Uniformity of Mantle Composition

Watt, J. Peter; Shankland, T. J.; Mao, Nai-Hsien

doi:10.1130/0091-7613(1975)3<91:uomc>2.0.co;2

Cited by 73 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Mao (1974) proposed an alternative to Birch's (1961) velocity -density correlation which incorporates Poisson's ratio. The correlation is at least as good as Birch's, and it has the interesting property of not requiring an increase in mean atomic weight in the transition zone (Ma0 1974;Watt, Shankland & Mao 1975).…”

Section: H E M I C a L Composftion D I F F E R E N C E Smentioning

confidence: 69%

Whole-mantle convection and plate tectonics

Davies¹

1977

Geophys J Int

133

View full text Add to dashboard Cite

It seems unlikely that the lower mantle is not involved in motions related to plate tectonics. The evidence relating to several alleged obstacles to lower-mantle convection is reviewed. The evidence for a chemical composition difference between the upper and lower mantle is not compelling. There now seems to be good evidence that the viscosity of the mantle is fairly uniform, contrary to a widely-held view, except possibly in the oceanic upper mantle. Phase changes may locally enhance or retard thermal convection, but are unlikely to prevent it. It is not necessary to assume that descending lithospheric slabs cannot penetrate below 700 km depth in order to explain the distribution and source mechanisms of deep earthquakes. Other recent seismic evidence suggests deep-mantle heterogeneities which may be related to large-scale flow.The behaviour of some simple layered-fluid models is analysed to show the effect of viscosity contrasts. The results indicate that the lower mantle would have to be 10000 times more viscous than the upper mantle in order to confine thermal convection to the upper mantle. The lower mantle would have to be at least 1000 times more viscous than the upper mantle in order to exclude viscous flow entrained by the moving plates. Alternatively, a shallow low-viscosity hyer would have to be about 10 000 times less viscous than the underlying mantle for the 'return' flow to be confined to it. These and other results indicate that only extreme viscosity contrasts (arising from non-linear or temperature-dependent rheology, for instance, as well as from different physical states) significantly affect large-scale flow patterns, and that flow patterns are strongly affected by moving boundaries.A version of whole-mantle thermal convection is proposed which seems to be capable of explaining the main features of plate tectonics: the buoyancy forces are concentrated in, but not confined to, the descending lithospheric slabs. Descending slabs would then cause their attached plates to move rapidly, as in the Elsasser model. The flow entrained by the fast plates would dominate the flow pattern in the mantle: convection cells under fast plates would be amplified; those under other plates would have smaller amplitudes and would fit between the fast cells. This model can qualitatively account for 460 G. F. Davies the large horizontal scale of the plates, the imperfect correlation between plate speed and the amount of attached descending slab, the motion of plates with 110 attached slab, and the opening of the Atlantic in particular. The model may provide a mechanism for the cyclic aggregation and dispersal of continents. The large-scale flow probably has smaller scale complications, especially in the upper mantle, and may be intrinsically unsteady. Narrow ascending thermal plumes could probably be incorporated into this model.

show abstract

Section: H E M I C a L Composftion D I F F E R E N C E Smentioning

confidence: 69%

Whole-mantle convection and plate tectonics

Davies¹

1977

Geophys J Int

133

View full text Add to dashboard Cite

show abstract

“…Although most cosmochemical models call for a perovskite plus magnesiowustite lower mantle, physical data permit considerable free silica. Thus, e.g., it has been shown that the mean atomic weight W is constant and close to 21 throughout the lower mantle [32]. For silica, W is 20.…”

Section: Discussionmentioning

confidence: 97%

Molecular dynamics modeling of stishovite

Luo

Çağın

Strachan

et al. 2002

Earth and Planetary Science Letters

View full text Add to dashboard Cite

“…Larger increases in density are implausible, since seismic observations limit the total density increase between the upper and lower mantle, and there is also a limit to the bulk earth iron content. If the density increase is due solely to an increase in iron concentration, the mean atomic weight r• increases by more than 2 g/mol of atoms, which may be detectable according to various formulations of velocity-density systematics [Watt et al, 1975].…”

Section: Changes In Iron And/or Silica Contentmentioning

confidence: 99%