Presenter: Ruei-Jiun Hung
Date: 2017/03/23
Abstract
The Taiwan Chelungpu-fault Drilling Project (TCDP) was operated to understand the fault zone characteristics after the 1999 Mw 7.6 Chi-chi earthquake. Seven Borehole Seismometers (TCDPBHS) were installed through the identified fault zone to monitor the microseismic activities and the fault-zone seismic structure properties. This study aims to understand the fault zone anisotropy after the Chi-chi earthquake. We first calculated cross correlation functions (CCFs) of the ambient noise field between every pair of the stations. Unfortunately, the absence of homogeneous noise energy distribution results in poor phase determination, and phase traveltimes from CCFs cannot provide explicit result to identify the fault-zone anisotropy. We, thus, focus on the understanding of noise configurations in this area. The probability density functions of seismic noise power spectra show diurnal variations in the frequency band of 1-25 Hz, suggesting human-generated sources are dominated in this frequency band. We also analyze the Fourier spectra by rotating every 5 degrees to search for the maximum background energy distribution. The result shows that the spectral amplitudes are stronger at NE-SW direction, with shallow incident angles. In order to suppress the possible phase traveltime errors due to the inhomogeneous noise distribution, CCFs analysis using coda waves from earthquakes are considered, where the coda waves could be feasible for retrieving the reliable empirical Green’s function. Better results are obtained, and we can identify the azimuthal anisotropy, especially in BHS4 (close to the slip-zone).The fast shear wave polarization (FSP) is about 110˚ from North. That is consistent with the stress regime in this area obtained from in-situ measurement. We also use shear wave splitting from micro-events recorded by TCDPBHS. The FSPs are mostly consistent with the direction of tectonic convergence, as well. However, FSPs recorded at fault-zone station show 90 degrees shifting compare to those recorded on hanging-wall side. Both two approaches enhance the consistency in results to give hints on fault zone anisotropy. Prolong studies of the fault zone anisotropy from TCDPBHS through the methodology stated above, we hope to resolve the possible temporal variation of the fault zone anisotropy 8-18 years since a nearly 12m large slip.