Modeling High Stress Drops, Scaling, Interaction and Irregularity of Repeating Earthquake Sequences Near Parkfield

Date: 4/26/2019

Time: 02:00 PM

Room: Elliott Bay

Repeating earthquake sequences have been actively investigated to clarify many aspects of earthquake physics and mechanics. In particular, small repeating earthquakes present a rare predictable opportunity for detailed observation and study because of their short recurrence times and known location. For example, two well-studied sequences along the San Andreas Fault, known as the Los Angeles and San Francisco repeaters, have several intriguing observations: (1) long (for the seismic moment) recurrence times that would suggest stress drops of 300 MPa based on typical assumptions, (2) near-synchronized timing prior to 2004, and (3) higher than typical inferred stress drops (of 25 to 65 MPa, up to 90 MPa locally), but not as high as the recurrence times suggest.

Our study shows that all these observations are self-consistent, i.e. they can be reproduced in a single fault model. The suitable models build on the standard rate-and-state fault models, with velocity-weakening patches imbedded into a velocity-strengthening region. To allow for the higher stress drops, either enhanced dynamic weakening during seismic slip or elevated normal stress, or both, are added to the velocity-weakening patches. Such models are able to match the observed average source properties of the San Francisco and Los Angeles repeaters, as well as the overall nontrivial scaling between the recurrence time and seismic moment exhibited by many repeating sequences as a whole, for reasonable parameter choices based on experiments and theoretical studies. One key finding is the substantial and variable occurrence of aseismic slip at the locations of the repeating sources among these models, which explains their atypical relation between recurrence interval and seismic moment, induces variability in the repeating source properties as observed, and results in their neither slip- nor time-predictable behavior.

Presenting Author: Semechah K. Y. Lui

Authors