关键词: 同震破裂, 时移地震, 浅源海域地震, 孔隙流体运移

Abstract:

Co-seismic near-surface rupture is one of the keys to the recognition of earthquake fault and the defense to seismic hazard. However, conventional investigation methods such as outcrop mapping and trenching, are often disturbed by the variation of capping formation. Besides, it’s difficult to apply these methods under the sea water. Drawing on the idea of time-lapse seismic techniques in the petroleum industry, we suggest identifying the buried co-seismic ruptures using two reflection seismic data sets acquired before and after earthquake, respectively. In this paper, a case study is presented.
On Nov. 26^th, 2018, a M_S6.2 earthquake occurred in the south Taiwan Straits. The focal mechanism of this event is dextral strike slip with a slight dip-slip component, and the aftershock distribution is E-W oriented. West of the epicenter, multi-channel seismic profiling was carried out twice under the direction of Fujian Earthquake Agency in 2017 and 2019. To avoid the influence of the difference in acquisition conditions, before comparing we reprocessed and cross-equalized the two data sets with the same de-noise method, illumination, migration algorithm and velocity field. The profile correlation in 20~50Hz shows that the dominant reflecting wave groups are coincident with the time-lapse ones, which means the two sections are in phase.
The comparison results show that: at about 25km west of the epicenter, the reflection profile met the earthquake fault inferred by focal mechanism, and the morphology of Fault F₁ did not change significantly after the earthquake, but at depths greater than 400m, the remarkable reflections near the strike-slip fault plane changed significantly. From 2017 to 2019, the strongest reflection in the hanging wall reduced in amplitude and shifted from near horizontal to a jagged fold in shape, besides, the polarity of two remarkable reflections reversed, and a piece of the basement reflection in the heading wall close to the fault plane subsided about 8milliseconds measured in two-way travel time. The remarkable reflections on the rest of the profile were aligned accurately as the control. These phenomena can be interpreted as a fluid migration through sandstone fractures model perfectly, which is also consistent with the petrological features in the study area. Since the gap between the two acquisition dates is only 26 months and there was no human activity affecting subsurface structure in the vicinity, the pore fluid migration is inferred to be related to the 2018 event.
This study demonstrates that although the vertical co-seismic displacement is smaller than the resolution of reflection seismic profiles in most cases, the near-surface fluid migration which accompanied the co-seismic rupture may cause significant impedance changes near the fault plane, and such changes can be reliably identified on time-lapse seismic profiles. Compared with the conventional investigation methods for co-seismic ruptures, the time-lapse seismic method can overcome the displacement absorption of capping formation and expand the identifiable scope of co-seismic ruptures. This method is practicable, especially for marine earthquake researches, because the acquisition repeatability and surface consistence of the marine reflection seismic data are relatively better than that of the land reflection seismic data. This study provides a new idea for recognizing the earthquake fault and slip distribution of shallow source earthquakes, which is especially important for the study of marine earthquakes with fewer geology and geodesy data available.

Key words: co-seismic rupture, time-lapse seismic, shallow-focus marine earthquake, pore fluid migration