When scientists reconstruct ancient climates from ice cores, how do they create an accurate timeline?

Researchers use chemical analyses, computer models, and other tools to date these valuable climate archives.

October 9, 2025

Beneath the snowy surfaces of our planet’s ice sheets and glaciers lies a trove of information about past climates. As snow falls year after year, it gets buried and compressed into a frigid chronicle of Earth history: Ice, and the impurities and air bubbles trapped inside it, preserve evidence of ancient temperatures, greenhouse gases, and even volcanic eruptions. To get at this record, scientists drill down and remove cylindrical “cores” of ice that can span hundreds of thousands of years.¹

That means ice core researchers have (so to speak) a lot of time on their hands. And building a reliable timeline for a core, linking what happened in Earth’s past to when it happened, is important—not just to make sense of the ice core itself, but to compare it with other ancient “archives,” like seafloor sediments or stalagmites.

“If you see something interesting 100,000 years ago in one archive, you have to be able to compare it confidently to another archive and know that you’re comparing the same time period,” says David McGee, professor of Earth and planetary sciences at MIT. And by placing key events in order, scientists can investigate causes and effects in the saga of Earth’s climate history—piecing together “not just how climate changed, but why it changed.”

So how do researchers date an ice core?

The most straightforward method is to count backward, one year at a time. For example, seasonal differences in snow and in impurities like dust can show up as alternating lighter and darker bands in a core²—helping divide the icy logbook into yearly chapters. Some aspects of snow that aren’t visible to the naked eye, like the molecular properties of water, also vary with the season.³ By counting off seasonal cycles, visible or invisible, scientists can in some cases pick out annual layers going back tens of thousands of years.⁴

That reveals the age of the ice itself. But the ice also contains another source of information: air bubbles, which bottle up ancient gases and show how the contents of Earth’s atmosphere have changed over time.

This direct access to ancient air is “really unusual” and deeply valuable, McGee says. But in a tricky twist, the air bubbles are younger than the ice around them, sometimes by thousands of years. As fallen snow is gradually compressed into ice by the weight of fresh layers, the air buried with it can still mix with the atmosphere until there’s been enough compaction to trap it in closed-off bubbles.

Dating these air bubbles, then, calls for additional tools, such as computer models of how snow becomes ice. Researchers also look at the nitrogen gas in the bubbles. Nitrogen comes in different “isotopes,” some of which are heavier and settle downward more readily. This means the mix of isotopes helps reveal the depth at which air was finally cut off from the atmosphere.⁵

More challenges arise deeper inside an ice core. Ice doesn’t simply pile up neatly “like layers of a cake,” McGee says. “It’s also flowing actively.” Think of pancake batter spreading on a griddle, he says. “As you go further and further down in an ice core, you’re sampling layers that have experienced more flow outward, and that’s caused them to thin.” In other words, “each centimeter is more and more time that’s been compressed.”

Eventually, that makes annual layers imperceptible. “As layers thin and thin, you lose the ability to continue to layer-count,” says McGee. “So then you have to use other methods to estimate the relationship between depth and age.”

Here again, computer models come in handy, simulating how ice flows and layers thin with depth. Researchers also hunt through an ice core for volcanic ashes, which McGee says can “anchor” the core in time. By examining their chemical makeup, scientists can link those ashes to an eruption whose age is known—for instance, based on the gradual decay of radioactive material, which provides a kind of ticking clock for the remnants near the volcano.⁶ (Radioactive decay can also help date the core directly; McGee says this is an important tool today as scientists hunt for super-old ice that’s been pushed up toward the surface.⁷) 

Still other methods rely on the “calendar” of Earth’s orbit, which shifts slightly over time. These shifts change how much sunlight reaches the poles, in predictable cycles that have been modeled deep into the past. Helpfully, these cycles go hand-in-hand with changes in the precise ratio of gases that get trapped in air bubbles. That means scientists can measure the changes in a core’s trapped air and match them up with the cycles of sunlight—effectively, situating data from the core within Earth’s orbital timeline.⁸

So researchers have assembled a toolkit of creative ways to date ice and trapped air. And McGee emphasizes that scientists aim never to rely on just one. By using several different tools—simulations of ice flow, annual layer counts, markers of volcanic eruptions—and gauging how well they agree, scientists can spot errors and build the most robust timeline possible.⁹

Ultimately, “age uncertainties do increase as we go back in time,” McGee says, and the record does lose some of its detail in deeper sections. Say, for example, that you want to know how high carbon dioxide (CO₂) levels were in the past (crucial data as we try to understand today’s climate change, which is driven by the vast amounts of CO₂ humanity is adding to the atmosphere). Near the top of the core, clear layers might let you track CO₂ backward year-by-year. But lower down, McGee says, you may have to settle for average levels over a decade, or a century, or even a millennium.

Despite the challenges, ice cores are incredibly rich sources of information about our planet’s past. And they’ve proven their reliability, McGee says, as researchers compare them to one another and to other records of Earth’s ancient climates. “We feel really confident in the information that comes from them.”

Thank you to Robert Miller of Masterton, New Zealand, for the question.

Submit your own question to Ask MIT Climate

Get the latest from Ask MIT Climate monthly in your inbox

Read more Ask MIT Climate