STATISTICAL MODELING TO MEASURE THE INFLUENCE OF MANTLE DYNAMIC TOPOGRAPHY ON CENOZOIC MID-ATLANTIC MARGIN SEA-LEVEL RECORDS

Global mean sea-level (GMSL) estimates derived from benthic foraminiferal δ¹⁸O and Mg/Ca-based temperatures show statistical similarities with “backstripped” sea-level estimates obtained by progressively accounting for compaction, loading, and thermal subsidence from the U.S. Mid-Atlantic margin. We statistically modelled these data using a Bayesian hierarchical model that is comprised of component processes that are each modeled with a Gaussian process prior. The structure of the statistical model we applied reflects the presence of the ice-volume based GMSL signal in both the δ¹⁸O_benthic-Mg/Ca GMSL proxy and the backstripping data, while also capturing the global, regional, and local signals exclusive to the backstripping data. Statistically modeling the data in this way allows us to assess the geometric structure of the statistically modeled estimates of differences between spatially distinct subsets of the backstripping record and the δ¹⁸O_benthic-Mg/Ca GMSL proxy and the variation/motion of these differences through space and time. Analysis of the result shows that the long-term differences between the backstripping records onshore New Jersey (NJ), offshore NJ, and onshore south of NJ can be fit with a sinusoid and move in a spatially coherent manner across the margin through the Cenozoic at a rate and direction of motion generally opposite to that of the North American Plate. The sinusoid we fit to the data has a 10³ km wavelength, which is similar to those described for processes like mantle dynamic topography. We assert that the geometric structure and the vector of motion of the modeled topography fit to the statistically derived backstripping residuals quantify the effects of MDT on the Mid-Atlantic margin over the past 66 Myr. We estimate that the topographical effect might have altered the sea-level records produced by backstripping cores collected onshore New Jersey by up to 40 meters in the Eocene, and those produced by backstripping cores collected offshore New Jersey by up to 25 m in the Miocene. Our results preclude hundreds of meters of Mid-Atlantic dynamic topographical effect on sea-level records through the Cenozoic as predicted by some models of MDT.