A late onset of plate tectonics as a solution of the New Core Paradox.

The Earth has sustained a magnetic field for at least 3.4 billion years, generated by convective motions of liquid iron within the outer core. Maintaining such a long-lived geodynamo requires efficient cooling of the core. However, a high core thermal conductivity as suggested by experiments and theoretical calculations reduces the convective power available prior to inner-core nucleation, making the continuous persistence of the magnetic field more difficult. This apparent incompatibility between high thermal conductivity estimates and evidence for a ~3.4 Ga-long geodynamo is known as the “New Core Paradox”. Because core cooling is primarily controlled by heat transfer through the overlying solid mantle, an accurate quantification of the heat extracted from the core via mantle convection is therefore essential to resolving this paradox.

Today, the mantle cools efficiently mainly through plate tectonics, via subduction of cold large plates. But when plate tectonics actually began is still debated. Did it start soon after Earth formed, around 4.5 billion years ago? Did it appear later, between 4 and 3 billion years ago? Or is it a more recent process, less than a billion years old?

We explored how different styles of mantle cooling would have influenced Earth’s thermal and magnetic history. We explored either a mobile surface like modern plate tectonics (i.e. mobile lid) or a less efficient, stagnant-lid-like regime (where the surface doesn’t move), or a transition from stagnant- to mobile-lid regime at a given time and with a given duration. This is important because how Earth’s mantle cooled over time is closely tied to its ability to keep generating a magnetic field.

We built a global model for the Earth coupling a core model including inner core formation and the possibility to form stably stratified layers, with a mantle model simulating different convective (hence cooling) regimes.

We performed a Markov Chain-Monte Carlo inversion using as constraints the present-day size of the inner core, the continuous 3.4 billion years old magnetic record, and the mantle potential temperature record. We inverted the viscosity parameters, tectonic transition parameters and core thermal conductivity. Our models successfully reproduce all the constraints for an onset between 4.0 Ga and 2.5 Ga, with a bimodal distribution characterized by a relatively early onset of mobile-lid convection with a long-duration transition, or a later onset with a more rapid transition to a mobile-lid regime. Our result show that the late and rapid transition case allows for a core thermal conductivity up to 110 W/m/K, providing a possible solution to the New Core Paradox.