SEISMIC TOMOGRAPHY INFERS A TRENCH MISSING FROM PLATE RECONSTRUCTIONS OF THE CRETACEOUS FARALLON OCEAN

Over the past five years, the USArray experiment has turned North America into the seismically best illuminated continent on earth. I present results and interpretations of finite-frequency body-wave tomography, a method capable of imaging to particularly deep mantle depths (1500-1800 km).

A surprising new observation is the vast amount of subducted slab in the lower mantle under western North America. Conflicting evidence seems to come from plate reconstructions: the Farallon/North America trench has traversed these longitudes too recently (only since ~50 Myr) to have allowed the slab to sink so deeply, according to commonly accepted notions on slab sinking velocities. At the reasonably required times (late Cretaceous), the reconstructed continental trench was located too far east, and copious volumes of lower-mantle slab do exist further east to account for its predicted activity. This leaves the deep western slab to be explained in some other way. A key observation is that the western slab directly connects upward to the present-day Cascadia trench, implying uninterrupted subduction from Cretaceous times until now.

Hence the global plate reconstructions are most likely missing an intra-oceanic trench in the northeastern Farallon basin during the Upper Cretaceous, of NW-SE strike as defined by the imaged lower-mantle slab. This trench and its associated island arc were accreted around 50 Myr to the westward-moving continent. Subduction continued largely uninterrupted into the newly-continental trench, whereas the previous continental trench was choked off at this time.

This interpretation of the tomographic image appears to be consistent with terrane accretion histories along the west coast of the U.S., which date the last big accretion event to ~50 Myr. It implies that North American land geology reflects the concurrent operation of two subduction systems prior to the Eocene, which may help explain some of the observed complexities, e.g. the Laramide episode.