Page 1 of 40 Theoretical Round

Short Questions
T1. Sundial (10 points)
The following diagram represents a sundial, for which the triangle, called a gnomon, casts a
shadow onto the surrounding surface, which has markings and numbers representing important
information. It is known that this sundial is located either between the Tropic of Cancer and
the Arctic Circle or between the Tropic of Capricorn and the Antarctic Circle.

In the image above, the angle between the dashed and solid lines for any given time is always
equal to the longitude difference between the time shown by the sundial and the civil time (the
time shown on your watch). For instance, the dashed line corresponding to 7h and the solid line
corresponding to 7h form an angle equal to the longitude difference between the location of this
sundial and the central meridian of the time zone.
Throughout the year, the shadow of the tip of the gnomon is always between curves A and C.
Read the following statements and indicate whether they are true or false. For each item, write
a T on the answer sheet if you think the statement is true and an F if you think the statement
is false. There is no need to explain your answers.
(a) This sundial will only function properly if it is located in the southern hemisphere.
(b) Curve A represents the trajectory of the shadow of the tip of the gnomon throughout the
winter solstice (in the hemisphere the sundial is located in).
(c) Line B represents the trajectory of the tip of the gnomon’s shadow throughout the equinoxes.
(d) The solid radial lines provide the mean local solar time.
(e) The analemma shape around the dashed line corresponding to 12h shows the position of
the tip of the gnomon’s shadow during the true solar noon at the central meridian of the
time zone throughout the year.

Solution:

1
Page 2 of 40 Theoretical Round

(a) True. The polar rod of the sundial faces the hemisphere that contains the south
cardinal point as the fundamental pole, so the sundial was designed for a location in
the southern hemisphere of the Earth.
(b) False. This curve represents the trajectory of the tip of the rod’s shadow throughout
the day of the summer solstice in the hemisphere in which the sundial is located.
(c) True. This straight line, which is perpendicular to the meridian line, represents
the trajectory of the tip of the rod’s shadow throughout the day of either southern
autumnal or vernal equinox.
(d) False. This set of radial lines corresponds to the true local solar time. The 12h line
coincides with north-south line, which must be the case for the true solar noon.
(e) True. The dashed lines and the solid lines form an angle corresponding to the lon-
gitude difference between the sundial and the center of the time zone. Therefore, the
dashed lines correspond to the true solar time of the center of the time zone. An
analemma shaped figure around the line that corresponds to a true local solar time
indicates the mean solar time.

T2. Galaxy Cluster (10 points)

An astrophysical survey mapped all the galaxies in a small region of the sky, of angular diameter
∆θ “ 0.01 rad, where many galaxies seemed to be concentrated around the central area of the
image. When the positions and redshifts of all the galaxies in this cluster were measured, an
interesting distribution emerged, which is shown in the plot below.

Using these observations, estimate the total mass of the galaxy cluster and express your answer
in solar masses. Assume that this galaxya cluster is in dynamical equilibrium, with a root-
mean-square redshift dispersion σz “ xpz ´ 0.7q2 y “ 0.0005. Feel free to make reasonable
approximations when considering the average velocities, masses, and spatial distribution of the
galaxies.
Consider that the distance to z̄ “ 0.7 in the standard cosmological model is DA “ 1500 Mpc.
Ignore cosmological effects on the distance.

Solution:
If U and K are the total gravitational potential energy and the total kinetic energy of the

2
Page 3 of 40 Theoretical Round

cluster, respectively, the virial theorem says that:

U ` 2K “ 0 2.0

Let’s say that there are N galaxies of masses mi , where i “ 1, ..., N , moving with velocities
~ui “ ~vi ´ ~v0 with respect to the cluster, where ~v0 is the velocity of the cluster as a whole.
Therefore, in the cluster frame, K is given by

N
ÿ 1
K“ mi u2i
i“1
2

As an estimation, consider that all galaxies have mass equal to the average mass xmy “
MT {N :
ÿ
N
1 MT ÿ 2 1 u2i 1
K“ ui “ MT “ MT σv2 2.0
2 N i“1 2 N 2

Where σv “ xu2i y is the root-mean-square galaxy speed, also known as velocity dispersion.
a

Notice, however, that only the dispersion in redshift is given, which translates into a radial
velocity dispersion:

σvr “ c ¨ σz “ 2.998 ˆ 108 m{s ˆ 0.0005 “ 1.499 ˆ 105 m{s 1.0

There is no penalization for students who use that σv “ cσz . For completeness of the
solution, however, we relate σv and σvr by assuming three-dimensional isotropy for velocities,
which yields σv2 “ 3σv2r . K is then given by:

3
K“ MT σv2r
2

The gravitational potential energy, on the other hand, is given by:

ÿ Gmi mj
U “´
i‰j
rij

This can be estimated in different ways. Usually, the spatial distribution of the large number
of galaxies may be approximated as uniform, so that one estimates U as the binding energy
of a homogeneous spherical mass distribution, which is given by:

3 GMT2
U “´ 2.0
5 R

And, assuming the cluster is spherical, its radius is:

DA ¨ ∆θ 1500 ˆ 106 ˆ 206265 ˆ 1.496 ˆ 1011 ˆ 10´2

R“ “ “ 2.3143 ˆ 1023 m 1.0
2 2

Now, using the virial theorem, it is possible to calculate the total mass:

3
Page 4 of 40 Theoretical Round

3 1 3
6 ´ GMT2 ` 2 ¨ MT σv2r “ 0
5 R 2
5Rσvr2
ñ MT “
G
5 ˆ 2.314 ˆ 1023 ˆ p1.499 ˆ 105 q2
MT “ kg
6.67 ˆ 10´11
MT « 3.9 ˆ 1044 kg “ 2.0 ˆ 1014 Md 2.0

Appendix on calculation of U : Other valid methods for estimating U include explicitly

writing the sum as:
˜ ¸2
1 ÿ 1
U “´ G mi x y
2 i
rij

1 GMT2
U “´
2 R
where the factor of 1/2 comes from counting all the unique pairs, and xrij
´1
y was taken (as
an approximation) as R .´1

Or, alternatively, using dimensional analysis to argue that:

GMT2
U “´
R
.

T3. Asteroid (10 points)

A peculiar asteroid of mass, m, was spotted at a distance, d, from a star with
b mass, M . The
magnitude of the asteroid’s velocity at the time of the observation was v “ GM d , where G is
the universal gravitational constant. The distance d is much larger than the radius of the star.
For both of the following items, express your answers in terms of M , d, and physical or mathe-
matical constants.
(a) (8 points) If the asteroid is initially moving exactly towards the star, how long will it take
for it to collide with the star?

Solution: In this scenario, the asteroid would fall directly towards the star. However,
in order to simplify the calculations, it is possible to consider that the asteroid would
be in a degenerate elliptical orbit. In that case, the semi-minor axis would be infinites-
imally small and the asteroid would still practically be moving on a straight line. It is
also important to notice that the focii would virtually be at the periapsis and apapsis
points. The degenerate elliptical orbit is shown in the following figure: 2.0

4
Page 5 of 40 Theoretical Round

The semi-major axis of this elliptical orbit can be obtained through the following ex-
pression, where m is the mass of the asteroid:

mv 2 GM m GM m
E“ ´ “´
2 d 2a
GM m GM m GM m
´ “´
2d d 2a
1 1
“
2d 2a
6d“a 2.0

Now, it is possible to use Kepler’s third law to find an expression for the period of the
orbit: c
T2 4π 2 d
3
“ ÝÑ T “ 2πd 1.0
d GM GM
If the asteroid is initially moving towards the star, it sweeps area I to reach the star.
Using Kepler’s second law, it is possible to calculate how long it takes for that to
happen:
πab
∆t AI ´ ab
“ “ 4 2
2.0
T AT otal πab
ˆ ˙
1 1
∆t “ ´ T
4 2π
´π ¯ c d
∆t “ ´1 d 1.0
2 GM

Note that the radius of the star is negligible in these calculations since it is significantly
smaller than d.

(b) (2 points) If the asteroid is instead initially moving exactly away from the star, how long
will it now take for it to collide with the star?

Solution: If the asteroid is initially moving away from the star, it sweeps areas II and
III before reaching the star. It is possible to again use Kepler’s second law to find the
time interval:

5
Page 6 of 40 Theoretical Round

ab
∆t AII ` AIII ` πab
4 `
πab
“ “ 2 2
1.5
T A πab
ˆ T otal ˙
3 1
∆t “ ` T
4 2π
ˆ ˙ c
3π d
∆t “ `1 d 0.5
2 GM

T4. White Dwarf (10 points)

The structure of a white dwarf is sustained against gravitational collapse by the pressure of
degenerate electrons, a phenomenon explained by quantum physics and related to the Pauli Ex-
clusion Principle for electrons. The equation of state of a gas made of non-relativistic degenerate
electrons is the following:
ˆ ˙2{3 2
3 h
P “ n5{3 ,
8π 5me e
where ne is the number of electrons per unit volume, which can be expressed in terms of the mass
density ρ using the dimensionless factor µe , the number of nucleons (protons and neutrons) per
unit electron. Also consider that the central pressure can be described by this equation of state.
In the condition of hydrostatic equilibrium, the pressure and gravitational forces balance each
other at any distance r from the centre of the star. This condition can be expressed by:

dP GM prqρprq
“´ ,
dr r2
where M prq is the mass contained in the sphere of radius r, and ρprq is the mass density of the
star at a radius r.
Assume that mp “ mn , the density of a white dwarf is roughly uniform, and the following
approximation is valid at the surface of the star:
ˇ
dP ˇˇ Pc
«´ ,
dr ˇr“R R

where Pc is the pressure at the center of the star, and R the star radius.
(a) (6 points) The relationship between the mass M and the radius R of a white dwarf can be
written in the form:

R “ a ¨ Mb
Find the exponent b and determine the coefficient a in terms of physical constants and µe .

Solution: Simplifying the expression given in the problem statement:

dP GM prqρ
“´
dr r2

6
Page 7 of 40 Theoretical Round

At r “ R, it is possible use the approximation provided and the expression ρ “

3M {p4πR3 q:

Pc GM t3M {p4πR3 qu
´ “´
R R2
2
3GM
Pc “ 1.5
4πR4
Furthermore, electron density ne can be related to mass density as:

ρ “ µe mp ne 1.5

Using now the equation of state:

˙2{3 ˙5{3
h2
ˆ ˆ
3 ρ
Pc “
8π 5me µe m p
˙2{3 ˙5{3
h2
ˆ ˆ
3 3M
“
8π 4πR3 µe mp
5me
2{3 ˙5{3 5{3
3GM 2 h2
ˆ ˙ ˆ
3 3 M
“
4πR4 8π 5me 4πµe mp R5
ˆ ˙ ˆ ˙2{3 ˙5{3
h2
ˆ
4π 3 3
R“ M ´1{3
3 8π 5Gme 4πµe mp

Therefore:
˙2{3 ˙5{3
h2
ˆ ˙ˆ ˆ
4π 3 3
a“ 1.5
3 8π 5Gme 4πµe mp

1
b“´ 1.5
3

(b) (4 points) Using the relationship found in the previous part, estimate the radius of a white
dwarf made of fully ionised carbon (12
6 C) with a mass of M “ 1.0 Md .

Solution: There are 2 nucleons (1 proton and 1 neutron) per unit electron for carbon,
so that µe “ 2.
1.0

7
Page 8 of 40 Theoretical Round

Using the expression for the radius found on the previous item:
˙2{3 ˙5{3
h2
ˆ ˙ˆ ˆ
4π 3 1 3 1 1
R«
3 8π 5 4π Gme pµe mp q5{3 M 1{3
h2 1 1
« 1.866 ¨ 10´2
Gme pµe mp q5{3 M 1{3
` ˘2
1.866 ˆ 10´2 ˆ 6.626 ˆ 10´34
« 5{3 1{3
6.67 ˆ 10´11 ˆ 9.11 ˆ 10´31 ˆ p2 ˆ 1.67 ˆ 10´27 q ˆ p1.988 ˆ 1030 q

R « 1.44 ˆ 106 m 3.0

T5. CMB (10 points)

The Cosmic Microwave Background (CMB) is radiation emitted during the early Universe. It is
reasonably homogeneous and isotropic, and well described by a black-body radiation spectrum.
Its emission spectrum today has a peak at a temperature of approximately Ttoday „ 3 K, as
measured by the COBE satellite FIRAS instrument.
(a) (3 points) What is the redshift (z) at which the CMB spectrum has a peak at the infrared
wavelength of λIR „ 0.1 mm?

Solution:
In order to find the redshift, we need λtoday . Since Ttoday refers to the temperature at
the emission peak, we use Wien’s law:

λtoday ¨ Ttoday “ b
λtoday « 1 mm 1.0
So the redshift of the CMB, when it is emmited at the infrared spectrum (with the
values mentioned) is
λtoday ´ λinfrared
z“ 1.0
λinfrared

1 ´ 0.1
z« “9 1.0
0.1
.
Note: It is also possible to find Tinfrared using Wien’s law, and then use the definition
of redshift to find z, which leads to the same answer.

(b) (7 points) By assuming a spatially flat, matter-dominated Universe, what is the age of the
Universe corresponding to the redshift of the previous part?

Solution:
As the Universe is assumed as composed of just matter (i.e. Ωm “ 1), so the expression

8
Page 9 of 40 Theoretical Round

´ ¯
for the age of the Universe is given, in terms of the scale factor a “ 1
1`z , by

1 1
air “ “ “ 0.1 1.0
1`z 1`9
da
Now, “ Hptq ¨ aptq 1.0
dt

And, from the first Friedmann equation for a matter-dominated universe, we know
that:
H02
H2 “ 2.0
a3

Therefore:
da H0 H0
“ 1.5 aptq “ a
dt a ptq aptq
ż 0.1 ?
a
6t“ da
0 H 0
ż 0.1
1 ? 2 1.5 0.1
“ a da “ a |0 2.0
H0 0 3 H0
2 ˆ 0.11.5
“
3 ˆ 70
“ 3 ˆ 10´4 s Mpc{km
3 ˆ 10´4 ˆ 3.086 ˆ 1022
“
1000 ˆ 86400 ˆ 365.2422
t « 0.3 Gyr 1.0

Medium Questions
T6. Cluster Photography (20 points)
An astronomer takes pictures, in the V-band, of a faint celestial target, from a place with no light
pollution. The selected target is the globular cluster Palomar 4, which has an angular diameter
of θ “ 72.02 and a uniform surface brightness in the V-band of mV “ 20.6 mag{arcsec2 . The
observation equipment consists of one reflector telescope, with diameter D “ 305 mm and F-
ratio f {5, and a prime focus CCD with quantum efficiency η “ 80% and square pixels with size
` “ 3.80 µm.
Given data:
• V-band central wavelength: λV “ 550 nm
• V-band bandwidth: ∆λV “ 88.0 nm
• Photons flux for a 0-magnitude object in the V-band: 10 000 counts{nm{cm2 {sq
(a) (3 points) Calculate the plate scale (the angle of sky projected per unit length of the sensor)
of the observation equipment in arcmin{mm.

9
Page 10 of 40 Theoretical Round

Solution:
The plate scale is simply the angular size of the image per unit length as projected at
the focal plane. Hence, for small angles,

`i
tan θi “ « θi 1.0
f
where θi is the angular size of the image, `i is the unit length, and f is the focal length
of the telescope, which is given by pD ˆ F-ratioq. 0.5

1 mm
θi “ “ 6.56 ˆ 10´4 rad “ 2.2541
5 ˆ 305 mm
P S « 2.25 arcmin{mm 1.5

(b) (4 points) Estimate the number of pixels, np , covered by the cluster image on the CCD.

Solution:
Given the plate scale, the diameter of the image on the focal plane is

72.02
dGC “ “ 0.532 mm 1.0
2.251 ˆ 60
This implies that the area covered by the image is
π 2
SGC “ ¨d “ 0.223 mm2
4 GC
which can be divided by the area of a single pixel to estimate the pixel coverage
0.223
np “ « 15 400 pixels 2.0
p3.8 ˆ 10´3 q2

Since there are other methods to estimate such quantity, the accepted range is

15300 ď np ď 15600 1.0

(c) (13 points) With an exposure time of t = 15 s, the astronomer obtains a signal-to-noise
ratio of S{N “ 225. Compute the brightness of the sky at the observation site, knowing
that the CCD has a readout noise (standard deviation) of 5 counts{pixel and dark noise of
6 counts{pixel{minute. Give your answer in mag{arcsec2 . You may find useful: σRON
2
“
np ¨ 1 ¨ RON and σDN “ np ¨ DN ¨ t.
2 2

Solution: The signal-to-noise ratio is given by the expression

NGC
S{N “ b
2
σGC ` 2
σsky 2
` σRON 2
` σDN
NGC
“a 3.0
NGC ` Nsky ` np ¨ RON 2 ` np ¨ DN ¨ t

10
Page 11 of 40 Theoretical Round

where,
NGC Ñ Total Source Count
$
’
’
σGC “ NGC Ñ Poisson Noise
’
’ a
’
’
’
&Nsky Ñ Total Sky Count
’
’
’
σsky “ Nsky Ñ Sky Noise
a
’
’
’ b
’σRON “ np ¨ 1 ¨ RON 2 Ñ Readout Noise
’
’
’
’
’
’
σDN “ np ¨ DN ¨ t Ñ Dark Noise
% a

The noise values associated with the CCD operation can be easily calculated:
&σRON “ 15400 ˆ 52 “ 385000 counts
$ 2

1
2
%σDN “ 15400 ˆ 6 ˆ ˆ 15 “ 23100 counts 2.0
60
Now, the apparent visual magnitude of the globular cluster must be calculated,
ˆ ˙
ΩGC
VGC ´ mV “ ´2.5 log
1
with ΩGC being the solid angle subtended by the globular cluster,
ˆ ˙2 ˆ ˙2
θ 72.02
ΩGC “ π ¨ “πˆ “ 4071.5 arcsec2 1.0
2 2
Hence,
VGC “ 20.6 ´ 2.5 log p4071.5q “ 11.6 mag 1.0
and the associated flux of photons can now be evaluated using the 0-magnitude as a
reference:
ˆ ˙
φGC
VGC ´ V0 “ ´2.5 log
φ0
φGC “ φ0 ¨ 10´VGC {2.5
« 0.234 counts{nm{cm2 {s 1.0
With the calculated parameters, the total source count is:
π
NGC “ η ¨ φGC ¨ ¨ D2 ¨ ∆λ ¨ t
4
π
NGC “ 0.80 ˆ 0.234 ˆ ˆ p30.5q2 ˆ 88.0 ˆ 15
4
“ 180 540 counts 1.0

By referring to the signal-to-noise ratio expression and substituting the values,

180540
225 “ a
180540 ` Nsky ` 385000 ` 23100
ˆ ˙2
180540
Nsky “ ´ 385000 ´ 180540 ´ 23100
225
Nsky “ 55250 counts 1.0

11
Page 12 of 40 Theoretical Round

The inverse procedure to determine the Total Source Count is taken:

Nsky
φsky “ π
η¨ ¨ D2 ¨ ∆λ ¨ t
4
55250
“
0.80 ˆ π4 ˆ 30.52 ˆ 88.0 ˆ 15
φsky “ 7.16 ˆ 10´2 counts{nm{cm2 {s 1.0

which implies ˆ ˙
φsky
Vsky “ V0 ´ 2.5 log “ 12.9 mag 1.0
φ0
and finally, since Ωsky “ ΩGC
ˆ ˙
Ωsky
msky “ Vsky ` 2.5 log
1
“ 12.8 ` 2.5 log p4071.5q
msky “ 21.9 mag{arcsec2 1.0

Considering the accepted range of pixels from the previous item, the accepted answers
for the brightness of the sky fall within the interval

21.7 ď msky ď 22.1mag{arcsec2

T7. Castaway (20 points)

After surviving a shipwreck and reaching a small island in the southern hemisphere, a sailor had
to estimate the island’s latitude using the Sun.
However, due to poor eyesight, the sailor couldn’t see the night stars very well, so his best option
was to rely on the Sun. He had no information about the date, but he realized the days were
longer than the nights.
(a) (7 points) The sailor noticed that on his first day on the island, the angle between the
positions of sunrise and sunset on the horizon was 120˝ . With this piece of information,
determine the range of possible values for the latitude of the island. Neglect the daily
motion of the Sun across the ecliptic.

Solution:
The first step is to find an expression for the azimuth of an object during sunrise:

12
Page 13 of 40 Theoretical Round

Using the law of cosines:

cosp90˝ ` δq “ cosp90˝ ´ aq cosp90˝ ` φq ` sinp90˝ ´ aq sinp90˝ ` φq cosp180˝ ´ Aq

6 ´ sinpδq “ ´ sinpaq cospφq ´ cospaq cospφq cospAq

Since the altitude a “ 0˝ at the sunrise:

sinpδq
cospφq “ 2.0
cospAq

Note that due to symmetry over the local meridian, the azimuth of sunrise corresponds
to 180˝ minus half of the angle given in the question statement (the 180˝ term comes
from the fact that the observer is on the southern hemisphere).
The scenarios that lead to the minimum and maximum values of latitude correspond to
the extreme values for the declination of the Sun in the southern hemisphere summer
(´23.5˝ and 0˝ ).
If the declination of the Sun is ´23.5˝ , the latitude has to be

sinp´23.5˝ q
cospφq “ ` ˝˘
cos 180˝ ´ 1202
cospφq “ ´2 sinp´23.5˝ q
φ “ ´37.1˝ 2.0

Only the negative solution of this equation is relevant in this case since the observer is
in the southern hemisphere.
If the declination is the Sun is 0˝ , simply plugging in this value to the formula will
result in a latitude of ´90˝ . Since celestial objects do not rise or set at the poles, this
approach would not be conceptually correct. However, at a point infinitesimally close
to the pole, the Sun would still mathematically rise and set, and it would be possible to
achieve the difference in azimuth from the problem statement with a declination very

13
Page 14 of 40 Theoretical Round

close to 0˝ . Thus, although a latitude of ´90˝ is impossible, any latitude infinitesimally

close to it would still mathematically be possible.
Therefore, the range of possible latitudes for the island is the following:

´90˝ ă φ ď ´37.1˝ 3.0

(b) (13 points) The angle between the positions of sunrise and sunset kept increasing daily.
After 40 days, this angle was equal to 163˝ . Estimate the latitude of the island. You may
neglect the eccentricity of the Earth’s orbit.

Solution:
Since the latitude is constant, the following expression must be true:
sinpδ0 q sinpδ40 q
cospφq “ “ 2.0
cospA0 q cospA40 q

It is also possible to apply the law of sines to the following triangle to derive another
expression for the declination of the Sun on each day:

1.0

sinpθ0 q sinp´δ0 q
˝
“
sinp90 q sinpq
sinpδ0 q “ ´ sinpθ0 q sinpq

Analogously:

sinpθ40 q sinp´δ40 q
“
sinp90˝ q sinpq
sinpδ40 q “ ´ sinpθ40 q sinpq 1.0

Combining the three previous expressions:

14
Page 15 of 40 Theoretical Round

sinpδ40 q cospA0 q
sinpδ0 q “
cospA40 q
´ sinpθ40 q sinpq cospA0 q
´ sinpθ0 q sinpq “
cospA40 q
sinpθ40 q cospA0 q
sinpθ0 q “
cospA40 q
sinpθ0 q “ k sinpθ40 q 2.0

cospA0 q cosp120˝ q
Where k “ “ “ 3.38.
cospA40 q cosp98.5˝ q
In order to replace one variable and solve for θ0 , it is possible to consider that the
angular velocity of the Sun on the Ecliptic is approximately constant due to the Earth’s
very low eccentricity. In that case, the following expression must be true:
40
θ40 “ θ0 ´ ˆ 360˝
Tyear
θ40 “ θ0 ´ β 2.0

40
Where β “ ˆ 360˝ “ 39.4˝ . Combining the expressions and solving for θ0 :
Tyear

sinpθ0 q “ k sinpθ0 ´ βq
sinpθ0 q “ k psinpθ0 q cospβq ´ sinpβq cospθ0 qq
sinpθ0 qpk cospβq ´ 1q “ k sinpβq cospθ0 q
k sinpβq 3.38 sinp39.4˝ q
6 tanpθ0 q “ “
k cospβq ´ 1 3.38 cosp39.4˝ q ´ 1
θ0 “ 53.1˝ 3.0

Solving for δ0 :

sinpδ0 q “ ´ sinpθ0 q sinpq

δ0 “ ´18.6˝ 1.0

Solving for the latitude of the island:

sinpδ0 q sinp´18.6˝ q
cospφq “ “ ` ˝˘
cospA0 q cos 180˝ ´ 1202

φ “ ´50.4˝ 1.0

Note that only the negative solution to the equation above is relevant since it is known
that the island is in the southern hemisphere.

T8. Binary Hardening (25 points)

Consider a binary system of black holes, both of equal mass M , separated by a distance a,
and revolving around their common center of mass (CM) in circular orbits. This binary system
moves against, and interacts with, a very large, uniform field of stars (each of mass m ! M )

15
Page 16 of 40 Theoretical Round
with number density n. Consider a star that approaches the system from infinity with speed v
and impact parameter b, in the reference frame of the CM (as shown in the figure below). Its
closest approach distance to the CM is rp « 12 a. For tasks (a) and (c), you should make use of
the fact that v 2 ! GM
a .

Figure 1: Diagram of the system

(a) (5 points) Obtain an expression for b, in terms of M , a, v, and physical constants. In this
task, assume that the star interacts with the binary as if its total mass was fixed at the
CM.

Solution: Let vp be the closest approach velocity of the star. From conservation of
mechanical energy:
GMT m 1 1
´ ` mvp2 “ mv 2 2.0
rp 2 2
where MT “ 2M is the total mass of the binary. From conservation of angular mo-
mentum, we obtain vp :
b
vp “ v 2.0
rp
setting rp « a{2 and inserting vp into the first equation, we solve for b:
c
1 2GM
b“a `
4 av 2
?
2GM a
b« 1.0
v

Where we have used v 2 ! GM a in the last step.

Note: The student can also solve it by exploring the geometry of the star’s orbit,
which is hyperbolic, by writing rp “ aorbit pe ´ 1q and finding aorbit and e by writing
out equations for the angular momentum and mechanical energy.

16
Page 17 of 40 Theoretical Round
After a complex interaction with the binary, the star is slingshot from the system. The exact
calculation of its ejection speed is complex, but the result can be estimated by considering that
the star only interacts with one of the components when near the system. As such, consider, in
part (b), only the gravitational interaction between the star and one of the components in the
binary.
(b) (6 points) The star approaches the component with an initial speed negligible compared
to the component’s orbital speed, and both are moving directly towards each other. After
interacting with the system, when the star is again far away from the black hole, we find
that the direction of its velocity vector is reversed and the final speed is vf . Determine vf ,
in terms of M , a, v, and physical constants. Assume that linear momentum and mechanical
energy are conserved during this interaction and that it takes place in a timescale much
smaller than the binary’s period. Recall that m ! M .

Solution: The speed of the black hole in circular orbit can be found by equating the
gravitational force to the centripetal net force:

GM 2 MV 2
“ a
a2
c2
GM
6V “ 1.0
2a
For the rest of the problem, we propose two different solutions:
Solution I:
Additionally, Let vf be the ejection speed of the star, and V 1 the final speed of the
black hole. To find vf , we use conservation of linear momentum and conservation of
mechanical energy for the two bodies:

M V “ M V 1 ` mvf 1.5
1 1 1
M V 2 “ mvf2 ` M V 12 1.5
2 2 2
1 2 1 2 1 ´ m ¯2
M V “ mvf ` M V ´ vf
2 2 2 M
m 2
0 “ vf2 ´ 2V vf ` v
M f
Since m ! M , we neglect the last term, yielding: 1.0
c
2GM
vf “ 2V “ 1.0
a

Because we are considering two moments where the distance between the star and the
black hole is very large, we disregard potential energy terms.
Solution II:
In the reference frame of the component, the star approaches the component with a
velocity ´V ~. 1.5
And, since mechanical energy and linear momentum are conserved, it can be shown that
relative speed between the components is the same before and after the interaction–
that is, restitution coefficient equal to unity). Furthermore, since M " m, the velocity
of the component can be considered unchanged (notice this is not valid in Solution I,
as there we deal with second order terms arising from energy considerations) in the

17
Page 18 of 40 Theoretical Round

interaction. Therefore, after the interaction, the star is ejected back with a velocity
~ in the opposite direction.
~vf1 “ V 1.5
To return to the CM frame, we write ~vf “ ~vf1 ` V
~ , such that

vf “ 2V 2.0

As we have previously found.

For the following task, assume that all stars approaching the system from infinity with an impact
parameter 12 b0 ď b ď 32 b0 (where b0 is the impact parameter of a star whose speed at infinity is
v0 ) attain a closest-approach distance rp « 12 a to the CM. Also assume that all stars exit
the system with the speed found in (b).
(c) (14 points) Upon each encounter, part of the total energy of the binary is transferred to
the kinetic energy of the star. Assume that the binary orbit remains circular. Knowing
this, using your results from previous tasks, and taking into account only encounters with
the stars within the specified range of impact parameters, show that the reciprocal of the
binary’s separation increases at a constant rate:
d 1
ˆ ˙
Gρ
“H
dt a v0
Here, ρ “ nm is the mass density of the star field, and G the universal gravitational
constant. Find the dimensionless constant H, which refers to hardening.

Solution: First, we know the kinetic energy acquired by a star during an encounter:
1 1 GM m
∆K˚ “ mvf2 ´ mv 2 « 1.0
2 2 a
Where we used that v 2 is negligible with respect to vf2 , since vf2 “ 2GM
a , and v ! a .
2 GM

This is also the energy lost by the binary during an encounter.

In order to find the rate of energy extraction, we must first obtain the rate of encounters,
which can be estimated as follows:
imagine the binary travelling with velocity v in the reference frame of a far away star
of impact parameter b. During a small interval dt, all the stars included inside the
annular cylinder of length L “ vdt bounded by radii of b and b ` db will interact with
the star. Call this number of stars dN . It is given by:

dN “ n ¨ Vcylinder
“ n ¨ Aannulus ¨ L
“ n ¨ π ¨ rpb ` dbq2 ´ b2 s ¨ v ¨ dt
« 2πn b v db dt 4.0

Defining dφ ” dN
dt to be the rate of encounters with stars of impact parameter between
b and b ` db and using b from part (a),

dφ “ 2πn b v db
?
2GM a
“ 2πn ¨ ¨ v
¨ db
v

?
dφ “ 2πn 2GM a ¨ db 1.0

18
Page 19 of 40 Theoretical Round

To find the total rate of encounters (φ “ dφ), we sum over all impact parameters in
ÿ

3 1
the interval 2 b0 ď b ď 2 b0 . Notice db “ b0 ´ b0 “ b0 , so that:
ÿ
1 3
2 2
?
φ “ 2πn 2GM a ¨ db
ÿ
?
? 2GM a
“ 2πn 2GM a ¨
v0
4πnGM a
“ 3.0
v0
Since each encounter takes away ∆K˚ and there are φ encounters per unit time, the
rate of change of energy for the binary is:

dEbin 4πnG2 M 2 m
“ p´∆K˚ q ¨ φ “ ´ 2.0
dt v0
A change in the binary total energy can also be written as:

dEbin d GM 2 GM 2 d 1
ˆ ˙ ˆ ˙
“ ´ “´ 2.0
dt dt 2a 2 dt a

Therefore, equating the two:

d 1
ˆ ˙
8πGnm
“
dt a v0
d 1
ˆ ˙
Gρ
“ 8π 1.0
dt a v0
Thus, H “ 8π

T9. Physics of Accretion (35 points)

The accretion of matter onto compact objects, such as neutron stars and black holes, is one of
the most efficient ways to produce radiant energy in astrophysical systems. Consider an element
of gas of mass ∆m in a stationary and geometrically thin disc of matter with a maximum radius
of Rmax and minimum stable orbital radius of Rmin (with Rmin {Rmax ! 1), in rotation around
a compact object of mass M and radius R.
(a) (6 points) Assuming that an element of gas in the disc follows an approximately Keplerian
circular orbit, find the expression for the total mechanical energy per unit mass ∆m
∆E
released
by this gas from the moment it starts orbiting at a radius Rmax until the moment it reaches
an orbit of radius r ! Rmax . This process occurs very slowly, transforming kinetic energy
into internal energy of the gas disc through viscous dissipation.
Note: Ignore the gravitational interaction between particles within the accretion disc and
give your final answer in terms of G, M and r.

Solution: For a Keplerian orbit:

vk2 GM ∆m GM
∆m “ 2
ñ vk2 “ 2.0
r r r

19
Page 20 of 40 Theoretical Round

Total energy:

1 GM ∆m 1 GM ∆m
E “ Ekin ` Epot ñ Eprq “ ∆mvk2 ´ ñ Eprq “ ´ 2.0
2 r 2 r
ˆ ˙ ˆ ˙
1 1 1 GM ∆m r GM ∆m
EpRmax q ´ Eprq “ GM ∆m ´ “ 1´ « 1.0
2 r Rmax 2r Rmax 2r
Therefore:

∆E GM
« 1.0
∆m 2r

(b) (5 points) Considering that the disc receives mass at an average rate of M9 , and assuming
that all the mechanical energy lost is ultimately converted into radiation, find an expression
for the total luminosity of the disc.

Solution: Variation in energy:

∆ET ot “ EpRmax q ´ EpRmin q

Luminosity:

∆ET ot ∆ET ot
LT ot “ “ ˆ M9 3.0
∆t ∆m

GM M9
LT ot “ 2.0
2Rmin

Where the result from (a) was used, allied with Rmin ! Rmax .

(c) (8 points) Consider now the ring composed of all mass elements from radius between r`∆r
to r. In this scenario, find an expression of the luminosity generated by the disc over its
small length ∆r at this radius, that is, find the expression for ∆t∆r
∆E
.

Solution: Variation in the energy of the tiny mass:

ˆ ˙ ˆ ˙
1 1 1 1 ∆r
∆E “ Epr ` ∆rq ´ Eprq “ GM ∆m ´ « GM ∆m
2 r r ` ∆r 2 r2 p1 ` ∆r{rq

∆E 1 1
ñ « GM ∆r 2 5.0
∆m 2 r
Multiplying this expression by the rate of mass variation per time, it is possible to
obtain an expression for ∆t∆r
∆E
:

20
Page 21 of 40 Theoretical Round

∆E ∆E GM M9 ∆r
“ ˆ M9 “
∆t ∆m 2r2

∆E GM M9
ñ “ 3.0
∆t∆r 2r2

(d) (10 points) Assuming that the gravitational energy released in this ring is locally emitted
by the surface of the ring in the form of black-body radiation, find an expression for the
surface temperature T of the ring.

Solution: Surface area of the ring:

A “ 2rπpr ` ∆rq2 ´ πr2 s “ 2pπr2 ` 2πr∆r ` π∆r2 ´ πr2 q 4.0

ñ A « 4πr∆r
Using Stefan-Boltzmann’s law:

AσT 4 “ L 2.0

∆E ∆E
4πr∆rσT 4 “ ñ 4πrσT 4 “ 2.0
∆t ∆r∆t
GM M 9
4πrσT 4 “
2r2
˜ ¸1{4
GM M9
ñ T “ 2.0
8πσr3

(e) (3 points) Consider that the central object is a stellar black hole with a mass of 3Md and
a rate of accretion of M9 “ 10´9 Md {year. Consider also that Rmin “ 3Rsch , where Rsch
is the Schwarzschild radius of the black hole. Determine the luminosity of the disc and the
peak wavelength of emission of its innermost part. Ignore gravitational redshift effects and
assume that the emission from the innermost part of the ring dominates the total emission.

Solution: Internal radius of the ring:

2GM 6GM
Rsch “ ñ Rmin “ 3Rsch “
c2 c2

6GM 6 ˆ 6.674 ˆ 10´11 ˆ 3 ˆ 1.988 ˆ 1030

Rmin “ “ “ 2.7 ˆ 104 m
c2 p2.998 ˆ 108 q2

Total luminosity:

21
Page 22 of 40 Theoretical Round

GM M9 M9 c2
LT ot “ “
2Rmin 12

10´9 ˆ 1.988 ˆ 1030 ˆ p2.998 ˆ 108 q2

ñ LT ot “
12 ˆ 365.2422 ˆ 24 ˆ 60 ˆ 60

ñ LT ot “ 5 ˆ 1029 W 1.5

Surface temperature:

˜ ¸1{4 ˙1{4
GM M9 6.674 ˆ 10´11 ˆ 3 ˆ 10´9 ˆ p1.988 ˆ 1030 q2
ˆ
T “ 3 “
8πσRmin 8π ˆ 5.670 ˆ 10´8 ˆ p2.7 ˆ 104 q3 ˆ 365.2422 ˆ 24 ˆ 60 ˆ 60

T “ 5.5 ˆ 106 K

Using Wien’s law:

b 2.898 ˆ 10´3
λ“ “
T 5.5 ˆ 106

ñ λ “ 5 ˆ 10´10 m 1.5

(f) (3 points) Now, considering another accretion system with M9 “ 1 Md {year and a peak
emission wavelength of λ “ 6 ˆ 10´8 m, estimate the mass of this black hole.

Solution: Using Wien’s law:

b b 2.898 ˆ 10´3
λ“ ñT “ “
T λ 6 ˆ 10´8

ñ T “ 4.8 ˆ 104 K 1.0

Considering this temperature to be that of the innermost of the accretion disk, and
using Rmin “ 3Rsch :

˜ ¸1{4 ˜ ¸ ˜ ¸
GM M9 4 GM M9 GM M9
T “ 3 ñT “ “
8πσRmin 8πσp3Rsch q3 8πσp6GM {c2 q3

22
Page 23 of 40 Theoretical Round

˜ ¸1{2
M9 c3
ñM “
8πσ63 GT 2

*1{2
1.988 ˆ 1030 {p3.15 ˆ 107 q p2.998 ˆ 108 q3
"
1 1
ñM “
p8πq1{2 63{2 5.670 ˆ 10´8 6.674 ˆ 10´11 ˆ p4.8 ˆ 104 q2

M « 2.5 ˆ 1039 kg « 1.3 ˆ 109 Md 2.0

Long Questions
T10. Greatest Eclipse (75 points)
The greatest eclipse is defined as the instant when the axis of the Moon’s shadow cone gets
closest to the centre of the Earth in a solar eclipse. This problem explores the geometry of this
phenomenon, using the solar eclipse of 29th May 1919 as an example, as it has great historical
significance for being the first time astronomers were able to observationally verify general
relativity. One of the scientific expeditions to observe this eclipse took place in the Brazilian
city of Sobral.
The two following tables show the Cartesian and spherical coordinates of the Sun and the Moon
at the time of the greatest eclipse. The system used for these coordinates is right-handed and
has the origin at the centre of the Earth, the positive x-axis pointing towards the Greenwich
meridian, and the positive z-axis pointing towards the North Pole. For the rest of this problem,
this will be referred to as system I.
Spherical coordinates:

Centre of the Sun Centre of the Moon

Radial Distance (r) 1.516 ˆ 1011 m 3.589 ˆ 108 m
Polar Angle (θ) 68˝ 291 44.12 68˝ 471 41.62
Azimuthal Angle (ϕ) ´1h 11m 28.2s ´1h 11m 22.9s

Cartesian coordinates:

Centre of the Sun Centre of the Moon

x 1.342 ˆ 1011 m 3.185 ˆ 108 m
y ´4.327 ˆ 1010 m ´1.025 ˆ 108 m
z 5.557 ˆ 1010 m 1.298 ˆ 108 m

For this problem, assume that the Earth is a perfect sphere.

Note: The spherical coordinates of a point P are defined as follows:

• Radial distance (r): distance between the origin (O) and P (range: r ě 0).

23
Page 24 of 40 Theoretical Round
• Polar angle (θ): angle between the positive z-axis and the line segment OP (range: 0˝ ď
θ ď 180˝ ).
• Azimuthal angle (ϕ): angle between the positive x-axis and the projection of the line
segment OP onto the xy-plane (range: ´12h ď ϕ ă 12h ).

Part I: Geographic Coordinates (25 points)

(a) (3 points) Determine the declination of the Sun and the Moon during the greatest eclipse
for a geocentric observer.

Solution: The declination is simply the complement of the polar angle:

δ “ 90˝ ´ θ 1.0
˝ 1
6 δd “ 21 30 15.9 2
1.0
δM oon “ 21˝ 121 18.42 1.0

(b) (3 points) Determine the right ascension of the Sun and the Moon at the time of the
greatest eclipse for a geocentric observer. The local sidereal time at Greenwich at that
same moment was 5h 32m 35.5s .

Solution: The local sidereal time at Greenwich corresponds by definition to the merid-
ian of right ascension that is right above Greenwich. Since Greenwich corresponds to
an azimuthal angle of 0h , the right ascension at any given point corresponds simply to

24
Page 25 of 40 Theoretical Round

the sum of the local sidereal time at Greenwich and the azimuthal angle:

α “ ϕ ` LSTGreenwich 1.0
6 αd “ 4 21 7.3h m s
1.0
αM oon “ 4h 21m 12.6s 1.0

(c) (4 points) Find a unit vector that indicates the direction of the axis of the Moon’s shadow
cone. This vector should point from the Moon to the vicinity of the centre of the Earth.

Solution: By subtracting the Cartesian coordinates of the Sun from the Cartesian
coordinates of the Moon, it is possible to obtain a vector for the direction of the axis
of the Moon’s shadow cone: 1.0

~u “ă ´1.339 ˆ 1011 , 4.317 ˆ 1010 , ´5.544 ˆ 1010 ą m 1.0

|~u| “ 1.512 ˆ 10
11
m 1.0
6 û “ă ´0.8855, 0.2855, ´0.3666 ą 1.0

(d) (15 points) Determine the latitude and the longitude of the point where the axis of the
Moon’s shadow cone crosses the surface of the Earth during the greatest eclipse.

Solution: If one draws a vector in the same direction as û up to the earth’s surface,
then its magnitude can be found by the formula, where M is the position vector of the
Moon, k is a constant, and R‘ is the radius of the Earth:

|M
~ ` k û|2 “ R2
‘ 5.0

Solving for k:

0 “ k 2 pû ¨ ûq ` 2kpû ¨ Mq

~ `M ~ ¨M~ ´ R2
‘

“ k 2 ` 2kpû ¨ Mq~ ` |M| ~ 2 ´ R2

‘

Now, |M|
~ 2 ´ R2 “ 1.288 ˆ 1017 m
‘

2pû ¨ Mq “ ´7.178 ˆ 108 m

~
6 0 “ k 2 ´ kp7.178 ˆ 108 q ` p1.288 ˆ 1017 q
a
7.178 ˘ p´7.178q2 ´ 4 ˆ 12.88
k“ ˆ 108
2
The vector will intersect the earth’s surface at two points. Hence two solutions. How-
ever, only the first intersection with the sphere is relevant in this case, so only the
smallest solution to the equation should be considered:
a
7.178 ´ p´7.178q2 ´ 4 ˆ 12.88
k“ ˆ 108
2
“ 3.528 ˆ 108 4.0

25
Page 26 of 40 Theoretical Round

The vector of location of the greatest eclipse corresponds to the position vector of the
Moon plus k times the direction vector û:

~p “ M
~ ` k û
“ă 6.091 ˆ 106 , ´1.828 ˆ 106 , 4.854 ˆ 105 ą 2.0

It is possible to use the following expressions to convert this vector to the spherical
system:
a
r “ x2 ` y 2 ` z 2
“ 6.378 ˆ 106 m
˜a ¸
x2 ` y 2
θ “ arctan (valid for z ą 0)
z
“ 1.495 rad “ 85˝ 381
´y¯
ϕ “ arctan (valid for x ą 0)
x
“ ´0.2915 rad “ ´16˝ 421 2.0

The latitude corresponds to the complement of the angle θ, and the longitude corre-
sponds to the angle ϕ:
4˝ 221 N, 16˝ 421 W 2.0

Part II: Duration of the Totality (50 points)

Precisely determining the duration of totality of a solar eclipse involves complex calculations
that would be beyond the scope of this problem. However, it is possible to obtain a reasonable
approximation for this value using the two following assumptions:
• The size of the umbra on the surface of the Earth remains roughly constant throughout
totality at a given location.
• The velocity of the umbra on the surface of the Earth remains roughly constant throughout
totality at a given location.

(e) (10 points) Estimate the radius of the umbra during the greatest eclipse. In order to
simplify the calculations, assume that the umbra is small enough that it can be considered
approximately flat and that the axis of the Moon’s shadow cone is extremely close to the
centre of the Earth during the greatest eclipse.

Solution: The approximate geometry described in the problem statement is shown

below:

26
Page 27 of 40 Theoretical Round

3.0

The angle β can be calculated as follows:

ˆ ˙
Rd ´ RM oon
β “ arccos
dd ´ dM oon
6.955 ˆ 108 ´ 1.737 ˆ 106
ˆ ˙
“ arccos
1.516 ˆ 1011 ´ 3.589 ˆ 108
“ 89.74˝ 3.0

Using this angle, it is possible to determine the radius of the the umbra:

RM oon ´ pdM oon ´ R‘ q ¨ cospβq

rumbra “ 3.0
sinpβq
rumbra “ 1.196 ˆ 105 m 1.0

(f) (3 points) Calculate the velocity of the Earth’s rotation at the latitude of the centre of the
umbra.

Solution: The Rotational velocity of the Earth at the latitude of the center of the
umbra is the following:
2πR‘
vrot “ cospφumbra q
23 56m 04s
h

2π ˆ 6.378 ˆ 106
“ cosp4˝ 221 q 2.0
23h 56m 04s
vrot “ 463.7 m{s 1.0

(g) (4 points) Determine the orbital velocity of the Moon at the instant of the greatest eclipse.
Neglect the changes in the semi-major axis of the Moon’s orbit.

27
Page 28 of 40 Theoretical Round

Solution: Using the vis-viva equation:

d ˆ ˙
2 1
vM oon “ GM‘ ´ 2.0
dM oon aM oon
d ˆ ˙
2 1
“ 6.674 ˆ 10´11 ˆ 5.972 ˆ 1024 ´
3.589 ˆ 108 3.844 ˆ 108
vM oon “ 1088 m{s 2.0

For the remaining items of this problem, assume that the tangential velocity of the Moon is
roughly the same as the orbital velocity and neglect its radial component.
In order to calculate the velocity of the umbra, it is convenient to define two new additional
right-handed coordinate systems. System II is defined as follows:
• Origin (OII ): position of the Moon at the instant of the greatest eclipse.
• Positive x-axis: Tangent to the declination circle. Points eastwards.
• Positive y-axis: Tangent to the meridian of right ascension. Points northwards.
System III is defined as follows:
• Origin (OIII ): centre of the umbra at the instant of the greatest eclipse.
• Positive x-axis: Tangent to the latitude circle. Points eastwards.
• Positive y-axis: Tangent to the meridian of longitude. Points northwards.
Note that in both systems, the xy-plane is tangent to the celestial sphere at the position of
the origin.
System III is similar to system II, with the only difference being that the origin (OIII ) is at
the centre of the umbra at the moment of the greatest eclipse.
(h) (14 points) Using System II, determine the velocity vector of the Moon during the greatest
eclipse. Note that the intersection between the Celestial Equator and the lunar orbit that
is closer to the position of the eclipse has a right ascension of 23h 07m 59.2s .

Solution: It is possible to use the following diagram to determine the value of the
angle κ between the x-axis and the velocity vector of the Moon:

28
Page 29 of 40 Theoretical Round

6.0

Note that the great circle that intersects the Moon’s meridian of right ascension at a
right angle has a tangent line that coincides with the x-axis, so the angle κ on the
figure is also the angle between the velocity vector of the Moon and the x-axis. Also
note that the Moon is moving from west to east in the celestial sphere (to the left in
the figure), so its velocity vector is on the first quadrant of the xy-plane on this system.
Using the four parts formula:

cospδM oon q cosp90˝ q “ sinpδM oon q cotpαM oon ` 24h ´ αintersection q

´ sinp90˝ q cotp90˝ ´ κq 3.0
6 0 “ sinpδM oon q cotpαM oon ´ αintersection q ´ tanpκq
sinpδM oon q
tanpκq “
tanpαM oon ´ αintersection q
„ 
sinpδM oon q
κ “ arctan
tanpαM oon ´ αintersection q
„ 
sinp21˝ 121 18.42 q
“ arctan
tanp4h 21m 12.6s ´ 23h 07m 59.2s q
κ “ 4˝ 171 2.0

Now it is possible to break down the velocity of the Moon into the x and y components:
» fi » fi
vM oon cospκq 1085 m{s
vII “ – vM oon sinpκq fl “ –81.25 m{sfl
0 0 3.0

The radial component of the velocity was neglected, so the velocity on the z-axis is
equal to zero.

(i) (10 points) Write the velocity vector of the Moon in System III. Note that in System I, the
azimuthal angle difference between the positions of the origins OII and OIII is negligible, so
you should only take into account the difference in the polar angles.

29
Page 30 of 40 Theoretical Round

Solution: It is possible to obtain system III by rotating system II from north to

south by an angle of θumbra ´ θM oon around the x-axis. The angles θumbra and θM oon
correspond to the polar angles of the umbra and the Moon on system I.
The following two figures illustrate this rotation.

Since the rotation is around the x-axis, the x component remains unchanged. The 3.0
component that was originally in the y-axis is broken down into two components on
the y-axis and the z-axis of the new system, which results in the following vector:
» fi
vM oon ¨ cospκq
vIII “ – vM oon ¨ sinpκq ¨ cospθumbra ´ θM oon q fl
´ vM oon ¨ sinpκq ¨ sinpθumbra ´ θM oon q 6.0
» fi
1085 m{s
vIII “ – 77.76 m{s fl
´ 23.54 m{s 1.0

(j) (6 points) Calculate the speed of the centre of the umbra along the surface of the Earth at
the instant of the greatest eclipse.

Solution: The velocity of the umbra corresponds to the vector sum of the component
caused by the rotation of the Earth and the component caused by the velocity of the
Moon.
Note that all points on the axis of the Moon’s shadow cone have the same velocity, vIII
is also the velocity of the axis of the Moon’s shadow cone at the center of the umbra.
Since the umbra moves along the surface of the Earth, so the zIII -component should
be set to zero.
» fi » fi » fi
vIII,x ´ vrot 1085 ´ 463.7 621.4 m{s
vumbra “ – vIII,y fl “ – 77.76 fl “ –77.59 m{sfl
0 0 0 4.0
The modulus of this vector is the following:
2.0
a
vumbra “ |vumbra | “ 621.42 ` 77.762 “ 626.3 m{s

30
Page 31 of 40 Theoretical Round
(k) (3 points) Estimate the duration of the totality of the eclipse at the location with the
coordinates found in part (d).

Solution: Considering the assumptions listed at the beginning of part II, it is possible
to estimate the length of the totality by dividing the diameter of the umbra by its
speed:

2 ¨ rumbra 2 ˆ 1.196 ˆ 105

∆ttotality “ “ 2.0
vumbra 626.3
∆ttotality “ 6min21.4s 1.0

T11. Ground Tracks (75 points)

The projection of a satellite’s orbit onto the Earth’s surface is called its ground track. At a
given instant, one can imagine a radial line drawn outward from the centre of the Earth to the
satellite. The intersection between the Earth’s spherical surface and this radial line is a point
on the ground track.
The location of this point is specified by its geocentric latitude and longitude. The ground track
is then essentially the figure traced by this point as the satellite moves around the Earth.
Part I: Sun-Synchronous Orbits (25 points)
It is particularly interesting to analyse the ground track of a so-called Sun-Synchronous orbit.
This is a nearly polar orbit around a planet where the satellite passes over any given point on
the planet’s surface at the same mean local solar time. This property is especially interesting
for satellite imaging, ensuring similar illumination conditions over different days.
The figure below shows the ground track of a satellite in a Sun-Synchronous orbit. Its inclination
angle (i) - the angle between the satellite orbital plane and the Earth’s equatorial plane - falls
within the range 90˝ ă i ă 180˝ . The graph depicts five complete orbits of the satellite.

Figure 2: Ground Track for five orbits of the satellite

31
Page 32 of 40 Theoretical Round
For the questions in Part I, assume that the Earth’s orbit around the Sun is circular.
(a) (3 points) Determine the nodal precession rate for the orbit in rad/s.

Solution: Since the orbit is Sun-Synchronous, this means that the satellite’s line of
nodes must drift 360˝ along one sidereal year. This ensures that the satellite will always
pass over a given meridian at the same local mean solar time. Hence, 1.0

360˝ 2π
Ω9 “ “ “ 1.991 ¨ 10´7 rad/s 2.0
Ts 365.2564 ¨ 24 ¨ 60 ¨ 60

(b) (8 points) Based on the ground track shown in Figure 2, determine the inclination of the
satellite’s orbit (in degrees) and estimate its orbital period (in minutes). Consider that the
orbital period of the satellite is shorter than one sidereal day.

Solution: After one orbit, the difference in longitude perceived by the satellite on the
ground track is associated with the combined effects of the rotation of the Earth and
nodal precession. This is expressed by

λ2 “ λ1 ´ ω‘ ¨ T ` Ω9 ¨ T 2.0

From the Ground Track, λ1 “ 90˝ and λ2 « ´135˝ for five orbits. Therefore, knowing
the value of ω‘ , the angular velocity of the rotation of our planet,

1 λ1 ´ λ2 1 225˝ ¨ π{180˝
T “ ¨ “ ¨ « 10800 s “ 180 min 2.0
5 ω‘ ´ Ω9 5 7.292 ¨ 10´5 ´ 1.991 ¨ 10´7

Although Ω9 is significantly smaller than ω‘ , it is conceptually important to include

the nodal precession rate in the formula.
The inclination of the orbit can be obtained directly from the Ground Track. Taking
any of the two points of the orbit with the largest (absolute) value of latitude, one finds

|φmax | « 55˝ 2.0

Since i ą 90˝ , |φmax | is the supplementary angle of the inclination:

i “ 180˝ ´ 55˝ “ 125˝ 2.0

(c) (2 points) Calculate the semi-major axis a of the orbit in km.

Solution: Having calculated the period associated with the orbit, the semi-major axis
is directly obtained from
˙1{3
T2 4π 2 T 2 GM‘
ˆ
3
“ ùñ a “ “ 1.06 ¨ 104 km 2.0
a GM‘ 4π 2

(d) (1 point) Determine the number of orbits completed by the satellite until it returns to the
same position on Earth.

32
Page 33 of 40 Theoretical Round

Solution: Since the orbit is sun-synchronous, the satellite completes an integer number
of orbits during one solar day. Hence,
24 ¨ 60
NS “ “ 8 orbits 1.0
180

(e) (11 points) As seen in Figure 2, the ground track crosses the Brazilian city of Maceió
pφ, λq “ p9.7˝ S; 35.7˝ W q and also Chorzów pφ, λq “ p50.3˝ N ; 19.0˝ Eq, in Poland. Knowing
that the ground track crosses Maceió at noon (local time), determine the local time that
the satellite track crosses Chorzów. Hint: specifically for this task, you may neglect the
effects of nodal precession.

Solution: The satellite takes more than 2 orbits and less than 3 orbits to go from
Poland to Brazil. Using as reference the point Po ahead of Maceió that is exactly 3
orbits after Chorzów,

λo “ λC ´ 3 ¨ 45˝

φ0 “ φC

Using sine law in the following spherical triangle,

sin φ
sin θ “ 2.0
sinp180˝ ´ iq

ˆ ˙
sin φo
θ0 “ arcsin “ 69.9˝
sin i

33
Page 34 of 40 Theoretical Round

ˆ ˙
sin φM
θM “ arcsin “ ´12.0˝
sin i

The angle the satellite traveled will depend on how many complete orbits it took and
the angle from the final whole orbit to Maceió:

∆θ “ n ¨ 360˝ ´ pθ0 ´ θM q 3.0

by the graph, n “ 3, which results in ∆θ “ 998.1˝ . Knowing its orbit is circular, 1.0

∆θ
∆t “ T ¨ “ 8.32 h 1.0
360˝
The local time at Chorzów over which the satellite passes by will be, considering λM “
35.7˝ W “ ´35.7˝ ,
´1
TC “ TM ´ ∆t ` ω‘ pλC ´ λM q 3.0

TC « 7h20min 1.0

Part II: Tundra orbits (50 points)

A Tundra orbit is a type of geosynchronous elliptical orbit characterised by a high incli-
nation. The apogee is positioned over a specific geographic region, allowing for prolonged
visibility and coverage over that area. This orbit ensures that a satellite spends the major-
ity of its orbital period over the northern - or southern hemisphere, making it particularly
useful for communications and weather observation over high-latitude regions.
The image below represents the ground track of a satellite in a Tundra orbit with an
argument of perigee equal to 270˝ . The satellite orbits the Earth in the same direction as
its rotation. For the following items, you can ignore the effects of the Earth’s oblateness.

Figure 3: Tundra Orbit Ground Track for one orbital period

34
Page 35 of 40 Theoretical Round
(f) (4 points) Based on the graph above, give the inclination of the satellite’s orbit i (in de-
grees), its orbital period T (in minutes), and its semi-major axis a (in km).

Solution: For the inclination, we need to take the maximum latitude reached by the
satellite, which gives us i “ 63˝ . For the period, we have to realize that in one orbit, 1.0
the Earth’s rotation does not cause a shift in the orbit, so it must be geosynchronous,
that is T “ T‘ “ 1436 min . For the semi-major axis, we apply Kepler’s Third Law: 1.0

a3 GM
2
“ Ñ a “ 42, 164 km 2.0
T 4π 2

(g) (12 points) Show that the time a satellite spends in the northern hemisphere is given by

1 sin´1 peq
ˆ ˙
1 e a 2
T “ ` ` ¨ 1´e T
2 π π
where e is the eccentricity of the orbit and T is its orbital period.

Solution: We will first find the expression for the area divided by the semi-latus
rectum as a function of eccentricity so then we can apply the Kepler’s Second Law 2.0
to find the required time. To do this, `we˘will use the projection of the area of a circle
when it is inclined by an angle arccos ab , as shown in the figure below:

Thus, we will first calculate the hatched area of the circle figure Acircle

θ ae ¨ b
Acircle “ Asector ´ Atriangle “ ¨ πa2 ´ 2.0
2π 2
We can also use that θ “ π
2 ´ α and α “ sin´1 peq. Therefore:

π ¨ a2 sin´1 peq ¨ a2
ˆ ˙
ab ¨ e
Acircle “ ´ ´
4 2 2
Now, we project this area,

35
Page 36 of 40 Theoretical Round

b
Aellipse “ Acircle ¨
a

πab sin´1 peq ¨ ab b2 e

ˆ ˙
Aellipse “ ´ ´ 4.0
4 2 2
Thus, the area covered by the satellite’s position vector from one semi-latus rectum to
the other will be given by twice the area found above, as we can see from the image,
so that the area on the perigee side is given by

πab
Aper “ ´ bpa sin´1 peq ` ebq
2
For the apogee side, it is enough to know that Aap “ πab ´ Aper , where πab is the area
of the whole ellipse. Therefore, we have:

πab
2.0
` ˘
Aap “ ` b a sin´1 peq ` eb
2
Using Kepler’s Second Law, we can write:

Aap T1
“ 2.0
Aellipse T
?
Using that Aellipse “ πab and b “ a 1 ´ e2 , we have:

T1 1 sin´1 peq e a
“ ` ` ¨ 1 ´ e2
T 2 π π
Alternative Solution

Source: https://en.wikipedia.org/wiki/Eccentric_anomaly

The relation between the true anomaly and the mean anomaly is given by

M “ E ´ e ¨ sinpEq

36
Page 37 of 40 Theoretical Round

From the figure (improve figure later), we observe that the satellite has the following
properties:

• At perigee:
– E“0
– θ“0

• At the semi-latus rectum:

– E “ cos´1 peq
– θ “ 90˝ 2.0

Therefore,?at the moment of passage through the semi-latus rectum, knowing that
sinpEq “ 1 ´ e2 , we have

4.0
a
M “ cos´1 peq ´ e ¨ 1 ´ e2

If we also use the perigee as the initial reference for the mean anomaly, we can associate
the time of passage from the perigee to a certain point in the orbit by
M
t“
ω
where ω is the mean angular velocity of the orbit, given by ω “ 2π
T , with T being the
orbital period. Thus, the time elapsed from the perigee to the semi-latus rectum is
given by ?
cos´1 peq ´ e ¨ 1 ´ e2
t“T 2.0
2π
Finally, the time T 1 spent in the northern hemisphere can be found as follows:

T 1 ` 2t “ T Ñ T 1 “ T ´ 2t

Thus, the desired ratio T 1 {T will be:

T1 2t
“1´ 2.0
T T

Substituting the expression for t, we get:

?
T1 cos´1 peq ` e ¨ 1 ´ e2
“1´
T π

If we use the relation cos´1 peq ` sin´1 peq “ 2,

π
we finally obtain:

T1 1 sin´1 peq e a
“ ` ` ¨ 1 ´ e2
T 2 π π

(h) (10 points) Estimate numerically the eccentricity e of its orbit. You can consider that the
eccentricity is so small that sinpeq « e and e2 ! 1.

37
Page 38 of 40 Theoretical Round

Solution: Between the two consecutive passages through the semi-latus rectum, we
can use the following equation:

∆λ “ π ´ ω‘ ¨ T 1 2.0
Where ∆λ is the variation in longitude of the satellite between its passages through
the semi-latus rectum, as shown in the image below.

Therefore, we can write using the expression of the last item

1 sin´1 peq
ˆ ˙
e a 2
∆λ ` ω‘ ¨ T ¨ ` ` ¨ 1´e “π
2 π π

As ω‘ “ 2π
T (the orbit is geosynchronous), we will get:
a
∆λ ` 2 ¨ sin´1 peq ` 2e ¨ 1 ´ e2 “ 0

∆λ
2.0
a
sin´1 peq ` e ¨ 1 ´ e2 “ ´
2
Considering that ∆λ « ´67.5˝ from the graph (it is important to note that this 2.0
value should be negative by the figure), we find from the above formula, by iteration,
e « 0.3 . If we use the approximations sinpeq « e and e2 ! 1 given in the statement,
we find:

∆λ
2e “ ´ 3.0
2
which gives us an approximate answer e « 0.295 « 0.3 1.0

(i) (18 points) From the ground track, we can observe that the satellite exhibits retrograde
motion in both its northern and southern hemisphere trajectories. Find the true anomaly
(in degrees) of the satellite at the beginning and end of its retrograde motion in the southern
hemisphere.

Solution: At the start and end points of retrograde motion, we have ωRA “ ω‘ , 2.0
where ωRA is the right ascension angular velocity of the satellite. To find this velocity,

38
Page 39 of 40 Theoretical Round

we use the following spherical triangle:

where θ2 “ θ ´ 90˝ , and it is known that:

cospθ2 q “ cospδq cospαq (Cosine Law) 1.0

sinpθ2 q cospδq
“ (Sine Law) 1.0
sinpαq cospiq
We substitute cospδq from the first expression into the second expression, obtaining:

sinpθ2 q cospθ2 q
“
sinpαq cospαq ¨ cospiq

tanpθ2 q ¨ cospiq “ tanpαq

We then derive the above expression with respect to time - it is possible to skip the
derivative step by just decomposing the satellite’s angular velocity vector.

cospiq 9 α9
θ“ 2.0
cos2 pθ2 q cos2 pαq
Using the Cosine Law expression to replace cospαq, we finally have:

cospiq 9
ωRA “ α9 “ θ 2.0
cos2 pδq

We can then calculate the value of θ9 as a function of the distance of the satellite from
the center of the Earth, using the conservation of angular momentum:

2
?
2
9θ “ 2π a 1 ´ e 2.0
T r 2

We still have the polar expression for the distance r as a function of the polar radius,
given by:

39
Page 40 of 40 Theoretical Round

ap1 ´ e2 q
r“ 2.0
1 ` e ¨ cospθq
Substituting everything into the expression that gives the condition for the start or end
of retrograde motion, we find:

2π p1 ` e cos θq2 cospiq 2π

“ ωearth “
T p1 ´ e2 q 32 cos2 pδq T

p1 ` e cos θq2 cospiq

3 “1 2.0
p1 ´ e2 q 2 cos2 pδq
From the graph, we find the latitude - or declination - of the start or end of retrograde
motion, given by δ « ´32˝ . Thus, substituting all the variables into the above equation 2.0
and solving for θ, we find: θ “ 54 ˝
and θ “ 306 ˝
as solutions. 2.0

(j) (6 points) It is also noticeable that the ground track of a Tundra orbit has the shape of a
figure-8, similar to an analemma, so that the satellite passes over the same point on Earth
in a single orbit. Calculate the minimum eccentricity the orbit would need to have for this
property to cease occurring. Use the same orbital inclination as the orbit in Figure 3.

Solution: For the figure-8 shape not to form, we need the inversion of motion in the
northern hemisphere to occur exactly at the apogee position . Therefore, we will use 2.0
exactly the same expression as the previous item, but with θ “ 180˝ and eccentricity as
the unknown variable . For the value of δ, as the satellite is at the position of maximum 1.0
declination, δ “ i. We have:

p1 ´ eq2 1
3 “1 2.0
p1 ´ e2 q 2 cos piq

Solving the above equation, we obtain: e « 0.4 1.0

40