DSpace Repository

The structure of the upper mantle, transition zone and uppermost lower mantle beneath southern Africa from broadband seismology

Show simple item record

dc.contributor.advisor Wright C en
dc.contributor.author Simon RES en
dc.date.accessioned 2016-09-22T11:14:12Z
dc.date.available 2016-09-22T11:14:12Z
dc.date.created 1998 en
dc.date.submitted 2003 en
dc.identifier.uri http://hdl.handle.net/20.500.11892/106369
dc.description.abstract Average body wave models from the surface to depths of 800 km have been derived for southern Africa and surrounding oceans using travel times and waveforms from earthquakes recorded at stations of the Kaapvaal craton and the South African National networks. Like most seismological techniques, our derived body wave models are prone to some averaging of differing structures. This is evident from the continental and oceanic regions where rays turned. The derived models averaged the wavespeed structure laterally across the Cape Fold belt, the Kaapvaal craton, Limpopo belt, the Namaqua-Natal mobile belt, the Zimbabwe craton, and portions of other surrounding belts (e.g Mozambique, Damara, Zambezi belts). However, most research work that discuss the deeper parts of the upper mantle and transition zone have to use an assumed model for the crust and shallow parts of the mantle, which is not the case in this work. In this first study of the transition zone for the central part of the African superswell, a damped least-squares inversion was used to minimize effects of origin time errors.<br><br> The Herglotz-Wiechert method combined with ray tracing was used to derive a preliminary P wavespeed model, followed by refinements using phase-weighted stacking and synthetic seismograms to yield the final model. A remarkably high quality set of shear wave (SV) travel times combined with ray tracing were used to derive the S wavespeed model, which was also refined using phase-weighted stacking and synthetic seismograms. These shear wave data represent not only the first SV record section for the region of study but also the first dense SV record section for the transition zone in the African continent. The S wavespeed model shows a good fit to both SH and SV travel times, indicating that anisotropy changes incoherently. Both average models are in good agreement with wavespeeds estimated from peridotite xenoliths found within the Kaapvaal craton. This agreement determined between seismology and petrology suggests that the nodules are representative of the upper mantle composition to minimum depths of at least 180 km in this region. Furthermore, the main features of the broadband data are reasonably well matched by both average models, as evidenced from the reduced travel times and synthetic seismograms.<br><br> Observations from broadband body waves confirm the presence of resolvable differences in the form of P and S wavespeed variations above the transition zone. We find evidence of a high wavespeed upper mantle lid in the S wavespeed model overlying a low wavespeed zone around 210 to -345 km depth that is not observed in the P wavespeed model. The base of the LWZ beneath southern Africa, which occurs at -345 km from our shear wave model is consistent with that at -350 km from results that satisfy both the regional seismic waveform data and the fundamental mode Rayleigh wave data for southern Africa. The derived average models both show a prominent 410 km discontinuity, in agreement with the SATZ model for a region of southern Africa to the north of the region covered by the present study. However, a weakly-defined 670 km discontinuity is observed in the P wavespeed whereas the S wavespeed shows a moderately sharp discontinuity.<br><br> Below the 410 km discontinuity throughout the transition zone down to the uppermost lower mantle, our P/S wavespeed ratios are very similar to those obtained from other stable continental regions. Above the transition zone, however, the P/S ratios from the different stable continental regions reveal variability in regional zones within the uppermost mantle structure. However, within the transition zone the P/S ratios show a slight increase, probably indicative of nothing anomalous within this region. Our average models reveal that the use of the IASP91 model in southern Africa as a basis for tomographic inversion is inadequate in the uppermost mantle, since our models show high wavespeeds compared with the global average. However, the average shear wavespeed structure beneath southern Africa within and below the transition zone is similar to that of the global average, but slightly higher P wavespeeds are observed especially below 500 km depth through to the base of the transition zone.<br><br> Our estimates of the transition zone thickness of 248 and 245 km from P and S waves respectively, are not only consistent with the global average thickness estimates obtained from recent SS precursor studies, but also with results of P-to-S conversions beneath southern Africa. Therefore, our seismic results indicate no regions of anomalous low wavespeeds within the uppermost 800 km of the mantle beneath southern Africa that could be associated with high temperatures and the uplift of the African superswell. We conclude that the transition zone and the uppermost lower mantle below southern Africa and surrounding oceans are consistent with a pyrolite composition, while the uppermost mantle is predominantly a refractory peridotite. en
dc.language English en
dc.title The structure of the upper mantle, transition zone and uppermost lower mantle beneath southern Africa from broadband seismology en
dc.type Doctoral degree en
dc.description.degree PhD (Geosciences) en

Files in this item

Files Size Format View

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record