Earth dynamics and the development of plate tectonics

The physics and chemistry of differentiation on different size scales determines the diversity of the observed planet population. Planetesimals likely differentiated early, while gas was still present in the disk. Fluid and magma mobilization and loss and impact erosion necessarily created a broad taxonomy of planetesimals, which contribute a different share of volatiles, metals, and silicates to growing embryos and planets. Accretionary impacts on larger planets create serial magma oceans associated with core formation and atmospheric outgassing. The evolving chemistry of the growing system can lead to divergent atmospheric compositions, crustal compositions and long term mantle and surface evolution. Forward models for magma ocean evolution match many planetary observations, including the differentiation of the lunar interior, the existence of large low-shear velocity provinces (LLSVPs) in Earth’s lower mantle, the oxidation state of the mantle, and the existence of clays in the primitive crust of Mars. The end of the magma ocean stage creates the initial conditions for later planetary tectonics: if a planet’s mantle solidified fractionally then compositional density stratification will suppress thermal convection for some time. Without vigorous mantle convection, the top thermal boundary layer may have a chance to thicken enough to suppress or delay plate tectonics.

As Earth’s twin in size, density, and heat producing elements, Venus was once predicted to have plate tectonics.  Early global surveys of Venus using data from Magellan found numerous possible subduction sites.  Many rejected this interpretation based on the apparently contradictory evidence for mantle plumes at the most of the subduction locations.  More recently plume-induced subduction was a proposed as a mechanism for initiation of plate tectonics on Earth. Laboratory simulations (Davaille et al., 2017) predict initial rifting of the lithosphere, a partial arc of subduction due to volcanic loading, and radial and concentric extension outboard of the trench as the lithosphere accommodates subduction along an arc. 

Quetzelpetlatl and Artemis Coronae exhibit all of these features, which were not fully predicted by prior models. Roll-back subduction may also explain why Artemis is twice as large as any other corona.  Other features suggest that under some circumstances the subduction may extend well beyond the initial plume diameter and become linear.  Venus’ hot lithosphere, like the Archean Earth’s, may inhibit the evolution to plate tectonics due to fault annealing (Bercovici and Ricard, 2014).  The detailed conditions that allow this process to occur offer lessons for plate tectonics initiation on rocky planets.

Since the discovery of plate tectonics it has been suspected, but never fully demonstrated, that a low-viscosity layer beneath the plates (i.e., the asthenosphere) may play a central role in facilitating their motions. This idea has cast a significant shadow not only on Earth dynamics but also on comparative planetology - in particular on Earth-Venus comparisons. We review constraints on the structure and viscosity of the asthenosphere that are provided by geoid and post glacial rebound studies. We also review previous numerical simulations that highlighted the potentially key role that a rheologically weak asthenosphere, interacting with plates above, could play in maintaining plate tectonics on Earth. We then present new models that expand explorations into the defining characteristics of the asthenosphere and its role for the mantle convection and plate tectonics. The collective works leads us to suggest a geodynamic definition of the asthenosphere that breaks from the traditional view that the principal role of the asthenosphere, for plate tectonics, is to provide basal lubrication for plates.

Venus provides a rich arena in which to stretch one’s tectonic imagination with respect to non-plate tectonic processes of heat transfer on an Earth-like planet. Venus is similar to Earth in density, size, and inferred composition and heat budget. The two planets, likely most similar early in their history, followed different evolutionary paths. Venus’ lack of plate tectonics and terrestrial surficial processes results in the preservation of a unique surface geologic record of non-plate tectonomagmatic processes. A global overview affords a rich spatial and temporal geologic perspective. This history is illustrated by structure-element maps of the Niobe Planitia–Aphrodite Terra area, encompassing >25% of Venus’ surface. Three tectonic domains, which extend globally, represent fundamental changes in global conditions and tectonic regimes through time, divided into eras. Impactors played a prominent role in the ancient era, characterized by thin global lithosphere. The Artemis superstructure era highlights sublithospheric flow processes related to a uniquely large super plume. The fracture zone complex era, marked by broad zones of tectonomagmatic activity, witnessed coupled spreading and underthrusting, since arrested. The three tectonic regimes provide possible analog models for terrestrial Archean craton formation, continent formation without plate tectonics, and mechanisms underlying the emergence of plate tectonics.

A variety of possible mechanisms have been proposed for the formation of ductile shear zones in the lithospheric mantle; such structures are essential for plate tectonics.  These mechanisms include thermal weakening due to viscous heating, localized water weakening coupled with fabric development, and decrease in grain size resulting from large-strain shear that leads to a transition from grain-size insensitive to grain-size sensitive deformation. In natural shear zones, the presence of mylonites and ultramylonites provide evidence of significant grain-size reduction due to dynamic recrystallisation and phase mixing. To investigate the processes involved in the formation of mylonitic rocks, we performed high-strain torsion experiments on two-phase materials to analyse the evolution of both microstructure and rheological behaviour. These experiments demonstrate that, as grain size decreases with increasing strain, mixing of small grains produces textures in which one phase pins the grain boundaries of the other phase, thus stabilising a grain size smaller than that anticipated from dynamic recrystallisation alone. The development of a well-mixed, two-phase, fine-grained rock facilitates a transition in deformation mechanism that causes weakening and strain localisation in rocks.

Moving beyond speculation and towards refined, and testable, models of early Earth evolution requires increasingly robust evidence. For geochemical approaches, obtaining such evidence is clouded by the small areal extent, typically poor exposure and extreme geological complexity of preserved early Archean lithosphere. If the Archean geological record is genuinely incomplete, as commonly claimed, then issues of preservation bias and the 'representativeness' of extant rocks need to be evaluated and quantified. Geodynamic inferences are impeded by the difficulty with establishing credible Phanerozoic analogues on Earth or elsewhere in the solar system. This presentation aims to discuss these uncertainties, and to consider what aspects of the Archean geological record provide the most reliable constraints on crust-mantle evolution in the ancient Earth, and that best inform geodynamic models. Key aspects of the geology and geochemistry of several Archean cratons are highlighted to this end.

Plate tectonics, involving a globally linked system of lateral motion of rigid surface plates, is a characteristic feature of our planet, but estimates of how long it has been the modus operandi of lithospheric formation and interaction range from the Hadean to the Neoproterozoic. Professor Cawood will review the geological proxies used to infer plate motion and conclude that significant changes in Earth behaviour occur in the mid- to late Archaean, between 3.2 Ga to 2.5 Ga. These data include a change from thin mafic crust to thicker intermediate continental crust, including the incoming of potassic and peraluminous granites. Dome and keel structures in granite-greenstone terranes, which relate to vertical tectonics, are replaced by linear thrust imbricated belts. Passive continental margins inferred to mark the onset of sea-floor spreading are no older than 3 Ga. The oldest foreland basin deposits associated with lithospheric convergence are a similar age. Palaeomagnetic data from the Kaapvaal and Pilbara cratons for the interval 2780-2710 Ma and from the Superior, Kaapvaal and Kola-Karelia cratons for 2700-2440 Ma suggest significant relative movements. This talk will consider these changes in the behaviour and character of the lithosphere to be consistent with a gestational transition from stagnant lid tectonics to the birth of sustained plate tectonics.

In order to understand the evolution of plate tectonics research needs to address three related problems of when, how, and what came before.  All three questions need to be answered before we can have any confidence that we have solved this important scientific problem: 1) what was Earth’s tectonic regime before the present episode of plate tectonics began? 2) given the preceding tectonic regime, how did plate tectonics become established? and 3) when did the present episode of plate tectonics begin?  The tripartite nature of the problem complicates solving it, but if and when research finds all three answers, the requisite consilience will provide greater confidence than if only focusing on the long-standing question of when did plate tectonics begin?   This effort requires an open mind and consideration all possible scenarios and lines of evidence.  Upon addressing these questions, there should not be an assumption that the Earth has only experienced two tectonic regimes: plate tectonics and whatever came before. There may have been multiple episodes of plate tectonic and non-plate tectonic convective styles on Earth.  Non-plate tectonic styles likely were dominated by “single lid tectonics” and this likely also evolved as Earth cooled and its lithosphere thickened.

Understanding when and how the continental crust was generated is an important step in unraveling the evolution of the Earth system, yet these questions remain a matter of great debate. Modelling of Hf isotopes in zircon and Nd isotopes in sediments suggest that new continental crust was generated continuously, but with a marked decrease in the growth rate at ~3 Ga, which we ascribe to the onset of plate tectonics. By 3 Ga at least ~65–75% of the present volume of the continental crust had been generated. There is increasing evidence that ~3 Ga also marked the transition between two different types of continental crust. Pre-3 Ga crust was on average mafic, dense and relatively thin (<20 km). In contrast, continental crust that formed after 3 Ga gradually became more intermediate in composition and thicker. The increase in crustal thickness is accompanied by increasing rates of crustal reworking and increasing input of sediment to the ocean. These changes may have been accommodated by a change in the lithosphere strength at around 3 Ga, as the latter became strong enough to support high relief crust. A box model is used to evaluate the timing of crust generation and destruction from 4–0 Ga. Relatively large volumes of crust were destroyed in the late Archaean and Early Proterozoic and over 100% of the present volume of continental crust has been recycled back into the mantle since 3 Ga.