Development of a Magnitude-Scaling Law for Non-Volcanic Tremors in Cholame Near Parkfield, California

Friday, 19 December 2014
Nadine Staudenmaier1, Ben Edwards1, Thessa Tormann1, Aurélie Guilhem2 and Stefan Wiemer1, (1)ETH Swiss Federal Institute of Technology Zurich, Zurich, Switzerland, (2)CEA/DAM/DIF, F-91297 Arpajon, France
Non-volcanic tremors (NVTs) are observed in transition zones between freely slipping and locked sections of faults, and in particular along subduction zones. Since 2001 NVTs have also been recorded in the Cholame segment of the San Andreas fault, where the Berkeley Seismological Laboratory (BSL) provides origin location and duration estimates for observed events. Due to their deep location below the seismogenic zone (i.e., about 15-35 kilometers), they have the potential to serve as substitutes for the lack of earthquakes, providing us with insights into the deep crust and its stress conditions.

Our first analysis of the BSL NVT catalogue showed that, similarly to the Gutenberg-Richter law for earthquake magnitude, the number of events in different duration classes of NVTs in this region follows a power law. Subsequently, using a spectral analysis technique, we sought to derive a magnitude scale for the tremors based on their radiated energy, Me, as typically applied to earthquakes. In order to obtain sufficiently low noise levels across a broad range of frequencies we used the BSL High Resolution Seismic Network (HRSN), a borehole network with high sensitivity seismometers. We use the entire duration of the tremor signal (e.g., 3-21.5 min), assuming that it originates at the source provided by the BSL. Analysis of the obtained Fourier spectra for more than 3000 events reveal distinct sets of NVT: Some with stronger, and some with weaker high-frequency falloff. Due to the absence of a correlation with propagation distance, we speculate that this effect must be related to either the source or recording site. The independence of the effect from the recording stations strongly suggests a source effect. We first present preliminary results of the magnitude scaling of NVTs and its relation to duration. We then discuss possible explanations for the significant distinction between the different types of tremors.