Resolving Earthquake Source Complexities in the Heterogeneous Earth

While the commonly used simple assumptions of sources and structures allows useful first-order approximation of earthquakes, they are increasingly insufficient in characterizing the complex earthquake ruptures and the seismic wave propagations. In this thesis, I present studies that address both the source and structural complexities, as well as their interactions, using flexible parameterizations and ideas. For large earthquakes, I develop a subevent inversion framework to determine their spatiotemporal rupture complexities, and applied it to multiple significant earthquakes. Our method does not assume a fault geometry and kinematic history, and incorporates Bayesian analysis for uncertainty assessments. In Chapter 2, I discovered that the 2018 Fiji Magnitude 8 deep earthquake doublet actually ruptured two slabs, which demonstrates local slab temperature as the critical factor for deep earthquakes, and reveals complex interactions between slabs. In Chapter 3, I determined that the 2019 Ridgecrest Magnitude 7 sequence coseismically ruptured orthogonal faults, and has superficially complex but in-depth simple fault geometries, which illustrates the fault geometrical control of the rupture behaviors which challenges traditional seismic hazard mapping. In Chapter 4, I found the 2021 South Sandwich Island Magnitude 8 earthquake consists of deep regular ruptures connected by a slow tsunamigenic event, which highlights the tsunami potential for deep initiations of megathrust earthquakes. For smaller earthquakes, I develop a Bayesian differential moment tensor inversion (diffMT) algorithm to remove the common earth structural effects, thus improving the earthquake focal mechanism resolvability. In Chapter 5, I demonstrated that diffMT reduces the moment tensor uncertainties substantially than traditional direct-inversion methods, and found that the North Korea nuclear tests in 2013-2016 are more dominated by explosive mechanism than previous understandings. Towards high resolution mapping of the Earth structure, I expand the seismic tomography for high resolution basin structure by combining the wide aperture of seismic stations and high density of industrial arrays. In Chapter 6, I applied this approach on the densely populated Los Angeles Basin, and found improved mapping of small scale heterogeneities, which can potentially promote earthquake ground motion assessments. In summary, I have developed research tools and applied novel ideas on complex seismic events and heterogeneous earth environments. The results illustrate the diverse controlling factors of complex earthquake ruptures, and reveal the complex interactions between earthquakes and earth structure.