Tiny Frigid Bubbles Get to the Core of Climate Change
As Michael Bender prepared to lead the way into the storage area of his lab at Princeton University, he gave a visitor a quizzical look. “You really might want to put these on,” he said, holding up a bulky red parka and a pair of thick gloves. “Oh, I’ll be fine,” said his guest. “No, really,” Bender insisted gently. “It would be a good idea.”
A minute later, it all made a lot more sense. The storage area is a refrigerator the size of a walk-in closet, chilled to minus 30°F, and with a powerful fan blowing just to ensure the frigid air circulates evenly to every corner of the cramped space. Plastic foam coolers and cardboard boxes lined with insulation cover most of the floor, with more piled on top. Bender reached into one of the coolers, pulled out a plastic bag with a lump of ice inside and held it to the light. On close inspection, you could see that the ice was permeated with tiny bubbles, as though it was a chunk of frozen Sprite — and if you chipped off a piece and dropped it into a glass of water, the ice would sizzle and hiss, as the bubbles escaped.
These bubbles didn’t come out of a soft-drink factory, however. They’re bits of ancient atmosphere, trapped in the spaces between fallen snowflakes that eventually became welded into a mass of solid ice in the world’s truly cold places. “This one is from Antarctica,” Bender said over the whirr of the fan. “And this,” he said, retrieving another sample, “comes from Greenland.”
The bubbles, preserved like flies in amber, are tiny time capsules that hold a record of what the air was like — its temperature, the gases it was made of, the tiny particles of dust and pollen and volcanic ash it carried — when the snow first fell. And because each year’s snowfall buries the snow from the previous year, which buries the snow from the year before, and so on into the past, the bubbles that come from deeper layers contain air that’s tens or even hundreds of thousands of years old.
By gently melting slices of ice from different depths to release and study this preserved air, scientists like Bender have teased out the story of a climate that has changed drastically, plunging into the frigid depths of ice ages and emerging into warm interglacial periods over at least the past 800,000 years.
In large part, their goal is to understand how the climate responds to changing concentrations of greenhouse gases such as carbon dioxide, which for the first time in the planet’s history are generated from human activity more than natural sources. They’re reading the past in order to understand what the future might hold.
What the past has told them already is that there’s been an intimate relationship between carbon dioxide and temperature as far back as they can see. When CO2 is high, so is the thermometer, and when it drops, the temperature goes with it. But the ice can tell them much more than that. It also carries information about what kinds of vegetation thrived in different eras, and whether the planet was moist or dry, and even how bright the Sun was.
All of that information and more is locked up deep ice; it’s Bender’s job, and that of his colleagues across the world, to unlock it.
The concept is simple enough, but the execution and analysis can be extremely complicated. The first step, Bender explained back outside the refrigerator, is to retrieve samples from sheets of ice that can be thousands of feet thick. U.S. scientists rely mostly on crews from Ice Core Drilling Services, based at the University of Wisconsin, Madison, who use custom-designed equipment to extract cylinders, or cores of ice, a little less than 5 inches across. Back in the lab, where the cylinders are shipped packed in dry ice, you can easily see the layers representing individual years, much as you can see each year of a tree’s growth in its annual rings.
You don’t just show up in Antarctica and drill anywhere, though. “The preferred place to work,” Bender said, “is at a dome.” These are the ice plateaus that mark the very highest points on the world’s highest continent. They’re ideal for two reasons. First, the ice sheet is thickest here, so you can drill most deeply into the past. Second, Bender said, “Once you get off the highest point, the ice is flowing laterally, trying to discharge into the ocean as bergs. The flow leads to the deepest layers being folded and mixed up.”
Even when the layers are nice and orderly, however, the information scientists can extract from the air bubbles, and also from the ice that surrounds them, isn’t much good if they don’t know how old a given layer is. They do it by comparing ice cores with other ancient records — sediments from the bottom of the sea, for example, where dust and organic matter, including shells of tiny plankton known as foraminifera, form their own layers.
The organic material’s age can be teased out with radioactive dating, and if you go far enough back, you can see a change in the orientation of tiny iron particles from a time when Earth’s north and south magnetic poles changed places about 780,000 years ago. (Contrary to what some poorly informed folks believe, these reversals, which happen every so often, have nothing to do with climate change). Scientists can also synchronize the sea floor and ice core records by looking for thin layers of ash that mark massive volcanic eruptions.
Once they’ve figured the age of a layer in an ice core, paleoclimatologists melt the ice and capture the trapped air. The meltwater tells them what the air temperature was at the time the original snow actually fell, based on the form of oxygen the water contains. The liberated air, meanwhile, tells the scientists how much CO2 was in the atmosphere at the time.
As climate skeptics love to point out, these measurements lead to an apparent paradox: if you look closely enough, you see that over and over, as ice ages gave way to warm interglacial periods, the temperature began to rise before the CO2. In fact, this makes perfect sense. Enormous amounts of CO2 are stored in the deep ocean, so when changes in Earth’s orbit bring more sunlight to the poles, the jolt of warmth liberates the stored gas, leading to more warming, and ultimately to the end of the ice age. A recent paper showed exactly how it might have played out.
Scientists like Bender aren’t content just to leave it at that, however. They’re constantly trying to determine new ways to slice and dice ancient air see what other stories they might tell of the ancient past. They look for traces of methane, for example, which naturally rise and fall as methane-burping wetlands spread during wetter times and shrink when it’s dry — a clue to average rainfall at different times in Earth’s history. They look for nitrous oxide, produced by bacteria in drier soils. They even look for changes in the mixture of gases that tell them how quickly the original snow grains welded themselves together, which tells them how bright the sun was at any given era.
All of that comes from the continuous ice-core record, which goes no more than 800,000 years into the past. But Bender is determined to break that barrier. His lab is now working with ice he believes to be more than a million years old. You can’t use conventional dating techniques to confirm its antiquity, but he and his colleagues think they’ve figured out a way (it has to do with radioactive argon).
It’s not just curiosity that drives him. For the past million years or so, ice ages have lasted about 100,000 years each (the information comes not just from ice cores, but also from geological records). But before that, Bender said, “the cycles lasted 40,000 years, and the ice volume was only half of what we’ve gotten more recently.” Nobody really knows why — but there was clearly something different going on, quite possibly having to do with a mix of greenhouse gases different from what came later.
Understanding what changed at a million years B.C. could help climate scientists better understand the climate system overall. That in turn will help climatologists to gauge the coming impacts of human-generated greenhouse gases more accurately. The better the information they have to feed into their models, the more we can trust the projections that come out — and plan for what’s on the way.