Seismic structure of Precambrian lithosphere: New constraints from broad-band surface-wave dispersion

Depth distributions of seismic velocities and their directional dependence (anisotropy) in the crust and mantle beneath cratons yield essential constraints on processes of their formation and evolution. Despite recent progress in mapping the lateral extent of cratonic roots around the globe, profiles of seismic velocities within them remain uncertain. In this study we employ a novel combination of waveform-analysis techniques and measure inter-station Rayleigh- and Love-wave phase velocities in broad period ranges that enable resolution from the upper crust to deep upper mantle. Sampling a selection of 10 Archean and Proterozoic locations, we derive new constraints on the isotropic and radially anisotropic seismic structure of Precambrian lithosphere.Shear-wave speed VS is consistently higher in the lithosphere of cratons than in the lithosphere of Proterozoic foldbelts. Because known effects of compositional variations in the lithosphere on VS are too small to account for the difference, this implies that temperature in cratonic lithosphere is consistently lower, in spite of sub-lithospheric mantle beneath continents being thermally heterogeneous, with some cratons underlain, as we observe, by a substantially hotter asthenosphere compared to others. Lithospheric geotherms being nearly conductive, this confirms that the stable, buoyant lithosphere beneath cratons must be substantially thicker than beneath younger continental blocks.An increase in VS between the Moho and a 100-150 km depth is consistently preferred by the data in this study and is present in seismic models of continents published previously. We argue that this is largely due to the transition from spinel peridotite to garnet peridotite, proposed previously to give rise to the “Hales discontinuity” within this depth interval. The depth and the width of the phase transformation depend on mantle composition; it is likely to occur deeper and over a broader depth interval beneath cratons than elsewhere because of the high Cr content in the depleted cratonic lithosphere, as evidenced by a number of xenolith studies. Seismic data available at present would be consistent with both a sharp and a gradual increase in VS in the upper lithosphere (a Hales discontinuity or a “Hales gradient”). The VS profile in the upper mantle lithosphere is not shaped by the temperature distribution only; this needs to be considered when relating seismic velocities to lithospheric temperatures.Radial anisotropy in the upper crust is observed repeatedly and indicates vertically oriented anisotropic fabric (VSH < VSV); this may yield a clue on how cratons grew, lending support to the view that distributed crustal shortening with sub-vertical flow patterns occurred over large scales in hot ancient orogens. In the lower crust and upper lithospheric mantle, radial anisotropy consistently reveals horizontal fabric (VSH > VSV); the fabric can be interpreted as a record of (sub-)horizontal ductile flow in the lower crust and lithospheric mantle at the time of the formation and stabilisation of the cratons. We also find indications for radial anisotropy below 200 km depth, corroborating recent evidence for anisotropy in the asthenosphere beneath cratons due to current and recent asthenospheric flow.