Monitoring of Relative Velocity Change Following the 2014 Mw 6.0 South Napa Earthquake

Temporal Evolution in Seismic Velocity¶

We estimate relative velocity changes (dv/v) with MSNoise [Lecocq et al., 2014]. Our result is published at Geophysical Research Letters [Taira et al., 2015]. There are several arhived webpages showig the previous results: SNapa_Noise.09-16-14 (last updated on 09-16-14), SNapa_Noise.03-30-15 (last updated on 03-30-15), and SNapa_Noise.08-10-15 (last updated on 08-10-15).

Following Brenguier et al. [2008], our analysis focuses on waveforms in a frequency range of 0.1-0.9 Hz. A set of 8 stations are selected based on the high rate of data completeness. We determine the reference Green’s function (GF) for each station pair as the average of 1-day GFs obtained in the time interval, January 2013 through July 2014. The time history of dv/v is determined by measuring delay times between 5-day stacks of GFs and the reference GF.

We include horizontal component data for the dv/v measurement. The maximum available number of channel pairs (e.g., ZZ, ZN, ZE) is 108. We obtain about 0.08% velocity reduction following the South Napa earthquake (Figure 1). Note that dv/v is plotted at the end of the 5-day-length time window that is used for computing 5-day stacks of GFs. Threfore, dv/v values plotted within the first 4 days following the South Napa mainshock included the pre- and post-mainshock data and the gradual decay of dv/v in this period is an artifact.

To monitor the relative velocity change, we implement an automated system that updates the time history of dv/v shown in Figure 1. This automated process makes use of waveforms archived at the Northern California Earthquake Data Center [NCEDC, 2014]. Figure 1 is currently updated every day.

A minor difference between Figure 1 and the result from Taira et al. [2015] is about the reference GF. The automated system uses the reference GF from the data between January 2013 through July 2014. Taira et al. [2015] make use of the data between September 2010 through July 2014 for computing the reference GF. However, this does not change the resultant dv/v time seriese significantly.

Figure 1. Top panel shows the number of channel pairs used to infer the relative velocity change (dv/v). The bottom panel shows the time history of dv/v. The error bars represent the two-sigma standard deviation. The dashed red line indicates the origin time of the South Napa earthquake. Note that the dv/v values plotted within the first 4 days following the South Napa mainshock included the pre- and post-mainshock data and the gradual decay of dv/v in this period is an artifact.

Map View¶

Figure 2. Map view of stations and the South Napa earthquake. The 8 stations used is a combination of the Berkeley Digital Seismic Network & the Northern Hayward Fault Network operated by UC Berkeley, the Northern California Seismic Network operated by U.S. Geological Survey, and the EarthScope Plate Boundary Observatory operated by UNAVCO, Inc.