| Data obtained by the newly installed large dynamic range digital strong-motion network from the 1993 Tapu, Taiwan earthquake and its aftershocks are used to examine the applicability of the empirical Green's function approach to near source rupture mapping. By summing the Green's functions over 16 subfaults distributed over the rupture plane of the mainshock, attempts are made to simulate the recordings of the mainshock, thereby extracting the rupture velocity, the slip time, and the moment release for each subfault of the source. With a simple source model, genetic algorithms are used to optimize the waveform inversion over the frequency band form 0.1 to 2 Hz. The method used in this study gives an alternative to estimate ground motions and the overall properties of source image.;The EMD/HSA method [Huang et al., 1998] is a powerful time series analysis tool for nonlinear, non-stationary, and dispersive data such as seismic signal. Based on local properties of the signal, sharp identifications of imbedded signal structures are given by the definition of the instantaneous frequency and amplitude. This method is utilized in several seismological applications in this study. Examples show that this new method can provide high-resolution dispersion information extraction, nonlinear seismic data analysis, and quantification of stress correlations. |
