Subsections

Tidal Triggering of LFEs near Parkfield, CA

Amanda M. Thomas, Roland Bürgmann, and David Shelly

Introduction

Studies of nonvolcanic tremor (NVT) in Japan, Cascadia, and Parkfield, CA have established the significant impact small stress perturbations, such as the solid earth and ocean tides, have on NVT generation (Thomas et al., 2009). Similar results irrespective of tectonic environment suggest that extremely high pore fluid pressures are required to produce NVT. Here we analyze the influence of the solid earth and ocean tides on a catalog of  500,000 low frequency earthquakes (LFE) distributed along a 150km section of the San Andreas Fault centered at Parkfield (Shelly and Hardebeck, 2010). LFEs comprising the tremor signal are grouped into families based on waveform similarity and precisely located using waveform cross-correlation. Analogous to repeating earthquakes, LFE families are thought to represent deformation on the same patch of fault. While the locations of repeating earthquakes are assumed to be coincident with the location of asperities in the fault zone, NVT occurs below the seismogenic zone, where fault zones behave ductilely. Here we explore the sensitivity of each of these LFE families to the tidally induced shear (right-lateral shear stress, RLSS), normal (fault-normal stress, FNS), and Coulomb (CS) stresses on the SAF.

Methods

Tidally induced strains are computed in the LFE source region using SPOTL. Assuming two-dimensional plane strain and linear elasticity, with an elastic modulus of 30 GPa and Poisson ratio of 0.25, strains are then converted to stresses and resolved into fault normal and parallel (shear) directions of the San Andreas fault (N$^{\circ}$45W), and the volumetric strain is converted to pressure. Timeseries of the resolved stresses are then used to compute the percent excess [=(actual number of LFEs that occur under a particular loading condition-expected number)/expected number] for each stress constituent. We compute the percent excess, or Nex, for each stressing condition. Additionally, relative tremor rates during times when the tides are encouraging or retarding failure can be used to estimate the effective normal stress (Dieterich, 1986). The precisely relocated LFE families allow us to map spatial variation of the aforementioned quantities in the deep San Andreas fault.

Preliminary Results

Preliminary results indicate that extremely small stresses induced in the lithosphere by the tides are sufficient to trigger/modulate LFE families on the deep SAF. Additionally, precise LFE locations coupled with tidal influence on LFE families can be used to produce maps of along-fault spatial variability in tidal sensitivity, friction, and effective normal stress (Figure 2.12). The tidally induced RLSS has the most robust influence on LFE generation, however many families also show statistically significant correlation or anti-correlation with FNS. All families exhibit near-lithostatic pore pressure.

Future research efforts will focus on using tidal sensitivity of LFEs to place controls on the mechanical properties and behavior of deep fault zones.

Acknowledgements

This work is funded by the United States Geological Survey and a National Science Foundation Graduate Research Fellowship.

References

Dieterich, J.H., Nucleation and triggering of earthquake slip: effect of periodic stresses , Tectonophysics, 144, 127-139, 1986.

Shelly, D.R. and J.L. Hardebeck, Precise tremor source locations and amplitude variations along the lower-crustal central San Andreas Fault, Geophys. Res. Lett., 37, L14301, 2010.

Thomas, A.M., R.M. Nadeau, and R. Bürgmann, Tremor-tide correlations and near-lithostatic pore pressure on the deep San Andreas fault, Nature, 462, 1048-1051, 2009.

Figure 2.12: (a) Along-fault cross section of the SAF viewed from the south-west. Vertically exaggerated topography is shown in grey. Local towns are marked by inverted triangles. Hypocenters of SAF seismicity, the 2004 Parkfield earthquake, and LFE locations are shown as blue dots, yellow star, and red circles respectively. Panels (b) and (c) are delineated by the green box. (b) LFE locations color coded by their FNS Nex (percent excess = [actual number of LFEs during times of positive FNS - expected number of LFEs during times of positive FNS]/expected number of LFEs during times of positive FNS). (c) LFE locations color coded by the RLSS Nex values. (See color version of this figure at the front of the research chapter.)
\begin{figure*}\begin{center}
\epsfig{file=amthomas10_1_1.eps, width=14.0cm}\end{center}\end{figure*}

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail: www@seismo.berkeley.edu
© 2007, The Regents of the University of California