| Summary: | Accurate absolute chronologies of ice cores are crucial to interpret paleoclimatic records and to decipher their relationship to external and internal forcings. The chronology of the deeper part of Antarctic ice cores is mainly constrained by orbital age markers inferred from tracers measured in air trapped in ice. In particular, age markers deduced from the alignment of the Total Air Content of the ice (TAC) on a local integrated summer insolation curve were used to constrain the AICC2012 chronology of the EPICA Dome C (EDC) ice core over the past 440 ka. However, a full understanding of the physical processes influencing TAC variations is still missing together with a quantitative estimate of the age uncertainties attached to TAC-based age makers. Here, we present new TAC measurements on EDC at a ~2 ka resolution extending the published record back to 800 ka. First, we investigate the imprint of orbital periodicities in the 800 ka TAC record using spectral analyses. We identify that the last 500 ka of the record is characterised by a dominant period related to obliquity but its strength decreases significantly in the oldest part. Second, we investigate the most relevant orbital tuning target to infer age markers based on a comparison between the TAC record and a selection of local summer insolation curves, and our understanding of how the insolation signal could get imprinted in TAC. Third, we quantify the age errors attached to TAC-based markers considering sources of uncertainties related to (i) the choice of the orbital target, (i) the filtering band for TAC, (iii) the record alignment method and (iv) the loss of visual resemblance between variations in TAC and its orbital target. Age uncertainties vary from ~2 ka to more than 6 ka. The comparison of a TAC-based chronology with the AICC2012 chronology over the last 600 ka suggests ages differences of 1.4 ka on average and they vary within the stated AICC2012 age uncertainties. Finally, we compare the TAC dating constraints with those deduced from O2/N2 and ...
|
|---|