How an ice core traps pieces of ancient atmosphere

In 1958, scientists working off Greenland hauled pieces of iceberg aboard a ship. The ice was cumbersome; the real prize was almost weightless. Inside were small pockets of compressed gas. The team released that gas and measured its carbon dioxide, oxygen and nitrogen. Their published account describes the shipboard analyses.

They were investigating a remarkable possibility: ice could keep a sample of air from long before anyone had instruments to measure it.

An ice core, a cylinder drilled from a glacier or ice sheet, can bring that air into a laboratory. But the atmosphere is not usually captured the moment a snowflake lands. The container takes shape slowly, as snow is buried and its connected air spaces close.

Snow begins with open doors

Fresh snow is a framework of ice grains with air between them. On a cold ice sheet where snow survives from year to year, new snowfall buries the older layers. The growing weight presses the grains closer together.

The buried snow also changes its crystal structure. Grains grow and join, leaving less empty space. This older, compacted snow is called firn. It is denser than fresh snow but still contains a network of passages through which air can move. This transformation does not require the snow to melt into water and freeze again.

Think of an open sponge gradually becoming a sealed container. The useful part of that comparison is the changing connection to outside air. The limit is that snow is not rubber being squeezed by a hand: its ice grains change shape and bond as burial continues.

Eventually, passages narrow and pinch shut. Isolated pockets become bubbles surrounded by ice. Researchers call this bubble close-off. It happens over a zone, not at one perfectly level boundary. Measurements of open and closed spaces in firn show why the distinction matters: a pore can contain air yet remain connected to the atmosphere.

The ice and its air keep different dates

Before sealing, air can continue moving through the firn. Wind mixes the upper layers, while the random motion of molecules spreads gases through deeper passages. This spreading is called diffusion. The spaces therefore do not simply preserve the air that arrived with the snow.

By the time the bubbles close, the surrounding ice has already spent years being buried. The trapped air is consequently younger than the ice enclosing it. Depending on the site and past conditions, that difference can range from tens to thousands of years. The South Pole ice-core chronology explains the two separate timescales.

That distinction matters when comparing greenhouse gases in bubbles with temperature clues in the ice itself. Giving both measurements the same date merely because they came from the same depth would misalign the records.

Nor is one sample a breath from one exact afternoon. Air moves through the firn at different rates, and bubbles seal progressively. A sample combines gas with a spread of ages. That mixing smooths changes in atmospheric composition. Research on low-snowfall Antarctic sites examines this smoothing. A brief atmospheric change can therefore appear less sharp in the recovered record.

Bringing a sealed sample home

Researchers choose drilling sites according to the history they want to recover. Snowfall, temperature, melting and ice movement all affect what survives. A hollow drill removes successive cylinders, which are recorded, kept cold and examined. The NASA introduction to ice-core work describes this journey from drilling to cold laboratory.

Depth alone is not a calendar. Scientists use preserved annual patterns where these can be identified, compare distinctive events between cores, and use models and other age constraints. The gas record then needs its own dating, accounting for the delay before sealing.

The laboratory must release the ancient air without adding a misleading amount of modern air. In one established method, a prepared ice sample goes into an evacuated metal container. Mechanical shaking rubs it against a perforated cylinder, rather like a cheese grater, opening bubbles while the ice remains frozen.

Instruments then measure the proportions of gases in the extracted air. Reference gases of known composition and tests using bubble-free ice help reveal contamination or changes introduced by the procedure. The Law Dome study documents the extraction controls. These are measurements of tiny gas samples, not estimates made from a bubble's appearance.

Why believe the air survived?

The early investigators did not begin with a finished climate record. Per Scholander's work on organisms surviving cold conditions led him to study how slowly gases move through solid ice. In 1956, he and colleagues proposed that glacier ice could preserve old air for millennia. A historical review traces the experiments behind that idea.

Later confidence came from comparisons, not simply from trusting ice to be a perfect bottle. Different cores and laboratories recovered consistent histories within their uncertainties. At Antarctica's Law Dome, rapid snowfall has sealed air recent enough to overlap with direct atmospheric measurements. That overlap lets researchers check whether the old-air method joins the instrumental record sensibly.

The method also opened a much longer window. In 2008, researchers reported carbon dioxide measurements extending the EPICA Dome C record to 800,000 years before present. At that time, the extension allowed comparisons across additional cycles of glacial climate.

A powerful archive, not a perfect bottle

Melting can move water through older layers and change the gas record. Impurities and chemical reactions can also affect particular gases. A review of melt-affected ice cores sets out these complications. Researchers therefore assess the site, sample and gas rather than treating every piece of ice as equally trustworthy.

Ice flow can stretch, thin or fold layers. Dating and gas-age uncertainty limit how precisely events can be ordered. Measuring past carbon dioxide also does not, by itself, identify every process that changed it.

What the method provides is more direct and more extraordinary: actual surviving gas, accompanied by a testable account of how it was trapped. Snow left its doors open; burial gradually closed them. A drill and a careful laboratory can open them again.