GEOSCAN, résultats de la recherche


TitreUncertainty estimation of shear-wave velocity structure from Bayesian inversion of microtremor array dispersion data
AuteurDosso, S E; Molnar, S; Cassidy, J
SourceAmerican Geophysical Union, Fall Meeting 2010, abstract volume; 2010, 1 pages
LiensOnline - En ligne
Séries alt.Secteur des sciences de la Terre, Contribution externe 20100162
RéunionAmerican Geophysical Union Annual Fall Meeting; San Francisco, CA; US; décembre 13-17, 2010
Mediaen ligne; numérique
Sujetsvitesse des ondes sismiques; vélocités acoustiques; sismo-sondages; ondes sismiques; ondes R; géophysique
ProgrammeÉvaluations ciblées des dangers dans l'Ouest du Canada, Géoscience pour la sécurité publique
Résumé(disponible en anglais seulement)
Bayesian inversion of microtremor array dispersion data is applied, with evaluation of data errors and model parameterization, to produce the most-probable shear-wave velocity (VS) profile together with quantitative uncertainty estimates. Generally, the most important property characterizing earthquake site response is the subsurface VS structure. The microtremor array method determines phase velocity dispersion of Rayleigh surface waves from multi-instrument recordings of urban noise. Inversion of dispersion curves for VS structure is a non-unique and nonlinear problem such that meaningful evaluation of confidence intervals is required. Quantitative uncertainty estimation requires not only a nonlinear inversion approach that samples models proportional to their probability, but also rigorous estimation of the data error statistics and an appropriate model parameterization. A Bayesian formulation represents the solution of the inverse problem in terms of the posterior probability density (PPD) of the geophysical model parameters. Markov-chain Monte Carlo methods are used with an efficient implementation of Metropolis-Hastings sampling to provide an unbiased sample from the PPD to compute parameter uncertainties and inter-relationships. Nonparametric estimation of a data error covariance matrix from residual analysis is applied with rigorous a posteriori statistical tests to validate the covariance estimate and the assumption of a Gaussian error distribution. The most appropriate model parameterization is determined using the Bayesian information criterion (BIC), which provides the simplest model consistent with the resolving power of the data. Parameter uncertainties are found to be under-estimated when data error correlations are neglected and when compressional-wave velocity and/or density (nuisance) parameters are fixed in the inversion. Bayesian inversion of microtremor array data is applied at two sites in British Columbia, the area of highest seismic risk in Canada, to study the ability to recover an accurate VS profile in different geological settings: thick (> 200 m) unconsolidated deposits on the Fraser River delta where the BIC indicates a power-law VS profile to 110 m depth, and a shallow soil layer over basement in Victoria is modeled as a weak linear gradient to 15-18 m depth over a half-space. The recovered VS profiles at both sites are in excellent agreement with invasive VS measurements from seismic cone penetration testing and/or surface-to-downhole methods.