Testing the Strain-rate Hypothesis for Deep Slab Seismicity

<p>The occurrence of deep earthquakes within subducting lithosphere (slabs) remains enigmatic because these earthquakes have many similarities to shallow earthquakes, yet frictional failure is strongly inhibited at high pressure. Regardless of depth, earthquakes occur where the temperature is cold enough that elastic deformation is accumulated over time: for frictionally controlled earthquakes at shallow depth, the rate of seismic moment release is correlated with the strain-rate. Comparison of spatial variation in strain-rate magnitude from 2D simulations of subduction to observed seismicity versus depth profiles suggest that strain-rate may also be a determining factor in the occurrence of deep slab seismicity (1). In addition, proposed mechanisms for deep earthquakes, including transformational faulting of metastable olivine and thermal shear instability, are known to depend directly on strain-rate. To test the hypothesis that strain-rate is a determining factor in the spatial distribution of deep earthquakes, we are creating 2D models of subduction with visco-elasto-plastic (VEP) rheology and a free surface in the software ASPECT (2). The 2D slab structure is constructed for specific locations in which the slab geometry is extracted from Slab 2.0 (3) and the plate age and convergence rate are used to define the thermal structure using a new mass-conserving slab temperature model (4) implemented in the Geodynamic WorldBuilder (5). The resulting strain-rate and stress, together with the pressure and temperature along multiple transects of the slab are used as input values for a 1D thermal shear instability model (6) using the same VEP rheology as the slab deformation models. &#160;Using this approach we can test whether the conditions in the slab favor failure through thermal shear instability and compare the spatial distibution to obsered seismicity. Initial results of this workflow will be presented, including how we have overcome some of the challenges in running VEP models for comparison to present-day slab seismicity. References: 1. Billen, M. I. , Sci. Advances, 2020. 2. Bangerth, W. et al., https://doi.org/10.5281/ZENODO.5131909, 2021. 3. Hayes, G.P. et al., Science, 2018. 4. Billen, M. I. and Fraters, M. R. T., EGU Abstract, 2022. 5. Fraters, M. R. T. et al., Solid earth, 2019. 6. Thielmann, M. Tectonophysics, 2018.&#160;</p><p>&#160;</p>