T53C-4699:
Investigating the Subduction History of the Southwest Pacific using Coupled Plate Tectonic-Mantle Convection Models

Friday, 19 December 2014
Kara J Matthews1, Nicolas E Flament1, Simon Williams1, Dietmar Müller1 and Michael Gurnis2, (1)University of Sydney, Sydney, NSW, Australia, (2)Caltech, Pasadena, CA, United States
Abstract:
The Late Cretaceous to mid Eocene (~85-45 Ma) evolution of the southwest Pacific has been the subject of starkly contrasting plate reconstruction models, reflecting sparse and ambiguous data. Disparate models of (1) west-dipping subduction and back-arc basin opening to the east of the Lord Howe Rise, (2) east-dipping subduction and back-arc basin closure to the east of the Lord Howe Rise, and (3) tectonic quiescence with no subduction have all been proposed for this time frame. To help resolve this long-standing problem we test a new southwest Pacific reconstruction using global mantle flow models with imposed plate motions. The kinematic model incorporates east to northeast directed rollback of a west-dipping subduction zone between 85 and 55 Ma, accommodating opening of the South Loyalty back-arc basin to the east of New Caledonia. At 55 Ma there is a plate boundary reorganization in the region. West-dipping subduction and back-arc basin spreading end, and there is initiation of northeast dipping subduction within the back-arc basin. Consumption of South Loyalty Basin seafloor continues until 45 Ma, when obduction onto New Caledonia begins. West-dipping Tonga-Kermadec subduction initiates at this time at the relict Late Cretaceous-earliest Eocene subduction boundary. We use the 3D spherical mantle convection code CitcomS coupled to the plate reconstruction software GPlates, with plate motions and evolving plate boundaries imposed since 230 Ma. The predicted present-day mantle structure is compared to S- and P-wave seismic tomography models, which can be used to infer the presence of slab material in the mantle at locations where fast velocity anomalies are imaged. This workflow enables us to assess the forward-modeled subduction history of the region.