Mantle Rheology and Plate Tectonics: Damage and Inheritance

Abstract:

The specific rheology of the lithosphere that allows the existence of plate tectonics on Earth is poorly understood. This rheology must explain why plate tectonics initiated on Earth but not on Venus, which has very similar size and gravity, but also how changes in plate motion can occur on time scales more rapid than that of the deeper convection (~100 myrs). We suggest that a key ingredient of this rheology is the coupling of the grain size evolution of the polycrystalline rocks that constitute the mantle, with the flow. We demonstrate this process using a grain evolution and damage mechanism (Bercovici and Ricard, 2012) with a composite rheology, which are compatible with field and laboratory observations. Our model predicts that the grain size is controlled by an dynamic equilibrium where deformation tends to reduce the grain sizes which otherwise would spontaneously increase. The presence of secondary phases is crucial as it eases the grain size reduction and inhibits their too fast recoveries. The resulting time-dependent rheology is non-linear, leads to localisation on time scales much shorter than those of the typical transit time of mantle convection but has also a long healing time. These characteristics of fast damage and long term memory is shown in simple models to spontaneously generate plates on a planet with Earth's parameters. For Venus hotter surface conditions, accumulation and inheritance of damage is negligible; hence only subduction zones survive and plate tectonics does not spread. This rheology also explains how subducting slabs can be rapidly detached when a buoyant crust is drawn into subduction and therefore how rapid plate changes can occur.