2023.04.21 15:00 星期五报告会
韦生吉 南洋理工大学亚洲环境学院 新加坡对地观测中心(EOS)Creep and seismic rupture of a serpentinite-rich Sumatran fault segment

2023-04-17

Creep and seismic rupture of a serpentinite-rich Sumatran fault segment

韦生吉 副教授

南洋理工大学 亚洲环境学院

新加坡对地观测中心(EOS)

报告摘要:

Earthquake is produced by shear dislocation of rocks across the fault, the frictional status and the area of locked/creeping patches on the fault thus govern the size and occurrence of damaging earthquakes. To better understand these fundamental earthquake physics issues, we deployed over 130 short period seismic nodal stations along the plate boundary type Sumatran fault in Aceh region to cover a segment that was reported to be creeping at various depths. We maintained the nodal array deployment from Jan 2020 to July 2021 by recharging the nodes every 35 days. A machine learning based earthquake detection algorithm was applied to the acquired dataset, which results in a high-resolution seismic catalog that has more than 8000 micro-seismic events. These events clearly delineate the subvertical creeping segment of the Sumatran fault and its Seulimeum branch to the northwest. The seismicity on the creeping segment is almost uniformly distributed from 3 to 12 km in depth, confirming the creeping nature of the fault segment as revealed by geodetic observations, but providing a much more accurate depth constraint. In contrast, the Seulimeum fault branch shows a much deeper seismicity at the depth range of 18 to 25 km, indicating the entire upper crust is fully locked. Sharp stepovers are observed along both strike (~10km) and strike-normal (~4km) directions between the seismicity on these fault segments. The creeping segment of the Sumatran fault, as defined by similar earthquake families, agrees well with the lenses of serpentinite, which has much smaller frictional coefficient that facilitates fault creep. Similar earthquake families show ~ km scale lineation along strike of the fault, where repeating earthquake pairs are identified. However, two shallow Mw6 earthquakes occurred on the creeping segment in the last 25 years. In particular, finite fault inversion of the 2013 Mw6.1 earthquake shows the rupture from 12 km to the surface. These observations suggest a partially creeping/locking or conditionally stable frictional status on the serpentinite-rich segment of the Sumatran fault, that should be considered in both single event and earthquake cycle simulations, as well as seismic hazard assessment.

报告人简介:

Wei Shengji has been a Principal Investigator at the Earth Observatory of Singapore (EOS) and Associate Professor at the Asian School of the Environment (ASE), both at Nanyang Technological University (NTU), Singapore, since 2014. He is also the Assistant Chair (Research) at ASE. Before he joined NTU, he completed his post-doctorate training at the Seismological Laboratory at the California Institute of Technology. He obtained his bachelor and PhD degrees in geophysics from the University of Science and Technology of China in 2004 and 2009, respectively.

Shengji’s research focusses on earthquake source studies both in real-time and after in-depth. He is also interested in the properties of structures to better understand earthquake sources and tectonic and geodynamic processes. His research experiences cover both natural and man-made earthquakes. Shengji’s approach to resolving the kinematic nature of earthquakes involves a combination of geodesy, geology, and seismology to better constrain the spatial-temporal evolution of seismic rupture properties. Besides reconciling various datasets, he has also developed techniques that allow waveform inversion/modelling to be extended to a higher frequency (>1Hz) range of relevance for damage assessment. He and his research team are investigating the following subtopics: earthquake rupture process imaging, strong ground motion simulations, earthquake focal mechanism inversion, crustal to upper mantle scale velocity structure inversion, induced seismicity and rock mechanics, Southeast Asia seismology, and volcano seismology.