SAFOD Collaboration

An intensive and ongoing effort by the SAFOD target event location working group is underway with the following goals: 1) the characterization of the detailed velocity and seismicity structure in the crustal volume containing the SAFOD main hole and 2) to determine the most accurate estimates of the absolute locations of SAFOD's target events. The HRSN data play a key role in this effort by providing low noise and high sensitivity seismic waveforms from active and passive sources and by providing a backbone of earthquake and tremor detection and waveforms from the numerous microearthquakes and tremors that are occurring in the general vicinity of SAFOD.

In a special section of Geophysical Research Letters from May of 2004, several papers make significant use of the HRSN data for characterizing the SAFOD area and illustrate the role that the HRSN data have played in the SAFOD effort over the past year (e.g., Oye et al., 2004; Roecker et al., 2004; Thurber et al., 2004; Nadeau et al., 2004).

In addition to the data collection, Berkeley's other contributions also include detailed monitoring of the repeating M2 SAFOD target sequences, and in particular, this year, their evolving pattern of behavior following the 28 September 2004, M6 Parkfield earthquake (Figure 6.4).

Figure 6.4: Upper right panel shows a 40 km along fault depth section of the San Andreas Fault that includes the epicenters of the 1966 and 2004 Parkfield M6 earthquakes (box and star respectively), HRSN catalog locations of seismicity in the region preceding the 2004 M6 event (black points) and the region containing the SAFOD target sequences of repeating M2 earthquakes (small box). Zoom box of the SAFOD region shows that the SAFOD target region actually contains numerous sites of repeating microearthquakes (circles and diamonds). The diamonds labeled NW and SE are the two primary M2 sequences being considered as targets for close-in monitoring and eventual penetration. Large circle labeled SW is also an $\sim$ M2 repeating site that lies $\sim$ 2-300 meters to the southwest of the primary targets and lies between the SAFOD drill site and the primary targets. Lower panel shows waveforms from the target sequence events occurring since 1984 and into the period immediately following the Parkfield mainshock. Waveforms on the left are from surface NCSN station PCA (VHZ channel) and the three on the right are from the HRSN. The much higher sampling rate and borehole sensors of the HRSN are critical in identifying and discriminating between repeats of the primary targets that are located only $\sim$ 70 m apart and whose waveforms among the NCSN recordings are virtually indistinguishable. Both the primary target (NW and SE) and SW sequences experienced several repeats during the year following the 2004 M6 (for comparison, prior to the M6 the recurrence times of sequence members was generally once every $\sim$ 3 years). During January of 2005 (between SAFOD's Phase 1 and Phase 2 drilling periods) a string of seismic sensors was deployed down the SAFOD hole and recorded one of the post-M6 SE sequence repeats. This close-in information in conjunction with recordings from the HRSN, NCSN, the temporary PASO deployment and the SAFOD pilot hole string led to some surprises regarding the accuracy of the absolute location estimates for the primary target sequences, and provided critical information for guiding drilling during phase 2 in the summer of 2005.