Recently,
we
have turned our attention to the interpretation of seismic amplitudes.
Amplitudes have long been neglected in seismic tomography, partly
because
measurements are not very stable, partly because the amplitude of a P
or
S wave is influenced by many factors: local impedance, attenuation,
source
radiation, scattering, focusing. Rather than taking classic
"peak-to-peak"
measurements of the amplitude, we have stabilized the observations by
computing
the RMS amplitude of the displacement signal of the P wave.
Ileana
Tibuleac (Tibuleac et al., 2003) has
shown that focusing and defocusing causes an rms fluctuation in the
amplitudes of teleseismic P waves of at least 39% (allowing
for
a maximized influence of all other factors).
Preliminary
results indicate that global models severely underestimate this
variance
in the observations, even when corrected for effects of attenuation.
Ying Zhou (Zhou et al., 2003) has shown that crustal
impedance effects can be handled by a station correction
if the signal is low passed at about 0.2 Hz. We are currently extending
the measurements to shallow earthquakes and more complicated sources.
This
is the PhD research project of Karin Sigloch, who designed a matched
filter that includes not only the source time-function, but
incorporates pP and sP reflections into the synthetic waveform
(see preprint ). Yue Tian is extending
Karin's method to (rotated) horizontal components for delay- and
amplitude measurements of S waves.
This
project has - over time - involved a large number of undergraduates who
helped in data processing and quality control, and through that gained
significant experience in data processing/mining.
Postdocs:
Ileana
Tibuleac
(now at University of Nevada, Reno)
Graduate students:
Karin Sigloch
Ying Zhou (now
at Virginia Tech)
Yue Tian
Undergraduates:
Pat Shamberger
(Senior thesis project)
Melissa
Ginsberg
(Freshman Scholar's Institute)
Obinna Eneanya
(Freshman Scholar's Institute)
Amanda Howard
(Freshman Scholar's Institute)
Marilyn Waite
(Freshman Scholar's Institute)
Jeffrey Tang
Funding:
NSF, Intel
(for FSI)