Tremorscope: Moore Foundation Reports
TremorSCOPE: Imaging the Deep Workings of the San Andreas Fault
Figure: Tremor source region southeast of the Parkfield seismic networks. Open circles indicate the most active zone of non-volcanic tremor at ~25 km depth. Triangles (HRSN and PBO borehole stations) and squares (BDSN and SCSN network broadband stations) show the location of existing seismic stations in the area used in the discovery and initial study of the tremor. Colored stars indicate the proposed location of new stations (orange, broadband; red, borehole). A TremorSCOPE of surface broadband stations and low-noise borehole instruments clustered around the tremor zone would greatly sharpen our ability to detect, examine and image the tremor events and surrounding crustal properties. We propose to install three-component accelerometers and velocity sensors at both the surface and borehole sites to optimize our ability to study all aspects of the tremor activity. The Cholame Segment (CS) forms the northernmost extent of the currently "locked" fault segment that last ruptured in California's great M7.8 Fort Tejon earthquake, in 1857. Figure modified from Nadeau and Dolenc (2005, Science).
A project idea submitted for consideration by the Moore Foundation by the Berkeley Seismological Laboratory (Richard Allen, Pascal Audet, Roland Bürgmann, Doug Dreger, Peggy Hellweg; Bob Nadeau, Barbara Romanowicz, Taka'aki Taira)
Until recently, active fault zones were thought to deform via seismic earthquake slip in an upper brittle section of the crust, and by steady, aseismic shear below. However, in the last few years, this view has been shaken by seismological observations of seismic tremor deep in the roots of active fault zones. First recognized on subduction zones in Japan and the Pacific Northwest, tremor has also been found to be very active on a short section of the San Andreas to the southeast of one of the most densely monitored fault segments in the world near Parkfield (Nadeau and Dolenc, 2005 Science). This deep (~20-30 km) zone of activity is located right below the nucleation zone of the great 1857 Fort Tejon earthquake. Thus understanding of the temporally and spatially complex faulting process in this zone may help us better understand the conditions that lead to such large ruptures. Should a great San Andreas earthquake occur during this experiment (it is considered "overdue" by paleoseismic studies of Scharer et al., 2010 Geology), the network would also provide unprecedented insights into the seismic rupture process. Despite the tremor source area being away from the existing Parkfield seismic networks (see Figure), early studies of this deep tremor have led to dramatic revisions in our views about how faults behave at depth (e.g., Nadeau and Guilhem, 2009 Science; Thomas, Nadeau and Bürgmann, 2009 Nature; Shelly, 2010 Nature; Shelly, 2010 Science). These studies reveal fascinatingly complex behavior of faulting in the deep crust that is dramatically different from "normal" earthquakes in the upper crust. By adding an array of about five broadband seismic stations and four borehole accelerometers around the tremor source zone, we can complement the existing monitoring activity at Parkfield to dramatically sharpen our TremorSCOPE to explore this fascinating natural laboratory of active lower crustal faulting. As described, the array of broadband and borehole seismometers we envision for TremorSCOPE will cost about $1m, including equipment, installation and telemetry. Insights from this project would also benefit the understanding of tremor and slip in other regions of the world where such phenomena have been observed, but are not nearly as accessible.
TremorScope Report: December 2010- December 2011
Figure 1: Map showing the region of study for the TremorScope project. Gold stars indicate locations of past tremor episodes determined by R. Nadeau at BSL. Prospective stations are marked as squares, purple will have surface installations and green will be borehole sites. Radio telemetry paths are marked as red lines. BLAK and HOGS are USGS microwave sites, through which data will be forwarded in real-time to UC Berkeley. ORPK will be a radio repeater for three stations. The precise location for station TCLK has not yet been selected. Seismic stations will have a four character name beginning with "T" for TremorScope. Radios for TremorScope telemetry have been installed at BLAK, and a temporary station has been deployed at TSCS.
Figure 2: Temporary station deployment at TSCS. This station is co-located with an Earth- scope PBO - GPS station. The digitizer, power and telemetry equipment are in the box protected by the brown tarp. The broadband seismometer is under the mound in the foreground, the accelerometer in the disturbed area on the right of the image. Data are telemetered in real-time through Black Mt.
The Berkeley Seismological Laboratory at the University of California Berkeley has receieved Moore Foundation support to develop a unique network of seismic instruments to study the recently discovered tremor activity in the deep roots of the San Andreas fault at Cholame, California. Funding is being used to build, collect and analyse data from this wide- aperture, low-noise, broad-band TremorScope network. This will improve our understanding of the nature of the tremor signals and its implications for fault mechanics and earthquake hazard. All data resulting from these efforts will be publicly available within a short time of their being collected.
Year 1 efforts have been directed toward Output 1.1 - To design, permit, and install the network of borehole and surface seismic stations to be used to study the deep tremor at Cholame, but the task has not yet been completed.
Network Design. Achieving our primary outcome requires the deployment of an integrated array of a variety of seismic instruments to focus with unprecedented precision on the distribution of tremors in space and time. The network will consist of eight seismic stations spread throughout the region (Figure 1). The stations will be of two types, surface stations and borehole stations. The surface stations will be concentrated in the region where tremor sources have been located. The borehole stations, which will have a lower detection threshhold due to the lower seismic noise at depth below the surface, will encircle the tremor region. Each of the eight sites will have a surface strong motion accelerometer to capture the ground motion of a large or great nearby earthquake on-scale, should an event like the 1857 rupture of the San Andreas occur.
Seismic equipment at each surface site (Figure 1) will consist of a three-component broadband seismometer with flat response to velocity between 8.3 mHz and 50 Hz, and a three- component, strong-motion accelerometer with flat response to acceleration from 0 to 50 Hz. Data from all 6 components will be digitized onsite by a 24-bit digitizer and both telemetered to our data center for archival and processing and stored on-site for retrieval should telemetry fail.
Our plan is to drill four boreholes that are 300 m deep at each of the borehole sites (Figure 1). At the bottom of each borehole, a three-component 2 Hz geophone package will be grouted in, with a cable to bring signals to the surface. They will be digitized, recorded and telemetered by a surface digitizer/data logger unit, which will also record data from the surface accelerometer. Just above the geophone, we will install a removable sensor package, which includes a three- component broadband seismometer, a three-component strong-motion accelerometer and a 24-bit digitizer. Digital data will be telemetered in optical fiber to a unit on the surface which will store them, as well as forewarding them through the telemetry to the data center.
We requested and received bids from the major manufacturers of seismic instrumentation for the TremorScope instrumentation. We selected Guralp Systems Ltd. as they are able to provide all the different elements required, and we would be able to install them without having to worry about interconnection problems.
Figure 3: Waveforms and spectrograms from a teleseism recorded by the temporarily installed equipment at TSCS. This quake occurred near the south coast of Papua, Indonesia (4.77S,133.66E), on 5 Dec 2011 at 19:28:35. The large amplitude waves are surface waves, which have travelled through the lithosphere across the Pacific Ocean.
All the seismic equipment has been ordered and received.
Network Permitting. Based on the distribution of tremor sources, the regional geology, site logistics and data quality from a number of existing seismic stations, we identified candidate site locations in the target region. During the spring and summer of 2011, we visited the TremorScope region several times, to evaluate prospective station sites, and to perform telemetry tests to ensure that we will be able to retrieve data in real-time. Our USGS colleagues in Menlo Park have a microwave telemetry system that extends to Black Mt. (BLAK, Figure 1) and Hog Canyon (HOGS). They have offered bandwidth for the TremorScope data and we have coordinated the locations of proposed TremorScope sites with them. For two of the borehole sites TPPG and TPBP, we will adopt and improve USGS short-period network sites to allow tremor to be detected. To retrieve data from those two sites and TRAM, we will install a radio repeater at Orchard Peak (ORPK, Figure 1).
After a telemetry test in August, we began the permitting process. We currently have signed permits for four TremorScope sites, THHS, TRAM, TSCN, TSCS. We are working to complete permitting for the other four sites and the telemetry repeater, ORPK.
Network Installation. Not as much progress as we had hoped has been made toward this goal. With USGS colleagues, we installed radios for the telemetry through Black Mt. We have begun purchasing and preparing infrastructure elements for the field stations, such as telemetry radios, solar panels and solar panel mounts. We have made inquiries about drilling companies with experience in drilling for scientific installations, but have not yet requested bids for the drilling effort. In late November, we installed a temporary seismic station at the southernmost TremorScope site, TSCS (Figure 2). We have been able to record several distant earthquakes (Figure 3). There has been very little local seismicity.
Next Steps: Our goal for the first six months of 2012 is to complete the permitting and installation of the TremorScope stations. As each station is installed, its data will be telemetered to our data center, and will be incorporated into our real-time earthquake processing. In addition, we will begin processing the data in search of tremor signals.