Large Scale Anisotropy Near the Core-Mantle Boundary from Global Waveform Inversion

Mark Panning and Barbara Romanowicz


The Earth's core-mantle boundary (CMB) is both a thermal and chemical boundary layer between the solid silicate mantle and the fluid iron outer core. The mantle-side portion of this layer (D"), is therefore the site of dynamic processes that may involve both thermal and chemical heterogeneity at various scales. Additionally, this area also functions as a mechanical boundary layer for the convection of the overlying mantle, leading to intense deformation. This deformation can lead to detectable seismic anisotropy, either through the alignment of anisotropic crystals in the strain field or through the alignment of layering or inclusions of materials with strongly contrasting elastic properties (Karato, 1998; Kendall and Silver, 1996).

Anisotropy in D" has been well established in several regions, including under the Pacific, northern Asia, Alaska, and central America, from the observation of S waves diffracting (Sdiff) or reflecting (ScS) at the CMB (Vinnik et al., 1989; Kendall and Silver, 1996; Lay et al., 1998). However, these studies only sample limited areas of D", and therefore interpretation is difficult. A more global picture of long-wavelength anisotropic D" structure would clearly aid interpretation both in terms of dynamic flow modeling as well as mineral physics.

With this in mind we have adapted our global waveform tomography approach (Mégnin and Romanowicz, 2000) to develop a 3D model of radial anisotropy throughout the mantle using a large dataset of three component time-domain waveforms of both surface and body waves.

Figure 34.1: Radially symmetric values of $\xi $ as a function of depth in the model. The values for PREM, the starting model, are shown by the dashed line, and the 670 discontinuity is shown by the dotted line. Notice the strong increase at the base of the mantle, similar but smaller in magnitude to that seen in the upper mantle.
\epsfig{file=panning03_1.epsi, width=8cm}\end{center}\end{figure}

D" Anisotropic Model

Figure: $\delta \ln{V_S}$ (A and B) and $\delta \ln{\xi}$ (C and D) shown at a depth of 2800 km. Slices are shown centered under the Pacific (A and C) and Africa (B and D).
\epsfig{file=panning03_2.epsi, width=8cm}\end{center}\end{figure}

The model is parameterized in terms of isotropic $V_S$ and the anisotropic $\xi $ parameter ( $\xi = V_{SH}^2 / V_{SV}^2$), which is directly related to radial anisotropy in shear velocity. In our model, D" is characterized by a strong radially symmetric signature of radial anisotropy, as seen in the uppermost mantle in previous anisotropic models such as PREM (Dziewonski and Anderson, 1981). Similar to the uppermost 200 km of the mantle, this signature is a positive $\delta \ln{\xi}$, indicating that horizontally polarized shear velocity, $V_{SH}$, is faster than vertically polarized shear velocity, $V_{SV}$ (Figure 34.1).

The 3D isotropic velocity imaged in D" in this study is consistent with earlier tomographic models of shear velocity in this depth range (Masters et al., 1996; Mégnin and Romanowicz, 2000), and is characterized by a strong degree 2 component representing a fast ring surrounding two low velocity features (often called superplumes) centered beneath the central Pacific and Africa (Figure 34.2, A and B). In the $\xi $ model, the strong degree 0 component appears to be limited to the lowermost 300 km (Figure 34.1, inset), but the regions that differ most from this average structure correlate well with the locations of the superplumes, with reduced values of $\delta \ln{\xi}$ under the central Pacific, Africa, and the south Atlantic, including patches with negative values ($V_{SV}>V_{SH}$).

The long-wavelength anisotropic features imaged in our model generally agree with more localized studies of D" anisotropy. Specifically, these studies imaged areas with positive $\delta \ln{\xi}$ beneath central America and Alaska. The central Pacific appears to be more variable with some areas showing negative $\delta \ln{\xi}$ (Lay et al., 1998).


Our study extends to a global scale the results obtained so far for limited regional sampling of D". The dominant $V_{SH} > V_{SV}$ found as one approaches the CMB suggests that the anisotropy observed in D" is related to the dominant horizontal flow in a mechanical boundary layer, analogous to the larger signal observed in the uppermost 200 km of the mantle. As one approaches regions of upwelling, the direction of flow changes and results in a different signature of anisotropy, as manifested in our study under the central Pacific and Africa. In reality, anisotropy in these regions bordering the large scale upwellings may be much more complex and include tilting of the vertical axis of symmetry assumed in our modeling. This could lead to azimuthal anisotropy which we do not yet attempt to model.

Although our model does not determine the microscopic causes of the observed anisotropy, the results clearly suggest that the dynamics of D" correspond with what would be expected in a boundary layer dominated by horizontal flow, and emphasize the unique character of the two superplume regions. Although mineral physics data are not yet available for the pressure and temperature conditions at the base of the mantle, our results suggest that similar relationships between anisotropic signature and flow prevail in the uppermost and lowermost mantle.


Dziewonski, A.M., and D.L. Anderson, Preliminary reference Earth model, Phys. Earth Planet. Int., 25, 297-356, 1981.

Karato, S.-I., Some remarks on the origin of seismic anisotropy in the D" layer Earth, Planets, Space, 50, 1019-1028, 1998.

Kendall, J.-M. and P.G. Silver, Constraints from seismic anisotropy on the nature of the lowermost mantle, Nature, 381, 409-412, 1996.

Lay, T., Q. Williams, E.J. Garnero, L. Kellogg, and M. Wysession, Seismic wave anisotropy in the D" region and its implications, in The core-mantle boundary region, M. Gurnis, M.E. Wysession, E. Knittle, B.A. Buffett, Eds., Amer. Geophys. Union, pp. 219-318, 1998.

Masters, G., H. Bolton, and G. Laske, Joint seismic tomography for P and S velocities: How pervasive are chemical anomalies in the mantle? EOS Trans. AGU, 80 S14, 1999.

Mégnin, C., and B. Romanowicz, The 3D shear velocity of the mantle from the inversion of body, surface, and higher mode waveforms, Geophys. Jour. Intl., 143, 709-728, 2000.

Vinnik, L., V. Farra, and B. Romanowicz, Observational evidence for diffracted SV in the shadow of the Earth's core, Geophys. Res. Lett., 16, 519-522, 1989.

Berkeley Seismological Laboratory
215 McCone Hall, UC Berkeley, Berkeley, CA 94720-4760
Questions or comments? Send e-mail:
© 2004, The Regents of the University of California