ame: _______________
N
Date: ________________
Lab (M/T): ___________
Exoplanets Lab
_____________________________________________________________________________
Part 1: The Radial Velocity Method
cience objective:find the period of the orbitalmotion.
S
Computational objective:Beginners: data visualization,plotting;Intermediate: data analysis:
goodness of fit;Advanced: fitting and exploring theparameter space to find the best-fit
parameters.
I n this section of the lab, you will use python to create plots of radial velocity versus time for a
known exoplanet system, and find its orbital period. The data set you will be using is
‘51_Peg.csv’. The goal of this lab is to familiarize yourself with the coding language Python -
which is widely used in Astronomy for handling data, making calculations, plotting, running
simulations, etc.
here are 3 versions of this part of the lab, depending on your comfort level with the coding
T
language Python, and your familiarity with numerical techniques (like fitting functions). Pick
whichever version you feel the most comfortable with. In all cases, you will open a python
notebook using Google Colab (a Google Documents/Sheets/Slides version of coding notebooks).
If you are a strong programmer but primarily use a different language, you can take this
opportunity to learn some Python. If you have plenty of experience with numerical techniques,
you can work through the most challenging version of the activity.
I f you create a new notebook or use Jupyter Notebooks instead, please indicate at the beginning
of your code which version you are doing.
on’t forget that you’ve done some coding in previous labs. While each project is unique, code
D
snippets are not. The best code is made of pieces you can copy, and use somewhere else. Start by
opening up those previous notebooks so you can use them.
1 - Beginner
(a) Review your past work
1 . S et yourself a 10 minute timer so you don’t spend too long on this part.
2. Open the previous notebook.
a. Identify the parts you can copy directly, like loading the packages or importing
files.
b. Look for the skills used in each, like making a graph, fitting a function, and doing
calculations. You may want to take a moment to make yourself a note of which
section to find which skill.
(b) Analysis of exoplanet data
3. Make acopyof thebeginner notebook. It has steps matching this procedure to guide you,
but no code.
4. Add the code to load in the applicable packages and data (‘51_Peg.csv’) into your python
notebook.
5. Save each of the columns of data as their own named data set.
6. Make a scatter plot (just points, no connecting line) of radial velocity (y-axis) versus date
(x-axis). Make sure to include x and y-labelswithunits, as well as a title. You’ll notice
that the date is in units of Reduced Julian Days(JD). 2450000 was subtracted from the
absolute date in order to express the values in more convenient numbers.
7. Recall from the intro that an exoplanet should induce a sinusoidal variation in the radial
velocity of the star. In your plot, there should be variation visible, but it should be
difficult to tell if it is sinusoidal, or what the period is, because the observations were
essentially random. Recall also that the “theoretical expectation” is a sine function with
some period and amplitude. A fitting code is usually used in order to find the best
solution, or the closest match between the data and the theoretical expectation from the
function, and find the best-fit period. We will doa simplified version of this type of
analysis; To find the period, we will try a few guessesand check by eye to see if the data
match the guess. For 51 Peg, we will examine two possible “guesses” for the period of
the planet: 4.2 days, and 6.7 days. To make it easier to see which guess is better, we want
to “fold” our data so all points fall within one orbital “phase”. For example, if the date of
the observation is 7.3, and the period being tested is 4.2, the “phase date” of this
datapoint is 7.3 - 4.2 = 3.1 days after the beginning of the cycle. Note the 3.05 is less than
the period being tested. If the observation date is 12.2, folding it once gives 8.0. We need
to fold it again, giving a phased date of 3.8 days after the beginning of the cycle.
Mathematically, this is the same as dividing by the proposed period and using the
remainder, so:
a. Calculate the phased date by dividing the date of each observation by the period
guess, and recording theremainder. Tip: Python hasa math function for this.
Replace the division command with ‘%’.
b. M ake two scatter plots of the radial velocity (y-axis) versus the phased date
(x-axis) - i.e. two sets of points, one for each phased date column.
c. Make sure to include x and y-labelswith units, aswell as a title that specifies
which data set refers to which period.
8. Of course only one of these guesses is correct. To tell which it is, look and see which set
of data gives a nice sinusoidal curve in your plot. Congratulations, this is the
unmistakable signature of an exoplanet! Based on your data, which is the correct period
for the planet?
(c) Check for goodness of fit
9. Let’s create a sine function
a. Write a function to calculate a sine function:𝑦 = 𝐴 𝑠 𝑖𝑛(ν𝑥) , where A is the
amplitude,ν is the frequency, and x is the inputx-value. In your function, A,ν,
and x should be variables that you can change.
b. Test out your function: using A=15, ν=3, and x=10,what value do you calculate
with your function? Calculate the answer for the same values analytically
(meaning, on paper, or calculator) and check that the code gives the expected
answer. If not, you have discovered a bug! Fix it!
10.Take a look at your plot of the radial velocity (y-axis) versus the correct phase date
(x-axis).
a. What is the amplitude?
b. What is the period in units of days?
c. What is the frequency (frequency = 2π/period)?
11.Create an array of 1000 data points that start at 0 and end at 35.
a. Use your sine function to calculate a sine curve using this array, your estimated
amplitude and frequency.
12.Plot your calculated sine curve on the same plot as radial velocity (y-axis) versus the
correct phase date (x-axis). Make sure to include x and y-labelswith units, and a title.
a. Zoom into the x-limits to better see the data points.
13.Plot your calculated sine curve on the same plot as radial velocity (y-axis) versus the
julian date (x-axis). Make sure to include x and y-labelswith units, and a title. Does the
model fit the data?
a. Zoom into the x-limits to better see the data points.
14.Use your sine curve model to estimate the following:
a. What is the expected radial velocity at a reduced Julian Date of 8?
b. What is the expected radial velocity at a reduced Julian Date of 20?
(d) Calculate the mass
15.Use equations 2 and 5 from the introduction to calculate the mass of the planet.
16.Check to make sure your answer is sensible (e.g. the planet should be big - maybe even
bigger than Jupiter, but not the size of a star.)
2 - Intermediate:
his is a similar activity to the Beginner, except with less instructions, and an added step of
T
calculating the goodness of fit between the data and a sine function.
(a) Review your past work
1 . S et yourself a 10 minute timer so you don’t spend too long on this part.
2. Open the previous notebook.
a. Identify the parts you can copy directly, like loading the packages or importing
files.
b. Look for the skills used in each, like making a graph, fitting a function, and doing
calculations. You may want to take a moment to make yourself a note of which
section to find which skill.
(b) Analysis of exoplanet data
3. Make acopyof theIntermediate python notebookand download the dataset
‘51_Peg.csv’.
4. Make a scatter plot (just points, no connecting line) of radial velocity (y-axis) versus date
(x-axis). Make sure to include the errors on the y-values, x and y-labelswith units, as
well as a title.
5. Recall from the figures in the pre-lab that an exoplanet should induce a sinusoidal
variation in the radial velocity of the tar. In your plot, there should be variation visible,
but it should be difficult to tell if it is sinusoidal, or what the period is. To find the period,
astronomers fit the data points with a function, or the “theoretical expectation”, in this
case, a sine function with some period and amplitude. A fitting code is usually used in
order to find the best solution, or the closest match between the data and the theoretical
expectation from the function, and find the best-fit period. We will do asimplified version
of this type of analysis; To find the period, we willtry only a few guesses, and not explore
the entire parameter space of possible periods.
For 51 Peg, we will examine two possible “guesses” for the period of the planet: 4.2
days, and 6.7 days.
a. Calculate the phased date by dividing the date of each observation by the period
guess, and recording theremainder. (Tip: use the“%” operator in Python).
b. Make two scatter plots of the radial velocity (y-axis) versus the phased date
(x-axis) - i.e. two sets of points, one for each phased date column.
c. Make sure to include errors on the y-values, x and y-labelswith units, as well as
a title.
6. Only one of these guesses is correct. To tell which it is, look and see which set of data
gives a nice sinusoidal curve in your plot. Congratulations, this is the unmistakable
signature of an exoplanet! Based on your data, which is the correct period for the planet?
(c)Check for goodness of fit
7. Write a function to calculate a sine function:𝑦 = 𝐴 𝑠 𝑖𝑛(ν𝑥). , where A is the
amplitude,ν is the frequency, and x is the inputx-value. In your function, A,ν, and x
should be variables that you can change.
8. To quantify the goodness of the fit, calculate the goodness of fit between the data and the
sine function using a fitting function of your choice (for example, a chi^2 estimator).
a. Report the goodness of fit of each of the guesses (4.2 and 6.7 days).
b. Is the result of your code consistent with what you determined by eye?
(d) Calculate the mass
9. Use equations 2 and 5 from the introduction to calculate the mass of the planet.
10.Check to make sure your answer is sensible (e.g. the planet should be big - bigger even
than Jupiter, but not the size of a star.)
3 - Advanced:
his activity asks you to write a fitting function and explore the parameter space to find the
T
best-fit period of 51 Peg. Please check with your GSI before attempting this version.
Tip: Advanced programmers can start with the “intermediate” activity, and once the code
to calculate the goodness of fit is in place, write a function that explores the parameter
space and finds the period and amplitude that give the best fit.
1. Make a copy ofthe advanced notebookand downloadthe dataset ‘51_Peg.csv’.
2. Make a scatter plot (just points, no connecting line) of radial velocity (y-axis) versus date
(x-axis). Make sure to include the errors on the y-values, x and y-labelswith units, as
well as a title.
3. Recall from the figures in the pre-lab that an exoplanet should induce a sinusoidal
variation in the radial velocity of the tar. In your plot, there should be variation visible,
but it should be difficult to tell if it is sinusoidal, or what the period is. To find the period,
astronomers fit the data points with a function, or the “theoretical expectation”, in this
case, a sine function with some period and amplitude.
Write a function that fits the data with a sine function with parameters P (period) and A
(amplitude) and reports back the orbital parameters of the exoplanet. Look for the best-fit
period in the range from 1 to 10. You can choose which range of amplitudes to explore.
a. For each value, report the goodness of fit of a sine function to the data.
b. What period results in the best fit of the data?
Calculate the mass
4. Use equations 2 and 5 from the introduction to calculate the mass of the planet.
5. C
heck to make sure your answer is sensible (e.g. the planet should be big - bigger even
than Jupiter, but not the size of a star.)
Part 2: The Transit Method
cience objective: calculate the effective temperatureof different planets.
S
Computational objective: use one code on several setsof input, thus saving time and effort by
letting the computer do the work for you instead of calculating the same thing for different inputs
by hand.
I n this part of the lab you will examine data outputs of simulations of Kepler lightcurves of main
sequence stars and the transits of planets around the stars, to see what they can tell you about
exoplanets.
1. L oad in the data set ‘Star_transit.csv’. This data set contains the day number on which a
planet transits its host star (counting starts at 0 for each case), the peak brightness drop of
the star in percent, the mass of the star (M_sun), the radius of the star (R_sun), and the
effective temperature (K).
2. With your group, pick a star to work on.
3. What is the period of the planet?
4. Using the data available in the csv file, calculate (in python) the semi-major orbit of the
planet.Make sure you are paying attention to units!
5. The drop in the star’s brightness as the planet passes in front of it is given by the ratio of
the planet’s area to the star’s area. Using the star’s radius and the peak drop in brightness
from the lightcurve, estimate the radius of the transiting exoplanet in cm
10
(𝑅
𝑠𝑢𝑛 = 6 𝑐𝑚).
. 96 × 10
8
a. What is the radius of the planet in Earth radii (𝑅𝐸𝑎𝑟𝑡ℎ = 6 𝑐𝑚)?
. 37 × 10
9
b. What is the radius of the planet in Jupiter radii (𝑅𝐽𝑢𝑝 = 6. 99 × 10
𝑐𝑚)?
6. W rite a function that will calculate the effective temperature of your exoplanet using
equation 10 from the introduction.
a. Determine the temperature of your planet.
b. Check your answers to make sure they make sense (e.g. a planet shouldn’t be
hotter than a star, or have Pluto-like temperatures at Mercury-like distances.)
7. In order for an exoplanet to be habitable, it needs to have liquid water. Water exists in the
liquid state between 273 and 373 K. Is your exoplanet in the right range to be considered
as “candidate” life-supporting planets? If not, is it too hot or too cold?
How to submit your lab
ou should be able to save the notebook as a pdf under the print option. Upload the pdf to
Y
Gradescope with the following outline:
1. Loading the libraries and data file (2 pts)
2. Scatter plot of data (5 pts)
3. Phased date calculation (3 pts)
4. Phased date plots (5 pts)
5. Goodness test [beginner steps 9 - 14, intermediate step 7, advanced step 3a] (5 pts)
6. Planet mass (5 pt)
7. Loading transit data and picking a star (2 pt)
8. determining the period and semi-major axis (4 pt)
9. Radii (4 pt)
10.Temperature calculation (5 pt)
11.Habitability (5 pt)