Dating an 800,000 year Antarctic ice core record using the isotopic composition of trapped air

Here we measure the isotopic composition of air trapped in the European Project for Ice Coring in Antarctica Dome C (EDC) ice core, and use this geochemical information to improve the ice core agescale and our understanding of air enclosure processes. A first result is the detection of a flow anomaly in the bottom 500m of the EDC ice core using the δ18O of atmospheric oxygen (noted δ18Oatm). By tuning the measured δ18Oatm to the orbital precession signal, we correct the EDC agescale over 400-800 ka for flow-induced distortions in the duration of events. Uncertainty in δ 18Oatm phasing with respect to precession limits the accuracy of the tuned agescale to ±6 ka. We use this improved agescale to date two 10Be peaks detected in the EDC ice core and associated with the Matuyama-Brunhes geomagnetic boundary. While the ice age of the "precursor" event agrees within uncertainty with the age of radioisotopically dated lavas, the volcanic age for the younger reversal is approximately 10 ka older than the mid-point of the 10 Be peak in the ice. Since 80% of the lavas recording the Matuyama-Brunhes reversal are located in the Central Pacific, the observed age difference may indicate that the magnetic field orientation at this location changed prior to the dipole intensity minimum recorded by the ice core 10Be, as suggested by recent geodynamo modeling. A particular challenge for ice core dating is accurately accounting for the age difference between the trapped air and surrounding ice. This gas age - ice age difference (noted Δage) depends on the age of the ice at the bottom of the firn. δ15N of N2 is constant in the atmosphere over the timescales considered here, so any deviation from atmospheric composition reflects fractionation processes in the firn. We show that δ15N is positively correlated with the ice deuterium content, a proxy for temperature, over the entire EDC record, and propose an accumulation-permeability-convection mechanism. While temporal resolution and noise in the available data limit our ability to constrain glacial Δage, these data suggest that δ15N may be used as a gas-phase climate proxy at EDC