HR: 11:35h
AN: T21F-06 INVITED [PDF]
TI: Seismological Constraints on Mantle Structure and Composition
AU: * van der Hilst, R D
EM: hilst@mit.edu
AF: Massachusetts Institute of Technology, 77 Mass Av., Cambridge, MA 02139 United States
AU: Trampert, J
EM: jeannot@geo.uu.nl
AF: Utrecht University, Budapestlaan 4, Utrecht, MA 3508 TA
Netherlands
AU: Stutzmann, E
EM: stutz@ipgp.jussieu.fr
AF: Institute de Physique du Globe, 4 Place Jussieu, Paris, 05 75252
France
AB:
Results of travel time tomography suggest that some slabs of subducted lithosphere are severely deformed between 400-1000 km
depth - which is loosely referred to as the transition zone (Layer "C" according to Bullen's classification) - and that
beneath some convergent margins subducted slab material has sunk to larger depths in the mantle, some perhaps to near the
CMB. The transition zone may be marked by enhanced levels of anisotropy. These inferences all suggest that mantle convection
is more complex than suggested by the canonical end-member models of either convective stratification at 660 km or unhindered
whole mantle overturn. Over the past 5 years the emphasis has shifted from mapping wavespeed patterns to constraining
spatial variations in temperature and composition, and there is now mounting seismological evidence for the existence of
compositional heterogeneity in the deep mantle. P and S wave travel times suggest that the ratio of relative variations in
shear- and compressional wavespeed (R=dlnVs/dlnVp) increases more rapidly with depth than can be explained by temperature
effects alone, in particular away from major convergent margins (i.e., in the regions with the lowest wavespeeds).
Furthermore, combined with constraints from mineralogy, normal mode studies based on full model space searches indicate that
in the deep mantle the most likely ratios between variations in elastic parameters and density require variable composition.
Both lines of study suggest that variations in perovskite and iron are needed to explain the seismological data. These
inferences are qualitatively consistent with main aspects of the thermo-chemical mantle model proposed by Kellogg et al.
(Science, 1999), but lack of observations of scattering off the implied interface (e.g., Castle \& Van der Hilst, JGR, 2003)
renders existence of a distinct layer in the deep mantle unlikely. Diffuse boundaries due to hitherto unknown phase
transitions or gradual, pressure induced changes in composition cannot yet be ruled out, however, and in thermo-chemical
convection vertical mixing gradients may exist over long periods of geological time, with rapid recycling in the shallow
mantle and a lowermost mantle that is much less often involved in convective overturn. We will speculate on the possible
origin of the compositional heterogeneity in the mantle beneath 1000 km depth.
DE: 1025 Composition of the mantle
DE: 7200 SEISMOLOGY
DE: 8120 Dynamics of lithosphere and mantle--general
SC: Tectonophysics [T]
MN: 2003 Fall Meeting