Page History

...

1. Stage 1: machine learning and transiting exoplanents. The objective of this stage is to minimize (hopefully to zero) the human interaction in the planetary candidate selection from the TESS photometry currently being used by the WINE collaboration. To date, the process is semi-automated in the sense that transiting exoplanet candidates are searched for in the TESS photometry using state-of-the-art algorithms, but before promoting the targets as promising/interesting for ground-based follow-up, visual inspection of the lightcurves is made in order to remove false-positives which are evident to the human eye but not to current detection algorithms. This has provided us with a large (~thousands) sample of false positives which we can use to train machine learning algorithms to perform the target selection in an automated fashion. The student will work on adapting already existing code to performing machine learning classification for the needs of the project, with the objective of incorporating these techniques into our planet detection workflow. We expect the student to publish this implementation, with a list of the detected candidates from our project which can form the basis for a sample of exoplanets that can be used to study planet ocurrence rates with the TESS mission.
   
   
2. Stage 2: constraining warm giant planet formation scenarios with TESS. With the process of detection and posterior follow-up and characterization of the exoplanets within the project fully automated, the student will work with the PIs of the project in performing injection and recovery tests on the TESS data in order to generate occurrence rates of warm giant exoplanets. The objective of this stage of the project will be to study (1) how efficient is our pipeline for the detection of warm giant exoplanets with TESS given they come from the different formation scenarios depicted in Figure 1 and (2) given this discovery efficiency and our candidate list, what is the real underlying occurrence rate of warm giant exoplanets, and what of the given formation scenarios is the most likely to give rise to the observed population of warm giant exoplanets. We expect the student to again lead a publication on this study.

View filenametess.pdfheight250Image Added

Figure 1: Warm giant exoplanets formation theories and example planet discoveries with our ongoing project. (a) The formation and posterior migration of warm giant exoplanets to the observed close-in orbits can happen through many mechanisms. Three of the most popular are depicted here: disk migration, in which planets migrate inwards through interactions with the disk which either damps their eccentricities if they are alone or forces resonances with nearby exoplanets; high-eccentricity migration in which planets interact with other exterior companions, which makes them undergo high-eccentricity variations which eventually puts them in close-in orbits and in-situ formation, in which planets are formed from collision of "seed" exoplanets that form the warm giant exoplanets at their currently observed close-in orbits. Note how each formation scenario predicts different outcomes. In particular, in-situ planet formation predicts the formation of Earth to super-Earth sized exoplanets, which we should be able to discover with our project. (b) Example discoveries that conform to some planet formation theories. The left panel shows transit timing variations observed in a Jupiter-sized exoplanet, which is evidence for dynamical interactions with a second non-transiting exoplanet. This conforms to planet formation via disk migration. The right panel shows the radial-velocity (RV) measurements of a star in which we detect both, the RV variation of a highly eccentric transiting exoplanet first detected in the TESS photometry and a long-term trend indicative of a long-period companion. This latter system conforms to planet formation via high-eccentricity migration.