HR: 1330h
AN: T32B-0927    [PDF]
TI: Heterogeneous lowermost mantle: compositional constraints and seismological observables
AU: * Samuel, H
EM: samuel@ipgp.jussieu.fr
AF: Institut de Physique du Globe de Paris, Laboratoire de Dynamique des Systemes Geologiques, 4 place jussieu, BP 89, Paris cedex 05, 75252 France
AU: Farnetani, C G
EM: cinzia @ipgp.jussieu.fr
AF: Institut de Physique du Globe de Paris, Laboratoire de Dynamique des Systemes Geologiques, 4 place jussieu, BP 89, Paris cedex 05, 75252 France
AU: Andrault, D
EM: andrault@ipgp.jussieu.fr
AF: Institut de Physique du Globe de Paris, Laboratoire des Geomateriaux, 4 place jussieu, Paris cedex 05, 75252 France
AB: Several seismological indicators strongly suggest the existence of compositional heterogeneity in the lowermost mantle. Indeed, the observed anticorrelation between bulk sound and shear wave velocity anomalies as well as the high values (i.e., $>2.7$) of the ratio $R=dlnV_P/dlnV_S$ are difficult to be explained by purely thermal convection. We use a compressible thermo-chemical convection model to study the stability of a chemically denser material, located in the lowermost mantle.Considering two phases for the lower mantle: (Fe,Mg)SiO3 perovskite and (Fe,Mg)O magnesiow\"ustite, we use geodynamical and seismological considerations and mineral physics data to constrain the composition of chemically denser material in the lower mantle. We show that the denser material has to be enriched in both iron and silica with respect to a pyrolitic lower mantle. The required enrichment is reduced if we consider that at high pressure Al-perovskite decreases the iron-magnesium partition coefficient between magnesiow\"ustite and perovskite. Our estimated composition of the dense material in then applied to the distribution of chemical heterogeneities calculated by our thermo-chemical convection model. In the deep mantle we predict broad seismic velocity anomalies and strong lateral velocity variations. Moreover, we find that areas of anticorrelation are associated with upwelling mantle flow. The calculated $R$ ratio varies laterally and may locally have values greater than 2.7, often associated with areas of anticorrelation. Our results compare well with seismic observations and provide a way to reconcile apparent discrepancies between global tomographic models. This suggests that only an enrichment in iron and silica in the lowermost mantle is required to explain seismological observations.
DE: 8121 Dynamics, convection currents and mantle plumes
DE: 8124 Earth's interior--composition and state (old 8105)
DE: 8125 Evolution of the Earth
DE: 8162 Rheology--mantle
DE: 8180 Tomography
SC: Tectonophysics [T]
MN: 2003 Fall Meeting