HR: 0800h
AN: GP21A-0149 [Abstracts]
TI: The Whole AMS Matrix: Using the Owens Lake, Ardath Slump, and Gaviota Slide cores to explore
classification of ellipsoid shapes
AU: * Schwehr, K
EM: kdschwehr@ucsd.edu
AF: Scripps Institution of Oceanography, Mail Code 0208
UCSD, La Jolla, CA 92093-0208
United States
AU: Driscoll, N
EM: ndriscoll@ucsd.edu
AF: Scripps Institution of Oceanography, Mail Code 0208
UCSD, La Jolla, CA 92093-0208
United States
AU: Tauxe, L
EM: ltauxe@ucsd.edu
AF: Scripps Institution of Oceanography, Mail Code 0208
UCSD, La Jolla, CA 92093-0208
United States
AB:
Categorizing sediment history using Anisotropy of Magnetic Susceptibility (AMS) has been a long standing challenge for the
paleomagnetic community. The goal is to have a robust test of shape fabrics that allows workers to classify sediments in
terms of being primary depositional fabric, deposition in with currents, or altered fabrics. Additionally, it is important
to be able to distinguish altered fabrics into such classes as slumps, crypto-slumps, drilling deformation (such as
fluidization from drilling mud and flow-in), and so forth.
To try to bring a unified test scheme to AMS interpretation, we are using three example test cases. First is the Owens Lake
OL92 core, which has provided previous workers with a long core example in a lacustrian environment. OL92 was classified
into five zones based on visual observations of the core photographs. Using these groupings, Rosenbaum et al. (2000) was
able to use the deflection of the minimum eigen vector from vertical to classify each individual AMS sample. Second is the
Ardath Shale location, which provides a clear case of a lithified outcrop scale problem that showed success with the
bootstrap eigen value test. Finally is the Gaviota Slide in the Santa Barbara Basin, which provides usage of 1-2 meter
gravity cores.
Previous work has focused on Flinn, Jelinek, and bootstrap plots of eigen values. In supporting the shape characterization we
have also used a 95% confidence F-Test by means of Hext's statistical work. We have extended the F-Test into a promising
new plot of the F12 and F23 confidence values, which shows good clustering in early tests. We have applied all of the
available techniques to the above three test cases and will present how each technique either succeeds or fails. Since each
method has its own strengths and weaknesses, it is clear that the community needs to carefully evaluate which technique
should be applied to any particular problem.
UR: http://schwehr.org/agufall2004
DE: 4219 Continental shelf processes
DE: 4558 Sediment transport
DE: 3022 Marine sediments--processes and transport
DE: 1518 Magnetic fabrics and anisotropy
DE: 1594 Instruments and techniques
SC: Geomagnetism and Paleomagnetism [GP]
MN: 2004 AGU Fall Meeting