Chapter-4 Cosmology (25 Lectures)

(a) Newtonian cosmology, Olber’s paradox, Hubble’s law and the expanding Universe, scale factor and co-
moving coordinate.
(b) Standard cosmology, the Friedmann equations from Newtonian cosmology, fluid equation , equation of state
for matter, dust etc. from basic thermodynamics, cosmological red shift, dark matter, dark energy and the
accelerating universe, tests and probes of Big Bang cosmology (the Cosmic Microwave Background, primordial
nucleosynthesis).

(A)
1. Introduction: Cosmology is the study of the universe as a whole. The common sense idea of the universe is
the totality of all that exists; which would in a sense encompass all that can influence or be influenced by us. In
other words, if there is something beyond our universe, we can, in principle, can get no information about it. A
scientist doesn’t deliberate on such things, they form part of metaphysics.
Cosmology was in the earlier times, the hobby of some philosophers, who gave a free play of their imagination
with very little factual basis. It was only after the improvement of the telescope towards the middle of the 18 th
century that astronomers began to think seriously about cosmology. They discovered three types of heavenly
objects:
a) The planets & comets which change their position in the background of the stars
b) The stars whose relative positions remain practically unaltered throughout the year, exceptions being some
nearby stars which have a measurable motion relative to the distant ones and
c) The nebulae which appear to be diffuse cloudlike objects, not forming sharp images even with the most
powerful telescopes then available.

2. The Olber’s paradox: On a dark & clear night, one can observe the Milky Way galaxy as a diffuse band
containing bright and dark patches stretching itself along the entire sky. We now know that it contains a large
number of stars along with cloud of gas and dust, weighing about 10 11M☼ and in the shape of a large lens or
cartwheel. The sun is situated at about 30,000 light years from the centre. A large number of galaxies were
discovered in due course and it was found that they form groups, clusters and superclusters. The universe is
supposed to have millions of galaxies and it is still a pertinent whether it is infinite and/or eternal. These are
closely related to the Olber’s Paradox, which asks, why is the night sky dark?
Let us assume that space is Euclidean, constituents are static (albeit incorrect), and the universe is infinite &
eternal. Then every line of sight must reach the photosphere of a star. If f * be the flux emerging at the star’s
photosphere, which has a radius ‘r*’ and situated at a distance ‘D*’ from us, then the flux reaching us is given by
fobs = f*(4πr*2/πD*2) = f* (dΩ/π), where dΩ is the solid angle subtended by the star on the sky. Since all the sight
lines eventually hit some solar surfaces, total flux from all 4π steradian is 4f*. This means that the whole sky
should radiate like a blackbody at temperature of a few thousand degrees, which is not observed actually.
One solution of the Olber’s paradox is that, the universe must be finite in extent or in age. Calculation for the
mean free path (l = 1/n*σ) until we reach the stellar surface yields 10 26 light years. Here n* is the density of stars
in the universe and σ is the stellar disc cross section. In fact that the night sky is dark means either the extent /
size of the universe is finite, that is, much less than 1026 light years or the age of the universe is finite, much less
than l/c or 1026 years. The distance beyond which light is yet to reach us is known as particle horizon.

3. Measurement of extra-galactic distances & Hubble’s law: In the cosmological scale, distances comprise of
rungs of a ladder, rising from nearby to distant objects. There are different methods to measure the separation
between objects; overlap in the ranges of applicability of two such methods help substantially in calibration. Yet
errors may propagate to a large extent. Fortunately, there are some direct methods which skip over few rungs of
the distance ladder and serve as a check to errors. The following table enumerates the methods which are used
generally:
DISTANCES METHODS
(a) Nearby stars (<1 Kpc) Trigonometric parallax, Main sequence fitting, Cepheid periodic-
luminosity relation.
(b) Large magellanic clouds, Measurement of orbital velocity from radial velocity curve of
local groups & Virgo Cluster spectroscopic binaries, Supernovae light echo,
galaxies (<20 Mpc)
(c) Distant galaxies (<100 Mpc) Tully Fisher relation, Faber Jackson relation, Measurement of surface
brightness fluctuations
(d) Beyond 100 Mpc Luminosity measurement, Sunyaev Zeldovich effect etc.

The Hubble’s law states that all galaxies recede from each other with a recession
velocity linearly proportional to the distance between them; v = H 0D, with the
Hubble parameter H 0 = 70±10 Km/s/Mpc, dimension T -1. Thus, observers in all
galaxies see exactly the same pattern of recession for other galaxies surrounding
them. v (km/s)
In the second Hubble diagram, consider ourselves to be situated at O, observing
 
Galaxy-1 at r1 and Galaxy-2 at r2 . The observer at Galaxy-2 would see Galaxy-1
     
receding from him at a velocity v12  v1  v2  H 0 r1  r2   H 0 r12 . We have D (Mpc)
used Euclidean space and non-relativistic velocities in our consideration, but G-1
results can be generalized to other regimes as well.
Hubble’s law thus leads to the fact that the universe is expanding and there is no G-2

centre of this expansion. Also in the past, the universe was dense and distance r1

between the galaxies was zero (Big Bang Singularity). Also, the universe has a r2 .
finite age. If it was not accelerating or decelerating, then the best estimated age O
of the Universe is close to t0 = 1/H0 = 14Gyr (Hubble time),

4. Estimation of the age of the Universe from natural/cosmic clocks: Since all elements heavier than oxygen
are produced in supernovae, the relative abundance ratio of isotopes reveal when they were produced. Modeling
of nuclear reactions show that initially, the two isotopes of uranium had ratio K0 = N0(U235) / N0(U238) = 1.16 –
1.34, but as of now, radioactive decay of U235 has resulted Kt = Nt(U235 ) / Nt(U238) = 0.00723. As Nt = N0 exp(-
t/τ), we have an estimation t = 6.2 Gyr. This is the lowest limit of the age of the Universe, and corrections for
metal enrichment of several stellar generations give the value of 10 Gyr. Same result is obtained from white
dwarf cooling data. Dating from globular clusters give a range of 10-15 Gyr for the age of the universe.

5. The Friedmann-Robertson-Walker(FRW) Metric: Homogeneity, which indicates translational symmetry,

doesn’t necessarily indicate isotropy, which on the other hand indicates rotational symmetry, and vice-versa. A
spherically symmetric universe with radially changing density will be isotropic but not homogeneous. A
rotating universe may be homogeneous, but not isotropic, as there is a preferred direction along the rotation
axis. The Cosmological Principle of Einstein states that the universe is both homogeneous and isotropic, which
is equivalent to the requirement that it appears the same from all directions. Again since the Hubble’s law v =
H0D leads to velocities even higher than ‘c’ at large distances, we need to invoke GR, relating the density of
mass energy (source of gravity) to curvature of space-time. Our aim is thus to find out a curvature or a metric
tensor which correctly specifies the line element of curved spacetime, with the only constraint that it must be
same everywhere for homogeneity.
There are three geometric possibilities, flat (plane), positively curved (as 2D surface of a 3D sphere) and
negatively curved (hyperboloidal surfaces where every point is a saddle point).
For a sphere, since x 2  y 2  z 2  R 2  constant, we have xdx  ydy  zdz  0 and line element
  xdx  ydy  
2

dl  dx  dy  dz .
2 2 2 2
Since z R x y ,
2 2 2 2
we can write dz  
2
 or
 z
 xdx  ydy 2 
dl 2  dx 2  dy 2   2 2 
. In spherical polar coordinates, the relation dl 2  dr 2  r 2d 2  r 2 sin 2 d 2
 R  x 2
 y 
needs to be modified as r = R, dr = 0 if we suppose to be on the surface of the sphere. Thus, finally we have
dl 2  R 2d 2  R 2 sin 2 d 2 representing the 2D homogeneous & isotropic universe that is unbounded yet finite,
with no preferred points or boundaries.
Since we live in three spatial dimensions, we extend this scheme to a 4D hypersphere in Euclidean space, to a
positively curved surface having four coordinates as x 2  y 2  z 2  w 2  R 2 , and the line element as
 xdx  ydy  zdz 2 
dl 2  dx 2  dy 2  dz 2   2 , dw  
xdx  ydy  zdz  & x 2  y 2  z 2  r2 giving
2 
finally
 R x y z 
2 2
w

r  2 dr  2 2 r 2 
dl 2  dr  2  r  2 d 2  r  2 sin 2 d 2   r  2
d 2
 r  2
sin 2
d 2
 dr  
 R 2  r  2 
1 
R 2  r 2  
 dr  2 R 2 r  2 d 2 r  2 sin 2 d 2 
 R 2     ,where in the denominator of the first term, we have introduced a
 1  kr  R
2 2
R2 R2 
curvature parameter k. In the present case k=1 (hypersphere). If k = 0, then dl 2  dr  2  r  2 d 2  r  2 sin 2 d 2
corresponds to a flat 4-D Euclidean space. For k = -1, we get the line element of a negatively curved space.
 dr 2 
If we define r  r  R, dr  dr  R, then the line element becomes dl 2  R 2   r 2 d 2  r 2 sin 2 d 2 . The
 1  kr
2

 dr 2 
interval between two events in spacetime would thus be ds 2  c 2 dt 2  R 2   r 2 d 2  r 2 sin 2 d 2  .
 1  kr
2

The coefficients of the above interval constitute the Friedmann-Robertson-Walker metric. The factor R that
multiplies the spatial part of it is known as the scale factor. If R(t) grows with time, every observer sees other
points in the Universe to recede radially, just as Hubble’s expansion. So, a galaxy at (r, θ, ϕ) remains at those
coordinates only as they are locked into themselves and expands according to R(t). Hence, (r, θ, ϕ) are called
co-moving coordinates.
r r
The instantaneous distance of a galaxy at commoving coordinate ‘r’ from us is l   dl  Rt 
dr
. This
1  kr 2 0 0

value, is (i) R sin-1r for k = 1, (ii) Rr for k = 0 and (iii) R sinh -1r for k = -1. Considering the first case,
r  sin l R becomes a maximum value of unity at l  R 2 . Galaxies beyond this point have consecutively
smaller value of r, reaching r = 0 at l  R, called antipode. Travelling continuously along one direction, we
pass through the antipode, and after travelling l  2R, we come back to the origin, facing the same direction.
The area of the sphere centered on us & passing through a galaxy at r’= rR is A  4r2  4R 2 sin 2 l R .
Beyond l  R 2 , the area decreases and at l  R, the sphere enclosing us and all previous ones has zero area.
This is analogous to drawing concentric circles on the 2D surface of a 3D sphere. Once
the circle passes through a point at a distance l  R, the circumference is zero, even
though it encloses all previous circles.
r r
Since l  Rt  , the velocity of a galaxy is v  l  R t 
dr dr l
 R , the
0 1  kr2 0 1  kr2 R
commoving co-ordinate ‘r’ being independent of time.
If we identify R R  H t , the Hubble parameter, then the Hubble’s law is recovered. Indeed, H must depend
on time, and roughly its reciprocal is the age of the Universe. With time, age increases and H decreases, or in
other words, if galaxies do not accelerate or decelerate, distances grow with time and H decreases, putting a
check over velocities to increase beyond c.

(B)
1. Newtonian Cosmology & derivation of Friedmann’s Equations: We have seen that the FRW metric can
describe three possible constant-curvature geometries of an isotropic & homogeneous universe; and even can
allow Hubble-like expansion by a scale factor R(t). To proceed and obtain the equations of motion that describe
the behavior of this scale factor, we need to specify the mass-energy distribution and relate it to the FRW
metric. This can even be done approximately based on local Newtonian arguments, without Einstein’s general
relativistic equations, which we see as below.
M We consider a spherical region of radius R, having mass M, constant density ρ and a
galaxy of mass m at the edge of the sphere. Conservation of energy yields
m mR 2 GMm 4R 3  mR 2 4GmR2
  E . Substituting M  , we get   E. Multiplying
2 R 3 2 3
2
 R  8G 2E
both sides by 2/mR , we get   
2  2
. Replacing 2E/m by –kc2, where k is
R 3 mR
the
2
 R  8G kc2
curvature parameter, we get what is known as the First Friedmann Equation:     2 . ………(I)
R 3 R
2 GMm 
From the energy equation, we can have R 2   E  . Differentiating both sides with respect to time, we
m R 

   2GM R  R   G  4R    R   4G . Since Newtonian gravity doesn’t account for
3
get 2R R
R2 R 2  3  R 3
gravitating effects of pressure, we need to add another term to the R.H.S of the equation to have a get a
  4G
complete physical picture and thus arrive at the Second Friedmann Equation: 
R
R 3c 2
 
c 2  3P . ……….(II)
From thermodynamic arguments, in an adiabatic environment, we have dU = -PdV. This is in conformity with
the cosmological principle, i.e., in a homogeneous & isotropic universe, there can be no net energy flow from
one region to another. Considering the internal energy of the system to be entirely due to the mass content, we
write U = ρc2 V. Taking derivative with respect to time,
d
dt
c 2V    P
dV
dt
 c 2V  c 2
dV
dt
 P
dV
dt
. In other
  V
words, c 2 
V
 
c 2  P . Since volume varies as 3 rd power of radius, dV/V=3dR/R and we finally arrive at

 3R
the Third Friedmann Equation: c 2 
R
 
c 2  P . ……….(III)
2
 R  8G kc2
The First Friedmann Equation:     2 is the gravitational form of conservation of mass-energy.
R 3 R
  4G
The Second Friedmann Equation: 
R
R 3c 2
 
c 2  3P is the equation of motion of a system under gravity.

 3R
The Third Friedmann Equation: c 2 
R
 
c 2  P is basically a restatement of energy conservation holding
separately for each co-spatial system, in thermal equilibrium with no energy exchange. For example, in a gas
composed of matter & radiation, it holds separately for density & pressure of each of the components.

2. History & Future of the Universe: Solving the Friedmann’s equations for R(t), one can have a description of
the history and fate of the universe , with of course, a specification for a proper equation of state P(ρ).
Intuitively, we have two different regimes or epochs to explore:
Matter dominated epoch: When pressure from all sources is much less than matter density, we have
from the 3rd F.E to be,    3R  R      3R R    R 3 .
Radiation dominated epoch: The pressure due to matter particles can be neglected but we have to take
into account the radiation pressure u/3=ρc2/3 of ultra-relativistic photons in the third F.E
 3R  2 c 2 
c 2   c   or     4R R    R 4 .
R  3 
(a) THE RECENT PAST: Let us trace the history of the scale factor R(t) with our knowledge of the
behavior of matter and radiation densities at different epochs. At early times, when R was not too small,
2
 R  8G
we can neglect the second term in the 1 st 
F.E , to get    . The R.H.S contains ρ, which varies
R 3
as R-3 in the matter dominated case, leading to the variation R ~ t 2/3. In radiation dominated case, ρ
varies as R-4 and hence we have R ~ t1/2. The expansion was slower in the later case as the gravitating
effect of radiation pressure also contributed.
(b) THE DISTANT PAST: At the earliest times, Lt Rt   0 and also Lt  t   4  . This is the Big Bang
1
t 0 t 0 R
Singularity, with which the common misconception is that, at t→0, all matter was concentrated at a
single point. This would violate the cosmological principle, as it would mean that the Big Bang occurred
at a particular point or location in space. The singularity was in the density, and it occurred everywhere
in the Universe at once. Even if it was open or infinite (indicating volume and mass to be infinite), no
matter how highly compressed mass we have in a given volume, there is infinitely more matter outside it
such that equally high density is achievable everywhere. From the Big Bang singularity, we can now
understand what is actually meant by the time coordinate. It is simply a universal or cosmic scale that
can be measured by all co-moving observers since the Big bang. As the universe is homogeneous &
isotropic, all co-moving clocks advance at the same rate. All such observers thus agree that at a time
t=t0, the local mean density measured by him/her reaches a particular value ρ 0.
(c) THE FUTURE: At a given point in future, the curvature term kc2/R2 cannot be neglected in the 1 st F.E.
We need to investigate separately the three cases that would arise with k=0, +1 & -1.
2
 R  8G 3H 2
For flat space (k=0), we would again have     H2    ; with our predefined value
R 3 8G
of the Hubble parameter. A flat universe thus indicates a critical density for every moment, including
now. With the current value of H 0, we have the critical density c,0  3H 0 8G  1.4  1011 M☼/Mpc3.
2

Since the typical density in L* type galaxies along with the dark halos is found to be about ten times less
than the critical density calculated above, we can define a parameter Ω = ρ/ρc. In this model, we are now
in matter dominated era where R varies as t2/3 and dR/dt varies as t-1/3, indicating that the scale factor
would continue to grow forever, stopping only at t=∞.
For a positively curved space (k=1), we consider a case where R continues to grow and thus ρ decreases
2 1/ 2
 R  8G c 2  3c 2 
until the L.H.S of the F.E becomes zero,   
1 st  2  0  R    . At this point, the
R 3 R  8G 
  4G
expansion halts. However, the rate of deceleration as obtained from the 2 nd F.E
R
R

3c 2

c 2  3P , 
doesn’t change sign. This indicates the onset of a collapse, in which the universe traces its past
expansion up to a “Big Crunch”. We also recall that for this recollapsing universe, the curvature is
positive; the volume is finite but unbounded.
For a negatively curved space (k=-1), the curvature term dominates over the density term in the 1 st F.E
2
 R  c2
in the course of time, and we may finally have    2  R  c, the expansion continuing forever at
R R
a constant “coasting” rate as compared to the flat case. An open universe thus expands eternally.
The transition from a radiation dominated universe (ρ ~ R -4, R ~ Scale factor ratio R/R0
k= -1
t ) to matter dominated one (ρ ~ R , R ~ t ) can be
1/2 -3 2/3
k =0
understood as follows -
The density of radiation and matter at present date, denoted by
the suffix 0 in the corresponding terms, is related to that at any
point of time by the two relations: rad  rad ,0 R0 R  &
4 k=1
R~t2/3

mc 2  m,0c 2 R0 R 3 . The two densities were equal when

R0 m,0c
2 R~t1/2
1
  , where the factor 1.7 accounts for the cosmic
R rad ,0 1.7 NOW
neutrino background, which, in spite of being non-zero mass,
Time (Gyr) in logarithmic scale
were relativistic at early times and behaved likely as radiation ,
(ρ ~ R -4). At present, we have ρrad,0c2 = σT04, T0=2.73K of CMBR and ρ m,0c2 = 0.3ρC. This clearly
indicates that we are now in the matter dominated era. Substituting the values, we have R 0/R = 3500 =
(t0/t)2/3 . The time of transition from matter dominated to radiation dominated era is
t  t0 35002 / 3  t0 2  105 . Thus the radiation dominated era lasted for a very small time, for only a
tiny fraction of the age of the Universe.
The schematic figure shows the two epochs, with the respective variation of R with time, and the future
of the universe as per the value of the curvature parameter. In the supercritical k=1 phase, R attains a
maximum value and the Universe then recollapses to a singularity, retracing its past evolution
 symmetrically in the reverse. The asymptotic appearance of this curve is due to the use of logarithmic
scale for visualizing the early-time behavior and various power law dependences.

4. Estimation of the age of the Universe: Since radiation dominance was only for a very small early
period, we can safely assume matter dominance throughout history (ρ ~ R -3, R ~ t2/3). Now, various
possibilities may arise as follows:
a) If ρ0 = ρc,0 (critical density now), and curvature parameter k = 0, then the first Friedmann
8G c , 0
2 2 3
 R  8G kc2  R  2  R0 
equation     2 reduces to    H t  
2
 R or H t   H 0   or
3 2

R 3 R R 3 R

R
3/ 2 t0 R0
R  1
 H 0  0    dt  3/ 2 
1
R1 / 2 dR  t0  2 H 0 3.
R R 0 H 0 R0 0
b) If we consider an empty universe (ρ=0), then with the curvature parameter k = -1, the second
t0 R0
dR
0 
  0, R  dR dt  HR  H R  constant, which gives dt 
Friedmann equation becomes R
0
H R
0 0 0
,

1
giving t0  H 0 .
Considering these two extreme cases, the density parameter Ω m = ρ/ρC lying between 1> Ω m > 0, we find that
the age of the universe lies between 2H 0-1/3 > t0 > H0 -1; substituting H0 -1=14 Giga years, we find that the age of
the Universe lies between 9 and 14 Gyrs. The transition from radiation dominance to matter dominance
occurred after 65,000 years of the Big Bang.

5. Dark Energy and the accelerating universe: The dynamics of the universe is solely controlled by
gravitational effects of matter & radiation. However, Einstein had initially used a term ‘Λg μν’ in his field
equation, which can be treated as a repulsive force that counteracts the conventional attractive gravitational
force in order to explain the existence of a static universe, an idea prevalent over those times.
After the Hubble expansion was discovered, Einstein discarded the term containing the cosmological constant
Λ, saying that it was the largest blunder in his life. But the Λ-like term is becoming necessarily important these
days in order to explain the dynamics of the universe and is often referred to as the dark/vacuum energy term.
2
 R  8G kc2 
With this included, the first and second Friedmann equations are modified as     2  and
R 3 R 3
  4G   3R
R
R

3c 2
c 2
 3 P
3
, while the third one 
 c 2

R
 
c 2  P remains unchanged.

(i) The modified second equation shows that a large enough positive value of Λ can cause R  to be
positive, i.e. to make the universe accelerate as opposed to the deceleration that always existed
without such term.
(ii) We can note that dimension of Λ is T -2.
(iii) From the first modified equation, we get that the cosmological constant effectively acts as an
additional energy density    c 2 8G. If Λ is a constant, the energy density is also constant; rather
than falling when R grows with time.
(iv) A counter-intuitive behavior results from the rather strange equation of state associated with the
cosmological constant, which relates a negative pressure P = - εΛ. When a volume element in the
universe grows due to expansion, work done by the negative pressure maintains the overall energy
density to be constant.
 (v) When R has grown enough, the cosmological constant dominates in the R.H.S of the 1 st F.E and we
 
may have, neglecting the other terms, H 2  R R   3  R   3 R. Integrating both sides of
2 1/ 2


1/ 2
  1 / 2 
    dt  Rt   exp   t   exp Ht  . Here the Hubble
dR
this equation, we would have 
R 3  3  
parameter becomes necessarily a constant. Thus, when the cosmological constant dominates, the
universe enters into an accelerating, exponentially expanding phase. If Λ is constant, this phase lasts
forever and the particle horizon tends to be a constant co-moving coordinate r h.
(vi) As opposed to the Universe without a cosmological constant, in which more & more volume
becomes visible as time progresses, in this case we have a fixed limit beyond which light will never
reach us. (Since at the time of emission, galaxies beyond that distance would be, in principle,
receding from us faster than the speed of light.)
(vii) Galaxies within the particle horizon will get more and more red-shifted with time, and therefore in
such a universe, observers would see more and more light sources blinking out, that is, being red-
shifted to infinity. Finally, such observers would be surrounded like a event horizon of black hole
type. This event horizon bounds the region of space in which causal communication is possible. T he
co-moving radial coordinate r H shrinks exponentially with time, and therefore all observers
eventually would lose contact with each other.
Observationally, the relevant model is now for a flat universe (k=0) which has a non-zero cosmological
2
 R  8G  H 2 8G 
constant. Thus, the first modified F.E,     on division by H0 2, gives 2
 2
 2
.
R 3 3 H0 3H 0 3H 0
Incorporating the present critical density that we require for closure, this gives
H2  
    m ,0    , 0 , writing finally in terms of density parameters. Considering the present value
H0
2
 C , 0 3H 0 2
of H = H0, we have Ω m,0 + ΩΛ,0=1 always for a flat, non zero-Λ universe. In a closed, positively curved
universe, Ω m,0 + ΩΛ,0 > 1 and for an open, negatively curved universe, Ω m,0 + ΩΛ,0 < 1. Recent measurements
have given the value very close to 1, assuming space to be near Euclidean. As k=0 marks the border between a
closed and an open universe, it may be difficult to determine whether space has finite or infinite volume.

Big Bang Cosmology or Modern Cosmology, as it is often referred to, was been subject to probes of
experimental predictions and has fortunately, passed the observational verification process. These tests also
provide information about the parameters that describe our present day Universe. Three such tests are described
in somewhat details below:

6. The Cosmological Redshift: Let us assume two wavefronts from a distant galaxy at commoving coordinate
re, emitted at te & te  te , arriving on earth at times t0 & t0  t0 . Since photons follow a null geodesic with
ds=0, in the Friedmann Robertson Walker metric, we have 0  c 2 dt 2  Rt  dr 2 1  kr2  , which implies,
2

cdt  Rt dr 1  kr2 . For the two wavefronts mentioned above, we have respectively, the following two
t0 r t 0  t 0 r
dt 1 e dr dt 1 e dr
integrals:  Rt  c 0 1  kr2
te
 and 
t e  t e
 
Rt  c 0 1  kr 2
. Since re is the commoving coordinate, the two

right hand side integrals are independent of time and hence equal. Also, since the time interval between two
successive light emissions as well as their receptions is very small compared to the dynamical time scales of the
universe (10-15 s for visible light against 10 17 s for Hubble’s time), we can safely assume R(t) to be a constant
between two successive emissions and two receptions.
t 0  t 0 t0 t e  t e t 0  t 0

Since the L.H.S of the second integral can be rewritten as                , we can finally have
t e  t e te te t0

Rt0 
t 0  t 0 t e  t e
t0 te t  
     
t0 te
. It follows that 
R(t0 ) Rte 
 0  0  e 
te e  0 Rte 
 1  z , where ‘z’ is known as the

cosmological redshift. It can now be understood that further in the past the received light was emitted, that is
more away the source is from us, the more it is red-shifted in proportion to the scale factors today and then. This
is the origin of Hubble’s law.
Cosmological redshift can be determined by obtaining the spectrum and measuring the wavelength of individual
features (either in emission or absorption) relative to their laboratory wavelengths. We also note that
cosmological redshift is different from Doppler, transverse Doppler & gravitational redshifts due to the fact that
it is a result of the expansion of the scale of the universe that takes place between the emission & the absorption
processes. In our expanding universe, R(t0) > R(te) always and ‘z’ is always a positive redshift. Indeed, it has
been experimentally found that all sources of light beyond 20 Mpc, are redshifted. Along with the main
cosmological shift, there are also present small red/blue shifts due to other effects.
The evolution of scale factor R(t) depend on the describing parameters of the universe, such as H 0, k, Ω m, ΩΛ
etc. So, though direct measurement is impossible, its evolution at different times in history can let us know
about the kind of universe we live in. From the cosmological redshift factor ‘z’, we can deduce the value of the
parameters that are related to the scale factor by measuring properties of distant objects. Two such properties
are flux from the object & its angular size.
Different cosmological models predict differently how these observables change with redshift. Measuring flux
from a standard candle to derive a distance and plotting distance against velocity is the whole idea behind the
Hubble diagram. In a curved and expanding space, distance can be defined in different ways, depending on the
property and history of evolution of the space. Nevertheless, observables can be calculated directly from the
FRW metric & by solving the Friedmann equations, and then compared. The luminosity distance obtained from
the Hubble’s diagram D L versus the redshift can be described best by a model in which (i) the universe is
currently in the accelerating stage, (ii) the transition from deceleration to acceleration occurred at a time
corresponding to z ~ 1. For a flat universe with k=0, with the given values of the density parameters, it is
suggestive of the fact that the universe is now dominated by dark energy of unknown sources, indicated by the
cosmological constant Λ.
A cosmological time dilation also accompanies the redshift. For example, a source at redshift ‘z’ emitting
photons at a certain wavelength and rate, will be seen by an observer as emitting wavelengths increased by a
factor (1+z) and photon arrival time lowered by a factor (1+z). Both these effects will reduce flux, in addition to
the reduction due to geometrical dilution (4π × distance2).

7. The Cosmic Microwave Background: As one goes back in time, the density of the universe is found to
increase monotonically with time. Models with large enough positive value of Λ indicate a minimum R > 0 in
the past and thus high density. At this point, the mean free path of photons was small, baryonic matter &
radiation was in thermodynamic equilibrium. Radiation field had a Planck spectrum and as it dominated, ρ~R -4 .
But the radiation density is also related to temperature as ρ~T 4, and we have T4~R-4 or temperature behaving
inversely as distance. When R was small, the early universe was intensely hot & dark. At such high
temperature, atoms were constantly being ionized and the universe was opaque mainly due to scattering of
electrons. As expansion occurred, the universe started to cool down. At T~3000K, photons of the radiation field
had enough energy to ionize H-atoms. Most of the protons & electrons could get recombined again. Once this
could happen, after 400,000 years of the Big Bang, the universe became transparent to radiation of most
frequencies, except the UV radiation at Ly-α and beyond. Looking at larger distances in any direction meant
looking backwards in time, and at some point, we would reach the surface of last scattering, beyond which
things become opaque again.
Photons emerging from the surface of last scattering undergo negligible additional scattering and absorption. As
the universe expands, number density decreases as R -3 , photon energy decreases as R -1 due to cosmological
redshift and thus photon energy density reduces as R -4. The Planck shape is maintained even though photons are
not in equilibrium with matter.
For a red-shifted photon, observed frequency     1  z , d   d 1  z  . From the Planck’s theory, we have
2h 3 d
the energy density of radiation in the range ν to ν+dν to be B  h / k B T
. Dividing by hν, the photon
2
c e 1
2 2 d
number density in the above frequency range is n  h / k B T
. As photon number is conserved, the
2
c e 1
n  2 2 d
density decreases with the universe’s expansion and is actually n   2 .
1  z  c 1  z  e
3 3 h / k B T
1
 2  d 
2
Substituting, we have finally n   2 h  / k BT  , with T   T 1  z . Spectrum thus keep the Planck’s form,
c e 1
but with reduced temperature, between the time of recombination and present and TCMBR  Trec 1  zrec , where
zrec is the redshift at which the recombination had occurred. A prediction of Big Bang cosmology is thus
therefore – “Space today should be filled with thermal photon distribution arriving from all directions of the
sky.”
As per Gamow’s prediction, that due to recombination at z~1000, we have a thermal spectrum corresponding to
a given temperature, the CMBR was discovered in 1965 with average temperature of 2.725K, with frequency in
the RJ side of the Planck’s spectrum. It thus solved the Olber’s paradox. Though every line of sight reaches the
ionized surface similar to the photosphere of a star, the expanding universe dilutes the radiation emitted by it
and shifts it to harmless microwave region.
The temperature of CMBR T = 2.725K is extremely uniform along the sky.
Only, a small dipole arises from Doppler effect due to the motion of local
group relative to the cosmological frame. Apart from this, there are small
temperature anisotropies, small angular sized regions having temperature
fluctuations δT ~ 29μK, such that δT/T ~ 10 -5. The extreme isotropy at z ~ 10 3
justifies the assumption of homogeneity & isotropy inherent to cosmological
principle. But the question is why & how. It raises the so called horizon problem.
At the time of recombination, the size of the horizon, i.e., region of space through which light can propagate
since the Big Bang corresponds to a physical region that subtends only 2 degrees on the sky today. Thus,
different regions separated by more than 2 degrees could not have been in causally connected by trec, and
thereby it is surprising that they would all have the same temperature within 10 -5. CMB photons from opposite
directions have never been in causal contact with each other till now, yet they have the same temperature.
The horizon problem can be explained in the following way: Very early during the evolution of the universe, in
the first small fraction of a second, there was an epoch of inflation. During this time, a vacuum energy density
with negative pressure caused an exponential expansion of the scale factor, much like the 2 nd acceleration epoch
of today. This inflationary expansion led causally connected regions to expand beyond the size of horizon at
that time. The theory also predicts that today, space is almost exactly flat, Ω m+ΩΛ is very close to 1. In fact, this
prediction is strongly confirmed by the observed anisotropies.

8. Primordial Nucleosynthesis: At earlier epochs, very high temperature should have rendered e-, e+, p, n in
thermal equilibrium. Since the difference of mass of neutron and proton is about 1.3MeV, at t<<1s, T>>10 10K,
we would have two reactions capable running in both directions; e   p     e  n and  e  p  n  e   n.

  mn  m p c 2 
3/ 2
N n  mn 
Now, from Saha equation, we have  exp   . At such high temperature, the ratio is
N p  m p   k BT 
close to unity. As the temperature gets to decrease, the ratio decreases and eventually the protons outnumber the
neutrons. This can’t continue indefinitely, and for k BT < 0.8MeV, the reaction time (9.46s) become larger than
the age of the universe at that epoch and no more neutrons can be created. Thus, thermal equilibrium can no
longer exist and this is called ‘neutron freeze-out’, and occurs at t=2s. So, the ratio N n/Np freezes at value of exp
(-1.3/0.8)=0.20. In the next minute, most of the neutrons become integrated to He-nuclei through the following
reactions: n  p  d   ; p  d 3He   ; d  d 3He  n;3He  n4He   ; d 3He4He  p. Since mean
lifetime of neutrons is about 15 minutes, they can undergo beta-decay into a proton & electron also, before
getting integrated to other elements.
Numerical calculations yield that on expansion of the universe, as the density & temperature decreases, all
parallel reactions ends up giving up the ratio between neutrons inside 4He and protons to be 1/7. Thus for every
two neutrons within 4 He, there are 14 protons, and thus 14-2=12 free protons. Finally, the ratio between He & H

is thus 1/12. The mass fraction of helium would be Y4 

 
4 N 4 He

4  1 12 1
 . Big bang
 
N H   4 N He 1  4  1 12 4
4

cosmology predicts that a quarter of mass in baryons was synthesized into Helium in first few minutes.
Measurements of helium abundances in stars, planetary nebulae etc. were consistent with this because the
amount could not have been produced in stars. Traces of deuterium, Helium-3, Lithium, and Beryllium etc.
were also produced in the first few minutes, amount depending on the baryon density at the time of
nucleosynthesis. As radiation energy density varies as R -4 and temperature in Planck spectrum varies as R -1 , the
photon number density varies as R-3. Also since baryon number density varies as R-3 , the baryon to photon ratio
(5×10-10) should not change with time. Observational data confirms this excellently.
The picture below summarizes the current view about the history of the Universe.