Skip to main content

Modelling the Archaean Subduction Environment

Final Report Summary - MASE (Modelling the Archaean Subduction Environment)

The ERC-funded project MASE (Modelling the Archaean Subduction Environment) looks into the geodynamical-petrological environment of early Earth subduction. Subduction zones today are major sites for continental crust production, and in this ERC project, this process has been further explored for early Earth conditions. This ERC project has so far resulted in 27 publications published or in press in peer-reviewed international journals. We developed numerical codes (both diversifying existing thermo-mechanical tectonic modeling tools and developing a new porous media fluid flow code) that describe geodynamics, and combined those with thermodynamical databases that quantify petrological processes. The outcomes of these combined tools are compared directly to available field data.

Direct comparison of large-scale numerical models of Archaean subduction dynamics and geochemical signatures from the Archaean rock record suggests that Archaean subduction was a less robust feature than it is today, with a more episodic appearance in which subduction starts and stops on a few-million year timescale. Today, subducting slabs tend to temporarily stagnate in the transition zone, and previous work has argued that in an early, hotter Earth, this stagnation was more pronounced. But our modelling shows that, instead, slab penetration is facilitated in a hotter mantle. This might resolve a long-standing paradox on the early-Earth heat budget. The cycling of volatiles (such as water) between the Earth surface and deep interior plays a key role in the operation of plate tectonics. Today, Earth probably takes more water into the deep Earth through subduction than it releases at mid-ocean ridges, but it is commonly assumed that this was not so in an early hotter Earth, and that the early Earth had very little internal water. We show through combined geodynamical and petrological modelling that also the early Earth was able to absorb significant amounts of water, and may therefore not have been as dry as commonly assumed.

Melting of the crust subducting plates is often invoked to explain the geochemical composition of Archaean crustal rocks. Our modeling results show that such slab melting, if it is at the origin of Archaean crust, must be a much more complex process than assumed previously, with episodic crustal delamination, and a complex interplay between slab dehydration and melting in the slab and overlying mantle wedge. Other processes, involving melt segregation and fractional crystallization in the mantle wedge and overlying plate, are probably at least as important for crust formation. Continental collision sites have a high potential for long-term preservation of the rock record. So it is important to understand how Archaean continental collision manifested itself in that rock record. We explored the dynamics and magmatic consequences of such collision process and conclude that throughout Earth history, in a hotter or hydrated mantle wedge, sublithospheric small-scale convection is likely to play an important part in the production of post-collisional magmatism, whereas the consequences of slab break-off appear to play a smaller role.

Archaean subduction and continental collision have been suggested as scenarios to build thick cratonic keels, and therefore understanding the dynamics of these cratonic keels provides useful insight in the operaton of Archaean subduction. Our work indicates that only a moderate increase in strength and perhaps buoyancy is required to build stable cratonic roots We show that Archean convergence (e.g. through subduction) is a viable model for craton formation, which leads to two distinct craton thickening stages, one driven by tectonic shortening, and one by gravitational thickening.

One of the most commonly cited marker of Archaean subduction is in the form of so-called "arc magmatism". A large part of Archaean rocks, now occuring as deformed, "grey gneisses", has been interpreted as representing ancient subduction zones. However, we show that a large part of the gneisses are the product of reworking of preexisting rocks. Often analogies with modern post-collisional settings can be made. Archean crustal rocks form by melting from <30 to ~60-75 km. We propose that this corresponds to a range of environments, from intracrustal to subduction. We demonstrate that proper subduction events are short-lived (a few Myrs typically), as recorded by magmatism. Collectively, the evidence suggests a regime where a partially molten crust evolves and differentiates above very unstable subduction zones.

In summary, this ERC project has added significant detail to the wide range of processes that is associated with subduction and crust formation throughout Earth history.