 |
| Titel |
Inversion of HâV in layered media from seismic ambient noise based on the diffuse field theory and on improved calculation of Green functions |
| VerfasserIn |
Francisco J. Sánchez-Sesma, José Piña, Antonio García-Jerez, Francisco Luzón, Mathieu Perton |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250097185
|
| Publikation (Nr.) |
 EGU/EGU2014-12735.pdf |
|
|
|
|
|
| Zusammenfassung |
| The microtremor H-V spectral ratio (MHVSR) is widely used to assess the dominant
frequency of soil sites. Measurements are relatively simple as only one station is needed. It
has been recently proposed a theoretical basis linking ambient noise vibrations with diffuse
field theory. In this theory the directional energy density computed as the average spectral
density of motion at a point, is proportional to the imaginary part of Green function at the
observation point. Appropriate normalization is crucial to make the experimental spectral
ratios closer to the theoretical counterpart. According to this theory the square of
H-V is twice the ratio ImG11 / ImG33, where ImG11 and ImG33 are the imaginary
part of Green functions at the load point for horizontal and vertical components,
respectively.
In order to efficiently compute the imaginary part of Green’s functions in a layered
medium we start from an integral on the complex k plane and, using Harkrider’s
nomenclature, separate formulae for body-, Rayleigh-, and Love-wave components to the
spectral densities are obtained. Then the poles allow for integration using the Cauchy residue
theorem plus some contributions from branch integrals.
It is possible to isolate pseudo reflections from ImG11 and thus constrain the inversion of
soil profile. We assess ImG11 removing the influence of illumination spectrum using the
H-V spectral ratio and an estimate of ImG33 (from an a priori model) by means of
ImG11=0.5(H-V )2*ImG33. It has been found that ImG33 is less sensitive to details of
stratigraphy. In fact, the Poisson ratio of the uppermost layer controls the slope in high
frequency. With the obtained model ImG33 can be updated and the estimate of ImG11 will be
improved.
ACKNOWLEDGEMENTS. This research has been partially supported by DGAPA-UNAM
under Project IN104712, by the MINECO research project CGL2010-16250, Spain, by the
EU with FEDER, and the AXA Research Fund. |
|
|
|
|
|
|