S41C-01 08:05h
Anomalous Intensities in the Gangetic Plain Following Great Himalayan Earthquakes
We have examined details of the intensity distribution of the sediments of the Punjab, Ganges and Brahmaputra river valleys following the Ms=8.1 Shillong (1897), Ms=7.8 Kangra (1905), and Ms=8.2 Bihar/Nepal (1934) earthquakes. For each earthquake we subtract the observed MSK intensities from a synthetic intensity derived from an inferred planar rupture model of the earthquake, combined with a homogeneous attenuation function derived from instrumentally recorded earthquakes. The resulting residuals are contoured to identify regions of anomalous intensity caused by local site effects. Observations indicative of liquefaction are treated separately from other indications of shaking severity lest they inflate inferred residual shaking estimates. Despite this precaution we find that intensites are 1-3 intensity units higher near the major rivers, as well as at the edges of the Ganges basin. We find evidence for a post-critical Moho reflection from the 1897 and 1905 earthquakes that raises intensities 1-2 units at distances of the order of 150 km from the rupture zone, and we find that the 1905 and 1934 earthquakes may have respectively triggered earthquakes at Dehra Dun and, more speculatively, near Agra. The 1934 earthquake shows an intensity low to the east suggestive of westward directivity. Four or more M=8 earthquakes are apparently overdue in the region based on seismic moment summation in the past 500 years. Results from the current study permit anticipated intensities in these future earthquakes to be refined to incorporate site effects derived from dense macroseismic data.
S41C-02 08:20h
Absolute site effects in Kachchh, India, determined from aftershocks of the 2002 Bhuj earthquake.
What can be learned about absolute site effects on ground motions from recordings of aftershocks at ten temporary seismic stations, none of which could be considered a "reference" (hard rock) site, and for which no geotechnical information is available? This challenge motivated our current study of Bhuj aftershocks; and our answer, briefly put, is: quite a bit. We started by constraining the regional attenuation and geometric spreading: this was the result of an earlier study [Bodin et al., BSSA 2004], the goal of which was to be able to reproduce the general character of the observations with a constrained set of stochastic synthetic ground motions. Our present work is based on the same aftershock data we used in the prior study. We first produced stable and reliable, unbiased source moment-rate spectra using the technique described by Mayeda et al., [BSSA, 2003]. With these known "absolute" source spectra, and the propagation terms we quantified in the previous study we inverted for the site response using only the largest ~200 earthquakes (M$>$2.8) in each of two depth ranges (0-25 km, and 20-40 km), to yield the "absolute" site terms for horizontal and vertical ground motions. We were able to obtain stable results in the 1-14 hz frequency band. The results reveal that the site terms generally share a common character: small amplifications (near unity) at the longer-period end of the pass-band, and decreases (perhaps due to attenuation or near-site scattering) at the higher frequency end. This character is evident in a similar study of earthquake ground motions in the Alps at sites on hard rock [Malagnini et al., BSSA 2004]. In contrast to Alpine hard rock sites, however, the vertical site terms at our sediment and soft-rock sites are generally rather flat and featureless. We observe differences in site response between stations which appeared to be on similar geologic conditions, and vice versa. For sites that appear to be on deep unconsolidated soils, amplification of horizontal motion is evident at the longer periods.
S41C-03 INVITED 08:35h
Site Reading Blues in the Mississippi Embayment
The Mississippi embayment of the central U.S. is the location of a major intraplate seismic zone with high earthquake strong motion hazards combined with unique geological structures that only serve to confound the usual notions of "source" excitation, wave propagation along the "path", and "site" response. Thick, unconsolidated sediments of the embayment are the usual suspects in assuming that strong ground motions from nearby earthquakes will be significantly amplified through their dramatic decrease in impedance from deeper rocks and through near-surface, site resonance effects. Near-surface resonance can be inferred from local microearthquake recordings and dominates the waveform of P and S waves. Indeed, the universal character of near-surface velocity structure has been inferred from detailed modeling of ground motions from an extraterrestrial source - the atmospheric acoustic shock wave from a large bolide that impacted the atmosphere last November in the region. Acoustic wave coupling with the ground shows that S wave and, often, P wave velocities are significantly less than the P wave velocity of air and that the substrate at depths of about 12 m produces a sharp boundary causing significant high frequency resonance effects. Nevertheless, engineering and seismological wisdom suggests that the thick unconsolidated sediments should be significantly anelastic at high frequency so that strong ground motions should be essentially damped away. Results of the Embayment Seismic Excitation Experiment demonstrate that previously inferred low seismological Qs's of 30 or less in the sediments are severely underestimated and that the bulk Qs is greater than 100 for Rayleigh waves that propagate entirely within the unconsolidated sedimentary column. Empirical analysis of regional S wave coda (Jemberie and Langston, this meeting) shows significant amplification of coda waves within the embayment of factors up to 18 relative to stations outside of the embayment. Although interpreted as a "site" effect, simple plane wave models for this amplification cannot predict the magnitude of the amplification. The long duration and amplitude of reverberations seen in the regional S waves also require significant deeper structure within the Paleozoic basins of the Reelfoot rift. Plane wave models demonstrate that "site" response is a complicated function of the underside reflectivity of structure in the upper 10 km of the crust and the decrease in impedance of the unconsolidated sediments. However, the distinction of source-path-site becomes completely blurred if earthquakes can rupture into the sediments exciting large, high frequency surface waves.
S41C-04 08:55h
Attenuation Within Sedimentary Basins, and the Shapes of Site Response Curves in the Puget Lowland, Washington State
Comparison of simple spectral ratio (SSR) and horizontal-to-vertical (H/V) site response estimates at 47 sites in the Puget Lowland of Washington State document significant attenuation of shear wave arrivals within the sedimentary basins there. Amplitudes of the horizontal components of shear-wave arrivals from three local earthquakes were used to compute SSRs with respect to the average of two bedrock sites, and H/V spectral ratios with respect to the vertical component of the shear-wave arrivals at each site. SSRs at thick basin sites show peak amplifications of 2 to 6 at frequencies of 3 to 6 Hz, and decreasing spectral amplification with increasing frequency above 6 Hz. In contrast, SSR site response curves at non-basin sites show a variety of shapes. We interpret the spectral decay at basin sites to attenuation within thick basin strata. Computing the frequency-independent, depth-dependent attenuation factor (Qs,int) from the SSR spectral decay gives values of about 30 for shallow sedimentary deposits and 220 for the deepest sedimentary strata. H/V site responses show less spectral decay than the SSR responses but contain many of the same resonance peaks. We hypothesize that the H/V method yields a flatter response because P-to-S mode conversion occurs at shallow horizons within the basin, such that attenuation in the deep basin strata influences both the horizontal and vertical components of the shear-wave arrivals to nearly the same degree. Correcting the SSR site responses for attenuation within the basins results in better agreement between SSR and H/V estimates of site response.
S41C-05 INVITED 09:10h
Seismic Wave Amplification in Las Vegas: Site Response and Empirical Estimates of Ground Motion
This presentation will summarize a multidisciplinary effort to understand seismic wave amplification in Las Vegas Valley. The project involves weak motion recording and analysis, geotechnical and seismic refraction field studies, geologic and lithologic interpretation and model building. We will provide a brief overview of the project, then focus on specifics of seismic wave amplification including observations and interpretations. We analyzed recordings of nuclear explosions from the Nevada Test Site (NTS) and regional earthquakes to estimate site response in Las Vegas. An empirical transfer function method was used to transform ground motion time-series at one (reference) station to other stations, using frequency dependent site response curves in the band 0.2-5.0 Hz. The method transforms the time-series to the frequency domain by Fast Fourier transform, multiplies the amplitude spectrum by the site response curve and inverse FFT's back to the time domain. The approach is validated by the ability to predict horizontal component S-wave ground motion measures, such as peak and rms ground velocities and accelerations. We then can provide empirical estimates of ground motion for a wider distribution of sites in Las Vegas. Frequency dependent amplifications (site response) and peak ground motions are strongly correlated with measures of shallow shear-wave (geotechnical) velocities. Details of the geotechnical measurements and models will be presented in a companion presentation. This work was performed under the auspices of the U.S. Department of Energy by University of California Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48.
S41C-06 09:30h
A Pilot Experiment in Collaborative Science and Engineering
We report on a pilot field experiment, meant to demonstrate the value of combining resources from multiple earthquake science and engineering facilities. Our initial intent was to employ a mobile shaker truck from the Network for Earthquake Engineering Simulation (NEES) facility with US Geological Survey (USGS) and Incorporated Research Institutions for Seismology (IRIS) seismographic instrumentation in a single study. The final experiment grew into four studies with additional participation from the Los Alamos National Lab and other NSF-sponsored consortia; the MAEC, SCEC, CENS, and HPWREN. The field experiment piggybacked on an earlier planned inaugural demonstration of NEES facilities, held at the NEES Garner Valley Digital Array (GVDA) site in southern California. We recorded ground motions generated by the `TRex' shaker truck at the GVDA and surrounding Garner Valley during August 18-22, 2004 and will continue to collect earthquake data for several months. Our first study tests the potential for generating and measuring nonlinear sediment response by recording TRex strong shaking on a temporary surface accelerometer micro-array and permanent GVDA down-hole accelerometers. Our second study focuses on ground motion site and basin effects by recording TRex signals at 20 temporary real-time, telemetered seismic stations deployed from the center to the edge of the basin. Earthquake signals recorded on this basin array will allow us to validate the process of extrapolating results from the artificial, surface source to those from natural earthquakes. The relatively small scale of the Garner Valley basin (4 km by 10 km, sediment depths less than 25 m) makes its characterization feasible, but scalable to larger basins elsewhere and motivates our third study in which we conduct a variety of imaging experiments. Basin array data will constrain broad-scale shear and compressional wave tomographic images of the basin structure. To constrain a higher-resolution image along a profile across the basin, we also collected reflection data generated by TRex recorded on densely spaced geophone strings, and conducted a sledge-hammer survey at even higher resolution to confirm the location of a suspected buried fault. For the fourth study, we assess the potential for shakers like TRex to do broad-scale, deep imaging as we stack TRex signals emitted repeatedly for nearly an hour, recorded on stations of the regional ANZA and statewide California Integrated Seismic Networks.
S41C-07 09:45h
Spectral Element Modeling of 3D Site Effects in the Alpine Valley of Grenoble, France.
Sitting on top of a 3D Y-shaped basin filled mostly with late quaternary deposits, the city of Grenoble (French Alps) is subject to strong amplification of seismic motion (see the SISMOVALP web site). In order to assess the magnitude and 3D complexity of these site effects, we propose a spectral element modeling approach previously applied to the prediction of strong ground motion in the Los Angeles sedimentary basin (Komatitstch et al., 2004). The spectral element method naturally accounts for depth variations of the free surface and of internal interfaces, such as the contact between the sediments and the bedrock. It is also well suited to model the propagation of surface waves generated at the basin edges. The 3D spectral element mesh honors the stiff surface topography of the mountains surrounding the city, as well as the bedrock depth obtained from extensive gravimetric measurements. In the basin, we use a generic 1D velocity model derived from geophysical measurements performed in a deep borehole that reached the substratum at 550 m depth in 1999. Results and comparison to data are shown in the time and frequency domain for small-size (Mw=2.5 and Mw=3.5) local events recorded in the past years. Then, a Mw=5.5 strike-slip event is simulated on the eastern border of the basin along the Belledonne fault, and the results are compared to those obtained by the method of Empirical Green Functions. References: http://www-lgit.obs.ujf-grenoble.fr/sismovalp/ Simulations of ground motion in the Los Angeles basin based upon the spectral- element method, Dimitri Komatitsch, Qinya Liu, Jeroen Tromp, Peter Süss, Christiane Stidham and John H. Shaw, Bulletin of the Seismological Society of America, vol. 94, p 187-206 (2004).