The air in ancient ice is younger than the ice
An ice-core bubble and the ice surrounding it usually have different ages. Snow becomes solid ice gradually, while air continues to diffuse through connected pores for decades or centuries. Only when those pores close are gases sealed, so the trapped atmosphere is younger than its icy container.
Annual snowfall is compacted into porous firn. At first, wind and molecular diffusion keep the pore air connected to the atmosphere. Deeper down, channels pinch off into isolated bubbles. The ice at that depth was deposited earlier than the mean age of the sealed gas. At Summit, Greenland, one study estimated an ice–gas age difference of about 210 years under then-current conditions; at NEEM, modelling yielded about 182 years. Colder, low-accumulation sites can have much larger offsets and broader age distributions.
Ice cores place greenhouse gases beside temperature and other climate proxies, but ‘beside’ is not automatically ‘same date’. Researchers must model gas transport and align separate timescales before asking which change led. A physical sample can contain two clocks because its components stopped exchanging with the world at different times.
Before close-off, gases diffuse at different rates and pore air mixes across depths. A sealed bubble therefore samples an age distribution, not a photograph of one day. Researchers use gases with known atmospheric histories, isotopic fractionation and firn density to tune transport models. Precision depends on how well those models describe diffusion, advection and the lock-in zone.
During abrupt climate transitions, a difference of centuries can reverse an apparent lead or lag between temperature and carbon dioxide. Ice temperature proxies belong to the ice timescale; enclosed gases belong to the gas timescale. Aligning them is an inference with uncertainty. That does not make the archive unreliable; it makes the archive’s formation process part of the evidence.
Could drilling or modern air contamination create the apparent younger gas?
Contamination is tested through duplicate samples, gas concentrations, isotopes and drilling controls. The age offset is predicted before drilling by compaction and diffusion, observed in open firn-air profiles, and reproduced across sites. Contamination can affect individual measurements, but it cannot explain the systematic transition from connected pores to sealed bubbles.
A site where gases sealed at snowfall yet retained an open, diffusive firn column would contradict the mechanism; present observations show those conditions are incompatible.