution to observations of old stars in globular clusters. For about 30
years, the ages of globular clusters have remained reasonable stable, at about 15 Gyr [74]. However,
recently these ages have been revised downward [75].
During the 1980s and 1990s, the globular cluster age estimates have improved as both new obser
vations have been made with CCDs, and since reﬁnements to stellar evolution models, including opaci
ties, consideration of mixing, and different chemical abundances have been incorporated [76]. From the
theory side, uncertainties in globular cluster ages come from uncertainties in convection models, opac
ities, and nuclear reaction rates. From the observational side, uncertainties arise due to corrections for
dust and chemical composition. However, the dominant source of systematic errors in the globular clus
ter age is the uncertainty in the cluster distances. Fortunately, the Hipparcos satellite recently provided
geometric parallax measurements for many nearby old stars with lowmetallicity, typical of glubular clus
ters, thus allowing for a new calibration of the ages of stars in globular clusters, leading to a downward
revision to 10 − 13 Gyr [76]. Moreover, there were very few stars in the Hipparcos catalog with both
small parallax erros and low metal abundance. Hence, an increase in the sample size could be critical in
reducing the statatistical uncertaintites for the calibration of the globular cluster ages. There are already
proposed two new parallax satellites, NASA’s Space Interferometry Mission (SIM) and ESA’s mission,
called GAIA, that will give 2 or 3 orders of magnitude more accurate parallaxes than Hipparcos, down
to fainter magnitude limits, for several orders of magnitude more stars. Until larger samples are avail
able, however, distance errors are likely to be the largest source of systematic uncertainty to the globular
cluster age [21].
Fig. 29: The recent estimates of the age of the universe and that of the oldest objects in our galaxy. The last three points
correspond to the combined analysis of 8 different measurements, for h = 0.64, 0.68 and 7.2, which indicates a relatively weak
dependence on h. The age of the Sun is accurately known and is included for reference. Error bars indicate 1σ limits. The
averages of the ages of the Galactic Halo and Disk are shaded in gray. Note that there isn’t a single age estimate more than
2σ away from the average. The result t0 > tgal is logically inevitable, but the standard EdS model does not satisfy this unless
h < 0.55. From Ref. [77].
The supernovae groups can also determine the age of the universe from their high redshift ob
servations. Figure 28 shows that the conﬁdence regions in the (Ω
M
, Ω
Λ
) plane are almost parallel to
the contours of constant age. For any value of the Hubble constant less than H
0
= 70 km/s/Mpc, the
implied age of the universe is greater than 13 Gyr, allowing enough time for the oldest stars in globular
clusters to evolve [76]. Integrating over Ω
M
and Ω
Λ
, the best ﬁt value of the age in Hubbletime units is
H
0
t
0
= 0.93 ± 0.06 or equivalently t
0
= 14.1 ±1.0 (0.65 h
−1
) Gyr [68]. The age would be somewhat
larger in a ﬂat universe: H
0
t
ﬂat
0
= 0.96
+0.09
−0.07
or, equivalently, [68]
t
ﬂat
0
= 14.4
+1.4
−1.1
(0.65 h
−1
) Gyr . (88)
Furthermore, a combination of 8 independent recent measurements: CMB anisotropies, type Ia
SNe, cluster masstolight ratios, cluster abundance evolution, cluster baryon fraction, deuteriumto
hidrogen ratios in quasar spectra, doublelobed radio sources and the Hubble constant, can be used to
determine the present age of the universe [77]. The result is shown in Fig. 29, compared to other recent
determinations. The best ﬁt value for the age of the universe is, according to this analysis, t
0
= 13.4±1.6
Gyr, about a billion years younger than other recent estimates [77].
We can summarize this Section by showing the region in parameter space where we stand nowa
days, thanks to the recent cosmological observations. We have plotted that region in Fig. 30. One could
also superimpose the contour lines corresponding to equal t
0
H
0
lines, as a cross check. It is extraordi
nary that only in the last few months we have been able to reduce the concordance region to where it
stands today, where all the different observations seem to converge. There are still many uncertainties,
mainly systematic; however, those are quickly decreasing and becoming predominantly statistical. In the
0
1
0 0
1
2
1 2
Ω
MATTER
Ω
C
O
S
M
O
L
O
G
I
C
A
L
C
O
N
S
T
A
N
T
3
Range of
Supernova
data
Range of
microwave
background
data
New preferred model
Old standard
model
Range
of cluster
data
R
u
l
e
d
o
u
t
b
y
a
g
e
t
0
<
9
G
y
r
Constant
expansion
Asymptote to
Einstein's original
static model
A
c
c
e
le
ra
tin
g
D
e
c
e
le
ra
tin
g
Steady Expansion
Recollapse
C
l
o
s
e
d O
p
e
n
F
l
a
t
Fig. 30: The concordance region. The sum ΩM + ΩΛ gives the total cosmic energy content and determines the geometry of
spacetime, whether spatially ﬂat, open or closed. Their difference, ΩM/2−ΩΛ, characterizes the relative strength of expansion
and gravity, and determines how the expansion rate changes with time, whether accelerating or decelerating. Furthermore, a
balance between the two densities determines the fate of the universe, whether it will expand forever or recollapse. These three
effects have been probed by recent observations, from large scale structure (cluster data), temperature anisotropies (microwave
background data) and the universe expansion (supernova data). Surprisingly enough, at present all observations seem to lie
within a naroow region in parameter space. The Einsteinde Sitter model is no longer the preferred one. The best model today
is a ﬂat model with a third of the energy density in the form of nonrelativistic matter and two thirds in the form of vacuum
energy or a cosmological constant. From Ref. [78].
near future, with precise observations of the anisotropies in the microwave background temperature and
polarization, to be discussed in Section 4.4, we will be able to reduce those uncertainties to the level of
one percent. This is the reason why cosmologists are so excited and why it is claimed that we live in the
Golden Age of Cosmology.
4 THE INFLATIONARY PARADIGM
The hot Big Bang theory is nowadays a very robust ediﬁce, with many independent observational checks:
the expansion of the universe; the abundance of light elements; the cosmic microwave background; a
predicted age of the universe compatible with the age of the oldest objects in it, and the formation of
structure via gravitational collapse of initially small inhomogeneities. Today, these observations are
conﬁrmed to within a few percent accuracy, and have helped establish the hot Big Bang as the preferred
model of the universe. All the physics involved in the above observations is routinely tested in the
laboratory (atomic and nuclear physics experiments) or in the solar system (general relativity).
However, this theory leaves a range of crucial questions unanswered, most of which are initial
conditions’ problems. There is the reasonable assumption that these cosmological problems will be
solved or explained by new physical principles at high energies, in the early universe. This assumption
leads to the natural conclusion that accurate observations of the present state of the universe may shed
light onto processes and physical laws at energies above those reachable by particle accelerators, present
or future. We will see that this is a very optimistic approach indeed, and that there are many unresolved
issues related to those problems. However, there might be in the near future reasons to be optimistic.
4.1 Shortcomings of Big Bang Cosmology
The Big Bang theory could not explain the origin of matter and structure in the universe; that is, the
origin of the matter–antimatter asymmetry, without which the universe today would be ﬁlled by a uniform
radiation continuosly expanding and cooling, with no traces of matter, and thus without the possibility
to form gravitationally bound systems like galaxies, stars and planets that could sustain life. Moreover,
the standard Big Bang theory assumes, but cannot explain, the origin of the extraordinary smoothness
and ﬂatness of the universe on the very large scales seen by the microwave background probes and the
largest galaxy catalogs. It cannot explain the origin of the primordial density perturbations that gave rise
to cosmic structures like galaxies, clusters and superclusters, via gravitational collapse; the quantity and
nature of the dark matter that we believe holds the universe together; nor the origin of the Big Bang itself.
A summary [79] of the problems that the Big Bang theory cannot explain is:
• The global structure of the universe.
 Why is the universe so close to spatial ﬂatness?
 Why is matter so homogeneously distributed on large scales?
• The origin of structure in the universe.
 How did the primordial spectrum of density perturbations originate?
• The origin of matter and radiation.
 Where does all the energy in the universe come from?
 What is the nature of the dark matter in the universe?
 How did the matterantimatter asymmetry arise?
• The initial singularity.
 Did the universe have a beginning?
 What is the global structure of the universe beyond our observable patch?
Let me discuss one by one the different issues:
4.1.1 The Flatness Problem
The Big Bang theory assumes but cannot explain the extraordinary spatial ﬂatness of our local patch of
the universe. In the general FRW metric (2) the parameter K that characterizes spatial curvature is a free
parameter. There is nothing in the theory that determines this parameter a priori. However, it is directly
related, via the Friedmann equation (8), to the dynamics, and thus the matter content, of the universe,
K =
8πG
3
ρa
2
−H
2
a
2
=
8πG
3
ρa
2
_
Ω −1
Ω
_
. (89)
We can therefore deﬁne a new variable,
x ≡
Ω −1
Ω
=
const.
ρa
2
, (90)
whose time evolution is given by
x
=
dx
dN
= (1 + 3ω) x , (91)
where N = ln(a/a
i
) characterizes the number of efolds of universe expansion (dN = Hdt) and
where we have used Eq. (30) for the time evolution of the total energy, ρa
3
, which only depends on the
barotropic ratio ω. It is clear from Eq. (91) that the phasespace diagram (x, x
) presents an unstable
critical (saddle) point at x = 0 for ω > −1/3, i.e. for the radiation (ω = 1/3) and matter (ω = 0) eras.
A small perturbation from x = 0 will drive the system towards x = ±∞. Since we know the universe
went through both the radiation era (because of primordial nucleosynthesis) and the matter era (because
of structure formation), tiny deviations from Ω = 1 would have grown since then, such that today
x
0
=
Ω
0
−1
Ω
0
= x
in
_
T
in
T
eq
_
2
(1 + z
eq
) . (92)
In order that today’s value be in the range 0.1 < Ω
0
< 1.2, or x
0
≈ O(1), it is required that at, say,
primordial nucleosynthesis (T
NS
· 10
6
T
eq
) its value be
Ω(t
NS
) = 1 ±10
−15
, (93)
which represents a tremendous ﬁnetuning. Perhaps the universe indeed started with such a peculiar
initial condition, but it is epistemologically more satisfying if we give a fundamental dynamical reason
for the universe to have started so close to spatial ﬂatness. These arguments were ﬁrst used by Robert
Dicke in the 1960s, much before inﬂation. He argued that the most natural initial condition for the
spatial curvature should have been the Planck scale curvature,
(3)
R = 6K/l
2
P
, where the Planck length
is l
P
= (¯ hG/c
3
)
1/2
= 1.62 10
−33
cm, that is, 60 orders of magnitude smaller than the present size
of the universe, a
0
= 1.38 10
28
cm. A universe with this immense curvature would have collapsed
within a Planck time, t
P
= (¯ hG/c
5
)
1/2
= 5.39 10
−44
s, again 60 orders of magnitude smaller than
the present age of the universe, t
0
= 4.1 10
17
s. Therefore, the ﬂatness problem is also related to the
Age Problem, why is it that the universe is so old and ﬂat when, under ordinary circumstances (based on
the fundamental scale of gravity) it should have lasted only a Planck time and reached a size of order the
Planck length? As we will see, inﬂation gives a dynamical reason to such a peculiar initial condition.
4.1.2 The Homogeneity Problem
An expanding universe has particle horizons, that is, spatial regions beyond which causal communica
tion cannot occur. The horizon distance can be deﬁned as the maximum distance that light could have
travelled since the origin of the universe [8],
d
H
(t) ≡ a(t)
_
t
0
dt
a(t
)
∼ H
−1
(t) , (94)
which is proportional to the Hubble scale.
16
For instance, at the beginning of nucleosynthesis the horizon
distance is a few lightseconds, but grows linearly with time and by the end of nucleosynthesis it is a
few lightminutes, i.e. a factor 100 larger, while the scale factor has increased only a factor of 10. The
fact that the causal horizon increases faster, d
H
∼ t, than the scale factor, a ∼ t
1/2
, implies that at any
given time the universe contains regions within itself that, according to the Big Bang theory, were never
in causal contact before. For instance, the number of causally disconnected regions at a given redshift z
present in our causal volume today, d
H
(t
0
) ≡ a
0
, is
N
CD
(z) ∼
_
a(t)
d
H
(t)
_
3
· (1 + z)
3/2
, (95)
which, for the time of decoupling, is of order N
CD
(z
dec
) ∼ 10
5
¸1.
This phenomenon is particularly acute in the case of the observed microwave background. Infor
mation cannot travel faster than the speed of light, so the causal region at the time of photon decoupling
could not be larger than d
H
(t
dec
) ∼ 3 10
5
light years across, or about 1
◦
projected in the sky today. So
why should regions that are separated by more than 1
◦
in the sky today have exactly the same tempera
ture, to within 10 ppm, when the photons that come from those two distant regions could not have been
in causal contact when they were emitted? This constitutes the socalled horizon problem, see Fig. 31,
and was ﬁrst discussed by Robert Dicke in the 1970s as a profound inconsistency of the Big Bang theory.
4.2 Cosmological Inﬂation
In the 1980s, a new paradigm, deeply rooted in fundamental physics, was put forward by Alan H.
Guth [81], Andrei D. Linde [82] and others [83, 84, 85], to address these fundamental questions. Ac
cording to the inﬂationary paradigm, the early universe went through a period of exponential expansion,
16
For the radiation era, the horizon distance is equal to the Hubble scale. For the matter era it is twice the Hubble scale.
T1
T1 = T2
T2
T
dec
= 0.3 eV
Our Hubble
radius at
decoupling
T
0
= 3 K
Universe
expansion
(z = 1100)
Our
observable
universe
today
Fig. 31: Perhaps the most acute problemof the Big Bang theory is explaining the extraordinary homogeneity and isotropy of the
microwave background, see Fig. 6. At the time of decoupling, the volume that gave rise to our present universe contained many
causally disconnected regions (top ﬁgure). Today we observe a blackbody spectrum of photons coming from those regions and
they appear to have the same temperature, T1 = T2, to one part in 10
5
. Why is the universe so homogeneous? This constitutes
the socalled horizon problem, which is spectacularly solved by inﬂation. From Ref. [80, 78].
driven by the approximately constant energy density of a scalar ﬁeld called the inﬂaton. In modern
physics, elementary particles are represented by quantum ﬁelds, which resemble the familiar electric,
magnetic and gravitational ﬁelds. A ﬁeld is simply a function of space and time whose quantum oscil
lations are interpreted as particles. In our case, the inﬂaton ﬁeld has, associated with it, a large potential
energy density, which drives the exponential expansion during inﬂation, see Fig. 32. We know from gen
eral relativity that the density of matter determines the expansion of the universe, but a constant energy
density acts in a very peculiar way: as a repulsive force that makes any two points in space separate at
exponentially large speeds. (This does not violate the laws of causality because there is no information
carried along in the expansion, it is simply the stretching of spacetime.)
This superluminal expansion is capable of explaining the large scale homogeneity of our observ
able universe and, in particular, why the microwave background looks so isotropic: regions separated
today by more than 1
◦
in the sky were, in fact, in causal contact before inﬂation, but were stretched to
cosmological distances by the expansion. Any inhomogeneities present before the tremendous expansion
would be washed out. This explains why photons from supposedly causally disconneted regions have
actually the same spectral distribution with the same temperature, see Fig. 31.
Moreover, in the usual Big Bang scenario a ﬂat universe, one in which the gravitational attraction
of matter is exactly balanced by the cosmic expansion, is unstable under perturbations: a small deviation
from ﬂatness is ampliﬁed and soon produces either an empty universe or a collapsed one. As we dis
cussed above, for the universe to be nearly ﬂat today, it must have been extremely ﬂat at nucleosynthesis,
deviations not exceeding more than one part in 10
15
. This extreme ﬁne tuning of initial conditions was
also solved by the inﬂationary paradigm, see Fig. 33. Thus inﬂation is an extremely elegant hypothesis
that explains how a region much, much greater that our own observable universe could have become
Inflation
reheating
the universe
end
inflation
inflaton
field
P
o
t
e
n
t
i
a
l
e
n
e
r
g
y
Value of
Fig. 32: The inﬂaton ﬁeld can be represented as a ball rolling down a hill. During inﬂation, the energy density is approximately
constant, driving the tremendous expansion of the universe. When the ball starts to oscillate around the bottom of the hill,
inﬂation ends and the inﬂaton energy decays into particles. In certain cases, the coherent oscillations of the inﬂaton could
generate a resonant production of particles which soon thermalize, reheating the universe. From Ref. [78].
smooth and ﬂat without recourse to ad hoc initial conditions. Furthermore, inﬂation dilutes away any
“unwanted” relic species that could have remained from early universe phase transitions, like monopoles,
cosmic strings, etc., which are predicted in grand uniﬁed theories and whose energy density could be so
large that the universe would have become unstable, and collapsed, long ago. These relics are diluted by
the superluminal expansion, which leaves at most one of these particles per causal horizon, making them
harmless to the subsequent evolution of the universe.
The only thing we know about this peculiar scalar ﬁeld, the inﬂaton, is that it has a mass and
a selfinteraction potential V (φ) but we ignore everything else, even the scale at which its dynamics
determines the superluminal expansion. In particular, we still do not know the nature of the inﬂaton ﬁeld
itself, is it some new fundamental scalar ﬁeld in the electroweak symmetry breaking sector, or is it just
some effective description of a more fundamental high energy interaction? Hopefully, in the near future,
experiments in particle physics might give us a clue to its nature. Inﬂation had its original inspiration in
the Higgs ﬁeld, the scalar ﬁeld supposed to be responsible for the masses of elementary particles (quarks
and leptons) and the breaking of the electroweak symmetry. Such a ﬁeld has not been found yet, and its
discovery at the future particle colliders would help understand one of the truly fundamental problems in
physics, the origin of masses. If the experiments discover something completely new and unexpected, it
would automatically affect the idea of inﬂation at a fundamental level.
4.2.1 Homogeneous scalar ﬁeld dynamics
In this subsection I will describe the theoretical basis for the phenomenon of inﬂation. Consider a scalar
ﬁeld φ, a singlet under any given interaction, with an effective potential V (φ). The Lagrangian for such
a ﬁeld in a curved background is
L
inf
=
1
2
g
µν
∂
µ
φ∂
ν
φ −V (φ) , (96)
Fig. 33: The exponential expansion during inﬂation made the radius of curvature of the universe so large that our observable
patch of the universe today appears essentialy ﬂat, analogous (in three dimensions) to how the surface of a balloon appears
ﬂatter and ﬂatter as we inﬂate it to enormous sizes. This is a crucial prediction of cosmological inﬂation that will be tested to
extraordinary accuracy in the next few years. From Ref. [84, 78].
whose evolution equation in a FriedmannRobertsonWalker metric (2) and for a homogeneous ﬁeld φ(t)
is given by
¨
φ + 3H
˙
φ + V
(φ) = 0 , (97)
where H is the rate of expansion, together with the Einstein equations,
H
2
=
κ
2
3
_
1
2
˙
φ
2
+V (φ)
_
, (98)
˙
H = −
κ
2
2
˙
φ
2
, (99)
where κ
2
≡ 8πG. The dynamics of inﬂation can be described as a perfect ﬂuid (7) with a time dependent
pressure and energy density given by
ρ =
1
2
˙
φ
2
+ V (φ) , (100)
p =
1
2
˙
φ
2
− V (φ) . (101)
The ﬁeld evolution equation (97) can then be written as the energy conservation equation,
˙ ρ + 3H(ρ + p) = 0 . (102)
If the potential energy density of the scalar ﬁeld dominates the kinetic energy, V (φ) ¸
˙
φ
2
, then we see
that
p · −ρ ⇒ ρ · const. ⇒ H(φ) · const. , (103)
which leads to the solution
a(t) ∼ exp(Ht) ⇒
¨ a
a
> 0 accelerated expansion. (104)
Using the deﬁnition of the number of efolds, N = ln(a/a
i
), we see that the scale factor grows expo
nentially, a(N) = a
i
exp(N). This solution of the Einstein equations solves immediately the ﬂatness
problem. Recall that the problem with the radiation and matter eras is that Ω = 1 (x = 0) is an unsta
ble critical point in phasespace. However, during inﬂation, with p · −ρ ⇒ ω · −1, we have that
1 + 3ω ≥ 0 and therefore x = 0 is a stable attractor of the equations of motion, see Eq. (91). As a con
sequence, what seemed an ad hoc initial condition, becomes a natural prediction of inﬂation. Suppose
that during inﬂation the scale factor increased N efolds, then
x
0
= x
in
e
−2N
_
T
rh
T
eq
_
2
(1 +z
eq
) · e
−2N
10
56
≤ 1 ⇒ N ≥ 65 , (105)
where we have assumed that inﬂation ended at the scale V
end
, and the transfer of the inﬂaton energy
density to thermal radiation at reheating occurred almost instantaneously
17
at the temperature T
rh
∼
V
1/4
end
∼ 10
15
GeV. Note that we can now have initial conditions with a large uncertainty, x
in
· 1, and
still have today x
0
· 1, thanks to the inﬂationary attractor towards Ω = 1. This can be understood very
easily by realizing that the three curvature evolves during inﬂation as
(3)
R =
6K
a
2
=
(3)
R
in
e
−2N
−→ 0 , for N ¸1 . (106)
Therefore, if cosmological inﬂation lasted over 65 efolds, as most models predict, then today the uni
verse (or at least our local patch) should be exactly ﬂat, see Fig. 33, a prediction that can be tested with
great accuracy in the near future and for which already seems to be some evidence from observations of
the microwave background [71].
Furthermore, inﬂation also solves the homogeneity problem in a spectacular way. First of all, due
to the superluminal expansion, any inhomogeneity existing prior to inﬂation will be washed out,
δ
k
∼
_
k
aH
_
2
Φ
k
∝ e
−2N
−→ 0 , for N ¸1 . (107)
Moreover, since the scale factor grows exponentially, while the horizon distance remains essentially
constant, d
H
(t) · H
−1
= const., any scale within the horizon during inﬂation will be stretched by the
superluminal expansion to enormous distances, in such a way that at photon decoupling all the causally
disconnected regions that encompass our present horizon actually come from a single region during
inﬂation, about 65 efolds before the end. This is the reason why two points separated more than 1
◦
in the sky have the same backbody temperature, as observed by the COBE satellite: they were actually
in causal contact during inﬂation. There is at present no other proposal known that could solve the
homogeneity problem without invoquing an acausal mechanism like inﬂation.
Finally, any relic particle species (relativistic or not) existing prior to inﬂation will be diluted by
the expansion,
ρ
M
∝ a
−3
∼ e
−3N
−→ 0 , for N ¸1 , (108)
ρ
R
∝ a
−4
∼ e
−4N
−→ 0 , for N ¸1 . (109)
Note that the vacuum energy density ρ
v
remains constant under the expansion, and therefore, very soon
it is the only energy density remaining to drive the expansion of the universe.
17
There could be a small delay in thermalization, due to the intrinsic inefﬁciency of reheating, but this does not change
signiﬁcantly the required number of efolds.
4.2.2 The slowroll approximation
In order to simplify the evolution equations during inﬂation, we will consider the slowroll approximation
(SRA). Suppose that, during inﬂation, the scalar ﬁeld evolves very slowly down its effective potential,
then we can deﬁne the slowroll parameters [86],
≡ −
˙
H
H
2
=
κ
2
2
˙
φ
2
H
2
¸ 1 , (110)
δ ≡ −
¨
φ
H
˙
φ
¸ 1 . (111)
It is easy to see that the condition
< 1 ⇐⇒
¨ a
a
> 0 (112)
characterizes inﬂation: it is all you need for superluminal expansion, i.e. for the horizon distance to grow
more slowly than the scale factor, in order to solve the homogeneity problem, as well as for the spatial
curvature to decay faster than usual, in order to solve the ﬂatness problem.
The number of efolds during inﬂation can be written with the help of Eq. (110) as
N = ln
a
end
a
i
=
_
te
t
i
Hdt =
_
φe
φ
i
κdφ
_
2(φ)
, (113)
which is an exact expression in terms of (φ).
In the limit given by Eqs. (110), the evolution equations (97) and (98) become
H
2
_
1 −
3
_
· H
2
=
κ
2
3
V (φ) , (114)
3H
˙
φ
_
1 −
δ
3
_
· 3H
˙
φ = −V
(φ) . (115)
Note that this corresponds to a reduction of the dimensionality of phasespace from two to one dimen
sions, H(φ,
˙
φ) → H(φ). In fact, it is possible to prove a theorem, for singleﬁeld inﬂation, which
states that the slowroll approximation is an attractor of the equations of motion, and thus we can al
ways evaluate the inﬂationary trajectory in phasespace within the SRA, therefore reducing the number
of initial conditions to just one, the initial value of the scalar ﬁeld. If H(φ) only depends on φ, then
H
(φ) = −κ
2
˙
φ/2 and we can rewrite the slowroll parameters (110) as
=
2
κ
2
_
H
(φ)
H(φ)
_
2
·
1
2κ
2
_
V
(φ)
V (φ)
_
2
¸ 1 , (116)
δ =
2
κ
2
H
(φ)
H(φ)
·
1
κ
2
V
(φ)
V (φ)
−
1
2κ
2
_
V
(φ)
V (φ)
_
2
≡ η − ¸ 1 . (117)
The last expression deﬁnes the new slowroll parameter η, not to be confused with conformal time (see
next Section). The number of efolds can also be rewritten in this approximation as
N = κ
2
_
φe
φ
i
V (φ) dφ
V
(φ)
, (118)
a very useful expression for evaluating N for a given effective scalar potential V (φ).
4.3 The origin of density perturbations
If cosmological inﬂation made the universe so extremely ﬂat and homogeneous, where did the galaxies
and clusters of galaxies come from? One of the most astonishing predictions of inﬂation, one that was not
even expected, is that quantumﬂuctuations of the inﬂaton ﬁeld are stretched by the exponential expansion
and generate largescale perturbations in the metric. Inﬂaton ﬂuctuations are small wave packets of
energy that, according to general relativity, modify the spacetime fabric, creating a whole spectrum of
curvature perturbations. The use of the word spectrum here is closely related to the case of light waves
propagating in a medium: a spectrum characterizes the amplitude of each given wavelength. In the
case of inﬂation, the inﬂaton ﬂuctuations induce waves in the spacetime metric that can be decomposed
into different wavelengths, all with approximately the same amplitude, that is, corresponding to a scale
invariant spectrum. These patterns of perturbations in the metric are like ﬁngerprints that unequivocally
characterize a period of inﬂation. When matter fell in the troughs of these waves, it created density
perturbations that collapsed gravitationally to form galaxies, clusters and superclusters of galaxies, with
a spectrum that is also scale invariant. Such a type of spectrum was proposed in the early 1970s (before
inﬂation) by Harrison and Zel’dovich [17], to explain the distribution of galaxies and clusters of galaxies
on very large scales in our observable universe. Perhaps the most interesting aspect of structure formation
is the possibility that the detailed knowledge of what seeded galaxies and clusters of galaxies will allow
us to test the idea of inﬂation.
4.3.1 Gauge invariant perturbation theory
Until now we have considered only the unperturbed FRW metric described by a scale factor a(t) and a
homogeneous scalar ﬁeld φ(t),
ds
2
= a
2
(η)[dη
2
−γ
ij
dx
i
dx
j
] , (119)
φ = φ(η) , (120)
where η =
_
dt/a(t) is the conformal time, under which the background equations of motion can be
written as
H
2
=
κ
2
3
_
1
2
φ
2
+a
2
V (φ)
_
, (121)
H
− H
2
=
κ
2
2
φ
2
, (122)
φ
+ 2Hφ
+ a
2
V
(φ) = 0 , (123)
where H = aH and φ
= a
˙
φ.
During inﬂation, the quantumﬂuctuations of the scalar ﬁeld will induce metric perturbations which
will backreact on the scalar ﬁeld. Let us consider, in linear perturbation theory, the most general line ele
ment with both scalar and tensor metric perturbations [87],
18
together with the scalar ﬁeld perturbations
ds
2
= a
2
(η)
_
(1 + 2A)dη
2
− 2B
i
dx
i
dη −
_
(1 + 2¹)γ
ij
+ 2E
ij
+ 2h
ij
_
dx
i
dx
j
_
, (124)
φ = φ(η) + δφ(η, x
i
) . (125)
The indices ¦i, j¦ label the threedimensional spatial coordinates with metric γ
ij
, and the [i denotes
covariant derivative with respect to that metric. The gauge invariant tensor perturbation h
ij
corresponds
to a transverse traceless gravitational wave, ∇
i
h
ij
= h
i
i
= 0. The four scalar perturbations (A, B, ¹, E)
are gauge dependent functions of (η, x
i
). Under a general coordinate (gauge) transformation [87, 88]
˜ η = η + ξ
0
(η, x
i
) , (126)
˜ x
i
= x
i
+ γ
ij
ξ
j
(η, x
i
) , (127)
18
Note that inﬂation cannot generate, to linear order, a vector perturbation.
with arbitrary functions (ξ
0
, ξ), the scalar and tensor perturbations transform, to linear order, as
˜
A = A− ξ
0
− Hξ
0
,
˜
B = B + ξ
0
−ξ
, (128)
˜
¹ = ¹− Hξ
0
,
˜
E = E − ξ , (129)
˜
h
ij
= h
ij
, (130)
where a prime denotes derivative with respect to conformal time. It is possible to construct, however,
two gaugeinvariant gravitational potentials [87, 88],
Φ = A+ (B − E
)
+H(B −E
) , (131)
Ψ = ¹+ H(B −E
) , (132)
which are related through the perturbed Einstein equations,
Φ = Ψ, (133)
k
2
−3K
a
2
Ψ =
κ
2
2
δρ , (134)
where δρ is the gaugeinvariant density perturbation, and the latter expression is nothing but the Poisson
equation for the gravitational potential, written in relativistic form.
During inﬂation, the energy density is given in terms of a scalar ﬁeld, and thus the gaugeinvariant
equations for the perturbations on comoving hypersurfaces (constant energy density hypersurfaces) are
Φ
+ 3HΦ
+ (H
+H
2
)Φ =
κ
2
2
[φ
δφ
−a
2
V
(φ)δφ] , (135)
−∇
2
Φ + 3HΦ
+ (H
+H
2
)Φ = −
κ
2
2
[φ
δφ
+ a
2
V
(φ)δφ] , (136)
Φ
+ HΦ =
κ
2
2
φ
δφ, (137)
δφ
+ 2Hδφ
− ∇
2
δφ = 4φ
Φ
−2a
2
V
(φ)Φ− a
2
V
(φ)δφ. (138)
This system of equations seem too difﬁcult to solve at ﬁrst sight. However, there is a gauge
invariant combination of variables that allows one to ﬁnd exact solutions. Let us deﬁne [88]
u ≡ aδφ +zΦ, (139)
z ≡ a
φ
H
. (140)
Under this redeﬁnition, the above equations simplify enormously to just three independent equations,
u
− ∇
2
u −
z
z
u = 0 , (141)
∇
2
Φ =
κ
2
2
H
a
2
(zu
− z
u) , (142)
_
a
2
Φ
H
_
=
κ
2
2
zu . (143)
From Equation (141) we can ﬁnd a solution u(z), which substituted into (143) can be integrated to give
Φ(z), and together with u(z) allow us to obtain δφ(z).
4.3.2 Quantum Field Theory in curved spacetime
Until now we have treated the perturbations as classical, but we should in fact consider the perturbations
Φ and δφ as quantum ﬁelds. Note that the perturbed action for the scalar mode u can be written as
δS =
1
2
_
d
3
x dη
_
(u
)
2
−(∇u)
2
+
z
z
u
2
_
. (144)
In order to quantize the ﬁeld u in the curved background deﬁned by the metric (119), we can write the
operator
ˆ u(η, x) =
_
d
3
k
(2π)
3/2
_
u
k
(η) ˆa
k
e
ik·x
+ u
∗
k
(η) ˆa
†
k
e
−ik·x
_
, (145)
where the creation and annihilation operators satisfy the commutation relation of bosonic ﬁelds, and the
scalar ﬁeld’s Fock space is deﬁned through the vacuum condition,
[ˆ a
k
, ˆa
†
k
] = δ
3
(k − k
) , (146)
ˆ a
k
[0) = 0 . (147)
Note that we are not assuming that the inﬂaton is a fundamental scalar ﬁeld, but that is can be written as
a quantum ﬁeld with its commutation relations (as much as a pion can be described as a quantum ﬁeld).
The equations of motion for each mode u
k
(η) are decoupled in linear perturbation theory,
u
k
+
_
k
2
−
z
z
_
u
k
= 0 . (148)
The ratio z
/z acts like a timedependent potential for this Schr¨ odinger like equation. In order to ﬁnd
exact solutions to the mode equation, we will use the slowroll parameters (110), see Ref. [86]
= 1 −
H
H
2
=
κ
2
2
z
2
a
2
, (149)
δ = 1 −
φ
Hφ
= 1 + −
z
Hz
. (150)
In terms of these parameters, the conformal time and the effective potential for the u
k
mode can be
written as
η =
−1
H
+
_
da
aH
, (151)
z
z
= H
2
[(1 + −δ)(2 − δ) +H
−1
(
−δ
)] . (152)
Note that the slowroll parameters, (149) and (150), can be taken as constant,
19
to order
2
,
= 2H( −δ) = O(
2
) , (153)
δ
= Hδ
_
+ δ +
...
φ
H
¨
φ
_
= O(
2
) . (154)
In that case, for constant parameters, we can write
η =
−1
H
1
1 −
, (155)
z
z
=
1
η
2
_
ν
2
−
1
4
_
, where ν =
1 + −δ
1 −
+
1
2
. (156)
19
For instance, there are models of inﬂation, like powerlaw inﬂation, a(t) ∼ t
p
, where = δ = 1/p < 1, that give constant
slowroll parameters.
We are now going to search for approximate solutions of the mode equation (148), where the
effective potential (152) is of order z
/z · 2H
2
in the slowroll approximation. In quaside Sitter there
is a characteristic scale given by the (event) horizon size or Hubble scale during inﬂation, H
−1
. There
will be modes u
k
with physical wavelengths much smaller than this scale, k/a ¸ H, that are well
within the de Sitter horizon and therefore do not feel the curvature of spacetime. On the other hand,
there will be modes with physical wavelengths much greater than the Hubble scale, k/a ¸H. In these
two asymptotic regimes, the solutions can be written as
u
k
=
1
√
2k
e
−ikη
k ¸aH , (157)
u
k
= C
1
z k ¸aH . (158)
In the limit k ¸ aH the modes behave like ordinary quantum modes in Minkowsky spacetime, ap
propriately normalized, while in the opposite limit, u/z becomes constant on superhorizon scales. For
approximately constant slowroll parameters one can ﬁnd exact solutions to (148), with the effective
potential given by (156), that interpolate between the two asymptotic solutions,
u
k
(η) =
√
π
2
e
i(ν+
1
2
)
π
2
(−η)
1/2
H
(1)
ν
(−kη) , (159)
where H
(1)
ν
(z) is the Hankel function of the ﬁrst kind [89], and ν is given by (156) in terms of the
slowroll parameters. In the limit kη →0, the solution becomes
[u
k
[ =
2
ν−
3
2
√
2k
Γ(ν)
Γ(
3
2
)
(−kη)
1
2
−ν
≡
C(ν)
√
2k
_
k
aH
_
ν−
1
2
, (160)
C(ν) = 2
ν−
3
2
Γ(ν)
Γ(
3
2
)
(1 −)
ν−
1
2
· 1 for , δ ¸1 . (161)
We can now compute Φ and δφ from the superHubblescale mode solution (158), for k ¸ aH.
Substituting into Eq. (143), we ﬁnd
Φ = C
1
_
1 −
H
a
2
_
a
2
dη
_
+ C
2
H
a
2
, (162)
δφ =
C
1
a
2
_
a
2
dη −
C
2
a
2
. (163)
The term proportional to C
1
corresponds to the growing solution, while that proportional to C
2
corre
sponds to the decaying solution, which can soon be ignored. These quantities are gauge invariant but
evolve with time outside the horizon, during inﬂation, and before entering again the horizon during the
radiation or matter eras. We would like to write an expression for a gauge invariant quantity that is
also constant for superhorizon modes. Fortunately, in the case of adiabatic perturbations, there is such a
quantity:
ζ ≡ Φ+
1
H
(Φ
+HΦ) =
u
z
, (164)
which is constant, see Eq. (158), for k ¸ aH. In fact, this quantity ζ is identical, for superhorizon
modes, to the gauge invariant curvature metric perturbation ¹
c
on comoving (constant energy density)
hypersurfaces, see Ref. [87, 90],
ζ = ¹
c
+
1
H
2
∇
2
Φ. (165)
Using Eq. (142) we can write the evolution equation for ζ =
u
z
as ζ
=
1
H
∇
2
Φ, which conﬁrms that
ζ is constant for (adiabatic
20
) superhorizon modes, k ¸aH. Therefore, we can evaluate the Newtonian
20
This conservation fails for entropy or isocurvature perturbations, see Ref. [90].
potential Φ
k
when the perturbation reenters the horizon during radiation/matter eras in terms of the
curvature perturbation ¹
k
when it left the Hubble scale during inﬂation,
Φ
k
=
_
1 −
H
a
2
_
a
2
dη
_
¹
k
=
3 + 3ω
5 + 3ω
¹
k
=
_
_
_
2
3
¹
k
radiation era ,
3
5
¹
k
matter era .
(166)
Let us now compute the tensor or gravitational wave metric perturbations generated during inﬂa
tion. The perturbed action for the tensor mode can be written as
δS =
1
2
_
d
3
x dη
a
2
2κ
2
_
(h
ij
)
2
−(∇h
ij
)
2
_
, (167)
with the tensor ﬁeld h
ij
considered as a quantum ﬁeld,
ˆ
h
ij
(η, x) =
_
d
3
k
(2π)
3/2
λ=1,2
_
h
k
(η) e
ij
(k, λ) ˆa
k,λ
e
ik·x
+ h.c.
_
, (168)
where e
ij
(k, λ) are the two polarization tensors, satisfying symmetric, transverse and traceless conditions
e
ij
= e
ji
, k
i
e
ij
= 0 , e
ii
= 0 , (169)
e
ij
(−k, λ) = e
∗
ij
(k, λ) ,
λ
e
∗
ij
(k, λ)e
ij
(k, λ) = 4 , (170)
while the creation and annihilation operators satisfy the usual commutation relation of bosonic ﬁelds,
Eq. (146). We can now redeﬁne our gauge invariant tensor amplitude as
v
k
(η) =
a
√
2κ
h
k
(η) , (171)
which satisﬁes the following evolution equation, decoupled for each mode v
k
(η) in linear perturbation
theory,
v
k
+
_
k
2
−
a
a
_
v
k
= 0 . (172)
The ratio a
/a acts like a timedependent potential for this Schr¨ odinger like equation, analogous to the
term z
/z for the scalar metric perturbation. For constant slowroll parameters, the potential becomes
a
a
= 2H
2
_
1 −
2
_
=
1
η
2
_
µ
2
−
1
4
_
, (173)
µ =
1
1 −
+
1
2
. (174)
We can solve equation (172) in the two asymptotic regimes,
v
k
=
1
√
2k
e
−ikη
k ¸aH , (175)
v
k
= C a k ¸aH . (176)
In the limit k ¸ aH the modes behave like ordinary quantum modes in Minkowsky spacetime, ap
propriately normalized, while in the opposite limit, the metric perturbation h
k
becomes constant on
superhorizon scales. For constant slowroll parameters one can ﬁnd exact solutions to (172), with effec
tive potential given by (173), that interpolate between the two asymptotic solutions. These are identical
to Eq. (159) except for the substitution ν →µ. In the limit kη →0, the solution becomes
[v
k
[ =
C(µ)
√
2k
_
k
aH
_
µ−
1
2
. (177)
Since the mode h
k
becomes constant on superhorizon scales, we can evaluate the tensor metric pertur
bation when it reentered during the radiation or matter era directly in terms of its value during inﬂation.
4.3.3 Power spectrum of scalar and tensor metric perturbations
Not only do we expect to measure the amplitude of the metric perturbations generated during inﬂation
and responsible for the anisotropies in the CMB and density ﬂuctuations in LSS, but we should also be
able to measure its power spectrum, or twopoint correlation function in Fourier space. Let us consider
ﬁrst the scalar metric perturbations ¹
k
, which enter the horizon at a = k/H. Its correlator is given
by [86]
¸0[¹
∗
k
¹
k
[0) =
[u
k
[
2
z
2
δ
3
(k −k
) ≡
T
R
(k)
4πk
3
(2π)
3
δ
3
(k − k
) , (178)
T
R
(k) =
k
3
2π
2
[u
k
[
2
z
2
=
κ
2
2
_
H
2π
_
2
_
k
aH
_
3−2ν
≡ A
2
S
_
k
aH
_
n−1
, (179)
where we have used ¹
k
= ζ
k
=
u
k
z
and Eq. (160). This last equation determines the power spectrum in
terms of its amplitude at horizoncrossing, A
S
, and a tilt,
n − 1 ≡
d lnT
R
(k)
d lnk
= 3 −2ν = 2
_
δ − 2
1 −
_
· 2η − 6 , (180)
see Eqs. (116), (117). Note from this equation that it is possible, in principle, to obtain from inﬂation a
scalar tilt which is either positive (n > 1) or negative (n < 1). Furthermore, depending on the particular
inﬂationary model [91], we can have signiﬁcant departures from scale invariance.
Let us consider now the tensor (gravitational wave) metric perturbation, which enter the horizon
at a = k/H,
λ
¸0[h
∗
k,λ
h
k
,λ
[0) = 4
2κ
2
a
2
[v
k
[
2
δ
3
(k −k
) ≡
T
g
(k)
4πk
3
(2π)
3
δ
3
(k − k
) , (181)
T
g
(k) = 8κ
2
_
H
2π
_
2
_
k
aH
_
3−2µ
≡ A
2
T
_
k
aH
_
n
T
, (182)
where we have used Eqs. (171) and (177). Therefore, the power spectrum can be approximated by a
powerlaw expression, with amplitude A
T
and tilt
n
T
≡
d lnT
g
(k)
d lnk
= 3 −2µ = −
_
2
1 −
_
· −2 < 0 , (183)
which is always negative. In the slowroll approximation, ¸ 1, the tensor power spectrum is scale
invariant.
4.4 The anisotropies of the microwave background
The metric ﬂuctuations generated during inﬂation are not only responsible for the density perturbations
that gave rise to galaxies via gravitational collapse, but one should also expect to see such ripples in
the metric as temperature anisotropies in the cosmic microwave background, that is, minute deviations
in the temperature of the blackbody spectrum when we look at different directions in the sky. Such
anisotropies had been looked for ever since Penzias and Wilson’s discovery of the CMB, but had eluded
all detection, until COBE satellite discovered them in 1992, see Fig. 6. The reason why they took so
long to be discovered was that they appear as perturbations in temperature of only one part in 10
5
. Soon
after COBE, other groups quickly conﬁrmed the detection of temperature anisotropies at around 30 µK,
at higher multipole numbers or smaller angular scales. There are at this moment dozens of ground
and balloonborne experiments analysing the anisotropies in the microwave background with angular
resolutions from 10
◦
to a few arc minutes in the sky, see Fig. 34.
4.4.1 Acoustic oscillations in the plasma
The physics of the CMB anisotropies is relatively simple [92]. The universe just before recombination
is a very tightly coupled ﬂuid, due to the large electromagnetic Thomson cross section (46). Photons
scatter off charged particles (protons and electrons), and carry energy, so they feel the gravitational
potential associated with the perturbations imprinted in the metric during inﬂation. An overdensity of
baryons (protons and neutrons) does not collapse under the effect of gravity until it enters the causal
Hubble radius. The perturbation continues to grow until radiation pressure opposes gravity and sets up
acoustic oscillations in the plasma, very similar to sound waves. Since overdensities of the same size will
enter the Hubble radius at the same time, they will oscillate in phase. Moreover, since photons scatter
off these baryons, the acoustic oscillations occur also in the photon ﬁeld and induces a pattern of peaks
in the temperature anisotropies in the sky, at different angular scales, see Fig. 34.
There are three different effects that determine the temperature anisotropies we observe in the
CMB. First, gravity: photons fall in and escape off gravitational potential wells, characterized by Φin the
comoving gauge, and as a consequence their frequency is gravitationally blue or redshifted, δν/ν = Φ.
If the gravitational potential is not constant, the photons will escape froma larger or smaller potential well
than they fell in, so their frequency is also blue or redshifted, a phenomenon known as the ReesSciama
effect. Second, pressure: photons scatter off baryons which fall into gravitational potential wells and the
two competing forces create acoustic waves of compression and rarefaction. Finally, velocity: baryons
accelerate as they fall into potential wells. They have minimum velocity at maximum compression and
rarefaction. That is, their velocity wave is exactly 90
◦
offphase with the acoustic waves. These waves
induce a Doppler effect on the frequency of the photons.
The temperature anisotropy induced by these three effects is therefore given by [92]
δT
T
(r) = Φ(r, t
dec
) + 2
_
t
0
t
dec
˙
Φ(r, t)dt +
1
3
δρ
ρ
−
r v
c
. (184)
Metric perturbations of different wavelengths enter the horizon at different times. The largest wave
lengths, of size comparable to our present horizon, are entering now. There are perturbations with wave
lengths comparable to the size of the horizon at the time of last scattering, of projected size about 1
◦
in the sky today, which entered precisely at decoupling. And there are perturbations with wavelengths
much smaller than the size of the horizon at last scattering, that entered much earlier than decoupling, all
the way to the time of radiationmatter equality, which have gone through several acoustic oscillations
before last scattering. All these perturbations of different wavelengths leave their imprint in the CMB
anisotropies.
The baryons at the time of decoupling do not feel the gravitational attraction of perturbations with
wavelength greater than the size of the horizon at last scattering, because of causality. Perturbations with
exactly that wavelength are undergoing their ﬁrst contraction, or acoustic compression, at decoupling.
Those perturbations induce a large peak in the temperature anisotropies power spectrum, see Fig. 34.
Perturbations with wavelengths smaller than these will have gone, after they entered the Hubble scale,
through a series of acoustic compressions and rarefactions, which can be seen as secondary peaks in
the power spectrum. Since the surface of last scattering is not a sharp discontinuity, but a region of
∆z ∼ 100, see Fig. 4, there will be scales for which photons, travelling from one energy concentration
to another, will erase the perturbation on that scale, similarly to what neutrinos or HDM do for structure
on small scales. That is the reason why we don’t see all the acoustic oscillations with the same amplitude,
but in fact they decay exponentialy towards smaller angular scales, an effect known as Silk damping, due
to photon diffusion [93, 92].
4.4.2 The SachsWolfe effect
The anisotropies corresponding to large angular scales are only generated via gravitational redshift
and density perturbations through the Einstein equations, δρ/ρ = −2Φ for adiabatic perturbations; we
Fig. 34: There are at present dozens of ground and balloonborne experiments looking at the microwave background tem
perature anisotropies with angular resolutions from 10
◦
to a few arc minutes in the sky, corresponding to multipole numbers
l = 2 − 3000. Present observations suggest the existence of a peak in the angular distribution, as predicted by inﬂation. The
theoretical curve (thick line) illustrates a particular model which ﬁts the data. From Ref. [94].
can ignore the Doppler contribution, since the perturbation is noncausal. In that case, the temperature
anisotropy in the sky today is given by [95]
δT
T
(θ, φ) =
1
3
Φ(η
LS
) Q(η
0
, θ, φ) + 2
_
η
0
η
LS
dr Φ
(η
0
−r) Q(r, θ, φ) , (185)
where η
0
is the coordinate distance to the last scattering surface, i.e. the present conformal time, while
η
LS
· 0 determines that comoving hypersurface. The above expression is known as the SachsWolfe
effect [95], and contains two parts, the intrinsic and the Integrated SachsWolfe (ISW) effect, due to
integration along the line of sight of time variations in the gravitational potential.
In linear perturbation theory, the scalar metric perturbations can be separated into Φ(η, x) ≡
Φ(η) Q(x), where Q(x) are the scalar harmonics, eigenfunctions of the Laplacian in three dimensions,
∇
2
Q
klm
(r, θ, φ) = −k
2
Q
klm
(r, θ, φ). These functions have the general form [96]
Q
klm
(r, θ, φ) = Π
kl
(r) Y
lm
(θ, φ) , (186)
where Y
lm
(θ, φ) are the usual spherical harmonics [89].
In order to compute the temperature anisotropy associated with the SachsWolfe effect, we have
to know the evolution of the metric perturbation during the matter era,
Φ
+ 3HΦ
+a
2
ΛΦ− 2K Φ = 0 . (187)
In the case of a ﬂat universe without cosmological constant, the Newtonian potential remains constant
during the matter era and only the intrinsic SW effect contributes to δT/T. In case of a nonvanishing
Λ, since its contribution is negligible in the past, see Eq. (21), most of the photon’s trajectory towards us
is unperturbed, and the only difference with respect to the Λ = 0 case is an overall factor [66]. We will
consider here the approximation Φ = const. during the matter era and ignore that factor, see Ref. [97].
In a ﬂat universe, the radial part of the eigenfunctions (186) can be written as [96]
Π
kl
(r) =
_
2
π
k j
l
(kr) , (188)
where j
l
(z) are the spherical Bessel functions [89]. The growing mode solutionof the metric perturbation
that left the Hubble scale during inﬂation contributes to the temperature anisotropies on large scales (185)
as
δT
T
(θ, φ) =
1
3
Φ(η
LS
) Q =
1
5
¹Q(η
0
, θ, φ) ≡
∞
l=2
l
m=−l
a
lm
Y
lm
(θ, φ) , (189)
where we have used the fact that at reentry (at the surface of last scattering) the gauge invariant Newtonian
potential Φ is related to the curvature perturbation ¹ at Hubblecrossing during inﬂation, see Eq. (166);
and we have expanded δT/T in spherical harmonics.
We can now compute the twopoint correlation function or angular power spectrum, C(θ), of the
CMB anisotropies on large scales, deﬁned as an expansion in multipole number,
C(θ) =
_
δT
T
∗
(n)
δT
T
(n
)
_
n·n
=cos θ
=
1
4π
∞
l=2
(2l + 1) C
l
P
l
(cos θ) , (190)
where P
l
(z) are the Legendre polynomials [89], and we have averaged over different universe realiza
tions. Since the coefﬁcients a
lm
are isotropic (to ﬁrst order), we can compute the C
l
= ¸[a
lm
[
2
) as
C
(S)
l
=
4π
25
_
∞
0
dk
k
T
R
(k) j
2
l
(kη
0
) , (191)
where we have used Eqs. (189) and (178). In the case of scalar metric perturbation produced during
inﬂation, the scalar power spectrum at reentry is given by T
R
(k) = A
2
S
(kη
0
)
n−1
, in the powerlaw
approximation, see Eq. (179). In that case, one can integrate (191) to give
C
(S)
l
=
2π
25
A
2
S
Γ[
3
2
] Γ[1 −
n−1
2
] Γ[l +
n−1
2
]
Γ[
3
2
−
n−1
2
] Γ[l + 2 −
n−1
2
]
, (192)
l(l + 1) C
(S)
l
2π
=
A
2
S
25
= constant , for n = 1 . (193)
This last expression corresponds to what is known as the SachsWolfe plateau, and is the reason why the
coefﬁcients C
l
are always plotted multiplied by l(l + 1), see Fig. 34.
Tensor metric perturbations also contribute with an approximately constant angular power spec
trum, l(l + 1)C
l
. The SachsWolfe effect for a gauge invariant tensor perturbation is given by [95]
δT
T
(θ, φ) =
_
η
0
η
LS
dr h
(η
0
− r) Q
rr
(r, θ, φ) , (194)
where Q
rr
is the rrcomponent of the tensor harmonic along the line of sight [96]. The tensor perturba
tion h during the matter era satisﬁes the following evolution equation
h
k
+ 3Hh
k
+ (k
2
+ 2K) h
k
= 0 , (195)
which depends on the wavenumber k, contrary to what happens with the scalar modes, see Eq. (187). For
a ﬂat (K = 0) universe, the solution to this equation is h
k
(η) = hG
k
(η), where h is the constant tensor
metric perturbation at horizon crossing and G
k
(η) = 3 j
1
(kη)/kη, normalized so that G
k
(0) = 1 at the
surface of last scattering. The radial part of the tensor harmonic Q
rr
in a ﬂat universe can be written
as [96]
Q
rr
kl
(r) =
_
(l − 1)l(l + 1)(l + 2)
πk
2
_
1/2
j
l
(kr)
r
2
. (196)
The tensor angular power spectrum can ﬁnally be expressed as
C
(T)
l
=
9π
4
(l − 1)l(l + 1)(l + 2)
_
∞
0
dk
k
T
g
(k) I
2
kl
, (197)
I
kl
=
_
x
0
0
dx
j
2
(x
0
−x)j
l
(x)
(x
0
− x)x
2
, (198)
where x ≡ kη, and T
g
(k) is the primordial tensor spectrum (182). For a scale invariant spectrum,
n
T
= 0, we can integrate (197) to give [98]
l(l + 1) C
(T)
l
=
π
36
_
1 +
48π
2
385
_
A
2
T
B
l
, (199)
with B
l
= (1.1184, 0.8789, . . . , 1.00) for l = 2, 3, . . . , 30. Therefore, l(l + 1) C
(T)
l
also becomes
constant for large l. Beyond l ∼ 30, the SachsWolfe expression is not a good approximation and the
tensor angular power spectrum decays very quickly at large l, see Fig.40.
4.4.3 The consistency relation
In spite of the success of inﬂation in predicting a homogeneous and isotropic background on which to
imprint a scaleinvariant spectrum of inhomogeneities, it is difﬁcult to test the idea of inﬂation. A CMB
cosmologist before the 1980s would have argued that ad hoc initial conditions could have been at the
origin of the homogeneity and ﬂatness of the universe on large scales, while a LSS cosmologist would
have agreed with Harrison and Zel’dovich that the most natural spectrumneeded to explain the formation
of structure was a scaleinvariant spectrum. The surprise was that inﬂation incorporated an understanding
of both the globally homogeneous and spatially ﬂat background, and the approximately scaleinvariant
spectrum of perturbations in the same formalism. But that could have been a coincidence, and is not
epistemologically testable.
What is unique to inﬂation is the fact that inﬂation determines not just one but two primordial
spectra, corresponding to the scalar (density) and tensor (gravitational waves) metric perturbations, from
a single continuous function, the inﬂaton potential V (φ). In the slowroll approximation, one determines,
from V (φ), two continuous functions, T
R
(k) and T
g
(k), that in the powerlaw approximation reduces
to two amplitudes, A
S
and A
T
, and two tilts, n and n
T
. It is clear that there must be a relation between
the four parameters. Indeed, one can see from Eqs. (199) and (193) that the ratio of the tensor to scalar
contribution to the angular power spectrum is proportional to the tensor tilt [86],
R ≡
C
(T)
l
C
(S)
l
=
25
9
_
1 +
48π
2
385
_
2 · −2π n
T
. (200)
This is a unique prediction of inﬂation, which could not have been postulated a priori by any cosmol
ogist. If we ﬁnally observe a tensor spectrum of anisotropies in the CMB, or a stochastic gravitational
wave background in laser interferometers like LIGO or VIRGO [99], with sufﬁcient accuracy to deter
mine their spectral tilt, one might have some chance to test the idea of inﬂation, via the consistency
relation (200). For the moment, observations of the microwave background anisotropies suggest that
the SachsWolfe plateau exists, see Fig. 34, but it is still premature to determine the tensor contribution.
Perhaps in the near future, from the analysis of polarization as well as temperature anisotropies, with the
CMB satellites MAP and Planck, we might have a chance of determining the validity of the consistency
relation.
Assuming that the scalar contribution dominates over the tensor on large scales, i.e. R ¸ 1, one
can actually give a measure of the amplitude of the scalar metric perturbation from the observations of
the SachsWolfe plateau in the angular power spectrum [97],
_
l(l + 1) C
(S)
l
2π
_
1/2
=
A
S
5
= (1.03 ±0.07) 10
−5
, (201)
n = 1.02 ± 0.12 . (202)
These measurements can be used to normalize the primordial spectrum and determine the parameters
of the model of inﬂation [91]. In the near future these parameters will be determined with much better
accuracy, as described in Section 4.4.5.
Fig. 35: Theoretical predictions for CMB temperature angular power spectra as a function of multipole number l for models
with primordial adiabatic perturbations. Each graph shows the effect of a variation in one of these parameters. From Ref. [100].
4.4.4 The acoustic peaks
The SachsWolfe plateau is a distinctive feature of Fig. 34. These observations conﬁrm the existence of a
primordial spectrum of scalar (density) perturbations on all scales, otherwise the power spectrum would
have started from zero at l = 2. However, we see that the spectrum starts to rise around l = 20 towards
the ﬁrst acoustic peak, where the SW approximation breaks down and the above formulae are no longer
valid.
As mentioned above, the ﬁrst peak in the photon distribution corresponds to overdensities that
have undergone half an oscillation, that is, a compression, and appear at a scale associated with the size
of the horizon at last scattering, about 1
◦
projected in the sky today. Since photons scatter off baryons,
they will also feel the acoustic wave and create a peak in the correlation function. The height of the peak
is proportional to the amount of baryons: the larger the baryon content of the universe, the higher the
peak. The position of the peak in the power spectrum depends on the geometrical size of the particle
horizon at last scattering. Since photons travel along geodesics, the projected size of the causal horizon
at decoupling depends on whether the universe is ﬂat, open or closed. In a ﬂat universe the geodesics
are straight lines and, by looking at the angular scale of the ﬁrst acoustic peak, we would be measuring
the actual size of the horizon at last scattering. In an open universe, the geodesics are inwardcurved
trajectories, and therefore the projected size on the sky appears smaller. In this case, the ﬁrst acoustic
peak should occur at higher multipoles or smaller angular scales. On the other hand, for a closed universe,
the ﬁrst peak occurs at smaller multipoles or larger angular scales. The dependence of the position of the
ﬁrst acoustic peak on the spatial curvature can be approximately given by [92]
l
peak
· 220 Ω
−1/2
0
, (203)
where Ω
0
= Ω
M
+ Ω
Λ
= 1 − Ω
K
. Present observations, specially the ones of the Mobile Anisotropy
Telescope (MAT) in Cerro Tololo, Chile, which produced two data sets, TOCO97 and TOCO98 [101],
and the recent balloonborne experiment BOOMERANG [71], suggest that the peak is between l = 180
and 250 at 95% c.l., with an amplitude δT = 80 ± 10 µK, and therefore the universe is most probably
ﬂat, see Fig. 36, and Ref. [102]. In particular, these measuremts determine that
0.85 ≤ Ω
0
≤ 1.25 (68% c.l.) (204)
That is, the universe is ﬂat, within 10% uncertainty, which is much better than we could ever do before.
In the near future we will measure Ω
0
to within 1%, with the new microwave anisotropy satellites.
Fig. 36: The left ﬁgure shows the power spectrum of the BOOMERANG experiment with 6 arcminute pixelization. The solid
curve is a marginally closed model with (ΩB, ΩM, ΩΛ, n, h) = (0.05, 0.26, 0.75, 0.95, 0.7). The dotted curve is Standard
CDM with (0.05, 0.95, 0.0, 1.0, 0.65). The dashed curves are open and closed models with ﬁxed Ω0 = ΩM + ΩΛ = 0.66 and
1.55, respectively. The right ﬁgure shows the likelihood function of Ω0 normalized to unity at the peak, after marginalizing
over the ΩM − ΩΛ direction. From Ref. [71].
At the moment there is not enough information at small angular scales, or large multipole numbers,
to determine the existence or not of the secondary acoustic peaks. These peaks should occur at harmonics
of the ﬁrst one, but are typically much lower because of Silk damping. Since the amplitude and position
of the primary and secondary peaks are directly determined by the sound speed (and, hence, the equation
of state) and by the geometry and expansion of the universe, they can be used as a powerful test of the
density of baryons and dark matter, and other cosmological parameters, see Fig. 35.
By looking at these patterns in the anisotropies of the microwave background, cosmologists can
determine not only the cosmological parameters, see Fig. 35, but also the primordial spectrum of den
sity perturbations produced during inﬂation. It turns out that the observed temperature anisotropies are
compatible with a scaleinvariant spectrum, see Eq. (202), as predicted by inﬂation. This is remarkable,
and gives very strong support to the idea that inﬂation may indeed be responsible for both the CMB
anisotropies and the largescale structure of the universe. Different models of inﬂation have different
speciﬁc predictions for the ﬁne details associated with the spectrumgenerated during inﬂation. It is these
minute differences that will allow cosmologists to differentiate between alternative models of inﬂation
and discard those that do not agree with observations. However, most importantly, perhaps, the pattern of
anisotropies predicted by inﬂation is completely different from those predicted by alternative models of
structure formation, like cosmic defects: strings, vortices, textures, etc. These are complicated networks
of energy density concentrations left over from an early universe phase transition, analogous to the de
fects formed in the laboratory in certain kinds of liquid crystals when they go through a phase transition.
The cosmological defects have spectral properties very different from those generated by inﬂation. That
is why it is so important to launch more sensitive instruments, and with better angular resolution, to
determine the properties of the CMB anisotropies.
Fig. 37: The left ﬁgure shows a simulation of the temperature anisotropies predicted by a generic model of inﬂation, as would
be seen by a satellite like COBE with angular resolution of 7
◦
. The right ﬁgure shows the same, but with a satellite like Planck,
with a resolution 100 times better. From Ref. [73].
4.4.5 The new microwave anisotropy satellites, MAP and Planck
The large amount of information encoded in the anisotropies of the microwave background is the rea
son why both NASA and the European Space Agency have decided to launch two independent satel
lites to measure the CMB temperature and polarization anisotropies to unprecendented accuracy. The
Microwave Anisotropy Probe [72] will be launched by NASA at the end of 2000, and Planck [73] is
expected in 2007.
As we have emphasized before, the fact that these anisotropies have such a small amplitude al
low for an accurate calculation of the predicted anisotropies in linear perturbation theory. A particular
cosmological model is characterized by a dozen or so parameters: the rate of expansion, the spatial cur
vature, the baryon content, the cold dark matter and neutrino contribution, the cosmological constant
(vacuum energy), the reionization parameter (optical depth to the last scattering surface), and various
primordial spectrum parameters like the amplitude and tilt of the adiabatic and isocurvature spectra, the
amount of gravitational waves, nonGaussian effects, etc. All these parameters can now be fed into a fast
code called CMBFAST [103] that computes the predicted temperature and polarization anisotropies to
1% accuracy, and thus can be used to compare with observations.
These two satellites will improve both the sensitivity, down to µK, and the resolution, down to arc
minutes, with respect to the previous COBE satellite, thanks to large numbers of microwave horns of var
ious sizes, positioned at speciﬁc angles, and also thanks to recent advances in detector technology, with
high electron mobility transistor ampliﬁers (HEMTs) for frequencies below 100 GHz and bolometers for
higher frequencies. The primary advantage of HEMTs is their ease of use and speed, with a typical sen
sitivity of 0.5 mKs
1/2
, while the advantage of bolometers is their tremendous sensitivity, better than 0.1
mKs
1/2
, see Ref. [104]. For instance, to appreciate the difference, compare the resolution in the temper
ature anisotropies that COBE and Planck would observe for the same simulated sky in Fig. 37. This will
allow cosmologists to extract information from around 3000 multipoles! Since most of the cosmological
parameters have speciﬁc signatures in the height and position of the ﬁrst few acoustic peaks, the higher
the resolution, the more peaks one is expected to see, and thus the better the accuracy with which one
will be able to measure those parameters, see Table 1. As an example of the kind of data that these two
satellites will be able to provide, see Fig. 38, which compares the present observational status with that
which will become available around 2008.
Fig. 38: The predicted angular power spectrum of temperature anisotropies, compared with the present data, see Fig. 34,
binned into 16 logarithmic intervals in multipole number between l = 2 and l = 1000. The right ﬁgure gives an estimate of the
accuracy with which the power spectrum will be measured by Planck. It is only limited by cosmic variance on all the angular
scales relevant to primary anisotropies. From Ref. [105].
Although the satellite probes were designed for the accurate measurement of the CMB tempera
ture anisotropies, there are other experiments, like balloonborne and ground interferometers, which will
probably accomplish the same results with similar resolution (in the case of MAP), before the satellites
start producing their own results [104]. Probably the most important objective of the future satellites will
be the measurement of the CMB polarization anisotropies, yet to be discovered. These anisotropies are
predicted by models of structure formation and are expected to arise at the level of microKelvin sensitivi
ties, where the new satellites are aiming at. The complementary information contained in the polarization
anisotropies will provide much more stringent constraints on the cosmological parameters than from the
temperature anisotropies alone. In particular, the curlcurl component of the polarization power spectra is
nowadays the only means we have to determine the tensor (gravitational wave) contribution to the metric
perturbations responsible for temperature anisotropies, see Fig. 39. If such a component is found, one
could constraint very precisely the model of inﬂation from its spectral properties, specially the tilt [100].
4.5 From metric perturbations to large scale structure
If inﬂation is responsible for the metric perturbations that gave rise to the temperature anisotropies ob
served in the microwave background, then the primordial spectrum of density inhomogeneities induced
by the same metric perturbations should also be responsible for the present large scale structure [106].
This simple connection allows for more stringent tests on the inﬂationary paradigm for the generation
of metric perturbations, since it relates the large scales (of order the present horizon) with the smallest
scales (on galaxy scales). This provides a very large lever arm for the determination of primordial spectra
parameters like the tilt, the nature of the perturbations, whether adiabatic or isocurvature, the geometry
of the universe, as well as its matter and energy content, whether CDM, HDM or mixed CHDM.
10 100 1000
10
11
10
10
10
9
10 100 1000
10
16
10
15
10
14
10
13
10
12
10 100 1000
10
16
10
15
10
14
10
13
10 100 1000
10
15
10
14
10
13
10
12
10
11
l l
l l
l
(
l
+
1
)
C
l
G
G
l
(
l
+
1
)
C
l
C
C
l
(
l
+
1
)
C
l
T
T
l
(
l
+
1
)

C
l
T
G

Fig. 39: Theoretical predictions for the four nonzero CMB temperaturepolarization spectra as a function of multipole mo
ment. The dotted curves are the predictions for a COBEnormalized scalar perturbation from an inﬂationary model with no
reionization and no gravitational waves for h = 0.65, ΩBh
2
= 0.024, and Λ = 0. The solid curves are the corresponding
predictions if the COBE anisotropy were entirely due to a stochastic gravitational wave background with a ﬂat scaleinvariant
spectrum (with the same cosmological parameters). The panel for C
CC
l
contains no dotted curve because scalar perturbations
produce no curl component of the polarization vector. From Ref. [100].
4.5.1 The galaxy power spectrum
As metric perturbations enter the causal horizon during the radiation or matter era, they create density
ﬂuctuations via gravitational attraction of the potential wells. The density contrast δ can be deduced from
the Einstein equations in linear perturbation theory, see Eq. (134),
δ
k
≡
δρ
k
ρ
=
_
k
aH
_
2
2
3
Φ
k
=
_
k
aH
_
2
2 + 2ω
5 + 3ω
¹
k
, (205)
where we have assumed K = 0, and used Eq. (166). From this expression one can compute the power
spectrum, at horizon crossing, of matter density perturbations induced by inﬂation, see Eq. (178),
P(k) = ¸[δ
k
[
2
) = A
_
k
aH
_
n
, (206)
with n given by the scalar tilt (180), n = 1 + 2η − 6. This spectrum reduces to a HarrisonZel’dovich
spectrum (49) in the slowroll approximation: η, ¸1.
Since perturbations evolve after entering the horizon, the power spectrum will not remain con
stant. For scales entering the horizon well after matter domination (k
−1
¸ k
−1
eq
· 81 Mpc), the metric
perturbation has not changed signiﬁcantly, so that ¹
k
(ﬁnal) = ¹
k
(initial). Then Eq. (205) determines
the ﬁnal density contrast in terms of the initial one. On smaller scales, there is a linear transfer function
T(k), which may be deﬁned as [86]
¹
k
(ﬁnal) = T(k) ¹
k
(initial) . (207)
To calculate the transfer function one has to specify the initial condition with the relative abundance
of photons, neutrinos, baryons and cold dark matter long before horizon crossing. The most natural
physical quantity symbol present range MAP Planck
luminous matter Ω
lum
h
2
0.001 −0.005 − −
baryonic matter Ω
B
h
2
0.01 −0.03 5% 0.6%
cold dark matter Ω
M
h
2
0.2 −1.0 10% 0.6%
hot dark matter Ω
ν
h
2
0 − 0.3 5% 2%
cosmological constant Ω
Λ
h
2
0 − 0.8 8% 0.5%
spatial curvature Ω
0
h
2
0.2 −1.5 4% 0.7%
rate of expansion h 0.4 −0.8 11% 2%
age of the universe t
0
11 −17 Gyr 10% 2%
spectral amplitude Q
rms
20 − 30 µK 0.5% 0.1%
spectral tilt n
S
0.5 −1.5 3% 0.5%
tensorscalar ratio r
ts
0 − 1.0 25% 10%
reionization τ 0.01 − 1.0 20% 15%
Table 1: The parameters of the standard cosmological model. The standard model of cosmology has around 12 different
parameters, needed to describe the background spacetime, the matter content and the spectrum of density perturbations. We
include here the present range of the most relevant parameters, and the percentage error with which the microwave background
probes MAP and Planck (without polarization) will be able to determine them in the near future. The rate of expansion is in
units of H0 = 100 h km/s/Mpc.
condition is that the abundances of all particle species are uniform on comoving hypersurfaces (with
constant total energy density). This is called the adiabatic condition, because entropy is conserved inde
pendently for each particle species X, i.e. δρ
X
= ˙ ρ
X
δt, given a perturbation in time from a comoving
hypersurface, so
δρ
X
ρ
X
+p
X
=
δρ
Y
ρ
Y
+ p
Y
, (208)
where we have used the energy conservation equation for each species, ˙ ρ
X
= −3H(ρ
X
+p
X
), valid to
ﬁrst order in perturbations. It follows that each species of radiation has a common density contrast δ
r
,
and each species of matter has also a common density contrast δ
m
, with the relation δ
m
=
3
4
δ
r
.
Within the horizon, the density perturbation amplitude evolves according to the following equa
tion, see Ref. [86],
H
−2
¨
δ
k
+ [2 − 3(2ω −c
2
s
)] H
−1
˙
δ
k
−
3
2
(1 − 6c
2
s
+ 8ω −3ω
2
) δ
k
= −
_
k
aH
_
2
δp
k
ρ
, (209)
where ω = p/ρ is the barotropic ratio, and c
2
s
= ˙ p/ ˙ ρ is the speed of sound of the ﬂuid.
Given the adiabatic condition, the transfer function is determined by the physical processes oc
curing between horizon entry and matter domination. If the radiation behaves like a perfect ﬂuid, its
density perturbation oscillates during this era, with decreasing amplitude. The matter density contrast
living in this background does not grow appreciably before matter domination because it has negligible
selfgravity. The transfer function is therefore given roughly by, see Eq. (52),
T(k) =
_
1 , k ¸k
eq
(k/k
eq
)
2
, k ¸k
eq
(210)
The perfect ﬂuid description of the radiation is far from being correct after horizon entry, because
roughly half of the radiation consists of neutrinos whose perturbation rapidly disappears through free
streeming. The photons are also not a perfect ﬂuid because they diffuse signiﬁcantly, for scales below
the Silk scale, k
−1
S
∼ 1 Mpc. One might then consider the opposite assumption, that the radiation
has zero perturbation after horizon entry. Then the matter density perturbation evolves according to
Eq. (209), with δ and ρ now referring to the matter alone,
¨
δ
k
+ 2H
˙
δ
k
+ (c
2
s
k
2
ph
− 4πGρ) δ
k
= 0 , (211)
which corresponds to the equation of a damped harmonic oscillator. The zerofrequency oscillator deﬁnes
the Jeans wavenumber, k
J
=
_
4πGρ/c
2
s
. For k ¸ k
J
, δ
k
grows exponentially on the dynamical
timescale, τ
dyn
= Imω
−1
= (4πGρ)
−1/2
= τ
grav
, which is the time scale for gravitational collapse.
One can also deﬁne the Jeans length,
λ
J
=
2π
k
J
= c
s
_
π
Gρ
, (212)
which separates gravitationally stable from unstable modes. If we deﬁne the pressure response timescale
as the size of the perturbation over the sound speed, τ
pres
∼ λ/c
s
, then, if τ
pres
> τ
grav
, gravitational
collapse of a perturbation can occur before pressure forces can response to restore hydrostatic equilibrium
(this occurs for λ > λ
J
). On the other hand, if τ
pres
< τ
grav
, radiation pressure prevents gravitational
collapse and there are damped acoustic oscillations (for λ < λ
J
).
We will consider now the behaviour of modes within the horizon during the transition from the
radiation (c
2
s
= 1/3) to the matter era (c
2
s
= 0). The growing and the decaying solutions of Eq. (211) are
δ = A
_
1 +
3
2
y
_
, (213)
δ = B
_
_
1 +
3
2
y
_
ln
√
1 +y + 1
√
1 +y − 1
−3
_
1 +y
_
, (214)
where A and B are constants, and y = a/a
eq
. The growing mode solution (213) increases only by a
factor of 2 between horizon entry and the epoch when matter starts to dominate, i.e. y = 1. The transfer
function is therefore again roughly given by Eq. (210).
Since the radiation consists roughly half of neutrinos, which free streem, and half of photons,
which either form a perfect ﬂuid or just diffuse, neither the perfect ﬂuid nor the freestreeming approx
imation looks very sensible. A more precise calculation is needed, including: neutrino free streeming
around the epoch of horizon entry; the diffusion of photons around the same time, for scales below Silk
scale; the diffusion of baryons along with the photons, and the establishment after matter domination of a
common matter density contrast, as the baryons fall into the potential wells of cold dark matter. All these
effects apply separately, to ﬁrst order in the perturbations, to each Fourier component, so that a linear
transfer function is produced. There are several parametrizations in the literature, but the one which is
more widely used is that of Ref. [107],
T(k) =
_
1 +
_
ak + (bk)
3/2
+ (ck)
2
_
ν
_
−1/ν
, ν = 1.13 , (215)
a = 6.4 (Ω
M
h)
−1
h
−1
Mpc , (216)
b = 3.0 (Ω
M
h)
−1
h
−1
Mpc , (217)
c = 1.7 (Ω
M
h)
−1
h
−1
Mpc . (218)
We see that the behaviour estimated in Eq. (210) is roughly correct, although the break at k = k
eq
is not
at all sharp, see Fig. 40. The transfer function, which encodes the soltion to linear equations, ceases to
be valid when the density contrast becomes of order 1. After that, the highly nonlinear phenomenon of
gravitational collapse takes place, see Fig. 40.
Fig. 40: The CDM power spectrum P(k) as a function of wavenumber k, in logarithmic scale, normalized to the local abun
dance of galaxy clusters, for an Einsteinde Sitter universe with h = 0.5. The solid (dashed) curve shows the linear (nonlinear)
power spectrum. While the linear power spectrum falls off like k
−3
, the nonlinear powerspectrum illustrates the increased
power on small scales due to nonlinear effects, at the expense of the largescale structures. From Ref. [47].
4.5.2 The new redshift catalogs, 2dF and Sloan Digital Sky Survey
Our view of the largescale distributionof luminous objects in the universe has changed dramatically dur
ing the last 25 years [19]: from the simple pre1975 picture of a distribution of ﬁeld and cluster galaxies,
to the discovery of the ﬁrst single superstructures and voids, to the most recent results showing an almost
regular weblike network of interconnected clusters, ﬁlaments and walls, separating huge nearly empty
volumes. The increased efﬁciency of redshift surveys, made possible by the development of spectro
graphs and – specially in the last decade – by an enormous increase in multiplexing gain (i.e. the ability
to collect spectra of several galaxies at once, thanks to ﬁbreoptic spectrographs), has allowed us not
only to do cartography of the nearby universe, but also to statistically characterize some of its properties,
see Ref. [109]. At the same time, advances in theoretical modeling of the development of structure, with
large highresolution gravitational simulations coupled to a deeper yet limited understanding of how to
form galaxies within the dark matter halos, have provided a more realistic connection of the models to the
observable quantities [110]. Despite the large uncertainties that still exist, this has transformed the study
of cosmology and largescale structure into a truly quantitative science, where theory and observations
can progress side by side.
For a review of the variety and details about the different existing redshift catalogs, see Ref. [19],
and Fig. 41. Here I will concentrate on two of the new catalogs, which are taking data at the moment and
which will revolutionize the ﬁeld, the 2degreeField (2dF) Catalog and the Sloan Digital Sky Survey
(SDSS). The advantages of multiobject ﬁbre spectroscopy have been pushed to the extreme with the
construction of the 2dF spectrograph for the prime focus of the AngloAustralian Telescope [111]. This
instrument is able to accommodate 400 automatically positioned ﬁbres over a 2 degree in diameter ﬁeld.
This implies a density of ﬁbres on the sky of approximately 130 deg
−2
, and an optimal match to the
galaxy counts for a magnitude b
J
· 19.5, similar to that of previous surveys like the ESP, with the
difference that with such an area yield, the same number of redshifts as in the ESP survey can be collected
in about 10 exposures, or slightly more than one night of telescope time with typical 1 hour exposures.
This is the basis of the 2dF galaxy redshift survey. Its goal is to measure redshifts for more than 250,000
galaxies with b
J
< 19.5. In addition, a faint redshift survey of 10,000 galaxies brighter than R = 21
will be done over selected ﬁelds within the two main strips of the South and North Galactic Caps. The
survey is steadily collecting redshifts, and there were about 93,000 galaxies measured by January 2000.
See also Ref. [111], where the survey is continuously updated.
CLUSTERS
IRAS
CFA2+SSRS2
LCRS
APM
Fig. 41: Compilation of largescale structure observations, showing the power spectrum P(k) as a function of wavenumber k.
No corrections for bias, redshift distortions, or nonlinear evolution have been made. Some of the redshift surveys have been
rebinned to make the points nearly independent. The black box comes from measurements of σ8 from presentday number
abundances of rich clusters, and the black point with error bars is from peculiar velocities. The height shows the 68%conﬁdence
interval. ΩM = 1 is assumed. The right panel shows a simulation of highprecision future CMB and LSS observations. MAP
(red boxes) and Planck (blue boxes) are simulated assuming that CHDM is the correct model. Green error bars show the
accuracy of the Sloan Digital Sky Survey and magenta error bars are for the 2 Degree Field Survey. No corrections are made
for redshift distortions or nonlinear evolution. The simulated data are indistinguishable from the underlying CHDM model for
a wide range of k. From Ref. [108].
The most ambitious and comprehensive galaxy survey currently in progress is without any doubt
the Sloan Digital Sky Survey [112]. The aim of the project is ﬁrst of all to observe photometrically the
whole Northern Galactic Cap, 30
◦
away from the galactic plane (about 10
4
deg
2
) in ﬁve bands, at limiting
magnitudes from 20.8 to 23.3. The expectation is to detect around 50 million galaxies and around 10
8
starlike sources. This has already led to the discovery of several highredshift (z > 4) quasars, including
the highestredshift quasar known, at z = 5.0, see Ref. [112]. Using two ﬁbre spectrographs carrying
320 ﬁbres each, the spectroscopic part of the survey will then collect spectra from about 10
6
galaxies
with r
< 18 and 10
5
AGNs with r
< 19. It will also select a sample of about 10
5
red luminous galaxies
with r
< 19.5, which will be observed spectroscopically, providing a nearly volumelimited sample
of earlytype galaxies with a median redshift of z · 0.5, that will be extremely valuable to study the
evolution of clustering. The data expected to arise from these new catalogs is so outstanding that already
cosmologists are making simulations and predicting what will be the scientiﬁc outcome of these surveys,
together with the future CMB anisotropy probes, for the determination of the cosmological parameters
of the standard model of cosmology, see Figs. 41 and 42.
As often happens in particle physics, not always are observations from a single experiment sufﬁ
cient to isolate and determine the precise value of the parameters of the standard model. We mentioned in
the previous Section that some of the cosmological parameters created similar effects in the temperature
anisotropies of the microwave background. We say that these parameters are degenerate with respect
to the observations. However, often one ﬁnds combinations of various experiments/observations which
break the degeneracy, for example by depending on a different combination of parameters. This is pre
Fig. 42: Constraint regions in the ΩM−H0 plane from various combinations of data sets. MAP data with polarization yields the
ellipse from upper left to lower right; assuming the universe ﬂat gives a smaller region (shortdashed line). SDSS (kmax = 0.1h
Mpc
−1
) gives the vertical shaded region; combined with MAP gives the small ﬁlled ellipse. A projecton of future supernovae
Ia results gives the solid vertical lines as bounds; combined with MAP gives the solid ellipse. A direct 10% measurement of
H0 gives the longdashed lines and ellipse. All regions are 68% conﬁdence. The ﬁducial model is the ΩM = 0.35 ﬂat ΛCDM
model. The right ﬁgure shows the same as before, but for constraints in the ΩM −ΩΛ plane. From Ref. [113].
cisely the case with the cosmological parameters, as measured by a combination of largescale structure
observations, microwave background anisotropies, Supernovae Ia observations and Hubble Space Tele
scope measurements, a feature named somewhat idiosyncratically as “cosmic complementarity”, see
Ref. [113]. It is expected that in the near future we will be able to determine the parameters of the stan
dard cosmological model with great precision from a combination of several different experiments, as
shown in Fig. 42.
5 CONCLUSION
We have entered a new era in cosmology, were a host of highprecision measurements are already posing
challenges to our understanding of the universe: the density of ordinary matter and the total amount
of energy in the universe; the microwave background anisotropies on a ﬁnescale resolution; primordial
deuteriumabundance from quasar absorption lines; the acceleration parameter of the universe from high
redshift supernovae observations; the rate of expansion from gravitational lensing; large scale structure
measurements of the distribution of galaxies and their evolution; and many more, which already put
constraints on the parameter space of cosmological models, see Fig. 30. However, these are only the
forerunners of the precision era in cosmology that will dominate the new millennium, and will make
cosmology a phenomenological science.
It is important to bear in mind that all physical theories are approximations of reality that can fail
if pushed too far. Physical science advances by incorporating earlier theories that are experimentally
supported into larger, more encompassing frameworks. The standard Big Bang theory is supported by a
wealth of evidence, nobody really doubts its validity anymore. However, in the last decade it has been
incorporated into the larger picture of cosmological inﬂation, which has become the new standard cosmo
logical model. All cosmological issues are now formulated in the context of the inﬂationary paradigm.
It is the best explanation we have at the moment for the increasing set of cosmological observations.
In the next few years we will have an even larger set of highquality observations that will test
inﬂation and the cold dark matter paradigm of structure formation, and determine most of the 12 or
more parameters of the standard cosmological model to a few percent accuracy (see table 1). It may
seem that with such a large number of parameters one can ﬁt almost anything. However, that is not
the case when there is enough quantity and quality of data. An illustrative example is the standard
model of particle physics, with around 21 parameters and a host of precise measurements from particle
accelerators all over the world. This model is, nowadays, rigurously tested, and its parameters measured
to a precision of better than 1% in some cases. It is clear that highprecision measurements will make the
standard model of cosmology as robust as that of particle physics. In fact, it has been the technological
advances of particle physics detectors that are mainly responsible for the burst of new data coming from
cosmological observations. This is deﬁnitely a very healthy ﬁeld, but there is still a lot to do. With
the advent of better and larger precision experiments, cosmology is becoming a mature science, where
speculation has given way to phenomenology.
There are still many unanswered fundamental questions in this emerging picture of cosmology.
For instance, we still do not know the nature of the inﬂaton ﬁeld, is it some new fundamental scalar
ﬁeld in the electroweak symmetry breaking sector, or is it just some effective description of a more
fundamental high energy interaction? Hopefully, in the near future, experiments in particle physics
might give us a clue to its nature. Inﬂation had its original inspiration in the Higgs ﬁeld, the scalar ﬁeld
supposed to be responsible for the masses of elementary particles (quarks and leptons) and the breaking
of the electroweak symmetry. Such a ﬁeld has not been found yet, and its discovery at the future particle
colliders would help understand one of the truly fundamental problems in physics, the origin of masses.
If the experiments discover something completely new and unexpected, it would automatically affect
inﬂation at a fundamental level.
One of the most difﬁcult challenges that the new cosmology will have to face is understanding the
origin of the cosmological constant, if indeed it is conﬁrmed by independent sets of observations. Ever
since Einstein introduced it as a way to counteract gravitational attraction, it has haunted cosmologists
and particle physicists for decades. We still do not have a mechanism to explain its extraordinarily small
value, 120 orders of magnitude below what is predicted by quantum physics. For several decades there
has been the reasonable speculation that this fundamental problem may be related to the quantization of
gravity. General relativity is a classical theory of spacetime, and it has proved particularly difﬁcult to
construct a consistent quantum theory of gravity, since it involves fundamental issues like causality and
the nature of spacetime itself.
The value of the cosmological constant predicted by quantum physics is related to our lack of
understanding of gravity at the microscopic level. However, its effect is dominant at the very largest
scales of clusters or superclusters of galaxies, on truly macroscopic scales. This hints at what is known
in quantum theory as an anomaly, a quantum phenomenon relating both ultraviolet (microscopic) and
infrared (macroscopic) divergences. We can speculate that perhaps general relativity is not the correct
description of gravity on the very largest scales. In fact, it is only in the last few billion years that the
observable universe has become large enough that these global effects could be noticeable. In its infancy,
the universe was much smaller than it is now, and, presumably, general relativity gave a correct descrip
tion of its evolution, as conﬁrmed by the successes of the standard Big Bang theory. As it expanded,
larger and larger regions were encompassed, and, therefore, deviations from general relativity would
slowly become important. It may well be that the recent determination of a cosmological constant from
observations of supernovae at high redshifts is hinting at a fundamental misunderstanding of gravity on
the very large scales.
If this were indeed the case, we should expect that the new generation of precise cosmological
observations will not only affect our cosmological model of the universe but also a more fundamental
description of nature.
ACKNOWLEDGEMENTS
I thank the organizers of the CERNJINR European School of High Energy Physics for a very warm
and friendly atmosphere. I also would like to thank my friends and collaborators Andrei Linde, Andrew
Liddle, David Wands, David Lyth, Jaume Garriga, Xavier Montes, Enrique Gazta˜ naga, Elena Pierpaoli,
Stefano Borgani, and many others, for sharing with me their insight about this fascinating science of
cosmology. This work was supported by the Royal Society.
References
[1] A. Einstein, Sitz. Preuss. Akad. Wiss. Phys. 142 (1917) (¸4); Ann. Phys. 69 (1922) 436.
[2] A. Friedmann, Z. Phys. 10 (1922) 377.
[3] E.P. Hubble, Publ. Nat. Acad. Sci. 15 (1929) 168.
[4] G. Gamow, Phys. Rev. 70 (1946) 572; Phys. Rev. 74 (1948) 505.
[5] A.A. Penzias and R.W. Wilson, Astrophys. J. 142 (1965) 419.
[6] S. Weinberg, “Gravitation and Cosmology”, John Wiley & Sons (1972).
[7] P.B. Pal, “Determination of cosmological parameters”, eprint Archive: hepph/9906447.
[8] E.W. Kolb and M.S. Turner, “The Early Universe”, Addison Wesley (1990).
[9] N. ArkaniHamed, S. Dimopoulos and G. Dvali, Phys. Lett. B 429 (1998) 263, eprint Archive:
hepph/9803315.
[10] S. Burles, K.M. Nollett, J.N. Truran, M.S. Turner, Phys. Rev. Lett. 82 (1999) 4176, eprint Archive:
astroph/9901157; S. Burles, K.M. Nollett, M.S. Turner, “BigBang Nucleosynthesis: Linking In
ner Space and Outer Space”, eprint Archive: astroph/9903300.
[11] K.A. Olive, G. Steigman and T. Walker, “Primordial Nucleosynthesis: Theory and Observations”,
eprint Archive: astroph/9905320.
[12] J.C. Mather et al., Astrophys. J. 420 (1994) 439; D.J. Fixen et al., Astrophys. J. 473 (1996) 576;
J.C. Mather et al., Astrophys. J. 512 (1999) 511, eprint Archive: astroph/9810373.
[13] R.H. Dicke, P.J.E. Peebles, P.G. Roll and D.T. Wilkinson, Astrophys. J. 142 (1965) 414.
[14] G.F. Smoot, Astrophys. J. 396 (1992) L1; C.L. Bennett et al., Astrophys. J. 464 (1996) L1, eprint
Archive: astroph/9601067.
[15] P.J.E. Peebles, “Principles of Physical Cosmology”, Princeton U.P. (1993).
[16] T. Padmanabhan, “Structure Formation in the Universe”, Cambridge U.P. (1993).
[17] E.R. Harrison, Phys. Rev. D 1 (1970) 2726; Ya. B. Zel’dovich, Astron. Astrophys. 5 (1970) 84.
[18] The IRAS Point Source Catalog Web page:
http://wwwastro.physics.ox.ac.uk/˜wjs/pscz.html
[19] L. Guzzo, “Largescale structure at the turn of the millennium”, 19th Texas Symposium, Paris
(1998), eprint Archive: astroph/9911115.
[20] P.J. Steinhardt, in Particle and Nuclear Astrophysics and Cosmology in the Next Millennium, ed.
by E.W. kolb and R. Peccei (World Scientiﬁc, Singapore, 1995).
[21] W.L. Freedman, “Determination of cosmological parameters”, Nobel Symposium (1998), eprint
Archive: hepph/9905222.
[22] G.G. Raffelt, “Dark Matter: Motivation, Candidates and Searches”, European Summer School of
High Energy Physics 1997. CERN Report pp. 235278, eprint Archive: hepph/9712538.
[23] S. Refsdael, Mon. Not. R. Astr. Soc. 128 (1964) 295; 132 (1966) 101.
[24] R.D. Blandford and T. Kundi´ c, “Gravitational Lensing and the Extragalactic Distance Scale”, e
print Archive: astroph/9611229.
[25] N.A. Grogin and R. Narayan, Astrophys. J. 464 (1996) 92, eprint Archive: astroph/9512156.
[26] M. Birkinshaw, Phys. Rep. 310 (1999) 97, eprint Archive: astroph/9808050.
[27] The Chandra Xray observatory Home Page:
http://chandra.harvard.edu/
[28] S. Sakai et al., “The Hubble Space Telescope Key Project on the Extragalactic Distance Scale
XXIV: The Calibration of TullyFisher Relations and the Value of the Hubble Constant”, eprint
Archive: astroph/9909269.
[29] F. Zwicky, Helv. Phys. Acata 6 (1933) 110.
[30] M. Davis and J. Huchra, Astrophys. J. 254 (1982) 437; R.D. Kirshner et al., Astron. J. 88 (1983)
1285.
[31] C.J. Copi and D.N. Schramm, Comm. Nucl. Part. Phys. 22 (1996) 1, eprint Archive: astro
ph/9504026.
[32] K.C. Freeman, Astrophys. J. 160 (1970) 811.
[33] K.G. Begeman, A.H. Broeils and R.H. Sanders, Mon. Not. R. Astr. Soc. 249 (1991) 523.
[34] C.M. Baugh et al., “Ab initio galaxy formation”, eprint Archive: astroph/9907056; Astrophys. J.
498 (1998) 405; eprint Archive: astroph/9703111.
[35] M. Fich and S. Tremaine, Annu. Rev. Astron. Astrophys. 29 (1991) 409.
[36] A.D. Linde, J. Garc´ıaBellido and D. Wands, Phys. Rev. D 54 (1996) 4060. eprint Archive: astro
ph/9605094.
[37] E. Roulet and S. Mollerach, Phys. Rep. 279 (1997) 68, eprint Archive: astroph/9603119.
[38] The MACHO Collaboration, C. Alcock et al., “Microlensing results from 5.7 years of LMC obser
vations”, eprint Archive: astroph/0001272.
[39] B. Paczy´ nski, Astrophys. J. 304 (1986) 1; Astrophys. J. Lett. 371 (1991) L63;
[40] The MACHO Collaboration, C. Alcock et al., Nature 365 (1993) 621; MACHO Home Page at:
http://wwwmacho.mcmaster.ca/
[41] The EROS Collaboration, E. Aubourg et al., Nature 365 (1993) 623; EROS Home Page at:
http://www.lal.in2p3.fr/recherche/eros/erosa.html
[42] The AGAPE Collaboration, P. Baillon et al., A & A 277 (1993) 1; AGAPE Home Page at:
http://cdfinfo.in2p3.fr/Experiences/AGAPE/frameen.html
[43] A. Dekel, Annu. Rev. Astron. Astrophys. 32 (1994) 371, eprint Archive: astroph/9401022.
[44] R.G. Carlberg et al., Astrophys. J. 462 (1996) 32, eprint Archive: astroph/9509034.
[45] A. Dekel, D. Burstein and S.D.M. White, “Measuring Omega”, in Critical Dialogues in Cosmology,
ed. N. Turok, World Scientiﬁc (1997); eprint Archive: astroph/9611108.
[46] C.L. Sarazin, Rev. Mod. Phys. 58 (1986) 1.
[47] M. Bartelmann et al., Astron. & Astrophys. 330 (1998) 1, eprint Archive: astroph/9709229; M.
Bartelmann and P. Schneider, “Weak Gravitational Lensing”, eprint Archive: astroph/9912508.
[48] Hubble Space Telescope Web Page:
http://oposite.stsci.edu/pubinfo/subject.html
[49] N.A. Bahcall, J.P. Ostriker, S. Perlmutter and P.J. Steinhardt, Science 284 (1999) 1481, eprint
Archive: astroph/9906463.
[50] N.A. Bahcall and X. Fan, Proc. Nat. Acad. Sci. 95 (1998) 5956, eprint Archive: astroph/9804082;
Astrophys. J. 504 (1998) 1, eprint Archive: astroph/9803277.
[51] Particle Data Group Home Page, http://www.cern.ch/pdg/1999/lxxx.html
[52] W. Hu, D.J. Eisenstein and M. Tegmark, Phys. Rev. Lett. 80 (1998) 5255, eprint Archive: astro
ph/9712057.
[53] S.D. Tremaine and J.E. Gunn, Phys. Rev. Lett. 42 (1979) 407; J. Madsen, Phys. Rev. D 44 (1991)
999.
[54] J. Primack, D. Seckel and B. Sadoulet, Ann. Rev. Nucl. Part. Sci. 38 (1988) 751; N.E. Booth, B.
Cabrera and E. Fiorini, Ann. Rev. Nucl. Part. Sci. 46 (1996) 471.
[55] S.M. Bilenky, “Neutrino masses, mixing and oscillations”, in these Proceedings; eprint Archive:
hepph/0001311.
[56] R. Bernabei et al., DAMA Collaboration, “Search for WIMP annual modulation signature: re
sults from DAMA/NaI3 and NaI4, and the global combined analysis”, preprint INFN/AE00/01.
DAMA Home Page, http://www.lngs.infn.it/lngs/htexts/dama/
[57] J. Ellis, “Limits on Sparticle Dark Matter”, talk presented at COSMO 98, Asilomar, California,
November 1998, eprint Archive: astroph/9903003.
[58] For a recent review, see A. Morales, “Direct Detection of WIMP Dark Matter”, eprint Archive:
astroph/9912554.
[59] UK Dark Matter Collaboration (J.J. Quenby et al.), “Dark Matter Experiments at the UK Boulby
mine”, at 26th International Cosmic Ray Conference (ICRC 99), in ‘Salt Lake City 1999, Cosmic
ray’, Vol. 2, pp. 269–272. Home Page at http://hepwww.rl.ac.uk/ukdmc/
[60] M. Bravin et al., Astropart. Phys. 12 (1999) 107; eprint Archive: hepex/9904005.
[61] J.I. Collar et al., “First Dark Matter Limits from a LargeMass, LowBackground Superheated
Droplet Detector”, eprint Archive: astroph/0001511.
[62] G. Jungman, M. Kamionkowski and K. Griest, Phys. Rep. 267 (1996) 195, eprint Archive: hep
ph/9506380.
[63] The Alpha Matter Spectrometer Home Page: http://ams.cern.ch/AMS/
[64] F. Halzen et al., Phys. Rep. 307 (1998) 243, eprint Archive: hepex/9804007.
[65] S. Weinberg, Rev. Mod. Phys. 61 (1989) 1.
[66] S.M. Carroll, W.H. Press and E.L. Turner, Ann. Rev. Astron. Astrophys. 30 (1992) 499.
[67] J.P. Ostriker and P.J. Steinhardt, Nature 377 (1995) 600.
[68] S. Perlmutter et al., The Supernova Cosmology Project, Bull. Am. Astron. Soc. 29 (1997) 1351,
eprint Archive: astroph/9812473; Astrophys. J. 517 (1999) 565, eprint Archive: astro
ph/9812133.
[69] A. Reiss et al., High Redshift Supernova Project, Astron. J. 116 (1998) 1009, eprint Archive:
astroph/9805201.
[70] C.S. Kochanek, Astrophys. J. 453 (1995) 545; Astrophys. J. 466 (1995) 638; E.E. Falco, C.S.
Kochanek and J.A. Mu˜ noz, Astrophys. J. 494 (1998) 47; eprint Archive: astroph/9707032.
[71] P. de Bernardis et al., “Mapping the CMB Sky: The BOOMERANG experiment”, eprint Archive:
astroph/9911461; A. Melchiorri et al., “A measurement of Omega from the North American test
ﬂight of BOOMERANG”, eprint Archive: astroph/9911445; P. D. Mauskopf et al.,“Measurement
of a Peak in the Cosmic Microwave Background Power Spectrum from the North American test
ﬂight of BOOMERANG”, eprint Archive: astroph/9911444.
[72] Microwave Anisotropy Probe Home Page: http://map.gsfc.nasa.gov/
[73] Planck Home Page: http://astro.estec.esa.nl/Planck/
[74] D.A. Vandenberg, M. Bolte and P.B. Stetson, Ann. Rev. Astron. Astrophys. 34 (1996) 461; eprint
Archive: astroph/9605064.
[75] L.M. Krauss, “The age of globular clusters”, eprint Archive: astroph/9907308.
[76] B. Chaboyer, P. Demarque, P.J. Kernan and L.M. Krauss, Science 271 (1996) 957; Astrophys. J.
494 (1998) 96; eprint Archive: astroph/9706128.
[77] C.H. Lineweaver, Science 284 (1999) 1503, eprint Archive: astroph/9911493.
[78] J. Garc´ıaBellido, Phil. Trans. R. Soc. Lond. A 357 (1999) 3237, eprint Archive: astro
ph/9906497.
[79] J. Garc´ıaBellido, “Inﬂationary Cosmology”, Vth Autumn School of Theoretical Physics, Santiago
de Compostela (1997).
[80] A. de R´ ujula, “Introduction to Cosmology”, CERN Summer Student Lecture Programme (1997).
[81] A. Guth, Phys. Rev. D 23 (1981) 347.
[82] A.D. Linde, Phys. Lett. 108B (1982) 389.
[83] A. Albrecht and P.J. Steinhardt, Phys. Rev. Lett. 48 (1982) 1220.
[84] For a personal historical account, see A. Guth, “The Inﬂationary Universe”, Perseus Books (1997).
[85] A.D. Linde, “Particle Physics and Inﬂationary Cosmology”, Harwood Academic Press (1990).
[86] A.R. Liddle and D.H. Lyth, Phys. Rep. 231 (1993) 1, eprint Archive: astroph/9303019.
[87] J.M. Bardeen, Phys. Rev. D 22 (1980) 1882.
[88] V.F. Mukhanov, H.A. Feldman and R.H. Brandenberger, Phys. Rep. 215 (1992) 203.
[89] M. Abramowitz and I. Stegun, “Handbook of Mathematical Functions”, Dover (1972).
[90] J. Garc´ıaBellido and D. Wands, Phys. Rev. D 53 (1996) 5437, eprint Archive: astroph/9511029.
[91] D.H. Lyth and A. Riotto, Phys. Rep. 314 (1999) 1, eprint Archive: hepph/9807278.
[92] D. Scott, J. Silk and M. White, Science 268 (1995) 829, eprint Archive: astroph/9505015; W. Hu,
N. Sugiyama and J. Silk, Nature 386 (1997) 37, eprint Archive: astroph/9604166. E. Gawiser and
J. Silk, “The Cosmic Microwave Background Radiation”, eprint Archive: astroph/0002044.
[93] J. Silk, Nature 215 (1967) 1155;
[94] M. Tegmark web page with latest experiments’ data:
http://www.hep.upenn.edu/˜max/cmb/experiments.html
[95] R.K. Sachs and A.M. Wolfe, Astrophys. J. 147 (1967) 73.
[96] E.R. Harrison, Rev. Mod. Phys. 39 (1967) 862; L.F. Abbott and R.K. Schaefer, Astrophys. J. 308
(1986) 546.
[97] E.F. Bunn, A.R. Liddle and M. White, Phys. Rev. D 54 (1996) 5917, eprint Archive: astro
ph/9607038.
[98] A.A. Starobinsky, Sov. Astron. Lett. 11 (1985) 133.
[99] LIGO Home Page: http://www.ligo.caltech.edu/
VIRGO Home Page: http://www.virgo.infn.it/
[100] M. Kamionkowski and A. Kosowsky, Ann. Rev. Nucl. Part. Sci. 49 (1999) 77, eprint Archive:
astroph/9904108.
[101] E. Torbet et al., Astrophys. J. 521 (1999) L79, eprint Archive: astroph/9905100; A. Miller et
al., Astrophys. J. 524 (1999) L1, eprint Archive: astroph/990642.
[102] L. Knox and L. Page, “Characterizing the Peak in the Cosmic Microwave Background Angular
Power Spectrum”, eprint Archive: astroph/0002162.
[103] CMBFAST code Home Page:
http://www.sns.ias.edu/˜matiasz/CMBFAST/cmbfast.html
[104] L.A. Page, “Measuring the anisotropy in the CMB”, eprint Archive: astroph/9911199.
[105] D. Scott, “New physics from the CMB”, eprint Archive: astroph/9911325.
[106] A.R. Liddle and D.H. Lyth, “Cosmological Inﬂation and Large Scale Structure”, Cambridge Uni
versity Press (2000).
[107] J.R. Bond and G. Efstathiou, Astrophys. J. 285 (1984) L45.
[108] E. Gawiser, J. Silk, Science 280 (1998) 1405, eprint Archive: astroph/9806197.
[109] G. Efstathiou et al. (Eds.) “Largescale structure in the universe”, Phil. Trans. R. Soc. Lond. A
357 (1999) 1198.
[110] B. Moore, Phil. Trans. R. Soc. Lond. A 357 (1999) 3259.
[111] The 2 Degree Field Home Page: http://www.mso.anu.edu.au/2dFGRS/
[112] The Sloan Digital Sky Survey Home Page: http://www.sdss.org/sdss.html
[113] D.J. Eisenstein, W. Hu and M. Tegmark, Astrophys. J. 518 (1999) 2, eprint Archive: astro
ph/9807130.
after the Big Bang, when the universe was a few times hotter than the core of the sun. Third, the cosmic microwave background (CMB), the afterglow of the Big Bang, discovered in 1965 by Arno A. Penzias and Robert W. Wilson [5] as a very isotropic blackbody radiation at a temperature of about 3 degrees Kelvin, emitted when the universe was cold enough to form neutral atoms, and photons decoupled from matter, approximately 500,000 years after the Big Bang. Today, these observations are conﬁrmed to within a few percent accuracy, and have helped establish the hot Big Bang as the preferred model of the universe. 2.1 Friedmann–Robertson–Walker universes ˇ Where are we in the universe? During our lectures, of course, we were in Casta Papierniˇ ka, in “the heart c of Europe”, on planet Earth, rotating (8 lightminutes away) around the Sun, an ordinary star 8.5 kpc1 from the center of our galaxy, the Milky Way, which is part of the local group, within the Virgo cluster of galaxies (of size a few Mpc), itself part of a supercluster (of size ∼ 100 Mpc), within the visible universe (∼ few × 1000 Mpc), most probably a tiny homogeneous patch of the inﬁnite global structure of spacetime, much beyond our observable universe. Cosmology studies the universe as we see it. Due to our inherent inability to experiment with it, its origin and evolution has always been prone to wild speculation. However, cosmology was born as a science with the advent of general relativity and the realization that the geometry of spacetime, and thus the general attraction of matter, is determined by the energy content of the universe [6], 1 Gµν ≡ Rµν − gµν R = 8πG Tµν + Λ gµν . 2 (1)
These nonlinear equations are simply too difﬁcult to solve without some insight coming from the symmetries of the problem at hand: the universe itself. At the time (19171922) the known (observed) universe extended a few hundreds of parsecs away, to the galaxies in the local group, Andromeda and the Large and Small Magellanic Clouds: The universe looked extremely anisotropic. Nevertheless, both Einstein and Friedmann speculated that the most “reasonable” symmetry for the universe at large should be homogeneity at all points, and thus isotropy. It was not until the detection, a few decades later, of the microwave background by Penzias and Wilson that this important assumption was ﬁnally put onto ﬁrm experimental ground. So, what is the most general metric satisfying homogeneity and isotropy at large scales? The FriedmannRobertsonWalker (FRW) metric, written here in terms of the invariant geodesic distance ds2 = gµν dxµ dxν in four dimensions, µ = 0, 1, 2, 3, see Ref. [6],2 ds2 = dt2 − a2 (t) dr2 + r2 (dθ2 + sin2 θ dφ2 ) , 1 − K r2 (2)
characterized by just two quantities, a scale factor a(t), which determines the physical size of the universe, and a constant K, which characterizes the spatial curvature of the universe,
(3)
6K R= 2 . a (t)
K = −1
Spatially open, ﬂat and closed universes have different geometries. Light geodesics on these universes behave differently, and thus could in principle be distinguished observationally, as we shall discuss later. Apart from the threedimensional spatial curvature, we can also compute a fourdimensional spacetime curvature, a ¨ a 2 ˙ K (4) R=6 +6 (4) +6 2. a a a
1 2
K = +1
K=0
OPEN FLAT CLOSED
(3)
One parallax second (1 pc), parsec for short, corresponds to a distance of about 3.26 lightyears or 3 × 10 18 cm. I am using c = 1 everywhere, unless speciﬁed.
Depending on the dynamics (and thus on the matter/energy content) of the universe, we will have different possible outcomes of its evolution. The universe may expand for ever, recollapse in the future or approach an asymptotic state in between. 2.1.1 The expansion of the universe In 1929, Edwin P. Hubble observed a redshift in the spectra of distant galaxies, which indicated that they were receding from us at a velocity proportional to their distance to us [3]. This was correctly interpreted as mainly due to the expansion of the universe, that is, to the fact that the scale factor today is larger than when the photons were emitted by the observed galaxies. For simplicity, consider the metric of a spatially ﬂat universe, ds2 = dt2 − a2 (t) dx2 (the generalization of the following argument to curved space is straightforward). The scale factor a(t) gives physical size to the spatial coordinates x, and the expansion is nothing but a change of scale (of spatial units) with time. Except for peculiar velocities, i.e. motion due to the local attraction of matter, galaxies do not move in coordinate space, it is the spacetime fabric which is stretching between galaxies. Due to this continuous stretching, the observed wavelength of photons coming from distant objects is greater than when they were emitted by a factor precisely equal to the ratio of scale factors, λobs a0 = ≡ 1+z, (5) λem a where a0 is the present value of the scale factor. Since the universe today is larger than in the past, the observed wavelengths will be shifted towards the red, or redshifted, by an amount characterized by z, the redshift parameter. In the context of a FRW metric, the universe expansion is characterized by a quantity known as the Hubble rate of expansion, H(t) = a(t)/a(t), whose value today is denoted by H0 . As I shall deduce ˙ later, it is possible to compute the relation between the physical distance dL and the present rate of expansion, in terms of the redshift parameter,3 1 H0 dL = z + (1 − q0 ) z 2 + O(z 3 ) . 2 (6)
At small distances from us, i.e. at z 1, we can safely keep only the linear term, and thus the recession velocity becomes proportional to the distance from us, v = c z = H0 dL, the proportionality constant being the Hubble rate, H0 . This expression constitutes the socalled Hubble law, and is spectacularly conﬁrmed by a huge range of data, up to distances of hundreds of megaparsecs. In fact, only recently measurements from very bright and distant supernovae, at z 1, were obtained, and are beginning to probe the secondorder term, proportional to the deceleration parameter q0 , see Eq. (22). I will come back to these measurements in Section 3. One may be puzzled as to why do we see such a stretching of spacetime. Indeed, if all spatial distances are scaled with a universal scale factor, our local measuring units (our rulers) should also be stretched, and therefore we should not see the difference when comparing the two distances (e.g. the two wavelengths) at different times. The reason we see the difference is because we live in a gravitationally bound system, decoupled from the expansion of the universe: local spatial units in these systems are not stretched by the expansion.4 The wavelengths of photons are stretched along their geodesic path from one galaxy to another. In this consistent world picture, galaxies are like point particles, moving as a ﬂuid in an expanding universe. 2.1.2 The matter and energy content of the universe So far I have only discussed the geometrical aspects of spacetime. Let us now consider the matter and energy content of such a universe. The most general matter ﬂuid consistent with the assumption of
3 4
The subscript L refers to Luminosity, which characterizes the amount of light emitted by an object. See Eq. (61). The local spacetime of a gravitationally bound system is described by the Schwarzschild metric, which is static [6].
pM = 0. that which in the absence of a cosmological constant would correspond to a ﬂat universe. In fact. H0 = 100 h km s−1 Mpc−1 . (7) where p(t) and ρ(t) are the pressure and energy density of the ﬂuid at a given time in the expansion. the Friedmann and the energyconservation equation give the evolution equation for the scale factor.77 h−1 1011 M /(h−1 Mpc)3 . these equations can be deduced from the Einstein equations (1).65. One can also deﬁne a critical density ρc . certainly a very dilute ﬂuid! In terms of the critical density it is possible to .88 h2 10−29 g/cm3 = 2. where we substitute the FRW metric (2) and the perfect ﬂuid tensor (7). one in which an observer comoving with the ﬂuid would see the universe around it as isotropic. a ¨ 4πG Λ =− (ρ + 3p) + . Let us now write the equations of motion of such a ﬂuid in an expanding universe. p = pM + pR . can be written in terms of the FRW metric and the perfect ﬂuid tensor (7) as d d 3 ρ a3 + p a = 0. in terms of which one can estimate the order of magnitude for the present size and age of the universe. satisfying U µ Uµ = −1.ν = 0). if it is nonzero. This constitutes today one of the most fundamental problems of physics. (14) (15) where M = 1. it can be associated with the vacuum energy of quantum ﬁeld theory. We can write the Hubble parameter today H0 in units of 100 km s−1 Mpc−1 .homogeneity and isotropy is a perfect ﬂuid. 3 3 a (8) where I have treated the cosmological constant Λ as a different component from matter. I will discuss those recent measurements in the next Section. ρ = ρM +ρR . The critical density ρc corresponds to approximately 4 protons per cubic meter. and U µ is the comoving fourvelocity. The µ = ν = 0 component of the Einstein equations constitutes the socalled Friedmann equation H2 = a ˙ a 2 = Λ K 8πG ρ+ − 2 . According to general relativity. The conservation of energy (T µν.ν = 0). The energy momentum tensor associated with such a ﬂuid can be written as [6] T µν = p g µν + (p + ρ) U µU ν .4 < h < 1 for decades. and only in the last few years has it been found to lie within 10% of h = 0. let alone cosmology. = 9. ρc ≡ 2 3H0 8πG = 1. with corresponding equations of state.989 × 1033 g is a solar mass unit. a direct consequence of the general covariance of the theory (Gµν. pR = ρR/3. Gyr . a 3 3 (10) I will now make a few useful deﬁnitions.773 h −1 The parameter h has been measured to be in the range 0. dt dt (9) where the energy density and pressure can be split into its matter and radiation components. although we still do not understand why should it have such a small value (120 orders of magnitude below that predicted by quantum theory). Together. −1 c H0 −1 H0 (11) (12) (13) = 3000 h −1 Mpc .
cosmological constant and spatial curvature today must add up to one. 0 (20) ΩΛ (a) = −→ 0 . dark matter or massive neutrinos) or by a cosmological constant. Making use of the cosmic sum rule today. in the context of a FRW universe. ρCMB = π (kTCMB )4/(¯ c)3 = 4. radiation. That is. their scale factor growing at a smaller speed with each time interval. for matter. today. a 1. Another relationship which becomes very useful is that of the cosmological deceleration parameter today.4 × 10−5 h−2 . (18). Uniform expansion corresponds to q0 = 0 and requires a precise cancellation: ΩM = 2ΩΛ . .5×10−34 g/cm3 . but we still recover the EdS model. cosmological constant and even curvature. corresponding to relativistic particles. ΩM = 8πG ρM 2 3H0 Λ ΩΛ = 2 3H0 ΩR = ΩK 8πG ρR 2 3H0 K =− 2 2. we can deduce the sum of the other two. a = a0. which is dominated either by nonrelativistic particles (baryons. we can write the matter and cosmological constant as a function of the scale factor (a0 ≡ 1) ΩM (a) = 8πG ρM ΩM = 3H 2(a) a + ΩM (1 − a) + ΩΛ (a3 − a) Λ ΩΛ a3 = 3H 2(a) a + ΩM (1 − a) + ΩΛ (a3 − a) a→∞ a→0 −→ −→ a→0 1 . (10). 2 (22) which is independent of the spatial curvature. say the spatial curvature. in terms of the matter and cosmological constant components of the universe. from 2 h the density of microwave background photons. a→∞ −→ 1 (21) This implies that for sufﬁciently early times. q0 ≡ − a ¨ aH 2 = 0 1 ΩM − Ω Λ . all matterdominated FRW universes can be described by Einsteinde Sitter (EdS) models (ΩK = 0. their scale factor growing at a greater speed with each time interval. and write the rate of expansion H 2 in terms of its value today. Accelerated expansion corresponds to q0 < 0 and comes about whenever ΩM < 2ΩΛ : spatial sections expand at an increasing rate. It represents spatial sections that are expanding at a ﬁxed rate. the vacuum energy will always dominate in the future. Decelerated expansion corresponds to q0 > 0 and occurs whenever ΩM > 2ΩΛ : spatial sections expand at a decreasing rate. the total fraction of matter density. see Eq. its scale factor growing by the same amount in equallyspaced time intervals. a0H0 (16) (17) We can evaluate today the radiation component ΩR . 5 Note that in the limit a → 0 the radiation component starts dominating. For instance. as a cosmic sum rule.5 On the other hand. Three massless neutrinos contribute an even smaller amount. 1 = Ω M + ΩΛ + ΩK . see Eq. q0 .deﬁne the ratios Ωi ≡ ρi /ρc. (18) An interesting consequence of these redeﬁnitions is that I can now write the Friedmann equation today. ΩΛ = 0). Therefore. (19) where we have neglected ΩR today. if we measure one of the three components. which 15 gives ΩCMB = 2. we can safely neglect the contribution of relativistic particles to the total density of the universe today. 2 H 2(a) = H0 ΩR a3 a2 a4 0 0 0 + Ω M 3 + ΩΛ + ΩK 2 a4 a a .
while the second one 4 3 2 to x = 2ΩK /3ΩM. Note that.2. (28) . One can see in Fig. a critical universe (H = H = 0) corresponds to those points 2 x ≡ a0/a > 0. for various pairs of values of (ΩM . while f (x) > 0. Let us rewrite Eq. corresponds. for which f (x) ≡ H (a) and f (x) vanish. 2. in the absence of a cosmological constant (Λ = 0). and ΩK < 0. we can write the energy density evolution as d d (ρa3) = −p (a3) = −3Hω (ρa3) . and a spatially closed universe (K = +1) to a recollapsing universe in the future. T dS = dU + pdV . Equation (9) implies that the expansion of the universe is adiabatic or isoentropic (dS = 0). f (x) = 3x2ΩM + 2xΩK = 0 f (x) = 6xΩM + 2ΩK = x=0 . through the Friedmann equations of motion. Equation (23) can 3 be understood as the energy conservation law E = T + V for a test particle of unit mass in the central potential 1 GM + k r2 . ˙ One can show that. Expanding around ΩM = 1. Such a well known (textbook) correspondence is incorrect when ΩΛ = 0: spatially open universes may recollapse while closed universes can expand forever.1. x = −2ΩK /3ΩM > 0 +2ΩK > 0 −2ΩK > 0 x=0 . For a barotropic ﬂuid. (29) where U = ρV is the total energy of the closed system and V = a3 is its physical volume. and ΩK > 0. (23) 2 a 6 2 where M ≡ 4π ρ a3 is the equivalent of mass for the whole volume of the universe. ΩΛ). satisfying the equation of state p = ωρ. In that case. for ΩM ≥ 1.4 Thermodynamical analogy It is also enlightening to ﬁnd an analogy between the energy conservation equation (9) and the second law of Thermodynamics. f (x) = x3 ΩM + x2 ΩK + ΩΛ = 0 .1. corresponding to a ﬂuid in thermal equilibrium at a temperature T. a critical universe. (24) V (r) = − r 2 corresponding to a Newtonian potential plus a harmonic oscillator potential with a negative spring constant k ≡ −Λ/3. and only in that case. we ﬁnd ΩΛ 27 (ΩM − 1) /ΩM . deﬁned as the borderline between indeﬁnite expansion and recollapse. These critical solutions are asymptotic to the Einsteinde Sitter model (ΩM = 1. (8) as Λ K 1 2 GM a − ˙ − a2 = − = constant .3 Mechanical analogy It is enlightening to work with a mechanical analogy of the Friedmann equation. precisely with a ﬂat universe (K = 0). 2. for ΩΛ = 0. 1 a range of possible evolutions of the scale factor. see Fig. we can write the solutions as ΩΛ = 0 The ﬁrst solution corresponds to the critical point x = 0 (a = ∞). dt dt (30) 4Ω sin3 M ΩM ≤ 1 1 3 arcsin(1 − Ω−1 ) M ΩM ≥ 1 . a spatially open universe (K = −1) corresponds to an eternally expanding universe. ΩΛ = 0). x = 2ΩK /3ΩM (25) (26) (27) Using the cosmic sum rule (19).
I=(.1. the energy density of radiation in thermal equilibrium can be written as [8] ρR = g∗ = π2 g∗ T 4 .5 0.9). K=(1.0 2. while the vertical axis is y = a/a0 in each case. N=(0. M=(0. For relativistic particles in thermal equilibrium.−1).5 2. The values of (ΩM .5). The horizontal axis represents τ = H0 (t − t0 ). D=(3.1.1).5 0. the trace of the energymomentum tensor vanishes (because of conformal invariance) and thus pR = ρR /3 ⇒ ω = 1/3.707).5 0.5.5).0.1. E=(0.0 0.5 1. In that case. which is conserved. we can write dp = (ρ + p)dT /T . 30 gi i=bosons (31) Ti T 4 + Ti 7 gi 8 i=fermions T 4 .0 2.0 0.5 0.59).0).0 1. J=(. the entropy per comoving volume is S = a3 (ρ + p)V /T . 1: Evolution of the scale parameter with respect to time for different values of matter density and cosmological parameter. (32) where g∗ is the number of relativistic degrees of freedom.0 0. and therefore dS = dT (ρ + p)V 1 d[(ρ + p)V ] − (ρ + p)V 2 = d + const.1).5 1.0 2.0 3 3.0 3 3.0 3 3.5 0.5 1. F=(0.1.0 0. [7].0 3 2 1 0 1 2 3 K L F 2 1 0 1 2 3 H 2 1 0 1 2 3 D B A C 3.2. up to an additive constant..1. L=(1. B=(0.2.0 3 M Bouncing 2 1 0 1 2 3 Fig.3. H=(3.0). G=(3.0 1.5 1.1. From Ref.1. T T T (33) That is.1).5 2. .0 1.. C=(1. The entropy per comoving volume is dominated by the contribution of relativistic particles.5 N 2 1 0 1 2 3 J I B 2 1 0 1 2 3 F E No Λ Flat A Closed Open Loitering 2.0 G D 1.5 1.0 0.5.0 2.0 2.5 2.2.0).0 0.5 2.0 3 3.5 2.0 2.5 0. ΩΛ) for different plots are: A=(1.0).0 1.5 1. coming from both bosons and fermions.0 1. Using the equilibrium expressions for the pressure and density.5).
0 0. during the adiabatic expansion of the universe.5 2. beyond which the universe has a bounce.5 3. all the way to high energy physics at the electroweak scale. S = g∗s 2π 2 g∗s (aT )3 = constant .3. separating eternal expansion from recollapse in the future. the scale factor grows inversely proportional to the temperature of the universe.0 1. which showed that.5 Recollapse os ed en 1. separating open from so that. and perhaps Nature has followed a different path.5 0. and that in the future it will become much colder and dilute. to atomic physics. the universe becomes hotter and hotter and thus the amount of energy available for particle interactions increases. The last two are still uncertain since we do not have any experimental evidence for those ultra high energy phenomena. see Eq. q0 = 0.0 0. the dotted line corresponds to t0 H0 = ∞. I will brieﬂy summarize the thermal history of the universe.0 ΩM closed universes. 45 Ti Ti 3 7 + = gi gi T 8 i=fermions T i=bosons (34) 3 . gran uniﬁcation (perhaps). Therefore.5 1. the nature of interactions goes from those described at low energy by long range gravitational and electromagnetic physics. to very good approximation.5 ng rati cele Ac ng rati cele De Expansion Cl Op u Bo nc e ΩΛ 1. separating accelerating from decelerating universes. (35) A consequence of Eq. Fig. the temperature of the radiation background scaled with redshift in the way predicted by the hot Big Bang model. 6 6 See the recent theoretical developments on large extra dimensions and quantum gravity at the TeV [9]. As we go back in time. from the Planck era to the present.0 2. indeed. The line ΩΛ = ΩM /2 corresponds to uniform expansion.5 2. and ﬁnally quantum gravity. 2: Parameter space (ΩM . the observational fact that the universe is expanding today implies that in the past the universe must have been much hotter and denser. ΩΛ ).0 1. a ∝ 1/T . (36) Such a relation has been spectacularly conﬁrmed with observations of absorption spectra from quasars at large distances. The dashed line corresponds to critical universes. Since the ratio of scale factors can be described in terms of the redshift parameter z. ΩK = 0. we can ﬁnd the temperature of the universe at an earlier epoch by T = T0 (1 + z) .0 2.5 0. The line ΩΛ = 1 − ΩM corresponds to a ﬂat universe. As a consequence.0 0. Finally. 2. .2 Brief thermal history of the universe In this Section. nuclear physics.0 1. (34) is that. (5).
It is the realm of long range gravitational physics. For instance. Much later. Since then. of crucial importance is the time at which certain particles decoupled from the plasma. which soon thermalized and became the origin of the hot Big Bang as we know it. which are unravelling the details of those fundamental interactions as we increase in energy. to become galaxies. and from a primordial soup of organic compounds. 3 × 105 yr). the universe must have originated at the Planck era (1019 GeV. At the end of inﬂation. photons decouple from the plasma. If conﬁrmed. What interests cosmologists is the statistical and thermal properties that such a plasma should have. 10−10 s). at about (1 eV. the universe became radiation dominated. and superclusters. Nevertheless. 10−43 s) from a quantum gravity ﬂuctuation. and biological life originated from previous generations of stars. 1 min) and all their energy goes into photons. the Earth.e. Note that although particle physics experiments have reached energies as high as 100 GeV. giving rise to the lightest elements. it may have gone through the quarkgluon phase transition (102 MeV. The furthest window we have on the early universe at the moment is that of primordial nucleosynthesis (1 − 0. characterizing the epoch of structure formation. Much later (∼ 1−10 Gyr). when protons and neutrons were cold enough that bound systems could form. However. It is probable (although by no means certain) that the asymmetry between matter and antimatter originated at the same time as the rest of the energy of the universe. There is still some speculation about the physics that took place in the universe above the energy scales probed by present colliders. one should bear in mind that the physical conditions that take place in our high energy colliders are very different from those that occurred in the early universe. According to the best accepted view. the Sun. in a process known as recombination: It is the realm of atomic physics. the small inhomogeneities generated during inﬂation have grown. Needless to say. Such a process is called reheating of the universe. Quantum ﬂuctuations of the inﬂaton ﬁeld most probably left their imprint then as tiny perturbations in an otherwise very homogenous patch of the universe. 10−35 s). Its consequences will be discussed below. i. It is a matter of speculation whether baryogenesis could have occurred at energies as low as the electroweak scale (100 GeV. The observed relative abundances of light elements are in agreement with the predictions of the hot Big Bang theory. Immediately afterwards. when baryons (mainly protons and neutrons) formed from their constituent quarks.1 MeV.The way we know about the high energy interactions of matter is via particle accelerators. we still do not have observational evidence that the universe actually went through the EW phase transition. and the role that causal horizons play in the ﬁnal outcome of the early universe expansion.3 eV. 10−5 s). travelling freely since then. respectively. clusters of galaxies. However. One can trace the evolution of the universe from its origin till today. from the leftovers of their annihilation with antibaryons. those experiments are crucial in understanding the nature and rate of the local fundamental interactions available at those energies. Finally (3K. we don’t have any experimental evidence for such a statement: Quantum gravity phenomena are still in the realm of physical speculation. Soon after.5 MeV. Immediately after. This process is known under the name of baryogenesis since baryons (mostly quarks at that time) must have originated then. 13 Gyr). As the universe cooled down. Soon after. when their interactions were not quick enough compared with the expansion of the universe. 1 s – 3 min). Those are the photons we observe as the cosmic microwave background. ∼ 105 yr). . it is plausible that a primordial era of cosmological inﬂation originated then. These machines could never reproduce the conditions of density and pressure in the rapidly expanding thermal plasma of the early universe. electrons become bound to nuclei to form atoms (0. perhaps dominated by a vacuum energy in the form of a cosmological constant. from the decay of the inﬂaton. the overall layout presented here is a plausible and hopefully testable proposal. the universe may have reached the Grand Uniﬁed Theories (GUT) era (1016 GeV. electronpositron annihilation occurs (0. via gravitational collapse. baryogenesis would constitute another “window” into the early universe. and they were left out of equilibrium with the plasma. matter and radiation have equal energy densities. Nevertheless. the huge energy density of the inﬂaton ﬁeld was converted into particles. soon after neutrino decoupling: It is the realm of nuclear physics.
Note the large range of scales involved.1 MeV. in 1946. however.2. in rough agreement with observations [6. so that the lighter elements could be built up quickly by succesive neutron captures. Gamow reasoned that. while in the latter the universe expansion cools the hot and dense plasma in just a few minutes. 2.I will now review some of the more robust features of the Hot Big Bang theory of which we have precise observational evidence. The abundances of the light elements would then be correlated with their neutron capture cross sections. Nevertheless. about the same density as the core of the Sun. 3: The relative abundance of light elements to Hidrogen. Fig. The detailed reactions by which stars burn hydrogen were ﬁrst laid out by Hans Bethe in 1939. These processes could take place when the universe had a temperature of around TNS ∼ 1 − 0. Soon afterwards. Big Bang nucleosynthesis (BBN) codes compute a chain of around 30 coupled nuclear . In the former. 10]. From Ref. In 1920 Eddington suggested that the sun might derive its energy from the fusion of hydrogen into helium. George Gamow realized that similar processes might have occurred also in the hot and dense early universe and gave rise to the ﬁrst light elements [4]. Nowadays. the physical conditions in star and Big Bang nucleosynthesis are very different. which last a few minutes and creates all the heavier elements beyond iron). gravitational collapse heats up the core of the star and reactions last for billions of years (except in supernova explosions. which is about 100 times the temperature in the core of the Sun. there was a large number of free neutrons at that time. Note.1 Primordial nucleosynthesis and light element abundance In this subsection I will brieﬂy review Big Bang nucleosynthesis and give the present observational constraints on the amount of baryons in the universe. while the density 2 4 is ρNS = π g∗ TNS ∼ 82 g cm−3 . [10]. although the early period of cosmic expansion was much shorter than the lifetime of a star. starting with the reaction n + p → D + γ. that 30 although both processes are driven by identical thermonuclear reactions.
and probably will remain undetected for a long time. and their temperature continues to decay inversely proportional to the scale factor of the universe.0190 ± 0. Nevertheless. 10]. Below this temperature. this number is well below closure density.l. the present baryon fraction of the critical density can be calculated from η10 as [10] ΩB h2 = 3.6 − 5. up to iron (Fe). F (39) while the rate of expansion of the universe at that time (g∗ = 10. [11. via W boson exchange in n + ν ↔ p + e− and p + ν ↔ n + e+ . Furthermore. and perhaps also 6Li. T > me = 0.8 MeV. since it would conﬁrm one of the most robust features of Big Bang cosmology.002 K. neutrinos are no longer in thermal equilibrium with the rest of the plasma. as a function of only one variable. and below 0.4 T 2/MP .) (38) Clearly. the number density fraction of baryons to photons. At temperatures T me . it would be fascinating if. much below that required for detection by present experiments (of order GeV). the number of light neutrinos and the homogeneous FRW expansion of the universe. In fact.75) was H 5. precisely because of the relative weakness of the weak interactions. η ≡ nB /nγ . heating up the plasma (but not the neutrinos. 7Li. 2. (37) Such a small value of η indicates that there is about one baryon per 109 photons in the universe today. Let us compute the difference. total 11 number of degrees of freedom g∗ = 2 .2 × 10−5 GeV−2 is the Fermi constant. and beyond Fe in novae and supernovae explosions. from the conservation of entropy. since they have an average energy of order 10−4 eV. σw G2 T 2 at ﬁnite F temperature T . so they decoupled. ingenious experiments were devised to detect such a background.8 MeV. Therefore. can be written as [8] ¯ Γν = nν σw v G2 T 5 . 7 Only the ﬁrst four or ﬁve elements can be computed with accuracy better than 1% and compared with cosmological observations. At temperatures above the the mass of the electron. with g∗ = 2 degrees of freedom. the present observations are only consistent.25 : 3 · 10−5 : 2 · 10−5 : 2 · 10−10] with various errors. (40) where I have used TCMB = 2.2 Neutrino decoupling Just before the nucleosynthesis of the lightest elements in the early universe. the cosmic background of neutrinos has a temperature today lower than that of the microwave background of photons.725 ± 0. Their observed relative abundance to hydrogen is [1 : 0. are produced in heavy stars.6271 × 10−3 η10 = 0. Neutrinos decouple when their interaction rate is slower than the universe expansion. mainly systematic. which had decoupled already). We can estimate the temperature at which decoupling occurred from the weak interaction cross section.reactions. equivalently. 3 and Ref.945 K . D. in the future. so baryons cannot account for all the matter in the universe. we ﬁnd that the ratio of Tγ and Tν today must be 11 Tγ = Tν 4 1/3 = 1.511 MeV. at Tν−dec 0. see Fig. 3He. At temperatures T < me . 7 The rest of nuclei. electrons and positrons annihilate into photons. Since neutrinos decoupled before e+ e− annihilation.401 ⇒ Tν = 1. as I shall discuss below. The BBN codes calculate these abundances using the laboratory measured nuclear reaction rates.22 × 10 GeV is the Planck mass. to produce all the light elements up to beryllium7. where 19 MP = 1.0024 (95% c. The neutrino interaction rate. only photons contribute to the entropy of the universe. These light elements are H. We still have not measured such a relic background of neutrinos. weak interactions were too slow to keep neutrinos in thermal equilibrium with the plasma. the only particle species contributing to 7 the entropy of the universe are the photons (g∗ = 2) and the electronpositron pairs (g∗ = 4 × 8 ). Any acceptable theory of baryogenesis should account for such a small number. with a very narrow range of values of η10 ≡ 1010 η = 4. 4He. the decay rate of the neutron. Γν ≤ H or.2. where GF = 1. .9 .
or about teq = 1.2.2.2 0.3. (44) This scale plays a very important role in theories of structure formation. electrons could eventually become bound to protons to form neutral hydrogen. 4: The equilibrium ionization fraction Xe as a function of redshift.65.6 − 5. (45) eq Xe π me . the size of causally connected regions in the universe.6 Xeeq 0.4 0. 2. (43) = c H0 ΩM a1/2 1 + dH = aH(a) a Thus the horizon size at matterradiation equality (a = aeq) is c H −1 −1/2 dH (aeq) = √0 ΩM a1/2 eq 2 12 (ΩMh)−1 h−1 Mpc . and thus 1 + zeq = a0 ΩM = = 3.3 Matterradiation equality Relativistic species have energy densities proportional to the quartic power of temperature and therefore scale as ρR ∝ a−4 . see Eq. Since then both decay differently. The ionization fraction of electrons in equilibrium with the plasma at a given temperature is given by [8] √ eq 4 2ζ(3) T 3/2 Eion /T 1 − Xe = √ η e . ΩM 0. As I will show later. i. there is always a nonzero probability that a rare energetic photon ionizes hydrogen and produces a free electron.5 0. Around the time of matterradiation equality. while nonrelativistic particles have essentially zero pressure and scale as ρM ∝ a−3 . dH ∼ H −1 . (30). the rate of expansion (18) can be written as (a0 ≡ 1) 1/2 aeq 1/2 1/2 . 1 0.9 0.4 Recombination and photon decoupling As the temperature of the universe decreased. there will be a time in the evolution of the universe in which both energy densities are equal ρR (teq) = ρM (teq ).1 × 104 ΩM h2 . Nevertheless.e.1 0 1000 1100 1200 1300 (1+z) 1400 1500 1600 eq Fig.9.3 0. and therefore (1 + zeq ) 3900. while h 0. aeq ΩR (41) where I have used ΩR h2 = ΩCMB h2 + Ων h2 = 3.24 × 10−5 for three massless neutrinos at T = Tν .8 0.2. (42) = H0 ΩM a−3/2 1 + H(a) = H0 ΩR a−4 + ΩM a−3 a The horizon size is the coordinate distance travelled by a photon since the beginning of the universe. Therefore. The two lines show the range of η10 = 4.7 0. the matter content of the universe today is below critical. The comoving horizon size is then given by aeq −1/2 c −1 −1/2 .2 × 103 (ΩM h2 )−2 7 × 104 years after the origin of the universe.
We can estimate this moment by evaluating Γγ = H at photon decoupling. the photon distribution will contain a sufﬁciently large number of highenergy photons to ionize a signiﬁcant fraction of hydrogen. Unfortunately.8 × 105 (ΩM h2 )−1/2 5 × 105 years old. and η is the baryontophoton ratio (37). where the error bars are smaller than the line width. Using ne = Xe η nγ . we deduce the corresponding redshift at recombination. Fig. the mean free path is much smaller that the causal horizon at that time and photons suffer multiple scattering: the plasma is like a dense fog. Comparing with the present temperature of the microwave background. when photons last scattered off protons and electrons and travelled freely ever since.3 eV Eion.26 eV. This background was predicted by George Gamow and collaborators in the 1940s. 2. The left panel corresponds to the monopole spectrum. they had doubts whether the radiation would have . λ−1 ∼ Γγ = ne σT . The mean free path of photons λγ in such a plasma can be estimated from the photon interaction rate. If eq we now use Eq. γ For temperatures above a few eV.725 ± 0. This decoupling occurred when the universe was approximately tdec = 1. The right panel shows the dipole spectrum.2. [12]. with cross section σT = 8πα2 = 6. 5: The Cosmic Microwave Background Spectrum seen by the FIRAS instrument on COBE.6 eV is the ionization energy of hydrogen.372 ± 0. one can compute 1100. (36). Note that the huge number of photons with respect to electrons (in the ratio 4He : H : γ 1 : 4 : 1010) implies that even at a very low temperature.014 mK. based on the consistency of primordial nucleosynthesis with the observed helium abundance. T0 = 2. for η10 5.where Eion = 13. δT1 = 3.5 The microwave background One of the most remarkable observations ever made my mankind is the detection of the relic background of photons from the Big Bang. They estimated a value of about 10 K.036 is the dimensionless electromagnetic coupling constant. (1 + zrec) 1270. we can compute the ionization fraction Xe as a function of redshift z.665 barn . the decoupling temperature as Tdec = 0. Photons remain in thermal equilibrium with the plasma of baryons and electrons through elastic Thomson scattering. deﬁning recombination eq as the time at which Xe ≡ 0. one ﬁnds that the recombination temperature is Trec = 0.1. and the corresponding redshift as (1 + zdec) This redshift deﬁnes the so called last scattering surface. Photons will decouple from the plasma when their interaction rate cannot keep up with the expansion of the universe and the mean free path becomes larger than the horizon size: the universe becomes transparent.65 × 10−25 cm2 = 0.002 K. From Ref. 3m2 e (46) where α = 1/137. In fact. see Fig. 4.2. although a somewhat more detailed analysis by Alpher and Herman in 1950 predicted Tγ ≈ 5 K.
the velocity of the Sun with respect to the CMB . or one part in 105. Roll and Wilkinson [13] studied the problem again in 1965. [12]. with spectral distortions below the level of 10 parts per million (ppm). Under the assumption that a Doppler effect is responsible for the entire CMB dipole. see Ref. Peebles. 5. From Ref.5 ± 1 K. The central region corresponds to foreground by the galaxy. 48. are of the order of 20 µK on large scales.) ± 7 µK (1σ statistical) (47) In fact.0035.14◦ ± 0. until Dicke. 6: The Cosmic Microwave Background Spectrum seen by the DMR instrument on COBE. with a resolution of about 7◦ in the sky.002 K. [14].). The most outstanding one has been the Cosmic Background Explorer (COBE) satellite. δT1 = 3.007 mK (95% c. [14]. The middle ﬁgure shows the dipole.002 K (systematic.e. and the lower ﬁgure shows the quadrupole and higher multipoles. b) = (264. the differential microwave radiometer (DMR) instrument on COBE. they learned that Penzias and Wilson had observed a weak isotropic background signal at a radio wavelength of 7.725 ± 0.372 ± 0.725 ± 0. see Fig. with a spectral resolution ∆ν = 0. They published their two papers back to back. whose FIRAS instrument measured the photon background with great accuracy over a wide range of frequencies (ν = 1 − 97 cm−1 ). Since then many different experiments have conﬁrmed the existence of the microwave background. has also conﬁrmed that it is an extraordinarily isotropic background. and this remarkable prediction slipped into obscurity. explaining the fundamental signiﬁcance of their measurement [6]. in fact. a dipole anisotropy of one part in 103. differences in the temperature of the blackbody spectrum measured in different directions in the sky. Fig.l.26◦ ± 0. 95% c. i. δT2 = 18 ± 2 µK. The top ﬁgure corresponds to the monopole. corresponding to a blackbody temperature of Tγ = 3. the photon spectrum is ν conﬁrmed to be a blackbody spectrum with a temperature given by [12] TCMB = 2. with that of Dicke et al. this is the best blackbody spectrum ever measured. see Ref.). There is. Moreover. T0 = 2.30.35 cm.014 mK. Nowadays. δT1 = 3.30) (95% c. Before they could measure the photon background.l. The deviations from isotropy. in the direction of the Virgo cluster.survived until the present.l. (l.372 ± 0.
a) − ρ(a) ¯ δ(x.5 km/s. i. clusters and superclusters on large scales. 7. or z < 0. that study the spatial distribution of hundreds of thousands of galaxies up to distances of a billion light years. (49) where the brackets · represent integration over an ensemble of different universe realizations. Zel’dovich. As I shall discuss below. etc. see Ref. we are left with a whole spectrum of anisotropies in the higher multipoles (quadrupole. see Ref. Those ripples must have left some trace as temperature anisotropies in the microwave background. Furthermore. When matter fell in the troughs of those waves.e. [12]. with a spectrum that is also scale invariant. This is a necessary requirement for any consistent theory of structure formation [16]. as matter fell into their troughs. and more recently the IRAS Point Source redshift Catalog. These structures are expected to arise from very small primordial inhomogeneities that grow in time via gravitational instability. and already put constraints on the theory of COBE even determined the annual variation due to the Earth’s motion around the Sun – the ultimate proof of Copernicus’ hypothesis.1. 8 . The reason why they took so long to be discovered was that they appear as perturbations in temperature of only one part in 105. 2. Today various telescopes – like the Hubble Space Telescope.8 When subtracted. that recede from us at speeds of tens of thousands of kilometres per second. at higher multipole numbers or smaller angular scales. we need a theory that will grow a density contrast ¯ with amplitude δ ∼ 10−5 at the last scattering surface (z = 1100) up to density contrasts of the order of δ ∼ 102 for galaxies at redshifts z 1. we know that the universe today is not exactly homogeneous: we observe galaxies. and independently by the Russian cosmologist Yakov B. δT2 = 18 ± 2 µK (95% c.l. a) ≡ = d3 k δk (a) eik·x . These catalogs are telling us about the evolution of clusters of galaxies in the universe. today. While the predicted anisotropies have ﬁnally been seen in the CMB. Harrison. it created density perturbations that collapsed gravitationally to form galaxies and clusters of galaxies. (48) ρ(a) ¯ where ρ(a) = ρ0 a−3 is the average cosmic density.). these anisotropies play a crucial role in the understanding of the origin of structure in the universe. but there are at present various catalogs like the CfA and APM galaxy catalogs. not all kinds of matter and/or evolution of the universe can give rise to the structure we observe today. and Las Campanas redshift surveys. [14] and Fig. [17]. to explain the distribution of galaxies and clusters of galaxies on very large scales in our observable universe. These inhomogeneities are like waves in the spacetime metric. Only a few galaxies are known at those redshifts. the anisotropies observed by the COBE satellite correspond to a smallamplitude scaleinvariant primordial power spectrum of inhomogeneities P (k) = δk 2 ∝ kn .). and indeed such anisotropies were ﬁnally discovered by the COBE satellite in 1992. the twin Keck telescopes in Hawaii and the European Southern Observatory telescopes in Chile – are exploring the most distant regions of the universe and discovering the ﬁrst galaxies at large distances. Such a type of spectrum was proposed in the early 1970s by Edward R.3 Largescale structure formation Although the isotropic microwave background indicates that the universe in the past was extraordinarily homogeneous. 6. Soon after COBE. see Fig. The furthest galaxies observed so far are at redshifts of z 5. see Ref. other groups quickly conﬁrmed the detection of temperature anisotropies at around 30 µK and above. octupole. whose light was emitted when the universe had only about 5% of its present age.rest frame is v = 371 ± 0. with n = 1. If we deﬁne the density contrast as [15] ρ(x. or 12 billion light years from the Earth. and that may have originated from tiny ripples in the metric.
have not yet gone nonlinear. [18]. like the Sloan Digital Sky Survey (SDSS) and the AngloAustralian two degree ﬁeld (2dF) Galaxy Redshift Survey. Before COBE discovered the anisotropies of the microwave background there were serious doubts whether gravity alone could be responsible for the formation of the structure we observe in the universe today. looking at millions of galaxies. The observed power spectrum of the galaxy matter distribution from a selection of deep redshift catalogs can be seen in Fig. so there must be some extra matter (dark since we don’t see it) to account for its gravitational pull.000 galaxies. ∼ over a large region of the sky. the largest structures. even deeper galaxy redshift catalogs are underway. How much there is is a matter of debate at the moment. If we want to make sense of the present observations. overdense regions decouple from the Hubble expansion to become bound systems. which start attracting eachother to form larger bound structures. 7: The IRAS Point Source Catalog redshift survey contains some 15. These important observations will help astronomers determine the nature of the dark matter and test the validity of the models of structure formation. the anisotropies were found with the right amplitude for structure to be accounted for by gravitational collapse of primordial inhomogeneities under the attraction of a large component of nonrelativistic dark matter. We know from Big Bang nucleosynthesis that all the baryons in the universe cannot account for the observed amount of matter. In the process. superclusters. That is. In order to resolve this issue. We show here the projection of the galaxy distribution in galactic coordinates. which are at this moment taking data. Some recent analyses suggest that there is not enough cold dark matter to reach the critical density required to make the universe ﬂat.Fig. and superclusters are forming now. covering over 83% of the sky up to redshifts of z ≤ 0. the standard theory of structure formation is a cold dark matter model with a non vanishing cosmological constant in a spatially ﬂat universe. In fact. Whether it is relativistic (hot) or nonrelativistic (cold) could be inferred from observations: relativistic particles tend to diffuse from one concentration of matter to another. Fortunately. This fundamental difference is an indication of the type of matter that gave rise to structure. Linear perturbation theory shows that the growing . we must conclude that some other form of energy permeates the universe. The primordial spectrum (49) is reprocessed by gravitational instability after the universe becomes matter dominated and inhomogeneities can grow. structure formation. From Ref. From these observations one can infer that most galaxies formed at redshifts of the order of 2 − 6. It seemed that a new force was required to do the job.05. thus transferring energy among them and preventing the growth of structure on small scales. and not the other way around.5. Nowadays. clusters of galaxies formed at redshifts of order 1. Gravitational collapse ampliﬁes the density contrast initially through linear growth and later on via nonlinear collapse. cosmic structure formed from the bottom up: from galaxies to clusters to superclusters. so we conclude that most of the matter responsible for structure formation must be cold. This is excluded by observations. 8. up to redshifts of z < 0.
mode 9 of small density contrasts go like [15. until photon decoupling. well before photons decoupled from baryons. slight deviations for a The important observation is that. The right panel shows a compilation of the most recent estimates of the power spectrum of galaxy clustering. since the density contrast at last scattering is of order δ ∼ 10−5 . 8: The left panel shows the matter power spectrum for clusters of galaxies. for radiation and matter. 16] δ(a) ∝ a1+3ω = a2 . On the other hand. enters the horizon before matterradiation equality. From Ref. electromagnetically). the fast radiationdriven expansion prevents darkmatter perturbations from collapsing. keq = d−1 (aeq) H 9 0. Since light can only cross regions that are smaller than the horizon. a < aeq a > aeq (50) in the Einsteinde Sitter limit (ω = p/ρ = 1/3 and 0. and the scale factor has grown since then only a factor zdec ∼ 103. radiation pressure eventually halts the contraction and sets up acoustic oscillations in the plasma that prevent the growth of perturbations. One can decompose the density contrast in Fourier components. (48). if ΩM = 1 or ΩΛ = 0. the suppression of growth due to radiation is restricted to scales smaller than the horizon. The resolution of this mismatch is one of the strongest arguments for the existence of a weakly interacting cold dark matter component of the universe. This is the reason why the horizon size at equality. (44). The reason why baryonic inhomogeneities cannot grow is because of photon pressure: as baryons collapse towards denser regions. This is very convenient since in linear perturbation theory individual Fourier components evolve independently. where δ ∼ 102 . of wavelength λ = k−1 < dH (aeq). while largescale perturbations remain unaffected. ﬂuctuations in that component could grow as soon as it decoupled from the plasma. even before matterradiation equality. Instead. a weakly interacting cold dark matter component could start gravitational collapse much earlier. A comoving wavenumber k is said to “enter the horizon” when k = d−1 (a) = aH(a). If a certain H perturbation.083 (ΩMh) h Mpc−1 . and thus reach the density contrast amplitudes observed today.Fig. one would expect a density contrast today of order δ0 ∼ 10−2. . from three different cluster surveys. respectively). for all ω. How much dark matter there is in the universe can be deduced from the actual power spectrum (the Fourier transform of the twopoint correlation function of density perturbations) of the observed large scale structure. sets an important scale for structure growth. from four of the largest available redshift surveys of opticallyselected galaxies. we observe structures like galaxies. If there is an additional matter component that only couples through very weak interactions. compared to the deprojected spectrum of the 2D APM galaxy survey. (51) The decaying modes go like δ(t) ∼ t−1 . a. There are aeq . see Eq. but we will not be concerned with them here. So how can this be possible? The microwave background shows anisotropies due to ﬂuctuations in the baryonic matter component only (to which photons couple. [19]. Eq.
I will devote. we expect to see a large fraction of those baryons constituting a hot ionized gas surrounding large clusters of galaxies. k k keq keq (52) This is precisely the shape that largescale galaxy catalogs are bound to test in the near future. 9: The power spectrum for cold dark matter (CDM). 3 DETERMINATION OF COSMOLOGICAL PARAMETERS In this Section. when baryons start to collapse onto dark matter potential wells.10 P ( k ) ( h3 Mpc 3) 10 5 4 1000 Microwave Background d ( h1 Mpc ) 100 10 Superclusters Clusters 1 Galaxies 1000 100 10 1 0. nonrelativistic Cold Dark Matter (CDM) allow structure to form on all scales via gravitational collapse. and the age of the universe t0 . we can infer relationships between the different cosmoWe will take the baryon fraction as given by observations of light element abundances. and conﬁrms the general picture of structure formation. hot dark matter (HDM). normalized to COBE. Using the homogeneity and isotropy on large scales observed by COBE.4. As a consequence. From Ref. ionizing the medium. they will wipe out small scale perturbations.8 HDM n=1 MDM n=1 0. since relativistic Hot Dark Matter (HDM) transfer energy between clumps of matter. However. they will convert a large fraction of their potential energy into kinetic energy of protons and electrons. and discussed in Section 4.10 These ﬁve basic cosmological parameters are not mutually independent. 9. The suppression factor can be easily computed from (50) as fsup = (aenter/aeq)2 = (keq/k)2. in accordance with Big Bang nucleosynthesis. The dark matter will then pull in the baryons. 10 . On the other hand. We will see that a large host of observations are determining the cosmological parameters with some reliability of the order of 10%. the matter content ΩM .1 B CO E CDM n=1 TCDM n = . for largescale structure formation. This is indeed what is observed. this Section to the more ‘classical’ measurements of the following cosmological parameters: The rate of expansion H0 . which will later shine and thus allow us to see the galaxies. however. [20].01 0. see Eq. tilted cold dark matter (TCDM). Naturally. see Fig. the cosmological constant ΩΛ . Most of the recent work in observational cosmology has been the search for virtually systematicfree observables.001 0. k−3 . the processed power spectrum P (k) will have the form: P (k) ∝ k. I will restrict myself to those recent measurements of the cosmological parameters by means of standard cosmological techniques. and this should be seen as a distinctive signature in the matter power spectra of future galaxy catalogs. In other words. Furthermore.1 k ( h Mpc ) 1 1 10 Fig. the spatial curvature ΩK . (38). together with a few instances of new results from recently applied techniques. and mixed hot plus cold dark matter (MDM). like those obtained from the microwave background anisotropies. the majority of these measurements are dominated by large systematic errors.
From Ref.0 and 5.0 2.0 0.0 1. 2n + 1 n=0 ∞ (58) . a 1 y≡ = . t0 = f (H0. ΩΛ).5.0 0. There are two speciﬁc limits of interest: an open universe with ΩΛ = 0.0 1.0 2. Another relationship between parameters appears for the age of the universe.5 1.5 1. for which the age can also be expressed in compact form. ΩΛ). Deﬁning a new time and normalized scale factor. (53) or viceversa.5 2. 3. determined from 1 t0 H0 = 0 dy 1 + (y −1 − 1)ΩM + (y 2 − 1)ΩΛ −1/2 .5 3. In a FRW cosmology.0 ΩM Fig. the cosmic expansion is determined by the Friedmann equation (8). 2. Therefore. if we determine that the universe is spatially ﬂat from observations of the microwave background.2. τ ≡ H0 (t − t0 ) .5 − 1. 1 + (1 − ΩM )1/2 2 2 ln t0 H0 = = 1/2 1/2 3 3(1 − ΩM ) Ω M (1 − ΩM )n .0. we can be sure that the sum of the matter content plus the cosmological constant must be one. (2n + 1)(2n + 3) (57) and a ﬂat universe with ΩΛ = 1 − ΩM . in parameter space (ΩM . 10 the contour lines for constant t0 H0 in parameter space (ΩM . 10: The contour lines correspond to equal t0 H0 = 0. 1. 1. ΩΛ). (55) with initial condition y(0) = 1. from bottom to top. y (0) = 1. (56) We show in Fig. The line t0 H0 = ∞ would be indistinguishable from that of t0 H0 = 5. the present age t0 is a function of the other parameters.logical parameters through the EinsteinFriedmann equations. In particular. 1 = Ω M + ΩΛ + ΩK .0 0.5 ΩΛ 1. we can deduce the value of the spatial curvature from the Cosmic Sum Rule. (54) a0 1+z we can write the Friedmann equation with the help of the Cosmic Sum Rule (19) as y (τ ) = 1 + (y −1 − 1)ΩM + (y 2 − 1)ΩΛ 1/2 . [7].5 0.0 0.0. ΩM .5 2. for which the age is given by t0 H0 = ΩM 1 1 + (1 − ΩM )1/2 − =2 ln 1/2 1 − ΩM (1 − ΩM )3/2 Ω M ∞ n=0 (1 − ΩM )n .
only due to an incorrect assessment of systematic errors [21]. As we will see. in several occasions. For example. ΩM .We have plotted these functions in Fig. [22]. one should also deﬁne what is known as the luminosity distance to an object in the universe. The energy ﬂux F received at the detector is the measured energy per unit time per unit area of the detector coming from that source. 11: The age of the universe as a function of the matter content. A standard candle is a luminous object that can be calibrated with some accuracy and therefore whose absolute luminosity is known. ΩΛ and t0 . dr 1+ 2 a2 H0 r2 ΩK 0 = 1 2 a2H0 0 dz . the fraction of the area of the 2sphere centered on the source that is covered by the detector is dA/4πa2 r2 (z). as a function of redshift z and the other cosmological parameters. dt0 = (1 + z)dt. but only the age of its constituents: stars. the photon energy on its way here will be redshifted. First. ds2 = 0. of course. light travels along null geodesics. the universe seemed to be younger than its constituents. we cannot measure the age of the universe directly. H0. within certain errors. for an open and a ﬂat universe. their calibration errors are within bounds. or. the rate of photon arrival will be timedelayed with respect to that emitted by the source. this is not a trivial bound and. etc. ΩΛ ). Fig.5 Gyr. Of course. 2 F ≡ L/4πdL. It is clear that in both cases t0 H0 → 2/3 as ΩM → 1. galaxies. i. see Eq. t0 > tgal + 1. ΩM. 11. and thus the observed energy E0 = E/(1 + z). From Ref. globular clusters. (2). Finally. The absolute luminosity L of such a source is nothing but the energy emitted per unit time. In a FriedmannRobertsonWalker universe. (1 + z)2(1 + zΩM ) − z(2 + z)ΩΛ (59) which determines the coordinate distance r = r(z. a logical inconsistency. Now let us consider the effect of the universe expansion on the observed ﬂux coming from a source at a certain redshift z from us. In order to understand those recent measurements. Second. Therefore. 4πa0 L . We can now use these relations as a consistency check between the cosmological observations of H0. the 0 total ﬂux detected is L L (60) F= 2 r 2 (z) ≡ 4πd2 . Imagine a source that is emitting light at a distance dL from a detector of area dA. ∼ during the progress towards better determinations of the cosmological parameters. Cepheid variable stars and type Ia supernovae are considered to be reasonable standard candles. Thus we can only ﬁnd a lower bound on the age of the universe. The luminosity distance dL is then deﬁned as the radius of the sphere centered on the source for which the absolute luminosity would give the observed ﬂux.e.
Fortunately. in clear conﬂict with geology. The arrival times of photons from two different gravitationally lensed images of the quasar depend on the different path lengths and the gravitational potential traversed. 3. as I will discuss shortly. First. still in conﬂict with ratios of certain unstable isotopes. parallax. which were previously used successfully in particle physics detectors. i. Therefore. 3.1. with systematic errors approaching the 10% level.4. b) SunyaevZel’dovich effect. mainly Cepheid variability and type Ia Supernovae. Higher order terms are not yet probed by cosmological observations. we obtain Eq. in 1954 Baade recalibrated the Cepheid distance and obtained a lower value.e. still with large (factor of two) systematic errors. technological. by the reﬁnement of existing methods for measuring extragalactic distances (e. Second. which is sensitive to the cosmological parameters ΩM and ΩΛ . since it involves knowledge about the primordial spectrum of inhomogeneities. offers tremendous potential because it can be applied at great distances and it is based on very solid physical principles [24].The ﬁnal expression for the luminosity distance dL as a function of redshift is thus given by [8] H0 dL = (1 + z) ΩK −1/2 sinn ΩK 1/2 z 0 dz (1 + z )2(1 + z ΩM ) − z (2 + z )ΩΛ . Later on. and the Hubble rate dropped down to H0 = 60 km s−1 Mpc−1 . There will be at least two different images of the same background variable point source. and leave the last method for Section 4. etc. Hubble’s data was based on Cepheid standard candles that were incorrectly calibrated with those in the Large Magellanic Cloud. in 1958 Sandage realized that the brightest stars in galaxies were ionized HII regions.). supernovae. with the development of completely new methods to determine H0 . Cepheids. H0 dL = z + ΩM 1 1− + ΩΛ z 2 + O(z 3 ) . which fall into totally independent and very broad categories: a) Gravitational lensing. proposed in 1964 by Refsdael [23]. H0 = 250 km s−1 Mpc−1 . Finally. Around 1929. Let us now pursue the analysis of the recent determinations of the most important cosmological parameters: the rate of expansion H0 . It is only recently that cosmological observations have gone far enough back into the early universe that we can begin to probe the second term. Hubble measured the rate of expansion to be H0 = 500 km s−1 Mpc−1 . and the age of the universe t0 . the matter content ΩM .1 The rate of expansion H0 Over most of last century the value of H0 has been a constant source of disagreement [21]. sin(x) if K = +1 and sinh(x) if K = −1. a measurement of the time delay and the angular separation of the different images of a variable quasar can be used to determine H0 with great accuracy. Finally. I will review here the ﬁrst three. . which implied an age of the universe of order t0 ∼ 2 Gyr.1 Gravitational lensing Imagine a quasistellar object (QSO) at large redshift (z 1) whose light is lensed by an intervening galaxy at redshift z ∼ 1 and arrives to an observer at z = 0. These improvements come from two directions. (6). corresponding to the Hubble law. c) Extragalactic distance scale. through the replacement of photographic plates (almost exclusively the source of data from the 1920s to 1980s) with charged couple devices (CCDs). in the past 15 years there has been signiﬁcant progress towards the determination of H0 . the cosmological constant ΩΛ . the spatial curvature ΩK . (61) where sinn(x) = x if K = 0. This method.g. but they would contribute as important consistency checks. solid state detectors with excellent ﬂux sensitivity per pixel. 2 2 (62) This expression goes beyond the leading linear term. Expanding to second order around z = 0. d) Microwave background anisotropies. into the second order term.
ν = ν0 (1 + z). in the near future. That is the reason why it has taken so much time since the original proposal for the ﬁrst results to come out. The inverseCompton scattering of microwave background photons off the hot electrons in the Xray gas results in a measurable distortion of the blackbody spectrum of the microwave background. the dark matter distribution in those systems is usually unknown. (∆ν/ν) (kB Tgas/me c2) ∼ 10−2 . and exactly compensates the redshift in energy of the photons that reach us). [27]. Fig. but those measurements are very difﬁcult and.25. . From Ref. Since the Xray ﬂux is distancedependent (F = L/4πd2 ).2 SunyaevZel’dovich effect As discussed in the previous Section. Measuring the spatial distribution of the SZ effect (3 K spectrum). seen here in an optical image (left) and an Xray image (right). for which there is a model of the lensing mass distribution that is consistent with the measured velocity dispersion. one can determine from there the distance to the cluster. see Fig. Since photons acquire extra energy from the Xray electrons. there are now very powerful telescopes that can be used for these purposes. see Ref.e. one can determine the density and temperature distribution of the hot gas. Nevertheless. 12. together with a high resolution Xray map (108 K spectrum) of the cluster. taken by the recently launched Chandra Xray Observatory. 12: The Coma cluster of galaxies. and thus the Hubble rate H0. one can determine [25] H0 = 72 ± 7 (1σ statistical) ± 15% (systematic) km s−1 Mpc−1 . (63) The main source of systematic error is the degeneracy between the mass distribution of the lens and the value of H0 . in the case of lensing by a cluster of galaxies. the gravitational collapse of baryons onto the potential wells generated by dark matter gave rise to the reionization of the plasma. This corresponds to a decrement of the microwave background temperature at low frequencies (RayleighJeans region) and an increment at high frequencies. Knowledge of the velocity dispersion within the lens as a function of position helps constrain the mass distribution. 3. known as the SunyaevZel’dovich (SZ) effect. associated with a complicated cluster potential. generating an Xray halo around rich clusters of galaxies. Fortunately.Unfortunately. [26]. while the SZ decrement is L not (because the energy of the CMB photons increases as we go back in redshift. The best candidate todate is the QSO 0957 + 561. we expect a shift towards higher frequencies of the spectrum. a known mass distribution of the intervening galaxy) and a variable background source with a measurable time delay.1. Assuming a ﬂat space with ΩM = 0. with the recent discovery of several systems with optimum properties. there are very few systems with both a favourable geometry (i. observed with the 10m Keck telescope. the prospects for measuring H0 and lowering its uncertainty with this technique are excellent. the method is just starting to give promising results and.
1.2 The matter content ΩM In the 1920s Hubble realized that the so called nebulae were actually distant galaxies very similar to our own. (61). together with precise Xray maps and spectra from the Chandra Xray observatory recently launched by NASA.03. and possible contaminations from point sources. see Eq. and 23 clusters within z < 0. which provides the ﬁducial comparison for Cepheids in more distant galaxies. lum + ΩB. 3. new possible extra sources to the matter content of the universe have been accumulating: ΩM = ΩB. (64) compatible with other determinations. as it contracts and expands. 3. dark + ΩCDM + ΩHDM (stars in galaxies) (MACHOs?) (weakly interacting : axion. for a set of 21 galaxies within 25 Mpc. However. A great advantage of this completely new and independent method is that nowadays more and more clusters are observed in the Xray. The main systematics come from possible clumpiness of the gas (which would reduce H0). in 1933. a possible metallicity dependence of the Cepheid periodluminosity relation and cluster population incompleteness bias. ∼ With better telescopes coming up soon. The Hubble Space Telescope (HST) was launched by NASA in 1990 (and repaired in 1993) with the speciﬁc project of calibrating the extragalactic distance scale and thus determining the Hubble rate with 10% accuracy. From the observed ﬂux one can then deduce the luminosity distance. and soon we will have highresolution 2D maps of the SZ decrement from several balloon ﬂights.The advantages of this method are that it can be applied to large distances and it is based on clear physical principles. Present measurements give the value [26] H0 = 60 ± 10 (1σ statistical) ± 20% (systematic) km s−1 Mpc−1 . (65) The main source of systematic error is the distance to the Large Magellanic Cloud. the assumption of hydrostatic equilibrium of the Xray gas. as well as from the European Xray satellite XMM launched a few months ago by ESA. and the Next Generation Space Telescope (NGST) proposed by NASA for 2008. Zwicky found dynamical evidence that there is possibly ten to a hundred times more mass in the Coma cluster than contributed by the luminous matter in galaxies [29]. neutralino?) (massive neutrinos?) (66) (67) (68) (69) . H0 could be larger). The most recent results from HST are the following [28] H0 = 71 ± 4 (random) ± 7 (systematic) km s−1 Mpc−1 . with 4 synchronized telescopes by the year 2005. projection effects (if the clusters are prolate. and thus the Hubble rate H0 . and the star’s absolute luminosity determined from the calibrated relationship.3 Cepheid variability Cepheids are lowmass variable stars with a periodluminosity relation based on the helium ionization cycles inside the star. At that time there was evidence that rotation curves of galaxies did not fall off with radius and that the dynamical mass was increasing with scale from that of individual galaxies up to clusters of galaxies. which will deliver orders of magnitude better resolution than the existing Einstein Xray satellite. details of models for the gas and electron densities. This time variability can be measured. as well as from future microwave background satellites. Since then. Other systematic uncertainties that affect the value of H0 are the internal extinction correction method used. Soon afterwards. it was not until the 1970s that the existence of dark matter began to be taken more seriously. it is expected that much better resolution and therefore accuracy can be obtained for the determination of H0. like the Very Large Telescope (VLT) interferometer of the European Southern Observatory (ESO) in the Chilean Atacama desert.
No such behaviour is observed. Later it became possible to measure galactic rotation curves far out into the disk. from microlensing and the direct search of Massive Compact Halo Objects (MACHOs).1 ± 2. The orbital velocity rose linearly from the center outward until it reached a typical value of 200 km/s. I(r) = I0 exp(−r/rD ). Therefore. one would expect that most of the galactic mass is concentrated within a few disk lengths rD . [32]. Υ = M/L. and ﬁnally from microwave background anisotropies.002 ≤ Ωlum h ≤ 0. so there must be a large fraction of baryons that are dark. M /L . and thus [31] 0. ˚ centered at ∼ 5500 A) is [30] LV = (1. Lgal = (1. from the cluster velocity dispersion with the use of the Virial theorem.The empirical route to the determination of ΩM is nowadays one of the most diversiﬁed of all cosmological parameters. (70) The contribution of galaxies to the luminosity density of the universe (in the visibleV spectral band. from the rotation curves of galaxies. such that the rotation velocity is determined as in a Keplerian orbit. see Ref. Moreover. so that for the sun Υ = 1.2) × 10−4 ΥV . etc. and a trend was found [32]. and then remained ﬂat out to the largest measured radii.0 ± 0. One can measure the orbital velocities of objects orbiting around the disk as a function of radius from the Doppler shifts of their spectral lines. the most convincing observations come from radio emission (from the 21 cm line) of . Spiral galaxies consist of a central bulge and a very thin disk. ΩM h = (6. from the stars in the disk. clusters of galaxies. from weak gravitational lensing. The rotation curve of the Andromeda galaxy was ﬁrst measured by Babcock in 1938. from the baryon fraction in the Xray gas of clusters.1 Luminous matter The most straight forward method of estimating ΩM is to measure the luminosity of stars in galaxies and then estimate the masstolight ratio. This ratio is usually expressed in solar units. stabilized against gravitational collapse by angular momentum conservation. deﬁned as the mass per luminosity density observed from an object.2. This was completely unexpected since the observed surface luminosity of the disk falls off exponentially with radius. In fact. 3. the luminous matter alone is far from the critical density. (72) (71) All the luminous matter in the universe. from direct detection of Weakly Interacting Massive Particles (WIMPs) at DAMA and UKDMC. which can be translated into a mass density by multiplying by the observed Υ in that band.7 ± 0. from the observed matter distribution of the universe via its power spectrum. 3.006 . we conclude that Ωlum ΩB.. For globular clusters and spiral galaxies we can determine their mass and luminosity independently and this gives Υ ≈ few.6) × 108 h L Mpc−3 . from direct detection of massive neutrinos at SuperKamiokande. The masstolight ratio of stars in the solar neighbourhood is of order Υ ≈ 3. comparing with the amount of baryons from Big Bang nucleosynthesis (38). and surrounded by an approximately spherical halo of dark matter. from galaxies. The matter content of the universe can be deduced from the masstolight ratio of various objects in the universe. For our galaxy. vrot = (GM/r)1/2 ∝ r−1/2. perhaps in the form of very dim stars. The luminosity of stars depends very sensitively on their mass and stage of evolution. from the cluster abundance and its evolution.2. account for Υ ≈ 10.3) × 108 h L Mpc−3 and Υgal = 6 ± 3 . I will review here just a few of them.2 Rotation curves of spiral galaxies The ﬂat rotation curves of spiral galaxies provide the most direct evidence for the existence of large amounts of dark matter. (73) As a consequence.
with 2 2 density proﬁle ρ(r) = ρc /(rc + r2 ). one should measure its rotation curve. one ﬁnds Υhalo ≥ 30 h. but its nature remains a mystery.3 Microlensing The existence of large amounts of dark matter in the universe. the existence of dark matter in our own galaxy. i.03 − 0. determined by radio observations of hydrogen gas in the disk [33]. about 13 disk lengths away. and in our own galaxy in particular. through direct detection. thousands of galactic rotation curves are known. Nowadays. it would be extraordinary if we could conﬁrm. 13 together with the relative components associated with the disk. Fig. where rD = 1. as well as problems with the determination of reliable galactocentric distances for the tracers. At large radii the dark matter distribution leads to a ﬂat rotation curve. For dark matter searches. the rotation curve of the Milky Way has been measured and conforms to the usual picture. the crucial quantity is the dark matter density in the solar neighbourhood. the dotdashed line is from the dark matter halo alone. in hydrostatic equilibrium. The dashed line shows the rotation curve expected from the disk material alone. is now established beyond any reasonable doubt. We will come back to direct searched of dark matter in a later subsection. which turns out to be (within a factor of two uncertainty depending on the halo model) ρDM = 0.22 kpc. A typical case is that of the spiral galaxy NGC 6503. which has been measured to much larger galactic radii than optical tracers. see Ref.2. This model is consistent with the universal rotation curve seen in Fig. while the furthest measured hydrogen line is at r = 22.73 kpc. which is much more difﬁcult because of obscuration by dust in the disk. 13: The rotation curve of the spiral galaxy NGC 6503. with v∞ equal to the plateau value of the ﬂat rotation curve.neutral hydrogen in the disk. with a plateau value of the rotation velocity of 220 km/s. the halo and the gas. and therefore Ωhalo ≥ 0. We have seen that baryons cannot account for the whole matter content of the universe. Adding up all the matter in galactic halos up to maximum radii. [35]. and all suggest the existence of about ten times more mass in the halos of spiral galaxies than in the stars of the disk. since the contribution of the halo . This picture has led to a simple model of darkmatter halos as isothermal spheres. The dark matter is supposed to undergo violent relaxation and create a virialized system. however. 3. 13. (74) Of course. Nevertheless. The measured rotation curve is shown in Fig. where rc is a core radius and ρc = v∞ /4πG.3 GeV/cm3 .e. For that purpose.05 . Recent numerical simulations of galaxy formation in a CDM cosmology [34] suggest that galaxies probably formed by the infall of material in an overdense region of the universe that had decoupled from the overall expansion.
very long exposure images of the Hubble Space Telescope restrict the possible Mdwarf contribution to the galaxy to be below 6%. 14. They cannot be normal stars since they would be luminous. while the overproduction of helium in the halo is strongly constrained. 14. in a direct search experiment. They could make a large contribution towards the total ΩM .e. see Fig. A ﬁnal possibility is primordial black holes (PBH). When the source is exactly aligned with the deﬂector of mass MD . The idea is based on the well known effect that a pointlike mass deﬂector placed between an observer and a light source creates two different images. with radius d1 d2 2 (75) rE = 4GMD d . and still be compatible with Big Bang nucleosynthesis. Whatever the arguments for or against baryonic objects as galactic dark matter.(74) is comparable in magnitude to the baryon fraction of the universe (38). and thus may be classiﬁed as nonbaryonic. stars with a mass M ≤ 0. nothing would be more convincing than a direct detection of the various candidates. . In other words. which could have been created in the early universe from early phase transitions [36]. an Einstein ring. i. one may ask whether the galactic halo could be made of purely baryonic material in some nonluminous form. Fortunately. even before baryons were formed. Fig.1 M which are intrinsically dim. neither hot gas since it would shine. The most plausible alternative is a halo composed of brown dwarfs with mass M ≤ 0. are MACHOs the nonluminous baryons ﬁlling the gap between Ωlum and ΩB ? If not. can be classiﬁed as lowmass brown dwarfs. If the two images cannot be separated because their angular distance α is below the resolving power of the observer’s telescope. [22]. Neutron stars or black holes would typically arise from Supernova explosions and thus eject heavy elements into the galaxy. the extrapolation of the stellar mass function to small masses predicts a large number of brown dwarfs within normal stellar populations. however. Could they be burntout stellar remnants? This seems implausible since they would arise from a population of normal stars of which there is no trace in the halo. stars not massive enough to reach supernova phase. and if so. They could be white dwarfs. the only effect will be an 11 A sometimes discussed alternative. how one should search for it. which never ignite hydrogen and thus shine only from the residual energy due to gravitational contraction.11 In fact. where d= d1 + d2 is the reduced distance to the source. Are they stars too small to shine? Perhaps Mdwarfs. in 1986 Paczy´ ski proposed a method for detecting faint stars in the halo of our n galaxy [39]. nor cold gas since it would absorb light and reemit in the infrared. or their exclusion. 14: Geometry of the light deﬂection by a pointlike mass which gives two images of a source viewed by an observer. what are they? Let us start a systematic search for possibilities. planetsize Jupiters. a halo composed by white dwarfs is not rigorously excluded.08 M . as shown in Fig. Despite some recent arguments. From Ref. the image would be an annulus.
50 kpc away. it has allowed the MACHO and EROS collaboration to draw exclusion plots for various mass ranges in terms of their maximum allowed halo fraction. for 10−4 M it is 1 day. and about 2 towards Andromeda. . the average microlensing time will be 3 months. if one looks simultaneously at several millions of stars in the LMC during extended periods of time. around 40 towards the bulge of our own galaxy. In order to be sure one has seen a microlensing event one has to monitor a large sample of stars long enough to identify the characteristic light curve shown in Fig. The MACHO Collaboration conclude in their 5year analysis. The ampliﬁcation factor A is shown in logarithmic scale to give the usual astronomical magnitude of an object. and for 10−6 M it is 2 hours. b) timesymmetric. 15: The apparent lightcurve of a source if a pointlike MACHO passes through the line of sight with a transverse velocity v and an impact parameter b. If the MACHO moves with velocity v transverse to the line of sight. and c) achromatic (because of general covariance). The ampliﬁcation factor is [39] 2 + u2 r A= √ . for 10−2 M it is 9 days. 2 towards the SMC. one has a good chance of seeing at least a few of them brightened by a dark halo object. one of them will occasionally pass near the line of sight and thus cause the image of the background star to brighten.e. apparent brightening of the star. where u≡ 2 rE u 4+u with r the distance from the line of sight to the deﬂector. For stars in the LMC one ﬁnds a probability. microlensing is a well established technique with a rather robust future. and the origin corresponds to the time of closest approach to the line of sight. the minimal distance to the line of sight. Thus. From Ref. Thus. seen by AGAPE [42]. of duration 34 days. The typical duration of the light curve is the time it takes a MACHO to cross an Einstein radius. The unequivocal signatures of such an event are the following: it must be a) unique (nonrepetitive in time). 17. see Fig. 41]. 22]. is shown in Fig. an optical depth for microlensing of the galactic halo. then one expects an apparent lightcurve as shown in Fig. with a slightly different technique based on pixel brightening rather than individual stars. The probability for a target star to be lensed is independent of the mass of the dark matter object [39. 15 for different values of b/rE. an effect known as gravitational microlensing. there are 12 candidates towards the LMC.e. is b. ∆t = rE /v. The natural time unit is ∆t = rE /v. and if its impact parameter. i. (76) . i. If the deﬂector mass is 1 M . A characteristic event. The ﬁrst microlensing events towards the LMC were reported by the MACHO and EROS collaborations in 1993 [40. If the galactic halo is ﬁlled with MACHOs. These signatures allow one to discriminate against variable stars which constitute the background. 15. Imagine an observer on Earth watching a distant star in the Large Magellanic Cloud (LMC). In particular.Fig. of approximately τ ∼ 10−6 . Nowadays. 16. [22].
see Ref. with several new experiments already underway.05.4 Virial theorem and large scale motion Clusters of galaxies are the largest gravitationally bound systems in the universe (superclusters are not yet in equilibrium). . [37].9 M . Zwicky noted in 1933 that these systems appear to have large amounts of dark matter [29].l. For a discussion of those new results. A maximum likelihood analysis gives a MACHO halo fraction of 20% for a typical halo model with a 95% conﬁdence interval of 8% to 50%. 16: The best candidate (LMC1) for microlensing from the MACHO Collaboration in the direction of the Large Magellanic Cloud. The lower mass is characteristic of white dwarfs. but a galactic halo composed primarily of white dwarfs is barely compatible with a range of observational constraints. depending on the halo model. one needs to appeal to a very nonstandard density and/or velocity distribution of these objects. The most likely MACHO mass is between 0. A 100% MACHO halo is ruled out at 95% c. if one wanted to attribute the observed events to brown dwarfs. consisting partially of compact objects. Measuring the velocity dispersion v 2 from the Doppler shifts of the spectral lines and estimating the geometrical size of the system gives an estimate of its total mass M . Nevertheless. see Ref. A recent reanalysis of this event suggested an ampliﬁcation factor Amax = 7. that the spatial distribution of events is consistent with an extended lens distribution such as Milky Way or LMC halo. to search for clear signals of parallax.Fig. for all except their most extreme halo model.20 ± 0. On the other hand. He used the virial theorem (for a gravitationally bound system in equilibrium). [37].8 ± 0. From Ref. where the degeneracy between mass and distance can be resolved. We know today several thousand clusters. where Ekin = 2 m v 2 is the average kinetic energy of one of the bound objects (galaxies) of mass m and Egrav = −m GM/r is the average gravitational potential energy caused by the attraction of the other galaxies.15 M and 0. [38].00 ± 0.09.2. and a duration of t = 34. 1 2 Ekin = − Egrav . 3. they have typical radii of 1 − 5 Mpc and typical masses of 2−9 ×1014 M . the ﬁeld is expanding. with achromaticity ˆ Ared/Ablue = 1. or binary systems. It is still unclear what sort of objects the microlensing experiments are seeing towards the LMC and where the lenses are.2.
l. it is known that galaxy clusters are the most powerful Xray sources in the sky [46]. where Xrays are produced by electron bremsstrahlung. The cosmic matter density inferred from such analyses is [43. (77) On scales larger than clusters the motion of galaxies is dominated by the overall cosmic expansion. together with the observed galaxy distributions.3 95% c.Fig. Since the 1960s. galaxies exhibit peculiar velocities with respect to the global cosmic ﬂow.09 (systematic) . at least within a 20% systematic error. As Zwicky noted. typically leading to a masstolight ratio Υcluster = 200 ± 70. Assuming that the average cluster Υ is representative of the entire universe 12 one ﬁnds for the cosmic matter density [44] ΩM = 0. The emission extends over the whole cluster and reveals the existence of a hot plasma with temperature T ∼ 107 − 108 K. when Xray telescopes became available. they scoop up mass over a large volume of space. Nevertheless. 95% and 99% probability. .05 (1σ statistical) ± 0. can then be translated into a measure for the masstolight ratio required to explain the largescale ﬂows. Assuming the 12 Recent observations indicate that Υ is independent of scale up to supercluster scales ∼ 100 h −1 Mpc. An example of the reconstruction of the matter density ﬁeld in our cosmological vicinity from the observed velocity ﬁeld is shown in Fig. the peculiar motions of galaxies are attributed to the action of gravity during the universe evolution.2.5 Baryon fraction in clusters Since large clusters of galaxies form through gravitational collapse. 90%.24 ± 0. From Ref. The panels are labeled according to different sets of selection criteria (A or B). this virial mass of clusters far exceeds their luminous mass. The plus sign shows the maximum likelihood estimate and the contours enclose regions of 68%. 45] ΩM > 0. caused by the matter density inhomogeneities which give rise to the formation of structure. (78) Related methods that are more modeldependent give even larger estimates. our Local Group of galaxies is moving with a speed of 627 ± 22 km/s relative to the cosmic microwave background reference frame. 3. and whether or not an LMC halo with MACHO fraction f is included. [38]. For example. 18. 17: Likelihood contours for MACHO mass m (in units of solar mass) and halo fraction f for a typical size halo. In the context of the standard gravitational instability theory of structure formation. towards the Great Attractor. The observed largescale velocity ﬁelds. and therefore the ratio of baryons over the total matter in the cluster should be representative of the entire universe.
From these estimates one can calculate the baryon fraction of clusters fB h3/2 = 0.01.05 ± 0.3 ± 0. giving general agreement (within a factor of 2) with the virial mass estimates.15 . (80) This value is consistent with previous determinations of ΩM . If some baryons are ejected from the cluster during gravitational collapse. Coordinates are in units of 10 h−1 Mpc. and using the Big Bang nucleosynthesis value of ΩB = 0. The marked structures are the Local Group (LG). From Ref.08 ⇒ ΩB ≈ 0. Assuming this fraction to be representative of the entire universe. or some are actually bound in nonluminous objects like planets. we ﬁnd ΩM = 0. . 18: The velocity and density ﬂuctuation ﬁelds in the Supergalactic Plane as recovered by the POTENT method from the Mark III velocities of about 3. gas to be in hydrostatic equilibrium and applying the virial theorem one can estimate the total mass in the cluster.03 − 0.Fig. for h = 0. then the actual value of ΩM is smaller than this estimate.1 (statistical) ± 20% (systematic) . The vectors are projections of the 3D velocity ﬁeld in the frame of the CMB. ΩM for h = 0.000 galaxies with 12 h−1 smoothing.65 . (79) which together with (73) indicates that clusters contain far more baryonic matter in the form of hot gas than in the form of stars in galaxies. the Coma cluster “Great Wall” (GW).65. the “Great Attractor” (GA). the PerseusPisces (PP) region and the “Southern Wall” (SW). [43].
Since the primordial spectrum is very approximately represented by a scaleinvariant Gaussian random ﬁeld. [48]. From a systematic study of the cluster mass distribution one can reconstruct the shear ﬁeld responsible for the gravitational distortion. 19: The most famous image of weak gravitational lensing around the Abell 2218 cluster. 3. NonGaussian effects are expected to arise from the nonlinear gravitational collapse of structure. At present. the estimates indicate ΩM = 0. the best way to present the results of structure formation is by working with the 2point correlation function in Fourier space (the equivalent to the Green’s function in QFT).3. the power spectrum is all we need to describe the galaxy distribution. gravitational instability increases the primordial density contrast. whether the universe has been dominated by cold or hot dark matter or by a cosmological constant since the beginning of structure formation. If the reprocessed spectrum of inhomogeneities remains Gaussian. whether cosmic strings or some other topological defect from an early phase transition are responsible for the formation of structure today. while ΩM = 0. whether they are Gaussian or nonGaussian. whether the primordial spectrum has tilt (deviations from scaleinvariance). whether adiabatic (perturbations in the energy density) or isocurvature (perturbations in the entropy density). From Ref.2. This analysis shows that there are large amounts of dark matter in the clusters. The power spectrum measures the degree of inhomogeneity in the mass distribution on different scales. 19.4 for the Corona Borealis ∼ supercluster. on scales of order 20 Mpc.7 Structure formation and the matter power spectrum One the most important constraints on the amount of matter in the universe comes from the present distribution of galaxies. and may be important at small scales [15]. made by the Hubble Space Telescope. etc. b) the recent creation of inhomogeneities. although the lensing masses tend to be systematically larger. [47]. and also depending on the rate of expansion of the universe.3.3 on scales < 6 h−1 Mpc. in rough agreement with the virial mass estimates..6 Weak gravitational lensing Since the mid 1980s. and c) the cosmic evolution of the inhomogeneity. It depends upon a few basic ingredientes: a) the primordial spectrum of inhomogeneities. seen at the last scattering surface as temperature anisotropies. As we mentioned in the Section 2. see for instance Fig. see Ref. The spectroscopic analysis showed that the cluster and the giant arcs were at very different redshifts. .2. into the present density ﬁeld responsible for the large and the small scale structure. deep surveys with powerful telescopes have observed huge arclike features in galaxy clusters. the socalled power spectrum. Fig. The usual interpretation is that the arc is the image of a distant background galaxy which is in the same line of sight as the cluster so that it appears distorted and magniﬁed by the gravitational lens effect: the giant arcs are essentially partial Einstein rings.2 − 0.
3 (deﬁned as ΛCDM and OCDM. on 10 Mpc scales.2. 2dF. At present.3 − 0. on 1000 Mpc scales. and several other large redshift catalogs.8 Cluster abundance and evolution Rich clusters are the most recently formed gravitationally bound systems in the universe. 8 and 9. compared with observations from massive clusters. In the near future. mainly from the observed cluster abundance. which determines the normalization of the spectrum. Abundance of massive clusters (clusters per [Mpc/h]3) 106 SCDM 108 OCDM 1010 ΛCDM TCDM 1012 0 0. to the CMB ﬂuctuations. The standard CDM model with ΩM = 1. The four models are normalized to COBE. 3. see Figs. A model that has become a working paradigm is a ﬂat cold dark matter model with a cosmological constant and ΩM = 0. respectively) can be normalized so that they agree with both the CMB and cluster observations. This model will soon be confronted with very precise measurements from SDSS. due to the low density. hardly any structure . Even though a large amount of work has gone into those analyses. deep redshift surveys are probing scales between 100 and 1000 Mpc. From Ref. The observational constraints on the power spectrum have a huge lever arm of measurements at very different scales. but with large uncertainties. [49]. is inconsistent with the cluster abundance. At present. The observed present (z ∼ 0) cluster abundance provides a strong constraint on the normalization of the power spectrum of density perturbations on cluster scales. The lowmass models (Open and ΛCDM) predict a relatively small change in the number density of rich clusters as a function of redshift because.5 Redshift 1. these measurements suggest a low value of ΩM . The evolution of the cluster abundance with redshift breaks the degeneracy among the models at z ∼ 0. see Section 4. produces too many clusters at all redshifts. Their number density as a function of time (or redshift) helps determine the amount of dark matter. 20. Both ΛCDM and OCDM are consistent with the observed cluster abundance at z ∼ 0.0 Fig.4. normalized to the CMB ﬂuctuations on large scales. while Standard CDM (EinsteinDe Sitter model. when normalized at COBE scales. we still have large uncertainties about the nature and amount of matter necessary for structure formation. The power spectra of both a ﬂat model with a cosmological constant or an open universe with ΩM = 0. which should begin to see the turnover corresponding to the peak of the power spectrum at keq.The working tools used for the comparison between the observed power spectrum and the predicted one are very precise Nbody numerical simulations and theoretical models that predict the shape but not the amplitude of the present power spectrum. see Fig. with ΩM = 1). galaxy survey observations will greatly improve the power spectrum constraints and will allow a measurement of ΩM from the shape of the spectrum. that are already taking data.5. 20: The evolution of the cluster abundance as a function of redshift.
as predicted in the simplest models of inﬂation. e. The highmass models (Tilted and Standard CDM) predict that structure has grown steadily and rich clusters only formed recently: the number density of rich clusters at z ∼ 1 is predicted to be exponentially smaller than today. Alternatively.4. 21. the theory of gravity (general relativity) may need to be modiﬁed on large scales.growth occurs since z ∼ 1. Note that although ΩB and Ωhalo coincide at H0 70 km/s/Mpc. the galactic halo component is the horizontal band. The need to introduce an effective cosmological constant on large scales is nowadays the only reason why gravity may need to be modiﬁed at the quantum level. 3. The second problem is what constitutes 90% of matter.g. that the universe is ﬂat. from BBN baryons to the mass inferred from cluster dynamics. a cosmological constant. i. the luminous matter Ωlum . (81) But one should be cautious. ΩM .2. ΩM = 0. (80). Note that in the range H0 = 70 ± 7 km/s/Mpc. There is the caveat that for this constraint it is assumed that the initial spectrum of density perturbations is Gaussian. the galactic halo component Ωhalo. There are four bands. (74). crossing the baryonic component from BBN. (73). 21: The observed cosmic matter components as functions of the Hubble expansion parameter. Ω0 = 1? One possibility could be that the universe is dominated by a diffuse vacuum energy. They could be in small clumps of hydrogen that have not started thermonuclear reactions and perhaps constitute the dark matter of spiral galaxies’ halos. due to quantum gravity effects. and most of the approaches simply consider the inclusion of a cosmological constant as a phenomenological parameter. for ΩM as a function of H0. Between the fraction predicted by BBN and that seen in stars and diffuse gas there is a huge fraction which is in the form of dark baryons. which only affects the very large scales. And ﬁnally.25 +0. this could be just a coincidence.15 −0. (38). This is the standard dark matter problem and could be solved by direct detection of a weakly interacting massive particle in the laboratory. what constitutes around 60% of the energy density. The observation of a single massive cluster is enough to rule out the ΩM = 1 model. Eq. such a proposal is still very speculative. and the dynamical mass from clusters. From this ﬁgure it is clear that there are in fact three dark matter problems: The ﬁrst one is where are 90% of the baryons.e. In fact. there are three dark matter problems.9 Summary of the matter content We can summarize the present situation with Fig. since we know from observations of the CMB. suggesting a low density universe [50]. and the dynamical mass component from large scale structure analysis is given by Eq. see Section 4. from dynamical mass to critical density. Fig. from BBN. .10 (1σ statistical) ± 20% (systematic) . but that has not yet been conﬁrmed observationally on cluster scales. the baryon content ΩB . Since we still do not have a quantum theory of gravity. Eq. see the text. The luminous matter component is given by Eq. three clusters have been seen.
for any of the three families of neutrinos. 22: The neutrino parameter space. 22.τ. for massless neutrinos. Supposing that the missing mass in nonbaryonic cold dark matter arises from a single particle dark matter (PDM) component. τ 2 6 7 8 9 10 Solar νe. and thus. Note the threshold of cosmologically important masses. as recent experiments seem to suggest. Note that this limit improves by six orders of magnitude the present bound on the tauneutrino mass [51].s BBN Limit 2 5 Solar νe−νµ. If neutrinos have a mass. the cosmic energy density in 3 massive neutrinos would be ρν = nν mν = 11 nγ mν . cosmologically detectable neutrinos (by CMB and LSS observations). Ων h2 = mν . 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 1 3 2 νe−νµ limit Cosmologically Excluded 1 νµ−ντ limit 0 LSND νµ. This allows one to compute the present number density in the form of neutrinos. nν (Tν ) = 11 nγ (Tγ ) = 112 cm−3 . If neutrinos have mass. see Section 2. mν ≤ 40 eV. They are the only candidates known to exist. including the results from the different solar and atmospheric neutrino oscillation experiments. From Ref. mixing angle against ∆m2 . per species of neutrino. 94 eV (82) The discussion in the previous Sections suggest that ΩM ≤ 0. 3 which turns out to be. 2 I will now go through the various logical arguments that exclude neutrinos as the dominant component of the missing dark matter in the universe. 21.2.2.4.2. and therefore its contribution today.νµ.05 ≤ ΩPDM h2 ≤ 0. its contribution to the critical density is bounded by 0.νµ.3.10 Massive neutrinos One of the ‘usual suspects’ when addressing the problem of dark matter are neutrinos.4. Is it possible that neutrinos with a mass 4 eV ≤ mν ≤ 40 . could they constitute the missing matter? We know from the Big Bang theory. and cosmologically excluded range of masses. [52]. see Fig.νe Cosmologically Detectable νµ−νs νe−νs Atmos νµ−ντ 2 3 ∆m (eV ) 4 Cosmologically Important Solar νe. see Fig.τ 11 10 4 10 3 10 2 10 1 10 0 sin 2θ Fig. that there is a cosmic neutrino background at a temperature of approximately 2K.
such collisions are essentially coherent over an entire nucleus. this constraint. is suppressed by a Boltzmann factor. 22. leading to an enhancement of the effective cross section.15. the deﬁnitive test to the possible contribution of neutrinos to the overall density of the universe would be to measure directly their mass in laboratory experiments. We must conclude that the simple idea that light neutrinos could constitute the particle dark matter on all scales is ruled out. and there is no theoretical reason to expect a massive sterile neutrino that does not oscillate into the other neutrinos.1(10 GeV/mν )2 for Majorana neutrinos and slightly smaller for Dirac neutrinos. Another possibility is that neutrinos have a large mass. like the tritium βspectrum and the neutrinoless doubleβ decay experiments. In that case. ∼ exp(−mν /Tdec).e. their number density at decoupling. it would be very implausible to have such a massive neutrino today. neutrinos scatter on nucleons via the weak axialvector current (spindependent) interaction. for this mass range. However.13 There are at present two types of experiments: neutrino oscillation experiments. which must be greater than the typical halo density ν −3 . either it is a sterile neutrino (it does not couple to the Z). For the small momentum transfers imparted by galactic WIMPs. against evidence from present observations. of order a few GeV. to the atmospheric neutrino oscillations from SuperKamiokande (∆ m2 3× ν 2 2 −5 −3 2 10 eV ) and the solar neutrino oscillations (∆ mν 10 eV ).8 MeV. Therefore. beyond closure density (82). their abundance. If it couples to the Z boson and has a mass below 45 GeV for Dirac neutrinos (39. see Fig. i. even for a single species. then it is ruled out by measurements at LEP of the invisible width of the Z. or it does couple but has a larger mass. [53]. Only the ﬁrst two possibilities would be cosmologically relevant. however. the present energy density has to be computed as a solution of the corresponding Boltzmann equation. which may provide a higher abundance (similar to the case of baryogenesis). which measure directly the mass of the electron neutrino and give a bound mνe < 2 ∼ eV. this mass.2. The relatively large detection rate in this case allowes one to exclude fourthgeneration Dirac neutrinos for the galactic dark matter [54]. which is far too high for structure formation. They could. In either case. In that case. the maximum local density in dark matter max 3 neutrinos is ρmax = nmax mν = m4 vesc/3π 2. known as the TremaineGunn limit. Neutrinos of such a low mass would constitute a relativistic hot dark matter component. see Bilenky’s contribution to these Proceedings [55]. In the case of a Majorana neutrino (its own antiparticle). Apart from a logarithmic correction. see Fig. If it were a Dirac neutrino there could be a lepton asymmetry. Neutrinos with such a mass could very well constitute the HDM component of the universe. could massive neutrinos constitute the dark matter halos of galaxies? For neutrinos to be gravitationally bound to galaxies it is necessary that their velocity be less that the escape velocity vesc . Laboratory limits for ντ of around 18 MeV [51]. nmax = p3 /3π 2. 22. How many neutrinos can be packed in the halo of a galaxy? Due to the Pauli exclusion principle.5. which measure only differences in squared masses. say the tauneutrino. 13 For a review of Neutrinos. However. with a lifetime greater than the age of the universe. Furthermore. exclude the known light neutrinos. ρhalo = 0. applying the same phasespace argument to the neutrinos as dark matter in the halo of dwarf galaxies gives mν ≥ 100 eV. see Ref. is too small for being cosmologically relevant. neutrinos could be the dark matter only if their mass was a few GeV. and much more stringent ones for νµ and νe . and thus their maximum momentum is pmax = mν vesc.005.5 GeV for Majorana neutrinos). ΩHDM < ∼ 0. Ων h2 ≤ 0. see Fig. For masses mν > Tdec 0. which would washout structure below the supercluster scale. There are two logical alternatives. one ﬁnds Ων h2 0.eV be the nonbaryonic PDM component? For instance. For a typical spiral galaxy. However. Anyway. gives a value for Ων h2 = 0. and direct masssearches experiments. The oscillation experiments give a variety of possibilities for ∆ m2 = 0. 22. see Section 2. since it would have to be stable.3 − 3 eV2 from LSND ν (not yet conﬁrmed). still play a role as a subdominant hot dark matter component in a ﬂat CDM model. . there is always the possibility of a fourth unknown heavy and stable (perhaps sterile) neutrino.2. the maximum number density is given by that of a completely degenerate Fermi gas with momentum pF = pmax. a neutrino mass of order 1 eV is not cosmological excluded. Of course.3 GeV cm gives mν ≥ 40 eV.
but they are essentially ruled out as a universal dark matter candidate. denote a thermal average at the freezeout temperature. . If WIMPs constitute the dominant component of the halo of our galaxy. see Carena’s contribution to these Proceedings. Tf mPDM /20. = 2 σ σannvrel H0 annvrel (83) Here vrel is the relative velocity of the two incoming dark matter particles and the brackets . The direct experimental search for them rely 14 For a review of Supersymmetry (SUSY). In fact.2 ×10−6 pb. 23: The maximum likelihood region from the annualmodulation signal consistent with a neutralino of mass m χ = 59 +17 −14 +0.3.4 GeV and a proton cross section of ξσp = 7. Something like a heavy stable neutrino. σannvrel ∼ α2 /8π mPDM ∼ 3 × 10−27 cm3 s−1 . The value of σannvrel needed for ΩPDM ≈ 1 is remarkably close to what one would expect for a WIMP with a mass mPDM = 100 GeV. There remains the mystery of what is the physical nature of the dominant cold dark matter component. it is expected that some may cross the Earth at a reasonable rate to be detected. From Ref. There are a few theoretical candidates for WIMPs.11 Weakly Interacting Massive Particles Unless we drastically change the theory of gravity on large scales.14 but at present there is no empirical evidence that such extensions are indeed realized in nature. . could be a reasonable candidate because its present abundance could fall within the expected range. Fig. see the text. like the neutralino. baryons cannot make up the bulk of the dark matter. We still do not know whether this is just a coincidence or an important hint on the nature of dark matter. The scatter plot represents the theoretical predictions of a generic MSSM. the nonobservation of supersymmetric particles at current accelerators places stringent limits on the neutralino mass and interaction cross section [57]. ΩPDM h2 ∼ 3 G3/2T0 h2 3 × 10−27 cm3 s−1 .2. even if they may play a subdominant role as a hot dark matter component. coming from supersymmetric extensions of the standard model of particle physics. . [56].0 −1. a generic Weakly Interacting Massive Particle (WIMP). Massive neutrinos are the only alternative among the known particles. when the dark matter particles go out of equilibrium with radiation.
depending on the model. use a cryogenic detector at 10 mK and search for a measurable temperature increase of the sample. see Fig.9 pb.15 Therefore. The DAMA/NaI experiment has actually reported such a modulation signal. There has been no conﬁrmation yet of this result from other dark matter search groups. and third. is to attribute a tentative signal unambiguously to galactic WIMPs rather than to some unidentiﬁed radioactive background. which provides a conﬁdence level of 99. 24: The DAMA experiment sees an annual variation. of course. The problem. the recoil energy of the nuclei in the elastic collision would be of order 10 keV. as well as from the NaI scintillation detectors of the UK dark matter collaboration (UKDMC) in the Boulby salt mine in England [59]. Therefore. see Fig. Current experiments already touch the parameter space expected from supersymmetric particles. If their mass is in the range 10 − 100 GeV. Fig. second. The expected behaviour of a WIMP signal is a cosine function with a minimum (maximum) roughly at the dashed (dotted) vertical lines. the net speed of the Earth relative to the galactic dark matter halo varies.4 × 10−6 −8 −0. and to reduce the background caused by cosmic rays requires that these experiments be located deeply underground. The CRESST experiment [60] uses sapphire crystals as targets and a new method to simultaneously measure 15 The time scale of the Sun’s orbit around the center of the galaxy is too large to be relevant in the analysis. and therefore there is a chance that they actually discover the nature of the missing dark matter. in the WIMP ﬂux due to the Earth’s motion around the Sun.3 GeV cm−3 is the local neutralino energy density in units of the galactic halo density. There are at present three different methods: First. see Fig. from the combined analysis of their 4year data [56]. The model independent residual rate in the lowest (2 − 6 keV) cumulative energy interval (in counts per day/kg/keV) is The best limits on WIMP scattering cross sections come from some germanium experiments [58]. with a typical number below 1 event/kg/day. From Ref. The main problem with such a type of experiment is the low expected signal rate. 24. of order 7%. search for an ionization signal in a semiconductor. one should be able to identify such energy depositions in a macroscopic sample of the target. [56]. . One speciﬁc signature is the annual modulation which arises as the Earth moves around the Sun. To reduce natural radioactive contamination one must use extremely pure substances. causing a modulation of the expected counting rate. typically a very pure germanium crystal.6% for a neutralino mass of mχ = 52 +10 GeV and a proton cross section of ξσp = 7.on elastic WIMP collisions with the nuclei of a suitable target. where ξ = ρχ /0. but hopefully in the near future we will have much better sensitivity at low masses from the Cryogenic Rare Event Search with Superconducting Thermometers (CRESST) experiment at Gran Sasso as well as at weaker cross sections from the CDMS experiment at Stanford and the Soudan mine.2 +0. 25. DAMA/ NaI1 DAMA/ NaI2 DAMA/ NaI3 DAMA/ NaI4 shown as a function of time since 1 January of the ﬁrst year of data taking. Dark matter WIMPs move at a typical galactic virial velocity of around 200 − 300 km/s. and the DAMA experiment in the Gran Sasso laboratory in Italy [56]. 23. one could search for scintillation light in NaI crystals or in liquid xenon.
The advantages of AMANDA are also directional. 25: Exclusion range for the spinindependent WIMP scattering cross section per nucleon from the NaI experiments and the Ge detectors. over a volume ∼ 1 km3 .3 The cosmological constant ΩΛ A cosmological constant is a term in the Einstein equations.3 km deep in very clear ice. It has always been assumed that quantum gravity effects. The search goals for the upcoming largescale cryogenic detectors CRESST and CDMS are also shown. From Ref. [65]. Alternatively. In other words. via some as yet unknown symmetry. Very recently there has been the interesting proposal of a completely new method based on a Superheated Droplet Detector (SDD). the AMANDA experiment at the South Pole [64]. The absence of such a background in both gamma ray satellites and the Alpha Matter Spectrometer [63] imposes bounds on their density in the halo. [22]. WIMPs traversing the solar system may interact with the matter that makes up the Earth or the Sun so that a small fraction of them will lose energy and be trapped in their cores. but this remains a downright speculation. Such a discrepancy is one of the biggest problems of theoretical physics [66]. which is expected to have 103 Cherenkov detectors 2. 3. There exist other indirect methods to search for galactic WIMPs [62]. which allows excellent background discrimination. one of the . see Ref. Such particles could selfannihilate at a certain rate in the galactic halo. These theories predict 4 a value of order ρv ∼ MP 5 × 1093 g/cm3.Fig. [61]. since the arrays of Cherenkov detectors will allow one to reconstruct the neutrino trajectory and thus its source. (1). In particular. that corresponds to the energy density of the vacuum of quantum ﬁeld theories. Moreover. the phonons and the scintillating light from particle interactions inside the crystal. which is about 123 orders of magnitude larger than the critical density (14). is competitive with the best direct searches proposed. In fact. see Ref. building up over the age of the universe. which claims to have already a similar sensitivity as the more standard methods described above. Their annihilation in the core would thus produce high energy neutrinos from the center of the Earth or from the Sun which are detectable by neutrino telescopes. Also shown is the range of expected counting rates for neutralinos in the MSSM. see Eq. would exactly cancel the cosmological constant. Λ ≡ 8πGρv . producing a potentially detectable background of high energy photons or antiprotons. neutrino telescopes are already competitive with direct search experiments. whether it comes from the Earth or the Sun. SuperKamiokande already covers a large part of SUSY parameter space.
neutrinos. this requires that the dark energy have negative pressure. In addition.1. this ethereal constant has been invoked several times in history to explain a number of apparent crises. together with strong indications of a low mass density universe (ΩM < 1). and kinetic . it is impossible for structure in the universe to grow. and that it will do so forever after. From Ref. ever since Einstein introduced it in 1917. that indicate that some kind of dark energy must make up the rest of the energy density up to critical. i. This argument [67] would rule out almost all of the usual suspects. Finally. such as cold dark matter. (20). The only known way to reconcile a low mass density with a ﬂat universe is if an additional “dark” energy dominates the universe today. if most of the energy of the universe resists gravitational collapse. always to disappear under further scrutiny [21]. Fig. A simple linear relation between the absolute magnitude and a “stretch factor” multiplying the light curve timescale ﬁts the data quite well.2 and Eq. Nevertheless. I will now discuss the different arguments one by one.difﬁculties with a nonzero value for Λ is that it appears coincidental that we are now living at a special epoch when the cosmological constant starts to dominate the dynamics of the universe. However. from observations of CMB anisotropies. radiation. since the ratio of dark energy to matter density goes like a(t)−3p/ρ. According to general relativity. 26: The Type Ia supernovae observed nearby show a relationship between their absolute luminosity and the timescale of their light curve: the brighter supernovae are slower and the fainter ones are faster. clusters and voids. there is growing evidence for an accelerating universe from observations of distant supernovae. see Section 2.e. [68]. It would have to resist gravitational collapse. there are new observational arguments for a nonzero value. In spite of the theoretical prejudice towards Λ = 0. the discrepancy between the ages of globular clusters and the expansion age of the universe may be cleanly resolved with Λ = 0. otherwise it would have been detected already as part of the energy in the halos of galaxies. ΩΛ = 1 − ΩM . This dilemma can be resolved if the hypothetical dark energy was negligible in the past and only recently became the dominant component. The most compelling ones are recent evidence that we live in a ﬂat universe. from the large scale distribution of galaxies.
0 redshift z Fig.energy. On the other hand. 1 ) (0. 11.12 (adding errors in quadrature). dominated today by a vacuum energy density. H0 = 70 ± 7 km/s/Mpc. A candidate that has recently been exploited with great success is a certain type of supernova explosions at large redshifts.93 ± 0. are conﬁdent that they can use this type of supernova as a standard candle.12 which is perfectly compatible with recent observations. Both groups conclude that distant supernovae are fainter than expected. By studying the characteristic light curves. where t0 H0 is shown as a function of ΩM . Since the light coming from some of these rare explosions has travelled for a large fraction of the size of the universe.5.02 0.5) (0. 1996) 18 16 14 0. see Fig. as was expected from the general attraction . their environment. called SN of type Ia.12 corresponds to ΩM = 0. in a ﬂat universe.5 1. marginally consistent with observations of large scale structure. 27. for a ﬂat universe with a cosmological constant. From Ref. see Fig. t0 H0 = 0. since they all have zero or positive pressure. These suggest that we probably live in a ﬂat universe that is accelerating. Thus.1 0. and the measured rate of expansion. the universe appears to be accelerating instead of decelerating. In fact. there is a pattern: brighter explosions last longer than fainter ones. cosmologists from two competing groups.2 0. p ≈ −ρ.20. 0) (1.0.24 for an open −0.05 +0. 0) ( 1.. see Fig. Although the maximum luminosity varies from one supernova to another. ΩΛ = 1 − ΩM . depending on their original mass. of a reasonably large statistical sample. These are white dwarf stars at the end of their life cycle that accrete matter from a companion until they become unstable and violently explode in a natural thermonuclear explosion that outshines their progenitor galaxy. to account for the missing energy. [68]. A similar diagram is found by the High Redshift Supernova Project [69]. one expects to be able to infer from their distribution the spatial curvature and the rate of expansion of the universe.34 +0. 0) Λ=0 Flat 22 effective mB 20 Calan/Tololo (Hamuy et al.5. with negative pressure. astronomers look for distant astrophysical objects that can serve as standard candles to determine the distance to the object from their observed apparent luminosity.J. 11. This negative pressure would help accelerate the universe and reconcile the expansion age of the universe with the ages of stars in globular clusters. This conclusions have been supported by growingly robust observational evidence from distant supernovae.5) (2. the Supernova Cosmology Project [68] and the Highredshift Supernova Project [69]. One of the surprises revealed by these observations is that high redshift type Ia supernovae appear fainter than expected for either an open (ΩM < 1) or a ﬂat (ΩM = 1) universe. 26 24 Supernova Cosmology Project (ΩΜ. and an open one. and this could be due to an accelerating universe. ΩΛ = 0.10 universe. −0. 26. A. we expect something like a cosmological constant. 0 ) (1. etc. one ﬁnds t0 H0 = 0.05 0. it takes about three weeks to reach its maximum brightness and then it declines over a period of months.ΩΛ) = ( 0. The intensity of the distant ﬂash varies in time. 27: Hubble diagram for the high redshift supernovae found by the SN Cosmology Project. For the present age of the universe of t0 = 13 ± 1 Gyr. see Fig. which corresponds to ΩM = 0.–0. In their quest for the cosmological parameters.93 ± 0.
5 Gyr 2 2 9.1 (1σ).4 Gyr op 1 2 3 1 0 1 2 3 ΩΜ ΩΜ Fig.4 Gyr 13.04 +0. (identiﬁed systematics) . as explained above. and. there are bounds on a cosmological constant that come from the statistics of gravitational lensing. which is used for calibration.28 +0.5 − 1. is now smaller than the distant one.09 +0.2±0. the intense efforts to search for highredshift objects have led to the peculiar situation where the nearby sample.l. absolute calibrations. are Ωﬂat = 0. There has been attempts to ﬁnd crucial systematic effects like evolution. H0 . completeness levels. Hence. The most natural explanation for this is the presence of a cosmological constant. something seems to be acting as a repulsive force on very large scales. for increasing the nearby supernovae sample will provide an important check. Moreover. for any value of h ≥ 0. ΩΛ ) plane. lensed by intervening galaxies. ΩΛ) plane. etc. but none of them are now considered as a serious threat.05 −0. and in particular to observations of supernovae at large redshifts.2 Gyr ΩΛ ΩΛ 1 1 0 Flat Λ=0 Universe expands forever lly recollapses eventua ting lera g cce eratin a cel de 0 ed os t cl fla en 1 0 7. of matter. in the supernovae observations that would invalidate the claims. for the high redshift supernovae results. The probability of ﬁnding a lensed image . because of possibly large systematic errors inherent to most cosmological measurements.9 Gyr 11. Gravitational lensing can be due to various accumulations of matter along the line of sight to the distant light sources. The systematic uncertainty is not shown.08 −0. the bestﬁt values for the combined analysis of both groups [68. The upperleft shaded region represents “bouncing universe” cosmologies with no Big Bang in the past. with the bestﬁt 68% and 90% conﬁdence regions in the (ΩM . [68]. For instance.72 +0. already underway.05 (identiﬁed systematics) . The right ﬁgure shows the isochrones of constant H 0 t0 . the age of −1 the universe in units of the Hubble time.) in the (Ω M . for a ﬂat universe (ΩM +ΩΛ = 1).0) are all measured relative to the same set of local supernovae (z < 0.5. (22).8ΩM−0.3). since the luminosities of the highredshift supernovae (z ∼ 0. 28: The left ﬁgure shows the bestﬁt conﬁdence regions (68% – 99% c. 69]. and would shift the ellipses vertically.08 = 0. The ﬁrst method uses the abundance of multiply imaged sources like quasars. The bestﬁt results from the Supernova Cosmology Project give a linear combination 0. chemical composition dependence. with two different methods.09 (1σ statistical) M −0. Further searches. a diffuse vacuum energy that permeates all space and.04 −0. Perhaps the most critical one today seems to be sampling effects. (84) (85) Ωﬂat Λ (1σ statistical) However.6ΩΛ = −0. see Eq. and any other systematic effects related to both data sets are critical. one may think that it is still premature to conclude that the universe is indeed accelerating.3 No Big Bang 3 99% 95% 90% 68% H0t0 65 km s1 Mpc1 = 18. reddening by dust. The lowerright shaded region corresponds to a universe that is younger than the oldest heavy elements. gives the universe an acceleration that tends to separate gravitationally bound systems from each other. Present observations disfavour the Eisnteinde Sitter model (circle) by several standard deviations. From Ref.
observations of the twopoint correlation function of temperature anisotropies in the microwave background provide a crucial test for the spatial curvature of the universe. and nuclear reaction rates. there were very few stars in the Hipparcos catalog with both small parallax erros and low metal abundance. From the observational side. For about 30 years. During the 1980s and 1990s.4 The spatial curvature ΩK As we will discuss in detail in Section 4. thus allowing for a new calibration of the ages of stars in globular clusters. However. 3. Furthermore. therefore.7. rule out the ΩM = 1 models and set an upper bound on the cosmological constant.4. Ω0 = ΩM + ΩΛ = 1. Moreover. to be launched by NASA at the end of year 2000 [72]. Hence. 3. consideration of mixing.) (86) has recently been obtained [70]. see Ref. this limit is very sensitive to the resolution of the numerical simulations. an increase in the sample size could be critical in reducing the statatistical uncertaintites for the calibration of the globular cluster ages. for ﬁxed H0. with the launch in 2007 of Planck satellite [73] we will be able to determine Ω0 with 1% accuracy. the dominant source of systematic errors in the globular cluster age is the uncertainty in the cluster distances. Those are believed to be the stars in the oldest clusters in the Milky Way. marginally consistent with the supernovae results.l. called GAIA. and different chemical abundances have been incorporated [76]. ΩΛ < 0. Very recent observations made by the balloon experiment BOOMERANG suggest that the universe is indeed spatially ﬂat (ΩK = 0) with about 10% accuracy [71].5 The age of the universe t0 The universe must be older than the oldest objects it contains. This distance. NASA’s Space Interferometry Mission (SIM) and ESA’s mission. but there are caveats to this powerful method due to uncertainties in the number density and lensing cross section of the lensing galaxies as well as the distant quasars. the globular cluster age estimates have improved as both new observations have been made with CCDs.0 ± 0. From those observations one can tell whether the photons that left the last scattering surface. which produces widely separated lensed images of quasars and distorted images of background galaxies.l. Using this method. including opacities. or in curved paths. uncertainties in globular cluster ages come from uncertainties in convection models. globular clusters. like in a ﬂat universe. the Hipparcos satellite recently provided geometric parallax measurements for many nearby old stars with low metallicity. Fortunately. at redshift z = 1100. and since reﬁnements to stellar evolution models. by both balloon experiments and by the Microwave Anisotropy Probe (MAP) satellite. when compared with numerical simulations. more lenses are predicted for ΩΛ > 0. uncertainties arise due to corrections for dust and chemical composition.is directly proportional to the number of galaxies (lenses) along the path and thus to the distance to the source. typical of glubular clusters. have travelled in straight lines. recently these ages have been revised downward [75]. [47]. The most reliable ages come from the application of theoretical models of stellar evolution to observations of old stars in globular clusters. increases dramatically for a large value of the cosmological constant: the age of the universe and the distance to the galaxy become large for ΩΛ = 0 because the universe has been expanding for a longer time. The observed statistics. From the theory side. There are already proposed two new parallax satellites. the ages of globular clusters have remained reasonable stable. opacities. However.) (87) These measuremnts are bound to be improved in the near future. an upper limit of ΩΛ < 0. A second method is lensing by massive clusters of galaxies.75 (95% c.1 (95% c. which are currently improving. at about 15 Gyr [74]. However. down . leading to a downward revision to 10 − 13 Gyr [76]. that will give 2 or 3 orders of magnitude more accurate parallaxes than Hipparcos. like in an open one.
0 (0.65 h−1) Gyr [68]. Note that there isn’t a single age estimate more than 2σ away from the average. The last three points correspond to the combined analysis of 8 different measurements.55. The age of the Sun is accurately known and is included for reference. One could also superimpose the contour lines corresponding to equal t0 H0 lines. mainly systematic. There are still many uncertainties. Fig. where all the different observations seem to converge.64.to fainter magnitude limits.1 (88) Furthermore. doublelobed radio sources and the Hubble constant. Error bars indicate 1σ limits. We have plotted that region in Fig.4 +1.2. The supernovae groups can also determine the age of the universe from their high redshift observations. a combination of 8 independent recent measurements: CMB anisotropies. for h = 0. Figure 28 shows that the conﬁdence regions in the (ΩM . We can summarize this Section by showing the region in parameter space where we stand nowadays. In the .4 (0. cluster masstolight ratios. according to this analysis. equivalently.06 or equivalently t0 = 14. which indicates a relatively weak dependence on h. The age would be somewhat larger in a ﬂat universe: H0 tﬂat = 0. cluster baryon fraction. as a cross check. 29. Integrating over ΩM and ΩΛ . thanks to the recent cosmological observations. however.07 tﬂat = 14. 29: The recent estimates of the age of the universe and that of the oldest objects in our galaxy. t0 = 13. but the standard EdS model does not satisfy this unless h < 0. [77]. can be used to determine the present age of the universe [77].68 and 7. cluster abundance evolution.4±1. type Ia SNe. 0. For any value of the Hubble constant less than H0 = 70 km/s/Mpc. those are quickly decreasing and becoming predominantly statistical. 30. allowing enough time for the oldest stars in globular clusters to evolve [76]. deuteriumtohidrogen ratios in quasar spectra. distance errors are likely to be the largest source of systematic uncertainty to the globular cluster age [21].93 ± 0. The best ﬁt value for the age of the universe is.1 ± 1. the implied age of the universe is greater than 13 Gyr. however.96 +0. From Ref. The result is shown in Fig. The averages of the ages of the Galactic Halo and Disk are shaded in gray.6 Gyr. about a billion years younger than other recent estimates [77]. Until larger samples are available. ΩΛ) plane are almost parallel to the contours of constant age. It is extraordinary that only in the last few months we have been able to reduce the concordance region to where it stands today. The result t0 > tgal is logically inevitable. the best ﬁt value of the age in Hubbletime units is H0 t0 = 0. for several orders of magnitude more stars.09 or.65 h−1) Gyr . compared to other recent determinations. 0 −1. [68] 0 −0.
Today. This assumption leads to the natural conclusion that accurate observations of the present state of the universe may shed light onto processes and physical laws at energies above those reachable by particle accelerators. and that there are many unresolved issues related to those problems. most of which are initial conditions’ problems. ΩM/2−ΩΛ . Their difference. we will be able to reduce those uncertainties to the level of one percent. We will see that this is a very optimistic approach indeed. . these observations are conﬁrmed to within a few percent accuracy. This is the reason why cosmologists are so excited and why it is claimed that we live in the Golden Age of Cosmology. 30: The concordance region. There is the reasonable assumption that these cosmological problems will be solved or explained by new physical principles at high energies. These three effects have been probed by recent observations. The Einsteinde Sitter model is no longer the preferred one. and determines how the expansion rate changes with time. near future. from large scale structure (cluster data). the abundance of light elements. this theory leaves a range of crucial questions unanswered. temperature anisotropies (microwave background data) and the universe expansion (supernova data). However. and have helped establish the hot Big Bang as the preferred model of the universe. whether spatially ﬂat. present or future. The best model today is a ﬂat model with a third of the energy density in the form of nonrelativistic matter and two thirds in the form of vacuum energy or a cosmological constant. with many independent observational checks: the expansion of the universe. From Ref. there might be in the near future reasons to be optimistic. characterizes the relative strength of expansion and gravity. at present all observations seem to lie within a naroow region in parameter space. The sum ΩM + ΩΛ gives the total cosmic energy content and determines the geometry of spacetime. All the physics involved in the above observations is routinely tested in the laboratory (atomic and nuclear physics experiments) or in the solar system (general relativity). However. Furthermore. whether accelerating or decelerating. with precise observations of the anisotropies in the microwave background temperature and polarization. to be discussed in Section 4. in the early universe. Surprisingly enough. a balance between the two densities determines the fate of the universe. a predicted age of the universe compatible with the age of the oldest objects in it. [78]. whether it will expand forever or recollapse. 4 THE INFLATIONARY PARADIGM The hot Big Bang theory is nowadays a very robust ediﬁce. the cosmic microwave background. open or closed.4. and the formation of structure via gravitational collapse of initially small inhomogeneities.2 Range of Supernova data Constant expansion Ω COSMOLOGICAL CONSTANT 1 New preferred model ng rati g ele n Acc lerati Asymptote to ce De Einstein's original static model Old standard model Steady Expansion 0 Range of cluster data Range of microwave background data Recollapse 0 se d 1 0 Ru led t0 out < 9 by Gy ag e r C Fl lo at n pe O 1 Ω MATTER 2 3 Fig.
nor the origin of the Big Bang itself.1 Shortcomings of Big Bang Cosmology The Big Bang theory could not explain the origin of matter and structure in the universe. . without which the universe today would be ﬁlled by a uniform radiation continuosly expanding and cooling. and thus without the possibility to form gravitationally bound systems like galaxies. the origin of the matter–antimatter asymmetry. 3 3 Ω We can therefore deﬁne a new variable. via gravitational collapse. it is directly related. but cannot explain. the origin of the extraordinary smoothness and ﬂatness of the universe on the very large scales seen by the microwave background probes and the largest galaxy catalogs. ρa3. Moreover. the quantity and nature of the dark matter that we believe holds the universe together. to the dynamics. (91) dN where N = ln(a/ai ) characterizes the number of efolds of universe expansion (dN = Hdt) and where we have used Eq.Did the universe have a beginning? . (92) . There is nothing in the theory that determines this parameter a priori.Why is the universe so close to spatial ﬂatness? . = Ω ρa2 (90) (1 + zeq) . K= x≡ whose time evolution is given by dx = (1 + 3ω) x . of the universe. tiny deviations from Ω = 1 would have grown since then.e. that is. which only depends on the barotropic ratio ω. such that today x = x0 = Tin Ω0 − 1 = xin Ω0 Teq 2 (89) Ω−1 const. . clusters and superclusters. 8πG 2 Ω − 1 8πG 2 ρa − H 2 a2 = ρa . It is clear from Eq. . . x ) presents an unstable critical (saddle) point at x = 0 for ω > −1/3.What is the nature of the dark matter in the universe? . . However. and thus the matter content.4. (91) that the phasespace diagram (x. stars and planets that could sustain life. via the Friedmann equation (8). A small perturbation from x = 0 will drive the system towards x = ±∞.How did the primordial spectrum of density perturbations originate? • The origin of matter and radiation. A summary [79] of the problems that the Big Bang theory cannot explain is: • The global structure of the universe. It cannot explain the origin of the primordial density perturbations that gave rise to cosmic structures like galaxies. i.How did the matterantimatter asymmetry arise? • The initial singularity. the standard Big Bang theory assumes. In the general FRW metric (2) the parameter K that characterizes spatial curvature is a free parameter.1 The Flatness Problem The Big Bang theory assumes but cannot explain the extraordinary spatial ﬂatness of our local patch of the universe. Since we know the universe went through both the radiation era (because of primordial nucleosynthesis) and the matter era (because of structure formation). with no traces of matter.Why is matter so homogeneously distributed on large scales? • The origin of structure in the universe.1. for the radiation (ω = 1/3) and matter (ω = 0) eras.What is the global structure of the universe beyond our observable patch? Let me discuss one by one the different issues: 4. (30) for the time evolution of the total energy.Where does all the energy in the universe come from? .
85]. 31. 1. again 60 orders of magnitude smaller than h the present age of the universe.38 × 1028 cm. i. a ∼ t1/2. the ﬂatness problem is also related to the Age Problem. or about 1◦ projected in the sky today. much before inﬂation. a new paradigm. the horizon distance is equal to the Hubble scale. 4. under ordinary circumstances (based on the fundamental scale of gravity) it should have lasted only a Planck time and reached a size of order the Planck length? As we will see. it is required that at. 84. that is. 16 For the radiation era. Therefore. but it is epistemologically more satisfying if we give a fundamental dynamical reason for the universe to have started so close to spatial ﬂatness. Information cannot travel faster than the speed of light. for the time of decoupling. t dH (t) ≡ a(t) 0 dt ∼ H −1 (t) . (95) which. were never in causal contact before.2 The Homogeneity Problem An expanding universe has particle horizons.1 × 1017 s. the early universe went through a period of exponential expansion. For the matter era it is twice the Hubble scale. t0 = 4. For instance. These arguments were ﬁrst used by Robert Dicke in the 1960s. A universe with this immense curvature would have collapsed within a Planck time.39 × 10−44 s.2 Cosmological Inﬂation In the 1980s. say. is of order NCD (zdec) ∼ 105 This phenomenon is particularly acute in the case of the observed microwave background. Guth [81].1. Andrei D. is NCD (z) ∼ a(t) dH (t) 3 (1 + z)3/2 .16 For instance. 4. that is. spatial regions beyond which causal communication cannot occur.1 < Ω0 < 1. and was ﬁrst discussed by Robert Dicke in the 1970s as a profound inconsistency of the Big Bang theory. Linde [82] and others [83. than the scale factor. primordial nucleosynthesis (TNS 106 Teq) its value be Ω(tNS ) = 1 ± 10−15 .e. to within 10 ppm. but grows linearly with time and by the end of nucleosynthesis it is a few lightminutes. a0 = 1.62 × 10 cm. tP = (¯ G/c5)1/2 = 5. see Fig. while the scale factor has increased only a factor of 10. a(t ) (94) which is proportional to the Hubble scale. The fact that the causal horizon increases faster.2. inﬂation gives a dynamical reason to such a peculiar initial condition.In order that today’s value be in the range 0. dH (t0 ) ≡ a0 . according to the Big Bang theory. where the Planck length 3 1/2 −33 is lP = (¯ G/c ) h = 1. why is it that the universe is so old and ﬂat when. the number of causally disconnected regions at a given redshift z present in our causal volume today. at the beginning of nucleosynthesis the horizon distance is a few lightseconds. According to the inﬂationary paradigm. was put forward by Alan H. The horizon distance can be deﬁned as the maximum distance that light could have travelled since the origin of the universe [8]. when the photons that come from those two distant regions could not have been in causal contact when they were emitted? This constitutes the socalled horizon problem. So why should regions that are separated by more than 1◦ in the sky today have exactly the same temperature. a factor 100 larger. dH ∼ t. implies that at any given time the universe contains regions within itself that. (93) which represents a tremendous ﬁnetuning. 60 orders of magnitude smaller than the present size of the universe. deeply rooted in fundamental physics. He argued that the most natural initial condition for the 2 spatial curvature should have been the Planck scale curvature. so the causal region at the time of photon decoupling could not be larger than dH (tdec) ∼ 3 × 105 light years across. or x0 ≈ O(1). (3)R = 6K/lP. . to address these fundamental questions. Perhaps the universe indeed started with such a peculiar initial condition.
From Ref. the volume that gave rise to our present universe contained many causally disconnected regions (top ﬁgure). Moreover. T1 = T2 . but a constant energy density acts in a very peculiar way: as a repulsive force that makes any two points in space separate at exponentially large speeds. Any inhomogeneities present before the tremendous expansion would be washed out. the inﬂaton ﬁeld has. but were stretched to cosmological distances by the expansion. 31. for the universe to be nearly ﬂat today. magnetic and gravitational ﬁelds. driven by the approximately constant energy density of a scalar ﬁeld called the inﬂaton. in particular. This extreme ﬁne tuning of initial conditions was also solved by the inﬂationary paradigm. it is simply the stretching of spacetime. one in which the gravitational attraction of matter is exactly balanced by the cosmic expansion. in the usual Big Bang scenario a ﬂat universe. which is spectacularly solved by inﬂation. in fact.3 eV Universe expansion (z = 1100) T0 = 3 K Our observable universe today T1 T2 T1 = T2 Fig. in causal contact before inﬂation. see Fig. 33. In modern physics. is unstable under perturbations: a small deviation from ﬂatness is ampliﬁed and soon produces either an empty universe or a collapsed one. 6. This explains why photons from supposedly causally disconneted regions have actually the same spectral distribution with the same temperature. see Fig. Today we observe a blackbody spectrum of photons coming from those regions and they appear to have the same temperature. At the time of decoupling. it must have been extremely ﬂat at nucleosynthesis. deviations not exceeding more than one part in 1015.Our Hubble radius at decoupling Tdec = 0. much greater that our own observable universe could have become . (This does not violate the laws of causality because there is no information carried along in the expansion. Why is the universe so homogeneous? This constitutes the socalled horizon problem. which resemble the familiar electric. As we discussed above. to one part in 105 . 78]. [80. why the microwave background looks so isotropic: regions separated today by more than 1◦ in the sky were. Thus inﬂation is an extremely elegant hypothesis that explains how a region much. a large potential energy density. which drives the exponential expansion during inﬂation. see Fig. 32. A ﬁeld is simply a function of space and time whose quantum oscillations are interpreted as particles. see Fig.) This superluminal expansion is capable of explaining the large scale homogeneity of our observable universe and. elementary particles are represented by quantum ﬁelds. associated with it. In our case. 31: Perhaps the most acute problem of the Big Bang theory is explaining the extraordinary homogeneity and isotropy of the microwave background. We know from general relativity that the density of matter determines the expansion of the universe.
even the scale at which its dynamics determines the superluminal expansion. etc.Inflation Potential energy Value of reheating the universe end inflation inflaton field Fig. In certain cases. These relics are diluted by the superluminal expansion. During inﬂation. the coherent oscillations of the inﬂaton could generate a resonant production of particles which soon thermalize. The Lagrangian for such a ﬁeld in a curved background is Linf = 1 µν g ∂µ φ∂ν φ − V (φ) . the inﬂaton. smooth and ﬂat without recourse to ad hoc initial conditions. 32: The inﬂaton ﬁeld can be represented as a ball rolling down a hill. The only thing we know about this peculiar scalar ﬁeld. reheating the universe. which leaves at most one of these particles per causal horizon. is that it has a mass and a selfinteraction potential V (φ) but we ignore everything else.2. or is it just some effective description of a more fundamental high energy interaction? Hopefully. When the ball starts to oscillate around the bottom of the hill. a singlet under any given interaction. inﬂation dilutes away any “unwanted” relic species that could have remained from early universe phase transitions. the origin of masses. we still do not know the nature of the inﬂaton ﬁeld itself. inﬂation ends and the inﬂaton energy decays into particles. driving the tremendous expansion of the universe. Inﬂation had its original inspiration in the Higgs ﬁeld. and collapsed. and its discovery at the future particle colliders would help understand one of the truly fundamental problems in physics. which are predicted in grand uniﬁed theories and whose energy density could be so large that the universe would have become unstable. In particular. making them harmless to the subsequent evolution of the universe. experiments in particle physics might give us a clue to its nature. the scalar ﬁeld supposed to be responsible for the masses of elementary particles (quarks and leptons) and the breaking of the electroweak symmetry. [78]. in the near future.1 Homogeneous scalar ﬁeld dynamics In this subsection I will describe the theoretical basis for the phenomenon of inﬂation. Furthermore. is it some new fundamental scalar ﬁeld in the electroweak symmetry breaking sector. If the experiments discover something completely new and unexpected. the energy density is approximately constant. long ago.. 4. cosmic strings. Consider a scalar ﬁeld φ. Such a ﬁeld has not been found yet. with an effective potential V (φ). like monopoles. 2 (96) . From Ref. it would automatically affect the idea of inﬂation at a fundamental level.
78]. . 2 (98) (99) where κ2 ≡ 8πG. (102) ˙ φ2 . (97) where H is the rate of expansion. 2 1 ˙2 φ − V (φ) . analogous (in three dimensions) to how the surface of a balloon appears ﬂatter and ﬂatter as we inﬂate it to enormous sizes. whose evolution equation in a FriedmannRobertsonWalker metric (2) and for a homogeneous ﬁeld φ(t) is given by ¨ ˙ φ + 3H φ + V (φ) = 0 . ρ + 3H(ρ + p) = 0 . together with the Einstein equations. ⇒ H(φ) const. From Ref. 33: The exponential expansion during inﬂation made the radius of curvature of the universe so large that our observable patch of the universe today appears essentialy ﬂat.Fig. This is a crucial prediction of cosmological inﬂation that will be tested to extraordinary accuracy in the next few years. [84. H2 = κ2 1 ˙ 2 φ + V (φ) . then we see (103) . 3 2 κ2 ˙ ˙ H = − φ2 . V (φ) that p −ρ ⇒ ρ const. ˙ If the potential energy density of the scalar ﬁeld dominates the kinetic energy. 2 (100) (101) The ﬁeld evolution equation (97) can then be written as the energy conservation equation. The dynamics of inﬂation can be described as a perfect ﬂuid (7) with a time dependent pressure and energy density given by ρ = p = 1 ˙2 φ + V (φ) .
during inﬂation. Note that we can now have initial conditions with a large uncertainty. for N for N 1. As a consequence. any scale within the horizon during inﬂation will be stretched by the superluminal expansion to enormous distances. see Fig. due to the superluminal expansion. (107) Moreover. thanks to the inﬂationary attractor towards Ω = 1. with p −ρ ⇒ ω −1. inﬂation also solves the homogeneity problem in a spectacular way. we see that the scale factor grows exponentially. However. 33. about 65 efolds before the end. This solution of the Einstein equations solves immediately the ﬂatness problem. 0. for N 1.. (91). (106) Therefore. Suppose that during inﬂation the scale factor increased N efolds. while the horizon distance remains essentially constant.which leads to the solution a(t) ∼ exp(Ht) ⇒ a ¨ >0 a accelerated expansion . This is the reason why two points separated more than 1◦ in the sky have the same backbody temperature. (104) Using the deﬁnition of the number of efolds. any relic particle species (relativistic or not) existing prior to inﬂation will be diluted by the expansion. a(N ) = ai exp(N ). then today the universe (or at least our local patch) should be exactly ﬂat. as observed by the COBE satellite: they were actually in causal contact during inﬂation. very soon it is the only energy density remaining to drive the expansion of the universe. if cosmological inﬂation lasted over 65 efolds. (108) (109) −→ Note that the vacuum energy density ρv remains constant under the expansion. as most models predict. First of all. in such a way that at photon decoupling all the causally disconnected regions that encompass our present horizon actually come from a single region during inﬂation. This can be understood very easily by realizing that the three curvature evolves during inﬂation as (3) R= 6K = a2 (3) Rin e−2N −→ 0. since the scale factor grows exponentially. N = ln(a/ai). 17 . for N 1. There could be a small delay in thermalization. then x0 = xin e−2N Trh Teq 2 (1 + zeq ) e−2N 1056 ≤ 1 ⇒ N ≥ 65 . and the transfer of the inﬂaton energy density to thermal radiation at reheating occurred almost instantaneously17 at the temperature Trh ∼ 1/4 Vend ∼ 1015 GeV. due to the intrinsic inefﬁciency of reheating. a prediction that can be tested with great accuracy in the near future and for which already seems to be some evidence from observations of the microwave background [71]. dH (t) H −1 = const. Recall that the problem with the radiation and matter eras is that Ω = 1 (x = 0) is an unstable critical point in phasespace. xin 1. becomes a natural prediction of inﬂation. Finally. any inhomogeneity existing prior to inﬂation will be washed out. ρM ∝ a−3 ∼ e−3N ρR ∝ a −4 ∼ e −4N −→ 0. see Eq. but this does not change signiﬁcantly the required number of efolds. we have that 1 + 3ω ≥ 0 and therefore x = 0 is a stable attractor of the equations of motion. δk ∼ k aH 2 Φk ∝ e−2N −→ 0. and therefore. 1. There is at present no other proposal known that could solve the homogeneity problem without invoquing an acausal mechanism like inﬂation. (105) where we have assumed that inﬂation ended at the scale Vend. Furthermore. and still have today x0 1. what seemed an ad hoc initial condition.
2 (φ) (113) which is an exact expression in terms of (φ). not to be confused with conformal time (see next Section). and thus we can always evaluate the inﬂationary trajectory in phasespace within the SRA. Suppose that. for the horizon distance to grow more slowly than the scale factor. . which states that the slowroll approximation is an attractor of the equations of motion. V (φ) (118) a very useful expression for evaluating N for a given effective scalar potential V (φ). the evolution equations (97) and (98) become H2 1 − H2 = κ2 V (φ) . In fact.e. δ ≡ − ˙ Hφ ≡ − It is easy to see that the condition a ¨ >0 (112) a characterizes inﬂation: it is all you need for superluminal expansion. as well as for the spatial curvature to decay faster than usual. φ) → H(φ). (110). H(φ. In the limit given by Eqs. the initial value of the scalar ﬁeld. then ˙ H (φ) = −κ2 φ/2 and we can rewrite the slowroll parameters (110) as = δ = 2 κ2 H (φ) H(φ) 2 1 2κ2 V (φ) V (φ) 2 1.4. (110) (111) Hdt = φe φi κdφ . 3 (114) (115) 3 δ ˙ 3H φ 1 − 3 ˙ 3H φ = − V (φ) . The number of efolds can also be rewritten in this approximation as N = κ2 φe φi V (φ) dφ . i. in order to solve the homogeneity problem. during inﬂation. then we can deﬁne the slowroll parameters [86]. we will consider the slowroll approximation (SRA). for singleﬁeld inﬂation. it is possible to prove a theorem. Note that this corresponds to a reduction of the dimensionality of phasespace from two to one dimen˙ sions. therefore reducing the number of initial conditions to just one. V (φ) V (φ) 2 (116) ≡ η− 1.2. the scalar ﬁeld evolves very slowly down its effective potential. in order to solve the ﬂatness problem.2 The slowroll approximation In order to simplify the evolution equations during inﬂation. If H(φ) only depends on φ. ˙ ˙ H κ2 φ2 = 2 H 2 H2 ¨ φ 1. (110) as N = ln aend = ai te ti 1. <1 ⇐⇒ The number of efolds during inﬂation can be written with the help of Eq. (117) 2 H (φ) κ2 H(φ) 1 V (φ) 1 − 2 2 V (φ) κ 2κ The last expression deﬁnes the new slowroll parameter η.
R. all with approximately the same amplitude. H2 = . to linear order. φ = φ(η) .4. with a spectrum that is also scale invariant. 4. ˜ Note that inﬂation cannot generate. (121) (122) (123) ˙ where H = aH and φ = aφ. The use of the word spectrum here is closely related to the case of light waves propagating in a medium: a spectrum characterizes the amplitude of each given wavelength. corresponding to a scaleinvariant spectrum. in linear perturbation theory. B. Let us consider. ds2 = a2 (η)[dη 2 − γij dxidxj ] . that is.18 together with the scalar ﬁeld perturbations ds2 = a2(η) (1 + 2A)dη 2 − 2Bi dxi dη − (1 + 2R)γij + 2Eij + 2hij dxidxj . where did the galaxies and clusters of galaxies come from? One of the most astonishing predictions of inﬂation. xi) . one that was not even expected. The four scalar perturbations (A. 2 φ + 2Hφ + a2V (φ) = 0 . is that quantum ﬂuctuations of the inﬂaton ﬁeld are stretched by the exponential expansion and generate largescale perturbations in the metric. (124) (125) The indices {i. and the i denotes covariant derivative with respect to that metric.3 The origin of density perturbations If cosmological inﬂation made the universe so extremely ﬂat and homogeneous. j} label the threedimensional spatial coordinates with metric γij . (126) (127) 18 . ˜ xi = xi + γ ij ξj (η. to explain the distribution of galaxies and clusters of galaxies on very large scales in our observable universe. These patterns of perturbations in the metric are like ﬁngerprints that unequivocally characterize a period of inﬂation. creating a whole spectrum of curvature perturbations. xi) . When matter fell in the troughs of these waves. clusters and superclusters of galaxies. the quantum ﬂuctuations of the scalar ﬁeld will induce metric perturbations which will backreact on the scalar ﬁeld. a vector perturbation. xi). it created density perturbations that collapsed gravitationally to form galaxies. where η = written as (119) (120) dt/a(t) is the conformal time. under which the background equations of motion can be κ2 1 2 φ + a2 V (φ) 3 2 κ2 2 H − H2 = φ . the inﬂaton ﬂuctuations induce waves in the spacetime metric that can be decomposed into different wavelengths. E) i are gauge dependent functions of (η. xi) . modify the spacetime fabric. Under a general coordinate (gauge) transformation [87. In the case of inﬂation. 88] η = η + ξ 0 (η. the most general line element with both scalar and tensor metric perturbations [87].3. according to general relativity. During inﬂation. i hij = hi = 0.1 Gauge invariant perturbation theory Until now we have considered only the unperturbed FRW metric described by a scale factor a(t) and a homogeneous scalar ﬁeld φ(t). Inﬂaton ﬂuctuations are small wave packets of energy that. Perhaps the most interesting aspect of structure formation is the possibility that the detailed knowledge of what seeded galaxies and clusters of galaxies will allow us to test the idea of inﬂation. The gauge invariant tensor perturbation hij corresponds to a transverse traceless gravitational wave. Such a type of spectrum was proposed in the early 1970s (before inﬂation) by Harrison and Zel’dovich [17]. φ = φ(η) + δφ(η.
a2 2 2 (131) (132) (133) (134) where δρ is the gaugeinvariant density perturbation. however. ξ). the energy density is given in terms of a scalar ﬁeld.with arbitrary functions (ξ 0. Φ = Ψ. κ2 k − 3K Ψ = δρ . which substituted into (143) can be integrated to give Φ(z). the above equations simplify enormously to just three independent equations. two gaugeinvariant gravitational potentials [87. and the latter expression is nothing but the Poisson equation for the gravitational potential. to linear order. z κ2 H 2 Φ= (zu − z u) . Let us deﬁne [88] u ≡ aδφ + zΦ . ˜ E = E −ξ. φ z≡a . = H 2 2 (139) (140) Under this redeﬁnition. However. Φ + HΦ = 2 Φ + 3HΦ + (H + H2 )Φ = δφ + 2Hδφ − 2 (135) (136) (137) (138) δφ = 4φ Φ − 2a2 V (φ)Φ − a2 V (φ)δφ . written in relativistic form. there is a gauge invariant combination of variables that allows one to ﬁnd exact solutions. h where a prime denotes derivative with respect to conformal time. 88]. This system of equations seem too difﬁcult to solve at ﬁrst sight. and thus the gaugeinvariant equations for the perturbations on comoving hypersurfaces (constant energy density hypersurfaces) are κ2 [φ δφ − a2 V (φ)δφ] . H z u = 0. u − u− (141) (142) (143) From Equation (141) we can ﬁnd a solution u(z). the scalar and tensor perturbations transform. which are related through the perturbed Einstein equations. 2 κ2 − 2 Φ + 3HΦ + (H + H2 )Φ = − [φ δφ + a2V (φ)δφ] . 2 a2 a2 Φ κ2 zu . Φ = A + (B − E ) + H(B − E ) . It is possible to construct. Ψ = R + H(B − E ) . During inﬂation. . ˜ B = B + ξ0 − ξ . (128) (129) (130) ˜ ij = hij . 2 κ2 φ δφ . ˜ R = R − Hξ 0 . as ˜ A = A − ξ 0 − Hξ 0 . and together with u(z) allow us to obtain δφ(z).
and the scalar ﬁeld’s Fock space is deﬁned through the vacuum condition. we will use the slowroll parameters (110). Note that the perturbed action for the scalar mode u can be written as δS = 1 2 d3x dη (u )2 − ( u)2 + z 2 u . In order to ﬁnd o exact solutions to the mode equation.4. we can write η= z 1 1 = 2 ν2 − . H2 2 a2 z φ = 1+ − δ = 1− . there are models of inﬂation. z (148) The ratio z /z acts like a timedependent potential for this Schr¨ dinger like equation. (153) (154) δ = Hδ +δ+ φ = O( 2 ) . The equations of motion for each mode uk (η) are decoupled in linear perturbation theory. a(t) ∼ tp . but that is can be written as a quantum ﬁeld with its commutation relations (as much as a pion can be described as a quantum ﬁeld). uk + k 2 − z uk = 0 . see Ref. . aH (151) (152) 2 z = H2 [(1 + − δ)(2 − δ) + H−1 ( − δ )] . that give constant slowroll parameters. a a† u(η. Hφ Hz = 1− (149) (150) In terms of these parameters.2 Quantum Field Theory in curved spacetime Until now we have treated the perturbations as classical.3. (149) and (150).. 1− 2 (156) where For instance. . x) = ˆ (2π)3/2 where the creation and annihilation operators satisfy the commutation relation of bosonic ﬁelds. ¨ Hφ In that case. z Note that the slowroll parameters.19 to order = 2H ( − δ) = O( 2 ) . for constant parameters. a ak ak 0 ˆ = 0. [86] H κ2 z 2 = . where = δ = 1/p < 1. [ˆk . (146) (147) Note that we are not assuming that the inﬂaton is a fundamental scalar ﬁeld.. . H 1− ν= (155) 1+ −δ 1 + . like powerlaw inﬂation. we can write the operator d3 k ∗ (145) uk (η) ˆ k eik·x + uk (η) ˆ k e−ik·x . z η 4 19 −1 1 . z (144) In order to quantize the ﬁeld u in the curved background deﬁned by the metric (119). but we should in fact consider the perturbations Φ and δφ as quantum ﬁelds. ˆ† ] = δ 3 (k − k ) . can be taken as constant. the conformal time and the effective potential for the uk mode can be written as η= −1 + H da .
a2 H . see Ref. u/z becomes constant on superhorizon scales. in the case of adiabatic perturbations. with the effective potential given by (156). There will be modes uk with physical wavelengths much smaller than this scale. (143). H −1 . In fact. during inﬂation. while in the opposite limit. and ν is given by (156) in terms of the slowroll parameters. the solution becomes 1 C(ν) k 2ν− 2 Γ(ν) −ν 2 ≡ √ uk  = √ 3 (−kη) 2k Γ( 2 ) 2k aH 3 ν− 1 2 . we ﬁnd Φ = C1 1 − δφ = C1 a2 H a2 a2 dη + C2 C2 . 90]. (164) H z which is constant. the solutions can be written as 1 uk = √ e−ikη 2k uk = C1 z k k aH . there will be modes with physical wavelengths much greater than the Hubble scale. 1 2 Φ. see Ref. appropriately normalized. a2 (162) (163) a2dη − The term proportional to C1 corresponds to the growing solution.δ 1. . and before entering again the horizon during the radiation or matter eras. Fortunately. k/a H. that are well within the de Sitter horizon and therefore do not feel the curvature of spacetime.We are now going to search for approximate solutions of the mode equation (148). On the other hand. (142) we can write the evolution equation for ζ = u as ζ = H 2 Φ. for k aH. to the gauge invariant curvature metric perturbation Rc on comoving (constant energy density) hypersurfaces. In these two asymptotic regimes. which conﬁrms that z 20) superhorizon modes. . We would like to write an expression for a gauge invariant quantity that is also constant for superhorizon modes. (159) e uk (η) = 2 (1) where Hν (z) is the Hankel function of the ﬁrst kind [89]. √ π i(ν+ 1 ) π 1/2 (1) 2 2 (−η) Hν (−kη) . For approximately constant slowroll parameters one can ﬁnd exact solutions to (148). In quaside Sitter there is a characteristic scale given by the (event) horizon size or Hubble scale during inﬂation. Therefore. for superhorizon modes. [87. there is such a quantity: 1 u ζ ≡ Φ+ (Φ + HΦ) = . [90]. for k Substituting into Eq. k/a H. while that proportional to C2 corresponds to the decaying solution. which can soon be ignored. (157) (158) In the limit k aH the modes behave like ordinary quantum modes in Minkowsky spacetime. where the effective potential (152) is of order z /z 2H2 in the slowroll approximation. In the limit kη → 0. We can now compute Φ and δφ from the superHubblescale mode solution (158). see Eq. this quantity ζ is identical. k aH. that interpolate between the two asymptotic solutions. (160) C(ν) = 2ν− 2 3 Γ(ν) ν− 1 2 3 (1 − ) Γ( 2 ) 1 for (161) aH. aH . (158). These quantities are gauge invariant but evolve with time outside the horizon. we can evaluate the Newtonian ζ is constant for (adiabatic 20 This conservation fails for entropy or isocurvature perturbations. (165) ζ = Rc + H2 1 Using Eq.
(173) (174) (175) (176) In the limit k aH the modes behave like ordinary quantum modes in Minkowsky spacetime.potential Φk when the perturbation reenters the horizon during radiation/matter eras in terms of the curvature perturbation Rk when it left the Hubble scale during inﬂation. e∗ (k. the potential becomes 1 1 a = 2 µ2 − . the metric perturbation hk becomes constant on superhorizon scales. 1 vk = √ e−ikη 2k vk = C a k k aH . We can now redeﬁne our gauge invariant tensor amplitude as a (171) vk (η) = √ hk (η) . λ) . = 2H2 1 − a 2 η 4 1 1 µ= + . 2 2κ with the tensor ﬁeld hij considered as a quantum ﬁeld. λ) are the two polarization tensors. λ) = e∗ (k. while in the opposite limit. δS = ˆ hij (η. with effective potential given by (173). the solution becomes C(µ) k vk  = √ 2k aH µ− 1 2 . These are identical to Eq.λ eik·x + h. that interpolate between the two asymptotic solutions. λ)eij (k. 2κ which satisﬁes the following evolution equation. decoupled for each mode vk (η) in linear perturbation theory. transverse and traceless conditions eij = eji . ij λ (169) (170) eij (−k. (167) (168) where eij (k. we can evaluate the tensor metric perturbation when it reentered during the radiation or matter era directly in terms of its value during inﬂation. . (177) Since the mode hk becomes constant on superhorizon scales. (146). a vk = 0 . (166) matter era .c. a (2π)3/2 λ=1. 3 5 Let us now compute the tensor or gravitational wave metric perturbations generated during inﬂation. λ) = 4 . analogous to the o term z /z for the scalar metric perturbation. For constant slowroll parameters one can ﬁnd exact solutions to (172). aH . satisfying symmetric. For constant slowroll parameters.2 ki eij = 0 . eii = 0 . x) = d3k hk (η) eij (k. (159) except for the substitution ν → µ. Φk = 1 − H a2 a2 dη Rk = 3 + 3ω Rk = 5 + 3ω 2 3 Rk Rk radiation era . (172) vk + k2 − a The ratio a /a acts like a timedependent potential for this Schr¨ dinger like equation. 1− 2 We can solve equation (172) in the two asymptotic regimes. . In the limit kη → 0. λ) ˆ k. appropriately normalized. ij while the creation and annihilation operators satisfy the usual commutation relation of bosonic ﬁelds. The perturbed action for the tensor mode can be written as a2 1 d3x dη 2 (hij )2 − ( hij )2 . Eq.
(182) where we have used Eqs. Let us consider ﬁrst the scalar metric perturbations Rk . The reason why they took so long to be discovered was that they appear as perturbations in temperature of only one part in 105.3. 2 a 4πk3 3−2µ (181) Pg (k) = 8κ2 k aH ≡ A2 T k aH nT . (116). AS . which enter the horizon at a = k/H. with amplitude AT and tilt nT ≡ d ln Pg (k) 2 = 3 − 2µ = − d ln k 1− −2 < 0 . Soon after COBE. . the power spectrum can be approximated by a powerlaw expression. Therefore. 2 z 4πk3 2 (178) . the tensor power spectrum is scale The metric ﬂuctuations generated during inﬂation are not only responsible for the density perturbations that gave rise to galaxies via gravitational collapse. (179) k3 uk 2 κ2 H = 2π 2 z 2 2 2π k aH 3−2ν ≡ A2 S k aH n−1 where we have used Rk = ζk = uzk and Eq. n−1≡ d ln PR (k) δ−2 = 3 − 2ν = 2 d ln k 1− 2η − 6 . but one should also expect to see such ripples in the metric as temperature anisotropies in the cosmic microwave background. (171) and (177). (117). that is. There are at this moment dozens of ground and balloonborne experiments analysing the anisotropies in the microwave background with angular resolutions from 10◦ to a few arc minutes in the sky. and a tilt.4. or twopoint correlation function in Fourier space. minute deviations in the temperature of the blackbody spectrum when we look at different directions in the sky. to obtain from inﬂation a scalar tilt which is either positive (n > 1) or negative (n < 1). until COBE satellite discovered them in 1992. we can have signiﬁcant departures from scale invariance. see Fig. 4.λ0 = 4 k. Furthermore. (160). 0h∗ hk . in principle. In the slowroll approximation. depending on the particular inﬂationary model [91]. but we should also be able to measure its power spectrum.3 Power spectrum of scalar and tensor metric perturbations Not only do we expect to measure the amplitude of the metric perturbations generated during inﬂation and responsible for the anisotropies in the CMB and density ﬂuctuations in LSS. 6. Note from this equation that it is possible. Such anisotropies had been looked for ever since Penzias and Wilson’s discovery of the CMB. Its correlator is given by [86] 0R∗ Rk 0 = k PR (k) = uk 2 3 PR(k) δ (k − k ) ≡ (2π)3 δ 3 (k − k ) . but had eluded all detection. This last equation determines the power spectrum in terms of its amplitude at horizoncrossing. which enter the horizon at a = k/H. Let us consider now the tensor (gravitational wave) metric perturbation.λ H 2π 2 λ Pg (k) 2κ2 vk 2 δ 3(k − k ) ≡ (2π)3 δ 3 (k − k ) . see Fig.4 The anisotropies of the microwave background 1. (180) see Eqs. (183) which is always negative. at higher multipole numbers or smaller angular scales. 34. invariant. other groups quickly conﬁrmed the detection of temperature anisotropies at around 30 µK.
4. pressure: photons scatter off baryons which fall into gravitational potential wells and the two competing forces create acoustic waves of compression and rarefaction. see Fig. so they feel the gravitational potential associated with the perturbations imprinted in the metric during inﬂation. their velocity wave is exactly 90◦ offphase with the acoustic waves. Finally. that entered much earlier than decoupling. which have gone through several acoustic oscillations before last scattering. δν/ν = Φ. If the gravitational potential is not constant.4. the photons will escape from a larger or smaller potential well than they fell in. so their frequency is also blue. All these perturbations of different wavelengths leave their imprint in the CMB anisotropies. are entering now. 4. 34. The largest wavelengths. through a series of acoustic compressions and rarefactions. An overdensity of baryons (protons and neutrons) does not collapse under the effect of gravity until it enters the causal Hubble radius. characterized by Φ in the comoving gauge. at decoupling.or redshifted. the acoustic oscillations occur also in the photon ﬁeld and induces a pattern of peaks in the temperature anisotropies in the sky. 4. The temperature anisotropy induced by these three effects is therefore given by [92] δT (r) = Φ(r. δρ/ρ = −2Φ for adiabatic perturbations. which can be seen as secondary peaks in the power spectrum. after they entered the Hubble scale. The universe just before recombination is a very tightly coupled ﬂuid. and as a consequence their frequency is gravitationally blue. t)dt + − . They have minimum velocity at maximum compression and rarefaction. and carry energy. And there are perturbations with wavelengths much smaller than the size of the horizon at last scattering. Since the surface of last scattering is not a sharp discontinuity. 92]. of projected size about 1◦ in the sky today. Perturbations with wavelengths smaller than these will have gone. Moreover. Perturbations with exactly that wavelength are undergoing their ﬁrst contraction. The perturbation continues to grow until radiation pressure opposes gravity and sets up acoustic oscillations in the plasma. Photons scatter off charged particles (protons and electrons). all the way to the time of radiationmatter equality. That is the reason why we don’t see all the acoustic oscillations with the same amplitude. see Fig. or acoustic compression. The baryons at the time of decoupling do not feel the gravitational attraction of perturbations with wavelength greater than the size of the horizon at last scattering. gravity: photons fall in and escape off gravitational potential wells. due to photon diffusion [93. of size comparable to our present horizon. see Fig. There are three different effects that determine the temperature anisotropies we observe in the CMB.or redshifted. First.1 Acoustic oscillations in the plasma The physics of the CMB anisotropies is relatively simple [92]. they will oscillate in phase. similarly to what neutrinos or HDM do for structure on small scales. because of causality. 3 ρ c tdec t0 (184) Metric perturbations of different wavelengths enter the horizon at different times. These waves induce a Doppler effect on the frequency of the photons. That is. will erase the perturbation on that scale. which entered precisely at decoupling. a phenomenon known as the ReesSciama effect. but in fact they decay exponentialy towards smaller angular scales. 34. Since overdensities of the same size will enter the Hubble radius at the same time. tdec) + 2 T 1 δρ r·v ˙ Φ(r. very similar to sound waves. an effect known as Silk damping. we . there will be scales for which photons. at different angular scales. but a region of ∆z ∼ 100. Those perturbations induce a large peak in the temperature anisotropies power spectrum. since photons scatter off these baryons. Second. travelling from one energy concentration to another. due to the large electromagnetic Thomson cross section (46). There are perturbations with wavelengths comparable to the size of the horizon at the time of last scattering. velocity: baryons accelerate as they fall into potential wells.4.2 The SachsWolfe effect The anisotropies corresponding to large angular scales are only generated via gravitational redshift and density perturbations through the Einstein equations.
the scalar metric perturbations can be separated into Φ(η. most of the photon’s trajectory towards us is unperturbed. x) ≡ Φ(η) Q(x).Fig. In case of a nonvanishing Λ. π (188) . where Q(x) are the scalar harmonics. The theoretical curve (thick line) illustrates a particular model which ﬁts the data. θ. since its contribution is negligible in the past. (187) (186) In the case of a ﬂat universe without cosmological constant. during the matter era and ignore that factor. 2 Qklm (r. the present conformal time. These functions have the general form [96] Qklm (r. φ) + 2 T 3 η0 ηLS dr Φ (η0 − r) Q(r. eigenfunctions of the Laplacian in three dimensions. (185) where η0 is the coordinate distance to the last scattering surface. see Ref. 34: There are at present dozens of ground and balloonborne experiments looking at the microwave background temperature anisotropies with angular resolutions from 10◦ to a few arc minutes in the sky. see Eq. Φ + 3H Φ + a2 Λ Φ − 2K Φ = 0 . φ) are the usual spherical harmonics [89]. In linear perturbation theory. [94]. the Newtonian potential remains constant during the matter era and only the intrinsic SW effect contributes to δT /T . since the perturbation is noncausal. corresponding to multipole numbers l = 2 − 3000. φ) . In that case. θ. while ηLS 0 determines that comoving hypersurface. φ) = Φ(ηLS) Q(η0. we have to know the evolution of the metric perturbation during the matter era. Present observations suggest the existence of a peak in the angular distribution. (21). θ. θ. can ignore the Doppler contribution. In order to compute the temperature anisotropy associated with the SachsWolfe effect. the intrinsic and the Integrated SachsWolfe (ISW) effect. due to integration along the line of sight of time variations in the gravitational potential. We will consider here the approximation Φ = const. [97]. φ) = Πkl (r) Ylm (θ. φ) . and the only difference with respect to the Λ = 0 case is an overall factor [66]. From Ref. i. where Ylm (θ. The above expression is known as the SachsWolfe effect [95]. In a ﬂat universe. φ). θ.e. the radial part of the eigenfunctions (186) can be written as [96] Πkl (r) = 2 k jl(kr) . as predicted by inﬂation. φ) = −k2 Qklm (r. the temperature anisotropy in the sky today is given by [95] δT 1 (θ. and contains two parts.
In that case. C(θ). see Eq. Tensor metric perturbations also contribute with an approximately constant angular power spectrum. The radial part of the tensor harmonic Qrr in a ﬂat universe can be written as [96] (l − 1)l(l + 1)(l + 2) 1/2 jl (kr) . θ. see Eq. The tensor perturbation h during the matter era satisﬁes the following evolution equation hk + 3H hk + (k2 + 2K) hk = 0 . In the case of scalar metric perturbation produced during inﬂation. in the powerlaw S approximation. The growing mode solution of the metric perturbation that left the Hubble scale during inﬂation contributes to the temperature anisotropies on large scales (185) as ∞ l δT 1 1 (θ. normalized so that Gk (0) = 1 at the surface of last scattering. AS 3 n−1 2 25 Γ[ 2 − 2 ] Γ[l + 2 − n−1 ] 2 (S) (192) l(l + 1) Cl 2π = A2 S = constant . φ) = T η0 ηLS dr h (η0 − r) Qrr (r. see Fig. the scalar power spectrum at reentry is given by PR (k) = A2 (kη0)n−1 . Since the coefﬁcients alm are isotropic (to ﬁrst order). one can integrate (191) to give Cl (S) = 3 2π 2 Γ[ 2 ] Γ[1 − n−1 ] Γ[l + n−1 ] 2 . see Eq. deﬁned as an expansion in multipole number. θ. (189) and (178). and we have expanded δT /T in spherical harmonics. contrary to what happens with the scalar modes. 34. (179). and we have averaged over different universe realizations.where jl (z) are the spherical Bessel functions [89]. l(l + 1)Cl . of the CMB anisotropies on large scales. φ) . (190) where Pl (z) are the Legendre polynomials [89]. we can compute the Cl = alm 2 as Cl (S) = 4π 25 ∞ 0 dk PR(k) jl2(kη0 ) . C(θ) = δT ∗ δT (n) (n ) T T = n·n =cos θ 1 4π ∞ l=2 (2l + 1) Cl Pl (cos θ) . and is the reason why the coefﬁcients Cl are always plotted multiplied by l(l + 1). φ) ≡ alm Ylm (θ. We can now compute the twopoint correlation function or angular power spectrum. (195) which depends on the wavenumber k. (196) Qrr (r) = kl πk2 r2 . For a ﬂat (K = 0) universe. The SachsWolfe effect for a gauge invariant tensor perturbation is given by [95] δT (θ. φ) = Φ(ηLS) Q = R Q(η0. (166). (193) This last expression corresponds to what is known as the SachsWolfe plateau. k (191) where we have used Eqs. 25 for n = 1 . (187). the solution to this equation is hk (η) = h Gk (η). where h is the constant tensor metric perturbation at horizon crossing and Gk (η) = 3 j1(kη)/kη. (194) where Qrr is the rrcomponent of the tensor harmonic along the line of sight [96]. (189) T 3 5 l=2 m=−l where we have used the fact that at reentry (at the surface of last scattering) the gauge invariant Newtonian potential Φ is related to the curvature perturbation R at Hubblecrossing during inﬂation. φ) .
n and nT . via the consistency relation (200). observations of the microwave background anisotropies suggest that the SachsWolfe plateau exists. two continuous functions. (x0 − x)x2 0 (T ) = ∞ 0 dk 2 Pg (k) Ikl . . A CMB cosmologist before the 1980s would have argued that ad hoc initial conditions could have been at the origin of the homogeneity and ﬂatness of the universe on large scales. but it is still premature to determine the tensor contribution. Assuming that the scalar contribution dominates over the tensor on large scales.e.3 The consistency relation In spite of the success of inﬂation in predicting a homogeneous and isotropic background on which to imprint a scaleinvariant spectrum of inhomogeneities. or a stochastic gravitational wave background in laser interferometers like LIGO or VIRGO [99]. In the slowroll approximation. . It is clear that there must be a relation between the four parameters. Beyond l ∼ 30. k (197) (198) where x ≡ kη. But that could have been a coincidence. 4. one determines. we can integrate (197) to give [98] l(l + 1) Cl (T ) = 48π 2 π 1+ A2 Bl . For a scale invariant spectrum.4.00) for l = 2.. we might have a chance of determining the validity of the consistency relation. . l(l + 1) Cl also becomes constant for large l. (199) and (193) that the ratio of the tensor to scalar contribution to the angular power spectrum is proportional to the tensor tilt [86]. The surprise was that inﬂation incorporated an understanding of both the globally homogeneous and spatially ﬂat background. 1.1184. i. For the moment. 0. with sufﬁcient accuracy to determine their spectral tilt. while a LSS cosmologist would have agreed with Harrison and Zel’dovich that the most natural spectrum needed to explain the formation of structure was a scaleinvariant spectrum. one can actually give a measure of the amplitude of the scalar metric perturbation from the observations of . . nT = 0. with the CMB satellites MAP and Planck. the SachsWolfe expression is not a good approximation and the tensor angular power spectrum decays very quickly at large l. Therefore. from the analysis of polarization as well as temperature anisotropies. and two tilts. 34. R≡ Cl (T ) (S) Cl = 48π 2 25 1+ 2 9 385 −2π nT . the inﬂaton potential V (φ). 30. and Pg (k) is the primordial tensor spectrum (182). it is difﬁcult to test the idea of inﬂation.The tensor angular power spectrum can ﬁnally be expressed as Cl 9π (l − 1)l(l + 1)(l + 2) 4 x0 j2(x0 − x)jl (x) dx Ikl = . . 3. Indeed.40. corresponding to the scalar (density) and tensor (gravitational waves) metric perturbations. and the approximately scaleinvariant spectrum of perturbations in the same formalism. that in the powerlaw approximation reduces to two amplitudes. Perhaps in the near future. R 1. AS and AT . see Fig. one might have some chance to test the idea of inﬂation. see Fig. . one can see from Eqs. from a single continuous function. PR (k) and Pg (k). from V (φ). . T 36 385 (T ) (199) with Bl = (1. and is not epistemologically testable. What is unique to inﬂation is the fact that inﬂation determines not just one but two primordial spectra. which could not have been postulated a priori by any cosmologist. (200) This is a unique prediction of inﬂation.8789. If we ﬁnally observe a tensor spectrum of anisotropies in the CMB.
These observations conﬁrm the existence of a primordial spectrum of scalar (density) perturbations on all scales. Since photons scatter off baryons. Since photons travel along geodesics. about 1◦ projected in the sky today. they will also feel the acoustic wave and create a peak in the correlation function. we see that the spectrum starts to rise around l = 20 towards the ﬁrst acoustic peak. the ﬁrst peak in the photon distribution corresponds to overdensities that have undergone half an oscillation. The position of the peak in the power spectrum depends on the geometrical size of the particle horizon at last scattering.03 ± 0. [100]. the projected size of the causal horizon at decoupling depends on whether the universe is ﬂat. In the near future these parameters will be determined with much better accuracy.07) × 10−5 .4. otherwise the power spectrum would have started from zero at l = 2. Fig. as described in Section 4.02 ± 0. 35: Theoretical predictions for CMB temperature angular power spectra as a function of multipole number l for models with primordial adiabatic perturbations. 5 (201) (202) n = 1. Each graph shows the effect of a variation in one of these parameters. 34.5. However. The height of the peak is proportional to the amount of baryons: the larger the baryon content of the universe.12 . 4. we would be measuring . by looking at the angular scale of the ﬁrst acoustic peak. and appear at a scale associated with the size of the horizon at last scattering. In a ﬂat universe the geodesics are straight lines and.the SachsWolfe plateau in the angular power spectrum [97]. the higher the peak. open or closed. From Ref.4 The acoustic peaks The SachsWolfe plateau is a distinctive feature of Fig. l(l + 1) Cl 2π (S) 1/2 = AS = (1. These measurements can be used to normalize the primordial spectrum and determine the parameters of the model of inﬂation [91]. a compression. As mentioned above. that is.4. where the SW approximation breaks down and the above formulae are no longer valid.
In an open universe. These peaks should occur at harmonics of the ﬁrst one. and the recent balloonborne experiment BOOMERANG [71].26. From Ref.55. respectively. 0. after marginalizing over the ΩM − ΩΛ direction. and gives very strong support to the idea that inﬂation may indeed be responsible for both the CMB . and therefore the universe is most probably ﬂat. with an amplitude δT = 80 ± 10 µK. In this case.7).65). and other cosmological parameters. 35.95. but also the primordial spectrum of density perturbations produced during inﬂation. TOCO97 and TOCO98 [101]. At the moment there is not enough information at small angular scales. 36. 0. the ﬁrst peak occurs at smaller multipoles or larger angular scales. Since the amplitude and position of the primary and secondary peaks are directly determined by the sound speed (and. with the new microwave anisotropy satellites. 0.l. to determine the existence or not of the secondary acoustic peaks. hence. specially the ones of the Mobile Anisotropy Telescope (MAT) in Cerro Tololo. which produced two data sets. 36: The left ﬁgure shows the power spectrum of the BOOMERANG experiment with 6 arcminute pixelization. suggest that the peak is between l = 180 and 250 at 95% c. 0. see Fig.05. the universe is ﬂat.95. (202).75. for a closed universe. The dashed curves are open and closed models with ﬁxed Ω 0 = ΩM + ΩΛ = 0. On the other hand.the actual size of the horizon at last scattering. but are typically much lower because of Silk damping. This is remarkable. [102]. see Fig. the geodesics are inwardcurved trajectories. 0. which is much better than we could ever do before. as predicted by inﬂation. h) = (0.0. It turns out that the observed temperature anisotropies are compatible with a scaleinvariant spectrum. The right ﬁgure shows the likelihood function of Ω0 normalized to unity at the peak. 35. see Fig. Fig.05. they can be used as a powerful test of the density of baryons and dark matter. 1. The dependence of the position of the ﬁrst acoustic peak on the spatial curvature can be approximately given by [92] lpeak 220 Ω0 −1/2 . The solid curve is a marginally closed model with (ΩB . and therefore the projected size on the sky appears smaller. ΩM .) (204) That is. the ﬁrst acoustic peak should occur at higher multipoles or smaller angular scales. n. In the near future we will measure Ω0 to within 1%. Present observations. ΩΛ. see Eq. and Ref. 0. The dotted curve is Standard CDM with (0.. within 10% uncertainty. cosmologists can determine not only the cosmological parameters. 0. these measuremts determine that 0. the equation of state) and by the geometry and expansion of the universe.l.0. Chile.85 ≤ Ω0 ≤ 1.66 and 1. (203) where Ω0 = ΩM + ΩΛ = 1 − ΩK . By looking at these patterns in the anisotropies of the microwave background. In particular.25 (68% c. [71]. or large multipole numbers.
5 The new microwave anisotropy satellites. textures. the cosmological constant (vacuum energy). down to arc minutes. and thus can be used to compare with observations. and Planck [73] is expected in 2007. with a resolution 100 times better. MAP and Planck The large amount of information encoded in the anisotropies of the microwave background is the reason why both NASA and the European Space Agency have decided to launch two independent satellites to measure the CMB temperature and polarization anisotropies to unprecendented accuracy. the reionization parameter (optical depth to the last scattering surface). analogous to the defects formed in the laboratory in certain kinds of liquid crystals when they go through a phase transition. thanks to large numbers of microwave horns of various sizes. perhaps. Different models of inﬂation have different speciﬁc predictions for the ﬁne details associated with the spectrum generated during inﬂation. 4. That is why it is so important to launch more sensitive instruments. vortices.4. the fact that these anisotropies have such a small amplitude allow for an accurate calculation of the predicted anisotropies in linear perturbation theory. most importantly. and various primordial spectrum parameters like the amplitude and tilt of the adiabatic and isocurvature spectra. with respect to the previous COBE satellite. the spatial curvature. and with better angular resolution. etc. with a typical sensitivity of 0. while the advantage of bolometers is their tremendous sensitivity. better than 0. These are complicated networks of energy density concentrations left over from an early universe phase transition. As we have emphasized before.1 . nonGaussian effects. However. the pattern of anisotropies predicted by inﬂation is completely different from those predicted by alternative models of structure formation. The cosmological defects have spectral properties very different from those generated by inﬂation. like cosmic defects: strings. These two satellites will improve both the sensitivity. the baryon content. The right ﬁgure shows the same. From Ref. as would be seen by a satellite like COBE with angular resolution of 7◦ . down to µK. 37: The left ﬁgure shows a simulation of the temperature anisotropies predicted by a generic model of inﬂation. to determine the properties of the CMB anisotropies. All these parameters can now be fed into a fast code called CMBFAST [103] that computes the predicted temperature and polarization anisotropies to 1% accuracy. It is these minute differences that will allow cosmologists to differentiate between alternative models of inﬂation and discard those that do not agree with observations. The Microwave Anisotropy Probe [72] will be launched by NASA at the end of 2000.anisotropies and the largescale structure of the universe. the amount of gravitational waves. the cold dark matter and neutrino contribution. [73]. with high electron mobility transistor ampliﬁers (HEMTs) for frequencies below 100 GHz and bolometers for higher frequencies. but with a satellite like Planck. positioned at speciﬁc angles. The primary advantage of HEMTs is their ease of use and speed.5 mKs1/2. and also thanks to recent advances in detector technology. and the resolution. A particular cosmological model is characterized by a dozen or so parameters: the rate of expansion. etc. Fig.
This simple connection allows for more stringent tests on the inﬂationary paradigm for the generation of metric perturbations. see Fig. [105]. The right ﬁgure gives an estimate of the accuracy with which the power spectrum will be measured by Planck. to appreciate the difference. which compares the present observational status with that which will become available around 2008. the higher the resolution. yet to be discovered. binned into 16 logarithmic intervals in multipole number between l = 2 and l = 1000. 38: The predicted angular power spectrum of temperature anisotropies. which will probably accomplish the same results with similar resolution (in the case of MAP). Probably the most important objective of the future satellites will be the measurement of the CMB polarization anisotropies. the nature of the perturbations.5 From metric perturbations to large scale structure If inﬂation is responsible for the metric perturbations that gave rise to the temperature anisotropies observed in the microwave background. as well as its matter and energy content. before the satellites start producing their own results [104]. This will allow cosmologists to extract information from around 3000 multipoles! Since most of the cosmological parameters have speciﬁc signatures in the height and position of the ﬁrst few acoustic peaks. see Ref. the geometry of the universe. whether CDM. In particular. compared with the present data. For instance. see Table 1. the curlcurl component of the polarization power spectra is nowadays the only means we have to determine the tensor (gravitational wave) contribution to the metric perturbations responsible for temperature anisotropies. From Ref. If such a component is found. 38. . there are other experiments. specially the tilt [100]. whether adiabatic or isocurvature. one could constraint very precisely the model of inﬂation from its spectral properties. see Fig. These anisotropies are predicted by models of structure formation and are expected to arise at the level of microKelvin sensitivities. HDM or mixed CHDM. see Fig. It is only limited by cosmic variance on all the angular scales relevant to primary anisotropies. 37. 39. The complementary information contained in the polarization anisotropies will provide much more stringent constraints on the cosmological parameters than from the temperature anisotropies alone. This provides a very large lever arm for the determination of primordial spectra parameters like the tilt. where the new satellites are aiming at. and thus the better the accuracy with which one will be able to measure those parameters. 4. since it relates the large scales (of order the present horizon) with the smallest scales (on galaxy scales). 34. As an example of the kind of data that these two satellites will be able to provide. then the primordial spectrum of density inhomogeneities induced by the same metric perturbations should also be responsible for the present large scale structure [106]. compare the resolution in the temperature anisotropies that COBE and Planck would observe for the same simulated sky in Fig. the more peaks one is expected to see. like balloonborne and ground interferometers. [104]. Although the satellite probes were designed for the accurate measurement of the CMB temperature anisotropies.mKs1/2. Fig.
(178). there is a linear transfer function T (k). baryons and cold dark matter long before horizon crossing.5. From this expression one can compute the power spectrum. 5 + 3ω (205) where we have assumed K = 0. The most natural . δk ≡ δρk = ρ k aH 2 2 Φk = 3 k aH 2 2 + 2ω Rk . and Λ = 0. n = 1 + 2η − 6 . which may be deﬁned as [86] Rk (ﬁnal) = T (k) Rk (initial) . Then Eq. see Eq. (134). they create density ﬂuctuations via gravitational attraction of the potential wells. (207) To calculate the transfer function one has to specify the initial condition with the relative abundance of photons. (206) with n given by the scalar tilt (180). 39: Theoretical predictions for the four nonzero CMB temperaturepolarization spectra as a function of multipole moment. (166). the metric perturbation has not changed signiﬁcantly. For scales entering the horizon well after matter domination (k−1 keq 81 Mpc). 4. (205) determines the ﬁnal density contrast in terms of the initial one.65. The solid curves are the corresponding predictions if the COBE anisotropy were entirely due to a stochastic gravitational wave background with a ﬂat scaleinvariant spectrum (with the same cosmological parameters). From Ref. so that Rk (ﬁnal) = Rk (initial). The density contrast δ can be deduced from the Einstein equations in linear perturbation theory.024. at horizon crossing. P (k) = δk 2 = A k aH n .10 9 11 10 l(l+1)Cl  l(l+1)Cl 10 10 TG TT 10 12 10 13 10 14 10 11 10 10 100 1000 15 10 100 1000 l 10 10 12 13 l GG l(l+1)Cl 10 14 l(l+1)Cl CC 10 13 10 14 10 15 10 15 10 10 16 16 10 100 1000 10 100 1000 l l Fig. On smaller scales. This spectrum reduces to a HarrisonZel’dovich spectrum (49) in the slowroll approximation: η. of matter density perturbations induced by inﬂation. The panel for ClCC contains no dotted curve because scalar perturbations produce no curl component of the polarization vector.1 The galaxy power spectrum As metric perturbations enter the causal horizon during the radiation or matter era. [100]. neutrinos. ΩB h2 = 0. Since perturbations evolve after entering the horizon. The dotted curves are the predictions for a COBEnormalized scalar perturbation from an inﬂationary model with no reionization and no gravitational waves for h = 0. see Eq. the power spectrum will not remain con−1 stant. 1. and used Eq.
2 − 1.0 0.5% 10% 15% Table 1: The parameters of the standard cosmological model. ρ (209) 2 where ω = p/ρ is the barotropic ratio.5 0 − 1.01 − 1. see Ref. The transfer function is therefore given roughly by. T (k) = 1. δρX = ρX δt. The rate of expansion is in units of H0 = 100 h km/s/Mpc. One might then consider the opposite assumption. This is called the adiabatic condition. see Eq.e. ˙ ¨ H −2 δk + [2 − 3(2ω − c2 )] H −1δk − s 3 k (1 − 6c2 + 8ω − 3ω 2 ) δk = − s 2 aH 2 δpk . needed to describe the background spacetime.001 − 0.01 − 0.3 0 − 0.8 11 − 17 Gyr 20 − 30 µK 0. (k/keq)2 . k k keq keq (210) The perfect ﬂuid description of the radiation is far from being correct after horizon entry. for scales below −1 the Silk scale.0 MAP − 5% 10% 5% 8% 4% 11% 10% 0. with the relation δm = 4 δr . given a perturbation in time from a comoving ˙ hypersurface. The photons are also not a perfect ﬂuid because they diffuse signiﬁcantly. (52). and cs = p/ρ is the speed of sound of the ﬂuid. 3 and each species of matter has also a common density contrast δm .0 0 − 0. its density perturbation oscillates during this era. The matter density contrast living in this background does not grow appreciably before matter domination because it has negligible selfgravity. [86]. valid to ˙ ﬁrst order in perturbations.7% 2% 2% 0. If the radiation behaves like a perfect ﬂuid. (208) ρX + pX ρY + p Y where we have used the energy conservation equation for each species. kS ∼ 1 Mpc. because roughly half of the radiation consists of neutrinos whose perturbation rapidly disappears through free streeming. and the percentage error with which the microwave background probes MAP and Planck (without polarization) will be able to determine them in the near future. ˙ ˙ Given the adiabatic condition. ρX = −3H(ρX + pX ). condition is that the abundances of all particle species are uniform on comoving hypersurfaces (with constant total energy density).5% 3% 25% 20% Planck − 0.5 0.005 0.6% 0. the density perturbation amplitude evolves according to the following equation. the transfer function is determined by the physical processes occuring between horizon entry and matter domination. the matter content and the spectrum of density perturbations. that the radiation . It follows that each species of radiation has a common density contrast δr .physical quantity luminous matter baryonic matter cold dark matter hot dark matter cosmological constant spatial curvature rate of expansion age of the universe spectral amplitude spectral tilt tensorscalar ratio reionization symbol Ωlum h2 Ω B h2 ΩM h2 Ω ν h2 Ω Λ h2 Ω0 h2 h t0 Qrms nS rts τ present range 0. We include here the present range of the most relevant parameters.2 − 1. Within the horizon. The standard model of cosmology has around 12 different parameters.03 0. so δρY δρX = .8 0. with decreasing amplitude.1% 0.5 − 1.4 − 0.5% 0.6% 2% 0. because entropy is conserved independently for each particle species X. i.
which is the time scale for gravitational collapse. ceases to be valid when the density contrast becomes of order 1. for scales below Silk scale. b = 3. if τpres > τgrav. 2 1+y . (211) which corresponds to the equation of a damped harmonic oscillator. to each Fourier component. which free streem. although the break at k = keq is not at all sharp. After that. All these effects apply separately. The zerofrequency oscillator deﬁnes the Jeans wavenumber. 40. gravitational collapse of a perturbation can occur before pressure forces can response to restore hydrostatic equilibrium (this occurs for λ > λJ ). [107]. (215) (216) (217) (218) h h −1 Mpc . There are several parametrizations in the literature. and y = a/aeq. the diffusion of photons around the same time. with δ and ρ now referring to the matter alone. as the baryons fall into the potential wells of cold dark matter. For k kJ .e. which either form a perfect ﬂuid or just diffuse. T (k) = 1 + ak + (bk)3/2 + (ck)2 a = 6. The transfer function is therefore again roughly given by Eq.4 (ΩMh)−1 h−1 Mpc . then. The growing mode solution (213) increases only by a factor of 2 between horizon entry and the epoch when matter starts to dominate. which encodes the soltion to linear equations. A more precise calculation is needed. ν = 1. 2 2 ¨ ˙ δk + 2H δk + (cs kph − 4πGρ) δk = 0 .has zero perturbation after horizon entry. the highly nonlinear phenomenon of gravitational collapse takes place. (211) are s δ = A 1+ δ=B 3 y .13 . One can also deﬁne the Jeans length.0 (ΩMh) c = 1. . Then the matter density perturbation evolves according to Eq. including: neutrino free streeming around the epoch of horizon entry. The transfer function. y = 1. If we deﬁne the pressure response timescale as the size of the perturbation over the sound speed. 40. to ﬁrst order in the perturbations. (210) is roughly correct. i. the diffusion of baryons along with the photons. and the establishment after matter domination of a common matter density contrast. (210). We will consider now the behaviour of modes within the horizon during the transition from the 2 radiation (cs = 1/3) to the matter era (c2 = 0). so that a linear transfer function is produced.7 (ΩMh) −1 ν −1/ν . δk grows exponentially on the dynamical s −1 −1/2 timescale. Since the radiation consists roughly half of neutrinos. (209). see Fig. and half of photons. The growing and the decaying solutions of Eq. neither the perfect ﬂuid nor the freestreeming approximation looks very sensible. if τpres < τgrav . see Fig. radiation pressure prevents gravitational collapse and there are damped acoustic oscillations (for λ < λJ ). Gρ (212) which separates gravitationally stable from unstable modes. (213) (214) √ 1+y+1 3 1 + y ln √ −3 2 1+y−1 where A and B are constants. −1 −1 We see that the behaviour estimated in Eq. λJ = 2π = cs kJ π . τdyn = Im ω = (4πGρ) = τgrav . τpres ∼ λ/cs . kJ = 4πGρ/c2. On the other hand. but the one which is more widely used is that of Ref. Mpc .
While the linear power spectrum falls off like k −3 . [19]. 2dF and Sloan Digital Sky Survey Our view of the largescale distribution of luminous objects in the universe has changed dramatically during the last 25 years [19]: from the simple pre1975 picture of a distribution of ﬁeld and cluster galaxies. similar to that of previous surveys like the ESP. in logarithmic scale. normalized to the local abundance of galaxy clusters. advances in theoretical modeling of the development of structure. 4. The advantages of multiobject ﬁbre spectroscopy have been pushed to the extreme with the construction of the 2dF spectrograph for the prime focus of the AngloAustralian Telescope [111]. see Ref.000 galaxies measured by January 2000. [111]. the ability to collect spectra of several galaxies at once. . This is the basis of the 2dF galaxy redshift survey. See also Ref.000 galaxies with bJ < 19. with large highresolution gravitational simulations coupled to a deeper yet limited understanding of how to form galaxies within the dark matter halos.5. and there were about 93. At the same time. the nonlinear powerspectrum illustrates the increased power on small scales due to nonlinear effects. to the discovery of the ﬁrst single superstructures and voids. to the most recent results showing an almost regular weblike network of interconnected clusters. a faint redshift survey of 10. and Fig. The increased efﬁciency of redshift surveys. [47]. thanks to ﬁbreoptic spectrographs). This instrument is able to accommodate 400 automatically positioned ﬁbres over a 2 degree in diameter ﬁeld. have provided a more realistic connection of the models to the observable quantities [110]. made possible by the development of spectrographs and – specially in the last decade – by an enormous increase in multiplexing gain (i. but also to statistically characterize some of its properties. Despite the large uncertainties that still exist. or slightly more than one night of telescope time with typical 1 hour exposures. Its goal is to measure redshifts for more than 250. 41.5. which are taking data at the moment and which will revolutionize the ﬁeld. has allowed us not only to do cartography of the nearby universe. [109]. The survey is steadily collecting redshifts.e.5. the same number of redshifts as in the ESP survey can be collected in about 10 exposures. In addition. with the difference that with such an area yield. The solid (dashed) curve shows the linear (nonlinear) power spectrum. and an optimal match to the galaxy counts for a magnitude bJ 19. Here I will concentrate on two of the new catalogs.000 galaxies brighter than R = 21 will be done over selected ﬁelds within the two main strips of the South and North Galactic Caps. where theory and observations can progress side by side. ﬁlaments and walls. at the expense of the largescale structures. the 2degreeField (2dF) Catalog and the Sloan Digital Sky Survey (SDSS). From Ref.5. where the survey is continuously updated. for an Einsteinde Sitter universe with h = 0. see Ref. This implies a density of ﬁbres on the sky of approximately 130 deg−2 .2 The new redshift catalogs.Fig. 40: The CDM power spectrum P (k) as a function of wavenumber k. this has transformed the study of cosmology and largescale structure into a truly quantitative science. separating huge nearly empty volumes. For a review of the variety and details about the different existing redshift catalogs.
No corrections are made for redshift distortions or nonlinear evolution. This is pre . which will be observed spectroscopically. ΩM = 1 is assumed. The most ambitious and comprehensive galaxy survey currently in progress is without any doubt the Sloan Digital Sky Survey [112]. This has already led to the discovery of several highredshift (z > 4) quasars.5. Green error bars show the accuracy of the Sloan Digital Sky Survey and magenta error bars are for the 2 Degree Field Survey. or nonlinear evolution have been made. The black box comes from measurements of σ 8 from presentday number abundances of rich clusters. together with the future CMB anisotropy probes. the spectroscopic part of the survey will then collect spectra from about 106 galaxies with r < 18 and 105 AGNs with r < 19. for the determination of the cosmological parameters of the standard model of cosmology. often one ﬁnds combinations of various experiments/observations which break the degeneracy. However. providing a nearly volumelimited sample of earlytype galaxies with a median redshift of z 0. MAP (red boxes) and Planck (blue boxes) are simulated assuming that CHDM is the correct model.8 to 23. at limiting magnitudes from 20. No corrections for bias. It will also select a sample of about 105 red luminous galaxies with r < 19. We mentioned in the previous Section that some of the cosmological parameters created similar effects in the temperature anisotropies of the microwave background. 30◦ away from the galactic plane (about 104 deg2 ) in ﬁve bands. As often happens in particle physics. that will be extremely valuable to study the evolution of clustering.CLUSTERS IRAS CFA2+SSRS2 LCRS APM Fig. We say that these parameters are degenerate with respect to the observations. 41: Compilation of largescale structure observations. [112]. Some of the redshift surveys have been rebinned to make the points nearly independent. and the black point with error bars is from peculiar velocities. redshift distortions. From Ref. Using two ﬁbre spectrographs carrying 320 ﬁbres each. at z = 5. not always are observations from a single experiment sufﬁcient to isolate and determine the precise value of the parameters of the standard model. for example by depending on a different combination of parameters. The right panel shows a simulation of highprecision future CMB and LSS observations. [108].3. see Figs. The data expected to arise from these new catalogs is so outstanding that already cosmologists are making simulations and predicting what will be the scientiﬁc outcome of these surveys. The simulated data are indistinguishable from the underlying CHDM model for a wide range of k. The expectation is to detect around 50 million galaxies and around 108 starlike sources. showing the power spectrum P (k) as a function of wavenumber k. The height shows the 68% conﬁdence interval.0. see Ref. 41 and 42. The aim of the project is ﬁrst of all to observe photometrically the whole Northern Galactic Cap. including the highestredshift quasar known.5.
were a host of highprecision measurements are already posing challenges to our understanding of the universe: the density of ordinary matter and the total amount of energy in the universe. The right ﬁgure shows the same as before. [113]. SDSS (k max = 0. that is not . [113]. as shown in Fig.Fig. Physical science advances by incorporating earlier theories that are experimentally supported into larger. primordial deuterium abundance from quasar absorption lines. However. All cosmological issues are now formulated in the context of the inﬂationary paradigm. The standard Big Bang theory is supported by a wealth of evidence. and determine most of the 12 or more parameters of the standard cosmological model to a few percent accuracy (see table 1). A projecton of future supernovae Ia results gives the solid vertical lines as bounds.35 ﬂat ΛCDM model. In the next few years we will have an even larger set of highquality observations that will test inﬂation and the cold dark matter paradigm of structure formation. more encompassing frameworks. a feature named somewhat idiosyncratically as “cosmic complementarity”. the microwave background anisotropies on a ﬁnescale resolution. 30. which already put constraints on the parameter space of cosmological models. cisely the case with the cosmological parameters. combined with MAP gives the solid ellipse. However. see Ref. MAP data with polarization yields the ellipse from upper left to lower right. large scale structure measurements of the distribution of galaxies and their evolution. and will make cosmology a phenomenological science. assuming the universe ﬂat gives a smaller region (shortdashed line). From Ref. nobody really doubts its validity anymore. 42. see Fig. A direct 10% measurement of H0 gives the longdashed lines and ellipse.1h Mpc−1 ) gives the vertical shaded region. It is important to bear in mind that all physical theories are approximations of reality that can fail if pushed too far. which has become the new standard cosmological model. 5 CONCLUSION We have entered a new era in cosmology. However. It is expected that in the near future we will be able to determine the parameters of the standard cosmological model with great precision from a combination of several different experiments. in the last decade it has been incorporated into the larger picture of cosmological inﬂation. It may seem that with such a large number of parameters one can ﬁt almost anything. as measured by a combination of largescale structure observations. All regions are 68% conﬁdence. Supernovae Ia observations and Hubble Space Telescope measurements. The ﬁducial model is the Ω M = 0. microwave background anisotropies. but for constraints in the ΩM − ΩΛ plane. 42: Constraint regions in the ΩM −H0 plane from various combinations of data sets. the acceleration parameter of the universe from highredshift supernovae observations. the rate of expansion from gravitational lensing. It is the best explanation we have at the moment for the increasing set of cosmological observations. these are only the forerunners of the precision era in cosmology that will dominate the new millennium. and many more. combined with MAP gives the small ﬁlled ellipse.
the scalar ﬁeld supposed to be responsible for the masses of elementary particles (quarks and leptons) and the breaking of the electroweak symmetry. with around 21 parameters and a host of precise measurements from particle accelerators all over the world. One of the most difﬁcult challenges that the new cosmology will have to face is understanding the origin of the cosmological constant. and its parameters measured to a precision of better than 1% in some cases. In fact. on truly macroscopic scales. the universe was much smaller than it is now. and it has proved particularly difﬁcult to construct a consistent quantum theory of gravity. In fact. It is clear that highprecision measurements will make the standard model of cosmology as robust as that of particle physics. However. ACKNOWLEDGEMENTS I thank the organizers of the CERNJINR European School of High Energy Physics for a very warm and friendly atmosphere. The value of the cosmological constant predicted by quantum physics is related to our lack of understanding of gravity at the microscopic level. experiments in particle physics might give us a clue to its nature. This model is. since it involves fundamental issues like causality and the nature of spacetime itself. There are still many unanswered fundamental questions in this emerging picture of cosmology. where speculation has given way to phenomenology. We still do not have a mechanism to explain its extraordinarily small value. Ever since Einstein introduced it as a way to counteract gravitational attraction. and its discovery at the future particle colliders would help understand one of the truly fundamental problems in physics. is it some new fundamental scalar ﬁeld in the electroweak symmetry breaking sector. An illustrative example is the standard model of particle physics. the origin of masses. This is deﬁnitely a very healthy ﬁeld. For instance. cosmology is becoming a mature science. Such a ﬁeld has not been found yet. We can speculate that perhaps general relativity is not the correct description of gravity on the very largest scales. This hints at what is known in quantum theory as an anomaly. nowadays. As it expanded. 120 orders of magnitude below what is predicted by quantum physics. In its infancy. presumably. If this were indeed the case. With the advent of better and larger precision experiments. deviations from general relativity would slowly become important. larger and larger regions were encompassed. but there is still a lot to do. if indeed it is conﬁrmed by independent sets of observations. we still do not know the nature of the inﬂaton ﬁeld. or is it just some effective description of a more fundamental high energy interaction? Hopefully. Andrew . in the near future. it would automatically affect inﬂation at a fundamental level. General relativity is a classical theory of spacetime. general relativity gave a correct description of its evolution. If the experiments discover something completely new and unexpected. it is only in the last few billion years that the observable universe has become large enough that these global effects could be noticeable. it has haunted cosmologists and particle physicists for decades. and. I also would like to thank my friends and collaborators Andrei Linde. we should expect that the new generation of precise cosmological observations will not only affect our cosmological model of the universe but also a more fundamental description of nature. as conﬁrmed by the successes of the standard Big Bang theory. therefore. rigurously tested. and. Inﬂation had its original inspiration in the Higgs ﬁeld. it has been the technological advances of particle physics detectors that are mainly responsible for the burst of new data coming from cosmological observations. It may well be that the recent determination of a cosmological constant from observations of supernovae at high redshifts is hinting at a fundamental misunderstanding of gravity on the very large scales. For several decades there has been the reasonable speculation that this fundamental problem may be related to the quantization of gravity.the case when there is enough quantity and quality of data. its effect is dominant at the very largest scales of clusters or superclusters of galaxies. a quantum phenomenon relating both ultraviolet (microscopic) and infrared (macroscopic) divergences.
Liddle, David Wands, David Lyth, Jaume Garriga, Xavier Montes, Enrique Gazta˜ aga, Elena Pierpaoli, n Stefano Borgani, and many others, for sharing with me their insight about this fascinating science of cosmology. This work was supported by the Royal Society. References [1] A. Einstein, Sitz. Preuss. Akad. Wiss. Phys. 142 (1917) (§4); Ann. Phys. 69 (1922) 436. [2] A. Friedmann, Z. Phys. 10 (1922) 377. [3] E.P. Hubble, Publ. Nat. Acad. Sci. 15 (1929) 168. [4] G. Gamow, Phys. Rev. 70 (1946) 572; Phys. Rev. 74 (1948) 505. [5] A.A. Penzias and R.W. Wilson, Astrophys. J. 142 (1965) 419. [6] S. Weinberg, “Gravitation and Cosmology”, John Wiley & Sons (1972). [7] P.B. Pal, “Determination of cosmological parameters”, eprint Archive: hepph/9906447. [8] E.W. Kolb and M.S. Turner, “The Early Universe”, Addison Wesley (1990). [9] N. ArkaniHamed, S. Dimopoulos and G. Dvali, Phys. Lett. B 429 (1998) 263, eprint Archive: hepph/9803315. [10] S. Burles, K.M. Nollett, J.N. Truran, M.S. Turner, Phys. Rev. Lett. 82 (1999) 4176, eprint Archive: astroph/9901157; S. Burles, K.M. Nollett, M.S. Turner, “BigBang Nucleosynthesis: Linking Inner Space and Outer Space”, eprint Archive: astroph/9903300. [11] K.A. Olive, G. Steigman and T. Walker, “Primordial Nucleosynthesis: Theory and Observations”, eprint Archive: astroph/9905320. [12] J.C. Mather et al., Astrophys. J. 420 (1994) 439; D.J. Fixen et al., Astrophys. J. 473 (1996) 576; J.C. Mather et al., Astrophys. J. 512 (1999) 511, eprint Archive: astroph/9810373. [13] R.H. Dicke, P.J.E. Peebles, P.G. Roll and D.T. Wilkinson, Astrophys. J. 142 (1965) 414. [14] G.F. Smoot, Astrophys. J. 396 (1992) L1; C.L. Bennett et al., Astrophys. J. 464 (1996) L1, eprint Archive: astroph/9601067. [15] P.J.E. Peebles, “Principles of Physical Cosmology”, Princeton U.P. (1993). [16] T. Padmanabhan, “Structure Formation in the Universe”, Cambridge U.P. (1993). [17] E.R. Harrison, Phys. Rev. D 1 (1970) 2726; Ya. B. Zel’dovich, Astron. Astrophys. 5 (1970) 84. [18] The IRAS Point Source Catalog Web page: http://wwwastro.physics.ox.ac.uk/˜wjs/pscz.html [19] L. Guzzo, “Largescale structure at the turn of the millennium”, 19th Texas Symposium, Paris (1998), eprint Archive: astroph/9911115. [20] P.J. Steinhardt, in Particle and Nuclear Astrophysics and Cosmology in the Next Millennium, ed. by E.W. kolb and R. Peccei (World Scientiﬁc, Singapore, 1995). [21] W.L. Freedman, “Determination of cosmological parameters”, Nobel Symposium (1998), eprint Archive: hepph/9905222.
[22] G.G. Raffelt, “Dark Matter: Motivation, Candidates and Searches”, European Summer School of High Energy Physics 1997. CERN Report pp. 235278, eprint Archive: hepph/9712538. [23] S. Refsdael, Mon. Not. R. Astr. Soc. 128 (1964) 295; 132 (1966) 101. [24] R.D. Blandford and T. Kundi´ , “Gravitational Lensing and the Extragalactic Distance Scale”, ec print Archive: astroph/9611229. [25] N.A. Grogin and R. Narayan, Astrophys. J. 464 (1996) 92, eprint Archive: astroph/9512156. [26] M. Birkinshaw, Phys. Rep. 310 (1999) 97, eprint Archive: astroph/9808050. [27] The Chandra Xray observatory Home Page: http://chandra.harvard.edu/ [28] S. Sakai et al., “The Hubble Space Telescope Key Project on the Extragalactic Distance Scale XXIV: The Calibration of TullyFisher Relations and the Value of the Hubble Constant”, eprint Archive: astroph/9909269. [29] F. Zwicky, Helv. Phys. Acata 6 (1933) 110. [30] M. Davis and J. Huchra, Astrophys. J. 254 (1982) 437; R.D. Kirshner et al., Astron. J. 88 (1983) 1285. [31] C.J. Copi and D.N. Schramm, Comm. Nucl. Part. Phys. 22 (1996) 1, eprint Archive: ph/9504026. [32] K.C. Freeman, Astrophys. J. 160 (1970) 811. [33] K.G. Begeman, A.H. Broeils and R.H. Sanders, Mon. Not. R. Astr. Soc. 249 (1991) 523. [34] C.M. Baugh et al., “Ab initio galaxy formation”, eprint Archive: astroph/9907056; Astrophys. J. 498 (1998) 405; eprint Archive: astroph/9703111. [35] M. Fich and S. Tremaine, Annu. Rev. Astron. Astrophys. 29 (1991) 409. [36] A.D. Linde, J. Garc´aBellido and D. Wands, Phys. Rev. D 54 (1996) 4060. eprint Archive: astroı ph/9605094. [37] E. Roulet and S. Mollerach, Phys. Rep. 279 (1997) 68, eprint Archive: astroph/9603119. [38] The MACHO Collaboration, C. Alcock et al., “Microlensing results from 5.7 years of LMC observations”, eprint Archive: astroph/0001272. n [39] B. Paczy´ ski, Astrophys. J. 304 (1986) 1; Astrophys. J. Lett. 371 (1991) L63; [40] The MACHO Collaboration, C. Alcock et al., Nature 365 (1993) 621; MACHO Home Page at: http://wwwmacho.mcmaster.ca/ [41] The EROS Collaboration, E. Aubourg et al., Nature 365 (1993) 623; EROS Home Page at: http://www.lal.in2p3.fr/recherche/eros/erosa.html [42] The AGAPE Collaboration, P. Baillon et al., A & A 277 (1993) 1; AGAPE Home Page at: http://cdfinfo.in2p3.fr/Experiences/AGAPE/frameen.html [43] A. Dekel, Annu. Rev. Astron. Astrophys. 32 (1994) 371, eprint Archive: astroph/9401022. [44] R.G. Carlberg et al., Astrophys. J. 462 (1996) 32, eprint Archive: astroph/9509034. astro
[45] A. Dekel, D. Burstein and S.D.M. White, “Measuring Omega”, in Critical Dialogues in Cosmology, ed. N. Turok, World Scientiﬁc (1997); eprint Archive: astroph/9611108. [46] C.L. Sarazin, Rev. Mod. Phys. 58 (1986) 1. [47] M. Bartelmann et al., Astron. & Astrophys. 330 (1998) 1, eprint Archive: astroph/9709229; M. Bartelmann and P. Schneider, “Weak Gravitational Lensing”, eprint Archive: astroph/9912508. [48] Hubble Space Telescope Web Page: http://oposite.stsci.edu/pubinfo/subject.html [49] N.A. Bahcall, J.P. Ostriker, S. Perlmutter and P.J. Steinhardt, Science 284 (1999) 1481, eprint Archive: astroph/9906463. [50] N.A. Bahcall and X. Fan, Proc. Nat. Acad. Sci. 95 (1998) 5956, eprint Archive: astroph/9804082; Astrophys. J. 504 (1998) 1, eprint Archive: astroph/9803277. [51] Particle Data Group Home Page, http://www.cern.ch/pdg/1999/lxxx.html [52] W. Hu, D.J. Eisenstein and M. Tegmark, Phys. Rev. Lett. 80 (1998) 5255, eprint Archive: astroph/9712057. [53] S.D. Tremaine and J.E. Gunn, Phys. Rev. Lett. 42 (1979) 407; J. Madsen, Phys. Rev. D 44 (1991) 999. [54] J. Primack, D. Seckel and B. Sadoulet, Ann. Rev. Nucl. Part. Sci. 38 (1988) 751; N.E. Booth, B. Cabrera and E. Fiorini, Ann. Rev. Nucl. Part. Sci. 46 (1996) 471. [55] S.M. Bilenky, “Neutrino masses, mixing and oscillations”, in these Proceedings; eprint Archive: hepph/0001311. [56] R. Bernabei et al., DAMA Collaboration, “Search for WIMP annual modulation signature: results from DAMA/NaI3 and NaI4, and the global combined analysis”, preprint INFN/AE00/01. DAMA Home Page, http://www.lngs.infn.it/lngs/htexts/dama/ [57] J. Ellis, “Limits on Sparticle Dark Matter”, talk presented at COSMO 98, Asilomar, California, November 1998, eprint Archive: astroph/9903003. [58] For a recent review, see A. Morales, “Direct Detection of WIMP Dark Matter”, eprint Archive: astroph/9912554. [59] UK Dark Matter Collaboration (J.J. Quenby et al.), “Dark Matter Experiments at the UK Boulby mine”, at 26th International Cosmic Ray Conference (ICRC 99), in ‘Salt Lake City 1999, Cosmic ray’, Vol. 2, pp. 269–272. Home Page at http://hepwww.rl.ac.uk/ukdmc/ [60] M. Bravin et al., Astropart. Phys. 12 (1999) 107; eprint Archive: hepex/9904005. [61] J.I. Collar et al., “First Dark Matter Limits from a LargeMass, LowBackground Superheated Droplet Detector”, eprint Archive: astroph/0001511. [62] G. Jungman, M. Kamionkowski and K. Griest, Phys. Rep. 267 (1996) 195, eprint Archive: hepph/9506380. [63] The Alpha Matter Spectrometer Home Page: http://ams.cern.ch/AMS/ [64] F. Halzen et al., Phys. Rep. 307 (1998) 243, eprint Archive: hepex/9804007.
Harwood Academic Press (1990). eprint Archive: astroph/9911493. Rev. eprint Archive: astroph/9911445. . Science 271 (1996) 957. “Particle Physics and Inﬂationary Cosmology”. Chaboyer.“Measurement of a Peak in the Cosmic Microwave Background Power Spectrum from the North American test ﬂight of BOOMERANG”. Melchiorri et al. Rep.H. [87] J. Astron. Vth Autumn School of Theoretical Physics. “The Inﬂationary Universe”. P. Astrophys. [84] For a personal historical account. [67] J. J.esa. 494 (1998) 47. Phys. eprint Archive: astroph/9805201. 494 (1998) 96.. “A measurement of Omega from the North American test ﬂight of BOOMERANG”.gsfc. Krauss. Astron. Bolte and P.nl/Planck/ [74] D.S. astro [79] J. 29 (1997) 1351. [69] A. eprint Archive: astroph/9812473. CERN Summer Student Lecture Programme (1997). Carroll. Reiss et al. eprint Archive: astroph/9812133. Lond. Bull. Bardeen. eprint Archive: astroph/9911444..B. 48 (1982) 1220. “Mapping the CMB Sky: The BOOMERANG experiment”.J. Lyth. 116 (1998) 1009.nasa. Soc. Ann... M. Am. Krauss.D.. Rev. E. Nature 377 (1995) 600. [82] A. 108B (1982) 389.J. Demarque. Phys.R. 34 (1996) 461. C. see A. [77] C. de Bernardis et al. Albrecht and P.L. Mu˜ oz. eprint Archive: astroph/9706128. Astron. Weinberg. [83] A. Soc. Lett. “Inﬂationary Cosmology”. Rev. 61 (1989) 1. [86] A. Astrophys. J. eprint Archive: astroph/9707032. P. eprint Archive: astroph/9605064.[65] S. Perseus Books (1997). [80] A. Vandenberg. n [71] P.M. Linde.M. Phys. 466 (1995) 638. Astrophys. eprint Archive: ı ph/9906497. Stetson. Science 284 (1999) 1503. J. Lett. A. D. D 22 (1980) 1882. Guth. Garc´aBellido. Liddle and D. [70] C. Mod. 231 (1993) 1. Ann.A.gov/ [73] Planck Home Page: http://astro. Rev.M. [78] J. Astrophys.estec. [85] A. [76] B.D. Falco. Press and E. Kochanek. “The age of globular clusters”. Phys. Astrophys. Phil.M. Garc´aBellido. Lineweaver. eprint Archive: astroph/9911461. Rev. “Introduction to Cosmology”.J. The Supernova Cosmology Project.P. eprint Archive: astroph/9303019. Astron.H. Trans. 517 (1999) 565. Rev.S. [72] Microwave Anisotropy Probe Home Page: http://map. High Redshift Supernova Project.A. u [81] A. R. de R´ jula. Steinhardt. [66] S. A 357 (1999) 3237. [75] L. W. P. 30 (1992) 499.H. Guth. Steinhardt. eprint Archive: astroph/9907308. Kochanek and J. 453 (1995) 545. Phys.E. Turner. Astrophys. Astrophys. Santiago ı de Compostela (1997). Perlmutter et al. Kernan and L. J. J. D 23 (1981) 347. [68] S. J. Phys. Linde. Mauskopf et al. Ostriker and P.
Hu. Phys. “New physics from the CMB”.F. Kamionkowski and A. R. Astrophys. Part. J. Soc. D 53 (1996) 5437. “The Cosmic Microwave Background Radiation”. Astrophys. eprint Archive: hepph/9807278. eprint Archive: astroph/9911199. [109] G. Harrison.H. [89] M. Abbott and R. 521 (1999) L79. Dover (1972).A.[88] V. Kosowsky. Phys.html [95] R. White. [93] J. Astrophys. Nature 215 (1967) 1155. (Eds. Sov. Silk. Astron. Rep. 314 (1999) 1.A.R. Phys. J.. Wands. [110] B. [105] D. J. “Measuring the anisotropy in the CMB”. Lyth and A. Phys. Stegun. eprint Archive: astroph/9511029. Scott.caltech.edu/˜max/cmb/experiments. E.F.edu/ VIRGO Home Page: http://www. Riotto. [92] D. eprint Archive: astroph/9604166. eprint Archive: astroph/9905100. [107] J. [101] E.R. Trans. 308 (1986) 546.upenn.R. Page.H. Astrophys. D 54 (1996) 5917. Trans. “Handbook of Mathematical Functions”. 147 (1967) 73.K. [98] A. W. Nature 386 (1997) 37. Silk. L. [108] E. J. Bunn.edu/˜matiasz/CMBFAST/cmbfast.. 39 (1967) 862. Silk. Schaefer. Science 280 (1998) 1405. Science 268 (1995) 829. Cambridge University Press (2000). [102] L. Rep. .hep. Page. [97] E. Knox and L. ı [91] D.A. Sugiyama and J. Bond and G.F. Sachs and A.ligo. eprint Archive: astroph/9904108. eprint Archive: astroph/990642. “Cosmological Inﬂation and Large Scale Structure”. Silk and M. Lett. Tegmark web page with latest experiments’ data: http://www. A 357 (1999) 3259. Garc´aBellido and D. 49 (1999) 77. Rev.H.ias. [106] A. J. “Characterizing the Peak in the Cosmic Microwave Background Angular Power Spectrum”.infn. Abramowitz and I. eprint Archive: astroph/0002162. Gawiser and J. 524 (1999) L1. [99] LIGO Home Page: http://www. 215 (1992) 203. Sci. Lond. Silk. A 357 (1999) 1198.M. Torbet et al. Lyth. Phys.virgo. Liddle and D. White. A. 11 (1985) 133. Rev. J. Phil. A. Rev. Miller et al. Moore. Nucl. [94] M.) “Largescale structure in the universe”. Scott. Astrophys. Soc. Phil. R. Ann. eprint Archive: astroph/9806197.html [104] L. Mukhanov. Efstathiou et al. Wolfe. [96] E. H.K. Lond. eprint Archive: astroph/0002044. Feldman and R. Brandenberger. 285 (1984) L45. Liddle and M. eprint Archive: astroph/9505015. Rev. Mod. Starobinsky.sns. [90] J. Gawiser. eprint Archive: astroph/9607038. [103] CMBFAST code Home Page: http://www. N.R. Efstathiou. eprint Archive: astroph/9911325.it/ [100] M. J.
mso.sdss.au/2dFGRS/ [112] The Sloan Digital Sky Survey Home Page: http://www. 518 (1999) 2.org/sdss. Hu and M.[111] The 2 Degree Field Home Page: http://www. eprint Archive: astroph/9807130. Eisenstein. J.anu. Astrophys.J. Tegmark. . W.html [113] D.edu.
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.