|
Thrust Area 3: Subsurface Imaging, Tectonics, Seismicity, and Source Parameters
The main activities have been in the areas of LARSE, the Standard 3D Velocity model, fault-zone trapped waves, and stress determinations.
LARSE II
Two of the three field components of the LARSE II experiment were conducted during the past year. The passive deployment consisted of some 90 broadband instruments deployed along a line from Santa Monica Bay to the Garlock Fault. The instruments recorded for 6 months during which there were many local and teleseismic events. The data from this experiment is being dubbed off and will be sent to the IRIS-DMC for archiving. A quick look at some of the events indicates that the array recorded some high-quality data.
The main part of the LARSE II experiment (see Figure 5)-- the active source survey, was done in October 1999. Some 1400 instruments were deployed along 6 lines to record the 93 explosive sources that were used. The main line went from Santa Monica Bay, through the Northridge epicenter, across the Transverse Ranges and Mojave desert to the Tehachapi Mtns. Dog-leg lines were run across the Santa Monica Mtns to examine the areas and strong focusing/defocusing that occurred during the Northridge event. The data from the active portion of the survey is still being downloaded from the various recording instruments. It will be placed in a common format, and made available to participating scientists by February 1, 2000.
Both the permitting of this survey and the actual data acquisition were significant tasks that were accomplished successfully. Permits for the 93 shots in LARSE II involved 43 landowners, and permits for the 1400 seismographs involved 332 landowners. The landowners included federal, state, county, and city governments, school districts, utility companies, and private landowners. The shots and seismographs were located in national forests, parks, golf courses, utility easements, schoolyards, and private fields. Drilling took approximately 2 months; loading, 3 weeks; deployment, 3 days, shooting, 5 nights; and cleanup, 3 weeks.
Preliminary plots indicate the data are generally quite good, and some are remarkable. For example, even several of the tiny shots (5-25 lbs) in the Santa Monica area and San Fernando Valley were visible as far northward as the Mojave Desert, 60-80 km away. Travel time and amplitude anomalies are visible in first arrivals near the San Andreas fault, and reflections are visible raw shot gathers from the middle and lower crust. We look forward to good resolution of the shapes of the sedimentary basins and faults along the LARSE II corridor.
The high-resolution part of LARSE II -- the 8 km Vibroseis line through Santa Monica and L.A. has been delayed until 2000, because of permitting problems with the City of Los Angeles. An analysis of aftershock data from the Northridge earthquake, indicates that the target structure may be somewhere in the 0 - 2 km depth range in the survey area.
LARSE I
The onshore-offshore data recorded on Line 1, along with the OBS data taken at the same time, has been combined into a model for borderland region which shows crustal thickening beneath the Los Angeles and San Gabriel Basins, and the presence of a significant amount of low-velocity material (possibly Pelona Schist) in the mid-crust beneath the San Gabriel Mtns.
Receiver function analysis of the multi-component passive data recorded has been analyzed to look for the structure of the Moho. The results show a significant offset (> 1 km) in the Moho beneath the San Andreas fault (Figure 6), and the presence of a root beneath the San Gabriel Mtns.
Standard 3D Velocity Model
In January 1999, Version 1 of the standard 3D velocity model was released. This version contained the major basins of the greater Los Angeles area. The rest of the model was a smoothed 1D background velocity model. The model is being checked by comparing it against waveforms, traveltime data, direct borehole measurements of velocity and gravity. The comparison with borehole observations is generally fairly good, but it indicates that a more complicated representation of the basins may be needed. Both the travel time and waveform data indicate that he model is too slow in S-wave speed, probably reflecting the need for a lower Poisson's ratio in the basins. The density component of the model does a good job of predicting gravity in the areas where it has detail. The model is available from the data center and includes software and data to generate the Vp, Vs, and density fields on arbitrary meshes.
A second version of the model has been proposed and is scheduled to be ready by January, 2000. This version will incorporate some of the changes suggested through the testing and will add several new components. In particular, the new model will have a laterally varying background velocity that is a modified version of the Hauksson model. It will have a laterally varying Moho based on receiver function analysis of TriNet data. The topmost layer of the model (20-500m) will be a geo-technical layer that is based on a new soil classification map and shallow borehole data. Finally, an Imperial Valley basin structure will be incorporated into the model.
In addition to the standard model, the Harvard group is working a more detailed model for the Los Angeles Basin and Santa Monica Bay region. This model is based on oil industry reflection seismic lines and on borehole observations. The first version of the model is due in early 2000. We are looking in ways of combining parts of this model with the standard 3D velocity model.
Fault-Zone Trapped Waves
This year the goal of the fault-zone trapped waves was to try and determine how well the depth-dependence of the fault zone parameters could be determined. An experiment on an exhumed fault (the Punchbowl fault) was proposed and started. However, two days after the experiment was started, the Hector Mines earthquake occurred. This earthquake combined with the large number of instruments that were in S. California because of LARSE experiment seemed like too good of an opportunity to pass up, and consequently the focus of the study was shifted to the Hector' area. Two survey areas were established, each in the shape of a '+'-sign, with a 1 km length along the fault and a 1/2 km perpendicular to it. The sites were on hard rock, along fairly simple segments of the fault. Each array had 31 stations, with the southern one having an additional 20 instrument array set up in one quadrant of the '+' sign. The data from these two areas should be available in early 2000.
Stress
Lateral variations in stress orientations determined from earthquake mechanisms are being used to look at complicated interactions of stress regimes. For example, the Los Angeles region exhibits a strike-slip near the surface and thrust at depth, indicating that simple 2D models of local tectonics may not be adequate. The San Andreas west of Cajon Pass is in a state of stress consistent with thrust faulting, while to the east it is in a strike-slip regime, implying that little convergence is taking place across the San Bernardino segment.
|
system
Southern California Earthquake Center
@ 