Ice Cores and the Long View — What Deep Time Tells Us About CO₂

The most powerful data set in climate science doesn't come from satellites or ground stations. It comes from cores of ice drilled out of the Antarctic and Greenland ice sheets — samples of frozen atmosphere going back 800,000 years, with some records extending further.

Ice cores work because when snow falls and compresses into ice, it traps small bubbles of the atmosphere as it was at the time of deposition. Pull a core from 3 kilometers deep in the Antarctic ice sheet, and you're looking at air that was captured when Homo sapiens had just emerged in Africa. The chemical composition of those bubbles tells you the CO₂ concentration. The ratio of oxygen isotopes in the ice itself tells you the temperature at the time of deposition. Match the two, and you have a record of how temperature and CO₂ have tracked together across hundreds of thousands of years.

The data from cores like EPICA Dome C shows a striking pattern. Over the past 800,000 years, the Earth has cycled through roughly eight ice age cycles, each lasting about 100,000 years. CO₂ concentrations ranged between about 180 ppm (glacial maximum) and 280 ppm (interglacial warm period). Temperature varied by roughly 6-10°C between the coldest and warmest periods.

What you see in the data is that CO₂ and temperature track together very closely. This isn't a surprise given the physics. But it also comes with an important nuance: CO₂ changes don't always lead temperature changes. In many ice age transitions, the temperature signal slightly precedes the CO₂ signal.

Climate skeptics have pointed to this as evidence against CO₂-driven warming. This is a misreading of the data. The primary driver of ice ages is Milankovitch cycles — variations in Earth's orbital parameters (eccentricity, axial tilt, and precession) that change the distribution of solar energy reaching Earth's surface. These orbital changes trigger initial warming or cooling, which then affects CO₂ (through ocean outgassing or absorption, permafrost dynamics, etc.), which then amplifies the initial orbital-driven signal through the greenhouse feedback. CO₂ can be an amplifier that follows an initial trigger rather than the original cause, and that's still consistent with CO₂ having a strong warming effect.

The Paleocene-Eocene Thermal Maximum (PETM), about 56 million years ago, is another instructive data point. A rapid release of carbon into the atmosphere — likely from volcanic activity or the destabilization of seafloor methane hydrates — raised global temperatures by 5-8°C over about 20,000 years. Sea levels rose significantly. Ocean acidification was substantial. Many species went extinct. It took about 200,000 years for the excess carbon to be absorbed and temperatures to return to baseline.

The current human-driven CO₂ increase is happening about 10 times faster than the PETM spike. We're not in the PETM — the starting conditions are different and the total carbon input will likely be lower — but the paleoclimate record makes it clear that rapid CO₂ increases produce large and long-lasting changes in temperature, sea level, and ocean chemistry. The deep time record isn't telling us something subtle or ambiguous. It's telling us that CO₂ is a strong climate driver and that atmospheric chemistry changes of the scale we're producing have, historically, led to significant and disruptive planetary changes.

Ice Cores and the Long View — What Deep Time Tells Us About CO₂

// COMMENTS

ON THIS PAGE