AST 251 | U of T | Dr.

Reid | 1
Project 1: Analyzing Transit Light Curves
• The three projects in this course form a progression from
present-day understanding of potentially habitable.
environments, to what we might expect to be able to learn
within around 50 years.
• Released today after class, due February 25 at 11:59 p.m.
• You will be given three data sets: pick TWO to analyze and
discard the other at your discretion.
• Everyone has unique data sets.
• I encourage you to come to my office hours to talk through your
dataset and methods if you are uncertain how to proceed.

AST 251 | U of T | Dr. Reid | 2

Searching for
Habitable Planets

AST 251 | U of T | Dr. Reid | 3

How do we find and characterize
potentially habitable planets?

AST 251 | U of T | Dr. Reid | 4

We’ll begin with planets in other solar systems to
develop a sense of the range of possible planet
types.

Later, we’ll examine individual planets closely,

including those in our own solar sytem.

AST 251 | U of T | Dr. Reid | 5

We refer to planets orbiting
stars other than the Sun as
extrasolar planets or
exoplanets.

Credit: NASA Ames/JPL-Caltech/T. Pyle

AST 251 | U of T | Dr. Reid | 6
The most obvious way to find an exoplanet is simply to
photograph another star and examine the image for
orbiting planets.

This technique is known as direct imaging.

AST 251 | U of T | Dr. Reid | 7

Direct imaging is very difficult for the same reason it is
difficult to photograph a firefly sitting on a car headlight: the
contrast between star and planet is too high.

Credit: Flickr user mikelietz

AST 251 | U of T | Dr. Reid | 8

So far, direct imaging has
detected only around 150
exoplanets. We can only see
very young Jupiter-type (gas
giant) planets orbiting
extremely far from their parent
stars, and we only see them as
single pixels.

They are not likely homes for

life.

Credit: Jason Wang (Caltech)/Gemini Planet Imager Exoplanet Survey

AST 251 | U of T | Dr. Reid | 9
Direct imaging is the only method in which
exoplanets are actually seen.

All of the other methods are indirect, meaning

that the existence and properties of the
planets are deduced from observations of the
host star.

AST 251 | U of T | Dr. Reid | 10

There are several other detection methods, but
the two most successful ones have been the
transit method and the radial velocity
method.

We will start with the transit method and later

combine it with the radial velocity method.

AST 251 | U of T | Dr. Reid | 11

A transit occurs when one astronomical object
passes in front of another from our point of view.

star

a transiting planet

AST 251 | U of T | Dr. Reid | 12

A transit of the Moon across
the Sun, as seen by the
STEREO spacecraft

Credit: NASA/STEREO

AST 251 | U of T | Dr. Reid | 13

Our current telescopes aren’t powerful enough to
actually photograph exoplanets transiting the way
you just saw the Moon transiting the Sun.

However, we can spot transits in a star’s light

curve.

AST 251 | U of T | Dr. Reid | 14

The light curve of a
star is a graph of the
star’s brightness
versus time

star’s brightness

time
AST 251 | U of T | Dr. Reid | 15
star’s brightness

?
time
AST 251 | U of T | Dr. Reid | 16
Here, ingress and
egress are too quick
for you to see the
slope (but you could
see them if you could
zoom the time axis).

Credit: NASA
AST 251 | U of T | Dr. Reid | 17
star’s brightness

1
Ingress and
egress take time,
0.9999
so the light curve
is sloped.

time
AST 251 | U of T | Dr. Reid | 18
The amount of light blocked by a planet varies with
the size of the planet.

So, we can use the transit method to directly

measure the size of the planet, even without seeing
the planet!

AST 251 | U of T | Dr. Reid | 19

 1

star’s brightness
Depth

Rplanet Rstar 0.99

time
cross−sectional area of planet
𝐃𝐞𝐩𝐭𝐡 =
cross−sectional area of star

AST 251 | U of T | Dr. Reid | 20

 1

star’s brightness
Depth

Rplanet Rstar 0.99

time
cross−sectional area of planet
𝐃𝐞𝐩𝐭𝐡 =
cross−sectional area of star
𝛑𝐑𝟐𝐩𝐥𝐚𝐧𝐞𝐭
𝐃𝐞𝐩𝐭𝐡 =
𝛑𝐑𝟐𝐬𝐭𝐚𝐫

AST 251 | U of T | Dr. Reid | 21

 1

star’s brightness
Depth

Rplanet Rstar 0.99

time
cross−sectional area of planet
𝐃𝐞𝐩𝐭𝐡 =
cross−sectional area of star
𝛑𝐑𝟐𝐩𝐥𝐚𝐧𝐞𝐭
𝐃𝐞𝐩𝐭𝐡 =
𝛑𝐑𝟐𝐬𝐭𝐚𝐫

𝐑 𝐩𝐥𝐚𝐧𝐞𝐭 = 𝐑 𝐬𝐭𝐚𝐫 𝐃𝐞𝐩𝐭𝐡

AST 251 | U of T | Dr. Reid | 22
 Measuring transit depths
allows us to measure the
sizes of each planet in an
exoplanetary system, even
though we can’t see the
planets themselves.

We then have artists

render what the system
could look like. Remember
we don’t actually
photograph these planets
directly.

Credit: NASA/JPL-Caltech AST 251 | U of T | Dr. Reid | 23

The light curves we’ve seen so far are highly idealized.
Among other idealizations, we’ve ignored noise.

Real light curves contain significant noise, originating

from both astronomical sources and the camera used to
record the data.

AST 251 | U of T | Dr. Reid | 24

(Muirhead et al., ApJ, 2015)

Real light curves are often very

noisy. The points must be
averaged in time to build up a
higher signal-to-noise ratio and
make the transit more
apparent.

Usually, light curves are “phase

folded”, meaning that many
different transits of the same
planet are combined to improve
the data quality.
(Zhou et al., AJ, 2017)
AST 251 | U of T | Dr. Reid | 25
star’s brightness

Phase folding is the

process of averaging
together multiple transits to
improve data quality.
time
I won’t require you to do it
yourself, but you’ll
sometime see it in plots
star’s brightness

from the scientific

literature.

phase AST 251 | U of T | Dr. Reid | 26

star’s brightness

Phase folding is the

process of averaging
together multiple transits to
improve data quality.
time
I won’t require you to do it
yourself, but you’ll
sometime see it in plots
star’s brightness

from the scientific

literature.

phase AST 251 | U of T | Dr. Reid | 27

Also, keep in mind that not everything that
produces a dip in the light curve of a star will be a
planet.

Can you think of other things that might produce a

dip in a star’s light curve?

AST 251 | U of T | Dr. Reid | 28

 A sunspot on the host
star can produce an
effect similar to a
transiting planet.

Credit: NASA/SDO AST 251 | U of T | Dr. Reid | 29

To confirm that a transit is
caused by a planet, we look for
multiple transits of the same
depth repeating at regular
intervals, corresponding to
successive orbits of the planet.

We also have to check that the

size of the object is within the
plausible range of sizes for a
planet.

AST 251 | U of T | Dr. Reid | 30

noise in the measurements

AST 251 | U of T | Dr. Reid | 31

 Typically, we would want to
noise in the measurements observe a planet transiting
at least three times to be
confident it exists.

If we only see one or two

transits, it’s possible that
something else is
responsible.

probable repeating transits AST 251 | U of T | Dr. Reid | 32

How to Decide Whether a
Transiting Planet is in the
Habitable Zone

AST 251 | U of T | Dr. Reid | 33

One critical piece of HZ
information is how far the
planet orbits from the star. Is
it in the habitable zone?

AST 251 | U of T | Dr. Reid | 34

The distance at which a planet orbits is
directly related to the time it takes to
orbit, called its orbital period.

AST 251 | U of T | Dr. Reid | 35

The duration of an
individual transit
(minutes to hours) is

star’s brightness (relative)

usually much shorter
than the orbital period of
the planet (days to
years). So, the transits
will appear highly zoom in on a
single transit
compressed on a light
curve without phase
folding.

time (days)
AST 251 | U of T | Dr. Reid | 36
star’s brightness (relative)

orbital
period

time (days)
AST 251 | U of T | Dr. Reid | 37
The relationship between a planet’s orbital
period and its distance from its parent star is
called Kepler’s Third Law.

AST 251 | U of T | Dr. Reid | 38

 German astronomer Johannes
Kepler (1571-1630) worked out
the three laws of planetary.

Kepler’s laws are actually just

consequences of Newton’s Law
of Gravity, which hadn’t been
discovered in Kepler’s time.

Credit: Public domain

AST 251 | U of T | Dr. Reid | 39
Kepler’s First Law
All planets orbit their parent stars in elliptical
orbits with the star* at one focus.
*Thisis technically not correct—the center of mass of the system is at one focus. However, in the
general case that the star is much more massive than any of its planets, the difference is small.
We will return to this point.

AST 251 | U of T | Dr. Reid | 40

 semi-minor axis
focus
center
semi-major axis, a

focus

This ellipse has eccentricity = c/a = 0.6

AST 251 | U of T | Dr. Reid | 41

 center
radius

This ellipse has eccentricity = 0 (i.e. it’s a circle)

AST 251 | U of T | Dr. Reid | 42

Orbits with different eccentricities
(thicker line à higher e)
AST 251 | U of T | Dr. Reid | 43
Credit: YouTube user Ms. Edge AST 251 | U of T | Dr. Reid | 44
Kepler’s Second Law*
As they orbit, planets sweep out equal
areas in equal times.
*This is the one that confuses everyone. It’s really a statement of the conservation of angular momentum.

AST 251 | U of T | Dr. Reid | 45

 perihelion distance = a(1 − e) aphelion distance = a(1 + e)

AST 251 | U of T | Dr. Reid | 46

Credit: Pearson Education
Snapshots of an orbiting planet spaced at equal
times—the planet moves faster at perihelion
then at aphelion.
AST 251 | U of T | Dr. Reid | 47
Kepler’s Third Law
The square of a planet’s orbital period
is proportional to the cube of its orbital
semi-major axis.

AST 251 | U of T | Dr. Reid | 48

 2a

𝟐
𝟒𝛑
𝐏𝟐 = 𝐚𝟑
𝐆(𝐌𝐩𝐥𝐚𝐧𝐞𝐭 + 𝐌𝐬𝐭𝐚𝐫 )

AST 251 | U of T | Dr. Reid | 49

If we assume the mass of the planet is much less than the
mass of the star, we can use Kepler’s Third Law to work out
the planet’s orbital semimajor axis from its orbital period:

𝟐
𝟒𝛑
𝐏𝟐 = 𝐚𝟑
𝐆(𝐌𝐩𝐥𝐚𝐧𝐞𝐭 + 𝐌𝐬𝐭𝐚𝐫 )

AST 251 | U of T | Dr. Reid | 50

If we assume the mass of the planet is much less than the
mass of the star, we can use Kepler’s Third Law to work out
the planet’s orbital semimajor axis from its orbital period:

𝟐
𝟒𝛑
𝐏𝟐 = 𝐚𝟑
𝐆(𝐌𝐩𝐥𝐚𝐧𝐞𝐭 + 𝐌𝐬𝐭𝐚𝐫 )
𝐢𝐟 𝐌𝐬𝐭𝐚𝐫 ≫ 𝐌𝐩𝐥𝐚𝐧𝐞𝐭

then 𝐌𝐩𝐥𝐚𝐧𝐞𝐭 + 𝐌𝐬𝐭𝐚𝐫 ≅ 𝐌𝐬𝐭𝐚𝐫

AST 251 | U of T | Dr. Reid | 51

If we assume the mass of the planet is much less than the
mass of the star, we can use Kepler’s Third Law to work out
the planet’s orbital semimajor axis from its orbital period:

𝟐
𝟒𝛑
𝐏𝟐 = 𝐚𝟑
𝐆(𝐌𝐩𝐥𝐚𝐧𝐞𝐭 + 𝐌𝐬𝐭𝐚𝐫 )
𝐢𝐟 𝐌𝐬𝐭𝐚𝐫 ≫ 𝐌𝐩𝐥𝐚𝐧𝐞𝐭
+
+
4π
P ≅ a0 then 𝐌𝐩𝐥𝐚𝐧𝐞𝐭 + 𝐌𝐬𝐭𝐚𝐫 ≅ 𝐌𝐬𝐭𝐚𝐫
GM,-./

AST 251 | U of T | Dr. Reid | 52

In summary: From the transit depth, we get the
planetary radius. From the orbital period, we get
the orbital semimajor axis.

Taking both factors into account, we can decide

from a transit light curve roughly what type of
planet we have detected and whether it is in the
habitable zone of its parent star. This is what you
will do for your first project.

AST 251 | U of T | Dr. Reid | 53

Interpreting Realistic Transit
Light Curves

AST 251 | U of T | Dr. Reid | 54

 1

star’s brightness

time
Very rarely can we monitor a single star continuously. We have to
deal with interruptions due to weather, interference from the Moon,
the day/night cycle, other observing priorities, and so on. So, light
curves are often sparsely sampled.

AST 251 | U of T | Dr. Reid | 55

 Could there be a pair of missing transits?

star’s brightness

time
Very rarely can we monitor a single star continuously. We have to
deal with interruptions due to weather, interference from the Moon,
the day/night cycle, other observing priorities, and so on. So, light
curves are often sparsely sampled.

AST 251 | U of T | Dr. Reid | 56

 Could there be a BUNCH of missing transits?

star’s brightness

time
Very rarely can we monitor a single star continuously. We have to
deal with interruptions due to weather, interference from the Moon,
the day/night cycle, other observing priorities, and so on. So, light
curves are often sparsely sampled.

AST 251 | U of T | Dr. Reid | 57

 A lower observing
cadence may mean that
missing missing
ingress or egress are
ingress egress absent from the data,
making it harder to
establish whether a given
dip in the light curve is
genuinely associated with
a planet.

AST 251 | U of T | Dr. Reid | 58

So far, we have looked at light curves with only a single
planet present.

If you were an alien looking at our own solar system, you

could see as many as 8 transiting planets plus
miscellaneous small objects, all in the same light curve!

AST 251 | U of T | Dr. Reid | 59

How many planets
do you see here?

Planets may come in

a range of sizes, and
orbital periods. Some
of them may be very
difficult to identify in
the light curve.

Some might be
buried in or close to
the noise.

AST 251 | U of T | Dr. Reid | 60

How many planets
do you see here?

Planets may come in

a range of sizes, and
orbital periods. Some
of them may be very
difficult to identify in
the light curve.

Some might be
buried in or close to
the noise.

AST 251 | U of T | Dr. Reid | 61

How many planets
do you see here?

Planets may come in

a range of sizes, and
orbital periods. Some
of them may be very
difficult to identify in
the light curve.

Some might be
buried in or close to
the noise.

AST 251 | U of T | Dr. Reid | 62

How many planets
do you see here?

Planets may come in

a range of sizes, and
orbital periods. Some
of them may be very
difficult to identify in
the light curve.

Some might be
buried in or close to
the noise.

AST 251 | U of T | Dr. Reid | 63

Planets themselves emit & reflect light.

If a transiting planet emits or reflects a large

amount of light, this may be noticeable during
the secondary eclipse when the planet passes
behind the star.

AST 251 | U of T | Dr. Reid | 64

 no eclipse
primary transit
(we see all of the light secondary eclipse
(planet covers star)
from the star and the (star covers planet)
planet)

star brightness

time
AST 251 | U of T | Dr. Reid | 65
JWST has recently
used a secondary
eclipse to measure
the surface
temperature of
rocky exoplanet
TRAPPIST-1b.

AST 251 | U of T | Dr. Reid | 66

AST 251 | U of T | Dr. Reid | 67
 Modelling of the light
curve before, during,
and after secondary
eclipse also allows
calculation of the
albedos of
exoplanets.

Credit: Adams et al., 2021, ApJ

AST 251 | U of T | Dr. Reid | 68

So far, we’ve been
representing stars as
perfectly uniform discs of
light.

Credit: NASA/SDO
AST 251 | U of T | Dr. Reid | 69
As shown in this
photograph of the Sun,
real stars are not
uniformly bright across
their surfaces.

Credit: NASA/SDO
AST 251 | U of T | Dr. Reid | 70
 Light from these regions is
coming from cooler layers of
the star, so they appear redder
and darker

Light from these regions is

coming from hotter layers
of the star, so they appear
yellower and brighter

Credit: NASA/SDO
AST 251 | U of T | Dr. Reid | 71
 Representing stars as
discs of uniform
brightness produces
1
transits that have flat
bottoms.

time
AST 251 | U of T | Dr. Reid | 72
 The surfaces of real stars
are brightest toward their
centres and darker toward
their edges, a phenomenon
called limb darkening.
1
Limb darkening smoothes
out the transit light curve
Limb darkening rounds
all the straight lines in
and rounds the bottom of
the light curve. the transit.
time AST 251 | U of T | Dr. Reid | 73
Planets may also
transit across any
part of the disk of
impact
their host star. The
parameter, b
impact parameter
measures how far a
transit is away from
the midplane of the
star.

AST 251 | U of T | Dr. Reid | 74

In some cases, only
part of the planet may
pass in front of the
star. We call this
grazing incidence.

AST 251 | U of T | Dr. Reid | 75

Limb darkening and grazing incidence can
affect your interpretation of the transit depth,
and hence your measurement of the planet’s
radius.

AST 251 | U of T | Dr. Reid | 76

For example, a grazing transit of
a large planet could be mistaken
for a full transit of a smaller
planet, if the amount of starlight
blocked is the same in both
cases.

In this case, the blue regions

have the same areas and would
produce the same transit depth,
but the planets are very different
sizes.

AST 251 | U of T | Dr. Reid | 77

Careful examination of the shape of the
transit in the light curve can distinguish these
cases.

AST 251 | U of T | Dr. Reid | 78

 Without limb darkening

star’s brightness
In these light curves,
we are varying the
impact parameter time
without limb darkening. Credit: planethunters.org
AST 251 | U of T | Dr. Reid | 79
 With limb darkening

star’s brightness
Light curves with the
same set of impact
parameters, but now time
with limb darkening. Credit: planethunters.org
AST 251 | U of T | Dr. Reid | 80
As we’ve learned, many
stars are variable,
meaning that their intrinsic
luminosity changes over
time.

This gives the light curve

an uneven baseline, and
makes it complicated to
determine the depth of
each transit.

AST 251 | U of T | Dr. Reid | 81

The transit method can also be used to
detect things orbiting an exoplanet, such as
rings or moons.

AST 251 | U of T | Dr. Reid | 82

Credt: NASA's Goddard Space Flight Center and J. Koynock
AST 251 | U of T | Dr. Reid | 83
 planet

n
oo
m
In October 2018, the first evidence of
an exomoon emerged. Credit: Dan Durda
(Teachey & Kipping, Science, 2018)
AST 251 | U of T | Dr. Reid | 84
We’ve already said that most of the places in our
own solar system that have enduring liquid water
are moons, not planets.

Might exomoons be important habitats for life

elsewhere in the universe?

We will return to this point.

AST 251 | U of T | Dr. Reid | 85

Appendix: Why Transit Depths Don’t Depend
on How Far the Planet is from its Host Star

The following slides are here to answer a question that often

comes up, but isn’t strictly necessary for you to understand. We
won’t go over these in class.

AST 251 | U of T | Dr. Reid | 86

At first, it may seem that it’s the distance between the star and
the exoplanet that determines how much of the star’s light is
blocked distant star
(not the Sun)
Earth

exoplanet close to star

AST 251 | U of T | Dr. Reid | 87

At first, it may seem that it’s the distance between the star and
the exoplanet that determines how much of the star’s light is
blocked distant star
(not the Sun)
Earth

exoplanet far from star

AST 251 | U of T | Dr. Reid | 88

 In reality, the Earth is so far away that the light rays
received from the other star are almost parallel. Larger
planets simply block more of them.
distant star (not
the Sun)
to distant Earth

smaller exoplanet

AST 251 | U of T | Dr. Reid | 89

 In reality, the Earth is so far away that the light rays
received from the other star are almost parallel. Larger
planets simply block more of them.
distant star (not
the Sun)
to distant Earth

larger exoplanet

AST 251 | U of T | Dr. Reid | 90

Appendix: Calculations With
Kepler’s Third Law
You may need to use Kepler’s third law for your
projects. I’m including a sample calculation to
show you how to work through it.

AST 251 | U of T | Dr. Reid | 91

Let’s say P = 2.1 years and Mstar = 2.0 M☉. To find a, first
convert everything to standard units (m, s, kg, etc.)

AST 251 | U of T | Dr. Reid | 92

Let’s say P = 2.1 years and Mstar = 2.0 M☉. To find a, first
convert everything to standard units (m, s, kg, etc.)

Convert P to seconds:
P = (2.1 y)(365.24 days/y)(86400 s/day) = 6.627 x 107 s

AST 251 | U of T | Dr. Reid | 93

Let’s say P = 2.1 years and Mstar = 2.0 M☉. To find a, first
convert everything to standard units (m, s, kg, etc.)

Convert P to seconds:
P = (2.1 y)(365.24 days/y)(86400 s/day) = 6.627 x 107 s

Convert Mstar to kg:

Mstar = 2MSun = (2)(1.989 x 1030 kg) = 3.978 x 1030 kg

AST 251 | U of T | Dr. Reid | 94

Let’s say P = 2.1 years and Mstar = 2.0 M☉. To find a, first
convert everything to standard units (m, s, kg, etc.)

Convert P to seconds:
P = (2.1 y)(365.24 days/y)(86400 s/day) = 6.627 x 107 s

Convert Mstar to kg:

Mstar = 2MSun = (2)(1.989 x 1030 kg) = 3.978 x 1030 kg
𝟏/𝟑
(𝟔. 𝟔𝟕𝟒×𝟏𝟎)𝟏𝟏 𝐦𝟑 𝐤𝐠 )𝟏 𝐬 )𝟐 )(𝟑. 𝟗𝟕𝟖×𝟏𝟎𝟑𝟎 𝐤𝐠) 𝟕 𝐬)𝟐
𝐚≅ (𝟔. 𝟔𝟐𝟕×𝟏𝟎
𝟒(𝟑. 𝟏𝟒𝟏𝟓𝟗)𝟐

AST 251 | U of T | Dr. Reid | 95

Let’s say P = 2.1 years and Mstar = 2.0 M☉. To find a, first
convert everything to standard units (m, s, kg, etc.)

Convert P to seconds:
P = (2.1 y)(365.24 days/y)(86400 s/day) = 6.627 x 107 s

Convert Mstar to kg:

Mstar = 2MSun = (2)(1.989 x 1030 kg) = 3.978 x 1030 kg
𝟏/𝟑
(𝟔. 𝟔𝟕𝟒×𝟏𝟎)𝟏𝟏 𝐦𝟑 𝐤𝐠 )𝟏 𝐬 )𝟐 )(𝟑. 𝟗𝟕𝟖×𝟏𝟎𝟑𝟎 𝐤𝐠) 𝟕 𝐬)𝟐
𝐚≅ (𝟔. 𝟔𝟐𝟕×𝟏𝟎
𝟒(𝟑. 𝟏𝟒𝟏𝟓𝟗)𝟐

𝐚 ≅ 𝟑. 𝟎𝟗×𝟏𝟎𝟏𝟏 𝐦 = 2.1 AU

AST 251 | U of T | Dr. Reid | 96

1 AU = 1 Astronomical Unit
1 AU ≈ average distance between
Earth and the Sun
≈ 150 million km

Credit: Wikimedia Commons user Huritisho

AST 251 | U of T | Dr. Reid | 97
For this class, you can also
use Wolfram Alpha or similar
tools to perform calculations or
check your results.

You have to be able to enter

mathematical formulas using
the notation it expects, but it is
convenient to ensure that you
have correctly accounted for all
of the units. It also often
provides answers in several
different types of units, so that
you can check that they make
sense.

AST 251 | U of T | Dr. Reid | 98