SEMINAR - Bridget Smith-Konter
Abstract
January 9, 2007 marked the passing of 150 years of relative seismic quiescence along California’s central and southern San Andreas Fault System (SAFS). For an active fault system with earthquake repeat times averaging roughly 150 years, the SAFS is likely primed for another major, and potentially devastating, seismic event. In an effort to identify inherent stress behaviors of individual fault segments of the SAFS, a 3-D time-dependent deformation model is used to simulate stress evolution throughout the earthquake cycle. The model spans the last 1000 years of earthquake history and combines an up-to-date set of paleoseismic, geologic, and geodetic constraints. Interseismic Coulomb stress rates generated by the model range from 0.2 to 7.2 MPa/100yrs and reflect variations in slip rate, fault orientation, and locking depth. Assuming a prescribed slip history based on paleoseismic evidence and the historical earthquake record, the model can also be used to estimate the magnitude of accumulated Coulomb stress on each fault segment spanning multiple earthquake cycles. These simulations reveal an evolving stress field through time, and in particular, a significant level of accumulated stress (~7.8 MPa) along the southernmost portion of the San Andreas (Coachella segment) at present day, where major earthquake activity has been absent for over 300 years. While models of this nature are highly dependent on an assumed time and slip history, they provide a quantitative foundation for advancing our ability to recognize zones of elevated seismic risk.
Bridget's seminar title: "Historical Deformation and Stress Evolution of the San Andreas Fault System"; Wednesday the 7th of March 2007
4 comments:
3D Semi-Analytic Elastic and Viscoelastic Fourier Deformation Models
Exploration of earthquake scenarios that span several thousand years, and deform over an equal number of kilometers, requires models that are three-dimensional, time-dependent, and computationally efficient. My Ph.D. thesis research was directed toward the development, verification, & application of a semi-analytical Fourier model describing the 3D response of both elastic and viscoelastic mediums to a distribution of body forces. Using Fourier analysis, the horizontal complexity of a given fault system has no effect on the speed of the computation; likewise, because the solution is analytic in time, no numerical time stepping is required. This approach allows for rapid computer model calculations that are over 20 times faster than previous methods (e.g., finite element methods). A single time-step for a mesh of 2048 by 2048 horizontal grid cells, containing over 400 fault patches, requires only 40 seconds of CPU time on a personal computer. Multiple time steps, including hundreds of years of earthquake history, can be computed in a matter of hours.
Model development involved extensive testing against analytic solutions including: 2-D analytic tests of a homogeneous elastic half-space [Weertman, 1964], a layered elastic half-space [Rybicki, 1971], non-surface observation planes [Savage and Lisowski, 1993], and a layered viscoelastic half-space [Nur and Mavko , 1977]; 2-D analytic Boussinesq tests for the point load solution [Love, 1944] and the thin-plate flexure solution [Le Pichon et al., 1973]; a 3-D elastic half space [Okada, 1985, 1992].
GPS & InSAR Applications
Space geodetic techniques, such as GPS and InSAR, provide valuable data that offer a detailed synoptic picture of the strain accumulation along Earth's plate boundaries. However, modeling of these data is critical in order to determine the corresponding tectonic stress and rheologic parameters. Accurate models must incorporate time-dependent interactions among complex 3-D fault systems. Using the 3D Fourier model described above, along with 1000+ GPS-derived horizontal velocity measurements, calculations of both secular and episodic deformation and stress due to plate boundary forces are feasible.
Likewise, InSAR data can also be efficiently investigated using 3D crustal deformation models. For example, ascending and descending interferograms derived from ERS satellites have been used to estimate surface slip and fault parameters along the Hector Mine earthquake rupture [Sandwell et al., 2002]. Large-scale synthetic interferograms can also be produced for the purpose of integrating GPS and InSAR data to provide both high spatial and high temporal resolution at the plate boundary
SRTM Topograph
I have also investigated the resolution quality of the Shuttle Radar Topography Mission data. The Shuttle Radar Topography Mission (SRTM) collected radar interferometry data over 80% of Earth's landmass from 60 ºN to 56 ºS latitude in February of 2000 from the Space Shuttle Endeavour. Both C-band and X-Band data were acquired simultaneously during the mission and subsequently processed by JPL and DLR, respectively. The completed SRTM Digital Elevation Model (DEM) provides a global topography data set critical for a number of scientific investigations, specifically in areas outside of the U.S. where the quality of data is typically poorer. Many of these applications require high horizontal resolution and vertical accuracy.
The primary objective of my SRTM research was to establish the resolution and accuracy of C-band SRTM-1 topography in order to provide a means of appropriate filter design and scientific application. This work involved cross-spectral analyses of the C-band 30-m SRTM DEM with both the National Elevation Dataset (NED) and Hector Mine Airborne Laser Swath Mapping (ASLM) dataset in order to identify horizontal resolution and any geo-location errors in the SRTM DEM. Spectral comparisons of the NED and SRTM data yield coherent results for wavelengths greater than 200 m. Two additional spectral comparisons made with the Hector Mine laser topography data suggest that the NED is of poorer quality for wavelengths longer than 350, and that SRTM topography is inferior for wavelengths shorter than 350 m. Additionally, a northeast phase shift of 11.87 m east and 10.58 m north was identified in the NED. No geo-location errors were identified in the SRTM DEM. From these results, low-pass filter/decimation algorithms can be designed in order to suppress the short wavelength noise and expedite large-area SRTM processing.
Bridget's - Education and Outreach
Lectures:
- Numerical Methods for Geophysical Partial Differential Equations
- SIO Birch Aquarium Public Lecture
- Earthquakes 101 (SIO Earthquake Education Workshop)
- San Andreas Overview (SIO 239)
- Shuttle Radar Topography Mission (STARS Workshop)
- Fourier Transform Fault Model
SIO Earthquake Education Workshop
COSMOS/SIO High School Student Visualization Scenes:
- Global Seismicity
- Southern California Seismicity Timeline
- Southern California Earthquake Magnitudes
- Parkfield Earthquakes (2004-2006)
- The 2004 Parkfield Earthquake
San Andreas Visualization Scenes:
- Model featured in April 2006 National Geographic: "The Next Big One"link) (pdf
- Model displayed in feature story on Apple Computer's website
href="http://topex.ucsd.edu/body_force/San_Andreas_Earthquake_Machine.mov">- San Andreas: Earthquake Machine (Quicktime Movie, 1.15 Gb)
Post a Comment