T11F-01 INVITED 08:00h
Site Characterization for the San Andreas Fault Observatory at Depth
The San Andreas Fault Observatory at Depth (SAFOD) is a comprehensive project to obtain rock and fluid samples, make in situ measurements, and install a permanent geophysical observatory in an inclined borehole crossing the San Andreas Fault Zone to a depth of 3.2 km. SAFOD is located near the town of Parkfield in central California, along a segment of the fault that fails through a combination of aseismic creep and regularly recurring microearthquakes. By passing through or very close to the rupture patch for a repeating magnitude 2 earthquake, SAFOD will present a unique opportunity to study the physical and chemical processes controlling earthquake generation and determine the composition, physical properties and mechanical behavior of an active, plate-bounding fault at depth. SAFOD is one element of the National Science Foundation's new EarthScope initiative (see www.earthscope.org). Drilling of SAFOD began in the summer of 2004, and will be completed in 2007. Since the SAFOD project was first proposed about 12 years ago, a wide variety of geophysical and geological investigations have been carried out in the region surrounding the SAFOD site. These studies include high-resolution seismic reflection and refraction profiles conducted through the SAFOD site and across the San Andreas Fault in 1998, 2002 and 2003, airborne and ground-based gravity and magnetic surveys, thermal and geochemical studies in shallow wells, geologic mapping, magnetotelluric profiling, microearthquake relocations, 3-D seismic tomography and fault-zone guided wave studies. These investigations were supplemented by drilling and testing of a 2.2-km-deep pilot hole at the SAFOD site in the summer of 2002, which was then instrumented with a 32-level, 3-component seismic array. This comprehensive suite of investigations surrounding the SAFOD site and in the pilot hole achieved a number of critical milestones. These include improving the location accuracy for the repeating microearthquakes being targeted by SAFOD and better defining the overall structure and geophysical setting of the San Andreas Fault Zone at Parkfield. These investigations are also providing critical information on the locations of secondary fault zones and velocity discontinuities that were used to design the current SAFOD drilling, sampling, testing and monitoring program. This geophysical and geological model is now being tested and refined as we drill toward the San Andreas Fault with SAFOD.
http://www.earthscope.org
T11F-02 08:15h
Seismic Velocity Structure From a Refraction - Reflection Survey Across the San Andreas Fault at SAFOD
Detailed characterization of the subsurface geology surrounding the San Andreas Fault Observatory at Depth (SAFOD) drill site is required to plan drilling and to interpret down-hole results. A 46-km long seismic refraction and reflection line was acquired in November 2003 perpendicular to the San Andreas Fault centered on the SAFOD drill site. The line was deployed straight across country assisted by helicopters. A fixed array of 912 three-component stations at a 25-50 m spacing recorded 68 explosive 25-100 kg shots at a 0.5-1 km spacing. Piggyback deployments recorded the shots on seismographs in the SAFOD pilot hole, on permanent and temporary earthquake arrays, and on a high-resolution (5-m) refraction array between the drill site and the fault. First arrival travel times from the main line were inverted to obtain a 2-D seismic velocity model of the upper crust. The model contains strong variations from $<$2 km/s at the surface to $>$6 km/s in granitic basement. Granitic rocks of the Salinian terrane west of the San Andreas Fault are substantially faster than adjacent predominantly sedimentary rock of the Franciscan terrane east of the fault. Salinian basement slopes westward from 0.7 km subsurface at SAFOD to $\sim$2.5 km depth at the Salinas River, with suggestions of two faulted steps. A small velocity contrast and reflections indicate the position of the Coast Range fault, which juxtaposes the Great Valley Sequence sedimentary rocks and the Franciscan terrane. A body of late Cenozoic sedimentary rocks indicated by low seismic velocity lies immediately west of the surface trace of the San Andreas Fault, between SAFOD and the fault. This body is bounded to the west by the Buzzard Canyon Fault or similar structure just east of SAFOD. It extends much deeper than sedimentary rocks at the SAFOD site and approaches the depth of the shallowest earthquakes. Saline water in these rocks could explain the previously observed zone of high electrical conductivity. In August, SAFOD drilled laterally from granitic into sedimentary rocks northeast of the drill site. Detailed structure of the San Andreas and Buzzard Canyon faults is critical to linking down-hole observations to surface geology.
T11F-03 INVITED 08:30h
SAFOD Site Characterization using the Pilot Hole Seismic Array
Since we installed it in July 2002 and its demise in July 2004, we used a 32 level array of 3-component, 15 Hz seismographs in the Pilot Hole to study the structure and properties of the SAFOD site. The array levels were spaced at 40 m intervals, with the deepest level at 2096 m. A catalogue of several hundred earthquakes and explosions was recorded with the array, mostly with a sampling rate of 2 kHz. We found that the seismograms from these events contain significant information about the structure and properties of the SAFOD site and San Andreas Fault. These data show source and receiver position-sensitive travel time variations in the P- and S-waves. They also commonly contain several secondary phases between the direct waves. By analyzing the data with a variety of travel time ratio, tomographic, and wave-field migration techniques, we have been able to map changes in Vp/Vs along the SAF and along the PH itself. Along the SAF, a northwest-to-southeast reduction in Vp/Vs correlates well with a reduction in fault creep. Between the SAF and PH, P-wave tomography indicates that a large body of rocks with slow velocities is present. By migrating the secondary phases with Kirchhoff methods, we have been able to image SAF-paralleling structures, both within the low velocity body, and on its edges. As one possible interpretation, we have suggested that these features represent a faulted and fault bounded sedimentary section. So far this interpretation has fit the results of the SAFOD 2004 drilling. The presence of these faults and the unexpected volume of sedimentary rocks, which also appear fluid saturated, have important implication how local stresses and strains distribute and build during the earthquake cycle.
T11F-04 08:45h
Defining the SAFOD Drilling Trajectory: Locating the Target Earthquakes
Our previous work on 3D seismic wavespeed models and earthquake locations near the SAFOD site contributed to defining the Phase I drilling plan for summer 2004. Two key results from that stage of our work were that the target earthquakes lie about a kilometer shallower than previously thought and that the SAFOD Pilot Hole, drilled vertically into a fractured granite body, was estimated to be situated several hundred meters southwest of a zone of substantially reduced seismic wavespeed. From this summer's drilling, we now know that the Pilot Hole is only about 200 meters away from the low velocity material at 2 km depth below surface. The current stage of our work involves the deployment of a dozen temporary seismic instruments around SAFOD and the execution of small shots adjacent to 6 of these sites in October-November 2004. The shots will be recorded by a geophone to be installed near the bottom of the Phase I SAFOD borehole and/or a multilevel geophone array. This will allow us to construct one or more additional "virtual earthquakes" (receiver gathers of surface shots treated as earthquakes) for testing absolute location accuracy, now at a position about a kilometer closer to the actual target earthquakes than the bottom of the Pilot Hole geophone array. We present the updated location and wavespeed model results obtained with the incorporation of the 2004 data. We also discuss a suite of tests of our absolute and relative location capability for the target region, and identify additional steps needed to reduce location uncertainties to approximately 10 meters in preparation for Phase III of SAFOD drilling, hitting the rupture patch of a magnitude 2 target earthquake.
T11F-05 09:00h
Multi-Scale Crustal Seismic Anisotropy in the Region Surrounding the San Andreas Fault Near Parkfield, CA.
The region surrounding the San Andreas Fault Observatory at Depth (SAFOD) near Parkfield, CA is an ideal location to study the effect of crustal structure and the state of stress on seismic velocity anisotropy because the direction of maximum horizontal compression is at a high angle to the predominantly northwest-southeast structural trend. Data from the 2.2-km-deep pilot hole and upper section of the main SAFOD borehole provides a unique opportunity for studying the in-situ physical properties of the crust adjacent to the San Andreas Fault Zone. To study seismic anisotropy in the crust at multiple scales, we utilize a suite of geophysical logs from the SAFOD boreholes, in addition to earthquake data recorded on the Pilot Hole array and on the regional High Resolution Seismic Network (HRSN) operated by U.C. Berkeley. At the smallest scale, dipole sonic logs in the SAFOD boreholes indicate that the shear-wave velocity anisotropy of the rocks within a few feet of the wellbore is on the order of 3 to 10% and controlled by the tectonic stress field. An analysis of earthquake seismograms shows that ray paths through the crust adjacent to the fault exhibit fast shear wave polarizations aligned with the direction of maximum horizontal compression, in agreement with the SAFOD measurements, whereas ray paths along the San Andreas fault yield fault-parallel fast directions. The delay times of the lagging shear wave are also much larger when the waves have traveled along the fault zone indicating that structural fabric has a stronger influence on velocity anisotropy than the regional stress field. We conclude that within the San Andreas Fault Zone, the structural fabric is the dominant mechanism responsible for velocity anisotropy whereas in the surrounding crust, the direction of maximum horizontal compression is the most important controlling factor.
T11F-06 INVITED 09:15h
Characterization of fault zone structure at the SAFOD site with magnetotelluric exploration
Magnetotelluric exploration has played an important role in site characterization at the SAFOD site. An initial magnetotelluric (MT) survey in 1994 revealed a fault zone conductor (FZC) extending to a depth of around 2-3 km. A more detailed MT survey in 1997 showed that this feature was typical of the Parkfield segment of the San Andreas Fault. Other MT surveys have shown that similar zones of low resistivity are present worldwide on other major strike-slip faults. These features are generally attributed to a zone of enhanced fracturing and elevated fluid contents, although clay mineralization can also decrease the resistivity. Some segments of major strike-slip faults also exhibit zones of low resistivity in the mid-crust, which may be directly related to crustal flow processes (e.g. creeping segment of the San Andreas Fault and the North Anatolian Fault at Izmit). Since the MT data collection at Parkfield in 1994-7, additional inversion and modeling studies have been undertaken and constrain the depth extent of the low resistivity wedge to around 1.5 to 2 km. High- resolution earthquake locations now reveal that seismicity begins at the base of the low resistivity wedge. The \textit{in situ} resistivity imaged with MT shows a significant correlation with velocities derived from seismic tomography. While the shallow geoelectric structure at the SAFOD site is relatively well understood, the deeper structure at seismogenic depths is less well constained. Synthetic inversion studies show that a longer MT profile could image features in the mid-crust that may be associated with crustal flow and deformation processes. It is possible that the width of a shear zone in the lower crust could be determined by a carefully executed MT survey.
T11F-07 09:30h
Crustal Structure Across the San Andreas Fault at the SAFOD Site, California, From Gravity and Magnetic Studies
Newly compiled gravity, magnetic and geologic datasets from the Parkfield region around the San Andreas Fault Observatory at Depth (SAFOD) help define the structure and geophysical setting of the San Andreas Fault (SAF). A previous 2-D model incorporating gravity, ground magnetic, and geology data along a profile passing within 200 m of the SAFOD pilot hole suggests that, as drilling proceeds NE toward the surface trace of the SAF, it will encounter a broad zone separating granitic basement on the SW and Franciscan basement rocks on the NE. Further to the NE the drill will probably encounter interleaved Franciscan rocks and serpentinite, and possibly a mass of serpentinite itself. Of particular significance are two magnetic bodies located just SW of the SAF which appear to be steeply dipping, fault-parallel, and likely fault-bounded slivers, although the exact nature and depth of these bodies is unclear. The zone between SAFOD and the SAF is also characterized by a dipping Tertiary sedimentary section overlying granitic basement rocks, which coincides with the low velocity zone observed with seismic measurements. This 2-D model was constrained by pilot hole measurements, where we observe a boundary between sediment and granitic basement at ~770 m depth and an order of magnitude downward increase in magnetic susceptibility at ~1400 m depth, which coincides with the steeply dipping northeastern edge of a magnetic granitic body within the granitic basement. New gravity data collected during the fall of 2003 along a 50-km-long seismic reflection/wide-angle refraction profile across the SAF extend through the SAFOD site. Stations were spaced 100 m apart along a 5-km-long portion of the profile centered about the surface trace of the SAF and then 300 m apart along the rest of the line; data were reduced to isostatic residual gravity anomalies. We use these data to produce a 2-D model which further constrains the geometry of the dipping sedimentary section just NE of SAFOD in order to determine if the drill will encounter sediments as it heads toward the SAF. Preliminary modeling shows that the dipping Tertiary sedimentary section NE of SAFOD is likely shallow ($<$ 2km); however, it is difficult to constrain the dip angle of the base of the section. These data will also address the extent of shallow Franciscan rocks NE of the SAF. Although Franciscan rocks crop out immediately NE of the SAF, previous gravity data suggests that the Franciscan rocks overlie low-density Cenozoic deposits. New ground magnetic data will be collected during the fall of 2004 along the profile extending through SAFOD and regions of the SAF just NW and SE of existing ground magnetic data. These observations will allow us to explore the extent of the unknown magnetic bodies along the SAF, their 3-D geometry, and thus their significance to future drilling at SAFOD. If these bodies are indeed steeply dipping, fault-parallel slivers, then as drilling proceeds NE toward the fault zone it is likely that the drill will encounter at least 4 faults that bound the two magnetic slivers.
T11F-08 INVITED 09:45h
Structure of the San Andreas Fault Zone and SAFOD Drill Site as Revealed by Surface Geologic Mapping and Seismic Profiling Near Parkfield, California
New geologic mapping, combined with high-resolution seismic reflection and refraction profiling, in the vicinity of the San Andreas Fault Observatory at Depth (SAFOD) deep bore hole near Parkfield, California indicates a wide and structurally complex San Andreas fault zone. The geologic and geophysical techniques both indicate the location and orientation of previously unmapped faults. The San Andreas fault zone near the SAFOD site is at least 5 km wide and is dominated by faults oriented subparallel to the surface trace of the San Andreas fault. Subsurface orientation of faults indicate that the San Andreas fault zone is composed of at least two flower structures in the upper 3 km, with the main trace approximately centered on the more easterly of these structures. The SAFOD drill site is located near the center of the other flower structure. A third flower structure may lie farther to the southwest of our studies. The flower structures likely merge at depth, but at least in the upper 3 to 4 km (the target depth of the SAFOD main hole), they are distinct features. Based on inferred ages of rock types juxtaposed across faults and geomorphic evidence along faults, it appears that the plate boundary (San Andreas fault zone) near the SAFOD site has migrated eastward during the past approximately 20 million years by forming new flower structures.