1. Using Short-Period Surface Waves to Study Seismic Source and Structure. 2. Source Complexity of Large Strike-Slip Earthquakes

Citation

Thio, Hong Kie (1996) 1. Using Short-Period Surface Waves to Study Seismic Source and Structure. 2. Source Complexity of Large Strike-Slip Earthquakes. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/s8s1-f531. https://resolver.caltech.edu/CaltechTHESIS:10102023-192037999

Abstract

The availability of high dynamic range very broad band seismic data in recent years has greatly increased the level of detail and the speed with which we can study the seismic source. The work presented in this thesis draws heavily on the deployment of broad-band seismometers, both on a worldwide scale, with networks like IRIS, IRIS/IDA and GEOSCOPE, as well as on a local scale, using data from the TERRAscope network. The routine study of seismicity in Southern California, like in other seismically active regions, has traditionally been carried out using dense arrays of high-gain short-period seismometers. With the addition of the very broad band instrumentation of TERRAscope we can improve this pursuit in several ways, one of which being the use of short period surface waves to study local earthquakes as described in chapter 1 of this thesis. Over the years, surface waves have proved to be very reliable and stable for moment tensor inversions. The method is very rapid, and because of the longer periods used they are more reliable for consistent estimation of earthquake moment. At short distances the surface waves arrive within a few minutes after an event has occurred at the stations, and with real-time telemetry we can obtain the size and mechanism for local earthquakes within minutes. The propagation corrections for surface waves are very straightforward so that this procedure can be made completely automatic. Armed with the results from above procedure, we can determine travel time residuals for a dense distribution of raypaths across Southern California. In chapter 2 we present tomographic inversions of these resid­uals, for Love and Rayleigh waves at periods between 10 and 50 seconds. The results indicate that lateral variations of phase velocity of up to 10% exist in the area, and that these anomalies can have relatively short wavelengths.

The 1994 Northridge earthquake provided a wealth of data to apply our moment tensor inversion to, and in chapter 3 we present a detailed analysis of the aftershock mechanisms in relation to the source complexity of the mainshock. We show that the orientation of the aftershock mechanisms changes away from the zone where rupture took place. We suggest that this change in mechanism reflects changes in fault geometry which have limited the extent of the Northridge rupture, leading to a high static stress drop.

The issue of source complexity is discussed further in chapter 4, where we present a systematic study of the rupture of three large strike-slip earthquakes and compare these results with observation on the surface rupture. We find a very good correlation which suggests that the source complexity can be attributed to fault geometry, which tends to become simpler as slip accumulates along a fault. This provides an explanation for the high stress drops that are observed for earthquakes which occur along faults with low strain rates.

Finally, in chapter 5 we compiled energy and moment estimates for earthquakes in Southern California, based on the results in the previous chapters. We find that the radiated seismic energy is not linearly related to the seismic moment, but that instead the energy-moment ratio increases as a function of moment. We provide some suggestions as to the cause of this relationship, including a moment dependence of the specific fracture energy and a non-similarity of the frictional stress between different size earthquakes.

Thesis Files