The Magnitude of Surface Ocean Acidification and Carbon Release During Eocene Thermal Maximum 2 (ETM‐2) and the Paleocene‐Eocene Thermal Maximum (PETM)

Abstract

Eocene Thermal Maximum 2 (ETM‐2; 54.1 Ma) was the second largest Eocene hyperthermal. Like the Paleocene‐Eocene Thermal Maximum (PETM), ETM‐2 was characterized by massive carbon emissions and several degrees of global warming and thus can serve as a case study for assessing the impacts of rapid CO2 emissions on ocean carbonate chemistry, biota, and climate. Marine carbonate records of ETM‐2 are better preserved than those of the PETM due to more subdued carbonate dissolution. As yet, however, the magnitude of this carbon cycle perturbation has not been well constrained. Here, we present the first records of surface ocean acidification for ETM‐2, based on stable boron isotope records in mixed‐layer planktic foraminifera from two midlatitude ODP sites (1210 in the North Pacific and 1265 in the SE Atlantic), which indicate conservative minimum global sea surface acidification of −0.20 +0.12/−0.13 pH units. Using these estimates of pH and temperature as constraints on carbon cycle model simulations, we conclude that the total mass of C, released over a period of 15 to 25 kyr during ETM‐2, likely ranged from 2,600 to 3,800 Gt C, which is greater than previously estimated on the basis of other observations (i.e., stable carbon isotopes and carbonate compensation depth) alone.

Proxy evidence for state-dependence of climate sensitivity in the Eocene greenhouse

“…2 , Supplementary Fig. 1 ) overlaps with existing low-resolution δ 11 B -based records from Tanzania 4 , 5 , and records from the Middle Eocene Climatic Optimum (MECO; ~40.1–40.5 Ma) 11 , Eocene Thermal Maximum 2 (ETM2; 54.1 Ma) 9 , and the Paleocene-Eocene Thermal Maximum (PETM; ~56 Ma) 9 , 10 (all re-calculated for consistency, see “Methods” and Supplementary Data 3 ), and demonstrates the validity of our multi-species treatment of δ 11 B in deriving mixed-layer pH and CO 2 concentrations. This continuous view of the evolution of CO 2 confirms that the highest CO 2 levels, outside of the short-lived increase in CO 2 at the PETM 9 , 10 , 17 , occurred during the Early Eocene Climatic Optimum (EECO; 49–53 Ma 18 ).…”

“…900 ± 100 ppm (±2 se, n = 14) 9 , 10 . For the EECO and the PETM 9 , 10 , the average CO 2 , calculated using the average δ 11 B and Mg/Ca-temperature estimates in each interval is 1470 (+360/−300) ppm (2 s.d.) and 1790 (+560/−380) ppm respectively (or 1980 (+510/−440) ppm and 2470 (+690/−540) (2 s.d.)…”

“…Identification of planktonic foraminiferal depth habitats used in this study are based on relationships between stable isotope foraminiferal geochemistry and ecology in relationship to δ 11 B offsets (e.g., refs. 4,9,10 , and references therein). Additional foraminifera species used here (Morozovella aragonensis, Acarinina quetra, A. pentacamerata, M. crater, A. cuneicamerata, A. pseudosubsphaerica) were cross-calibrated against previously known species (A. pseudotopilensis, A. praetopilensis, A. soldadoensis, Guembelitrioides nuttalli, Pearsonites broedermanni) for their δ 11 B behaviour collected from the same time interval and core site 4,5,11,17 and site-specific species offsets in δ 11 B were not identified.…”

“…Therefore, we excluded the M. velascoensis record of the PETM 17 , since this species is randomly offset from our tested species A. soldadoensis when comparing five samples from site 1209 and at similar ages (Δδ 11 B M.velascoensis – A. soldadoensis = 0.8 ± 0.6‰ (2 s.d.) 9 . Also, both G. index and G. kugleri records of the MECO 11 are excluded, because the former showed variable habitat depth and δ 11 B offsets in TDP 4 , and the latter is not sufficiently tested for within site inter species offsets.…”

“…1 ). This record, coupled to existing δ 11 B-CO 2 reconstructions 4 , 5 , 9 – 11 and novel Global Mean Temperature (GMT) estimates, is used to provide proxy evidence of the state dependency in climate sensitivity, with higher sensitivity during the warm period of the early Eocene, and lower towards the transition to the colder, late Eocene.

Fig.

…”