Constraining the chemical, rheological and electromagnetic properties of the lowermost mantle (D") is important to understand the formation and dynamics of the Earth's mantle and core. To explain the origin of the variety of characteristics of this layer observed with seismology, a number of theories have been proposed¹, including core–mantle interaction, the presence of remnants of subducted material and that D" is the site of a mineral phase transformation. This final possibility has been rejuvenated by recent evidence for a phase change in MgSiO₃ perovskite (thought to be the most prevalent phase in the lower mantle²) at near core–mantle boundary temperature and pressure conditions³. Here we explore the efficacy of this 'post-perovskite' phase to explain the seismic properties of the lowermost mantle through coupled ab initio and seismic modelling of perovskite and post-perovskite polymorphs of MgSiO₃, performed at lowermost-mantle temperatures and pressures. We show that a post-perovskite model can explain the topography and location of the D" discontinuity, apparent differences in compressional- and shear-wave models¹ and the observation of a deeper, weaker discontinuity^{4, 5}. Furthermore, our calculations show that the regional variations in lower-mantle shear-wave anisotropy are consistent with the proposed phase change in MgSiO₃ perovskite.

1. School of Earth and Environment, University of Leeds, Leeds LS2 9JT, UK
2. Department of Earth Sciences, University College London, Gower Street, London WC1E 6BT, UK
3. Department of Earth Sciences, University of Bristol, Wills Memorial Building, Queens Road, Bristol BS8 1RJ, UK
4. †Present address: Department of Earth Sciences, University of Bristol, Wills Memorial Building, Queens Road, Bristol BS8 1RJ, UK

Correspondence to: James Wookey^1,4 Correspondence and requests for materials should be addressed to J.W. (Email: j.wookey@bristol.ac.uk).

Received 26 July 2005 | Accepted 18 October 2005

MORE ARTICLES LIKE THIS

These links to content published by NPG are automatically generated.