Insights on Earthquakes and Thermochemical Heterogeneity in Earth's Deep Interior: Generation and Propagation of Seismic Waves

Abstract

Seismic waves deliver valuable messages about earthquake faulting processes and Earth's structures. This thesis includes seismological studies that address two frontier research questions: how do earthquake faults rupture and how can seismic waves be used to map the subsurface velocity structures. Comprising three chapters dedicated to the investigation of earthquake source parameter estimation and an additional three chapters focused on the depth inversion of layered structures, this work advances our understanding of the source and structures spanning from Earth’s deep interior to the surface.

In Chapter 2, we focused on estimating corner frequencies and stress drops for deep-focus earthquakes. By applying spectral ratio analysis based on empirical Green's function, the median stress drop estimates for deep-focus earthquakes are estimated to be one order of magnitude higher than those for shallow earthquakes. This difference suggests that the shear stress of faults in the mantle transition is on average higher the crust by an order of magnitude, indicative of the coexistence of multiple physical mechanisms in Earth's deep ruptures in the mantle.

In Chapter 3, we explore the potential biases in corner frequency estimates when utilizing the Brune source model for earthquakes with multiple subevents. We adopted a source time function decomposition approach that treats complex seismic sources as a composite of multiple Brune sources. We found that earthquake corner frequency correlates best with the corner frequency of the subevent with the highest moment release. This observation implies that when employing the Brune model, the estimated corner frequency, and consequently the stress drop of a complex earthquake, is predominantly determined by the most substantial subevent rather than the overall rupture area.

In Chapter 4, we introduced a point-wise stacking method for the precise estimation of stress drop. This approach optimizes the utilization of source spectra, resulting in a more robust estimate of corner frequencies, holding the potential to significantly mitigate the substantial variation often observed in stress drop estimates. In synthetic tests, the point-wise method yields stable less sensitive estimates to the quality of source spectra. This is promising for deep-focus earthquakes with limited high quality of source spectra.

In Chapter 5, we investigated the influence of velocity heterogeneity on imaging of the mantle transition zone using long-period SH-wave reverberations and assessed the method efficacy using synthetic waveforms simulated based on the spectral element method. The depth difference of 410-km discontinuity beneath the western US than the central-eastern US disappeared after we corrected travel times using a 3-D shear wave velocity model, highlighting the importance of accounting for 3-D velocity variations in subsurface imaging.

In Chapter 6, we investigated the global and regional cross-correlation of earthquake coda waves. Using Global Seismic Network, we established global correlograms that align with the correlograms derived from synthetic waveforms. Using the Southern California Seismic Network, we identified plausible reflection signals from the mantle transition zone. To enhance our interpretation of these reflection signals, we plan to conduct synthetic tests involving high-frequency coda waves at higher frequencies up to 0.1 Hz.

Finally, in Chapter 7, we conducted a preliminary exploration of Distributed Acoustic Sensing (DAS) data in Cordova, Alaska. We detected tidal signals strongly correlated with tide heights. We further applied surface wave inversion through cross-correlation to image the subsurface velocity structure. Our future work will focus on improving long-period signal detection and refining subsurface imaging techniques.