Outer Core, SmKS, AxiSEM3D
Study the structure of the outermost outer core by SmKS phases
The Earth’s magnetic field, which protects us from solar radiation, is generated by fast convection in the liquid metal Outer Core (OC) at 2891 km to 5150 km depth. However, the driving force behind this convection is still uncertain due to the challenges in direct observation. Utilizing the seismic waves generated by large earthquakes that penetrate the OC, seismologists provide the most important reference, the seismic velocity profile, for other geoscientists to constrain their chemical composition and convection models. Seismic observation of the OC is always limited by low data quality and complicated lower mantle structure. The motivation of this project is to overcome these difficulties and constrain the velocity structure in the outermost OC with a high-quality dataset and a new measurement method based on full 3-D elastic wavefield simulation by AxiSEM3D (Leng et al., 2019).
Seismicity, Cascade Volcano, Machine Learning
Investigating Tectonic, Magmatic, and Hydrothermal Micro-earthquakes in the Three Sisters Volcanic Complex
The Three Sisters Volcanic Complex, located in the Central Orogen, is considered one of the most threatening volcanoes in the US (wiki). Continuous ground uplift and magmatic volatile emissions west of South Sister suggest the possibility of ongoing magma intrusion into the upper crust (USGS report). Existing public seismicity catalogs show the area as nearly “silent” aside from a swarm in 2001 near Middle Sister, but long-term monitoring has been challenged by the presence of only two seismometers within the vast (287,000 acres) wilderness area. In this project, we used the data from a newly deployed dense array (5G) to better understand the seismogenic structure underlying the Three Sisters volcanoes. We applied a Machine Learning (PhaseNet, PyOcto) and Match Filtering (MF) workflow to detect and locate small seismic events that may have been overlooked by the existing seismic network.
Induced seismicity, Nearest-Neighbor-Distance clustering, earthquake swarms
Study the clustering features of induced seismicity in the Sichuan Basin, China.
We utilized the nearest-neighbor distance clustering approach (Zalipin and Ben-Zion, 2013a, b) to analyze earthquake interaction features in the Sichuan basin and conduct a careful comparison with the surrounding regions. After distinguishing background events from earthquake clusters, we measured background seismicity rates and categorized clusters into repeat, couple, star, chain, and complex based on their structure. We proposed and modified three quantitative attributes to measure the swarm-like feature of earthquake clusters. We found that the seismicity in the Sichuan Basin has more foreshocks in its earthquake clusters than the surrounding region. We conclude that earthquake interactions, especially foreshocks, play a pivotal role in triggering seismic events in the Sichuan basin and should be carefully incorporated into future models of induced earthquake rupture and risk assessment.