
Selma worked with Dr. Kunimoto on exoplanet research.
What was a highlight(s) of your summer research experience?
One of the biggest highlights was diving into massive datasets from NASA's TESS mission and finding a "hot Neptune" candidate. These planets sit in the "Neptune Desert," a region close to the host star where Neptunian planets are exceptionally rare. With only seven similarly sized hot Neptunes having measured masses and densities so far, confirming and characterizing this candidate is crucial for refining theories on planetary formation and evolution. It's even a candidate target for the Ariel Space Mission.
To measure its mass and work toward confirming this discovery, we needed high-precision radial velocity data. We worked on a proposal for the MAROON-X spectrograph on Gemini North, an 8-meter class telescope. Finding out our proposal was not only approved but scored in the first quartile of all submissions was an incredible feeling. Earning observation time on a world-class telescope was a massive highlight.
What was the most challenging part of your research?
The nature of research is that there is no answer key; you are stepping onto paths that haven't been taken, and when a bug pops up or a computational model fails, you have to creatively troubleshoot your own way out. As of now, only one transiting exoplanet around a white dwarf has ever been discovered. Not because they're extremely rare, but because traditional search tools and detection algorithms simply weren't built to handle white dwarf transits, which are ultra-fast and break standard algorithms.
My main challenge was building and optimizing a data pipeline engineered specifically for these signals. One of the biggest hurdles was that, apart from a single confirmed example, we had no real-world benchmark data to test our code on. I had to architect and generate synthetic dataset simulations to stress-test our algorithms, debug edge cases, and teach the system to recognize true planetary signals.
What are some surprising facts about your research field and/or project?
First, how insanely fast white dwarf transits are! When a planet passes in front of a normal main-sequence star, the telescope captures a continuous "U" shaped curve with plenty of data points. But because white dwarfs are stellar remnants roughly the size of Earth, a transit might only last long enough to capture one or two data points. This completely breaks traditional transit-fitting models, forcing you to engineer new ways to parse the data and teach the code what a "real" signal looks like.
Second, the mystery of the "Neptune Desert." Finding our hot Neptune candidate is like finding a missing puzzle piece. Because they are so rare, characterizing this planet's mass and density will help us figure out how it survived: whether it formed a massive, gas-poor core through giant impacts, or if it's the exposed interior of a former gas giant that had its outer layers stripped away.
How do you think this experience will help you in your developing career?
As someone studying both computer science and physics, this summer was an incredible opportunity to bring my two academic passions together. Moving into my second year of undergrad, I now have highly practical experience architecting code to solve physical problems and handling real-world, messy datasets rather than just clean textbook examples. This intersection of computation, data analysis, and astronomy is exactly where I want to build my career, and this experience cemented my confidence in navigating it.
What have you gained from this experience?
I received a massive technical upgrade. I became highly proficient in Python, specifically for complex data analysis and wrangling huge datasets. I also learned to run GPU-accelerated code on TESS's remote payload data operations (PDO) servers, which was an absolute game-changer for processing speed and scale. Beyond the hard coding skills, I gained a deep resilience for troubleshooting. When you are managing hundreds of thousands of stars and data points, things will inevitably break. Learning how to efficiently track down a bug in a massive codebase and turn a sea of raw numbers into clear, scientific insights was the most valuable takeaway.