Hierarchical Modeling of Terrestrial Climates and Planetary System Architectures

Abstract

This dissertation investigates the physical processes governing the climate, surface evolution, and orbital characterization of rocky planets as components of a single coupled system. Using mathematical and numerical models spanning molecular spectroscopy, three-dimensional fluid dynamics, coupled volatile-tectonic evolution, and statistical inference, it develops new physical constraints on how rocky worlds evolve and how they can be characterized from afar. The first part establishes the radiative-convective foundation for hot, dense rocky planet atmospheres. Using PCM-HiPT, a one-dimensional line-by-line radiative-convective model developed here for high-pressure, high-temperature conditions, I show that strong shortwave absorption by H2O suppresses near-surface convection, reducing surface temperatures by up to 2000K relative to fully convective predictions. The compositional sensitivity is strongly nonlinear: pure CO2 atmospheres are typically 1000K cooler near the surface than pure H2O atmospheres, and introducing just a few percent of H2O into an otherwise CO2-dominated atmosphere raises surface temperatures by hundreds of kelvin. These results revise predictions for magma ocean longevity and early planetary thermal evolution. The second part investigates how hothouse radiative profiles modify convective organization in three dimensions. Using cloud-resolving simulations, I show that planetary rotation drives a sharp transition from an episodic deluge regime, in which long periods of suppressed convection are punctuated by violent domain-scale precipitation events, to a quasi-steady state dominated by giant, broad-core tropical cyclones with radii of maximum wind nearly four times those of modern Earth storms and total power throughputs an order of magnitude larger. The third part couples the atmospheric physics of the first chapter to a model of volcanic outgassing, atmospheric escape, and lithospheric mechanics to investigate Venus's resurfacing history. I demonstrate that trace H2O enrichment of Venus's CO2-dominated atmosphere is a physically necessary trigger for episodic lithospheric overturns, and that present-day surface temperature carries almost no discriminating information about tectonic history. These results make specific testable predictions for the DAVINCI, EnVision, and VERITAS missions. The fourth part addresses the observational challenge of recovering planetary masses from transit timing variations. I show that the standard eccentricity parameterization used in Markov Chain Monte Carlo fitting introduces severe systematic biases at signal-to-noise levels typical of Kepler observations, and demonstrate that a square-root parameterization reduces these biases across the full noise range explored. Together, these results argue that the climates and architectures of terrestrial planets cannot be understood in isolation from one another, and that moving toward a predictive framework for rocky world evolution requires models that are physically connected across scales and honest about the degeneracies that limit what observations can distinguish.