Generating and Characterizing Terrestrial Exoplanet Atmospheres through Computational Modeling and Observations

Abstract

The field of exoplanetary science and astrobiology entered a golden era when the first exoplanet around a Sun-like star, 51 Pegasi-b, was discovered in 1995. In the last 30 years alone, the number of confirmed exoplanets has grown from 1, to nearly 6000. With the recent launch of JWST, the community is once again on the precipice of rapid growth and discovery. For the first time, JWST has enabled the potential observation of high mean molecular weight atmospheres on terrestrial exoplanets, enabling the foremost searches for atmospheric biosignatures and habitability indicators on Earth-sized exoplanets. However, given technological limitations, these surveys are constrained to terrestrials with M dwarf hosts, in what has become colloquially known as "The M-dwarf Opportunity". While these planets may be Earth-sized, they exist in radically different stellar and system environments than our home planet. The extreme stellar activity, extended pre-main sequence phase, and close proximity of the habitable zone to M dwarf hosts leads to radiative challenges for planetary companions. Moreover, the long lifetimes and strong gravitational interactions in M dwarf systems suggest many of these planets are tidally locked, potentially leading to steep temperature gradients across the surface, powerful winds, and in extreme cases, atmospheric collapse. Given the vast range of evolutionary pathways possible for terrestrial planets in these systems, robust evolutionary modeling is becoming increasingly vital to interpreting JWST observations, predicting observational discriminants, identifying likely false positive biosignatures, and connecting the extant atmospheres discovered to likely evolutionary histories. The body of research described in this dissertation is intended to prepare the scientific community to analyze data, select subsequent observations, and connect observational results to likely evolutionary histories of terrestrial exoplanets. Furthermore, we demonstrate that tenuous ($\sim$0.1 bar) atmospheres are possible and may be likely for irradiated M dwarf planets, sustained by constant geochemical outgassing against extreme expected stellar-driven escape. In this work, we focus on the high priority TRAPPIST-1 system of seven Earth-sized planets, which has been the target of early and ongoing JWST campaigns. Through this work, we provide comprehensive analyses of past and present atmospheric evolution on the TRAPPIST-1 planets, along with suggestions for future observations. However, while we apply our methods to TRAPPIST-1, the tools, data pipelines, and experimental methodology we develop are broadly applicable to general terrestrial exoplanet characterization, which has been identified as a NASA priority research area through their selection of JWST DDT time and the conception of the Habitable Worlds Observatory. To predict and constrain the generation and evolution of secondary atmospheres in this work, we employ computational models for hydrodynamic thermal atmospheric escape, atmospheric photochemistry, planetary climate, JWST noise modeling, JWST observation planning, and line-by-line radiative transfer. Additionally, we develop and test several approaches to modeling pipelines to achieve varying balances of computational efficiency, simulation detail, and statistical rigor. Through our extensive use of computational modeling, we characterize the past and present atmospheric evolution of the TRAPPIST-1 planets, and provide a real world example of how to conduct terrestrial exoplanet characterization with JWST. We begin by considering the early evolution of the TRAPPIST-1 system as a whole, driven by the extreme radiation from the host star's pre-main sequence phase. During this time, we find that all 7 planets undergo periods of hydrodynamic thermal atmospheric escape which typically lead to desiccation on the interior planets and water loss of up to nearly 10 Earth oceans on planets in the habitable zone. Furthermore, by connecting this work with the pioneering JWST observations of TRAPPIST-1b and c, we show that atmospheric oxygen on the interior planets could provide an upper limit on the initial water content for all planets in the system of approximately 8 Earth oceans, suggesting the outer planets may have retained surface water when entering the habitable zone. With this important historical context of water loss, we find that oxygenated atmospheres are theoretically justifiable TRAPPIST-1 system. With the likely history of volatiles on these planets informed by the first project in this work, we then look to current observing efforts to constrain the plausible extant atmospheres on the TRAPPIST-1 planets. As JWST observations continued, it became clear that thin oxygen-dominated atmospheres are among the only atmospheric scenarios still consistent to $<$1$\sigma$ for TRAPPIST-1b, c, and d observations. Moreover, significant upcoming observation time allocated to TRAPPIST-1e motivates the need to understand potential atmospheric compositions on the habitable zone planets that could be uncovered by future data. Guided by the currently available and future observations, we also conduct comprehensive parameter sweeps of realistic outgassing, surface deposition, and top-of-atmosphere escape rates to identify self-consistent atmospheres for all 7 planets. For TRAPPIST-1b, c, d, and e, we identify several atmospheric compositions and climate profiles that are still consistent with current data. For the outer planets of the system, we provide the community with atmospheric spectra that represent a range of environmental archetypes that may be useful in planning and justifying upcoming observing campaigns. Finally, we apply advanced Bayesian nested sampling algorithms with a photochemical model to select realistic boundary conditions for TRAPPIST-1c and fit for the source and sink rates that best match current JWST thermal emission observations with quantified statistical uncertainty. Overall, we support the conclusion that thin ($\sim$1 mbar -- 0.5 bar) O$_{2}$-dominated atmospheres provide the best fitting atmospheric solution for the emission data of TRAPPIST-1c. This body of work is one of the first ever examples of how to combine observations, atmospheric models, and statistical methods to discern constraints on planetary source and sink fluxes for terrestrial exoplanets. It pioneers methodology for how to best fit self-consistent atmospheric profiles to data, which has broad implications for atmospheric retrieval in the future. For the first time, we show that tenuous atmospheres can be generated and maintained by geochemical outgassing on the irradiated M dwarf planets of the TRAPPIST-1 system, despite the expected high atmospheric escape rates.