Event Locations with a Long Borehole Array: An Example from SAFOD
Abstract
In the summer of 2007 the SAFOD project completed its third phase of drilling multilateral wells through the San Andreas Fault in Parkfield, California. Prior to this drilling, numerous monitoring activities took place with purpose of characterizing the area of the drilling target, i.e. repeating magnitude 2 microearthquakes, and the associated fault structure. This characterization was conducted using surface networks of seismometers and borehole networks deployed in the Pilot Hole and within the Main Hole during the earlier stages of the project. One of these borehole networks consisted of 80 three-component seismometers and was deployed at depths ranging from 2.7 to 1.5km below the surface. This array operated in the highly deviated section of the well (~35 deg) within the low noise environment of the Main Hole for a period of two weeks. It also recorded the general microseismic activity of the SAF as well as surface explosions. Analysis of the active data has resulted on a more accurate velocity model as well as partial images of the San Andreas Fault at depth. In addition the recording of one of the target events by the array was critical for the planning of the final well trajectory. In the current study we analyze the remaining microseismic data recorded by the array. This includes mostly local events (i.e. aftershocks of the 2006 Parkfield event), SAFOD’s target event, in addition to larger regional ones. The data was recorded continuously at high sampling rates of 4kHz with the purpose of capturing the full frequency spectrum of small size events that may present larger corner frequencies. A preliminary catalog of the events was prepared. The location made use of a representative 3D velocity model of the study area. The automatic event location procedure followed basics steps of event detection, P- and S-wave onset picking P-wave polarization analysis and directed grid search inversion to obtain the best location within the 3D velocity model. Basic source parameters were also estimated to investigate scaling relations. Given the deviated nature of the well we have examined the effect of the geometry in the overall event location of the data. A number of events recorded by the array were also recorded by the surface networks. However, the quiet environment in the borehole enabled us to record microearthquakes that in some cases were not even recorded along the entire array. Besides its high level of event detection, the high vector fidelity of the 3C geophones allowed for precise particle motion analysis of first arrivals to determine the location of microearthquakes recorded during this effort. The use of the long array was investigated as a tool for locating events in comparison to shorter arrays of seismometers. Here we analyze how a long antenna can provide wider angular apertures with which more accurate event locations can be conducted.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2009
- Bibcode:
- 2009AGUFM.S23B1750C
- Keywords:
-
- 7215 SEISMOLOGY / Earthquake source observations;
- 7219 SEISMOLOGY / Seismic monitoring and test-ban treaty verification;
- 7250 SEISMOLOGY / Transform faults;
- 7294 SEISMOLOGY / Seismic instruments and networks