(1) Geophysical Institute, Academy
of Science of the Czech Republic, Prague
The SVEKALAPKO broadband part of the seismic array was used to study seismic anisotropy. A detection of splitting of shear waves of teleseismic events generally evidences upper mantle anisotropy. To model the anisotropy we evaluate two splitting parameters - the time delay between the fast and slow shear waves, and direction of the fast S polarisation. Teleseismic core shear waves (SKS and SKKS) for several selected high-quality events were used to evaluate the splitting parameters.
At first, parts of seismograms containing shear waves were denoised and filtered. We assume that: (1) sufficiently long part of the seismogram before the arrival of the SKS wave contains only noise (without any other seismic phases close to the SKS) and (2) the noise loaded to the SKS wave has the same pattern as that part analysed in (1). To remove noise from the analysed shear waveform, we apply characteristics of the noise in all three components separately. The denoised seismograms were filtered with Butterworth band-pass 3-60s of the third order. To determine the splitting parameters we minimise energy on the transversal component. Based on the generalised method by Silver and Chan (1991), we search for direction of the fast shear wave polarisation in plane (Q-T), perpendicular to the vertical ray path plane (L-Q) of the quasi-shear phase. A rotation of the co-ordinate system in the plane (Q-T) by an angle psi, and a time shift dt imposed on the shear wave components, yield the splitting parameters (polarisation vector) determined in 3D (Sileny and Plomerova, 1996). Measurement errors were estimated by the bootstrap method (Sandvol and Hearn, 1994) which consists of multiple inversions of simulated data that imitate the original data with different noise sequences. This allows us to calculate directly variance and covariance for all model parameters. The method enables us to distinguish among data with no apparent splitting, reliably split shear waves and noisy data.
We present evaluation of shear wave splitting for SKS waves of four high-quality records at the SVEKALAPKO array. We observe two kinds of variations of splitting parameters: (1) in dependence on direction of propagation, i.e., at a station the splitting parameters vary especially with back-azimuth of events (including 'null' measurements for certain back-azimuths) and (2) lateral variations, i.e., the splitting parameters of an event vary across the array. In general, shear wave splitting was detected at most of the stations and measured time delays were between 1 and 2 s, on average. Evaluation of shear-wave anisotropy indicates that lithospheric blocks with different anisotropic structures were found. The splitting parameters are coherent within the blocks. We demonstrate the lateral variations of the splitting parameters on two examples, as follows. For event 980928 (back-azimuth ~ 93 degs) we found 'null' splitting in the north-eastern block and the fast S wave consistently polarised approximately in the SWW direction in the central block. On the contrary, for event 990206 (back-azimuth ~ 45 degs) the fast S is oriented approximately to the south in the NE block and 'null' splitting was evaluated in the central block. The particle motion at stations close to the boundary between these blocks is complicated for both events. Similarly, stations close to the coast of the Baltic sea (FK11, FF15, FD13) seem to be located in another lithospheric block. The observed variations of splitting parameters are related to the surface trace of the Archean-Proterozoic boundary in Fennia, indicating a different anisotropic structure of the lithosphere in the two regions. To explain variations of the splitting parameters within the lithospheric blocks we need more complex models than a one-layered homogeneous anisotropy with horizontal symmetry axes.
Not available.
References
[1] Silver, P.G. and W.W. Chan, Shear wave splitting and subcontinental mantle
deformation., J. Geophys. Res., Vol. 96, pp. 16429-16454, 1991.
[2] Sileny, J. and J. Plomerova, Inversion of shear-wave splitting parameters
to retrieve three-dimensional orientation of anisotropy in continental
lithosphere, Phys. Earth. Plant. Int., Vol. 95, pp. 277-292, 1996.
[3] Sandvol, E. and T. Hearn, T., Bootstrapping Shear-Wave Splitting Errors.
Bull. Seismol. Soc. Am., Vol. 84, pp. 1971-1977, 1994.