Linking the stratigraphic record with mantle convection in time and space (Invited)

The influence of mantle convection on basin subsidence, the stratigraphic record, and sea level change has been discussed over the last decade and a number of connections made between the observational record and models. We are now making this linkage more explicit to both better constrain mantle processes and for use as a tool for the interpretation of vertical motions with sparse and incomplete observations. We provide an overview and discuss a number of concrete results that have emerged from our collaboration. One of the most important technical issues considered, but which is ignored in general, is that the stratigraphic record is attached to the moving plates and that plates differentially under go vertical motions as they move with respect to a changing dynamic topography field. We have embodied these ideas in a workflow using a linked paleogeographic-spherical mantle convection system (using the open source packages GPlates and CitcomS). Examples are taken from Australia, North America, and Antarctic-New Zealand. Using inferred paleoshorelines, we show that Australia tilted downward at long wavelengths by 300 m since the Eocene. In a high-resolution model of Australia embedded into a global model, we show that the tilting is consistent with motion toward the slabs in Melanesia but that an additional source of buoyancy beneath Antarctic is required to fit the total subsidence. For North America since the Cretaceous we have developed adjoint models in which we start from the present seismic structure beneath the continent, the robust Farallon slab in the lower mantle, and infer the initial condition and mantle parameters. The model that best fits an extensive set of tectonic subsidence curves and paleo-shorelines is consistent with the putative flat lying slab associated with the Laramide orogeny. Our prediction of a significant east-ward shift of a depocenter has been independently verified by a new high resolution set of tectonic subsidence curves from Utah to Colorado. The dynamic topography effect continued to shift eastward, such that today we predict that the east coast of the US should be dynamically subsiding, potentially consistent with the mismatch between relative sea level curves on the east coast and putative eustatic curves. With a forward approach, we have been able to integrate anomalous observations from the New Zealand-Antarctica conjugate margins including the large Ross Sea geoid low and 0.5-0.9 km of excess tectonic subsidence since the Cretaceous. The time-dependent topography and geoid provide powerful, complementary constrains and we find a strong mid-mantle upwelling associated with the geoid low.