par O'Neill, C.;Debaille, Vinciane ;Griffin, W.
Référence American journal of science, 313, page (912-932)
Publication Publié, 2013
Article révisé par les pairs
Résumé : The Hadean mantle was efficiently heated from high internal heat production, high rates of impact bombardment, and primordial heat from accretion. As a result a strong case is made for extremely high internal temperatures, low internal viscosities, and extremely vigorous mantle convection. Previous studies of mixing in such high-Rayleigh number convective environments indicate that chemically heterogeneous mantle anomalies should have efficiently remixed into the mantle on timescales of less than 100 Myr. However, 142Nd and 182W isotope studies indicate that heterogeneous mantle domains survived, without mixing, for over 2 Gyr-at odds with expected mixing rates. Similarly, concentrations of platinum group elements in Archean komatiites, purportedly due to the later veneer of meteoritic addition on the Earth, only achieve current levels at 2.7 Ga-indicating a time lag of almost 1 to 2 Gyr in mixing this material thoroughly into the mantle. Previous studies have sought to explain slow Archean mantle mixing via mantle layering due to endothermic phase changes, or anomalously viscous blobs of material, with limited efficacy. Here we pursue another explanation for inefficient mantle mixing in the Hadean: tectonic regime. A number of lines of evidence suggest that resurfacing in the Archean was episodic, and extending these models to Hadean times implies the Hadean was characterized by long periods of tectonic quiescence. We explore mixing times in 3D spherical-cap models of mantle convection, which incorporate vertically stratified and temperature-dependent viscosities. We show that mixing in stagnant-lid regimes is, at the extreme, over an order of magnitude less efficient than mobile-lid mixing, and for plausible Rayleigh numbers and internal heat production, the lag in Hadean convective recycling can be explained. The attractiveness of this model is that it not only explains the long-lived 142Nd and 182W anomalies, but also posits an explanation for the delay between accretion of the late veneer-between 4.5 to 3.8 Ga on a stagnant surface- and its fully mixed signature apparent in elevated PGEs in 2.7 Ga komatiites. It also provides an explanation for the 400 Myrs of immobility of the mafic protolith from which the Jack Hill zircons were sourced, and retards early heat loss from the mantle, providing a solution to the "Archean thermal catastrophe" of parameterized Earth evolution models.