News & Views

What controlled the evolution of Plate Tectonics on Earth?

Stephan Sobolev - edited by Anne Glerum August 14, 2019

Great Unconformity - Immensity River, Grand Canyon

The Great Unconformity–boundary between Lower Paleozoic and Proterozoic rocks indicating the global surface erosion event following Snowball Glaciations that might have kick-started the contemporary phase of active plate tectonics. Photo of the Grand Canyon, Immensity River, Colorado (public domain).

Prof. Dr. Stephan Sobolev. Head of the Geodynamic Modelling section of GFZ Potsdam.

Plate tectonics is a key geological process on Earth, shaping its surface, and making it unique among the planets in the Solar System. Yet, how plate tectonics emerged and which factors controlled its evolution remain controversial. The recently published paper in Nature by Sobolev and Brown suggests new ideas to solve this problem….

What makes plate tectonics possible on contemporary Earth?

It is widely accepted that plate tectonics is driven by mantle convection, but is the presence of said convection sufficient for plate tectonics? The answer is no, otherwise plate tectonics would be present on Mars and Venus and not only on Earth. The geodynamic community recognized that another necessary condition for plate tectonics is low strength at plate boundaries and particularly along the plate interfaces in subduction zones (e.g. Zhong and Gurnis 1992, Tackley 1998, Moresi and Solomatov 1998, and Bercovici 2003). To quantify the required strength at subduction interfaces, we have used global models of plate tectonics (Fig. 1A) that combine a finite element numerical technique employing visco-elasto-plastic rheology to model deformation in the upper 300 km of the Earth (Popov and Sobolev 2008) with a spectral code to model convection in the deeper mantle (Steinberger and Calderwood 2006). The model reproduces well present-day plate velocities if the effective friction at convergent plate boundaries is about 0.03 (Fig.1B). Low strength corresponds to subduction interfaces that are well lubricated by continental sediments (low friction; Lamb and Davis 2003, Sobolev and Babeyko 2005, or low viscosity; Behr and Becker 2018). In case of sediment shortages in the trenches (corresponding to a friction coefficient of 0.07-0.1), plate velocities would first decrease about two times (Fig. 1C) and then even more because of less negatively buoyant material having subducted into the mantle, leading to less convection driving force.

Effects of sediments on contemporary subduction according to global numerical models.

Figure 1. Global numerical model showing the effect of sediments on contemporary subduction. (A) The global model combines two computational domains coupled through continuity of velocities and tractions at 300 km depth. (B) NUVEL 1A plate velocities in a no-net-rotation reference frame (black arrows) versus computed velocities (blue arrows) for the global model with a friction of 0.03 at convergent plate boundaries. (C) Root mean square of computed plate velocities in the global model versus friction coefficient at convergent plate boundaries.

Hypothesis and its testing

Based on the previous discussion, we infer that continental sediments in subduction channels act as a lubricant for subduction. In addition, the presence of these sediments in trenches is a necessary condition for the stable operation of plate tectonics, particularly earlier in Earth’s evolution when the mantle was warmer and slabs were relatively weak. With this hypothesis we challenge the popular view that secular cooling of the Earth was the only major control on the evolution of plate tectonics on Earth since about 3 Ga. The hypothesis predicts that periods of stable plate tectonics should follow widespread surface erosion events, whereas times of diminished surface erosion should be associated with reduced subduction and possibly intermittent plate tectonics.

We test this prediction using geological proxies believed to identify plate tectonics activity (supercontinental cycles) and geochemical proxies that trace the influence of the continental crust on the composition of seawater (Sr isotopes in ocean sediments; Shields 2007) and continental sediments in the source of subduction-related magmas (oxygen and Hf isotopes in zircons; Cawood et al. 2013, Spencer et al. 2017). All three geochemical markers indeed show that just before or in the beginning of supercontinental cycles the influence of sediments is increasing, while it decreases before periods of diminished plate tectonic activity, like the boring billion period between 1.7 and 0.7 Ga (Cawood and Hawkesworth 2014; Fig. 2). The largest surface erosion and subduction lubrication events were likely associated with the global glaciation evens identified in the beginning (2.5-2.2 Ga) and at the end (0.7-0.6 Ga) of the Proterozoic Era (Hoffman and Schrag 2002). The latter snowball Earth glaciation event terminated the boring billion period and kick-started the modern phase of active plate tectonics.

Another prediction of our hypothesis is that in order to start plate tectonics, continents had to rise above sea level and provide sediments to the oceans. This prediction is again consistent with observations: there are many arguments for the beginning of plate tectonics between 3 and 2.5 Ga (see the review of Condie 2018) and, at the same time, this period is most likely when the continents rose above sea level (Korenaga et al. 2017).

Figure 2. Cartoon summarizing the factors that control the emergence and evolution of plate tectonics on Earth. Enhanced surface erosion due to the rise of the continents and major glaciations stabilized subduction and plate tectonics for some periods after 3 Ga and particularly after 0.7 Ga after the cooling of the mantle. Blue boxes mark major glaciations; transparent green rectangles show the time intervals when all three geochemical proxies consistently indicate increasing sediment influence (major lubrication events); and, a thick black dashed curve separates hypothetical domains of stable and unstable plate tectonics. The reddish domain shows the number of passive margins (Bradley 2008), here used as a proxy for plate tectonic intensity.

What was before plate tectonics?

The earlier geodynamic regime could have involved episodic lid overturn and resurfacing due to retreating large-scale subduction triggered by mantle plumes (Gerya et al. 2015) or meteoritic impacts (O’Neill et al. 2017). Retreating slabs would bring water into the upwelling hot asthenospheric mantle, generating a large volume of magma that formed protocontinents. Extension of the protocontinental crust could have produced nascent subduction channels (Rey et al. 2014) along the edges of the protocontinents lubricated by the sediments. In this way, a global plate tectonics regime could have evolved from a retreating subduction regime.

What is next?

Despite of the support from existing data, more geochemical information is required to conclusively test our hypothesis about the role of sediments in the evolution of plate tectonics. Additionally, this hypothesis must be fully quantified, which in turn will require coupled modeling of mantle convection and plate tectonics, surface processes and climate.

References
Behr, W. M. and Becker, T. W. Sediment control on subduction plate speeds. Earth Planet. Sci. Lett. 502, 166-173 (2018).

Bercovici, D. The generation of plate tectonics from mantle convection. Earth Planet. Sci. Lett. 205, 107–121 (2003).

Bradley, D. C. Passive margins through earth history. Earth Sci. Rev. 91, 1-26 (2008).

Cawood, P. A., Hawkesworth, C. J. and Dhuime, B. The continental record and the generation of continental crust. Geol. Soc. Amer. Bull. 125, 14-32 (2013).

Cawood, P. A. and Hawkesworth, C. J. Earth's middle age. Geology 42, 503-506 (2014).

Condie, K. C. A planet in transition: The onset of plate tectonics on Earth between 3 and 2 Ga? Geosci. Front. 9, 51-60 (2018).

Gerya, T.V. et al. Plate tectonics on the Earth triggered by plume-induced subduction initiation, Nature 527, 221-225 (2015).

Hoffman, P. F. and Schrag, D. P. The snowball Earth hypothesis: testing the limits of global change. Terra Nova 14, 129–155 (2002).

Korenaga, J., Planavsky, N. J. and Evans, D. A. D. Global water cycle and the coevolution of the Earth's interior and surface environment. Phil. Trans. R. Soc. Am. 375, 20150393 (2017).

Lamb, S. and Davis, P. Cenozoic climate change as a possible cause for the rise of the Andes. Nature 425, 792-797 (2003).

Moresi, L. and Solomatov, V. Mantle convection with a brittle lithosphere: Thoughts on the global tectonic style of the Earth and Venus. Geophys. J. Int. 133, 669-682 (1998).

O’Neill, C. et al. Impact-driven subduction on the Hadean Earth. Nature Geosci. 10, 793-797 (2017).

Popov, A.A. and Sobolev, S. V. SLIM3D: A tool for three-dimensional thermo mechanical modeling of lithospheric deformation with elasto-visco-plastic rheology, Phys. Earth Planet. Inter. 171, 55-75 (2008).

Rey, P. F., Coltice, N. and Flament, N. Spreading continents kick-started plate tectonics. Nature 513, 405–408 (2014).

Shields, G. A. A normalised seawater strontium isotope curve: possible implications for Neoproterozoic-Cambrian weathering rates and the further oxygenation of the Earth. eEarth 2, 35-42 (2007).

Sobolev, S. V. and Babeyko, A. Y. What drives orogeny in the Andes? Geology 33, 617-620 (2005).

Spencer, C. J., Roberts, N. M. W. and Santosh, M. Growth, destruction, and preservation of Earth's continental crust. Earth. Sci. Rev. 172, 87-106 (2017).

Steinberger, B. and Calderwood, A. Models of large-scale viscous flow in the Earth’s mantle with constraints from mineral physics and surface observations. Geophys. J. Intern., 167 1461–1481 (2006).

Tackley, P. J. Self-consistent generation of tectonic plates in three-dimensional mantle convection. Earth Planet. Sci. Lett. 157, 9-22, (1998).

Zhong, S. and Gurnis, M. Viscous flow model of a subduction zone with a faulted lithosphere: long and short wavelength topography, gravity and geoid. Geophys. Res. Lett. 19, 1891–1894 (1992).

Searching for future directions in tectonic modelling

John Naliboff & Jolante van Wijk - edited by Diogo Lourenço August 2, 2019

Searching for future directions in tectonic modelling. Boundary images by domdomegg - Own work, CC BY 4.0, in https://en.wikipedia.org/wiki/Plate_tectonics.

Geoscientists frequently use forward geodynamic simulations to test hypotheses derived from geophysical and geologic observations. While numerical simulations of lithospheric deformation have lead to key advances in our understanding of tectonic processes, in many cases it remains difficult to ascertain whether numerical models reproduce observations for the correct underlying regions. This week, John Naliboff and Jolante van Wijk discuss this issue, and talk about a White Paper being prepared by the Computational Infrastructure for Geodynamics (CIG) Long-Term Tectonics Working Group on this topic.

John Naliboff. Assistant Research Scientist in the Department of Earth and Planetary Sciences, UC Davis.

In recent years, advances in numerical methodology, high-performance computing and elucidation of complex geologic observations have enabled 3-D simulations of long-term lithospheric deformation at kilometer-scale resolution and with complex non-linear material behavior. The lithosphere models generally include rheological and compositional layering that delineate a brittle upper crust, ductile (viscous) to brittle mid- to lower crust and a brittle to ductile lithospheric mantle overlying a purely viscous asthenosphere. Inherently, the rheological behavior of distinct layers varies temporally through complex feedbacks between temperature, grain-size, strain-rate, phase transitions, flexural stresses and finite deformation. Across a wide range of tectonic settings, numerical investigations incorporating this type of behavior have qualitatively and in some cases quantitatively reproduced key first- and second-order geologic observations.

Jolante van Wijk. Associate Professor in the Department of Earth and Environmental Science, New Mexico Tech.

Despite these successes, numerous significant challenges remain as the computational tectonics community looks toward investigations that account for physical processes acting across a wide range of spatial and temporal scales (Figure 1). Geodynamic model development currently evolves around modifying existing models to include surface processes, thermodynamically consistent melt and volatile transport, metamorphic reactions, and brittle failure to reproduce characteristic features of the seismic cycle.

While many of these processes or features are active areas of research and have been addressed on an individual basis, it often remains unclear at best as to how one should numerically validate even the simplest models of lithospheric deformation. In other words, one can ask whether numerical models of lithospheric processes are reproducing key observations for the correct underlying reasons. Significantly, this question of validation equally applies to observational studies: given that many geologic processes contain significant feedbacks across vast spatial and temporal scales, to what degree can a set of specific observations be interpreted to meaningfully reflect first-order processes?

Continued close collaboration between observational, experimental, and computational Earth scientists is needed to overcome these challenges. At present, the CIG Long-Term Tectonics Working Group is preparing a white paper draft that outlines a 5-10 year vision for collaboration between computational Earth scientists and experimental and observational communities. Given the vast series of topics and disciplines associated with lithospheric dynamics, the White Paper will be organized around Transitional Domains within the lithosphere. The Transitional Domains include Earth’s surface, the brittle-ductile transition, the Moho, the mid-lithosphere discontinuity, and the lithosphere-asthenosphere boundary. For each of these domains we will address the following questions:

How are these domains characterized in the Earth’s lithosphere?
(How) have these domains been modeled previously?
What steps can we take to improve the characterization of these transitional domains in numerical models?
What new methods need to be developed to implement the domains?

Figure 1. Spatial and temporal scales associated with distinct lithospheric processes, which was published in Cooper et al., 2015, GSA Today, v. 25(6), pp. 42-43 (Figure 1a).

A new generation of geodynamic models will be developed to include the transitional domains. These models will need to be validated, using datasets from the observational and experimental communities, with newly developed techniques.

The White Paper will also include sections on increasing the value of lithospheric models for other scientific communities, and on a pathway toward increasing societal relevance of our modeling efforts.

Please contact John Naliboff (jbnaliboff@ucdavis.edu) and Jolante van Wijk (jolante.vanwijk@nmt.edu) for suggestions or questions. A draft White Paper will be presented at CIG’s annual meeting at AGU 2019 in San Francisco.

Let’s talk about plagiarism

Anonymous - edited by Tobias Meier July 17, 2019

Hey you! Do you have 5 minutes to talk about plagiarism?
Have you ever wondered if some parts of a thesis that you have supervised are simply a copy-paste from another thesis or article? This week, an anonymous guest author will tell us about their personal experience with plagiarism in science and what can be done against it.

Granted, it is not the most fascinating topic. Until recently, I really thought there was nothing to say about it. Everybody agrees that plagiarism is bad, and one shouldn’t do it, right? Plagiarism is just for a pair of lazy bachelor students or maybe one or two entitled old professors who believe they are untouchable, right? Right?! Oh boy, was I naive!

For me, it all started with reading a few words that do ring a bell on a master student thesis that I had co-supervised. After some more investigation, I realized that this student did indeed copy and paste sentences and even paragraphs from my PhD thesis, as well as from other articles. He did also plagiarize in former assignments and in a scientific article he published in a journal at the beginning of the year. Uh uh. At this point, the student had already defended his thesis and just got his master degree validated. In the process, the thesis had been evaluated by two independent reviewers and also had been read by my two PhD advisors. Nobody suspected anything. And this happened at THE top Earth science research institution of a country which is renowned for the quality of its research. No problem, I think, I contact the co-supervisor and the director of studies. For sure they’ll know what to do. Hahaha. I spare you the details, but, to sum it up, the master degree had already been awarded, so there was no way whatsoever to change anything about it.

I didn’t make friends this past few weeks by insisting and playing the self-righteous scientist card. The student still got his master and will soon be enrolled in the PhD program of the same institution. However, my complaining seems to have had some effect. In the institution in question, they will buy the rights to a plagiarism scanner software and create a special commission to deal with plagiarism cases. From now on, master students will have to include a declaration of originality for their master theses, and they will have a course on research integrity. If the same situation arises, there will be official tools to deal with it, and hopefully the education the students will receive will help prevent plagiarism.

So yes, sometimes it’s worth it to be (a bit) annoying. Here are a few other things you might want to consider in order to avoid this kind of situation.

Plagiarism and “self-plagiarism” (also called text recycling) are not allowed by most journals, however, there is quite a large part of the scientific community that does not see the problem with self-plagiarism and does it regularly in articles. Some copy whole paragraphs from former articles of theirs and, sometimes, these articles pass the plagiarism scan that journals generally do. So it is really worth it to scan for plagiarism every paper you receive to review. That’s how I gave my fastest peer review ever: 5 minutes to scan the article, 5 min to realize that a whole section was a copy and paste from another article, and 5 min to write a rejection message.

Check every thesis, every draft and every paper you receive with a plagiarism software. You might have some surprises. If you do so, you’re making students/co-authors a favour. Had I done that check with my student prior to his thesis submission, he could have had the chance to make things right, avoided cheating on an exam, and got his master degree fair and square. Instead of this, he has to walk around with a master diploma he didn’t really earn. Not a good start in one’s professional life. Same with co-authors: if you catch their plagiarism, you save all your team the embarrassment of getting your paper rejected by a journal because of this.

It might be a good idea to check the policy of your institution on plagiarism before you’re faced with the situation I described earlier. If there is nothing planned, urge people in charge to set up some procedure. You don’t want to be in the situation of catching a student after his master has been validated and not being able to do anything about it.

Finally, to people who practice text recycling: if you want to copy a sentence from another article because it is the best sentence to describe your thoughts… Why not putting quotes? If you don’t, you’re just being dishonest.

Production and recycling of Archean continental crust

Charitra Jain - edited by Luca Dal Zilio June 19, 2019

Beautiful swirly mantle. What kind of secrets are hidden here?

Continents are essential for the development and survival of life on Earth. However, as surprising as it may sound, there did not exist a planetary scale numerical model to show the formation of the oldest continents until the recent study ‘Growing primordial continental crust self-consistently in global mantle convection models‘ in Gondwana Research by Jain et al., 2019. Hot off the press, the first author of this study himself, Charitra Jain, Post Doctoral Research Associate in the Department of Earth Sciences at Durham University, shares the scoop in our News & Views!

Why do we care?

Uniquely positioned within the habitable zone [1], Earth is the sole planet within our solar system that sustains life. Understanding the factors that make a planet habitable [2,3] are becoming increasingly relevant with the rapidly expanding catalogue of extrasolar planets over the last decade [4,5]. The operation of plate tectonics and the formation and stability of continental landmasses have played a crucial role in the atmospheric evolution and development of life on Earth [6]. Plate tectonics is a theory where the outer surface of the Earth (lithosphere) is fragmented into a number of mobile plates that drift at a speed of few centimetres per year relative to each other, atop a convecting mantle. Mountains, volcanoes, and earthquakes are found at the boundaries of these plates. Continents cover about a third of the planet’s surface area and are made of thin crust overlying a thick and undeformable cratonic lithosphere [7,8]. Even after having assembled a reasonable model for its origins and internal workings, many fundamental questions pertaining to Earth sciences remain unresolved, for example:

How did the first continents form and what accounts for their stability over billions of years? Was their growth an episodic process [9] or a continuous process with a significant drop in crustal growth rate around 3 Ga (billion years ago) [10]? How much of the continental crust has recycled into the mantle [11]?
What was the global geodynamic regime exhibited by early Earth? When and how did the subduction-driven plate tectonics commence [12]? Did the presence of continents play a role in a regime transition from vertical tectonics to horizontal tectonics around 3Ga [13]?

Owing to the increasing paucity of natural observational data as we go back further in time [14], numerical modelling constrained by experimental and field data has thus become indispensable to uncover the secrets of Earth’s evolutionary history. Our recent study [15] is a step in the right direction where we have developed a new two-stage melting algorithm to create oceanic (basaltic/mafic) and continental (felsic) crust in self-consistent global mantle convection models for the first time.

What’s new in these models?

Generally, two stages of mantle differentiation are inferred to generate continental crust as shown in the schematic in Fig. 1A. First, basaltic magma is extracted from the mantle. Second, it is buried and partially melts to form felsic continental crust. During the much hotter Archean conditions [16,17], majority of continental crust was made of Tonalite-Trondhjemite-Granodiorite (TTG) [18,19] rocks. Experimental data suggests that TTGs are formed when hydrated basalt melts at garnet-amphibolite, granulite or eclogite facies conditions [20,21] and specific P-T conditions [22] have been employed as a criterion for generating TTGs by the authors in their models.

Interested in the long-term planetary evolution, we parametrised the processes of melt generation and melt extraction. If the melt is generated within the top 300 km of the mantle (Fig.1B1), it is instantaneously removed from the depth (Fig.1B2) [23,24] and transported both to the bottom of the crust (plutonism/intrusion) and to the top of the model domain (volcanism/eruption) (Fig. 1B3). The intruded melt stays molten while a temperature adjustment to account for adiabatic decompression is applied, and tends to result in a warm, weak lithosphere. The erupted melt is rapidly solidified by setting its temperature to the surface temperature (300K), resulting in a strong and cold lithosphere [25]. The mass ratio of erupted to intruded melt is controlled by a parameter called eruption efficiency, which is tested extensively in our simulations. Geological data suggests that the majority of mantle-derived melts intrude at a depth, corresponding to an eruption efficiency between 9% and 20% [26].

Figure 1: A, One-dimensional compositional variation with basalt-harzburgite continuum consisting of a mixture of olivine (ol) and pyroxene-garnet (px-gt) mineralogies in different proportions. Upon initialisa- tion, the whole mantle has a pyrolytic composition. B, Cartoon depicting a section of a mesh column (not to scale) in a stage: B1, where crustal production has already happened; B2, after melt removal but before compaction or opening gaps in lithosphere for magmatism; B3, with the eruption and intrusion of the melt with the white downgoing arrows representing compaction of tracers/markers.

How do these results stack up against data?

By varying initial core temperature, eruption efficiency, and limiting the mass of TTG that can be generated in our simulations to 10% and 50% of basalt mass, we present results from two sets of simulations. The crustal volumes have the same order of magnitude and the crustal composition follows similar trends as reported from geological and geochemical data [10,27,28]. Interestingly, we report two stages of TTG production without needing a significant change in convection regime: a period of continuous linear growth with time and intense recycling fuelled by strong plume activity and lasting for around 1 billion years, followed by a stage with reduced TTG growth and moderate recycling. The production of TTGs happen at the tip of deformation fronts driven by the lateral spreading of plumes (mantle upwellings) that rise to the surface (Fig. 2). These results indicate that the present-day slab- driven subduction was not required for the genesis of Archean TTGs [29,30] and early Earth exhibited a “plutonic squishy lid” or vertical-tectonics geodynamic regime [31–33].

Figure 2: Thermal (top) and compositional (bottom and zoom-in) evolution with time for a simulation with initial core temperature of 6000 K, mantle potential temperature of 1900 K, eruption efficiency of 40%, and TTG mass limited to 10% of basalt mass. The lighter shades of teal in the composition field represent progressive mantle depletion (higher harzburgite content) with time.

What’s coming next?

These results are parameter-dependent to some extent and they would change in a 3D domain as that would limit the impact of plumes on lithosphere dynamics. Future models will aim for forming the viscous cratonic roots and incorporating the effects of water [34,35] and grain-size [36–38] on mantle rheology. Going forward, modelling coupled core-mantle-atmosphere systems will shed more light on the role of different tectonic modes towards planetary habitability [39,40] and help solve these contentious aspects.

References:
[1]  Huang, S.-S. The Problem of Life in the Universe and the Mode of Star Formation. Publications of the Astronomical Society of the Pacific 71, 421 (1959).
[2]  Kasting, J. F. & Catling, D. Evolution of a Habitable Planet. Annual Review of Astronomy and Astrophysics 41, 429–463 (2003).
[3]  Zahnle, K. et al. Emergence of a Habitable Planet. Space Science Reviews 129, 35–78 (2007).
[4]  Marcy, G. W. & Butler, R. P. Detection of extrasolar giant planets. Annual Review of Astronomy and Astrophysics 36, 57–97 (1998).
[5]  Gillon, M. et al. Seven temperate terrestrial planets around the nearby ultracool dwarf star TRAPPIST-1. Nature 542, 456–460 (2017).
[6]  Korenaga, J. Plate tectonics and planetary habitability: current status and future challenges. Annals of the New York Academy of Sciences 1260, 87–94 (2012).
[7]  Goodwin, A. M. Precambrian geology: the dynamic evolution of the continental crust (Academic Press, 1991).
[8]  Hoffmann, P. F. Precambrian geology and tectonic history of North America. Geology of North America—An Overview (Geological Society of America, 1989).
[9]  Condie, K. C. Supercontinents and superplume events: distinguishing signals in the geologic record. Physics of the Earth and Planetary Interiors 146, 319–332 (2004).
[10]  Dhuime, B., Hawkesworth, C. J., Delavault, H. & Cawood, P. A. Continental growth seen through the sedimentary record. Sedimentary Geology 357, 16–32 (2017).
[11]  Spencer, C. J., Roberts, N. M. W. & Santosh, M. Growth, destruction, and preservation of Earth’s continental crust. Earth Science Reviews 172, 87–106 (2017).
[12]  Korenaga, J. Initiation and Evolution of Plate Tectonics on Earth: Theories and Observations. Annual Review of Earth and Planetary Sciences 41, 117–151 (2013).
[13]  van Hunen, J., van Keken, P. E., Hynes, A. & Davies, G. F. Tectonics of early Earth: Some geodynamic considerations. In Special Paper 440: When Did Plate Tectonics Begin on Planet Earth?, 157–171 (Geological Society of America, 2008).
[14]  Gerya, T. Precambrian geodynamics: Concepts and models. Gondwana Research 25, 442–463 (2014).
[15]  Jain, C., Rozel, A. B., Tackley, P. J., Sanan, P. & Gerya, T. V. Growing primordial continental crust self-consistently in global mantle convection models. Gondwana Research 73, 96–122 (2019).
[16]  Labrosse, S. & Jaupart, C. Thermal evolution of the Earth: Secular changes and fluctuations of plate characteristics. Earth and Planetary Science Letters 260, 465–481 (2007).
[17]  Herzberg, C. & Gazel, E. Petrological evidence for secular cooling in mantle plumes. Nature 458, 619–622 (2009).
[18]  Jahn, B.-M., Glikson, A. Y., Peucat, J. J. & Hickman, A. H. REE geochemistry and isotopic data of Archean silicic volcanics and granitoids from the Pilbara Block, Western Australia: implications for the early crustal evolution. GEOCHIMICA ET COSMOCHIMICA ACTA 45, 1633–1652 (1981).
[19]  Drummond, M. S. & Defant, M. J. A model for Trondhjemite-Tonalite-Dacite Genesis and crustal growth via slab melting: Archean to modern comparisons. Journal of Geophysical Research 95, 21503–21521 (1990).
[20]  Barker, F. & Arth, J. G. Generation of trondhjemitic-tonalitic liquids and Archean bimodal trondhjemite-basalt suites. Geology 4, 596 (1976).
[21]  Martin, H. Effect of steeper Archean geothermal gradient on geochemistry of subduction-zone magmas. Geology 14, 753 (1986).
[22]  Moyen, J.-F. The composite Archaean grey gneisses: Petrological significance, and evidence for a non-unique tectonic setting for Archaean crustal growth. LITHOS 123, 21–36 (2011).
[23]  Christensen, U. R. & Hofmann, A. W. Segregation of subducted oceanic crust in the convecting mantle. Journal of Geophysical Research 99, 19867–19884 (1994).
[24]  Xie, S. & Tackley, P. J. Evolution of helium and argon isotopes in a convecting mantle. Physics of the Earth and Planetary Interiors 146, 417–439 (2004).
[25]  Rozel, A. B., Golabek, G. J., Jain, C., Tackley, P. J. & Gerya, T. Continental crust formation on early Earth controlled by intrusive magmatism. Nature 545, 332–335 (2017). 4
[26]  Crisp, J. A. Rates of Magma Emplacement and Volcanic Output. Journal of Volcanology and Geothermal Research 20, 177–211 (1984).
[27]  Armstrong, R. L. Radiogenic Isotopes: The Case for Crustal Recycling on a Near-Steady-State No-Continental-Growth Earth. Philosophical Transactions of the Royal Society of London Series A: Mathematical Physical and Engineering Sciences 301, 443–472 (1981).
[28]  Tang, M., Chen, K. & Rudnick, R. L. Archean upper crust transition from mafic to felsic marks the onset of plate tectonics. Science 351, 372–375 (2016).
[29]  Bédard, J. H. A catalytic delamination-driven model for coupled genesis of Archaean crust and sub-continental litho- spheric mantle. GEOCHIMICA ET COSMOCHIMICA ACTA 70, 1188–1214 (2006).
[30]  Johnson, T. E., Brown, M., Gardiner, N. J., Kirkland, C. L. & Smithies, R. H. Earth’s first stable continents did not form by subduction. Nature 543, 239–242 (2017).
[31]  Van Kranendonk, M. J., Collins, W. J., Hickman, A. & Pawley, M. J. Critical tests of vertical vs. horizontal tectonic models for the Archaean East Pilbara Granite–Greenstone Terrane, Pilbara Craton, Western Australia. Precambrian research 131, 173–211 (2004).
[32]  Fischer, R. & Gerya, T. Early Earth plume-lid tectonics: A high-resolution 3D numerical modelling approach. Journal of Geodynamics 100, 198–214 (2016).
[33]  Lourenco, D. L., Rozel, A. B., Gerya, T. & Tackley, P. J. Efficient cooling of rocky planets by intrusive magmatism.Nature Geoscience 215, 1–6 (2018).
[34]  Hirth, G. & Kohlstedt, D. L. Water in the oceanic upper mantle: implications for rheology, melt extraction and the evolution of the lithosphere. Earth and Planetary Science Letters 144, 93–108 (1996).
[35]  Mei, S. & Kohlstedt, D. L. Influence of water on plastic deformation of olivine aggregates: 1. Diffusion creep regime.Journal of Geophysical Research 105, 21457–21469 (2000).
[36]  Hall, C. E. & Parmentier, E. M. Influence of grain size evolution on convective instability. Geochemistry, Geophysics, Geosystems 4, 469 (2003).
[37]  King, S. Archean cratons and mantle dynamics. Earth and Planetary Science Letters 234, 1–14 (2005).
[38]  Rozel, A. Impact of grain size on the convection of terrestrial planets. Geochemistry, Geophysics, Geosystems 13 (2012).
[39]  Gillmann, C. & Tackley, P. Atmosphere/mantle coupling and feedbacks on Venus. Journal of Geophysical Research: Planets 119, 1189–1217 (2014).
[40]  Foley, B. J. & Driscoll, P. E. Whole planet coupling between climate, mantle, and core: Implications for rocky planet evolution. Geochemistry, Geophysics, Geosystems 17, 1885–1914 (2016).