Miller, Kenneth G; Browning, James V; Schmelz, W John; Kopp, Robert E; Mountain, Gregory S; Wright, James D (2020): Smoothed Cenozoic sea-level relative to modern from deep-sea geochemical and continental margin records [dataset]. PANGAEA, https://doi.org/10.1594/PANGAEA.923139

We use published Pacific benthic foraminiferal oxygen isotope data and Mg/Ca records to derive a Cenozoic (66 Ma) global mean sea level (GMSL) estimate. This paper is novel in providing the first Pacific benthic foraminiferal oxygen isotopic splice for the entire Cenozoic, a detailed (Myr scale) sea-level record for the last 48 Ma based on the benthic foraminiferal oxygen isotopic and Mg/Ca approach (Mg/Ca records older than 48 Ma are uncertain). We use the 2012 Geological Time Scale (GTS), a 2-Myr smoothed paleotemperatures (Cramer et al., 2011) who used a low-pass filter that passes >80% of the amplitude for frequencies <0.5/Myr (wavelength >2 Myr), ramping down to <20% of the amplitude for frequencies >1.25/Myr (wavelength <0.8 Myr). We used equation 7b Cramer et al. (2011) and a simplified paleotemperature equation for benthic foraminifera T = 16.1– 4.76 [δ18Obenthic – (δ18Oseawater – 0.27)] to solve for oxygen isotopic changes of seawater. We assume that shorter term (<2 Myr) temperature changes comprise ~20% of the oxygen isotopic changes of seawater changes. The resultant oxygen isotopic changes of seawater estimate was scaled to GMSL changes using a revised seawater oxygen isotopes to sea-level calibration of 0.13‰/10 m of Winnick and Caves (2015). Because of temperature effects notable during peak Pleistocene interglacials, we iteratively fit the last interglacial cycle to known sea level during MIS5e and applied these temperatures (1.8°C) to major Middle to Late Pleistocene peak interglacials, tapering the temperature from the long term estimates for the peak interglacials using a Gaussian filter. We applied an empirically correction for carbonate ion change across the Eocene-Oligocene transition, to remove an apparent warming effect of ~1.5°C; we applied their empirical correction to the sea-level curve, reducing the amplitude by 28 meters from 34.17 to 34.30 Ma.

Sea level in has been obtained by interpolating to 20-ka intervals and using a 49-point Gaussian convolution filter, removing periods shorter than 490 ka.