1. Searching for “Old Ice” in Allan Hills, East Antarctica 

Fig 1. Schematics of the glaciological setting near Allan Hills Blue Ice Areas, modified from Whillans and Cassidy (1983).

Fig 1. Schematics of the glaciological setting near Allan Hills Blue Ice Areas, modified from Whillans and Cassidy (1983).

Ice cores from Antarctica record eight continuous glacial-interglacial cycles over the past 800 thousand years (ka; EPICA Members, 2004; Petit et al, 1999; Kawamura et al, 2007). Based on the composition of air trapped in polar ice cores and the isotopic composition of the ice, it has been established that Antarctic temperature, global ice volume, and the concentration of carbon dioxide (CO2) are tightly correlated on glacial-interglacial timescale.

Recent discoveries of 1 Ma ice from shallow (depth < 200 m) ice cores drilled in the Allan Hills Blue Ice Areas (BIA) demonstrate the potential to retrieve stratigraphically discontinuous old ice (Higgins et al, 2015). BIAs are regions where net ablation leads to the exposure of ancient glacial ice at the surface of an ice sheet. Here in Allan Hills BIAs, ancient ice is brought up to the surface by ice flow guided by rising bedrock topography (Fig 1; Whillans and Cassidy, 1983; Bintanja, 1999). This unique glaciological setting makes Allan Hills BIAs one of the well-studied meteorite concentration sites in Antarctica (Whillans and Cassidy, 1983). Intriguingly, one meteorite (ALH88019) collected in the Allan Hills BIAs has a terrestrial age of 2 million years (Scherer et al, 1997), hinting at the existence of ice older than 2 Ma in this area.

During the 2015-16 Antarctic field season, we drilled three ice cores in Allan Hills BIAs. Two cores (ALHIC1502 and ALHIC1503) contain ice with exceptionally old age (ALHIC1503 was drilled in the same borehole in Higgins et al, 2015). By precisely measuring the isotopic composition of argon (40Aratm) in the trapped air, we established the age of the ice and discovered ice as old as 2 Ma, with a single unreplicated sample dating back to 2.7±0.3 Ma. The dating method is based on the premise that 40Ar is produced by the radioactive decay of 40K, whereas 38Ar and 36Ar are stable through geologic time. Thus, the 40Ar/38Ar ratio will increase over time with a rate of 0.066±0.007‰ (Bender et al, 2008). We also measured CO2 and other gases in these unique samples, in which the oldest pristine CO2 samples date back to 1.5 Ma. The older ice samples are affected by respiratory CO2 from the bedrock.

 
Fig 2. Depth-age relationship of ALHIC1502 (blue) and ALHIC1503 (red) ice cores. Data in circles are measured at Princeton University (PU; Higgins  et al , 2015). New measurements made at Scripps Institution of Oceanography (SIO) are plotted in squares. Error bars represent the the external reproducibility (1σ; 110 kyr and 220 kyr for samples measured at SIO and PU, respectively) of the measurement or 10% of the sample age, whichever is greater. Grey bar highlights the section where &gt;1 Ma ice is present. In the right inset, Ar-ages are plotted against distance from the bedrock.

Fig 2. Depth-age relationship of ALHIC1502 (blue) and ALHIC1503 (red) ice cores. Data in circles are measured at Princeton University (PU; Higgins et al, 2015). New measurements made at Scripps Institution of Oceanography (SIO) are plotted in squares. Error bars represent the the external reproducibility (1σ; 110 kyr and 220 kyr for samples measured at SIO and PU, respectively) of the measurement or 10% of the sample age, whichever is greater. Grey bar highlights the section where >1 Ma ice is present. In the right inset, Ar-ages are plotted against distance from the bedrock.