UNITED PLATES OF AMERICA REVISITED: DID SUBDUCTION-INDUCED MANTLE FLOW BRING THE CRATONS TOGETHER? (Invited Presentation)

Interior (pre-Grenville) Laurentia is a tectonic aggregate of six or more Archean cratons, plus non-uniformly distributed Paleoproterozoic juvenile crust. How and why did all (but one) of the cratons coalesce convulsively in geons 19 and 18? The age and subduction polarity of their mutual geosutures provides a clue. If slab roll-back dominated the dynamics of plate convergence and suturing, then we should expect that the aggregation process should have started around the Superior craton, because it occupied the subducting plate with respect to neighboring cratons (Hearne, Rae, North Atlantic, Marshfield). Conversely, if trench suction was dominant—traction by convergent and sinking flow in the ambient mantle in response to cooling by slabs—we should expect that aggregation began around the Rae craton, which was the upper plate with respect to neighboring cratons (Slave, Hearne, Superior, North Atlantic). Collision ages have been determined by U-Pb dating of tuffs on the lower plate at the passive-margin to foredeep stratigraphic transition, and on the upper plate from igneous rocks in transition from arc to slab-failure petrogeneses. Minimum age constraints are obtained from metamorphic and exhumation ages of deformed passive-margin rocks. The data indicate that the oldest geosutures bound the Rae craton, 1.97 Ga for Rae-Slave and 1.92 Ga(?) for Rae-Hearne. The other geosutures are all younger than 1.90 Ga. The geosutures become systematically younger with distance from the center of the Rae craton, with a correlation coefficient r² ≈ 0.7. There is no correlation whatsoever between geosuture age and distance from the center of the Superior craton. These results indicate that proto-Laurentia aggregated around the Rae craton, and are therefore suggestive of a positive feedback involving subduction, mantle flow and convergent plate motion. They support the orthoversion model of Mitchell, Kilian and Evans (Nature 482: 208, 2012) for supercontinental episodicity. The analysis needs to be extended to Nuna overall, and to other supercontinents.