Planet Hunting: Searching for Long Period Exoplanets Utilizing the Transiting Exoplanet Survey Satellite and the Hungarian-made Automated Telescope

Abstract

The science of extra-solar planets (or exoplanets) has taken off in the past three decades or so, and shows no sign of slowing. Astronomers have hopes to eventually be able to measure the abundances of elements in exoplanetary atmospheres in the near future. However, the sample of planets shows a bias towards large, small-orbit planets, likely due to the nature of the methods used to detect them. In this senior thesis, we attempt to address part of this selection bias by developing the methods to detect long-period exoplanets, with periods up to 100 days. We do so by combining the light curves produced from images in the Transiting Exoplanet Survey Satellite (TESS) and the Hungarian-made Automated Telescope-South (HATSouth), and then running a Box-fitting Least-Squares algorithm to look for periodic dips in brightness within that data. The results from that algorithm were then fed into a Random Forest Classifier as features, used to determine which sources may necessitate follow-up observation or analysis. The classifier returned 69 sources as objects of interest, three of which had already been noted by astronomers using data from TESS. We concluded by noting 14 objects of interest on which follow-up observations and analysis would be particularly interesting due to the potential presence of exoplanets.