Chemicals trapped in ancient glacial or polar ice can move substantial distances within the ice, according to new evidence from University of Washington researchers.
That means past analyses of historic climate changes gleaned from ice core samples might not be entirely accurate.
"The ice cores themselves are wonderful records of climate. Nobody is questioning that," said Alan Rempel, a postdoctoral research scientist in the UW Applied Physics Laboratory.
In fact, the research shows that the fingerprint of chemical variations within ice cores is much sharper than had previously been expected. But it also shows that substances that are climate signatures -- from sea salt to sulfuric acid -- travel through the frozen mass along microscopic channels of liquid water between individual ice crystals, away from the ice on which they were deposited.
The movement becomes more pronounced over time as the flow of ice carries the substances deeper within the ice sheet, where it is warmer and there is more liquid water between ice crystals. By contrast, oxygen isotopes that can indicate past temperatures are carried mostly within the ice.
The possible movement of chemical signatures away from the ice on which they were deposited means scientists must re-examine questions such as whether warm summers coincided with high levels of sea salt in the air, Rempel said. But that is only true for ancient ice, since little movement is shown in ice less than 100,000 years old.
The findings by Rempel; John Wettlaufer, a senior physicist at APL; Edwin Waddington, a UW professor of Earth and space sciences; and APL visiting scientist Grae Worster from the University of Cambridge in England are published in today's (May 31) issue of the journal Nature.
Ice sheets (large polar glaciers) are built by thousands of years of accumulated snowfall, to depths of thousands of meters. Each season's snowfall forms a distinctive layer that can be analyzed chemically after being extracted in a core sample.
Certain impurities serve as markers that can tell scientists what was going on climatically at various times. Those substances are found principally in unfrozen liquid that accumulates at the boundaries of individual crystals within the ice sheets.
The new research shows those substances migrate deeper into the ice sheet, where it is warmer, faster than the ice on which they were deposited, said Wettlaufer, an ice physics expert. The displacement is larger at greater depths. The result is that substances found 3 kilometers (1.9 miles) deep could be 50 centimeters (20 inches) or more away from the ice on which they were deposited many thousands of years ago, a distance that accounts for about 100 years of snowfall.
"The point of the paper is to suggest that the ice core community go back and redo the chemistry," said Wettlaufer. "That's a lot of work, and we're hoping to be involved in that."
The Nature article notes that the best high-resolution climate records over the past few hundred thousand years have come from ice cores taken from Greenland and Antarctica.
A core from interglacial ice in central Greenland suggests that a sudden cooling took place in the Eemian period 115,000 to 125,000 years ago. However, the new study shows that impurities used as climate markers may have moved as much as 20 inches, a distance large enough to offset the resolution at which the core was examined and alter the interpretation of the ice-core record.
The Vostok core from Antarctica, which goes back some 450,000 years, contains even greater displacement because of the greater depth, but it has not been examined at even the close spatial resolutions of the Greenland core, Wettlaufer said.
Rempel said the researchers hope to devise models that can help scientists account for the relative movement of different impurities when making their ice core measurements.
But in the meantime, said Wettlaufer, scientists doing that climate work have to take into account how much their measurements might be skewed, and adjust their findings accordingly.
"That would be what we most would want to influence -- the way people make their observations," he said. "Since they are doing all that work, they can't afford to neglect these important physical processes in their interpretation."
[Contact: Alan Rempel, John Wettlaufer, Sandra Hines, Vince Stricherz]