Earthquakes occur in a variety of geologic settings throughout the Earth: from volcanic summits all the way down to depths of 660 km at the bottom of the mantle transition zone. Though they pose significant hazard in some regions, earthquakes are also scientific opportunities to gain insight about different processes within the Earth system. Earthquakes all share the common behavior of suddenly releasing energy that had been stored within rock, but the driving forces and physical conditions of individual systems where they occur varies greatly. Earthquake observations that characterize rupture source properties reveal both the similarities and differences between shallow and deep earthquakes. These are key to differentiating the types of processes that could be leading to earthquake failure in different systems under varying physical conditions. Our general understanding of earthquake proprieties are used in the studies presented here to help guide investigations of both volcanic systems and subducting slabs. Here we present two types of earthquake analyses: statistical modeling of shallow volcanic earthquakes and geodynamic modeling for deep earthquake investigation.

In a volcanic setting, the driving forces behind the stress changes in the system leading to seismicity can be associated with magma migration and pressure changes within the plumbing of a volcano. Improving our understanding of this relationship could help us to utilize future earthquake occurrences in interpreting changes in a volcano's behavior. Here we use earthquake statistics and statistical hazard modeling to investigate earthquakes occurring during the active 2018 Kīlauea Volcano eruption and find changes in earthquake behavior that temporally align with physical changes in the behavior of the larger volcanic system. Much deeper below Earth's surface, in subducting lithosphere, the driving forces leading to deep (>70 km) earthquakes are associated with the long-term deformation of the subducting slab. Although they do not pose the seismic hazard that the Kīlauea earthquakes do, improving our understanding of how these unusual earthquakes can occur at high temperatures and pressures found in the deep slab can help us better understand the rheology of Earth's lithosphere as well as what processes may be going on within it (e.g., phase transitions, thermal shear heating). For this portion of the thesis we present novel numerical subduction modeling methods developed specifically to be paired with deep earthquake observations to investigate the role long-term strain-rates play in where deep earthquakes occur.

Understanding both the commonalities and differences between the types of earthquakes observed at volcanic summits and within subducting lithosphere is helpful for understanding more general earthquake behavior. Earthquakes are both a hazard that needs improved understanding, as well as a vital tool themselves to better understanding different lithospheric processes of varying temporal and spatial scales across Earth’s system.