SCEC Award Number 12024 View PDF
Proposal Category Collaborative Proposal (Integration and Theory)
Proposal Title A Search for Seismicity Patterns that Reflect Mechanical Erosion of Locked Patches Due to Creep
Name Organization
Allan Rubin Princeton University Maximilian Werner Princeton University
Other Participants Maximilian J. Werner is a post doctoral researcher who works with Allan Rubin. He will be involved and funded from this proposal
SCEC Priorities 2b, 2c, 4e SCEC Groups EFP, Seismology, FARM
Report Due Date 03/15/2013 Date Report Submitted 10/14/2015
Project Abstract
The purpose of this project was to illuminate the relationship between microseismicity ``streaks" and neighboring, predominantly aseismic ``holes" that are observed on creeping faults in northern California. Two hypotheses have been proposed to describe this relationship: (some) streaks mark the boundary between frictionally locked and aseismically creeping patches; and/or some are sandwiched between two creeping patches. Simple analytical expressions from fracture mechanics suggest that bounding streaks experience a growing stress concentration, a change in conditions that might be reflected in microseismicity patterns such as increasing rates, moment release or maximum magnitudes. We identified dozens of streaks on the creeping section of the San Andreas fault and the Calaveras fault and analyzed their microseismicity for such patterns, including a systematic study of the behavior of repeating earthquake sequences. We found no obvious signals on streaks that would betray mechanical erosion of locked patches by aseismic creep. In particular, streaks with moderate earthquakes that are inferred to have ruptured beyond the streak and thus indicate a frictionally locked neighboring hole do not show evidence for accelerating seismicity, increasing moment release or maximum magnitudes, or systematically increasing moment release rates in repeating earthquake sequences. As a result, we are reassessing how locked patches respond to frictional erosion by aseismic creep through numerical simulations on rate-state faults. These simulations suggest that the simple analytical expressions do not reflect the complexity of the response, and thus could explain why no obvious signal was found in the data.
Intellectual Merit The results of this project have contributed to the empirical characterization of streaks and holes on northern California’s creeping faults and to our mental picture of the microseismicity patterns that might reflect the gradual
loading of locked fault patches due to aseismic creep. We showed that patterns of increasing seismic rates, moment release rates or maximum magnitudes (patterns expected from considerations of the growing stress concentration at the locked/creeping boundary) are not robustly observed on individual streaks that were inferred to be at the boundary of locked and creeping patches. This lack of a signal has prompted an analysis of simulated microseismicity patterns in numerical models of the boundary of locked/creeping fault patches endowed with rate-state friction and may advance our understanding of how frictional erosion of locked fault patches by creep is reflected in microseismicity.
Broader Impacts This project has supported the training of a postdoctoral researcher. In addition, the development of methods to characterize seismicity patterns that reveal mechanical erosion of locked fault patches by creep from below might
eventually help in identifying those strike-slip faults that are nearing the end of their earthquake cycle and that are therefore more likely to rupture in large earthquakes than others.
Exemplary Figure Figure 5: Streaks and predominantly aseismic holes along a section of the northern San Andreas fault. Solid symbols denote identified repeating earthquake sequences (RES). Red symbols denote RES with shortening recurrence
interval trends and increasing moment trends. Blue symbols indicate RES with lengthening recurrence intervals
and decreasing moments. Black symbols denote the remaining RES. We observe no spatially coherent evidence
for increased moment release rates of RES on contiguous streaks. Such a pattern would be expected from simple analytical expressions of the growing shear stress in response to loading from aseismic creep at the boundary of
locked and aseismically creeping fault patches. Recent numerical simulations by Werner and Rubin (2013) of the
mechanical erosion on deformable rate-state faults show a more complex response that could explain the absence
of the expected patterns. [Figure from Werner and Rubin, 2011]