[S-CG45] Science of slow-to-fast earthquakes

[SCG45-04] Near-field seismic observations of slip pulse durations and rupture stopping phases from large continental earthquakes

*Jesse Ryan Kearse¹, Yoshihiro Kaneko¹ (1.Kyoto University)

Keywords:slip pulse , rupture dynamics, strong motion

Whether seismic slip in natural earthquakes propagates as a narrow pulse (short rise time) or a crack (long rise time) is a fundamental question in seismology. There is compelling evidence from many large earthquakes that demonstrates that on average, local rise times are short compared to total event duration (~10-20%), thus supporting a pulse model. What is not clear, however, is the mechanism responsible for slip arrest. Does slip arrest occur due to the local frictional properties of the host fault (self-healing) or is the slipping fault arrested by the dynamic stress field returning from the fault edges (stopping phase)? To address this issue, we take advantage of the recent abundance of seismic observations from large continental earthquakes (Mw 7-8) to generate a catalog of slip-parallel ground velocity pulses in the near-field of the slipping fault (<3 km). Informed by dynamic rupture models, these records suggest local slip rise time on the fault is short (1 – 9 s) and is arrested by stopping phases originating from low velocity layers near the surface, and from lateral fault boundaries. Records from elongated strike-slip ruptures show the influence of rupture segmentation. Slip and rise time near the segment boundaries are truncated by the arrival of stopping phases, and are larger near the center of the fault segments, consistent with a crack model of rupture. We suggest that during large continental earthquakes, short rise times, and thus pulse-like rupture behavior is controlled by the propagation and arrest of crack-like rupture cascade across segmented fault networks.