UC San Diego

Earthquakes produce damaging ground motions generally increasing with their size, or moment. Moment reflects the low-frequency energy that an earthquake radiates, but earthquakes of similar moment can produce very different amounts of high-frequency shaking that is destructive to infrastructure. Stress drop, the change in average shear stress across a fault due to a rupture, is a geophysical parameter often used to quantify this variation. For earthquakes without surface rupture or deformation, stress drop must be inferred from the earthquake’s radiated waves. This inference requires assumptions about rupture models, attenuation models, and signal-processing parameter choices, such that studies using similar methods can differ by an order of magnitude or more in their estimates of the same event. Relative stress drop between events is better constrained, and can reveal robust spatial variations in high-frequency content, but is still susceptible to bias due to model and parameter assumptions. The normalized frequency index (nFI) is an alternative measure of the relative amount of high-frequency radiation, computed directly from observed spectra and corrected empirically for path, station, and magnitude effects. The nFI requires fewer assumptions and is simpler to compute, especially for large datasets, yet still correlates with stress drop. Determining spatial patterns in high-frequency radiation content has direct implications for improving seismic hazard models. In Chapter 2, we compute stress drops from P- and S-wave spectra of the 2019 Ridgecrest earthquake sequence. In Chapter 3, we develop the normalized frequency index, which quantifies the variation in high-frequency radiation. We apply this method to the same dataset as in Chapter 2. In Chapter 4, we compute the normalized frequency index for 30 years of California seismicity, explore its spatial variation, and compare it to several geophysical parameters. Finally, Chapter 5 discusses conclusions and possible future work.