Seismicity Behavior Within Rock Valley Illuminated by a Dense Nodal Deployment and Machine-Learning Methods
Description:
A dense 48-node array was deployed within Rock Valley in Nevada and crosses a number of faults that are within the valley. The nodal array has been deployed for roughly a whole year providing an extensive view of the current seismicity. The primary goals of the deployment were to detect microseismic events and to allow for the detailed mapping of the spatiotemporal evolution of seismicity on these fault structures. We expand the number of events in this catalog by using Earthquake Transformer (Mousavi et al., 2020) and alternatively PhaseNet (Zhu and Beroza, 2019), a machine-learning based event detector and phase picker. We improve the model used in the detection of events via Transfer-Learning to update the model using analyst picks made on the nodal deployment. For the detected phases we associate them with the REAL (Rapid Earthquake Association and Location) (Zhang et al., 2019). We initially use Hypoinverse to locate the events, and later relocate them with GrowClust (Trugman and Shearer, 2017). The refined locations illuminate many fault structures. We also use the machine-learning based method of Ross et al. (2018) to determine the first motion polarity of the P-wave arrivals, which we use to compute focal mechanisms. The dense nodal array’s density provides the azimuthal coverage needed to obtain accurate focal mechanisms even for microseismicity. Using the expanded catalog of earthquakes and focal mechanisms, we examine the spatiotemporal variability in focal mechanisms for different fault strands for the array. With this level of detail, we can determine, which fault strands are active currently.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
Session: Opportunities and Challenges for Machine Learning Applications in Seismology [Poster]
Type: Poster
Date: 4/19/2023
Presentation Time: 08:00 AM (local time)
Presenting Author: Colin N. Pennington
Student Presenter: No
Invited Presentation:
Authors
Colin Pennington Presenting Author Corresponding Author pennington6@llnl.gov Lawrence Livermore National Laboratory |
Qingkai Kong kong11@llnl.gov Lawrence Livermore National Laboratory |
William Walter walter5@llnl.gov Lawrence Livermore National Laboratory |
|
|
|
|
|
|
Seismicity Behavior Within Rock Valley Illuminated by a Dense Nodal Deployment and Machine-Learning Methods
Category
Opportunities and Challenges for Machine Learning Applications in Seismology