GEOSCAN Search Results: Fastlink


TitleProbabilistic site characterization based on Bayesian inversion of ambient vibration array recordings in SW British Columbia, Canada
AuthorMolnar, S; Ventura, C E; Finn, W D L; Taiebat, M; Dosso, S E; Cassidy, J F
SourceProceedings of the 15th World Conference on Earthquake Engineering; 2012 p. 1-10
Alt SeriesEarth Sciences Sector, Contribution Series 20120134
Meeting15th World Conference on Earthquake Engineering; Lisbon; PT; September 24-28, 2012
ProvinceBritish Columbia
AreaFraser River Delta
Lat/Long WENS-123.5000 -123.0000 49.2500 49.0000
Subjectsgeophysics; seismic velocities; velocity surveys; seismic waves; r waves
ProgramTargeted Hazard Assessments in Western Canada, Public Safety Geoscience
One of the most important properties for characterizing earthquake site response is the shear-wave velocity (VS) profile. The microtremor array method extracts phase velocity Rayleigh wave dispersion curves from recordings of ambient vibrations for inversion of VS structure at a site. A variety of sediment sites with diverse amplification behaviour are chosen to evaluate the ability of this passive-source method to provide VS profiles for site response characterization in SW British Columbia, Canada. A Bayesian inversion method, with evaluation of data errors and model parameterization, is applied here to produce the most-probable VS profile together with quantitative uncertainty estimates. The resulting VS profile probability distribution at each site generally demonstrates: (1) a well resolved VS gradient within the upper sediment layer; and (2) poor resolution of the basement layer VS, consistent with surface seismic methods in general. Probabilistic site response analyses are performed, in which common predictors of site amplification, including VSZ (harmonic average VS to a depth z) and linear 1D amplification spectra, are computed using a sample of VS profiles drawn from the posterior probability density of the Bayesian inversion. Site amplification probability distributions are shown to be more informative than amplification estimated for a single best-fit VS profile by characterizing the uncertainty and therefore level of confidence in the predictions. Amplification probability spectra are evaluated by comparison to empirical earthquake and microtremor spectral ratios, with generally good agreement. This result provides confidence that when the VS profile probability distribution from Bayesian inversion of dispersion data is included in theoretical site amplification calculations the primary influence of site-specific structure is accounted for appropriately.