Discovering new solar systems : Jupiter analogs and the quest to find another Earth

Robertson, Paul Montgomery
Journal Title
Journal ISSN
Volume Title

Exoplanets are now known to be ubiquitous throughout the Galaxy. From the Kepler survey, we expect nearly every main-sequence star to form planetary systems during its formation phase. However, the detection limits of Kepler are confined to planets with short orbital periods, comparable to those in the inner solar system. Thanks to the long observational time baseline of the McDonald Observatory Radial Velocity (RV) Survey, we can identify gas giant planets in the outer regions of extrasolar planetary systems. The statistics of such planets are not well known, and are important for understanding the physics behind planet formation and migration. In this dissertation, I detail the discovery of five giant exoplanets on long-period orbits–so-called “Jupiter analogs.” For two systems of giant planets discovered through our survey, pairs of planets follow closely-packed orbits, creating the possibility for dynamical instability. I therefore examine the orbital resonances that allow these planets to avoid gravitational disruption. Because we see an abundance of small, potentially habitable exoplanets in the Kepler data set, current and upcoming exoplanet surveys concentrate on finding Earth-mass planets orbiting stars near enough to facilitate detailed follow-up observations. Particularly attractive targets are cool, low-mass “M dwarf” stars. Their low masses (and thus higher RV amplitudes from exoplanets) and close-in habitable zones allow for relatively quick detection of low-mass planets in the habitable zone. However, the RV signals of such planets will be obscured by stellar magnetic activity, which is poorly understood for M stars. In an effort to improve the planet detection capabilities of our M dwarf planet survey, I have conducted a detailed investigation of the magnetic behavior of our target stars. I show that, while stellar activity does not appear to systematically influence RV measurements above a precision level of ∼ 5 m/s, activity cycles can occasionally produce RV signals in excess of 10 m/s. Additionally, I show that long-term, solar-type stellar activity cycles are common amongst our M dwarf targets, although they are significantly less frequent than for FGK stars. In the case of GJ 328, I have discovered a magnetic activity cycle that appears in the RV data, causing the giant planet around the star to appear to be on a more circular orbit than indicated by the activity-corrected data. Such corrections are essential for the discovery of Earthlike exoplanets.