ANIMA the journey: how we model olivine CPO-related anisotropic viscosity

The long-term fluid-like movements in the Earth’s mantle largely depend on the rheological behaviour of olivine, the main rock-forming mineral in the upper mantle. Although the average viscosity of the mantle can be estimated from post-glacial rebound or geoid anomalies, the micromechanical mechanisms that facilitate the deformation of the solid mantle have been identified from rock mechanics experiments. Dislocation creep emerges as the predominant deformation mechanism in the uppermost mantle, aligning olivine crystals into a crystallographic preferred orientation (CPO) parallel to the flow, while this alignment of crystals also results in anisotropic viscous behaviour. Thus, anisotropic viscosity and CPO evolve hand in hand, and this interaction may impact many geodynamic processes. For example, beneath tectonic plates CPO evolves parallel to the plate motion direction, weakening the asthenosphere in that direction. However, if the plate motion direction changes, the asthenosphere will resist this change, leading to smaller velocities, less deformation and therefore a slow evolution of the CPO towards the new plate motion direction. In the ANIMA project, we aimed to find an efficient way of modelling CPO evolution and the related anisotropic viscosity in a fully coupled way within a geodynamic simulation. We developed a method that tracks CPO evolution on advected particles based on the D-REX method and utilizes the eigenvalues of the mean CPO orientation matrices to predict the anisotropic viscous parameters. These parameters allow us to calculate a tensor form of the viscosity, which we then feed back into our model solution. This method can be applied in combination with other rheologies, although with a cost of having to represent the viscosity as a tensor in the entire model domain, regardless of the dominant deformation mechanism. Despite an estimated increase in computational cost by up to an order of magnitude, incorporating anisotropic viscosity coupled to CPO evolution stands feasible for regional geodynamic models. This development will facilitate the study of a broad new range of geodynamics problems that involve olivine texture and anisotropic viscosity.