Using an inverse mantle convection model that assimilates seismic structure and plate motions, we reconstruct Farallon plate subduction back to 100 million years ago. Models consistent with stratigraphy constrain the depth dependence of mantle viscosity and buoyancy, requiring that the Farallon slab was flat lying in the Late Cretaceous, consistent with geological reconstructions. The simulation predicts that an extensive zone of shallow-dipping subduction extended beyond the flat-lying slab farther east and north by up to 1000 kilometers. The limited region of flat subduction is consistent with the notion that subduction of an oceanic plateau caused the slab to flatten. The results imply that seismic images of the current mantle provide more constraints on past tectonic events than previously recognized.

No abstract

A B S T R A C TDynamic earth models are used to better understand the impact of mantle dynamics on the vertical motion of continents and regional and global sea level change since the Late Cretaceous. A hybrid approach combines inverse and forward models of mantle convection and accounts for the principal contributors to long-term sea level change: the evolving distribution of ocean floor age, dynamic topography in oceanic and continental regions, and the geoid. We infer the relative importance of dynamic versus other factors of sea level change, determine time-dependent patterns of dynamic subsidence and uplift of continents, and derive a sea level curve.We find that both dynamic factors and the evolving distribution of sea floor age are important in controlling sea level. We track the movement of continents over large-scale dynamic topography by consistently mapping between mantle and plate frames of reference, and we find that this movement results in dynamic subsidence and uplift of continents. The amplitude of dynamic topography in continental regions is larger than global sea level in several regions and periods, so that it has controlled regional sea level in North and South America and Australia since the Late Cretaceous, northern Africa and Arabia since the late Eocene, and Southeast Asia in the Oligocene-Miocene. Eastern and southern Africa have experienced dynamic uplift over the last 20 to 30 m.y., whereas Siberia and Australia have experienced Cenozoic tilting. The dominant factor controlling global sea level is a changing oceanic lithosphere production that has resulted in a large amplitude sea level fall since the Late Cretaceous, with dynamic topography offsetting this fall. returned to Caltech where he is currently John E. and Hazel S. Smits Professor of Geophysics and director of the Seismological Laboratory. His effort is divided between linking geodynamics and the rock record (especially stratigraphy), the basic physics of geophysical phenomena, studies of the deep interior of the Earth, and the development of new computational methods, including the GPlates package for plate tectonic modeling.

[1] We apply adjoint models of mantle convection to North America since the Late Cretaceous. The present-day mantle structure is constrained by seismic tomography and the time-dependent evolution by plate motions and stratigraphic data (paleoshorelines, borehole tectonic subsidence, and sediment isopachs). We infer values of average upper and lower mantle viscosities, provide a synthesis of North American vertical motions (relative sea level) from the Late Cretaceous to the present, and reconstruct the geometry of the Farallon slab back to the Late Cretaceous. In order to fit Late Cretaceous marine inundation and borehole subsidence, the adjoint model requires a viscosity ratio across 660 km discontinuity of 15:1 (reference viscosity of 10 21 Pa s), which is consistent with values previously inferred by postglacial rebound studies. The dynamic topography associated with subduction of the Farallon slab is localized in western North America over Late Cretaceous, representing the primary factor controlling the widespread flooding. The east coast of the United States is not stable; rather, it has been experiencing continuous dynamic subsidence over the Cenozoic, coincident with an overall eustatic fall, explaining a discrepancy between sea level derived from the New Jersey coastal plain and global curves. The east coast subsidence further constrains the mantle viscosity structure and requires an uppermost mantle viscosity of 10 20 Pa s. Imposed constraints require that the Farallon slab was flat lying during Late Cretaceous, with an extensive zone of shallow dipping Farallon subduction extending beyond the flat-lying slab farther east and north by up to 1000 km than previously suggested.

[1] The dynamic subsidence of the United States east coast is addressed using the discrepancy between regional and global estimates of sea level, elevation of paleoshorelines, and adjoint models of mantle convection that assimilate plate motions and seismic tomography. The positions of Eocene and Miocene paleoshorelines are lower than predicted by global sea levels, suggesting at least 50 m, and possibly as much as 200 m of subsidence since the end of the Eocene. Dynamic models predict subsidence of the east coast since the end of Eocene, although the exact magnitude is uncertain. This subsidence has been occurring during an overall global sea-level fall, with the eustatic change being larger than the dynamic subsidence; this results in a regional sea-level fall in the absence of land subsidence. Dynamic subsidence is consistent with the difference between eustasy and regional sea level at the New Jersey coastal plain. Citation: Spasojević, S., L. Liu, M. Gurnis, and R. D. Müller (2008), The case for dynamic subsidence of the U.S. east coast since the Eocene, Geophys.

West Antarctica and adjacent seafl oor have topographic elevations 0.5-1.2 km greater than expected from models of lithospheric age and crustal structure. Ocean crust near New Zealand has no equivalent depth anomaly, but tectonic subsidence histories from Campbell Plateau petroleum wells show anomalously high subsidence rates during the Paleogene, and total subsidence 0.5-0.9 km greater than expected from rift basin models. Geophysical and geochemical anomalies suggest that upward mantle fl ow supports the anomalous topography beneath Antarctica, and the Campbell Plateau subsidence history indicates that topographic support mechanisms were long lived (>80 Ma) and recoverable over a period of ~30 m.y. as plate motions moved New Zealand from Antarctica. We construct models of Late Cretaceous and Cenozoic mantle fl ow with a slab graveyard and upwelling above that is initially rooted at 1000-1500 km depth. Our models match topography and subsidence history anomalies, and are consistent with mantle seismic wave speed anomalies and the geoid. We suggest that when thermally driven slab downwelling ceased ca. 100 Ma, low-density material that was fertilized within a broad zone in the lower mantle during the previous ~400 m.y. of Gondwana subduction was released and able to rise. Mantle upwelling from depths of 700-1500 km, lasting for periods of ~100-200 m.y., with enriched chemistry related to the prior subduction history may be a general process that follows subduction death, and has not previously been recognized.

Supplementary Figure S1. Details of velocity structure for models S20RTS 1 (first column) and SB4L18 2 (second column), and model TX2005 3 (third column) in the zones of geoid minima. (a) Integrated tomography in depth range 2050-2850 km; (b) Integrated tomography in depth range 300-1000 km; (c) Cross section from North America to Ross Sea; (d) Cross section from central Asia to Ross Sea; (e) Correlation coefficient between 18 observed geoid and tomography calculated at every 100 km depth; (f) Observed geoid. Semi-transparent outlines on (a-b) cover zone of global geoid high; blue and red dashed lines (c-d) indicate lower mantle high velocity and upper-to-mid mantle low velocity 21 anomalies of interest, respectively; black, blue and red lines on (e) show mean correlation 22 coefficient between whole geoid, areas of negative geoid and areas of positive geoid, respectively; red lines on (f) indicate position of great circle cross-sections intersecting geoid low. Correlation coefficient between observed geoid and tomography models (e) is calculated as a product of values divided by product of standard deviations at every 100 km depth. The mean value of correlation coefficient is than calculated for whole geoid, areas of geoid lows, and geoid highs. 28 Integrated tomography plots (a-b) show a correlation between geoid low (f) and both 30 seismically fast regions in lower mantle (a) and slow regions in upper-to-mid mantle (b) for all investigated tomography models. In the region of NE Pacific geoid low, all models define a seismically fast anomaly in the lower 500-1000 km of mantle (c), which is 33 probably related to an ancient subducted slab that hasn't been recognized previously, 34 except by analysis of SKS-SKKS splitting discrepancies 4. A zone of upper-mid mantle 35 seismically slow velocities is located above this fast anomaly (c), model SB4L18 2 defines 2 nature geoscience | www.nature.com/naturegeoscience

[1] We show that time-dependent models of mantle upwellings above a cold downwelling in the New Zealand-Antarctica region since 80 Ma can explain anomalous geophysical observations: ∼1.0 km of positive residual bathymetry at the Antarctica margin, a large Ross Sea geoid low, 0.5-0.9 km of excess tectonic subsidence of the Campbell Plateau since 80 Ma, and several seismic wave speed anomalies. Model results indicate that the largest mantle upwelling, centered in the Ross Sea, has an average temperature anomaly of 200°C and density anomaly of 0.6%, and it rose from midmantle depths at 80 Ma to a present depth of 400-1000 km. Anomalous Campbell Plateau subsidence requires a smaller hot anomaly evolving within the upper mantle under the region of the reconstructed Late Cretaceous Campbell Plateau. The excess subsidence of the plateau results from northward drift of New Zealand away from the dynamic topography high created by the smaller hot anomaly. To fit present-day geoid and residual topography observations, we require a large lower:upper mantle viscosity ratio of 100:1. We suggest that the distribution of temperature and viscosity is related to long-lived Gondwana subduction that accumulated high-density, high-viscosity lower mantle below a chemically altered upper mantle with anomalously low density and/or high temperature. Time-dependent observations enable constraints on absolute viscosities of 10 23 Pa s and 10 21 Pa s for the lower and upper mantle, respectively.Citation: Spasojevic, S., M. Gurnis, and R. Sutherland (2010), Inferring mantle properties with an evolving dynamic model of the Antarctica-New Zealand region from the Late Cretaceous,