Figure 41.1 shows the location of the network, its relationship to the San Andreas fault, sites of significance from previous and ongoing research using the HRSN, double-difference relocated earthquake locations from 1987-1998, routine locations of seismicity from August 2002 to July 2003, nonvolcanic tremor locations from January 2001 through April 2005, and the epicenter of the 1966 and 2004 M6 earthquakes that motivated much of the research. The HRSN records exceptionally high-quality data, owing to its 13 closely spaced three-component borehole sensors (generally emplaced in the extremely low attenuation and background noise environment at 200 to 300 m depth (Table 41.1), its high-frequency wide bandwidth recordings (0-100 Hz; 250 sps), and its low magnitude detection threshold (below magnitude Ml).
Several aspects of the Parkfield region make it ideal for the study of small earthquakes and nonvolcanic tremors and their relationship to tectonic processes and large earthquakes. These include the fact that the network spans the SAFOD (San Andreas Fault Observatory at Depth) experimental zone, the nucleation region of earlier repeating magnitude 6 events and a significant portion of the transition from locked to creeping behavior on the San Andreas fault, the availability of three-dimensional P and S velocity models (Michelini and McEvilly, 1991), the existing long-term HRSN seismicity catalogue that is complete to very low magnitudes and that includes at least half of the M6 seismic cycle, a well-defined and simple fault segment, the existence of deep nonvolcanic tremor (NVT) activity, and a homogeneous mode of seismic energy release as indicated by the earthquake source mechanisms (over 90 right-lateral strike-slip).
In a series of journal articles and Ph.D. theses, we have presented the cumulative, often unexpected, results of UC Berkeley's HRSN research efforts (see: http://www.seismo.berkeley.edu/seismo/faq/parkfield_bib.html). They trace the evolution of a new and exciting picture of the San Andreas fault zone responding to its plate-boundary loading, and they are forcing new thinking on the dynamic processes and conditions within the fault zone at the sites of recurring small earthquakes and deep nonvolcanic tremors (Nadeau and Dolenc, 2005).
The Parkfield area has also become an area of focus of the EarthScope Project (http://www.earthscope.org) through the SAFOD experiment (http://www.icdp-online.de/sites/sanandreas/news/news1.html), and the HRSN is playing a vital role in this endeavor. SAFOD is a comprehensive project to drill into the hypocentral zone of repeating M 2 earthquakes on the San Andreas Fault at a depth of about 3 km. The goals of SAFOD are to establish a multi-stage geophysical observatory in close proximity to these repeating earthquakes, to carry out a comprehensive suite of down-hole measurements in order to study the physical and chemical conditions under which earthquakes occur, and to monitor and exhume rock, fluid, and gas samples for extensive laboratory studies (Hickman et al., 2004).
The data acquisition system operated quite reliably until late 1996, when periods of unacceptably high down time developed. During this period, as many as 7 of the remote, solar-powered telemetered stations were occasionally down simultaneously due to marginal solar generation capacity and old batteries, and recording system outages of a week or more were not uncommon. In July 1998, the original data acquisition system failed permanently. This system was a modified VSP recorder acquired from LBNL, based on a 1980- vintage LSI-11 cpu and a 5 MByte removable Bernoulli system disk with a 9-track tape drive, configured to record both triggered microearthquake and Vibroseis data (Vibroseis discontinued in 1994, Karageorgi et al., 1997). The system was remote and completely autonomous, and data tapes were mailed about once a month to Berkeley for processing and analysis. The old system also had a one-sample timing uncertainty and a record length limitation because the tape write system recovery after event detection was longer than the length of the record, leaving the system off-line after record termination and until write recovery was completed.
In July 1999, the network was reduced, due to limited instrument availability, to four RefTeks at critical sites that would ensure continuity in monitoring at low magnitudes and the archive of characteristic events for studying the evolution of their recurrence intervals. Properties of the 10 original sites are summarized in Table 41.2.
We have also added three new borehole stations, with NSF support, at the NW end of the network as part of the SAFOD project to improve resolution of the structure, kinematics and monitoring capabilities in the SAFOD drill-path and target zones. Figure 41.1 illustrates the location of the drill site, the new borehole sites, and locations of earthquakes recorded by the initial and upgraded/expanded HRSN.
The three new SAFOD stations have a similar configuration as the original upgraded 10 station network and include an additional channel for electrical signals. Station descriptions and instrument properties are summarized in Tables 41.1 and 41.2. All HRSN Q730 dataloggers employ FIR filters to extract data at 250 and 20 Hz (Table 41.3).
The remoteness of the drill site and new stations required an installation of an intermediate data collection point at Gastro Peak, with a microwave link to the CDF facility. The HRSN stations use SLIP to transmit TCP and UDP data packets over bidirectional spread-spectrum radio links between the on-site data acquisition systems and the central recording system at the CDF. Six of the sites transmit directly to a router at the central recording site. The other seven sites transmit to a router at Gastro Peak, where the data are aggregated and transmitted to the central site over a 4 MBit/second digital 5.4 GHz microwave link. All HRSN data are recorded to disk at the CDF site.
The upgraded and expanded system is compatible with the data flow and archiving common to all the elements of the BDSN/NHFN and the NCEDC (Northern California Earthquake Data Center), and is providing remote access and control of the system. It has also provided triggered data with better timing accuracy and longer records, which are to eventually flow seamlessly into NCEDC. The new system also helps minimize the problems of timing resolution, dynamic range, and missed detections, in addition to providing the added advantage of conventional data flow (the old system recorded SEGY format).
At present, all 38 continuous 20 sps data streams are telemetered to the BSL. All continuous 250 sps data are migrated periodically from HRSN computer in Parkfield to DLT tape. These tapes are then mailed periodically to the BSL and are then processed and archived at the NCEDC. Seven vertical 250 sps channels are also telemetered to the NCEDC for purposes of quality control and SAFOD related activities. These data are archive temporarily (for 10 days) and then removed. Copies of the data are later restored for permanent archiving during uploading of the 38 250 sps continuous data streams from the DLT tapes.
A feature of the new system that has been particularly useful both for routine maintenance and for pathology identification has been the Internet connectivity of the central site processing computer and the station dataloggers with the computer network at BSL. Through this connection, select data channels and on-site warning messages from the central site processor are sent directly to BSL for evaluation by project personnel. If, upon these evaluations, more detailed information on the HRSN's performance is required, additional information can also be remotely accessed from the central site processing computer at Parkfield. Analysis of this remotely acquired information has been extremely useful for trouble shooting by allowing field personnel to schedule and plan the details of maintenance visits to Parkfield. The connectivity also allows certain data acquisition parameters to be modified remotely when needed, and commands can be sent to the central site computer and dataloggers to modify or restart processes when necessary.
The network connectivity also allows remote monitoring of the background noise levels being recorded by the HRSN stations. For example, shown in Figure 41.3 are power spectral density plots of background noise for vertical components of the 7 HRSN stations that are most critical for monitoring seismicity in the region containing SAFOD. The PSD analysis gives a rapid assessment of the HRSN seismometer responses across their wide band-width. By routinely generating these plots with data telemetered from Parkfield, changes in the seismometer responses, often indicating problems with the acquisition system, can be easily identified, and corrective measures can then be planned and executed on a relatively short time-frame.
Shortly after being recorded to disk on the central site HRSN computer, event triggers for the individual station data are determined, and a multi-station trigger association routine then processes the station triggers and identifies potential earthquakes. For each potential earthquake that is detected, a unique event identification number (compatible with the NCEDC classification scheme) is assigned. Prior to San Simeon earthquake of December 22, 2003, 30 second waveform segments were then collected for all stations and components and saved to local disk as an event gather, and event gathers were then periodically telemetered to BSL and included directly into the NCEDC earthquake database (dbms) for analysis and processing.
Because of its mandate to detect and record very low magnitude events in the Parkfield area, the HRSN is extremely sensitive to changes in very low amplitude seismic signals. As a consequence, in addition to detecting very small local earthquakes at Parkfield, the HRSN also detects numerous regional events. For example spot checks of aftershocks following the M6.5 San Simeon earthquake of December 22, 2003 using continuous data and HRSN event detection listings have revealed that the overwhelming majority of HRSN detections following San Simeon resulted from seismic signals generated by San Simeon's aftershocks despite the HRSN's 50 km distance from the events. Data from the California Integrated Seismic Network (CISN) show that there were 1,150 San Simeon aftershocks with magnitudes 1.8 in the week following San Simeon, and during this same period, the number of HRSN event detections was 10,500 (compared to an average weekly rate before San Simeon of 115 detections) This suggests that despite the 50 km distance the HRSN is detecting San Simeon aftershocks well below magnitude 1.
Since the beginning of the network's data collection in 1987 and up until recently, the local and regional events were discriminated based on analyst assessment of S-P times, and only local events with S-P times less than 2.5 sec at the first arriving station were picked and located as part of the HRSN routine catalog. However, because of the large swarms of aftershocks from the San Simeon and M6 Parkfield earthquake of September 2004 and because of declining funding levels, this approach has had to be abandoned.
Current Processing. Subsequent to the M6.5 San Simeon earthquake on December 22, of 2003, our long-standing data handling procedure was no longer viable due to the enormous rate of San Simeon aftershock detections (Figures 41.2) In the first 5 months following the San Simeon mainshock, over 70,000 event detections were made by the HRSN system (compared to an average 5 month detection rate of 2500 prior to San Simeon). In the first month following the 28 September 2004 Parkfield M6 quake, over 40,000 detections were also made. Numerous additional (false) detections have also been occurring as a result of drilling activities associated with SAFOD drilling.
The dramatic increase in event detections vastly exceed the HRSN's capacity to process both the continuous and triggered event waveform data. To prevent the loss of seismic waveform coverage, processing of the triggered waveform data has been suspended to allow archiving of the 250 sps continuous data to tape to continue uninterrupted. Cataloging of the event detection times from the modified REDI real-time system algorithm is also continuing, and the continuous 250 sps waveform data is currently being periodically uploaded from the DLT tape archive onto the NCEDC for access to the research research community.
Funding to generate catalogs of local events from the 10's of thousands of aftershock detections has not been forthcoming, and as a consequence major changes in our approach to cataloging events have had to be implemented, which involves integration of HRSN data into NCSN automated event detection and cataloging (with no analyst review) combined with a high resolution procedure now being developed to automatically detect, pick, locate and determine magnitudes for similar and repeating events down to very low magnitudes (i.e., below magnitude -1.0Ml). These new schemes are discussed in more detail in the activities section below.
Remote monitoring of the networks health using the Berkeley Seismological Laboratory's SeisNetWatch software are also performed to identify both problems that can be resolved over the Internet (e.g. rebooting of data acquisition systems due to clock lockups) and more serious problems requiring field visits.
Over the years, such efforts have paid off handsomely by providing exceptionally low noise recordings (Figure 41.3) of very low amplitude seismic signals produced by microearthquakes (below magnitude 0.0Ml) and nonvolcanic tremors (Nadeau and Dolenc, 2005).
Over the previous several years, we have had a serious decline in the robustness of the power system components (primarily the aging solar panels and batteries that have been in use since initiation of the network in 1987) of the network. Simultaneous outages at multiple stations are now becoming an all too frequent occurrence and are seriously affecting efforts to monitor tremor and micro- and repeating earthquake activity in the Parkfield area.
For example, during the winter of late 2004/early 2005, monitoring for nonvolcanic tremor activity using a standard detection set of 8 HRSN channels revealed significant (and sometimes catastrophic) gaps in the data. Figure 41.4 illustrates the seriousness of the problem with an example from tremor monitoring during periods of overcast weather. During the 7 day period shown, all 8 stations used for monitoring tremor activity were out simultaneously for over 50% of the time. The remaining 50% of the time, outages occurred for at least some of these 8 stations, resulting in significantly degraded capability for unambiguous detection of the low-amplitude tremor activity.
As suspected, further investigation, both remotely and on site, showed that these gaps occurred due to insufficient battery re-charge at many of the network's stations, which are remote solar powered installations. In previous years, similar but less severe data gaps have occurred during the winter months and have been attributed to overcast skies during the rainy season. In the winter of 2005 exceptionally heavy rainy season exacerbated the outage problem to an intolerable level, and to avoid a potential repeat of the situation, efforts were undertaken to refurbish and upgrade the solar power systems.
Specifically, the following steps were and continue to be taken:
1) replacement of the oldest batteries and switching of the remaining old batteries to the less power consuming pre-amplifiers;
2) improvement of the wiring scheme along the lines suggested by the solar power representative;
3) upgrade/replacement of solar panels. (Solar panels degrade at 1% per year, and newer versions have improved output. Since the installation of the HRSN over 18 years ago, the same size/format panel has gone from 40 watts to 55). This is a relatively easy field task, and should gain us 20-30% capacity at each site.
Among the three newer sites (CCRB, SCYB, LCCB), both the batteries and solar panels are relatively new. Nonetheless, stations CCRB and LCCB both had some outages last winter, which is most likely explained by the limited sunlight in these areas due to hilly terrain. We have, therefore, added one more solar panel at each of these sites to enhance their power system robustness.
The table shown in figure 41.4 summarizes the tasks of the power system upgrade effort, and shows the state of completion of the tasks as of the end of 2005. To date all tasks have now been completed.
This approach has been successful at discriminating small events in the local Parkfield area from other types of event detections and for providing automated locations of a significantly increased number of small events in the local area (approx. double that of the NCSN network alone). However, the rate of local events from the HRSN sensitized NCSN catalog is still only catching about 1/2 the number of local events previously cataloged by the HRSN, and waveforms for the small events are not typically made available. In addition, unlike the previous HRSN catalog, the additional events added by the NCSN-HRSN integration are not reviewed by an analyst nor do they generally have magnitude determinations associated with them. In some cases, the selection rules used for the integrated catalog also result in exclusion of events that are otherwise included by the NCSN.
These limitations severely hamper efforts relying on similar and characteristically repeating microearthquakes. They also reduce the effectiveness of research relying on numerous very small magnitude events in the SAFOD zone (e.g. for targeting the SAFOD targets).
To help overcome these limitations, we have embarked on an effort to develop an automated similar event cataloging scheme based on cross-correlation and pattern scanning of the continuous HRSN data now being archived. The method uses a small number of reference events whose waveforms, picks, locations, and magnitudes have been accurately determined, and it automatically detects, picks, locates and determines magnitudes for events similar to the reference event to the level of accuracy and precision that only relative event analysis can bring.
The similar event detection is also remarkably insensitive to the magnitude of the reference event used, allowing similar events ranging over several magnitude units to be fully cataloged using a single reference event. It also does a remarkably good job even when seismic energy from multiple events is superposed. Once a cluster of similar events has been cataloged, it is a relatively straight forward process to identify characteristically repeating microearthquake sequences within the cluster (frequently a single similar event "cluster" will contain several sequences of repeating events).
Application of the method using one of the SAFOD target events as a reference is illustrated in Figure 41.6. The magnitude of the reference event is 2.2. This event was scanned through 5 years of continuous data, and 67 other events occurring within a zone of 150 m were detected (including 3 very small quakes that were not even by the HRSN REDI-type system). The magnitudes of these events ranged down to magnitude -1.2 Ml. In addition to the SAFOD target sequence from which the reference was derived, several other repeating sequences within the 150m zone were also identified (5 of which had not previously been known to exist).
The procedure is still being refined to capture even smaller events, events over a larger area and for increased processing speed. Eventually, a composite catalog of similar event groups from throughout the HRSN coverage zone is planned.
The approach also holds promise in other applications where automated and precise monitoring of bursts of seismic activity to very low magnitudes is desirable (e.g. in aftershock zones or in volcanic regions) or where automated updates of preexisting repeating sequences and their associated deep slip estimates are desired.
At this stage SAFOD drilling has penetrated the fault with a sub-horizontal hole slightly beneath the SAFOD target sequences, and current efforts have been focused on obtaining final estimates of the targets relative location to the existing hole to accuracies of meters if possible. This high degree of accuracy is required in order to target accurately three multi-lateral side cores for sampling and monitoring within the final target zone.
HRSN Activities this year have contributed in three principal ways to these and longer-term SAFOD monitoring efforts:
1) In collaboration with the USGS, we have integrated the 7 vertical HRSN channels telemetered from Parkfield into the NCSN triggering scheme (described above) to increase the sensitivity of NCSN detection in the SAFOD area. This has effectively doubled the number of small events the target location working group has for constraining the relative location of the target sequences.
2) Again in collaboration with the USGS, we have nearly completed a telemetry upgrade that will allow all 38 channels of the HRSN data (both 20 sps and 250 sps data streams) to flow directly from Parkfield, through the USGS Menlo Park processing center, and also to the BSL for near-real-time processing and archiving on the web based NCEDC. This will provide near immediate access of the HRSN data to the community without the week's to month's delay associated with having to transport DLT tapes to Berkeley, upload, and quality check the data.
3) We have also applied our prototype similar event automated catalog
approach to the primary and two secondary SAFOD target zones and
were able to provide the SAFOD event location working group with
rapid and precise double-difference and relative magnitude catalogs of 82
similar events in the zone immediately surrounding target region
occurring between 2001 day 178 and 2006 day 218 (August 6 of this
Figure 41.6 shows the double difference locations and estimated rupture dimensions (based on Nadeau and Johnson, 1998) of 67 of these events that were derived using one event from the SAFOD primary target sequence as the reference. Other primary target events are shown in green, and events from a secondary target located 40 m to the southeast are shown in blue. Several other suspected repeating sequences can be seen as tight clusters of similarly sized events. We are in the process of confirming these events as characteristically repeating sequences members.
The SAFOD similar event catalogs are now being used by the working group to extract data from the corresponding PASO array, Pilot Hole, NCSN and mainhole data sets for integration with the HRSN data to provide as much and as detailed information as possible in the final push at locating the target sequence for the lateral side core drilling.
Daley, T.M. and T.V. McEvilly, Shear wave anisotropy in the Parkfield Varian Well VSP, Bull. Seism. Soc. Am., 80, 857-869, 1990.
Hickman, S., M.D. Zoback and W. Ellsworth, Introduction to special section: Preparing for the San Andreas Fault Observatory at Depth, Geophys. Res. Lett., 31, L12S01, doi:10.1029/2004GL020688, 2004.
Karageorgi, E., R. Clymer and T.V. McEvilly, Seismological studies at Parkfield. IV: Variations in controlled-source waveform parameters and their correlation with seismic activity, 1987-1994, Bull. Seismol. Soc. Am., 87, 39-49, 1997.
Michelini, A. and T.V. McEvilly, Seismological studies at Parkfield: I. Simultaneous inversion for velocity structure and hypocenters using B-splines parameterization, Bull. Seismol. Soc. Am., 81, 524-552, 1991.
Nadeau, R.M. and D. Dolenc, Nonvolcanic Tremors Deep Beneath the San Andreas Fault, SCIENCE, 307, 389, 2005.
Nadeau, R. M., and L. R. Johnson, Seismological Studies at Parkfield VI: Moment Release Rates and Estimates of Source Parameters for Small Repeating Earthquakes, Bull. Seismol. Soc. Amer., 88, 790-814, 1998.
Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail: email@example.com
© 2006, The Regents of the University of California