This thesis focuses on two research topics: investigating dynamically generated seismic structures at the topmost outer core and studying dynamically triggered seismic events in the Earth’s crust. Earth’s liquid-iron outer core convects vigorously and contains buoyant light elements that accumulate at the top of the outer core. Modeling of seismic SmKS phases shows that these light elements can explain puzzling observations of delays between SmKS phases in a manner that is consistent with the light elements being entrained in sub-horizontal convection currents, which help create the Earth’s magnetic field and help the Earth to lose its heat. The Earth’s crust is solid and subject to brittle failure when stressed. An investigation of intra- and inter-plate seismicity that occurs in the USA during the passage of strongly deforming seismic surface waves confirms that these deformations cause transient stresses and trigger tremor. However, with the help of a decision-tree algorithm, it is found that the peak size of these stresses does not correlate with the occurrence of local earthquakes. To facilitate the gathering of more such data, a simple algorithm as well as a deep-learning algorithm were developed and implemented to automatically detect potentially triggered earthquakes. The deep-learning algorithm more reliably detected small earthquakes than the simple SNR algorithm. Moreover, a citizen-science project was developed and launched, called Earthquake Detective, through which thousands of volunteer scientists helped detect potentially triggered earthquakes and tremor, primarily in Alaska. The volunteers generally agree more with each other when identifying the presence of earthquake signals and the absence of signals than when identifying tremor or other signals. Volunteer scientists and the machine-learning algorithm perform comparably when detecting earthquake signals. In addition, volunteer scientists are naturally able to detect tremor signals with minimal training. Overall, the dynamics of the Earth’s shallow, solid, and brittle, as well as its deep, liquid, and convecting boundary layers have been studied and are somewhat better understood via the analyses of seismic waveforms described in this thesis.

Creator

• Tang, Vivian

DOI

• https://doi.org/10.21985/n2-pg4w-bk38

Subject

• Geophysics

Language

• en

Alternate Identifier

• http://dissertations.umi.com/northwestern:15357
• etdadmin_upload_774452

Keyword

• Outer Core
• Machine-learning
• Dynamically triggered seismic events

Date created

• 2020-01-01

Resource type

• Dissertation

Rights statement

• In Copyright