In this study we systematically examined deep tremor triggered by large teleseismic earthquakes between 2001 and 2010 in three tectonically active regions along the SAF system. A total of 42 large teleseismic earthquakes [Table S1] were used in this study. In Central California (CC), 12 large teleseismic events triggered tremor along the Parkfield-Cholame segment [Table S2] of the San Andreas Fault (SAF) during the study period. On the other hand, only the 2002 Mw7.9 Denali Fault earthquake triggered clear tremor in Northern California (NC) and Southern California (SC) [Table S2]. Even though the 2010 Mw8.8 Chile earthquake triggered local earthquake in the Coso geothermal fields, the Chile event did not trigger tremor in NC and SC [Figure S1, Figure S2]. In addition, we found a positive correlation between the median tremor amplitudes and the Peak Ground velocity (PGV) of surface waves [Figure S2]. Regarding the tremor amplitude, we also corrected for the effects of geometrical spreading and attenuation with a constant Q of 100. Their correlations remain largely unchanged with and without corrections [Figure S3].
Our results revealed a marked difference in triggering behavior in CC, NC, and SC. The result suggests that the background noise level likely contributes to, but is not the primary cause of the different triggering behaviors [Figure S2]. In addition to background noise, we also considered the effects of path and site effects on tremor amplitudes. We randomly chose 2 micro earthquakes in three regions [Table S3] and computed the S-wave spectra [Figure S4]. The results suggest that the path and site effects also cannot completely explain the difference in tremor amplitudes for the frequency range of 2-8 Hz in these regions. Because ambient tremor is very active in CC and no ambient tremor reports in NC and SC, we suggest the lack of widespread triggering of tremor in NC and SC is likely related to the low background tremor rate in these regions.
Table S1. List of the information and measured parameters for all 42 teleseismic earthquakes used in this study. See the title within the file for the details of each field.
Table S2. List of triggered tremor locations used in this study.
Table S3. List of local earthquakes used in Figure S4.
Figure S1. (a) (Top) 2-8 Hz band-pass-filtered vertical seismograms showing the triggered earthquake signals in Coso by the 2010 Mw8.8 Chile earthquake at the Calaveras Fault (CF) in Northern California (NC). The seismograms are plotted along the strike of the CF, with northwest at the top and southeast on the bottom. The along-strike distance to the tremor source in NC triggered by the 2002 Mw 7.9 Denali Fault earthquake, and the station names are marked by the seismograms. The event name and the occurrence year, its magnitude (M), and the epicenter distance (Dist) and back azimuth (BAZ) relative to the broadband station are shown on the top. (Bottom) A comparison between the instrument-corrected transverse (T), radial (R), and vertical (Z) velocity seismograms and the 2-8 Hz band-pass-filtered seismograms recorded at the broadband station MHC. The zero time corresponds to the origin time of the mainshock. The velocity seismograms have been time-shifted back to the tremor sources. The adjusted times of Love waves (in T component), Rayleigh waves (in R and Z components), and tremor are marked below the station names. The thick vertical bar marks the amplitude scale of surface waves. (b) (Top) 2-8 Hz band-pass-filtered vertical seismograms showing the triggered earthquake signals in Coso by the 2010 Mw8.8 Chile earthquake at the San Jacinto fault (SJF) in the Anza network in Southern California (SC). The seismograms are plotted along the strike of the SJF to the tremor source in SC triggered by the 2002 Mw 7.9 Denali Fault earthquake, with northwest at the top and southeast on the bottom. (Bottom) A comparison between the velocity and the 2-8 Hz band-pass-filtered seismograms recorded at the broadband station RDM. Other notations are the same as in Figure S1a.
Figure S2. Maximum peak ground velocity (PGV) of surface wave (horizontal axis) in vertical component versus median amplitude (vertical axis) of the 2-8 Hz 3-components band-pass-filtered envelope functions during the surface waves (a-c) and signal to noise ratio (SNR) (d-f) at broadband stations PKD, RDM, and MHC in Central California (CC), Southern California (SC), and Northern California (NC), respectively. The event number with star symbol (*) indicates the measurement is from the nearby station FROB in CC and DNR in SC (Table S1, available as an electronic supplement to this paper). The tremor-triggering and non-triggering events were marked by shaded and open symbols, respectively. A total of 42 earthquakes were selected among the three regions. The numbers mark the measurements of 12 triggering events in CC. The background noise level of each event is calculated from a 600-s time window before the arrival of the predicted P-wave. The correlation coefficient (c.c.) and slope of the fitting line between the median tremor amplitudes and the PGVs for tremor and non-tremor in each region are marked in (a-c). The light gray background shows the median noise level for all events. The horizontal dotted lines in (a-c) mark the 1.5 times of the median noise level in each region. The vertical dotted lines in (a, d) show the apparent triggering amplitude threshold of 0.03 cm/s in CC (Peng et al., 2009). The horizontal dotted lines in (d-f) marks the 1.5 of SNR.
Figure S3. Tremor amplitude versus PGV in Central California (CC) before (a) and after (b) distance correction.
Figure S4. S-wave spectra of local earthquakes (thick lines) and the corresponding noise spectra (dotted lines) recorded at stations PKD, RDM, and MHC in Central California (CC), Southern California (SC), and Northern California (NC), respectively.
[ Back ]
