ISSN:
1420-9136
Schlagwort(e):
Lower mantle
;
seismic tomography
;
thermoelasticiti
;
composition models
;
geodynamics
;
shear modulus
Quelle:
Springer Online Journal Archives 1860-2000
Thema:
Geologie und Paläontologie
,
Physik
Notizen:
Abstract We estimate (ϖμ/ϖT) P of the lower mantle at seismic frequencies using two distinct approaches by combining ambient laboratory measurements on lower mantle minerals with seismic data. In the first approach, an upper bound is estimated for |(ϖμ/ϖT) P | by comparing the shear modulus (μ) profile of PREM with laboratory room-temperature data of μ extrapolated to high pressures. The second approach employs a seismic tomography constraint (ϖ lnV S /ϖ lnV P ) P =1.8–2, which directly relates (ϖμ/ϖT) P with (ϖK S /ϖT) P . An average (ϖK S /ϖT) P can be obtained by comparing the well-established room-temperature compression data for lower mantle minerals with theK S profile of PREM along several possible adiabats. Both (ϖK S /ϖT) and (ϖμ/ϖT) depend on silicon content [or (Mg+Fe)/Sil of the model. For various compositions, the two approaches predict rather distinct (ϖμ/ϖT) P vs. (ϖK S /ϖT) P curves, which intersect at a composition similar to pyrolite with (ϖμ/ϖT) P =−0.02 to −0.035 and (ϖK S /ϖT) P =−0.015 to −0.020 GPa/K. The pure perovskite model, on the other hand, yields grossly inconsistent results using the two approaches. We conclude that both vertical and lateral variations in seismic velocities are consistent with variation due to pressure, temperature, and phase transformations of a uniform composition. Additional physical properties of a pyrolite lower mantle are further predicted. Lateral temperature variations are predicted to be about 100–250 K, and the ratio of (ϖ lnp/ϖ lnV S ) P around 0.13 and 0.26. All of these parameters increase slightly with depth if the ratio of (ϖ lnV S /ϖ lnV P ) P remains constant throughout the lower mantle. These predicted values are in excellent agreement with geodynamic analyses, in which the ratios (ϖ ln ϱ/ϖ lnV S ) P and (ϖ ϱ/ϖ lnV S ) P are free parameters arbitrarily adjusted to fit the tomography and geoid data.
Materialart:
Digitale Medien
URL:
http://dx.doi.org/10.1007/BF00874732
Permalink
