First views from assimilation of a new ‘tomographic’ circum-Pacific plate reconstruction into mantle circulation models

One of the greatest challenges of modeling the plate tectonic history of Earth during the Mesozoic and Cenozoic eras lies in reconstructing the Pacific Ocean and its predecessor ocean, Panthalassa.  A major reason for the plate tectonic uncertainty in this region is extensive subduction, which has consumed most (>95%) of the Pacific-Panthalassan ocean lithosphere formed since 150 Ma (Törsvik et al., 2019) and recycled it into the mantle, destroying the information on past plate motions recorded by seafloor magnetic lineations.  Consequently, many circum-Pacific margin plate tectonic models, including the popular GPlates models (e.g. Matthews et al., 2016; Müller et al., 2019), necessarily extrapolate 1000’s of km of subducted seafloor (i.e. synthetic seafloor isochrons).  Given the limited constraints, it is understandable that such models also prefer more straightforward solutions with a smaller number of larger plates, avoiding the complexities of modeling intra-oceanic subduction despite geological evidence from accreted circum-Pacific oceanic terranes.

Here we build the first topologically-closed, global plate tectonic model of the circum-Pacific using structurally-restored slabs from mantle seismic tomography as our primary constraint.  We use the numerical code TERRA to assimilate three variants of our ‘tomographic’ global plate model into mantle circulation forward models and assimilate the default GPlates model as a reference.  We show our preliminary geodynamic modeling results and test our model predictions against observed mantle structure, Earth’s geoid, and oceanic realm dynamic topography. 

All cases favor plate models that incorporate intra-oceanic subduction within Pacific-Panthalassa, particularly within the northern Pacific.  We find robust support for significant slab lateral advections (i.e. non-vertical slab sinking) under NW Pacific basin.  We discuss similarities and differences between our new ‘tomographic’ plate models and the GPlates model, which has been used for almost all geodynamic studies of the circum-Pacific to date.