Analytical bias in the oldest section of the Dome C CO2 record

Thursday, 18 December 2014
Bernhard Bereiter1, Sarah Eggleston2, Jochen Schmitt2, Christoph Nehrbass-Ahles2, Thomas F Stocker2, Hubertus Fischer2, Sepp Kipfstuhl3 and Jerome Aime Chappellaz4, (1)Scripps Institution of Oceanography, La Jolla, CA, United States, (2)University of Bern, Bern, Switzerland, (3)Alfred Wegener Institute Helmholtz-Center for Polar and Marine Research Bremerhaven, Bremerhaven, Germany, (4)LGGE, St Martin d Heres, France
The EPICA Dome C (EDC) ice core has allowed to reconstruct atmospheric CO2 concentrations for the last 800,000 years. Here, we revisit the oldest part of the EDC CO2 record using different air extraction methods and sections of the core. For our standard system, we found an analytical artifact, which increases over the deepest 200 m and reaches 10.1±2.4 ppm in the oldest/deepest part. The mechanism behind is not unambiguously known yet, but it is related to insufficient gas extraction of our standard system, suboptimal storage of one section of the core, and structural changes of the ice with depth. Based on the re-measurements presented here we derive a correction for the original data to account for the analytical bias and revisit previous conclusions drawn from the biased data by Lüthi et al. (2008). First, the record low CO2 concentration ever found in an ice core in the period MIS 16 still remains as such, but requires slight correction by 2 ppm. Second, the period from MIS 16 to 19 as a whole does not show anymore unexpected low CO2 concentrations as expected from the CO2-temperature relation of MIS 1 to 15. Based on the corrected data of this study, this previous finding applies only for the periods MIS 16 and 17, but not for 18 and 19.