Using microearthquakes as probes of larger earthquake rupture

Although they garner very little attention from the public, microearthquakes are an extraordinarily valuable tool that seismologists use to understand better the physics underpinning faulting and earthquake rupture. Microearthquakes may not influence fault behavior on a large scale, but their ubiquitous nature has allowed me to determine precisely the time-dependent behavior of earth materials and to compute detailed descriptions of wave propagation, tasks which have significantly improved our understanding of fault slip and earthquake strong ground motion.; The majority of this dissertation is dedicated to understanding nonlinear strong ground motion. To this end, I identify reductions in the near-surface seismic velocity coincident with four moderate and large earthquakes. Applying moving-window cross correlation on multiple repeating earthquake sequences allows me to identify these time dependent changes in seismic velocity. There are multiple lines of evidence that suggest that velocity reductions are evidence of nonlinear strong ground motion induced damage. First, velocity always decreases following earthquakes, which suggests a damage mechanism. The size of the velocity reductions that I observe are correlated with the strength of shaking for the earthquake that caused them, indicating a cause and effect relationship between strong ground motion and velocity reductions. For multiple earthquakes, I also identify a correlation between the size of velocity reductions and site conditions. This is expected; soft rocks are easier to damage with strong ground motion than hard rocks. The healing behavior of these velocity reductions also parallels the behavior of velocity reductions observed in laboratory studies of the recovery of materials from transient nonlinear strain.; In the final chapter of this dissertation, I develop a new earthquake location technique. This method takes advantage of the expected similarity of the waveforms of nearby earthquakes to determine wave propagation parameters for many windows of time in a seismogram. This allows me to locate earthquakes that were sparsely recorded. I apply this technique to three medium magnitude earthquakes on the Calaveras Fault near streaks of seismicity. The new locations of these events suggest that streaks represent seismicity induced by the interaction between zones of a fault that accommodate slip differently, i.e., aseismically and seismically.