Interpretation of ice core records – University of Copenhagen

29 September 2008

Interpretation of ice core records

I am trying to establish an overview of what the ice cores from around the world are believed to be telling us about the past climate. My question is how should I intepret ice core records? And . . . should I extrapolate the concept of temperature preceding carbon dioxide phenomena to today's carbon dioxide buildup. That is, will today's carbon dioxide buildup contribute to global warming or is nature running the show and carbon dioxide has little or nothing to do with global warming?

In a natural state with no human impact on greenhouse gasses etc., ice core evidence tell us that during abrupt climate warmings (transitions from glacial to interglacial conditions), temperature probably leads CO2 (see the note below).

Now, let's say that temperature starts to change. What initiates a warming is still much debated, but variations in how much energy different parts of the Earth's receive from the Sun (so‐called Milankovitch forcing) most certainly plays a role. So do the feedbacks on the climate system, of which you outline one. There are many other feedbacks in the climate system, of which some are well‐understood, and some are not. CO2 plays a role in many of these, and will help amplify the initial temperature change. So although the temperature starts to change first, CO2 is not just changing as a consequence of the temperature change, CO2 is an integral part of the dynamical system which is changing.

During past climate change, the temperature was the initial driver, and currently CO2 is. So we cannot directly use evidence from past climate change as analogues to future change, where mankind is forcing CO2 change. But we can use ice core studies (and studies of other records of past climate) to improve understanding of the dynamics of climate, thereby improving our ability to predict what the future will bring.

Note: The reason why I say "probably" is that it is not always easy to objectively find the point where something (temperature, CO2 etc) starts to change and because there is an ice core analysis difficulty, known as the "firn age ‐ gas age problem". The CO2 is measured in the bubbles in the ice, while the temperature is derived from measurements on the ice samples itself. In the upper layers of an ice cap, the snow is gradually transformed to snow under the overburden from later years' of snow. The upper layers are called "the firn", and the firn is permeable to air down to 50‐80 meters' depth. This means that todays' atmosphere is not locked up in the ice at surface, but at a depth of ~60 meters, surrounded by much older ice. In the driest parts of Antarctica (producing the longest ice cores records of close to a million years), this offset of bubble age and the age of the surrounding ice is several millennia. Researchers model and correct for this offset, but there is a significant uncertainty of the magnitude of the correction, leading to uncertainty in the assessment of the phasing of temperature and CO2 changes. However, based on the data we have today, it seems most likely that the temperature starts to changes prior to the CO2 concentration.

-- Sune O. Rasmussen, post doc