Seismic discontinuity structure beneath the Canadian Shield and the signature of continental roots

• Main Author: Thompson, David Andrew
• Published: University of Bristol 2011
• Subjects: 551.22
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.559492

The Canadian Shield is one of the largest exposures of Precambrian geology on the Earth, and is underlain by a deep (>250 km) and laterally extensive cratonic root. The processes which formed the crust during this period of Earth history remain a matter of considerable debate, as do the events which led to the creation and stabilisation of the sub-continental lithospheric mantle. The thermal effect that these cold, refractory roots have on the underlying asthenosphere and mantle transition zone is also unclear. The Hudson Bay intracratonic basin lies at the core of stable continental North America with several hypotheses, including crustal thinning through extension or a surface-coupled mantle downwelling, being postulated for its existence. The reason for several of these problems being unanswered is due to the regions inaccesiblilty and harsh climate. However, recent seismic deployments across the Hudson Bay region have allowed these hypotheses to be tested through a teleseisrnic receiver function study. Bulk crustal properties (thickness, Vp/Vs) determined using the H-}( stacking method (Zhu and Kanamori, 2000) indicate that the Meso- to Paleoarchean crust of the Rae domain is strikingly simple with uniform thickness (~37 km), a sharp Moho and low Vp/Ys ratio « 1.73) indicative of a felsic-to-intermediate composition. There is little evidence from the geological record of the central Rae domain for collisional processes, and the spatial extent of felsic crust is more consistent with models of crustal formation involving vertical tectonic processes. The thickest crust observed lies on southern Baffin Island (~43 km), coincident with the Paleoproterozoic Trans-Hudson Orogen. The results suggest that crusta I thicknesses could have been originally in excess of 65 km, similar to what is seen in modern collisional environments. The results point toward secular crusta I evolution. from non-plate tectonic prior to 3.0Ga through to fully developed Wilson cycle style plate tectonics by the Paleoproterozoic. Heterogeneous (dipping and/or anisotropic) structure within the shallow lithospheric mantle is observed on both the radial and tangential RF components beneath the Melville Peninsula and northern Baffin Island. The pattern of periodicity on the tangential component suggests either an interface or anisotropic layer dipping towards at 240°/60°, agreeing well with dike swarms located on central Baffin Island, believed to be emplaced during localised intrabasinal rifting during the Paleopro- terozoic. This makes extensional processes a prime candidate for the structure primarily observed in the lithospheric mantle, possibly favouring this type of deformation as a mechanism for subsidence of the Hudson Bay intracratonic basin. Pds arrival times and differential times (P660s-P4lOs, indicative of transition zone thickness) from the 41 O-km (the '410') and 660-km (the '660') seismic discontinuities are analysed for stations located across the entire Canadian Shield. 1-0 time migrations and 3-~ depth migrations show that the transition zone discontinuities are unperturbed beneath one of the largest and deepest cratonic roots on the Earth, with the implication that the root has little «SO K) or no thermal effect at transition zone depths. The internal discontinuity structure is also simple, meaning the observations can be reconciled with phase changes in the olivine system alone. Frequency dependence of amplitude from the '410' show that the pattern with frequency is indicative of a gradational increase in velocity with depth across the '410'. Comparing these observations to synthetic RFs suggests that the sharpness of the '410' is in the region of 20-34 km, depending on the nature of the velocity increase. This sharpness estimate is consistent with water contents of 550 pprn to greater than 1200 ppm. These values lie above the saturation point of peridotite at these depths meaning a region containing dehydration melting, and hence a seismic low velocity layer, should be present above the '410'. This is inconsistent with much of the migrated data, and other avenues for broadening may need to be explored. Velocity jumps constrained by the modelling are 5-7%, above that of the ak 135 model (4.4%). The higher velocity jump appears to be consistent with a mantle of pyrolitic composition. Collectively, the results provide new fundamental constraints on the processes that were occurring during the Precambrian from a previously unstudied region. The study has also highlighted the simplicity of the mantle transition zone beneath a vast cratonic root, raising questions regarding the presence of small-scale convection beneath continents and their affect on the Earth's heat budget.