Precise earthquake source parameter estimation
Abstract/Contents
- Abstract
- The question of whether earthquakes, as they increase in size, radiate seismic waves more efficiently is at the core of our understanding of the physics of faulting, as well as our ability to mitigate the effects of strong ground motion. If earthquakes have some intrinsic time or length scale that could be observed or modeled, then seismologists could determine the ultimate size of an earthquake just as it begins to rupture. On the other hand, if earthquakes are self-similar, with no intrinsic time or length scale, than any information learned about the plethora of small and intermediate earthquakes can simply be scaled up to predict parameters, such as ground motion, for larger, more devastating earthquakes. Many studies find that apparent stress and stress drop increase with seismic moment, yet others find an independence of these parameters with moment, obeying self-similar earthquake source physics. Source measurements are controversial due to the inherent difficulty in correcting the radiated waves to negate path and site effects, such as attenuation, scattering or amplification. Independent studies of the same earthquake may find seismic energies that differ by an order of magnitude. Methods to estimate source parameters need to account for these effects, or quantify the range of validity for estimates made with uncorrected seismic records. In this work, I precisely estimate the source parameters radiated seismic energy, apparent stress and stress drop, using both relative spectral measures from empirical Green's functions, and close distance acceleration records. Using relative empirical Green's functions, I can handily negate source and path effects, without explicit consideration of anelastic attenuation. Working with data from 8 sequences of earthquakes in the western US and Honshu, Japan, ranging from M 1.8 to Mw 7.1, I find no clear trend of a dependence of apparent stress or stress drop with moment, finding a constant scaled energy, ER/Mo of 3.5x10-5, or apparent stress of ~ 1 MPa, to fit the data well. The average Brune stress drop for these data is ~5 MPa. By using many stations and relative measures, I statistically show self-similar earthquake scaling. However, there are anomalous enervated and energetic events that show individual departure from the overall trend, representing the true variability in earthquake source parameters. I revisit the aRMS stress drop using recent broadband stations and strong motion accelerometers. The aRMS stress drop samples an inherently different portion of the earthquake spectrum than the Brune stress drop, and can be directly related to PGA and hence high-frequency ground motion. While the aRMS stress drop is much simpler and faster to measure, it does not model attenuation, and hence suffers from loss of signal at distances > ~20 km. At close stations, and for large earthquakes, the aRMS stress drop values are very similar to those of the Brune stress drop, yet with reduced error base on corner frequency uncertainty. That the aRMS[not] method yields stable stress drops supports the assumptions behind the formulation: that earthquake acceleration records can be considered random, band-limited, white Gaussian noise, and overall, a self-similar earthquake model. The last portion of this work focuses on five great earthquakes, Mw > 8.5, over the past 7 years. Because they are so rare, seismologists don't have much information about these devastating events. Understanding how they relate to smaller earthquakes will aid in hazard mitigation. I estimate the radiated seismic energy and apparent stress, using a novel, teleseismic empirical Green's function deconvolution. At near distances, great earthquake are too big to model, as effects from one end of the rupture will interfere with those from other parts, and local recordings are often saturated. But at far distances, ~3000 km -- 9000 km, I show that moderate earthquakes, Mw 6.5 -- 7.5 can be used as Green's functions, and are used to correct the mainshocks from path and site effects. Use of several different eGf earthquakes demonstrates the limitations on the method, but also increases the precision of the energy estimates. I find that both P and S waves give consistent energy estimates when using eGf events. Azimuthal dependence of radiated energy indicates expected rupture directivity, and can be modeled using Haskell line sources to understand the rupture process.
Description
| Type of resource | text |
|---|---|
| Form | electronic; electronic resource; remote |
| Extent | 1 online resource. |
| Publication date | 2011 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Associated with | Baltay, Annemarie Susan |
|---|---|
| Associated with | Stanford University, Department of Geophysics |
| Primary advisor | Beroza, Gregory C. (Gregory Christian) |
| Primary advisor | Zoback, Mark D |
| Thesis advisor | Beroza, Gregory C. (Gregory Christian) |
| Thesis advisor | Zoback, Mark D |
| Thesis advisor | Dunham, Eric |
| Thesis advisor | Hanks, Thomas C |
| Advisor | Dunham, Eric |
| Advisor | Hanks, Thomas C |
Subjects
| Genre | Theses |
|---|
Bibliographic information
| Statement of responsibility | Annemarie Susan Baltay. |
|---|---|
| Note | Submitted to the Department of Geophysics. |
| Thesis | Thesis (Ph. D.)--Stanford University, 2011. |
| Location | electronic resource |
Access conditions
- Copyright
- © 2011 by Annemarie Susan Baltay
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...