Introduction¶
The first-order geological feature of our planet must unarguably be plate tectonics. It is responsible for the partitioning of the Earth into terrestrial and marine environments; it drives the periodic gathering and scattering of the continents; it erects and tears down the great mountain ranges; it recycles and perpetually renews the chemical inventory of the surface world. Plate tectonics has even been implicated as a prerequisite [Foley, 2015, Stern, 2016] or co-requisite [Dorn et al., 2017, Lenardic and Crowley, 2012] for the emergence of life on our planet. Alongside the theory of evolution, the theory of plate tectonics underpins the deepest meaningful chapters of the human story.
It is provocative, then, to note that, sixty years since its advent, there remain deep unsolved questions regarding the theory of plate tectonics. It is not yet known how or when it originated [Cawood et al., 2018, Condie, 2018, Davies, 2006], nor when and under what conditions it might cease [O'Neill et al., 2016, Sleep, 2000, Stevenson, 2003]. It is not yet clear why it operates on Earth and not, or no longer, on any other comparable planet [Lourenço et al., 2016, Ruiz et al., 2010, Stern, 2018]. Most significantly, there is as yet no single framework that can successfully unify the kinematic and dynamic aspects of plate tectonics [Bercovici et al., 2015, Coltice et al., 2017, Combes et al., 2012, Yoshida and Santosh, 2011].
The need for a satisfactory solution to the conundrum of plate tectonics has been sharpened of late by the advent of large-scale exoplanet astronomy. As of writing, NASA’s exoplanet database enumerates some four thousand confirmed planet sightings beyond our solar system [NASA Exoplanet Science Institute, 2019], of which more than fifty could be considered ‘Earth-like’ even under the strictest definition of the term [Kopparapu et al., 2014]. Factoring in observational biases toward smaller stars and larger, closer-orbiting planets, the total number of so-called ‘Earth-likes’ within fifty light-years alone could number in the thousands [HEASARC, 2017]. There is an increasingly strong case to be made that such planets may ultimately prove to make up the numerical bulk of all planets in our universe [Petigura et al., 2013].
Such buoyant results suggest the Earth is likely to find itself in good company in the cosmos. However, a ‘comparative planetology’ approach to our own solar system supports the opposite conclusion. Here, at least four planetary bodies qualify as ‘Earth-like’ according to the most common astronomical definitions (CITATION); yet, of these, only one has a youthful and dynamic surface capable of supporting the lengthy chemical reaction chains that sustain and constitute life as we know it [Stern, 2016]. Meanwhile, some planetary bodies formerly viewed as too small or too cold to host life have proved to be tantalisingly dynamic, with Saturn’s moon Titan [Liu et al., 2016], Jupiter’s moon Europa [Riley et al., 2006], and even the spurned ‘dwarf’ planet Pluto [Dalle Ore et al., 2019] exhibiting clear signs of recent geological activity.
Clearly the view that planetary behaviours are predominantly a function of bulk planetary parameters wants nuance. The situation immediately suggests two hypotheses, both as plausible as they are unalike: either planets are deterministic heat engines driven rigidly and precisely by their bulk parameters - henceforth the ‘Sensitivity Hypothesis’ [Sleep, 2000, Valencia et al., 2007, van Heck and Tackley, 2011]; or they are fundamentally non-linear systems governed by chaotic attractors and the exigencies of chance and history - the ‘Stochastic Hypothesis’ [Höink et al., 2013, Lenardic and Crowley, 2012, Weller et al., 2015]. The contest of these two ideas goes to the heart of planetary science.
This thesis sets out the reasoning, methodology, and outcomes of a massive suite-modelling program designed to test this question at a scale never previously attempted. First, a rigorous literature review and theoretical discussion is accompanied by the results of preliminary one-dimensional modelling, designed to illustrate the fundamental issue and provide supporting quantitative constraints for further analysis (Chapter 1). Next, we describe the novel methodologies and new tools devised for this program, together with a brief account of the technical and theoretical challenges involved in numerical modelling on the scale required for this work (Chapter 2). In subsequent chapters we lay out the surprising results of a comprehensive benchmarking study (Chapter 3), thoroughly replicate and extend an iconic viscoplastic rheology paper (Chapter 4), and explore the impact of material heterogeneities on a planet’s choice of tectonic mode (Chapter 5). In the final results chapter (Chapter 6), we apply our model system to the Early Earth to critically evaluate the role that oceanic plateaux could have played in the rise of plate tectonics on our planet. This is followed by two discussion chapters in which we review the theoretical and methodological learnings of the study as a whole (Chapter 7), and endeavour to situate the work in a ‘comparative planetology’ framework, with implications for the projected likelihood of finding substantially Earth-like planets beyond our own (Chapter 8). With the essential business of our study complete, we take an opportunity to consider the status and future of our discipline as a whole from the point of view of the history and philosophy of science (Chapter 9), before bringing matters to a close with some very brief final remarks (Conclusion).
Naturally, references and certain supplementary materials may be found in their proper place (Bibliography and Appendices); in addition to these we also direct the interested reader to the website for our modelling tool, PlanetEngine (planetengine.info), to the interactive digital version of our thesis (https://rsbyrne.github.io/thesis), and to our popular science blog, which navigates related terrain (buildmeaplanet.com). A detailed point-for-point overview of the entire thesis is also provided immediately following this introduction (Summary).
It has been the ambition of this author to attempt to make a significant, and indeed, foundational contribution to the young discipline of fundamental planetary science: what we argue will in the fullness of time be called ‘planetology’ (Conclusion). Such efforts are almost always doomed, and it will be the judgement of the reader whether this attempt falls into that category; but, for the progress of science as a whole, we must on occasion invest in a little fancy, whose unbidden fruit is so often the sweetest. Along this lofty and perhaps foolhardy road, the author has begged and received the indulgence, aid, favour, and patience of many kind and wise people. Though we ask no forgiveness for our impossible dream, we do extend our gratitude to those without whom we could not even have made the attempt. The rather lengthy list concludes the prefatory section of this thesis (Acknowledgements).