Marylou Fournier Tondreau MSc, MSc, PharmD

Abstract:

The characterisation of exoplanetary atmospheres has entered a new era with the advent of the James Webb Space Telescope (JWST), which enables the detection of molecular species in the atmospheres of both gas giants and smaller, potentially rocky planets. This thesis focuses on the development and application of atmospheric retrieval methods to interpret transmission spectra from JWST observations, with particular attention to assumptions in atmospheric models and their impact on the resulting inferences. The first part of this thesis examines how specific modelling choices can influence atmospheric retrievals. I investigate the role of planetary rotation in modifying effective gravity and demonstrate that the inclusion of centrifugal forces can alter retrieved atmospheric properties, particularly for low-density, fast-rotating planets that are common targets for transmission spectroscopy. I also explore the implementation of Centered Log-Ratio (CLR) priors for atmospheric composition, a method designed to better represent the statistical priors of gas abundances in retrievals. This is particularly relevant for smaller exoplanets, where the dominant atmospheric constituents are often unknown and conventional priors can lead to biased results. The second part of the thesis presents retrievals on JWST Early Release Science observations for the hot Saturn WASP-39 b. I apply the NEMESISPY retrieval framework to transmission spectra obtained using multiple JWST instruments using both free retrieval and equilibrium chemistry approaches. These analyses highlight the strengths and limitations of each method and illustrate how the choice of parameterisation and integration of chemical equilibrium codes can impact inferences about the atmospheric conditions on a planet. The final chapters focus on a 1.5 R_earth exoplanet L 98-59 d. Using a JWST transmission spectrum obtained with the NIRSpec G395H instrument, I perform a range of retrievals to test what can be inferred about its atmosphere. My analysis suggests the potential presence of a high mean molecular weight atmosphere dominated by the sulfur-bearing gases H2S and SO2. While these initial results are tentative, I outline an observational strategy targeting additional transits using complementary wavelengths to constrain its composition more robustly. Taken together, the work presented in this thesis examines assumptions in the technique of atmospheric retrieval and demonstrates its application to a diverse set of exoplanets. It shows how careful treatment of model assumptions and priors can improve the robustness of interpretations and enhance the scientific return from JWST observations. As the focus of exoplanet science shifts toward smaller and cooler planets, these tools and approaches will be essential for addressing fundamental questions about atmospheric diversity and the potential for habitability beyond our Solar System

Abstract:

Our ability to detect and characterize small planets in diverse environments is expanding rapidly with the development and continued improvement of the transit and radial velocity methods. Better models, instruments, and telescopes are producing greater planet yields and tighter planetary radius and mass constraints, which in turn provide new targets for atmospheric characterization and produce new insights on planet composition, formation, and evolution. In this thesis, I present work on the characterization and mass determination of small planets with the radial velocity method, the detection of new planets via the transit method, and the study of a planet’s atmosphere through transmission spectroscopy and its implications for planet formation and planet population features.First, I report on mass estimation and characterization of the long-period exoplanet Kepler- 538b. This sub-Neptune with a period of P = 81.7 days is the only planet known to be orbiting its Sun-like star (0.892 M⊙). Simultaneously modeling Kepler photometry and radial velocities (RVs) yields a semi-amplitude of 1.68 ± 0.39 m s−1 and a planet mass of 10.6 ± 2.5 M⊕, which made Kepler-538b the smallest planet beyond P = 50 days with an RV mass measurement at the time of publication. Precise mass measurements on long-period planets will not only directly address questions about the long-period planet population, but also draw comparisons and shed light on aspects of the short-period planet population like the planetary radius occurrence gap and the impact of high stellar irradiation on exoplanet compositions and atmospheres.Next, I discuss K2-136c, a sub-Neptune with a period of P = 17.3 days and the largest of three transiting planets orbiting a late-K dwarf (0.742 M⊙) in the young Hyades open cluster (650 ± 70 Myr). Collecting and analyzing RV data from the HARPS-N and ESPRESSO spectrographs jointly with photometry from the K2 and TESS space telescopes yielded an RV semi-amplitude of 5.46 ± 0.45 m s−1 for K2-136c, corresponding to a mass of 18.0 ± 1.7 M⊕. K2-136c is now the smallest planet to have a measured mass in an open cluster and one of the youngest planets ever with a mass measurement. As a result, this system adds an important new window into young small planet compositions, atmospheric mass loss constraints around young active stars, and planetary evolution at relatively unexplored ages.I then present the TATER planet detection pipeline and apply it to high-cadence photometry of 914 known planet systems observed during TESS Cycle 3. This work has led to the new validation of 4 short-period planets. This study provides independent modeling and vetting of hundreds of planet candidates while also expanding the known planet population and providing updated transit ephemerides and planet radii.Finally, I report on the atmospheric characterization of WASP-166b, a short-period super- Neptune (P = 5.44 d, Mp = 32.1 ± 1.6 M⊕, Rp = 7.1 ± 0.3 R⊕). WASP-166b is located near the edge of the Hot Neptune Desert, a sparse region of exoplanet parameter space at high stellar irradiation and intermediate planet radii. Using transmission spectroscopy of WASP-166b (two transit observations with the James Webb Space Telescope), initial analyses show evidence of H2O and CO2; no evidence of SO2, NH3, or a cloud deck; constraints on planetary metallicity and the C/O ratio; and a plausible formation pathway that includes planetesimal accretion followed by core erosion or photoevaporation. This in turn points to mechanisms that can create substellar or stellar C/O ratios and superstellar metallicities, like photoevaporation and core erosion, as feasible components of the formation of the Hot Neptune Desert