SHIPS - Amplification
Urban Seismic Experiments Investigate Seattle Fault and Basin
Seismic Wave Amplification in Seattle Basin
Radial (east-west) horizontal component recordings for the Mw = 7.6 Chi-Chi (Taiwan) main shock of September 20, 1999 [Shin et al., 2000], show amplification of the shear-wave arrivals in the Seattle basin relative to sites in the Olympic Mountains (Figure 3, below, enlarged figure and caption).
The (east-west) horizontal component corresponds to the radial component because the azimuth of propagation of these arrivals was 89_E. The seismic recordings were made using identical RefTek models, geophones, and gains and hence the amplitudes of the recordings are directly comparable. Waveforms in Figure 3 are aligned on the predicted time for the S-wave arrival (calculated from the iasp91 standard earth model) and then shifted using cross correlation for optimal alignment of the waveforms. Traces are shown in true relative amplitude and have been low-pass filtered with a corner at 0.25 Hz (4 s period).
Similar amplification in the Seattle basin is observed for the shear-waves recorded on the transverse (N-S) horizontal components and for compressional-wave arrivals recorded on the vertical component (Figure 3). Note the large (factor between 6 and 10) amplification of the signal in Seattle, in the middle of the Seattle basin, relative to sites in the Olympic Mountains west of Hood Canal and to sites in the Cascade foothills, outside of the Seattle basin. Figure 3 shows in addition that the duration of compressional and shear wave arrivals at stations in the Seattle basin (approaching 100 seconds in Seattle) is significantly longer than for stations located outside of the Seattle basin (Figure 3).
In Figure 3 we also show the approximate E-W geometry of the base of the Seattle basin, compiled from Wet SHIPS and gravity results and preliminary inversions of Dry SHIPS first arrival times. There are two possible explanations for the strong correlation between the geometry of the basin filled with sediments and the amount of amplification. The first possibility is that the amplification results from focusing associated with the entire basin. The second possibility is that the amplification results from resonances within specific layers in the basin, probably the uppermost, lower velocity Quaternary deposits, whose geometry may mirror the geometry of the entire basin.
Dry SHIPS also provided a new understanding of the eastern end of the Seattle basin and Seattle fault.
Due to the sparsity of receivers in this region during Wet SHIPS, tomography models resulting from our Wet SHIPS study failed to image this end of the basin. Gravity inversions for the depth to the top of volcanic basement rocks in the Seattle basin, have greater uncertainty in the eastern end of the basin. This greater uncertainty is a result a lower density contrast between the basin fill and the Tertiary basement rocks in the east relative to the western part of the basin, where higher density Crescent Formation volcanics form the basement.Fig 4. Reduced record section for Shot point 1 in the Olympic Peninsula showing vertical component traces only.
Note the large time delays associated with the Seattle basin and evidence for vertical steps (breaks) in the basin floor.
Record sections from Dry SHIPS, such as Figure 4 (above), clearly show a prominent travel time delay introduced by the lower velocities of the basin filling sediments extending eastward to the foothills of the Cascades. The distance over which travel times are delayed by basin sediments (Figure 4) implies that the Seattle basin is approximately 75 km wide, as opposed to the 60 km inferred from gravity data.
Thus the length of the Seattle fault is at least 75 km, assuming that the Seattle basin was formed by the fault [Johnson et al., 1994]. This length is in fairly close agreement with the 85-km length assumed by Pratt et al. .
Standard fault-length earthquake-magnitude relations [Wells and Coppersmith, 1994] suggest that the Seattle fault could produce a magnitude 7.2 to 7.6 earthquake, for a fault length of 75 km, fault plane depths between 20 and 30 km and fault plane dips between 35° and 60°.
The dip of the Seattle fault has not been well resolved by SHIPS, and has been estimated to lie between 20° and 70° [Johnson et al., 1994, 1999; Pratt et al., 1997]. Although uncertainty in the dip of the fault has only marginal significance for calculations of earthquake magnitude, it is vital for calculations of the total fault slip rate and for the response of the basin to strong shaking.
Determining the origin of the amplification of the seismic waves within the Seattle basin remains a high-priority research topic.
Is the basin amplification mainly determined by the thickness and seismic velocities of the upper, unconsolidated sediments, or is the geometry and seismic velocities of the entire basin fill important?
Merging of the data from all three SHIPS experiments remains to be completed. This merging will provide the data needed for final 3-D tomographic imaging of the entire Seattle basin, particularly that part of the basin to the east of Seattle which was not imaged by Wet SHIPS. A site response study based on the comparison of the seismic amplitudes generated by the Kingdome implosion and our four shots as recorded during Kingdome SHIPS is just underway.
Finally, analysis of our recordings of the Japanese Volcano Island earthquake will provide detailed information about variations in site response in the Seattle area to compare to our basin-scale results from the Chi Chi earthquake recorded during Dry SHIPS.