Investigating the effects of plate-driving forces on observed surface deformation using global mantle flow models

Geodynamic models based on seismic tomography have been utilized to understand a wide range of physical processes in the Earth's mantle ranging from lithospheric stress states to plate-mantle interactions. However, the influence of various model components and the associated physical properties of the mantle on the observed surface deformation is still an open question and requires further research. In this study, we develop global mantle flow models based on high-resolution seismic tomography to quantify the relative importance of the plate driving and resisting forces on the surface motions. Our models include temperature and density variations based on seismic tomography, lithospheric structure, and the observed locations of subducted slabs, using the geodynamics software ASPECT. We use a diffusion/dislocation creep rheology with different parameters for the major mantle phases. To facilitate plate-like deformation, we prescribe weak plate boundaries at the locations given by global fault databases. We resolve the resulting strong viscosity variations using adaptive mesh refinement such that our global models have a minimum resolution of <10 km in the lithosphere. We analyze the influence of slab viscosity, plate boundary friction, asthenospheric viscosity, and plate boundary geometry on reproducing the observed GPS surface velocities. Our parameter study identifies model configurations that have up to 85% directional correlation and a global velocity mean within 10% difference with the observed surface motions. Our results also suggest that the modeled velocities are very sensitive to the plate boundary friction, particularly to variations in viscosity, dip angles, and the plate boundary geometry, i.e., open vs closed boundaries, or localized vs. diffused deformation zones. These models show the relative influence of plate-driving forces on the surface motions in general, and in particular the importance of using accurate models of plate boundary friction for reproducing observed plate motions. In addition, they can be used as a starting point to separate the influences of lithospheric structure and mantle convection on surface observables like strain rate, stress field, and topography.