 |
| Titel |
Linking seismology, mineralogy and geodynamics with seismic anisotropy in the lowermost mantle |
| VerfasserIn |
Andy Nowacki, Andrew Walker, James Wookey, Jack Walpole, Guy Masters, J.-Michael Kendall |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250082614
|
|
|
|
|
|
| Zusammenfassung |
| The core-mantle boundary (CMB) is the site of the largest change in properties in the Earth,
where the liquid outer core and solid mantle meet. Forming the lower boundary layer in the
convecting mantle, D″ (the lowermost mantle) may hold the key to understanding
dynamics both above and below. One property of the region which holds much potential
to advance this understanding is its seismic anisotropy, which may be caused by
factors such as the alignment of anisotropic mineral grains in response to mantle
flow.
Anisotropy is widely observed in D″, yet not in the overlying mantle more than a few
hundred kilometres above the CMB, as evidenced by numerous tomographic and waveform
studies. Shear wave splitting is an unambiguous indicator of the presence of anisotropy and
measurements thereof need not make any simplification regarding the kind of anisotropy.
Such measurements therefore allow us to test the widest range of candidate processes which
might cause D” anisotropy. Ultimately, if one cause such as mineral alignment is more likely
than others, we can then use seismic anisotropy to directly infer flow in the lowermost
mantle.
In order to test candidate processes for D″ anisotropy, we construct a series of elastic
models of the lowermost mantle. Each is based on a different assumption regarding the
cause of lowermost mantle anisotropy, concentrating thus far on the development of
lattice-preferred orientation in dislocation creep in lower mantle mineral phases such as
perovskite, post-perovskite and (Mg,Fe)O (and mixtures thereof). In order to do
this, for these phases we require mineral physical data regarding the single-crystal
elasticity and deformation mechanisms. Whilst there exists some uncertainty in these
parameters, we can nevertheless test what effect these have on our final models. We then
use a steady-state mantle flow field retrieved from seismic, geodetic and mineral
physical observables, and calculate the texturing along pathlines in the lowermost
mantle, eventually producing a three-dimensional model of completely general
elasticity.
Observations of seismic anisotropy in ScS waves are then re-created for our candidate
models and direct comparison can be made with the data. A complicating factor is that the
ray-theoretical assumption may not accurately capture the sensitivity of the waves to varying
D″ elastic structure, and thus we use a spectral-element approach to calculate synthetic
seismograms at the same frequency as the observations (~0.2Â Hz). The calculations
involve thousands of processors and terabytes of memory, but are necessary for
retrieving the wavefield in a fully anisotropic medium. We compare a new set of global
observations of shear wave splitting in ScS, corrected for upper mantle anisotropy,
and can potentially rule in or out different causative mechanisms for anisotropy in
the lowermost mantle. More constraints can be incorporated in the future as our
method allows the measurement of any seismic phase and any causative mechanism. |
|
|
|
|
|
|