I don’t see anyone else blogging this, so I will, even though it is old news. Record-shattering 2.7-million-year-old ice core reveals start of the ice ages says Science, but you have to immeadiately know the caveat: it isn’t a continuous 2.7 Myr core. In fact, before long you discover that it’s kinda the same thing as Atmospheric composition 1 million years ago from blue ice in the Allan Hills, Antarctica from PNAS, 2015, which I also unaccountably failed to blog or see blogged.
So, stepping back: rather than take a core vertically through relatively undisturbed ice, which is what Vostok or EPICA did, these folks have deliberately selected an area where the ice is flowing, bringing old ice near to the surface. And as it happens they drilled horizontally. So, you kinda get potluck for whatever you pull out. And as you can see from the figure (from the 1 Myr 2015 paper) it is much less fun to interpret (if it isn’t obvious: what you get is the box with the rounded edges, and the dots inside): their Argon dating only gives them age to within ~100 kyr; and they can’t pull out a continuous record, they just get a range.
1 Myr is only just outside the previous 800 kyr. 2.7 Myr is older; still, you’ll notice that the Science article, while gushing, doesn’t actually say much. Neither does their conference abstract: The old ice can be binned into three age groups: 1 Ma, 1.5 Ma, and 2+ Ma, disturbed by layers of ≤800 ka ice. This age-depth relationship indicates large-scale disturbance in the ice stratigraphy, reinforcing the concept of climate snapshots instead of time-series. Three climate proxies (Xe/Kr, δDice, and pCH4) fall within the range of variations in the recent 100-kyr glacial cycles, but with reduced variability. Hopefully, there will be more later.