A Caltech Library Service

Upper mantle anisotropy: evidence from free oscillations

Anderson, Don L. and Dziewonski, Adam M. (1982) Upper mantle anisotropy: evidence from free oscillations. Geophysical Journal of the Royal Astronomical Society, 69 (2). pp. 383-404. ISSN 0016-8009. doi:10.1111/j.1365-246X.1982.tb04956.x.

PDF - Published Version
See Usage Policy.


Use this Persistent URL to link to this item:


Isotropic earth models are unable to provide uniform fits to the gross Earth normal mode data set or, in many cases, to regional Love-and Rayleigh-wave data. Anisotropic inversion provides a good fit to the data and indicates that the upper 200km of the mantle is anisotropic. The nature and magnitude of the required anisotropy, moreover, is similar to that found in body wave studies and in studies of ultramafic samples from the upper mantle. Pronounced upper mantle low-velocity zones are characteristic of models resulting from isotropic inversion of global or regional data sets. Anisotropic models have more nearly constant velocities in the upper mantle. Normal mode partial (Frediét) derivatives are calculated for a transversely isotropic earth model with a radial axis of symmetry. For this type of anisotropy there are five elastic constant. The two shear-type moduli can be determined from the toroidal modes. Spheroidal and Rayleigh modes are sensitive to all five elastic constants but are mainly controlled by the two compressional-type moduli, one of the shear-type moduli and the remaining, mixed-mode, modulus. The lack of sensitivity of Rayleigh waves to compressional wave velocities is a characteristic only of the isotropic case. The partial derivatives of the horizontal and vertical components of the compressional velocity are nearly equal and opposite in the region of the mantle where the shear velocity sensitivity is the greatest. The net compressional wave partial derivative, at depth, is therefore very small for isotropic perturbations. Compressional wave anisotropy, however, has a significant effect on Rayleigh-wave dispersion. Once it has been established that transverse anisotropy is important it is necessary to invert for all five elastic constants. If the azimuthal effect has not been averaged out a more general anisotropy may have to be allowed for.

Item Type:Article
Related URLs:
URLURL TypeDescription
Additional Information:© 1982 Royal Astronomical Society. Received 1981 September 21; in original form 1981 May 22. This research was supported by National Science Foundation Grants No. EAR77-14675, and EAR78-05353 and National Aeronautics and Space Administration Grant No. NSG-7610. We thank Brian Mitchell for preprints of his work with G. Yu. Contribution No. 3592, Division of Geological and Planetary Sciences, California Institute of Technology.
Funding AgencyGrant Number
Other Numbering System:
Other Numbering System NameOther Numbering System ID
Caltech Division of Geological and Planetary Sciences3592
Issue or Number:2
Record Number:CaltechAUTHORS:20140507-111433818
Persistent URL:
Official Citation: Don L. Anderson and Adam M. Dziewonski Upper mantle anisotropy: evidence from free oscillations Geophys. J. Int. (1982) 69 (2): 383-404 doi:10.1111/j.1365-246X.1982.tb04956.x
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:45572
Deposited By: Ruth Sustaita
Deposited On:07 May 2014 19:54
Last Modified:10 Nov 2021 17:14

Repository Staff Only: item control page