Exploring Alternative Techniques for Exoplanet Detection Around Host Stars




Torres-Quijano, Amílcar Rafael

Journal Title

Journal ISSN

Volume Title



The search for exoplanets has revolutionized our understanding of the universe, revealing a vast diversity of planets unlike those in our Solar System. Established techniques like measuring transits, radial velocity, and direct imaging have confirmed over 5,600 exoplanets. Technological advancements and alternative techniques, such as machine learning and drift scanning, offer potential enhancements to current methods. I explore these alternatives to improve exoplanet detection around distant stars and deepen our understanding of planetary formation.

Stars and their planets form from the same gas cloud, making a star's composition an indirect indicator to planetary interior compositions. Machine learning algorithms trained on the abundances of known exoplanet-hosting stars, allows us to identify significant "features'' (abundances or molar ratios) that may indicate the presence of a small planet. I test three distinct groups of exoplanets (general small planets, sub-Neptunes, super-Earths) and find that Na and V are key elements regardless of planetary radii. I also create a list of stars with a ≥90% probability of hosting small planets across each ensemble which indicates a chemically unique population.

Detecting planets from ground-based observations in the mid-infrared (MIR, 7-25 µm) is challenging due to the need of removing fast time variable components, array background, in addition to significant overhead. Drift scanning addresses these issues eliminating the need for chopping and nodding. I analyze drift scan observations from the CanariCam instrument on the 10.4 m Gran Telescopio Canarias, report a significantly improved Strehl ratio (54%) compared to chop/nod (35%) and explore applications for future missions.



Computational methods, Exoplanets, Infrared astronomy, Mini Neptunes, Stellar abundances, Super Earths



Physics and Astronomy