The Three Fates of Planetary Siblings: Nonlinear Cloud Feedbacks and Climate Evolution on Earth, Venus, and Mars

Abstract

Despite forming from common materials and orbiting the same star, Earth, Venus, and Mars have starkly different modern climates, yet each may have experienced habitable periods in the past. Understanding the divergence in climate between these three planetary siblings is fundamental to our concept of planetary habitability and the search for life in the universe. In this thesis, I investigate the role of nonlinear cloud feedbacks in shaping the climate and evolution of Earth, Venus, and Mars. In Part I, I utilize a hierarchical modeling approach to investigate the role of a subtropical stratocumulus cloud break-up in past and future warm climates on Earth. In Chapter 2, I develop a stacked mixed-layer model that reproduces multiple equilibria and hysteresis found in a Large Eddy Simulation and demonstrate that the critical CO2 at which break-up occurs is highly sensitive to microphysical processes but that break-up is likely to occur at CO2 levels relevant to the Eocene warm climate (~ 50 Mya). In Chapter 3, I quantify the maximum warming a stratocumulus break-up could induce by performing cloud-locking simulations in a state-of-the-art GCM, in which I impose and completely remove subtropical low-clouds. I find that nonlocal stabilizing clouds elsewhere provide a negative feedback that greatly mitigates warming, and that this feedback may contribute to warming but is not enough on its own to explain key features of the Eocene. In Part II, I examine the co-evolution of spin and climate on Venus to constrain the duration of a past putative habitable period. Venus may have been habitable for billions of years if it always had a slow rotation rate conducive to the formation of strong substellar cloud cover. However, spin is coupled to climate through the atmospheric thermal tide, so past changes in atmospheric state would drive dramatic spin evolution. In Chapter 4, I perform targeted GCM simulations varying atmospheric properties to explore the dependence of the atmospheric thermal tide on climate state. I find that the atmospheric tide is extremely sensitive to atmospheric state, and that a habitable atmosphere produces a much stronger tide than Venus's modern atmosphere. A stronger thermal tide would lead to a faster rotation than today. In Chapter 5, I develop a novel spin-climate evolution model and demonstrate that spin-up of planetary rotation due to the strong thermal tide in the habitable state weakens cloud cover and leads to a runaway greenhouse. This self-terminating habitable state has a maximum duration of 0.3 Gyr, with decreasing duration for faster initial rotation. These results sharply constrain the possible lifetime of a habitable state on Venus and suggest that Venus-like planets around Sun-like stars are unlikely to be observed in habitable states. In Part III, I explore the efficacy of high-altitude ice clouds in warming the Martian surface during a chaotic obliquity evolution. In Chapter 6, I conduct GCM simulations of an ancient Mars at varying obliquity and initial surface ice distribution and find multiple equilibria at low obliquity depending on surface ice due to differences in atmospheric moisture availability. At high obliquity, warming is limited by extensive surface ice that promotes low-altitude cloud formation. I then develop a low-complexity model capable of simulating tens of thousands of years of Martian obliquity, surface ice, and high-altitude cloud evolution. This model captures the first-order physics of the GCM results, and the dramatic increase in computational efficiency over the GCM allows an exploration of Martian evolution on geologic timescales. I use the low-complexity model to perform an initial investigation of transient warm periods on early Mars due to the coupling of obliquity, ice, and clouds.