REALLY NOW, HOW THICK ARE THE CRATONS?

While it is generally agreed that the cratons are underlain by thick, basalt-depleted keels in near-isopycnic balance, the maximum thickness of the keels has remained controversial since this theory of the continental tectosphere was proposed in the mid-1970s. Recent reviews have combined tomographic models with other information to conclude that cratonic keels achieve maximum depths of 200-250 km. Here we show that this inference is not consistent with some first-order aspects of upper-mantle tomography. In a meta-analysis of more than 20 whole-mantle tomographic models published by various research groups, we estimate the correlation between upper-mantle seismic structure and crustal tectonic structure by projecting the tomographic models onto the GTR1 global tectonic regionalization. These projections show consistent and statistically significant cratonic signatures extending to depths exceeding 300 km. In particular, the average shear velocities of stable continental regions in the depth interval 200-400 km are higher than those of the convecting oceanic mantle, and the corresponding shear-velocity gradients are distinctly lower, in good agreement with high-resolution regional models derived from multi-phase waveform inversions. We conclude that the average thickness of the cratonic tectosphere is unlikely to be less than 300 km; indeed, beneath some (if not most) cratons, the keels nearly fill the entire upper mantle. The isopycnic requirements for thick cratonic tectosphere place strong constraints on models of continental formation and evolution.