Critical Parameters of the Initiation Zone for Spontaneous Dynamic Rupture Propagation

Tuesday, 16 December 2014
Martin Galis1, Christian Pelties2, Jozef Kristek3,4, Peter Moczo3,4, Jean Paul Ampuero5 and Paul Martin Mai1, (1)King Abdullah University of Science and Technology, Thuwal, Saudi Arabia, (2)Ludwig Maximilian University of Munich, Munich, Germany, (3)Comenius University, Bratislava, Slovakia, (4)Slovak Academy of Sciences, Geophysical Institute, Bratislava, Slovakia, (5)California Institute of Technology, Pasadena, CA, United States
Numerical simulations of rupture propagation are used to study both earthquake source physics and earthquake ground motion. Under linear slip-weakening friction, artificial procedures are needed to initiate a self-sustained rupture. The concept of an overstressed asperity is often applied, in which the asperity is characterized by its size, shape and overstress. The physical properties of the initiation zone may have significant impact on the resulting dynamic rupture propagation. A trial-and-error approach is often necessary for successful initiation because 2D and 3D theoretical criteria for estimating the critical size of the initiation zone do not provide general rules for designing 3D numerical simulations. Therefore, it is desirable to define guidelines for efficient initiation with minimal artificial effects on rupture propagation. We perform an extensive parameter study using numerical simulations of 3D dynamic rupture propagation assuming a planar fault to examine the critical size of square, circular and elliptical initiation zones as a function of asperity overstress and background stress. For a fixed overstress, we discover that the area of the initiation zone is more important for the nucleation process than its shape. Comparing our numerical results with published theoretical estimates, we find that the estimates by Uenishi & Rice (2004) are applicable to configurations with low background stress and small overstress. None of the published estimates are consistent with numerical results for configurations with high background stress. We therefore derive new equations to estimate the initiation zone size in environments with high background stress. Our results provide guidelines for defining the size of the initiation zone and overstress with minimal effects on the subsequent spontaneous rupture propagation.