Cosmology

Faraz Ahmed
May 2023

Table of Contents
1 Introduction 2

2 Cosmological Principle 2

3 Scale factor 2
3.1 Revisiting redshift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
3.2 Velocities in an expanding universe . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

4 Densities in an expanding universe 4

4.1 Matter density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.2 Radiation density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.3 Spooky densities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

5 Distances in an expanding universe 5

6 Temperature in an expanding universe 5

7 Friedmann’s Equation 6
7.1 Shape of the universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
7.2 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

8 Model Universes 9
8.1 Radiation-dominated universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
8.2 Matter-dominated universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
8.3 Dark-energy-dominated universe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

9 Age of the universe 10

10 Conclusion 11

1
USAAAO Guide Faraz Ahmed

1 Introduction
Building on observations, cosmology helps us to understand and predict everything about the
universe, including its evolution, age, shape, and more!

2 Cosmological Principle
The cosmological principle states that at large distances, our universe is homogeneous and isotropic.
What it means is that, the universe appears to be the same from every point. Crudely, a homoge-
nous universe is one where moving to a different point wont affect your view and isotropy means
that seeing in a a different direction wont affect your view. To help you understand what it means
for a surface to be perfectly homogeneous and isotropic (aka maximally symmetric), think of an
infinite flat plane. This is an example of a flat maximally symmetric surface. Can you think of a
curved maximally symmetric surface?

3 Scale factor
Consider lattice points on a screen. You observe these points to be separated with each other by
some specified distance χ. Now, if you zoom out of the screen, the distance between each point
is reduced. If you zoom in, you’ll observe that the distance between each point has increased.
In both of these instances, the distance between points is scaled by some ratio (also called scale
factor). Our universe expands in a similar way. Out of convention, we say that the present-day
physical distance is equal to the coordinate distance between the points, or that the current value
of scale factor is 1. As we expand or contract, the physical distance is related to the comoving
distance as
R = a(t)χ
Where a(t) is the scale factor for that instant, χ is the comoving distance, and R is the physical
distance.

Figure 1: Yo, Mr. White! The universe is expanding! (S is the scale factor)

2
USAAAO Guide Faraz Ahmed

3.1 Revisiting redshift

Consider a light source emitting photons of wavelength λ. As the photons travel towards us, the
universe also expands. This expansion also results in the stretching of the wavelength of light. As
a result, when light reaches us, it has the wavelength,

λ = λo

How would you relate it with the emmitted wavelength? Remember that the ratio of wave-
lengths and scale factor remains constant because that’s the “comoving wavelength”, so we have,
λe
λo =
a(t)
and finally, we end up with the familiar looking expression,
λo 1
= =1+z
λe a(t)
where z is the redshift you encountered in the previous chapter. This is powerful because we
haven’t considered what happens between these two instances, the scale factor has allowed us to
elegantly relate the two wavelengths without carrying out any long calculations such as integrating
over the path.

3.2 Velocities in an expanding universe

We would like to calculate the velocity of an object resulting from the expansion of the universe.
From the previous part,
R = aχ
Taking the time-derivative,
dχ
Ṙ = ȧχ + a
dt
Due to how it’s defined, χ is a constant. So we get,
ȧ
Ṙ = · a(t)χ
a
Where I’ve multiplied and divided by a(t) on the RHS to get a nicer looking expression.
ȧ
Ṙ = ·R
a
Does it look familiar? Of course it does! This is the general form of Hubble’s law (v = H0 d), which
is valid for present-day calculations. We define,
ȧ
=H
a
where H is the Hubble parameter for the universe1 .
1
Current value, H0 = 73 ± 1 km/s/Mpc

3
USAAAO Guide Faraz Ahmed

4 Densities in an expanding universe

We’ll be mostly dealing with matter density, radiation density, and dark matter and energy den-
sities (boo!). There are lots more but our universe is/was mostly dominated by these.
We represent these quantities as: ρM , matter density, ρrad , radiation density, ρΛM , dark matter
density, and ρΛ , dark energy density.

4.1 Matter density

As its name suggests, matter density is simply the density of matter in the universe.
So, intuitively,
1 1
ρM ∝ 3 ∝ 3
R a
So, in terms of current matter density, at a particular moment in time, we can write
ρ0,M
ρt,M = = ρ0,M · (1 + z)3
a(t)3
Or to put it in words, the matter density falls off by the third power of the scale factor.

4.2 Radiation density

Radiation density is the energy density of photons in the universe. Which is given by,
nhc
ρrad =
λ
Where λ is the wavelength (which we previously showed scales with a), h is Planck’s constant, c is
the speed of light, and n is the number density of photons in the universe, which as you might’ve
guessed, falls off just like the matter density. So we can write,
1
ρrad ∝ 4
a
And similarly,
ρ0,rad
ρt,rad = = ρ0,rad · (1 + z)4
a(t)4

4.3 Spooky densities

Crudely, dark energy is the energy of empty space, and dark matter is well, matter we can’t see
and therefore can only detect gravitationally and the spooky part is that it doesn’t interact with
radiation! Explaining dark matter and energy in detail is beyond the scope of this guide so you’re
expected to just take these results as is (sorry!). You can look into WIMPs and MACHOs for more
info.
Let’s talk about their scaling relations,
ρΛ = constant
Since dark matter is just a form of matter,
1
ρΛM ∝
a3
ρ0,ΛM
ρt,ΛM = = ρ0,ΛM · (1 + z)3
a3

4
USAAAO Guide Faraz Ahmed

5 Distances in an expanding universe

Because we can measure them, we define two distances in the universe: Dang , known as the angular
diameter, and DL , known as the luminosity distance.
d
Dang = = R · a(t)
θ
DL is defined as, r
L R
DL = =
4πF a(t)
Hence, these distances are usually related as2 ,
DL
Dang · (1 + z) = =R
(1 + z)

Figure 2: Dang! (Source: CERN)

6 Temperature in an expanding universe

This relation is something that you’d want to derive on your own but I’ve given the proof here for
completeness.
We know,
1
T 4 ∝ ρrad ∝ 4
a
Hence,
1
T ∝
a
And we can then write,
T0
Tt = = T0 · (1 + z)
a(t)
What it also means is that our universe cools down as it expands. How cool!
2
No proof. Sorry, not sorry. This is cosmology for high-schoolers.

5
USAAAO Guide Faraz Ahmed

Example 6.1: (IOAA 2008)

The average temperature of the Cosmic Microwave Background is currently T = 2.73 K.
It yields the origin of CMB to be at redshift zCMB = 1100. The current densities of the
Dark Energy, Dark Matter, and Normal Matter components of the Universe as a whole are,
ρDE = 7.1·10−30 g/cm3 , ρDM = 2.4·10−30 g/cm3 , and ρN M = 0.5·10−30 g/cm3 , respectively.
What is the ratio between the density of Dark Matter to the density of Dark Energy at the
time CMB was emitted, if we assume that the dark energy is the energy of empty space?

Solution: We know that dark energy density is constant and dark matter density is given by the
relation,
ρt,ΛM = ρ0,ΛM · (1 + z)3
So the ratio in question is just,
   
ρΛM ρΛM
= · (1 + zCMB )3
ρΛ CMB ρΛ 0

Which on calculating gives us,  

ρΛM
≈ 4.5 · 108 □
ρΛ CMB

7 Friedmann’s Equation
I could write the equation directly or give you a bogus Newtonian derivation, you know what,
I’ll give you a bogus derivation at least you’ll remember how to derive in the likely case that you
forget the formula in-contest. Alright, roll your sleeves up and consider a particle at the edge of a
sphere of radius R. We can write the energy of this particle as,
1 GM m
E = mv 2 −
2 R
where M is the mass of our sphere and v is the velocity of our particle. Let’s rewrite this by
writing M = ρV and the expression for velocity we got from subsection 3.2,
 2
1 ȧ 4πGρmR2
E = m· R2 −
2 a 3
mR2
We will divide both sides by 2
 2
2E ȧ 8πG
= − ρ
mR2 a 3

6
USAAAO Guide Faraz Ahmed

It turns out3 that we can write the LHS as −k a2

which is also called the curvature term.
And finally, we arrive at Friedmann’s equation,
 2
ȧ 8πG k
= ρ− 2
a 3 a

We can also expand the density into its constituents and write the equation equivalently as,
 2
ȧ 8πG k
= (ρM + ρrad + ρΛ ) − 2
a 3 a

Note that,  
ȧ
H=
a
where H is the Hubble Parameter as discussed in subsection 3.2

We can also write the densities as a function of initial densities, which gives us,
8πG k
H2 = (ρM,0 (1 + z)3 + ρrad (1 + z)4 + ρΛ ) − 2
3 a

7.1 Shape of the universe

We said in the previous part that − ak2 is known as the curvature and rightfully so, because,

k<0 =⇒ open universe

k=0 =⇒ flat universe
k>0 =⇒ closed universe
As usual, no proofs.
For the special case of k = 0, we define the corresponding density as ρcrit :

3H 2
ρcrit =
8πG

7.2 Parameters
We define the density parameter Ω as,
ρ
Ω=
ρcrit
Let’s rewrite the Friedmann’s equation in terms of Ω
8πG ρ k
H2 = ρcrit · − 2
3 ρcrit a
k
H 2 = H 2Ω −
a2
3
from General Relativity

7
USAAAO Guide Faraz Ahmed

Because we like our equations nice and clean, we define a new dummy variable,
k
Ωk = −
H 2 a2
Let’s substitute things to get,
1 = Ω + Ωk
Expand stuff to get,
1 = ΩM + Ωrad + ΩΛ + Ωk
This is interesting to us because it tells us that any point in time, the sum of the density parameters
is just 1.
(ΩM + Ωrad + ΩΛ + Ωk )0 = (ΩM + Ωrad + ΩΛ + Ωk )t = 1
Or,
Ωk = 1 − Ω

Figure 3: Cosmological society and its future (Source: WMAP-NASA)

Let’s do an example problem to make sure you’re getting the hang of things.

Example 7.1: (Folklore)

Consider a flat universe with only matter and dark energy. For this universe, write ΩM,t as
a function of ΩM,0 and z

Solution: Because this is a flat universe, we have,

Ωk = 0

and so we can write,

ΩM,t + ΩΛ,t = ΩM,0 + ΩΛ,0 = 1

8
USAAAO Guide Faraz Ahmed

we can write ΩΛ,t as,

H02
ΩΛ,t = ΩΛ,0 ·
H2
and H 2 as,
8πG
H2 = (ρM,t + ρΛ,t )
3
now we write the densities in terms of initial densities,
8πG
H2 = (ρM,0 (1 + z)3 + ρΛ,0 )
3
get the Ω in your line,
H 2 = H02 (ΩM,0 (1 + z)3 + ΩΛ,0 )
we substitute this in to get,
1
ΩΛ,t = ΩΛ,0 ·
ΩM,0 (1 + z)3 + ΩΛ,0
and,
ΩΛ,t = 1 − ΩM,t ; ΩΛ,0 = 1 − ΩM,0
Finally,
ΩM,0 (1 + z)3
ΩM,t = □
ΩM,0 (1 + z)3 + 1 − ΩM,0

8 Model Universes
In what follows, the universe will be flat. To obtain relations for the scale factor with time, we
need to solve the Friedmann equation for some simple cases, that is, cases where a certain type of
density is dominant.
  r
da 1 8πG ρ0
· = · 3γ
dt a 3 a
Z
3γ
a 2 −1 da = Ct

8.1 Radiation-dominated universe

For radiation,
4
γ=
3
and solving the integral gives us,
1
a ∝ t2

9
USAAAO Guide Faraz Ahmed

8.2 Matter-dominated universe

For matter,
γ=1
and solving the integral gives us,
2
a ∝ t3
Since, radiation falls off faster than matter, it’s fair to assume that a once radiation-dominated
universe would turn into a matter-dominated one.

8.3 Dark-energy-dominated universe

For dark-energy,
γ=0
and solving the integral gives us,
a ∝ eCt
Which is exponential expansion!
Turns out that this model very accurately models our universe. Our universe is flat, was dominated
by radiation after the big bang, then became matter-dominated and is now currently dominated
by dark energy.

9 Age of the universe

We know ȧa for the current day, so we just need to substitute for a and ȧ, which will allow us to
get the age of the universe (started from the bottom now we here?)
I intend to leave the math as an exercise for you so I’ll just spam the results on your screen. NOW.
1
Dark-energy-dominated t0 = H0
; also known as Hubble time
1
Radiation-dominated t0 = 2H0
2
Matter-dominated t0 = 3H0

Example 9.1: (IOAA 2008)

Consider a type Ia supernova in a distant galaxy which has a luminosity of 5.8 · 109 L⊙ at
its maximum light. Suppose you observe this supernova using your telescope and find that
its brightness is 1.6 · 10−7 times the brightness of Vega. The redshift of its host galaxy is
known to be z = 0.05. Calculate the distance of this galaxy (in pc) and also the Hubble
time. Given m⊙ = −26.72

Solution: We know that Vega is our zero-point, so we’ll find the brightness of Vega in terms of
the Sun.  
F⊙
−26.72 − 0 = −2.5 log
Fvega
Solving, we get
Fvega ≈ 2.09 · 10−11 F⊙

10
USAAAO Guide Faraz Ahmed

Let’s substitute Vega’s brightness to find the brightness of the supernova in terms of sun’s bright-
ness.
FSN ≈ 3.34 · 10−18 F⊙
We know the formula for flux,
LSN L⊙
2
≈ 3.34 · 10−18
4πdSN 4πd2⊙
we can now solve for dSN .
dSN ≈ 202 Mpc
1
Now, we need to find H0
, which is not a hard task per se but it does require some unpleasant
calculations.
v = cz = H0 · d
d
t= ≈ 13.2 ·109 yrs □
cz

10 Conclusion
You’re now set to tackle most problems on Cosmology that appear in the AOs or learn more
Cosmology from Intro textbooks by authors like Barbara Ryden or Andrew Liddle.
Cosmology is my favourite part of Astronomy Olympiads and I hope I was able to present it in an
attractive and interesting manner.

11
 Cosmic Distance Ladder
Faraz Ahmed
May 2023

Table of Contents
1 Introduction 2

2 Parallax 2

3 HR-Diagrams 4

4 Standard Candles 4
4.1 Type 1a Supernovae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
4.2 Cepheid Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

5 Hubble’s Law 5

6 Conclusion 7

1
USAAAO Guide Faraz Ahmed

1 Introduction
In this handout, we will learn how to measure distances to any object in the universe. Different
methods have been developed that work best for different distance scales, so we often group all
these methods together into a handy toolkit known as the distance ladder.

Figure 1: Distance Ladder (Source: David Darling)

2 Parallax
Hold your index finger in front of your eyes at a reasonable
distance. Now, close your left eye and observe the finger and
its background through your right eye. Do the same, closing
your right eye and seeing with your left eye. What do you ob-
serve? The optical background behind your finger appears to
have shifted. This is because the position of the observer (your
eye) has changed. If you repeat this experiment, focusing on ob-
jects farther away from you, you would notice that this change
becomes more and more negligible as the distance from the ob-
ject being viewed becomes larger.

Now, let’s extend this to an observer on earth viewing a star

at two different times in the year that are six months apart.
This situation is exactly analogous to the finger example we
gave in the previous paragraph: here we have the diameter of
earth’s orbit as the distance between your eyes, and your finger
is the star. It’s important that our stellar background is much Figure 2: Parallax of a nearby
more distant to us than the star we’re observing (that way the star (Source: John McGraw)
background doesn’t also shift).

2
USAAAO Guide Faraz Ahmed

Let’s get into some terminologies now. The angle p in the diagram is termed as the parallax
angle, it’s usually measured in arcseconds and the distance to the star, d, is measured in units of
parsecs1 . From the diagram, we can write,

1 AU = d tan p

By definition,
1 AU
= 1 pc
tan(1” )
Since the angles are small,2
206265 AU = 1 pc
Let’s continue to write down the equation for d in terms of p, which is just,
1 206265
d= pc = AU
p p

Example 2:1 (USAAAO Round One 2023)

Friedrich Bessel was the first person to quantitatively measure the annual change in stellar
positions due to the motion of the Earth around the Sun. This change is known as stellar
parallax. Bessel determined the stellar parallax for 61 Cygni to be about 1/3 arc-seconds.
What is the distance to 61 Cygni from Earth?

(a) 1/3 light years

(b) 3 light years
(c) 1/3 parsec
(d) 3 parsecs
(e) 3 kiloparsecs

Solution: GG, this is what we just discussed.

1
d= pc = 3 pc
p

So the answer is just (d) 3 parsecs.

Unforunately, it’s pretty hard to make parallax measurements for most of our neighbourhood
because measuring small angles is hard. For example, ESA’s Gaia can measure parallaxes to the
order of 10-100 nano-arcseconds accurately (What’s that in parsecs assuming Gaia to be at L1 ?)
or just about 1% of our neighbourhood.
1
1 pc = 3.26 ly = 3.086 · 1016 m
2
206265 arc seconds = 1 radian

3
USAAAO Guide Faraz Ahmed

3 HR-Diagrams
Consider two clusters of stars that have around the same age. This would mean that the HR-
Diagrams of these clusters should be about identical (see Stellar Evolution handout for an expla-
nation of the HR diagram). If for one cluster, we use mV on y because we don’t know its MV , the
difference that arises would be the uniform vertical shift due the variation in distance.3 Which is
then calculated by using (see Stellar Flux handout for an explanation of the distance modulus),
 
d
mV − MV = −5 log
10

A better known version of this technique is called Main-sequence fitting, where we fit the Main-
sequence branch observed in the cluster to the actual known Main-sequence branch and get the
distance as we described previously. However, this only works if the cluster has a significant
diversity of stars.

Figure 3: Main-Sequence fitting (Source: GSU.edu)

4 Standard Candles
Standard Candles are classes of objects that have a calibrated absolute magnitude MV . That is,
we know their absolute magnitude by some physical relation. Since the apparent magnitude is
simply what we observe, then all we need to do is just find the distance modulus and ta-da!

4.1 Type 1a Supernovae

A Type Ia Supernova is the result of the explosion of a white dwarf after it exceeds the Chan-
drasekhar limit. Since all white dwarfs share the same limit to their mass, when they explode the
3
We’re drawing the observer’s version of an HR-Diagram so the y and x-axes represent MV and MB−V

4
USAAAO Guide Faraz Ahmed

conditions are always fairly similar. As a result, the peak absolute magnitude of every Type 1a
supernovae is the same:
MV = −19.3
And then we use distance-modulus and determine the distance to the supernova, yay.

4.2 Cepheid Variables

Cepheid variables display a unique property known as the period-luminosity relationship,
discovered by Henrietta Leavitt. By observing many cepheids and plotting their properties on
a diagram, astronomers found out that all cepheids exhibit a linear relationship between their
luminosity and their period. In other words, cepheids with longer periods have a higher absolute
magnitude. As a result, just by observing the periodic variation in brightness of a Cepheid, you
can figure out its absolute magnitude!

Figure 4: Period-Luminosity relation for several types of variable stars. Type I cepheids are often used.
(Source: Australia Telescope)
Note that the graph above has multiple lines. This is because cepheids come in two types: Type
I (Classical) and Type II (W Virginis). Recall from Stellar Evolution that Population I stars are
brighter than Population II stars because they are, on average, younger and bluer. This is an
important distinction that we will come back to in the next section.

5 Hubble’s Law
The universe is expanding, but how did we figure that out? Redshift. When an object has a
recessional velocity it will cause a redshift due to the Doppler effect. Mathematically, we can

5
USAAAO Guide Faraz Ahmed

express this shift as follows:

∆λ
=z
λe
where z is the redshift (It’s named that because as ob-
jects move away from us, the wavelength is stretched,
which means that it gets longer, or more redder for
the visible spectrum). To relate it with the velocity,
we will simply use the doppler-shift formula,
∆λ v
=
λ c
Note that this only holds true when v ≪ c. When v
is comparable to c, we use,
s
∆λ 1+β
= −1
λ 1−β
v
where β = c Figure 5: Redshifting of absorption lines
(Source: Wikipedia)
Edwin Hubble used the Period-Luminosity relations to estimate distances to Cepheids and noticed
something interesting. When he plotted the velocity-distance curve, he arrived at this:

Figure 6: Hubble’s graph (Source: NASA)

Objects farther from us, move away faster from us. Hubble explained this phenomenon with the
following idea: the universe is expanding. Since the universe expands, the objects within it will
appear to be pulled away from us in all directions since the physical distance to them is increasing.
In addition, if an object is farther from us, that means there is more space in-between to expand,
causing it to move away from us even faster and resulting in a higher redshift!

This relationship between distance and recessional velocity is:

v = H0 · d

6
USAAAO Guide Faraz Ahmed

Where H0 is Hubble’s constant. It is generally said to be about 70km/s/Mc (Let’s analyze

the units. km/s are the units for recessional velocity and Megaparsecs is a measure of distance.
Therefore, the units are simply velocty/distance), but more recent estimates show that its value
comes out to be around H0 = 73 ± 1 km/s/Mpc.

Example 5.1: (USAAAO Round One 2023)

How far from the Solar System would a galaxy with a redshift of 0.035 be?

(a) 150 Mpc

(b) 200 Mpc
(c) 250 Mpc
(d) 300 Mpc
(e) 350 Mpc

Solution: We know,
∆λ v
z= =
λ c
So,
v = cz ≈ 0.105 · 105 km/s
Now, we’ll use Hubble’s law,
v = H0 · d
Where, H0 = 72 km/s/Mpc . So,
v
d= ≈ 150M pc
H0
or option (a).

6 Conclusion
With the techniques we learned in this handout, we can now find the distance to any object in the
universe using nothing but our observations! Of course, that is not entirely true. Some stars are
too dim to be seen, especially for those far away subject to lots of interstellar extinction. However,
the cosmic distance ladder is still an incredibly useful tool that you will see frequently used.