The issue of tectonic contamination upon geological inferences of relative sea level history is an important one. The issue arises on timescales that range from the 21-26 kyrs that have passed since the Last Glacial Maximum, to the most recent time when periods as warm as the present were expected to have existed - such as the mid-Pliocene. The coral based record from Barbados, for example, is known to be contaminated by continuing tectonic uplift of the island at a rate of approximately 0.34 mm/yr. For the Pliocene warm period at ~3 Myr, records from geological sites, such as the Orangeburg Scarp in North Carolina, have played a prominent role in arguments underpinning the design of the ongoing international PlioMIP program. In connection with the latter site, Rowley et al (2013) have recently argued that this record is contaminated by a tectonic imprint sufficiently strong to suggest that the usual inferences of Pliocene eustatic sea level based upon it (eg. Miller et al, 2012) must be seen as highly suspect.

Here we employ a tomographically constrained model of the mantle convection process to revisit the issue of tectonic imprint on relative sea level at the Orangeburg site, as well as other similar locations. Our analysis is based upon the inferred time dependence of dynamic topography forced by the mantle’s internal density heterogeneities delivered by the S20RTS seismic tomography model. We begin by comparing the static, present day dynamic topography predicted by the (linear) internal loading theory based on the formalism of Pari and Peltier (2000) with that predicted using the full nonlinear time-dependent mantle convection model of Shahnas and Peltier (2010, 2011). We demonstrate first that these two methodologies produce extremely similar results for the static field. We then proceed to run the nonlinear convection model in data assimilation mode while continuously nudging the internal density field back towards the structure inferred from tomography. Following a transient shock associated with the assimilation process, the model makes rather stable predictions for the time dependence of dynamic topography at a number of important locations from where data have been selected for the purpose of inferring the mid-Pliocene eustatic sea level. At Orangeburg where the inferred rates of tectonic uplift have ranged from 0.005 to 0.02 mm/yr (Dowsett and Cronin (1990), Soller (1989)) our model predicts an uplift rate of 0.024 mm/yr. This is sufficiently high to leave little room for any significant increase in eustatic sea level beyond what is expected to have existed as a consequence of the fact that the Greenland ice sheet had not yet been formed.

Dowsett and Cronin (1990), Geology, 18, 435-438

Miller et al (2012), Geology, 40, 407-410

Pari and Peltier (2000), J. Geophys. Res., 105, 5635-5662

Rowley et al (2013), Science, 340, 1560-1563

Shahnas and Peltier (2010), J. Geophys. Res., 115, B11

Shahnas and Peltier (2011), J. Geophys. Res., 116, B8

Soller (1989), USGS professional paper, 1466-A