HR: 1340h
AN: S43B-1312    [Abstracts]
TI: Towards tomographic imaging with microtremor arrays: the application of Hilbert-Huang Transform
AU: * Chen, Y
EM: yoc03001@engr.uconn.edu
AF: University of Connecticut, 261 Glenbrook Road, Storrs, CT 06268,
AU: Liu, L
EM: lanbo@engr.uconn.edu
AF: University of Connecticut, 261 Glenbrook Road, Storrs, CT 06268,
AU: Mehl, R
EM: Robert.Mehl@uconn.edu
AF: University of Connecticut, 261 Glenbrook Road, Storrs, CT 06268,
AB: Using microtremor measurements to study subsurface structure is very attractive for both earthquake engineering and earth-scale seismology. Measurement of microtremors allows for is relatively easy and can be widely applied. Research in this area has gained great momentum in the last few years. Nevertheless, on the microtremor signal processing front, the conventional fast Fourier transform (FFT) remains the foremost tool for fundamental time-frequency analysis. We introduce a novel approach for microtremor analysis by employing the so-called ~{!0~}Hilbert-Huang transform (HHT)~{!1~} to extract more information from the microtremor array records. HHT is a combination of the process of empirical mode decomposition and the Hilbert transform to construct the Hilbert-Huang spectrum that contains the time-frequency information of the recorded signals. HHT does not require the signals under analysis to be linear and stationary. Unlike FFT, the HHT does not assume the time-domain signal has harmonic components over the entire frequency-domain; the energy only concentrates in a few intrinsic mode functions (and this is essentially true for many natural signals or noises). In this study, we applied HHT to the empirical Green~{!/~}s functions extracted from cross-correlation between microtremor station pairs and obtained the group and phase velocity dispersion information for the fundamental and higher mode Rayleigh waves. The inversion of shear-wave velocity profile based on these dispersion curves are in agreement with geological structure revealed by other methods with higher order accuracy. With a dense array it is possible to tomographically image the shear wave velocity structure at a given site.
DE: 3255 Spectral analysis (3205, 3280)
DE: 3270 Time series analysis (1872, 4277, 4475)
DE: 7290 Computational seismology
DE: 7294 Seismic instruments and networks (0935, 3025)
SC: Seismology [S]
MN: 2007 Fall Meeting