13:30 〜 13:45
[S03-1-01] Envelopes of scalar plane wavelets propagating through 2-D random media with power-law spectra
The envelope broadening and the peak delay of the high-frequency S wavelets from regional earthquakes are caused by scattering due to random velocity inhomogeneities in the earth. Several stochastic models have been proposed so far to describe these phenomena, such as the radiative transfer theory, the diffusion approximation, and the Markov approximation. Recently, Sato (2016, Geophys. J. Int.) proposed a new stochastic model to synthesize the envelope of a scalar wavelet radiated from a point source in a 3-D von Karman-type random medium having a power-law spectrum. The point is to split the spectrum into two components using the center wavenumber of the wavelet; the long-scale component produces the envelope broadening and the peak delay by multiple forward scattering, and the short-scale component causes amplitude attenuation by wide-angle scattering. The former and latter phenomena are evaluated by the Markov and Born approximations, respectively. Later, Sato and Fehler (2016, Geophys. J. Int.) extended Sato's theory into the case of a wavelet from a point source in a 2-D random medium. They showed that the synthetic envelopes, except for coda, match well the results of finite difference (FD) simulations. Emoto and Sato (2016a, Abstr. JpGU Meet.; 2016b, Abstr. Seism. Soc. Jpn. Fall Meet.) performed 3-D FD simulations and obtained the results consistent to the theory. In this study, we conducted FD simulations of scalar plane wavelets propagating through 2-D von Karman-type random media. We followed Sato and Fehler (2016) concerning the choice of the model parameters. We produced 15 media with different random seeds, and set 7 receivers at each propagation distance in every medium. The simulated seismograms were squared and stacked to obtain FD envelopes. We found they totally agree well with the present theory of a plane-wavelet version, except for the small von Karman order and the low center wavenumber of the wavelets.