n.Newfeed

History · 6 min read

How Ice Cores Became Records of Past Atmospheres

How drilling, gas analysis, isotope measurement, and dating turned polar ice into a climate archive.

How Ice Cores Became Records of Past Atmospheres

An ice core is a cylinder drilled from a glacier or ice sheet. Its scientific value comes from several records occupying the same material. Snow preserves ratios of stable oxygen and hydrogen isotopes related to past temperature and moisture transport. Dust, sea salt, sulfate, ash, and human pollutants arrive at the surface and become buried. Most unusually, bubbles close within compacting snow and retain small samples of ancient air.

The result is not a year-by-year thermometer sealed without disturbance. Researchers must date the ice, account for gas being younger than the surrounding ice, model flow and compression, and distinguish local snowfall conditions from wider climate. Ice cores became authoritative archives through decades of drilling, laboratory improvement, cross-checking, and comparison among sites—not because the layers interpret themselves.

From shallow samples to deep polar drilling

Scientists had examined glacier layers and trapped bubbles before deep coring became practical. Twentieth-century electromechanical drills made it possible to recover long, continuous cylinders rather than fragments from pits and tunnels. During the International Geophysical Year of 1957–58, polar research programs expanded logistics and geophysical measurements in Greenland and Antarctica.

In Greenland, the Camp Century project recovered a core to bedrock in 1966. Willi Dansgaard and colleagues used oxygen-isotope measurements to infer large climatic variations through the last glacial period. The interpretation depended on a physical relation: evaporation, condensation, and temperature fractionate water molecules containing different isotopes. That relation is strong enough to make isotope ratios a climate proxy, but its calibration varies with site and atmospheric circulation.

Later Greenland programs—including Dye 3, the Greenland Ice Sheet Project 2, and the Greenland Ice Core Project—improved resolution and replication. Their records revealed abrupt shifts during the last glacial period, now called Dansgaard-Oeschger events. Agreement between independently drilled cores showed that many changes were not artifacts of a single damaged sequence. Differences near the base also demonstrated where ice flow had folded or disturbed old layers.

Antarctica offered slower accumulation and a longer record. Soviet and later Russian, French, and American teams drilled at Vostok Station. A 1999 synthesis led by Jean-Robert Petit presented about 420,000 years of climate and atmospheric history. It showed carbon dioxide and methane varying with Antarctic temperature across four glacial cycles. The curves were closely related, but their alignment alone did not assign a single direction of causation. Orbital forcing, ocean circulation, ice sheets, greenhouse gases, and feedbacks interact over glacial cycles.

Air and ice have different clocks

Fresh snow contains connected pore spaces. As more snow accumulates, grains compact into firn, an intermediate material denser than snow but still permeable. Air continues to mix through much of the firn column. Only deeper down do pores pinch off into isolated bubbles. The trapped gas is therefore younger than the ice around it.

Researchers call the difference delta age. Estimating it requires models of firn densification informed by accumulation rate and temperature, as well as measurements of gases with known atmospheric histories. At a high-accumulation Greenland site, closure occurs relatively quickly. At a cold, low-accumulation Antarctic site, the difference can span much longer. Reports that compare a gas change with a temperature proxy must place both on a common chronology and carry the age uncertainty.

Annual layers can be counted where accumulation is sufficient. Seasonal variations in isotopes, dust, chemistry, or electrical conductivity provide markers. Volcanic eruptions leave sulfate or ash horizons that can be matched among cores and sometimes to historically dated events. Farther back, thinning makes annual layers too narrow, and researchers rely more on ice-flow models and orbital or stratigraphic tie points.

This layered dating is one reason cores are more than simple depth records. Deeper usually means older, but not at a constant rate. Ice deforms and flows toward the margins. Near bedrock, folding, melting, or mixing can reverse or obscure the sequence.

Laboratory techniques expanded the archive

Early studies consumed relatively large samples. Modern continuous-flow analysis can melt a core progressively and measure chemical species at fine resolution. Laser spectroscopy measures greenhouse-gas concentrations and isotopic composition. Mass spectrometry resolves stable isotopes. Microscopy and particle counters characterize dust. Each technique has blanks, calibration standards, and contamination controls, especially important for trace concentrations.

The air bubbles provide a direct sample of past atmospheric composition, though altered by enclosure and extraction processes that scientists test. Carbon dioxide measured in bubbles is not inferred from a biological proxy in the same way as temperature. It is ancient air, with corrections and age modeling. That distinction underlies the importance of the records for greenhouse-gas history.

Other components are more indirect. Oxygen isotopes in snowfall reflect temperature but also the moisture source, path, elevation, and seasonality. Dust can reflect aridity, wind strength, source area, and transport. Sulfate can indicate volcanic deposition, yet connecting a peak to a particular eruption may require geochemical matching. Multiple indicators reduce ambiguity.

EPICA extended and replicated the glacial record

The European Project for Ice Coring in Antarctica coordinated drilling at Dome C and Dronning Maud Land. At Dome C, the team reached bedrock after drilling more than three kilometers. A 2004 Nature paper reported a climate record spanning roughly 740,000 years; subsequent measurements extended the usable record to about 800,000 years.

EPICA showed eight glacial cycles and an important change in their character. Earlier cycles in the record had lower amplitude than the four most recent ones. Greenhouse-gas measurements found that Antarctic temperature, carbon dioxide, and methane covaried over these long cycles. Results from different laboratories and comparison with marine sediments, orbital calculations, and other archives strengthened the chronology.

The core also exposed limits. Dome C has excellent age reach but lower snowfall than many Greenland sites, so a given year occupies less ice. Greenland cores provide finer detail for some abrupt Northern Hemisphere changes but do not extend as far continuously. There is no single best core; site selection trades time span, resolution, preservation, and geographic relevance.

What the archive established—and what it did not

Ice cores established that atmospheric carbon dioxide and methane changed substantially with glacial cycles long before direct monitoring began. They documented abrupt climate changes, volcanic fallout, dust variation, and industrial-era increases in several pollutants. Because trapped air and climate proxies appear in the same core, their relative timing can be studied more closely than when archives are widely separated.

They did not show that carbon dioxide was the initial trigger of each deglaciation. Orbital changes redistribute sunlight and can initiate regional warming; greenhouse gases then act as global feedbacks. Nor does a lag at one location negate the radiative effect of carbon dioxide. Timing questions must distinguish Antarctic local temperature, global mean temperature, ocean carbon release, and the age models used.

The search for older ice aims to reach beyond 800,000 years and examine the transition from roughly 41,000-year to 100,000-year glacial pacing. Projects such as Beyond EPICA have targeted very old, slow-moving Antarctic ice. Basal material may be old without being orderly, so age alone is insufficient; researchers need a sequence whose stratigraphy can be reconstructed.

What remains unknown includes how faithfully a small number of polar sites represent regional processes, how some abrupt changes propagated, and whether the oldest recoverable ice will preserve an undisturbed atmospheric sequence. The history of ice coring is a history of making those uncertainties measurable. A frozen cylinder becomes an archive only through physics, chemistry, dating, and comparison.

Sources consulted

4 sources, every one with a link you can follow.

  1. EPICA Community Members, “Eight Glacial Cycles from an Antarctic Ice Core,” Nature (2004).
  2. Jean-Robert Petit and colleagues, “Climate and Atmospheric History of the Past 420,000 Years from the Vostok Ice Core,” Nature (1999).
  3. U.S. National Centers for Environmental Information, “Ice Core,” paleoclimatology archive guide.
  4. British Antarctic Survey, “Ice Cores and Climate Change,” methods synthesis.

Want the reader to remember this spot? Sign in — it is free.

Next Hammurabi's Stele as Law, Literature, and Royal Monument

The app

Keep going, or take it with you.

Free accounts read twenty pieces a day. Pro is US$1.99 a month or US$11.99 a year for unlimited reading.