4 Cosmological Perturbation Theory

So far, we have treated the universe as perfectly homogeneous. To understand the formation
and evolution of large-scale structures, we have to introduce inhomogeneities. As long as these
perturbations remain relatively small, we can treat them in perturbation theory. In this chapter,
we will develop the formalism of cosmological perturbation theory.
The Einstein equations couple perturbations in the stress-energy tensor to those in the metric,
so the two need to be studied simultaneously. We write the small perturbations of the metric
and the stress-energy tensor as

gµ⌫ (⌘, x) = ḡµ⌫ (⌘) + gµ⌫ (⌘, x) ,

Tµ⌫ (⌘, x) = T̄µ⌫ (⌘) + Tµ⌫ (⌘, x) .

To avoid clutter we will often drop the argument (⌘, x) on the perturbations.

4.1 Metric Perturbations

To avoid unnecessary technical distractions, we will take the background metric ḡµ⌫ to be the
flat FRW metric. The perturbed spacetime can then be written as
h i
ds2 = a2 (⌘) (1 + 2A)d⌘ 2 2Bi dxi d⌘ ( ij + hij )dxi dxj , (4.1.1)

where A, Bi and hij are functions of space and time. We shall adopt the useful convention that
Latin indices on spatial vectors and tensors are raised and lowered with ij , e.g. hi i = ij hij .
It will be extremely useful to perform a scalar-vector-tensor (SVT) decomposition of the
perturbations. For 3-vectors, this should be familiar. It simply means that we can split any
3-vector into the gradient of a scalar and a divergenceless vector

Bi = @i B + B̂i , (4.1.2)
|{z} |{z}
scalar vector

with @ i B̂i = 0. Similarly, any rank-2 symmetric tensor can be written

hij = 2C ij + 2@hi @ji E + 2@(i Êj) + 2Êij , (4.1.3)

| {z } | {z } |{z}
scalar vector tensor

where
⇣ 1 ⌘
2
@hi @ji E ⌘ @i @j ij r E, (4.1.4)
3
1⇣ ⌘
@(i Êj) ⌘ @i Êj + @j Êi . (4.1.5)
2
As before, the hatted quantities are divergenceless, i.e. @ i Êi = 0 and @ i Êij = 0. The tensor
perturbation is traceless, Ê i i = 0. The 10 degrees of freedom of the metric have thus been
decomposed into 4 + 4 + 2 SVT degrees of freedom:

71
72 4. Cosmological Perturbation Theory

• scalars: A, B, C, E

• vectors: B̂i , Êi

• tensors: Êij

What makes the SVT-decomposition so powerful is the fact that the Einstein equations for
scalars, vectors and tensors don’t mix at linear order and can therefore be treated separately.
In these lectures, we will mostly be interested in scalar fluctuations and the associated density
perturbations. Vector perturbations aren’t produced by inflation and even if they were, they
would decay quickly with the expansion of the universe. Tensor perturbations are an important
prediction of inflation and we will discuss them briefly in Chapter 6.

Gauge problem.—The metric perturbations in (4.1.1) aren’t uniquely defined, but depend on our
choice of coordinates or the “gauge choice”. In particular, when we wrote down the perturbed metric,
we implicitly chose a specific time slicing of the spacetime and defined specific spatial coordinates on
these time slices. Making a di↵erent choice of coordinates, can change the values of the perturbation
variables. It may even introduce fictitious perturbations. These are fake perturbations that can arise
by an inconvenient choice of coordinates even if the background is perfectly homogeneous.
Fictitious perturbations.—For example, consider a flat FRW spacetime and make the following change
of the spatial coordinates, xi 7! x̃i = xi + ⇠ i (⌘, x). We assume that ⇠ i is small, so that it can also be
treated as a perturbation. Using dxi = dx̃i @⌘ ⇠ i d⌘ @k ⇠ i dx̃k , the line element becomes
⇥ ⇤
ds2 = a2 (⌘) d⌘ 2 2⇠i0 dx̃i d⌘ ij + 2@(i ⇠j) dx̃i dx̃j , (4.1.6)

where we have dropped terms that are quadratic in ⇠ i and defined ⇠i0 ⌘ @⌘ ⇠i . We apparently have
introduced the metric perturbations Bi = ⇠i0 and Êi = ⇠i . But these are just fictitious gauge modes
that can be removed by going back to the old coordinates.
As another example, consider a change in the time slicing, ⌘ 7! ⌘ + ⇠ 0 (⌘, x). The homogeneous
density of the universe then gets perturbed, ⇢(⌘) 7! ⇢(⌘ + ⇠ 0 (⌘, x)) = ⇢¯(⌘) + ⇢¯ 0 ⇠ 0 . Even in an
unperturbed universe, a change of the time coordinate can therefore introduce a fictitious density
perturbation
⇢ = ⇢¯ 0 ⇠ 0 . (4.1.7)
Conversely, we can also remove a real perturbation in the energy density by choosing the hypersurface
of constant time to coincide with the hypersurface of constant energy density. We then have ⇢ = 0
although there are real inhomogeneities.
These examples illustrate that we need a more physical way to identify true perturbations. One
way to do this is to define perturbations in such a way that they don’t change under a change of
coordinates.

We need to face that fact that the metric perturbations can be changed by a change of
coordinates. Consider the following transformation

X µ 7! X̃ µ ⌘ X µ + ⇠ µ (⌘, x) , where ⇠0 ⌘ T , ⇠ i ⌘ Li = @ i L + L̂i . (4.1.8)

We have split the spatial shift Li into a scalar, L, and a divergenceless vector, L̂i . In the next
insert, I will show how the metric transforms under this change of coordinates. In terms of the
73 4. Cosmological Perturbation Theory

SVT-decomposition, we get

A 7! A T0 HT , (4.1.9)

B 7! B + T L0 , B̂i 7! B̂i L̂0i , (4.1.10)

1 2
C 7! C HT r L, (4.1.11)
3
E 7! E L, Êi 7! Êi L̂i , Êij 7! Êij . (4.1.12)

Derivation.—The trick is to exploit the invariance of the spacetime interval,

ds2 = gµ⌫ (X)dX µ dX ⌫ = g̃↵ (X̃)dX̃ ↵ dX̃ , (4.1.13)

where I have used a di↵erent set of dummy indices on both sides to make the next few lines clearer.
Writing dX̃ ↵ = (@ X̃ ↵ /@X µ )dX µ (and similarly for dX ), we find

@ X̃ ↵ @ X̃
gµ⌫ (X) = g̃↵ (X̃) . (4.1.14)
@X µ @X ⌫
This relates the metric in the old coordinates, gµ⌫ , to the metric in the new coordinates, g̃↵ .
Let us see what (4.1.14) implies for the transformation of the metric perturbations in (4.1.1). I
will work out the 00-component as an example and leave the rest as an exercise. Consider µ = ⌫ = 0
in (4.1.14):
@ X̃ ↵ @ X̃
g00 (X) = g̃↵ (X̃) . (4.1.15)
@⌘ @⌘
The only term that contributes to the l.h.s. is the one with ↵ = = 0. Consider for example ↵ = 0
and = i. The o↵-diagonal component of the metric g̃0i is proportional to B̃i , so it is a first-order
perturbation. But @ X̃ i /@⌘ is proportional to the first-order variable ⇠ i , so the product is second
order and can be neglected. A similar argument holds for ↵ = i and = j. Eq. (4.1.15) therefore
reduces to ✓ ◆2
@ ⌘˜
g00 (X) = g̃00 (X̃) . (4.1.16)
@⌘
Substituting (4.1.8) and (4.1.1), we get
2
a2 (⌘) 1 + 2A = 1 + T 0 a2 (⌘ + T ) 1 + 2Ã
2
= 1 + 2T 0 + · · · a(⌘) + a0 T + · · · 1 + 2Ã

= a2 (⌘) 1 + 2HT + 2T 0 + 2Ã + · · · , (4.1.17)

where H ⌘ a0 /a is the Hubble parameter in conformal time. Hence, we find that at first order, the
metric perturbation A transforms as

A 7! Ã = A T0 HT . (4.1.18)

I leave it to you to repeat the argument for the other metric components and show that

Bi 7! B̃i = Bi + @i T L0i , (4.1.19)

hij 7! h̃ij = hij 2@(i Lj) 2HT ij . (4.1.20)

In terms of the SVT-decomposition this leads to (4.1.9)–(4.1.12).

One way to avoid the gauge problems is to define special combinations of metric perturbations
74 4. Cosmological Perturbation Theory

that do not transform under a change of coordinates. These are the Bardeen variables

⌘ A + H(B E 0 ) + (B E 0 )0 , ˆ i ⌘ Ê 0 B̂i , Êij , (4.1.21)

i
1
⌘ C H(B E 0 ) + r2 E . (4.1.22)
3

Exercise.—Show that , and ˆ i don’t change under a coordinate transformation.

These gauge-invariant variables can be considered as the ‘real’ spacetime perturbations since
they cannot be removed by a gauge transformation.
An alternative (but related) solution to the gauge problem is to fix the gauge and keep track
of all perturbations (metric and matter). For example, we can use the freedom in the gauge
functions T and L in (4.1.8) to set two of the four scalar metric perturbations to zero:

• Newtonian gauge.—The choice

B = E = 0, (4.1.23)
gives the metric
⇥ ⇤
ds2 = a2 (⌘) (1 + 2 )d⌘ 2 (1 2 ) ij dx
i
dxj . (4.1.24)

Here, we have renamed the remaining two metric perturbations, A ⌘ and C ⌘ , in

order to make contact with the Bardeen potentials in (4.1.21) and (4.1.22). For perturba-
tions that decay at spatial infinity, the Newtonian gauge is unique (i.e. the gauge is fixed
completely). In this gauge, the physics appears rather simple since the hypersurfaces of
constant time are orthogonal to the worldlines of observers at rest in the coordinates (since
B = 0) and the induced geometry of the constant-time hypersurfaces is isotropic (since
E = 0). In the absence of anisotropic stress, = . Note the similarity of the metric to
the usual weak-field limit of GR about Minkowski space; we shall see that plays the role
of the gravitational potential. Newtonian gauge will be our preferred gauge for studying
the formation of large-scale structures (see Chapter 5).

• Spatially-flat qauge.—A convenient gauge for computing inflationary perturbations is

C = E = 0. (4.1.25)

In this gauge, we will be able to focus most directly on the fluctuations in the inflaton
field (see Chapter 6) .

4.2 Matter Perturbations

We write the perturbed stress-energy tensor as

T 0 0 = ⇢¯(⌘) + ⇢ , (4.2.26)
T i 0 = [¯
⇢(⌘) + P̄ (⌘)] v i , (4.2.27)
T ij = [P̄ (⌘) + P ] i
j ⇧i j , (4.2.28)
75 4. Cosmological Perturbation Theory

where vi is the bulk velocity and ⇧i j is a transverse and traceless tensor describing anisotropic
stress. We will use q i for the momentum density (¯⇢ + P̄ )v i . In case there are several contributions
to the stress-energy tensor (e.g. photons, baryons, dark matter, etc.), they are added: Tµ⌫ =
P (a)
a Tµ⌫ . This implies
X X X X ij
⇢= ⇢a , P = Pa , q i = i
q(a) , ⇧ij = ⇧(a) . (4.2.29)
a a a a

We see that the perturbations in the density, pressure and anisotropic stress simply add. The
velocities do not add, but the momentum densities do. Finally, we note that the SVT decom-
position can also be applied to the perturbations of the stress-energy tensor: ⇢ and P have
scalar parts only, qi has scalar and vector parts,

qi = @i q + q̂i , (4.2.30)

and ⇧ij has scalar, vector and tensor parts,

ˆ j) + ⇧
⇧ij = @hi @ji ⇧ + @(i ⇧ ˆ ij . (4.2.31)

It is also convenient to write the density perturbations in terms of the dimensionless density
contrast ⌘ ⇢/⇢. In summary, scalar perturbations of the total matter are described by
four perturbation variables, ( , P, v, ⇧). Similarly, the perturbations of distinct species a =
, ⌫, c, b, · · · are represented by ( a , Pa , va , ⇧a ).

Under the coordinate transformation (4.1.8), the stress-energy tensor transform as

@X µ @ X̃
T µ ⌫ (X) = T̃ ↵ (X̃) . (4.2.32)
@ X̃ ↵ @X ⌫
Evaluating this for the di↵erent components, we find

⇢ 7! ⇢ T ⇢¯ 0 , (4.2.33)
0
P 7! P T P̄ , (4.2.34)
qi 7! qi + (¯
⇢+ P̄ )L0i , (4.2.35)
vi 7! vi + L0i , (4.2.36)
⇧ij 7! ⇧ij . (4.2.37)

Exercise.—Confirm eqs. (4.2.33)–(4.2.37). [Hint: First, convince yourself that the inverse of a matrix
of the form 1 + ", were 1 is the identity and " is a small perturbation, is 1 " to first order in ".]

There are various gauge-invariant quantities that can be formed from metric and matter
variables. One useful combination is

⇢¯ ⌘ ⇢ + ⇢¯ 0 (v + B) , (4.2.38)

where vi = @i v. The quantity is called the comoving-gauge density perturbation.

Above we used our gauge freedom to set two of the metric perturbations to zero. Alternatively,
we can define the gauge in the matter sector:
76 4. Cosmological Perturbation Theory

• Uniform density gauge.—We can use the freedom in the time-slicing to set the total density
perturbation to zero
⇢ = 0. (4.2.39)

• Comoving gauge.—Similarly, we can ask for the scalar momentum density to vanish,

q = 0. (4.2.40)

Fluctuations in comoving gauge are most naturally connected to the inflationary initial
conditions. This will be explained in §4.4 and Chapter 6.

There are di↵erent versions of uniform density and comoving gauge depending on which of the
metric fluctuations is set to zero. In these lectures, we will choose B = 0.

4.3 Equations of Motion

In this section, we will determine the linearized equations of motion for the matter and metric
perturbations. We will work in Newtonian gauge where the metric takes the following form
!
2 1+2 0
gµ⌫ = a . (4.3.41)
0 (1 2 ) ij

Exercise.—Show that the connection coefficients associated with the metric (4.3.41) are
0 0
00 =H+ , (4.3.42)
0
i0 = @i , (4.3.43)
i ij
00 = @j , (4.3.44)
0
⇥ 0
⇤
ij = H ij + 2H( + ) ij , (4.3.45)
i
⇥ 0
⇤ i
j0 = H j , (4.3.46)
i i il
jk = 2 (j @k) + jk @l . (4.3.47)

4.3.1 Conservation Equations

The equations of motion for the matter perturbations follow from conservation of the stress
tensor, rµ Tµ⌫ = 0. If there is no energy and momentum transfer between the di↵erent compo-
(a)
nents, then the species are separately conserved and we also have rµ Tµ⌫ = 0. Equipped with
the perturbed connection, we can derive the perturbed conservation equations from

rµ T µ ⌫ = 0
= @µ T µ ⌫ + µ ↵
µ↵ T ⌫
↵ µ
µ⌫ T ↵ . (4.3.48)

• Continuity equation.—For ⌫ = 0, we find

0
@⌘ ⇢ = 3H( ⇢ + P ) + 3 (¯
⇢ + P̄ ) @i q i , (4.3.49)

This is the continuity equation describing the evolution of the density perturbation. The
first term on the right-hand side is just the dilution due to the background expansion [as in
77 4. Cosmological Perturbation Theory

⇢¯0 = 3H(¯ ⇢ + P̄ )], the @i q i term accounts for the local fluid flow due to peculiar velocity,
and the ˙ term is a purely relativistic e↵ect corresponding to the density changes caused
by perturbations to the local expansion rate [(1 )a is the “local scale factor” in the
spatial part of the metric in Newtonian gauge].

• Euler equation.—For ⌫ = i, we find

@⌘ q i = 4Hq i ⇢ + P̄ )@ i
(¯ @i P @j ⇧ij . (4.3.50)

This is the Euler equation for a viscous fluid, i.e. the “F = ma” of the fluid. In §1.2,
we showed that peculiar velocities decay as a 1 . We therefore expected the momentum
density to scale as q / a 4 . This explains the first term on the rhs of (4.3.50). The
remaining terms are force terms.

Derivation.—Consider first the ⌫ = 0 component of (4.3.48),

µ µ
@0 T 0 0 + @i T i 0 + 0
µ0 T 0 + i
µi T 0
0 0
00 T 0
0
T i0 i
T 0i i j
j0 T i =0. (4.3.51)
| {z } | i0{z } | 00{z }
O(2) O(2) O(2)

Substituting the perturbed stress-energy tensor and the connection coefficients gives
0 0
⇢ + ⇢) + @i q i + (H +
@0 (¯ + 3H 3 )(¯
⇢ + ⇢)
0 0 i
⇥ j⇤
(H + )(¯
⇢ + ⇢) (H ) j (P̄ + P ) i = 0, (4.3.52)

and hence

⇢¯0 + @0 ⇢ + @i q i + 3H(¯
⇢ + ⇢) 3¯
⇢ 0
+ 3H(P̄ + P ) 3P̄ 0
= 0. (4.3.53)

Writing the zeroth-order and first-order parts separately, we get

⇢¯0 = 3H(¯
⇢ + P̄ ) , (4.3.54)
0
@⌘ ⇢ = 3H( ⇢ + P ) + 3 (¯
⇢ + P̄ ) r·q. (4.3.55)

The zeroth-order equation (4.3.54) is simply the conservation of energy in the homogeneous back-
ground. The first-order equation (4.3.55) is the continuity equation for the density perturbation ⇢.
Next, consider the ⌫ = i component of (4.3.48), @µ T µ i + µ ⇢
µ⇢ T i
⇢ µ
µi T ⇢ = 0, and hence
µ µ j j
@0 T 0 i + @j T j i + 0
µ0 T i + j
µj T i
0 0
0i T 0
0 j
ji T 0
0
0i T j
k
ki T j =0. (4.3.56)

Substituting the perturbed stress-energy tensor [with T 0 i = qi ] and the connection coefficients gives
h i
@ 0 qi + @ j (P̄ + P ) ij ⇧j i 4Hqi (@j 3@j ) P̄ ij @i ⇢¯
⇣ ⌘
H ji q j + H ij qj + j
2 (i @k) + ki jl @l P̄ jk = 0 . (4.3.57)
| {z }
3@i P̄

Cleaning this up, we find

@ ⌘ qi = 4Hqi (¯
⇢ + P̄ )@i @i P @ j ⇧ij , (4.3.58)

which confirms the form of the Euler equation (4.3.50).

78 4. Cosmological Perturbation Theory

It is instructive to evaluate (4.3.49) and (4.3.50) for a few special cases:

• Matter.—For a non-relativistic fluid (i.e. matter), we have Pm = 0 and ⇧ij

m = 0. The
continuity and Euler equations then simplify considerably. Writing m ⌘ ⇢m /⇢m , we
obtain
0 0
m = r · vm + 3 , (4.3.59)
0
vm = Hv m r . (4.3.60)

Each term in these equations should be rather intuitive. Combining the time derivative of
(4.3.59) with the divergence of (4.3.60), we find
00 0 00 0
m +H m = r2 + 3( +H ). (4.3.61)
" "
friction gravity

In Chapter 5, we will apply this equation to the clustering of dark matter perturbations.

• Radiation.—For a relativistic fluid (i.e. radiation), we have Pr = 13 ⇢r and ⇧ij

r = 0. The
continuity and Euler equations become

0 4
r = r · vr + 4 0 , (4.3.62)
3
1
v r0 = r r r . (4.3.63)
4
Combining the time derivative of (4.3.62) with the divergence of (4.3.63), we get

00 1 2 4 00
r r r = r2 + 4 . (4.3.64)
3 3
" "
pressure gravity

In Chapter 5, we will show how this equation leads to the oscillations in the observed
spectrum of CMB anisotropies.

Exercise.—Show that the most general forms of the continuity and Euler equations are
✓ ◆ ✓ ◆
0 P̄a i 0 Pa P̄a
a = 1 + (@ v
i a 3 ) 3H a , (4.3.65)
⇢¯a ⇢¯a ⇢¯a
✓ ◆
P̄a0 1
vai 0 = H+ vai @ i Pa @j ⇧ija @i . (4.3.66)
⇢¯a + P̄a ⇢¯a + P̄a

Confirm that these expressions reduce to the equations for matter and radiation in the appropriate
limits.

Comments.—The two equations (4.3.65) and (4.3.66) aren’t sufficient to completely describe the
evolution of the four perturbations ( a , Pa , va , ⇧a ). To make progress, we either must make further
simplifying assumptions or find additional evolution equations. We will do both.
79 4. Cosmological Perturbation Theory

• A perfect fluid is characterized by strong interactions which keep the pressure isotropic, ⇧a = 0.
In addition, pressure perturbations satisfy Pa = c2s,a ⇢a , where cs,a is the adiabatic sound
speed of the fluid. The perturbations of a perfect fluid are therefore described by only two
independent variables, say a and va , and the continuity and Euler equations are sufficient for
closing the system.
• Decoupled or weakly interacting species (e.g. neutrinos) cannot be described by a perfect fluid
and the above simplifications for the anisotropic stress and the pressure perturbation do not
apply. In that case, we can’t avoid solving the Boltzmann equation for the evolution of the
perturbed distribution function fa .
• Decoupled cold dark matter is a peculiar case. It is collisionless and has a negligible velocity
dispersion. It therefore behaves like a pressureless perfect fluid although it has no interactions
and therefore really isn’t a fluid.

The di↵erent matter components are gravitationally coupled through the metric fluctuations
in the continuity and Euler equations. The dynamics to the perturbed spacetime is determined,
via the Einstein equations, by the perturbations of the total stress-energy tensor.

4.3.2 Einstein Equations

Let us compute the linearised Einstein equation in Newtonian gauge. This is conceptually
straightforward, although algebraically a bit tedious.1 We require the perturbation to the Ein-
stein tensor, Gµ⌫ ⌘ Rµ⌫ 12 Rgµ⌫ , so we first need to calculate the perturbed Ricci tensor Rµ⌫
and Ricci scalar R.
Ricci tensor.—We recall that the Ricci tensor can be expressed in terms of the connection as
⇢ ⇢
Rµ⌫ = @ µ⌫ @⌫ µ + ⇢ µ⌫ µ ⌫⇢ . (4.3.67)

Substituting the perturbed connection coefficients (4.3.42)–(4.3.47), we find

R00 = 3H 0 + r2 + 3H( 0
+ 0
)+3 00
, (4.3.68)
R0i = 2@i ( 0 + H ) , (4.3.69)
⇥ ⇤
Rij = H 0 + 2H2 00
+ r2 0
2(H + 2H )( + 2
) H 0
5H 0
ij (4.3.70)
+ @i @j ( ).

I will derive R00 explicitly and leave the other components as an exercise.

Example.—The 00 component of the Ricci tensor is

⇢ ⇢ ↵ ⇢ ↵ ⇢
R00 = @⇢ 00 @0 0⇢ + 00 ↵⇢ 0⇢ 0↵ . (4.3.71)

1
Once in your life you should do this computation by hand. After that you can use Mathematica: an example
notebook can be downloaded here.
80 4. Cosmological Perturbation Theory

The terms with ⇢ = 0 cancel in the sum over ⇢, so we only need to consider summing over ⇢ = i,
i i ↵ i ↵ i
R00 = @i 00 @0 0i + 00 ↵i 0i 0↵
i i 0 i j i 0 i j i
= @i 00 @0 0i + 00 0i + ji
00
| {z } | 0i{z 00} 0i 0j

O(2) O(2)
2 0 0 0 0 2 j i
=r 3@0 (H ) + 3(H + )(H ) (H ) i j
= 3H 0 + r2 + 3H( 0
+ 0
)+3 00
. (4.3.72)

This confirms the result (4.3.68).

Exercise.—Derive eqs. (4.3.69) and (4.3.70).

Ricci scalar.—It is now relatively straightforward to compute the Ricci scalar

R = g 00 R00 + 2 g 0i R0i +g ij Rij . (4.3.73)

| {z }
O(2)

It follows that

a2 R = (1 2 )R00 (1 + 2 ) ij Rij
⇥ ⇤
= (1 2 ) 3H 0 + r2 + 3H( 0 + 0
)+3 00
⇥ 0 00
⇤
3(1 + 2 ) H + 2H2 + r2 2(H 0 + 2H2 )( + ) H 0
5H 0

(1 + 2 )r2 ( ). (4.3.74)

Dropping non-linear terms, we find

a2 R = 6(H 0 + H2 ) + 2r2 4r2 + 12(H 0 + H2 ) +6 00

+ 6H( 0
+3 0
). (4.3.75)

Einstein equations.—We have done all the work to compute the Einstein equation

Gµ ⌫ = 8⇡GT µ ⌫ . (4.3.76)

We chose to work with one index raised since that simplifies the form of the stress tensor
[see §4.2]. We will first consider the time-time component. The relevant component of the
Einstein tensor is

0 00 1
G 0=g R00 g00 R
2
1
= a 2 (1 2 )R00 R, (4.3.77)
2
where we have used that g 0i vanishes in Newtonian gauge. Substituting (4.3.68) and (4.3.75),
and cleaning up the resulting mess, we find
0
G0 0 = 2r2 6H( + H ). (4.3.78)

The 00-Einstein equation therefore is

0
r2 3H( + H ) = 4⇡Ga2 ⇢ , (4.3.79)
81 4. Cosmological Perturbation Theory

P
where ⇢ ⌘ a ⇢a is the total density perturbation. Equation (4.3.79) is the relativistic gener-
alization of the Poisson equation. Inside the Hubble radius, i.e. for Fourier modes with k H,
we have |r | 2 ˙
3H| + H |, so that eq. (4.3.79) reduces to r 2 2
⇡ 4⇡Ga ⇢. This is the
Poisson equation in the Newtonian limit. The GR corrections in (4.3.79) will be important on
scales comparable to the Hubble radius, i.e. for k . H.
Next, we consider the spatial part of the Einstein equation. The relevant component of the
Einstein tensor is

1
Gi j = g ik Rkj gkj R
2
1 i
= a 2 (1 + 2 ) ik Rkj R. (4.3.80)
2 j
From eq. (4.3.70), we see that most terms in Rkj are proportional to kj . When contracted with
ik this leads to a myriad of terms proportional to i . We don’t want to deal with this mess.
j
Instead we focus on the tracefree part of Gi j . We can extract this piece by contracting Gi j with
the projection tensor P j i ⌘ @ j @i 13 ij r2 . Using (4.3.70), this gives
2 2 4
P j i Gi j =
a r ( ). (4.3.81)
3
This should be equated to the tracefree part of the stress tensor, which for scalar fluctuations is
2 4
P j iT ij = P j i ⇧i j = r ⇧. (4.3.82)
3
Setting (4.3.81) and (4.3.82) equal, we get

= 8⇡Ga2 ⇧ , (4.3.83)
P
where ⇧ ⌘ a ⇧a . Dark matter and baryons can be described as perfect fluids and therefore
don’t contribute to the anisotropic stress in (4.3.83). Photons only start to develop an anisotropic
stress component during the matter-dominated era when their energy density is subdominant.
The only relevant source in (4.3.83) are therefore free-streaming neutrinos. However, their e↵ect
is also relatively small, so to the level of accuracy that we aspire to in these lectures they can
be ignored. Equation (4.3.83) then implies ⇡ .

Exercise.—By considering the time-space Einstein equation, show that

0
+H = 4⇡Ga2 q . (4.3.84)

With this, the Poisson equation (4.3.79) can be written as

r2 = 4⇡Ga2 ⇢¯ , (4.3.85)

where is the comoving-gauge density contrast (4.2.38).

By considering the trace of the space-space Einstein equation, we can derive the following
evolution equation for the metric potential
00 0
+ 3H + (2H 0 + H2 ) = 4⇡Ga2 P , (4.3.86)
where P is the total pressure perturbation. If we hadn’t ignored the anisotropic stress,
then (4.3.86) would have an additional source term.
82 4. Cosmological Perturbation Theory

4.3.3 Metric Evolution

It is interesting to consider the evolution of during both the radiation-dominated era and the
matter-dominated era:

• During the matter era, we have

2H 0 + H2 = 8⇡Ga2 P̄ = 0 , (4.3.87)

so that eq. (4.3.86) reduces to

00 0
+ 3H = 0, (4.3.88)
where H = 2/⌘. This equation has two solutions: = const. and / ⌘ 5 / a 5/2 .
Notice that the growing mode solution is = const. independent of the wavelength of the
fluctuation.

• During the radiation era, we have

8⇡G 2
2H0 + H2 = a ⇢¯ = H2 , (4.3.89)
3
8⇡G 2 1 0
4⇡Ga2 P = a ⇢ = r2 3H( +H ) , (4.3.90)
3 3
so that eq. (4.3.86) becomes
001 0
= r2 ,
+ 4H (4.3.91)
3
where H = 1/⌘. Substituting the Fourier expansion of the field,
Z
d3 k ik·x
(⌘, x) ⌘ k (⌘) e , (4.3.92)
(2⇡)3/2
we get
00 4 0 1
k + k + k2 k = 0, (4.3.93)
⌘ 3
This equation has the following exact solution
j1 (y) n1 (y) 1
k (⌧ ) = Ak + Bk , y ⌘ p k⌘ , (4.3.94)
y y 3
where the subscript k indicates that the solution can have di↵erent amplitudes for each
value of k. The size of the initial fluctuations as a function of wavenumber will be a
prediction of inflation. The functions j1 (y) and n1 (y) in (4.3.94) are the spherical Bessel
and Neumann functions
sin y cos y y
j1 (y) = 2
= + O(y 3 ) , (4.3.95)
y y 3
cos y sin y 1
n1 (y) = 2
= + O(y 0 ) . (4.3.96)
y y y2
Since n1 (y) blows up for small y (early times), we reject that solution on the basis of initial
conditions, i.e. we set Bk ⌘ 0. We match the constant Ak to the primordial value of the
potential, k (0) = 23 Rk (0), see eq. (4.4.110) below. Using (4.3.95), we find
✓ ◆
sin y y cos y
k (⌘) = 2R k (0) . (4.3.97)
y3
83 4. Cosmological Perturbation Theory

Figure 4.1: Numerical solutions for the linear evolution of the gravitational potential.

Notice that (4.3.97) is valid on all scales. Outside the (sound) horizon, y = p13 k⌘ ⌧ 1,
the solution approaches = const., while on subhorizon scales, y 1, we get

cos p1 k⌘
3
k (⌘) ⇡ 6Rk (0) (subhorizon) . (4.3.98)
(k⌘)2

During the radiation era, subhorizon modes of therefore oscillate with frequency p1 k
3
and an amplitude that decays as ⌘ 2 / a 2 . Remember this.

Figure 4.1 shows the evolution of the gravitational potential for three representative wavelengths.
As predicted, the potential is constant when the modes are outside the horizon. Two of the modes
enter the horizon during the radiation era. While they are inside the horizon during the radiation
era their amplitudes decrease as a 2 . The resulting amplitudes in the matter era are therefore
strongly suppressed. During the matter era the potential is constant on all scales. The longest
wavelength mode in the figure enters the horizon during the matter era, so its amplitude is only
suppressed by the factor of 9/10 coming from the radiation-to-matter transition (see §4.4.2).

4.4 Initial Conditions

At sufficiently early times, all scales of interest to current observations were outside the Hubble
radius. On super-Hubble scales, the evolution of perturbations becomes very simple, especially
for adiabatic initial conditions.

4.4.1 Adiabatic Fluctuations

Adiabatic perturbations have the property that the local state of matter (determined, for exam-
ple, by the energy density ⇢ and the pressure P ) at some spacetime point (⌘, x) of the perturbed
universe is the same as in the background universe at some slightly di↵erent time ⌘ + ⌘(x). (No-
tice that the time shift varies with location x.) We can therefore think of adiabatic perturbations
as arising because some parts of the universe are “ahead” and others “behind” in the evolution.
If the universe is filled with multiple fluids, adiabatic perturbations correspond to perturbations
84 4. Cosmological Perturbation Theory

induced by a common, local shift in time of all background quantities; e.g. adiabatic density
perturbations are defined as

⇢a (⌘, x) ⌘ ⇢¯a (⌘ + ⌘(x)) ⇢¯a (⌘) = ⇢¯a0 ⌘(x) , (4.4.99)

where ⌘ is the same for all species a. This implies

⇢a ⇢b
⌘= 0
= 0 for all species a and b . (4.4.100)
⇢¯a ⇢¯b

Using ⇢¯a0 = 3H(1 + wa )¯

⇢a , we can write this as

a b
= for all species a and b . (4.4.101)
1 + wa 1 + wb

Thus, for adiabatic perturbations, all matter components (wm ⇡ 0) have the same fractional
perturbations, while all radiation perturbations (wr = 13 ) obey

4
r = m. (4.4.102)
3
P
It follows that, for adiabatic fluctuations, the total density perturbation, ⇢ ⌘ a⇢
¯a a , is
dominated by the species that carries the dominant energy density ⇢¯a , since all the a ’s are
comparable. At early times, the universe is radiation dominated, so it natural to set the initial
conditions for all super-Hubble Fourier modes then. Equation (4.3.86) implies that = const.
on super-Hubble scales, while equation (4.3.79) leads to

⇡ r = 2 = const. (4.4.103)

Equations (4.4.103) and (4.4.102) show that, for adiabatic initial conditions, all matter pertur-
bations are given in terms of the super-Hubble value of the potential . In these lectures, we
will be concerned with the evolution of photons, baryons and cold dark matter (CDM). Their
fractional density perturbations will satisfy the relation (4.4.102) on super-Hubble scales, but
will start to evolve in distinct ways inside of the horizon.

4.4.2 Curvature Perturbations

The gravitational potential is only constant on super-Hubble scales if the equation of state
of the background is constant. Whenever the equation of state evolves (e.g. in the transitions
from inflation to radiation domination or from radiation to matter domination), so will the
gravitational potential (cf. Fig. 4.1). It will be convenient to identify an alternative perturbation
variable that stays constant on large scales even in these more general situations. Such a variable
is the comoving curvature perturbation:
H
R= + q, (4.4.104)
⇢¯ + P̄

where T 0 j ⌘ @j q. Defining the initial conditions in terms of R will allow us to match the
predictions made by inflation to the fluctuations in the primordial plasma most easily.
85 4. Cosmological Perturbation Theory

Proof.—The following is a proof that R is conserved on super-Hubble scales, i.e. for modes with k ⌧
H. First, it is useful to note that on large scales k ⌧ H, the Einstein equations imply H q = 13 ⇢.
In this limit, the curvature perturbation can be written as

k⌧H ⇢
R ! . (4.4.105)
3(¯
⇢ + P̄ )

To find the evolution of R, we consider the continuity equation [eq. (4.3.49)]:

@ ⇢ @
+ 3H( ⇢ + P ) + @i q i = 3(¯
⇢ + P̄ ) , (4.4.106)
@⌘ @⌘

On large scales, @i q i is of order k 2 and can be dropped relative to terms of order H2 . Solving (4.4.105)
for , and substituting it into (4.4.106), we get
✓ ◆
@ ⇢ @R @ ⇢
+ 3H( ⇢ + P ) = 3(¯ ⇢ + P̄ ) + (¯
⇢ + P̄ ) . (4.4.107)
@⌘ @⌘ @⌘ ⇢¯ + P̄

The time derivatives of ⇢ cancel on both sides and we are left with

@R ⇢¯0 + P̄ 0
3(¯
⇢ + P̄ ) = 3H( ⇢ + P ) ⇢. (4.4.108)
@⌘ (¯⇢ + P̄ )

Using ⇢¯0 = 3H(¯

⇢ + P̄ ), this becomes
✓ ◆
R0 P̄ 0
(¯
⇢ + P̄ ) = P ⇢ . (4.4.109)
H ⇢¯0

For adiabatic perturbations, the right-hand side vanishes and we have established the conservation
k⌧H
of the comoving curvature perturbation, R 0 ! 0.

Exercise.—Substituting (4.3.84) into (4.4.104) gives

H( 0 + H ) k⌧H 5 + 3w
R= ! . (4.4.110)
4⇡Ga2 (¯
⇢ + P̄ ) 3 + 3w

Use this to show that the amplitude of super-Hubble modes of drops by a factor of 9/10 in the
radiation-to-matter transition.

Another advantage of the variable R is that it is gauge-invariant, i.e. its value does not depend
on the choice of coordinates. Although (4.4.105) was written in terms of variables defined in
⇥ ⇤
Newtonian gauge, it applies in an arbitrary gauge if we write gij = a2 (1 2 ) ij + @i @j E for
the metric and T 0 j = @j q for the momentum density. While we will mostly stick to Newtonian
gauge, the computation of inflationary fluctuations in Chapter 6 turns out to be simplest in
spatially flat gauge [ = E = 0]. The curvature perturbation R provides the “bridge” between
results obtained in Chapter 6 and the analysis in the rest of the notes.

4.4.3 Statistics
Quantum mechanics during inflation only predicts the statistics of the initial conditions, i.e. it
predicts the correlation between the CMB fluctuations in di↵erent directions, rather than the
specific value of the temperature fluctuation in a specific direction. For Gaussian initial condi-
86 4. Cosmological Perturbation Theory

tions, these correlations are completely specified by the two-point correlation function

hR(x)R(x0 )i ⌘ ⇠R (x, x0 ) = ⇠R (|x0 x|) , (4.4.111)

where the last equality holds as a consequence of statistical homogeneity and isotropy. The
Fourier transform of R then satisfies
2⇡ 2
hR(k)R⇤ (k0 )i = 2
R (k) D (k k0 ) , (4.4.112)
k3
where 2 (k) is the (dimensionless) power spectrum.
R

Exercise.—Show that Z
dk
⇠R (x, x0 ) = 2
R (k) sinc(k|x x0 |) . (4.4.113)
k

In Chapter 6, we will compute the form of 2 (k) predicted by inflation. We will find that the
R
spectrum takes a power law form
✓ ◆n s 1
2 k
R (k) = As . (4.4.114)
k⇤
9 1
This agrees with observational constraints if As = 2 ⇥ 10 and ns = 0.96, for k⇤ = 0.05 Mpc .

4.5 Summary
We have derived the linearised evolution equations for scalar perturbations in Newtonian gauge,
where the metric has the following form
⇥ ⇤
ds2 = a2 (⌘) (1 + 2 )d⌘ 2 (1 2 ) ij dxi dxj . (4.5.115)

In these lectures, we won’t encounter situations where anisotropic stress plays a significant role,
so we will always be able to set = .
• The Einstein equations then are
0
r2 3H( + H ) = 4⇡Ga2 ⇢ , (4.5.116)
0 2
+H = 4⇡Ga (¯
⇢ + P̄ )v , (4.5.117)
00 0 0 2 2
+ 3H + (2H + H ) = 4⇡Ga P. (4.5.118)

The source terms on the right-hand side should be interpreted as the sum over all relevant
matter components (e.g. photons, dark matter, baryons, etc.). The Poisson equation takes
a particularly simple form if we introduce the comoving-gauge density contrast

r2 = 4⇡Ga2 ⇢¯ . (4.5.119)

• From the conservation of the stress-tensor, we derived the relativistic generalisations of

the continuity equation and the Euler equation
✓ ◆ ✓ ◆
0 P P̄ P̄
+ 3H = 1+ r·v 3 0 , (4.5.120)
⇢ ⇢¯ ⇢¯
✓ ◆
0 1 P̄ 0 r P
v + 3H 0
v = r . (4.5.121)
3 ⇢¯ ⇢¯ + P̄
87 4. Cosmological Perturbation Theory

These equations apply for the total matter and velocity, and also separately for any non-
interacting components so that the individual stress-energy tensors are separately con-
served.

• A very important quantity is the comoving curvature perturbation

H( 0 + H )
R= . (4.5.122)
4⇡Ga2 (¯
⇢ + P̄ )

We have shown that R doesn’t evolve on super-Hubble scales, k ⌧ H, unless non-adiabatic

pressure is significant. This fact is crucial for relating late-time observables, such as the
distributions of galaxies (Chapter 5), to the initial conditions from inflation (Chapter 6).