SCEC Award Number 16001 View PDF
Proposal Category Individual Proposal (Integration and Theory)
Proposal Title Nonlinear attenuation of strong seismic waves, intact rock as a fragile geological feature, and modulation of tectonic strain by strong ground motions
Investigator(s)
Name Organization
Norman Sleep Stanford University
Other Participants
SCEC Priorities 6e, 6b, 2d SCEC Groups Geology, GMP, Seismology
Report Due Date 03/15/2017 Date Report Submitted 03/08/2017
Project Abstract
 Strong seismic waves bring the subsurface beyond its elastic limit. In particular, surface waves produce strain that varies slowly with depth. The dynamic stress is strain times the stiffness. Failure occurs when this stress exceeds the frictional strength, cracking the rock and reducing the stiffness. The stiffness self-organizes with depth in the damaged zone so that failure is just reached in strong events, yielding an estimate of past shaking. We have used borehole data to confirm the predicted effects on stiffness from finite past water table depth and coefficients of friction depending on clay content occur. We have numerically modeled strong vertical S-waves at shallow depths. We used scaling relationships to study the rupture tip and the near fault environment of large earthquakes. We documented nonlinear interaction of S-waves with Rayleigh waves, which constrains three-dimensional frictional failure criteria.
Intellectual Merit We studied nonlinear attenuation and the associated damage from rock failure to constrain past and present shaking from strong events. We confirmed that the predicted effects on past rock damage from variations of the coefficient of friction from clay content and finite past water table depth do occur. The method yields past peak dynamic strain and with input from SCEC community models peak ground velocity (PGV). We obtained past PGV of 1-2 m/s in Los Gatos, Parkfield, and near LAX airport. We obtained 0.5 m/s west of Whittier Narrows in agreement with SCEC nonlinear models. The method also constrains future PGV by constraining the threshold of strong nonlinear attenuation. We modeled nonlinear vertical S-waves in the shallow layered subsurface. We used heat flow to constrain fault-normal stress near Parkfield and its effects on the rupture tip of large San Andreas events. We continue to study how the effective slip-weakening distance may self-organize on major faults.
Broader Impacts Our work centers on the amplitudes of past and future shaking and the physics of the rupture tip, this is, central SCEC priorities. Our scaling relationships help in planning massive numeral calculations and understanding their results. Our work on the fine-scale features of the rupture tip may help with finding equivalent larger-scale material properties for dynamic rupture models.

Our nonlinear scaling relationships are applicable to the interaction of strong tidal stresses and modest tectonic stresses on icy satellites.

The PI with Peter Olds and junior college students (3 from underrepresented minorities) studied K-Pg outcrops near Trinidad CO. We found two exposures of a normal fault that slipped about ~1 m just once from extreme seismic waves from the impact. This observation confirms the reality of extreme shaking and is an example of an induced earthquake. In addition, such induced faults from tectonic earthquakes are quite rare or nonexistent. Thus, extreme seismic waves are also quite rare.

We have flagged the importance of maintaining water levels at Whittier Narrows so that strong Love waves from San Andres events continue to nonlinearly attenuate reducing the shaking in Downtown Los Angeles. The Watermaster is already doing this for other reasons and earthquake risk is already a legal criterion for California water management. It should be possible to overpressure sealed aquifers in the Los Angeles Basin to nonlinearly damp strong S-waves and thus base isolate the city. However, it is royally premature to consider putting a plan into action.

Our paleoseismic methods provide estimates of past peak dynamic strain at several sites near Parkfield. We will compare these predictions with SCEC community models of large earthquakes that break the Parkfield segment. It may be possible to constrain whether the creeping zone breaks. Our methods also constrain the occurrence of rogue events that break deep into the ductile zone.
Exemplary Figure Figure 3. Resolved acceleration for nonlinear models: shallow layer over half space with no fluid pressure (thin lines) and with an artesian aquifer between 100 and 150 m depth (thick lines). Longer signal is windowed between 3 and 4 s. Nonlinear attenuation in the model without an aquifer clips an acceleration equal to the effective coefficient of friction. Nonlinear attenuation occurs within the deep aquifer but the signal is not simply clipped. Deliberately overpressuring a deep aquifer could partially base isolate the surface from strong S waves, for example, within Downtown Los Angeles From [40].

[40] Sleep, N. H., and N. Nakata (submitted) Nonlinear attenuation S-waves by frictional failure at shallow depths. Bull. Seismological Soc. Am., 107, XXXX–XXXX, doi:XXXX. SCEC contribution number 7204.