Modelling Effects of Olivine Grain Size Evolution on Upper Mantle Dynamics

Abstract:

In geodynamic numerical models, grain-size-independent dislocation creep is commonly defined as the governing viscous flow law for the upper mantle. However, grain-size-dependent diffusion creep may become the dominant deformation mechanism if grain size is sufficiently small. Studies that utilize composite diffusion-dislocation creep rheologies and fixed grain size suggest that the upper mantle is stratified, with the dominant mechanism being dislocation creep at shallow depths and diffusion creep farther down. Studies with dynamic grain size evolution implemented for the upper mantle demonstrate that the contrary might be the case, where diffusion creep acts within the mantle lithosphere (due to smaller grain sizes produced under lower temperature/higher stress conditions) and dislocation creep in the asthenosphere below. Whether, and where diffusion creep dominates has important implications for the overall strength of the lithosphere and the dynamic evolution of lithospheric-scale extension and orogeny.

To investigate the importance of grain size and the effects of resulting viscous creep within the upper mantle, we developed a two-dimensional thermo-mechanical numerical code based on the finite difference method with a fully-staggered Eularian grid and freely-advecting Lagrangian markers. The model implies a composite diffusion-dislocation creep rheology and a dynamic grain size evolution model based on the paleowattmeter and including recently published olivine grain growth laws.

Preliminary results of upper mantle extension indicate olivine grain sizes of ~3 cm for large parts of the asthenosphere while in the lithosphere they range from ~1 μm at the Moho to ~1 cm at the LAB. This grain size distribution indeed indicates that diffusion creep dominates deformation in the lithosphere and dislocation creep in the asthenosphere. We furthermore test the implications of wet and dry olivine rheology and respective grain growth laws and interpret their effects on large-scale tectonic processes. Our results help explain deformation localization during extension by strain weakening related to grain size reduction and consequent diffusion creep activation.