Bounding of near-fault ground motion based on radiated seismic energy with a consideration of fault frictional mechanisms

The energy radiated as seismic waves strongly depends on the fault rupture process associated with rupture speed and dynamic frictional mechanisms involved in the fault slip motion. Following McGarr and Fletcher approach, we derived a physics-based relationship of the weighted average fault slip velocity vs apparent stress, rupture speed and static stress drop based on a dynamic circular fault model. The resultant function can be approximately used to bound near-fault ground motion and seismic energy associated with near-fault coseismic deformation. Fault frictional overshoot and undershoot mechanisms governed by a simple slip-weakening constitutive relation are included in our consideration by using dynamic rupture models named as M- and D-models and proposed by Madariaga (1976) and Boatwright. We applied the above function to the 2008 great Wenchuan earthquake and the 1999 Jiji (Chi-Chi) earthquake to infer the near-fault ground motion called slip weighted average particle velocity and obtained that such model-dependent prediction of weighted average ground velocities is consistent to the results derived from the near-fault strong motion observations. Moreover, we compared our results with the results by McGarr and Fletcher approach, and we found that the values of the weighted average particle velocities we obtained for these two earthquakes are generally smaller and closer to the values by direct integration of strong motion recordings of the near-fault particle velocity waveform data. In other words, if this result comes to be true, it would be a straightforward way used to constrain the near-fault ground motion or to estimate source parameters such as rupture speed, static and dynamic stress drops.