Global plate kinematic reconstructions and mantle flow models incorporating lithospheric deformation

The effect of mantle flow on surface topography has been the subject of considerable interest over the last few years. A common approach to the problem is to link plate tectonic reconstructions and global geodynamic models. An important limitation of this approach is that traditional plate tectonic reconstructions do not take the deformation of the lithosphere into account. Plates are represented as rigid blocks, resulting in continental overlap in full-fit reconstructions. Models that use topological polygons avoid continental overlaps, but plate velocities are still derived on the basis of Euler poles for rigid blocks. Our objective is to develop quantitative models of surface plate kinematics that include areas of deforming continental crust. We generated a series of global reconstructions including deforming plates in key areas, derived using tools developed within the open source reconstruction software GPlates. We used geological and geophysical data to define the areal and temporal extent of major crustal deformation phases. For convergent plate boundaries, we incorporated quantitative estimates of deformation to model the time-varying geometries of subduction zones such as the Andean margin of South America and east of the Lord Howe Rise. In reconstructions of continental breakup, for example between Australia and Antarctica or the opening of the South and Equatorial Atlantic, the timing and the intensity of continental extension is imposed by the progressive, diachronous breakup and initiation of seafloor spreading for each major margin system. We used these models as a boundary condition for geodynamic models. For each deforming domain, a topological mesh was generated such that surface velocity fields within the deforming regions are calculated by linear interpolation from velocities at the boundaries and from additional constraining points within the deforming regions. The velocity field derived from the plate reconstructions were used as a time-dependent surface boundary condition in mantle convection models that include compositionally distinct crust and continental lithosphere embedded within the thermal lithosphere. We computed forward global mantle flow models using 3D-spherical finite-element code CitcomS to simultaneously quantify the relative contributions of lithospheric stretching and deep mantle flow to the subsidence of passive margins. This workflow allowed us to investigate the interaction between mantle flow and lithospheric stretching and their relative contributions to surface topography in specific passive margins systems, and compare them to observations. In the South Atlantic, the anomalously deep Argentine Basin contrasts with the Elevated Passive Continental Margins (EPCMs) of Northeast Brazil and Southern Africa. Our models reproduce this first-order topographic asymmetry of the margins of the South Atlantic Ocean. We attribute the large subsidence of the Argentine margin to the dynamic topography low induced by ongoing subduction along the narrow southern portion of South America and suggest that part of the uplift of Southern Africa can be attributed to its relative motion away from a dynamic topography low. The Brazilian Highlands are also predicted to lie over a dynamic topographic high. These results underline the importance of dynamic topography to the total subsidence at passive margins.