HR: 10:35h
AN: V32B-02 [Abstracts]
TI: Lower mantle structure and composition: insights from generalized inversions of radial seismic
profiles
AU: * Matas, J
EM: jan.matas@ens-lyon.fr
AF: Laboratoire de Sciences de la Terre, UMR CNRS 5570, Ecole normale superieure de Lyon, 46, allee
d'Italie, Lyon, 69007
France
AU: Bass, J D
EM: jaybass@uiuc.edu
AF: Department of Geology, University of Illinois, Urbana-Champaign, 1301 W Green St, Urbana, IL 61801
United States
AU: Ricard, Y
EM: yanick.ricard@ens-lyon.fr
AF: Laboratoire de Sciences de la Terre, UMR CNRS 5570, Ecole normale superieure de Lyon, 46, allee
d'Italie, Lyon, 69007
France
AU: Mattern, E
EM: estelle.mattern@ens-lyon.fr
AF: Laboratoire de Sciences de la Terre, UMR CNRS 5570, Ecole normale superieure de Lyon, 46, allee
d'Italie, Lyon, 69007
France
AU: Mattern, E
EM: estelle.mattern@ens-lyon.fr
AF: Department of Geology, University of Illinois, Urbana-Champaign, 1301 W Green St, Urbana, IL 61801
United States
AB:
A large number of studies have already been dedicated to the interpretation
of 1-D radial seismological profiles of seismic wave velocities, density,
and elastic properties in terms of mantle mineralogy and thermal structure.
No clear consensus has, however, emerged from these studies. We show results
for lower mantle structure and bulk composition obtained by using a generalized
inverse technique together with the most recent and accurate values of physical
properties (including, of particular importance, shear moduli) of relevant
lower mantle minerals. We have designed an iterative numerical code that
allows us to precisely evaluate the a posteriori uncertainties, correlations
and resolution of best matching chemical composition and geotherm. We discuss
the effect of various sources of uncertainties and study the effect of
experimental uncertainties in the physical properties, of various averaging
schemes for the composite mantle assemblages, and of different formulations
of equation of state. We compare the results obtained from inversions of density
and bulk sound velocity to those obtained from density and the seismic
velocities Vp and Vs. We show that including the shear properties significantly
reduces the a posteriori uncertainties on the resulting geotherm models, and
we evaluate the correlations between the thermal structure and averaged
depth-dependent bulk composition. We also test the compatibility of radial
seismic profiles with the hypothesis of a chemical discontinuity in the
deep lower mantle. We show that although the a posteriori uncertainties do
not allow us to identify a unique compositional-thermal model, several robust
conclusions can be drawn. (1) Average lower mantle compositional models that
are not pyrolitic-like in composition are consistent with seismological
observations. (2) There are large trade-off's between thermal structure and
bulk composition. (3) Regardless of the origin of the 660-km interface, other
(chemical) discontinuities may be located at greater depth in the lower mantle.
The upper-part of the lower mantle presents a lithospheric signature
while the underlying layer is consistent with silica enrichment. (4) The role
of minor elements (such as aliminium, calcium, water) is difficult to
constrain but not remains significant.
DE: 1025 Composition of the mantle
DE: 3260 Inverse theory
DE: 3919 Equations of state
DE: 3924 High-pressure behavior
DE: 7208 Mantle (1212, 1213, 8124)
SC: Volcanology, Geochemistry, Petrology [V]
MN: Fall Meeting 2005