Seismic Tomographic Images of the Cratonic Upper Mantle Beneath the Canadian Western Superior Province: a Remnant Archean Slab?

\def\twist{\raise 0.1em ~} Broadband digital data of the Lithoprobe TW{\twist}ST experiment were tomographically inverted for upper--mantle P and S velocity structures beneath the Western Superior Province. The 17 portable broadband stations were arrayed in northern Ontario, Canada, so as to crosscut the strike of many subprovinces as well as the boundary with the Proterozoic Trans--Hudson Orogen to the north. The resulting tomographic images reveal three robust velocity anomalies: (i) a dipping tabular high--velocity anomaly; (ii) a relatively shallow low--velocity anomaly directly above the positive anomaly; and (iii) a deep low--velocity body. The first anomaly could be interpreted as a 30--50 km thick eclogite/dunite layer representing a remnant subducted oceanic lithosphere. The presence of this slab--shaped body within the cratonic root suggests its origin at around 2.7 Ga and its SE--NW strike is perceptibly oblique to the main E--W trend of the subprovince boundaries. Resolution tests suggest that this dipping high--velocity anomaly could extend to the vicinity of the 660 km seismic discontinuity. The low--velocity anomalies may be related to processes that occurred at the edges of the descending slab or may be expressions of later upwelling material. The presence of a thick cratonic root (≈300 km) may also be revealed by the tomographic images. Previous SKS--splitting large delay times (1.1 to 2.1s) are consistent with a thick root, and fast polarization trends, which are broadly subparallel to the lithotectonic boundaries, appear at the southern end of the array to be somewhat rotated as if influenced by the presence of an `oblique' structure. It may be wrong to assume the apparent slab structure is stationary in the upper mantle. Helmstaedt and Schulze's [1989] model of continental--root formation by imbrication of subducted Archean plates suggests another possibility, namely subsequent delamination and sinking of one or more plates due to changing conditions. This could alleviate the need to explain an extreme longevity for an apparent slab structure at asthenospheric depths. Indeed, there are other scenarios once we consider the possibility of active structures. For example, if the apparent transition at around 300 km does represent the continental root extent, then the deeper lateral variations may be unrelated structures over which the craton has moved. While we must remain open to such possibilities, we prefer, at least for the purposes of future hypothesis testing, to focus on interpretations that may relate more directly to the surface geology. Overall, these traveltime inversions are interpreted as providing evidence at depth for late Archean structures resembling those of modern subduction tectonics.