Lookback at our 2021 summer internships

2021 iREx interns (from left to right, from top to bottom): Samantha Lambier, Leslie Moranta, Kim Morel, Luc Bazinet, Marylou Fournier-Tondreau, Michael Matesic, Amalia Karalis, Alexandrine L’Heureux, Nicholas Swidinsky, Jacob Kennedy, Dhvani Doshi, Maigan Devries, Lan Xi Zhu, Patrick Horlaville, Samson Mercier, Sarah Thiele, Michael Poon, Alexander Gass, Maude Larivière, Xueying Li, Thomas Villeneuve.

Autumn is approaching, and it is the end of the summer internships at the Institute. For our 21 interns, it is time to go back to their undergraduate studies, or to embark on the adventure of graduate studies. Before leaving us, they agreed to answer a few questions.

Learn more about Luc Bazinet, Maigan Devries, Dhvani Doshi, Marylou Fournier-Tondreau, Alexander Gass, Patrick Horlaville, Amalia Karalis, Jacob Kennedy, Alexandrine L’Heureux, Samantha Lambier, Maude Larivière, Xueying Li, Michael Matesic, Samson Mercier, Leslie MorantaKim Morel, Michael PoonNicholas SwidinskySarah Thiele, Thomas Villeneuve, Lan Xi Zhu in these short interviews, which touch on their project and their experience at iREx.

Luc Bazinet

Intern from the University of Ottawa who worked with Björn Benneke at the Université de Montréal

What was the topic of your internship?

I improved the code for SCARLET, a program that creates atmospheric models for exoplanets. SCARLET uses a table to find how much of each molecule is in the atmosphere of the model. My job was to create a bigger table that would create models for a bigger range of exoplanets, notably hot Jupiters. Furthermore, the new table has more than double the number of chemical species which included ions which were not considered before.

What was interesting about it?

Having discrepancies between the model’s atmosphere and the real exoplanet’s atmosphere is exciting. It means that there is a chemical process present on the exoplanet. This process may be caused by the star, by the planet’s surface or, more exciting still, by life. The implementation of the new table might be an important step in finding life outside our Solar System!

What did you discover? What was your most important result?

With my improvement to SCARLET, we created a model for the hot Jupiter WASP-76 b. With this model and transits from the telescope MAROON-X, we were able to detect heavy metals on this exoplanet which were predicted in the literature. In the future, the improved table will be used to create even more exoplanet atmospheres that were impossible to model before.

What did you learn this summer?

I learned a lot about exoplanets in general. I learned how different kinds of light help us detect as well as characterise exoplanets. I learned how chemistry and chemical processes determine the atmospheric composition and thus dictates our detections here on Earth. Also, I acquired a lot of skills in programming, which will certainly be useful in my future.

What was the biggest challenge during your internship?

Working online was difficult. Help was harder to access than if we were in person. If I had a desk with my co-workers, I could ask questions with less effort and I would have been more motivated. Also, I would have liked to move to Montréal to explore this city and to see my supervisor and co-workers in person.

What did you like most about your internship?

I loved learning about a subject that has always interested me. I’ve always loved astronomy, so learning about exoplanets in such detail was a dream come true. I liked the presentations that would not only talk about exoplanetary atmospheres, but other subjects in astronomy that I know less about, for example protoplanetary discs and exoplanet evolution. Furthermore, being around some of the brightest experts in exoplanetary research was truly incredible.

Maigan Devries

Intern from the University of Northern British Columbia who worked with Jason Rowe at Bishop’s University

What was the topic of your internship?

Transits of Kepler Objects of Interest (KOI) can be used to determine values for exoplanet parameters such as orbital period. Markov Chain Monte Carlo (MCMC) has been used to update these values, and I worked on validating the new results. To validate these results, I compared the new and old parameter values and visually inspected transits when there was a significant difference between the current and past values. I also looked at the auto-correlation of the MCMC chains to ensure adequate chain length.

What was interesting about it?

It was interesting to see the variations of different transits. I expected them all to look quite similar, but there was some variation. It was also interesting to see the differences between the new results and prior results – some results are in close agreement with prior values, and other results have varied significantly.

What did you discover? What was your most important result?

The most important result was determining the final KOI values. Along the way, I discovered how to determine whether the KOIs had satisfactory results and how to ‘fix’ them in each case if needed.

What did you learn this summer?

I learned how to identify KOIs with transits that were in poor agreement with the data and how to re-run them using Compute-Canada. I also learned how to implement auto-correlation for MCMC chains and how to use it to determine if the MCMC chains are long enough.

What was the biggest challenge during your internship?

My biggest challenge was figuring out the best way to use auto-correlation on the MCMC chains and how to best restart the chains when needed.

What did you like most about your internship?

I really enjoyed working with Dr. Rowe and his research team again this summer. It is a great way to learn new things – both related to my project and not – and overall the group has a great dynamic.

Dhvani Doshi

Trottier Intern from the University of Waterloo who worked with Nicolas Cowan at McGill University

What was the topic of your internship?

The goal of my internship was to see how Earth-like clouds would affect our observations of exoplanetary atmospheres. There have been multiple studies that have shown that clouds would make it hard to understand these atmospheres so we wanted to see if that would also be the case with Earth-like clouds.

What was interesting about it?

It was interesting to learn how, just by having clouds in an atmosphere, it would completely change our idea of that planet’s atmosphere. It could change how we identify molecules, the planet’s overall radius, and much more. It was also fun to see how well our current technological capabilities would be at understanding an Earth-like planet.

What did you discover? What was your most important result?

My most important result was finding out that if an exoplanet had an atmosphere identical to Earth and had Earth-like clouds, we likely wouldn’t be able to detect those clouds for certain star systems. This was a very optimistic result, as before we had feared that clouds of any kind would massively impact how much we could learn from other planets.

What did you learn this summer?

I learned the importance of statistics this summer and how it can help us dictate the significance of our observations. Another thing I was able to learn was the process of preparing a manuscript as well as the importance of a thorough literature review to understand where your research stands in the community.

What was the biggest challenge during your internship?

The most difficult thing this summer was developing a methodology of handling my data that I was confident in. It was important for me to verify the accuracy of my methodology with different types of testing and toy models. However, it was hard to gauge how I could create these tests in the first place.

What did you like most about your internship?

I enjoyed working at the Institute, because of the environment it fostered. I loved being in an environment with so many researchers that were extremely motivated to discover more about exoplanets.

Marylou Fournier-Tondreau

Trottier Intern from the Université de Montréal who worked with Björn Benneke at the Université de Montréal

What was the topic of your internship?

I worked on a project to model reflected light during secondary eclipses more accurately. The DISORT program is one of the best ways to calculate a planet’s scattered radiation at the moment, but the computation is time-consuming. I used unsupervised learning, more specifically a k-means clustering algorithm, to reproduce the reflected light spectrum with less wavelength data.

What was interesting about it?

My project allowed me to consolidate and broaden my knowledge of planetary spectra. During my internship, I was working with SCARLET, a framework for analysing and modelling atmospheres. It was interesting to learn how the theory surrounding exoplanet atmospheres is implemented in code.

What did you discover? What was your most important result?

Unsupervised learning with data clustering can be used to drastically accelerate the modelling of planetary spectra. We can reproduce virtually the same reflected light spectrum with around 3% of the wavelength data, that is 34 times fewer calculations for the DISORT program. The results are promising and the method could be refined to further decrease the number of clusters and the computation time.

What did you learn this summer?

I had the opportunity to work with data clustering for my project. During my studies, I never had the chance to take a course in machine learning, but I knew that it was widely used in astrophysics. I was really excited to learn about unsupervised learning and apply it to my project. The potential in atmosphere retrieval is definitely huge.

What was the biggest challenge during your internship?

During my studies, I had only developed basic skills in coding, so I would say working programming was challenging. For my project, I had to convert the Python wrapper of DISORT to Python 3. At first, I felt like an impostor, but that challenge definitely allowed me to learn a lot.

What did you like most about your internship?

It was my first full-time experience with research in astrophysics, which has been a dream for many years. It was amazing to be a part of research in exoplanetary atmospheres within Prof. Benneke’s group. I had the chance to work with a very enthusiastic and insightful supervisor. I am also grateful for all the support I had from the graduate students.

Alexander Gass

Intern from McGill University who worked with Nicolas Cowan at McGill University

What was the topic of your internship?

My internship was focused on doing a re-analysis of the lava planet 55 Cancri e. The goal of the analysis was to acquire a map of the planets temperature, mainly to verify a previously published map that had some unexpected features. To do the analysis I used an open-source pipeline (Spitzer Phase Curve Analysis) designed by two iREx graduate students, in order to take data from the Spitzer Space Telescope and to create orbital phase curves, from which we could obtain a temperature map.

What was interesting about it?

Besides the fact that studying a planet far outside of our Solar System is very cool in itself, it was interesting to get a hands-on look at the software tools used to process telescope images. It was also fun to explore previous research and try to think of creative ways to make our analysis different but still robust.

What did you discover? What was your most important result?

I discovered that data reduction is not as easy as it may seem! There is a LOT to correct for in telescope data – mostly related to the telescope hardware. Although we have not yet obtained a temperature map, we are in the final stages of our reduction. Our most important realisation was that detector systematics have a huge impact on our data.

What did you learn this summer?

I’m not even sure where to begin! I became very familiar with the Python programming language libraries and tools used in exoplanet studies, and got comfortable with modifying existing Python packages (mainly our data pipeline, SPCA). Along with the coding and data analysis skills I developed, I learned a tonne about how we observe exoplanets, and how we can learn about the temperature of a planet by simply looking at how bright it is throughout its orbit.

What was the biggest challenge during your internship?

The biggest challenge by far was posed by our observation data for 55 Cancri e. Because the 55 Cancri system is so bright, the Spitzer telescope wasn’t able to continually observe it over a complete orbit. This meant our data was compiled from eight separate observations done months apart. Our data pipeline was not built to handle unusual data like this, and so our biggest challenge was figuring out how to modify our pipeline to get an accurate model for this dataset.

What did you like most about your internship?

What I enjoyed most was brainstorming and problem solving with other members of the research group when I hit a roadblock. It was a great way to bond, as well as learn more about other areas in astrophysics besides the data analysis aspect. There was also a very satisfying and clear sense of progression that came from working on the same project for an entire summer.

Patrick Horlaville

Intern from McGill University who worked with Björn Benneke at the Université de Montréal

What was the topic of your internship?

Finding exoplanets’ occurrence rates surrounding M stars (the least massive) with TESS Space Telescope data.

What was interesting about it?

Exoplanets occurrence rate projects give an insight about actual populations of exoplanets surrounding particular types of stars, beyond the data given by the telescope.

What did you discover? What was your most important result?

I was able to replicate results from a newly-graduated student, Merrin Peterson, who initiated the planets’ occurrence rates project with TESS data for M stars during her Master’s. It took me a lot of effort getting through it, but it worked!

What did you learn this summer?

I learned a good deal about exoplanetary science, bash script troubleshooting (which was very time consuming), and more particularly about Merrin Peterson’s exoplanets occurrence rates project. Through that project, I deepened my understanding of the Python programming language and exoplanetary sciences applied to this specific context.

What was the biggest challenge during your internship?

Setting my environment. I have spent most of my time trying to create an environment on my computer from which I would be able to investigate scripts (mainly Python) and do science. The first part was long and annoying, but the second was fascinating.

What did you like most about your internship?

The collaboration between students and professors, the discussions with other interns, investigating intriguing concepts which become clearer as I take the time to look into them and the outreach opportunities that were given to us!

Amalia Karalis

Intern from McGill University who worked with Eve Lee at McGill University

What was the topic of your internship?

The radius valley is a gap in the radius distribution of exoplanets found by Kepler Space Telescope, between super-Earths and sub-Neptunes. My project used planet formation models to see if this gap is created from formation (primordially), rather than from mass loss. More specifically, we varied a subset of parameters in our model to see which would give the best fit when compared to observational data.

What was interesting about it?

It was interesting to investigate a different theory than the one that is currently accepted for the existence of this valley. It was also cool to learn a bit more about exoplanet formation.

What did you discover? What was your most important result?

We are currently still in the process of running some of the models, but we do see that the model distributions do better when we have a top-heavy core-mass distribution and later times for core formation.

What did you learn this summer?

I developed my coding skills this summer since my project focused a lot on statistical analysis. I also got to learn about exoplanets and their formation process, as well as the leading theories explaining the existence of the radius valley.

What was the biggest challenge during your internship?

Working from home was definitely a challenge. It was very isolating to be alone at my desk all day. Weekly Zoom meetings and the intern catch ups definitely helped, but I’m excited to be going back in person in the Fall.

What did you like most about your internship?

Finding a community of people with interests similar to mine! It was great to get to interact with other interns from all over the country who were interested in the same field as I am. I also loved being able to learn more about exoplanets and coding, while also doing a small part in contributing to the field.

Jacob Kennedy

Intern from McGill University who worked with Björn Benneke at the Université de Montréal

What was the topic of your internship?

My internship primarily focused on studying how clouds will appear in exoplanet transmission spectra with the James Webb Space Telescope (JWST), and whether or not it will be possible to characterise cloud composition and structure using this new data. To do this, I simulated a number of different cloud scenarios with variable composition, vertical extent, altitude, and particle sizes using the atmospheric retrieval and modelling framework SCARLET, that was developed by Prof. Björn Benneke. From there, we could compare the theoretical spectra with real data and simulated JWST data to get an idea of what clouds will look like in the JWST spectrum and what we could potentially constrain moving forward.

What was interesting about it?

Working in SCARLET allowed me to interact with the complete exoplanet study pipeline, from theoretical modelling to retrieval and instrument simulations. Understanding how the interplay of different atmospheric processes come together to produce the final transmission spectrum of a planet was really interesting. Further, having the opportunity to study first-hand how JWST will contribute to the future of exoplanet science was incredibly exciting.

What did you discover? What was your most important result?

Via an extensive exploration of different cloud scenarios, we were able determine what types of clouds are likely to be visible with JWST data on certain exoplanets. More generally, using the SCARLET package we were able to develop a systematic framework to infer the presence and properties of clouds once JWST data is available.

What did you learn this summer?

As expected, I learned quite a lot about the exoplanet science field as a whole, although more generally, I learned a great deal about the research pipeline and group collaboration. Drawing on the experience/expertise of the graduate students in the group really helped me develop as a scientist.

What was the biggest challenge during your internship?

Remembering to remind myself of how the immediate task tied into the goal of the greater project was at times difficult. The first couple months I often felt as if the final goal was out of reach, although with time as the pieces fell into place, it became clear how those initial steps fed into the grander design of the project.

What did you like most about your internship?

I really enjoyed being able to work in a group environment with weekly and sometimes bi-weekly meetings. This allowed me to do my own independent work, while still consistently receiving feedback and direction from both the graduate students and my supervisor.

Alexandrine L’Heureux

Trottier Intern from the Université de Montréal who worked with René Doyon at the Université de Montréal

What was the topic of your internship?

The topic of my internship was to analyse radial velocity (RV) data from SPIRou, an infrared spectropolarimeter, to detect planetary signals from the TRAPPIST-1 system. This system is well-known for having seven transiting Earth-size exoplanets with at least three in the habitable zone. All were detected using transit-timing variations (TTV), but no RV detection has been made before.

What was interesting about it?

When collecting RV data, it is the gravitational effect of the planets on the star that is measured. Because of the large number of exoplanets, the RV signal from TRAPPIST-1 is quite messy and it is difficult to untangle the effect of each planet. Moreover, the stellar activity has an RV signature that needs to be subtracted from the data. I found this puzzle very interesting to work through!

What did you discover? What was your most important result?

I managed to detect the RV signal from TRAPPIST-1b, the planet closest to the star! To achieve this result, I analysed the signal with a joint model: a Keplerian model for the planets (using mass ratios from previous TTV detections) and a Gaussian Process (GP) model for the stellar activity (trained on data sets proxy to the activity). A GP is a good way to model correlated noise such as stellar activity.

What did you learn this summer?

I learned a lot about GP modeling, a method I hadn’t even heard of before, and its usefulness! I also had the chance to attend some talks and meetings that gave me a better idea of what it’s like to do research on exoplanets.

What was the biggest challenge during your internship?

The biggest challenge I faced was to wait. During the first months of my internship, my codes were taking close to an hour to run which was a lot of waiting before seeing if my results had improved. Thankfully, I learned how to use multiprocessing, cutting the run time significantly!

What did you like most about your internship?

I had a really great experience working with my supervisor, Prof. René Doyon, and collaborators, Olivia Lim and Étienne Artigau. Our frequent meetings kept my work structured and I appreciate how they trusted me to present the result of our work to the SPIRou team.

Samantha Lambier

Trottier Intern from Western University who worked with Jonathan Gagné at the Université de Montréal and Planétarium Rio Tinto Alcan d’Espace pour la vie

What was the topic of your internship?

The topic of my internship was determining the ages of young stars (under 1 billion years) by writing a code to calculate the lithium present in them. I was particularly focused on stars in Young Stellar Associations (YSAs), which are groups of stars that formed relatively recently from the same molecular cloud.

What was interesting about it?

This project is interesting because YSAs are excellent environments for observing exoplanet formation due to their young ages. Ultimately, determining star ages with lithium measurements can aid in choosing which stars should be observed in order to have the greatest probability of finding an extrasolar system in the process of being formed.

What did you discover? What was your most important result?

My most important result was successfully developing a code to calculate the amount of lithium from the spectrum of a star. Going forward, this code can now be used by others!

What did you learn this summer?

This summer, I really learned a lot more about coding in the Python programming language. I had taken a computer simulations course last year in school, and I thought I learned a lot from it, but actually applying what I learned in that course to a research project is what solidified my coding knowledge.

What was the biggest challenge during your internship?

In order to try and avoid essentially “reinventing the wheel”, I had begun my project trying to use a pre-existing program to measure lithium. This involved me spending a long time learning how to code directly in the terminal to install it, then even longer to get it to work the way I wanted it to. In the end, it was way easier to just write my own code, and I scrapped it!

What did you like most about your internship?

It’s hard to pick just one thing that I liked best, I really enjoyed pretty much everything about the internship (minus the fact that it was remote)! I especially liked the coding (particularly when it worked), the meetings with Dr. Gagné’s team, and the weekly catch-up meetings with the other interns. It was nice to see how everyone’s summer was going even though we couldn’t be all together.

Maude Larivière

Trottier Intern from McGill University who worked with David Lafrenière at the Université de Montréal

What was the topic of your internship?

I worked on developing a new method for depolluting the lightcurves of transiting exoplanets. Basically, when we observe the light of a star to detect planets, the signal is altered by external factor such as the atmosphere or the camera itself. I worked on trying to build a computer program to remove these factors to be able to detect the transit signal of exoplanetary candidate and possibly help with the confirmation of their status as exoplanets!

What was interesting about it?

I think the completeness of this project was really interesting. I started off by developping this code dealing with actual real-life stars and then got to apply it to current exoplanetary candidate. Seeing what I developped in action and even having detections worthy of submission to a follow-up group was gratifying. I got to experience different aspects of research and it was definitely an amazing opportunity.

What did you discover? What was your most important result?

Well, having my code working and useful was an important result for this project, since the key question was to determine if the method I used (PCA) was valuable for lightcurve clean-up. On top of that, we have analysed over 30 observations and are ready to submit about 9 observations with interesting transits to the TESS Space Telescope follow-up group!

What did you learn this summer?

I definitely learned more about programming. I also learned about exoplanets in general, but I would say that one of the most important things I learned was that research conclusions are not straightforward at all. It can be tricky to determine when you have the best possible result, and I definitely learned about the intricacy of analysis.

What was the biggest challenge during your internship?

I would say time management / keeping up with a schedule. It’s not like school where you have classes and deadlines everywhere or even friends to motivate you. Getting to work and staying motivated from home with all the distractions was definitely a challenge!

What did you like most about your internship?

I really enjoyed the community spirit of the iREx members. It was great to be part of such a dynamic team. My supervisors were so nice and encouraging, I got to pursue some outreach opportunities as well as learn about a field that is fascinating. Overall amazing experience!

Xueying Li

Intern from McGill University who worked with Nicolas Cowan at McGill University

What was the topic of your internship?

My internship was about using the Slepian functions, which have the property of being orthonormal both on a full sphere and a chosen domain, to solve the problem that eclipse mapping provides no information about the far side of a planet.

What was interesting about it?

Spherical harmonics are usually used as a basis to describe surface maps, but we adopted Slepian functions as an alternative. These functions have the desired property of being optimally concentrated in the region they are defined. This property should reduce the degeneracies caused by the incomplete observation information of eclipse mapping.

What did you discover? What was your most important result?

We discovered that Slepian functions are able to recover surface maps and lightcurves of planets well, which proves they form a consistent and complete basis just as spherical harmonics do. Also, previous studies say that the atmospheric structure of some hot Jupiters are influenced by their magnetic field strength, and we found Slepian functions can be used to distinguish models of a planet with or without a magnetic field.

What did you learn this summer?

I learned the basics of exoplanet mapping techniques, some about data analysis and a lot programming. More importantly, now I know what a real research is like. I learned how to efficiently organise my work and how to present my work in a clear way to different audiences.

What was the biggest challenge during your internship?

For me, the biggest challenge was programming. I did not have much experience on organising code and debugging, so sometimes I had to spend hours solving errors caused by some typos.

What did you like most about your internship?

I liked our weekly group meetings where we could share our research updates and set new goals. These meetings were useful to keep me on the right track and focus on the most important problem to be solved. I also liked the weekly iREx Café where we can hear about other interesting research in astrophysics.

Michael Matesic

Trottier Intern from the University of Waterloo who worked with Jason Rowe at Bishop’s University

What was the topic of your internship?

Our objective was to develop software that will determine the frequency of Earth analogues, called ηE (“eta-Earth”). These planets are habitable zone Earth-sized planets orbiting Sun-like stars. We aimed to accomplish this by using statistical methods to analyse and model transit events, where a planet crosses in front of its star, thereby dimming the light we observe. Our method allows us to confidently rule out false-positives and provide the most likely planetary candidates to the James Webb or Hubble space telescopes for follow-up observations.

What was interesting about it?

We currently have zero confirmed Earth analogues, but we do have a short list of promising candidates. Obviously, the prospect of finding Earth 2.0 and the implications/possibilities that come with such a discovery are truly exciting. Best-case scenario, we find (intelligent) life elsewhere in the Universe and worst-case scenario, we have a planet that could be colonised in the future with relative ease.

What did you discover? What was your most important result?

The project is nearing completion, so you can expect answers to our main objective within a couple of months. At this point in time, the most important results would be that we have the bulk of the analysis software tested and working. We will be performing our analysis on the eta-Earth candidate list very soon, so keep an eye out.

What did you learn this summer?

In addition to further growing my Python programming language experience and picking up some new coding tricks from Dr. Jason Rowe, I had the opportunity to learn much more about transit modelling and was able to develop a working intuition for the two statistical techniques upon which our project was based, these being Gaussian processes and nested sampling.

What was the biggest challenge during your internship?

The beginning was probably the most challenging, as I had to bring myself up to speed on a lot of information in a short amount of time. This includes understanding the theory and inner workings of the aforementioned transit modelling, Gaussian processes, and nested sampling. Following this, the rest of the project flowed smoothly.

What did you like most about your internship?

The entire experience was fantastic. It was great making new connections within my field while working on a project that I am very passionate about, but what I appreciated most was the willingness and availability on Dr. Rowe’s part to offer guidance throughout the summer. Not only would he readily answer any conceptual questions that I had, but would happily go into deeper discussion about them. All in all, I couldn’t have asked for a better supervisor.

Samson Mercier

Intern from McGill University who worked with Nicolas Cowan at McGill University

What was the topic of your internship?

My internship was about exoplanets and our ability to detect and characterise their respective atmospheres, as a condition (in our current understanding) to the emergence of life. More than 4000 exoplanets have been discovered thus far, with 55 Cancri e being one of the most well-known. This summer, my colleague Alex Gass and I worked on the re-analysis of the orbital phase curve of 55 Cancri e. To do so, we used datasets (one per observational period) collected by the Spitzer Space Telescope for a study done by Demory et al. in Nature in 2016, and the SPCA (Spitzer Phase Curve Analysis) processing pipeline to extract flux values from raw images and correct for the telescope’s detector systematics. Detector systematics help quantify the uncertainties of our measurements due to the differences between the detector’s response in simulation and in reality.

What was interesting about it?

55 Cancri e is an oddity in the field of exoplanets due to its intriguing thermal properties and an orbit close to its star, making it hard to observe. Re-analysing the data associated with this planet allows us to verify the robustness of Demory et al.’s results. If our work confirms their conclusions, it would mean that the planet has a thick enough atmosphere to shift the hottest point by approximately 400km to the east. Otherwise, as a rebuttal to Demory et al., we would conclude on the absence of an atmosphere on 55 Cancri e, diminishing the interest for this exoplanet among the scientific community.

What did you discover? What was your most important result?

By studying 55 Cancri e with the transit photometry method, I discovered that a typical model contains four terms : 1. a model describing the detector systematics, 2. a sinusoidal model for the light emitted by the planet and the star, 3. a transit model, when the planet passes in front of the star, and 4. an eclipse model, for when the planet passes behind the star. So far, we have used a BLISS (BiLinearly-Interpolated Subpixel Sensitivity) mapping model for the first term, which we fitted to a pooled dataset (concatenating the 8 observational periods). The residuals for this model revealed limits in capturing the astrophysical signal and suggested a more tailored approach with one model per observational period, which we will implement in the upcoming months.

What did you learn this summer?

Seeing as SPCA was written in the Python programming language, I spent a lot of my time learning how to write code and debug in this language. This was very useful as I needed to be able to understand and modify the processing pipeline to work with the 55 Cancri e data. I also had the opportunity to attend the iREx Cafés, which allowed me to have a broader knowledge of the topics covered by the exoplanet community, and I got to discover what projects the other members of iREx were working on. Finally, I had the privilege to be in a welcoming research group that organised fun social events while still helping me develop my professional skills such as learning how to synthesise my findings in preparation of my weekly meetings.

What was the biggest challenge during your internship?

Since I had never worked in the field of exoplanets before, I believe that applying my theoretical knowledge from my academic courses to a real case study was the biggest challenge I faced. Furthermore, in the project itself, we spent a lot of time figuring out which method was best for the data processing and how to navigate a remote server in order to deal with large amount of data (around 70 GB of raw images).

What did you like most about your internship?

I had a lot of fun working with Prof. Cowan’s research group, as the weekly group meetings and one-on-one meetings were a stimulating and engaging environment. Through our supervisors’ guidance we were able to break down our project into multiple steps which allowed us to progressively get closer to our goal. Finally, the picnics were a great way for me to get to know the other researchers in the team and what projects they were working on.

Leslie Moranta

Intern from the Université de Montréal who worked with Jonathan Gagné at the Université de Montréal and Planétarium Rio Tinto Alcan d’Espace pour la vie

What was the topic of your internship?

My internship focused on the detection and characterisation of young star associations in the solar neighborhood. The methods used in the literature have faced projection problems that have prevented finding associations within a 100 parsecs radius of our Solar System, so we developed an algorithm that fixes this issue.

What was interesting about it?

The young stars of these associations are ideal candidates for detecting exoplanets, so finding a way to locate associations near us can greatly contribute to finding exoplanets and better understand the Universe that surrounds us.

What did you discover? What was your most important result?

It has been possible to find new clusters of stars that are not or very poorly documented. We are still in the analysis phase, but the first results are very promising. In addition, we noticed that our algorithm also finds already known associations, which is very reassuring and confirms that our algorithm works well.

What did you learn this summer?

This internship was a good introduction to the research process in astrophysics, whether it was mathematical problem solving, data analysis or writing and editing scientific papers. Moreover, this internship confirmed my desire to pursue my graduate studies in exoplanet research at the Université of Montréal.

What was the biggest challenge during your internship?

My biggest challenge was to start the project from scratch after a year of work to switch to a new method. Obviously, all the work done was not lost because I had the opportunity to learn a lot about several aspects of the project, but it was still demotivating at the moment. However, it did not last long when we saw the encouraging results of our new algorithm.

What did you like most about your internship?

My internship as a whole was a very pleasant experience. It was a good opportunity to learn about several branches of astrophysics and to make rewarding encounters. I also had the good fortune to be well-guided by my research supervisor, which greatly contributed to the smooth running of my internship. I will always be grateful for this incredible research experience within iREx.

Kim Morel

Intern from the Université de Montréal who worked with David Lafrenière at the Université de Montréal

What was the topic of your internship?

Validation of the Single-Object Slitless Spectroscopy (SOSS) mode simulator for the James Webb Space Telescope (JWST) NIRISS (Near-Infrared Imager and Slitless Spectrograph) instrument.

What was interesting about it?

I got acquainted with the format of spectral traces obtained with the SOSS mode on NIRISS. I also allowed me to understand the importance of simulating data before a launch in order to be able to work with the right tools once real data is available.

What did you discover? What was your most important result?

I found many problems that allowed us to fix some issues in the simulator and upgrade performances of a spectrum extraction. Nonetheless, the SOSS simulator is able to retrieve transit spectra as expected.

What did you learn this summer?

I understood how the SOSS mode on JWST’s NIRISS instrument works, as well as the complexity behind the photon collection of an optical instrument on a telescope leading to the production of a stellar spectrum. I also learned a lot about data analysis.

What was the biggest challenge during your internship?

We encountered more problems than expected, and some of them required many weeks of research to find the solution. Actually, a few of them still need to be investigated. It is not always easy to find exactly where the problem comes from.

What did you like most about your internship?

I got the chance to work on a telescope whose launch is long-awaited (JWST), which is really exciting. I also enjoyed racking my brain to find solutions for problems, and I got plenty of that this summer!

Michael Poon

Trottier Intern from the University of Toronto who worked with Eve Lee at McGill University

What was the topic of your internship?

HR 8799 is a Canadian-discovered planetary system of four Jupiter-like planets. This spectacular system is over 100 light years away and was among the first to be detected by direct imaging. The difficulty of this discovery can be compared to spotting fireflies (dim planets) next to a spotlight (bright star). This summer, I studied how planets in the HR 8799 system spin and wobble due to the gravity of their star and each other. This is similar to how Earth is tilted on its axis causing seasons as we go around the Sun. The goal of the project is to create a 3D picture of each planet’s orbit and spin using state of the art observations.

What was interesting about it?

If you’ve ever gone stargazing, you will see planets like Saturn and Jupiter move along the night sky. Using a telescope, we can trace out the three-dimensional orbits of these planets in our Solar system. However, for planets outside our Solar system (called exoplanets), this is cutting-edge science. It requires a combination of very precise observations and theory to understand the three-dimensional evolution of these faraway systems. It’s exciting to work on theoretical predictions of planet dynamics that were once thought as impossible to observe.

What did you discover? What was your most important result?

I discovered that a certain theory named ‘secular dynamics’ is able to explain the dynamical details of planets in the HR 8799 system. Soon, we will be able to test the predictions of this theory as new observations are made in the near future.

What did you learn this summer?

I learned how to do theoretical astrophysics over Zoom. It’s quite tricky and I got better at writing equations with my mouse, but I wouldn’t recommend it. I learned that I really miss working in person with others in front of a whiteboard!

What was the biggest challenge during your internship?

The biggest challenge was to install Linux as dual-boot on my Windows system in order to run a certain code. It was a little nerve-wracking, but it worked out all right!

What did you like most about your internship?

It’s amazing to work at iREx where everyone is so excited about exoplanets. The environment was very friendly and I really enjoyed working with my supervisors Prof. Eve Lee and Dr. JJ Zanazzi.

Nicholas Swidinsky

Intern from the University of Lethbridge who worked with Jason Rowe at Bishop’s University

What was the topic of your internship?

During my internship I worked on the POET micro-satellite currently being designed by Dr. Jason Rowe and his team. Specifically I worked on the Exposure Time Calculator (ETC) which is a vital part of the mission. The ETC simulates photon count rates, which can then be used to simulate data results, and calculate associated noise with these simulations.

What was interesting about it?

There were a few things that were really interesting about this project. First, it was very interesting being part of the POET team, I got to meet many professionals as well as learn from them throughout the project. I was also interested in learning how to do data simulations, since I had never done anything similar before.

What did you discover? What was your most important result?

Unlike some of the other interns this summer, the nature of my work didn’t lead to any huge discoveries in astrophysics. Instead, I finalised versions 1.0 and 1.5 of the ETC, as well as used my ETC in a few other simulations. I also documented all of my work in technical reports for POET.

What did you learn this summer?

This summer I mostly improved upon previous skills that I learned throughout my undergrad as well as previous internships. All of my work on the ETC was coded in Python, which is a coding language that I had learned during my undergrad, and have used during my previous internships. I did learn new skills within Python, such as how to code simulations, which is something I have never done previously.

What was the biggest challenge during your internship?

The biggest challenge for me was having to work remotely for the entire summer. I find it very easy to get distracted while working from home, which makes it hard to have a productive work day. I found it especially difficult to stay focused when I encountered a difficulty in my work since, if I asked for help, I would have to wait for a response which could take a long time.

What did you like most about your internship?

I enjoyed getting to meet new people, and seeing all of the wonderful research that is being done on exoplanets. Within my lab group, I got to experience more of the research process by listening to some of the grad students progress, as well as the other undergrads I was working with. Additionally every Monday there was the iREx Café which gave me the opportunity to see different research that is being conducted outside of my lab group.

Sarah Thiele

Trottier Intern from the University of British Columbia who worked with Andrew Cumming McGill University

What was the topic of your internship?

When a planet is embedded in a star’s protoplanetary disc of gas and dust, there is a period of time when the planet slowly accretes disc material and increases in mass. My internship focused on “atmospheric recycling,” when a portion of the planet’s gaseous envelope gets continuously replaced by a flow of disc material, rather than remaining bound to the planet. This process hinders accretion and thus the mass that the planet can achieve. My project involved developing numerical methods to solve for these flow patterns around a planet.

What was interesting about it?

The most dominant type of exoplanets are called super-Earths, which have radii between one and fours times the size of Earth. Although they dominate the exoplanet population, their formation is still not well understood, as models predict that they should accrete enough mass to become gas giants. The restricting of mass accretion through atmospheric recycling may explain their formation channels. The better we can understand this process, the more we can learn about how these planets form!

What did you discover? What was your most important result?

We found that the simplicity of the numerical methods needed to solve for the flow patterns was highly dependent on the assumptions made about the relationship between state variables of the disc flow like pressure, density, and temperature. If we assume temperature to be constant for example, we can model the disc flow in different regions around the plane for various pressure-density relations. However, to better study the depth to which atmospheric recycling can penetrate into the planet’s envelope, we are working towards incorporating variations in all three flow variables.

What did you learn this summer?

Firstly, I gained an understanding of the basics in exoplanet science. This was great as I am new to this field. Secondly, I got to learn about multiple numerical methods that are used to solve a variety of physics problems. These techniques are great to have in your computational toolbox since they are so widely applicable! Lastly, the physics background needed for this project had an emphasis on fluid dynamics. I got to DIVE into the mathematics and concepts used to describe fluid flows… pun intended.

What was the biggest challenge during your internship?

I found the biggest challenge to be time: in short-term internships like this one, devoting long periods of time to problem-solving can take away from the potential progress of the project. There were multiple instances during the summer where I became stuck trying to get various pieces of code to work properly, which can be very time consuming. However, it was an educational process – learning to be patient and tolerate setbacks is an essential skill in the research world!

What did you like most about your internship?

While I greatly enjoyed learning the physics background for this project, my favourite part was the people I got to meet. Although the internship was done virtually, the student interns I got the chance to talk to online were all vibrant and fun people. The group of graduate students and my supervisor, Andrew Cumming, were lovely to work with. Prof. Cumming explains concepts extremely well, which made for a rewarding learning curve during my internship. I had a lot of fun this summer!

Thomas Villeneuve

Intern from McGill University who worked with Nicolas Cowan at McGill University

What was the topic of your internship?

The topic of my internship was exoplanet eclipse mapping. During ingress and egress, an exoplanet becomes partially obscured by its host star which allows us to retrieve thermal flux light curves originating from discrete sections of the exoplanet. With this in mind we suspect that the eclipse light curves may be the best source from which to retrieve detailed thermal emission maps. My research focused on deriving and validating an alternative basis, the Slepian functions on a hemisphere, to use for eclipse mapping.

What was interesting about it?

The notion that we can create maps for the surfaces of exoplanets that may be hundreds of lightyears away is pretty ridiculous. When we point our telescopes towards these planets, at best we see points of light, with most of that light originating from the planet’s host star. The planets themselves only contribute tenths or hundredths of a percentage of the total thermal flux, yet we can measure to great accuracy any change in the planetary light curves.

What did you discover? What was your most important result?

We found that, for low-band limits at least, the Slepian functions serve as a better basis for eclipse mapping than the spherical harmonics. When used to retrieve the map coefficients from the light curves of realistic test planets, the Slepian functions were able to provide better uncertainty estimates with similar or superior fitting accuracy. On the other hand we found that for low-band limits, phase curves were able to retrieve the map coefficients more accurately than eclipse mapping using either basis.

What did you learn this summer?

This project was my first introduction to many new topics such as the spherical harmonics, Bayesian statistics, exoplanet photometry and spectroscopy, MCMC simulations, and general coding in Python. I learned how to conduct research, how to read papers, how to plot nice figures, how to give presentations, and how to collaborate in a group environment. A university course may be a good way to learn about a specific topic in depth, but I believe that the breadth of experiences you pick up during research cannot be matched by any classroom experience.

What was the biggest challenge during your internship?

The largest challenge was likely my lack of coding experience. When I started this project, I was first tasked with deriving the Slepian functions. This was more in line with my background as an undergrad studying math and physics. After this, I wrote code to reproduce the Slepian functions which I used in conjunction with MCMC simulations to fit light curves. I found this was a huge jump for someone who had barely touched Python. It was difficult, but rewarding in the end.

What did you like most about your internship?

Aside from learning a ton of new physics, interacting with Prof. Cowan and his group was a really good experience for me. His graduate students were veritable role models. They were always willing to help and answer my endless questions. They fostered a great sense of community that made myself and the other undergrads feel very comfortable and included. Nick himself takes great pride in the work his students do which motivated me to try and do my absolute best.

Lan Xi Zhu

Intern from McGill University who worked with Nicolas Cowan at McGill University

What was the topic of your internship?

I worked on developing an interactive tool that implements an 1D radiative-convective model to study the thermal structure of planetary atmosphere. That model itself was described in a paper written by Robinson & Catling in 2012. This tool generates the temperature-pressure profile of an atmosphere specified by a series of user-input parameters.

What was interesting about it?

This project is actually along the same lines as my 2019 summer internship project, the Climate App, except that this new version is more professional. I really loved the experience of working on an ongoing project through several years and gradually coming up with new ideas and new features.

What did you discover? What was your most important result?

I made a preliminary version of the interactive tool that successfully reproduces temperature-pressure plots with some realistic sets of atmospheric parameters that are mentioned in Robinson & Catling 2012.

What did you learn this summer?

I learned how to make interactive modules in Python (Jupyter Notebook). In particular, I was able to generate a plot that immediately changes following user inputs, as was done using Javascript in Climate App. I also learned about and applied various problem solving strategies given the complexity of the model.

What was the biggest challenge during your internship?

The model that we tried to implement contains some mathematically heavy components. Since the model was originally developed to study Solar System planets/moons, it was quite challenging to study whether the model could be safely applied to exoplanets that are very different from Solar System worlds.

What did you like most about your internship?

I liked the uncertainty in this project, in the sense that we did not know at the beginning whether this model could actually be made into an interactive program. It was both challenging and exciting when I was not exactly sure where I was heading.