Download Volume II Accomplisments (28 Mb pdf). - IRIS

Imaging Attenuation in the Upper Mantle with the GSN
Colleen A. Dalton (Boston University), Goran Ekstrom (LDEO, Columbia University), Adam M. Dziewonski (Harvard
University)
Seismic-wave attenuation (1/Q) is thought to be highly sensitive to variations in temperature, and joint interpretation of
attenuation and velocity models should aid in distinguishing between thermal and chemical heterogeneity in the mantle.
However, imaging attenuation remains a challenging problem, because factors other than attenuation influence wave amplitude.
Principally, amplitudes are affected by focusing due to lateral velocity variations, but uncertainties in the calculation of source
excitation as well as inaccuracies associated with the instrument response can also obscure the attenuation signal in the data. We
have developed a method to remove these extraneous effects and isolate the signal due to attenuation. We invert a large data set
of fundamental-mode Rayleigh wave amplitudes, measured from waveforms recorded by the IRIS GSN and other global networks,
in the period range 50--250 seconds simultaneously for a 3-D model of upper-mantle shear attenuation, maps of phase
velocity, and amplitude correction factors for each source and receiver in the data set. The new three-dimensional attenuation
model (QRFSI12) contains large lateral variations in upper-mantle attenuation, 60% to 100%, and exhibits strong agreement
with surface tectonic features at depths shallower than 200 km. At greater depth,QRFSI12 is dominated by high attenuation in
the southeastern Pacific and eastern Africa and low attenuation along many subduction zones in the western Pacific. QRFSI12
is found to be strongly anti-correlated with global velocity models throughout the upper mantle, and individual tectonic regions
are each characterized by a distinct range of attenuation and velocity values in the shallow upper mantle. By comparing with
laboratory experiments, we have found that lateral temperature variations can explain much of the observed variability in velocity
and attenuation, although oceanic and cratonic regions appear to be characterized by different major-element composition
or volatile content for depths < 225 km.
References
Dalton, C.A., G. Ekstrom, and A.M. Dziewonski. The global attenuation structure of the upper mantle, J. Geophys. Res., 113,
doi:10.1029/2007JB005429, 2008.
Kustowski, B., G. Ekstrom, and A.M. Dziewonski. Anisotropic shear-wave velocity structure of the Earth's mantle: A global model, J. Geophys.
Res., 113, doi:10.1029/2007JB005169, 2008.
Faul, U.H., and I. Jackson. The seismological signature of temperature and grain size variations in the upper mantle, Earth Planet. Sci. Lett.,
234, 119-134, 2005.
Comparison of global shear-attenuation (a) and shear-velocity (b) models at 100-km and 400-km depth. The attenuation model is from Dalton et al. (2008) and the
velocity model is from Kustowski et al. (2008). (c) Grey points show seismologically observed shear velocity (S362ANI) and attenuation (QRFSI12) models sampled at
5762 points at 150-km depth. Colored contours enclose 75% of points from oceanic regions of age < 70 Myr (red) and > 70 Myr (blue) and from old-continental regions
(green). Black lines show the predicted relationship between shear velocity and attenuation using the experiment-derived model of Faul and Jackson (2005).
2010 IRIS Core Programs Proposal | Volume II | Upper Mantle Structure and Dynamics | II-229