Charles University, Faculty of Mathematics and Physics, Prague, Czech Republic
E-mail: io_at_karel.troja.mff.cuni.cz
, jiri.zahradnik_at_mff.cuni.cz http://geo.mff.cuni.cz
I. Oprsal, J. Zahradnik
Charles University, Faculty
of Mathematics and Physics, Prague, Czech Republic
E-mail:
io_at_karel.troja.mff.cuni.cz , jiri.zahradnik_at_mff.cuni.cz
http://geo.mff.cuni.cz
In Proceedings of Third
International Symposium on the ‘Effects of Surface Geology on Seismic
Motion’, Grenoble, France, 30 August – 1 September 2006,
Laboatoire Central des Ponts et Chausees, Paris, vol. 2, Paper No 075,
10pp.
Introduction
International experiment of strong motion prediction at the
Turkey Flat test site near Parkfield, California (M6, Sep. 28, 2004
earthquake) was a two-phase experiment, where local site-specific
structural data and two strong motion records were released to assist
in the ground motion simulation. In our approach, record R1 (rock site)
was firstly fit by finite-extent source modelling. Afterwards, using
individual 1D crustal models and the same source model, complete
numerical simulation was accomplished for sites R2, V1, V2, D1, D2, D3
in the 0 to 40 Hz band, comprising required acceleration time
histories, Fourier spectral ratios, peak ground motion values and
response spectra. In the second phase of the experiment the D3
(downhole) record was released to possibly modify the prediction.
Since, however, we found the previous blind simulation of the D3 record
(that from the 1st phase) acceptable up to 10 Hz, our 2nd phase
prediction results remained unchanged compared to the 1st phase. This
was the only technique among those applied by the participants in the
present blind prediction experiment which focused on the finite-extent
source effect.
download
article (PDF)
I. Oprsal, J. Zahradnik
Charles University, Faculty
of Mathematics and Physics, Prague, Czech Republic
E-mail:
io_at_karel.troja.mff.cuni.cz , jiri.zahradnik_at_mff.cuni.cz
http://geo.mff.cuni.cz