The two end-member schools of thought explaining observations of seismic anisotropy of the upper mantle are that the anisotropy results from frozen-in fabrics created at formation or subsequent deformation (Silver model) or that the anisotropy results from modern day plate motion at the lithosphere-asthenosphere boundary (Vinnik model). Regions of the world can be found for which either of these alternately is an adequate explanation of the observations. However, other regions exist that defy these simplistic models and for which more elaborate explanations are required.
Early SKS observations on the Kaapvaal craton have been interpreted to support the Vinnik model (Vinnik et al., 1992, 1995) whereas more recently, new observations based on SKS results from the Southern African Seismic Experiment (SASE) data to support the Silver model (Silver et al., 2001, 2004). We re-analysed the SASE data seeking to establish, in particular, that the assumption of a single layer of anisotropy is valid. With the sole exception of one event at one station, we are unable to reject this hypothesis, however the data are poor for higher complexity studies that sinply a single horizontal layer of anisotropy. Generally our results are in agreement with those of Silver et al. (2001, 2004), except that our error estimates are a factor of 3-5 larger.
We demonstrate that for some of the SASE sites there is excellent agreement between the fast direction of anisotropy and the modelled absolute plate motion, and that those sites lie in the centres of the Kaapvaal and Zimbabwe cratons. Conversely, the sites on the Proterozoic modile belts exhibit directions that are at a high oblique angle to APM.
Based on these correlations with tectonics, we propose a new model to explain SKS anisotropy. We suggest that Archean lithosphere was formed in a manner that did not generate coherent deformation, thus Archean lithosphere has no distinct anisotropy. In contrast, Proterozoic lithosphere was created with inherent coherent deformation, resulting in lattice preferred orientation (LPO) not related to modern APM. However, the thicker Archean roots compared to thinner Proterozoic lithosphere results in faster asthenospheric flow, resulting in the generation of LPO in the APM direction at the base of the lithosphere.