Evolving viscous anisotropy in the upper mantle and its geodynamic implications

Abstract:

Asthenospheric shear causes some minerals, particularly olivine, to develop anisotropic textures that can be detected seismically. In laboratory experiments, these textures are also associated with anisotropic viscous behavior, which should also be important for geodynamic processes. To examine the role of anisotropic viscosity in asthenospheric deformation, we developed a numerical model of coupled anisotropic texture development and anisotropic viscosity, both calibrated with laboratory measurements of olivine aggregates. This model characterizes the time-dependent coupling between large-scale formation of lattice preferred orientation and changes in asthenospheric viscosity for a series of deformation paths that roughly represent upper-mantle geodynamic processes. Consistent with previous studies, we find that texture development beneath a moving plate tends to align the a-axes of olivine with the plate motion direction, reducing the effective viscosity in this direction and increasing the plate velocity for a given driving force. We demonstrate that, in association with this texture development, the effective viscosity increases for horizontal shear in the direction perpendicular to the a-axes. This increase in viscosity in the perpendicular direction should provide resistance to changes in plate motion and associated texture development over periods of 10s of Myrs. However, the same well-developed texture due to horizontal shearing results in a viscosity decrease for vertical shearing. Therefore, a sub-lithospheric olivine texture with horizontal a-axes can quickly evolve to have vertical a-axes in response to vertical shearing. These results illustrate the coupled nature of anisotropic viscosity and texture formation. With these simplistic first-order models, we demonstrate the importance of anisotropic viscosity and olivine texture development on geodynamic processes and provide a stepping stone for integrating viscous anisotropy into larger scale geodynamic models.