Self-Consistent Generation of Continents and Their Influence on Global Mantle Dynamics

Abstract

A reasonable model for Earth's origins and its internal workings exists on the basis of tested theories, natural data, geophysical and -chemical investigations, modelling and high-pressure experiments. Nonetheless, many fundamental questions driving Earth sciences remain unresolved. Considering that most of the Earth is inaccessible to direct measurements and there is a paucity of natural evidence from its early stages of formation, geodynamic modelling efforts have become essential to further our understanding of planetary evolution. In this thesis, with the help of the mantle convection code StagYY, I have worked on the following two themes: (i) identifying the qualitative and quantitative correlation between continents and elevated temperatures in the mantle, and (ii) creating continental crust in global models in a self-consistent manner. Continents are the landmasses that cover about a third of the planet's surface and float atop the convecting mantle. It has been suggested that the compositionally evolved continental crust at the top and the melt-depleted cratonic roots underneath it evolved simultaneously during Earth's early history and have survived for billions of years. The existence of dynamic feedback between mantle convection and continents is indisputable and is evident from numerical and analogue modelling done previously. Continents have been shown to affect mantle's convective wavelength, however, whether they insulate the underlying mantle or not remains a matter of debate. In Chapter 3, I provide qualitative observations on the nature of this correlation by conducting a systematic parameter study in 2D global models with mobile (and prescribed) continents. The results of my models show that downwellings bring cold material down into the mantle along continental margins and thermal anomalies are pronounced underneath the continents. Spectral decomposition of temperature and composition fields output by these models gives the dominant degree and amplitude of this correlation. The dominant degree of correlation is shown to evolve with time and continental configuration. Using analytical scaling laws, it is quantitatively shown that correlation decreases with increasing core temperature, number of continents, internal heating, and Rayleigh number. Additionally, the results show for the first time, that melting-induced crustal production (MCP) events resulting from this correlation tend to break the continents apart, thereby destroying the correlation and acting as a negative feedback. In Chapter 2, I developed a new melting parameterisation that has the capability to create continental crust self-consistently. It is a two-step differentiation process to generate continental crust. The basaltic magma is extracted from the mantle, it gets hydrated, and then partially melts to form felsic crust. Formation of continental crust in global models with material recycling and secular cooling of the mantle has never been attempted before. In Chapter 4, I use this newly developed parameterisation to generate primordial continental crust and investigate the global geodynamic regime of early Earth. It is often accepted that subduction zones and intra-plate tectonic settings are the loci of present-day continental crust formation. However, the majority of continental crust from the Archean Eon (4.0-2.5 Ga) was made of Tonalite-Trondhjemite-Granodiorite (TTG) rocks and it was formed in a tectonic regime that still remains an enigma. Parameters such as core temperature, internal friction coefficient, and the ratio of intrusive (plutonism) and eruptive (volcanism) magmatism are varied systematically in global models with radiogenic heat production and core cooling. The results from my simulations and analytical scaling laws show two distinct stages of TTG production: a period of continuous linear growth with time and intense recycling similar to `plume-lid' tectonics that lasts until 1 billion years, followed by a stage with the TTG growth proportional to cubic root of time and moderate recycling. Most importantly and surprisingly, my results show that a drop in TTG production can happen without a major shift in the global geodynamic regime. This is in contradiction with previous suggestions that subduction and onset of plate tectonics were required to explain the decline of continental crust growth around 3.5 Ga. Overall, this work demonstrates the important coupling between continents and mantle dynamics, improves our understanding of early terrestrial planetary evolution, and offers a state-of-the-art melting parameterisation to generate continents.