Impact = heat?

Some 635 million years ago, Earth was engulfed in a deep glaciation. Ice up to 800 meters thick covered much, if not all, of the planet for more than 10 million years. But suddenly — over a period of a few hundred to a few thousand years — the planet warmed, the ice melted and Earth returned to a more “normal” state. How Earth came out of its snowball state so quickly has long puzzled researchers. One possibility, some suggest, is that something massive — a meteorite or asteroid — struck the planet and thawed the ice.

For years, scientists have debated what could have brought the planet out of this ice phase. Some researchers think that an increase in carbon dioxide in the atmosphere from volcanic eruptions was enough, says Jim Kasting of Pennsylvania State University in University Park. “If a snowball can form, it can melt,” he says. “Earth can deglaciate itself, work its way out — it’s just physics. You have to get the carbon dioxide absorption coefficients right, though, including so-called ‘pressure-induced’ absorption, which is missing from many climate models,” he says. Another suggestion, he says, is that the ice in the tropics was thin. Such thin ice is also much darker than thick ice, and this would mean that Earth could deglaciate at relatively low carbon dioxide levels.

Other researchers, however, don’t buy those hypotheses. They think that some sort of catastrophic trigger was needed to produce such a drastic, quick change, Kasting says. Think of peeling an orange: Once the prick of a knife breaks the seal, peeling off the skin — like melting the ice off the planet — is fairly easy. These researchers think there must have been such a knife prick during snowball Earth.

During a chance meeting between Jason Goodman, a climatologist now at Wheaton College in Norton, Mass., and Christian Koeberl, an impact expert at the University of Vienna in Austria, Goodman says he mentioned that the “carbon dioxide story has always kind of troubled me” as the deglaciation explanation. They began to discuss the possibility of an impact having caused the deglaciation, Goodman says — a fairly novel idea, Kasting adds.

Koeberl went home and, along with Boris Ivanov of the Russian Academy of Sciences in Moscow, ran impact models to test this hypothesis. He and Ivanov found that statistically there could have been at least one and up to five impacts over any given 10-million-year period large enough to have ended snowball Earth. And an impact striking the ocean would produce almost four times as much water vapor as one striking land, Koeberl noted at the annual meeting of the American Geophysical Union in San Francisco, Calif., in December.

He and Ivanov also determined that the amount of water vapor, salt and dust that would have entered the atmosphere increased exponentially, not linearly, with increasing size of the impact. Even a modest impact with a diameter of just five kilometers — less than half the size of the Chicxulub impact 65 million years ago that killed the dinosaurs — would have caused a local hotspot of a few hundred degrees, Koeberl says, melting the ice locally before radiating outward like the orange peel. And anything larger would have had a much larger effect.

“It could work,” Koeberl says. “But at this point, our simulations do not allow us to firmly conclude if an impact of a realistic magnitude could have caused deglaciation of snowball Earth,” he adds.

What might have happened climatically given the various impact scenarios is ongoing research, Goodman says. One thing to figure out is how an impact could cause warming — impacts generally produce cooling periods during which the ejecta from the impact blocks the sun’s radiation. The impact would have added clouds to snowball Earth, helping warm the planet. Goodman also intends to determine how the darkening of the planet’s surface — caused by dust covering the ice and by the melting ice exposing dark ocean water — affected albedo, which would determine how much light and heat were reflected back into space rather than captured on Earth.

“From my work, it’s clear that you need something more than carbon dioxide to get you out of snowball Earth,” says Ray Pierrehumbert, a climatologist and geophysicist at the University of Chicago in Illinois. The idea of an impact as the trigger “has long been in the back of my mind,” he says, “but I didn’t know how to do the calculations. So this team did it — and this is a big step on the way to evaluating the impact hypothesis.” However, he cautions, “there are a whole lot of issues to be worked out yet,” such as ice flow: Moving ice could potentially close up any hole made by an impact.

Another issue is that there were at least two periods of snowball Earth in history: one at about 635 million years ago, and another at 710 million years ago that is less well studied. “So you have to get out of snowball conditions twice,” Pierrehumbert says. “Is an impact going to explain it both times, or does it even need to, to be right for one of the snowballs?” he says. Nonetheless, there is one piece of additional evidence for an impact, he says, and that’s a layer with anomalously high contents of iridium that was found in rocks dated to the second snowball period. Still, although such iridium layers often indicate impacts, they can also be explained by cosmic dust, so it’s not a slam dunk, Koeberl adds.

“It is conceivable that an impact got us out of snowball Earth,” Kasting says, “but that’s not what I think happened.” Finding further evidence of an impact, such as shocked minerals, whose structures are deformed by an impact, would help the hypothesis, he says — but not finding that evidence doesn’t eliminate the possibility either.

Megan Sever

Links:
"Escape from snowball Earth," Geotimes, February 2007

Back to top