Date of Award
Doctor of Philosophy (PhD)
The past several decades have seen an explosion in humanity's knowledge about the existence of distant worlds. Thousands of exoplanets have been now been discovered thanks to the development and refinement of the radial velocity (RV) and transit techniques. We are beginning to piece together an understanding of the diversity of these worlds and of the mechanisms that drive planet formation and evolution. While this progress has been titanic, Earth-sized planets in the habitable zones of Sun-like stars (the places most suited to host life as we know it) remain out of our reach. Noise created by stellar activity along with insufficient instrumental precision conspire to obscure the minuscule signals created by these putative worlds. In this thesis, I chart a course from the discovery of giant exoplanets, to the novel applications of statistical techniques and improved instrumentation, which together show promise of allowing us to finally detect true Earth analogs. In Chapter 2, I present the discovery and confirmation of a pair of transiting hot Jupiters, both identified as planet candidates by the TESS space telescope. Our team collected ground-based photometry, conducted high-contrast imaging, and performed Doppler spectroscopy with CHIRON to verify the planetary nature of both candidates. TOI 564 b is nearly 50% more massive than Jupiter and orbits its star in about 1.7 days. It exhibits a rare grazing transit, making its planetary radius difficult to ascertain. TOI 905 b is slightly larger than Jupiter but only about two thirds its mass, and it orbits its host star in about 3.7 days. Both targets are good candidates for follow-up characterization. In particular, as the one of the brightest known stars to host a grazing transiting planet, TOI 564 is a prime target for future observations that could leverage its planet's transit geometry to constrain the presence of additional planets in the system. In Chapter 3, I present a novel application of principal component analysis (PCA) to the problem of stellar activity. As RV precision has improved, astronomers are now having to contend with astrophysical noise sources that contaminate RV observations. Photospheric features such as spots and faculae create time-varying sources of RV noise that collectively dwarf the RV signals induced by small (Earth-like) exoplanets. Overcoming this noise source is a high priority for the astronomical community because it would enable the detection of potentially habitable planets around bright nearby stars. When PCA is applied to realistic simulated time-series spectra, the fingerprints of stellar activity are revealed. Stellar activity manifests itself throughout stellar spectra on a pixel-by-pixel basis. My simulations show that the signatures of stellar activity are better recovered with high resolution than with high signal-to-noise ratios, even after the spectral lines are fully resolved. This prediction has contributed to the design of future planet-hunting spectrographs. Finally, in Chapter 4, I introduce a new spectrograph designed and built by the Yale Exoplanet Lab, the EXtreme PREcision Spectrograph (EXPRES). I compare EXPRES's performance to that of its predecessor, CHIRON, by focusing on two well-studied stars: Tau Ceti (a chromospherically quiet star) and Epsilon Eridani (a chromospherically active star). EXPRES outperformed CHIRON in terms of its single-measurement RV precision for both targets by a factor of 3. EXPRES's improvement factor over CHIRON was smaller, however, when considering the root-mean square velocity error. I conducted a periodogram analysis and a collection of planet injection and recovery simulations to further measure the relative performance of these instruments. The results suggest that EXPRES is delivering exquisite instrumental precision, but that addressing the impact of stellar activity remains essential to reach truly extreme on-sky precision for real stars. The future course of exoplanet science hinges upon whether it is possible to identify numerous nearby potentially habitable planets. The RV technique, already enormously successful at identifying and characterizing large planets, requires significant improvement in order to overcome the daunting challenge of noise produced by stellar activity. Taken together, recent innovations in the design of new high-precision spectrographs along with the burgeoning community-wide interest in the use of statistical techniques to decorrelate stellar activity RVs from planetary RVs, afford many reasons to be optimistic about the future of exoplanetary science and about the search for life beyond the Solar System.
Davis, Allen Bradford, "Wobbling Towards the Future: Applications of the Radial Velocity Technique to Detect Ever-Smaller Exoplanets" (2021). Yale Graduate School of Arts and Sciences Dissertations. 34.