SNOWBALL EARTH

Muse, Jan 2004 by Walker, Gabrielle

Picture a world entirely sheathed in ice, its oceans smothered by a freezing white blanket nearly a mile thick. Vast glaciers creep across the continents. Nothing else moves. There are no clouds, save perhaps a handful of high wispy streaks made from frozen crystals of carbon dioxide. With the temperature a chilling 40 degrees below zero, only a few living things survive. Algae cling to the meager warmth of volcanic springs, and bacteria eke out a living around hot-water vents deep in the ocean. For millions of years, nothing changes.

This is no far-off planet or alien moon. It's a view of Earth just a few hundred million years ago. A growing number of scientists believe that, after billions of years of comfortable existence, Earth suddenly plunged into a winter so extreme that it makes recent ice ages look warm. Then just as suddenly the ice melted again, and the planet sweltered in a climatic backlash of intense heat. Between 750 million and 590 million years ago, supporters of the "Snowball Earth" theory say, the climate may have swung back and forth between deep freeze and hothouse as many as four times. What's more, they say, the ice might be the reason complex animals like us are around today.

Two geologists, Joe Kirschvink of the California Institute of Technology and Paul Hoffman of Harvard University, started hammering out the basic ideas behind Snowball Earth in the early 1990s. But for at least 30 years before that, scientists around the world had been puzzling over rocks that seemed to show traces of ancient ice. Some were jumbles of boulders and silts, the boulders scraped and scratched from being dragged along by giant rivers of ice. Others contained "dropstones," rocks that drifting icebergs had carried out to sea and then dropped into the muck of an ancient seafloor.

Glacial rock deposits and dropstones are still being made in cold parts of the world today. The mystery rocks, however, were ancient. They had formed about 700 million years ago, at a time when the only living things were simple, single-celled bacteria and algae. Because the ice rocks were squeezed up against another type of rock that is normally laid down only in the tropics, a few geologists suggested that some of the glacial deposits had formed close to the equator-the hottest part of Earth. The idea was disturbing. If ice could exist under a hot tropical sun, it could exist anywhere. Could the whole planet have been covered in ice?

The answer, most other scientists agreed, was no. If Earth had ever frozen solid, they believed, it would never have melted. Computer models showed that the ice would have reflected sunlight back into space, keeping Earth's surface frozen forever. Since we're not living in a global icehouse today, something is wrong with this story. Either ice wasn't responsible for the rocks, or it had formed them closer to the poles, where you would expect ice to be.

Joe Kirschvink, however, didn't buy that. An expert in paleomagnetism (traces of ancient magnetic fields preserved inside rocks), Kirschvink had spent months studying 700-million-year-old glacial rocks that other scientists had found in Australia in 1986. By looking at magnetic minerals trapped in the rocks, Kirschvink became convinced that the rocks had been formed in the tropics. "It is absolutely, utterly bulletproof," he says. There must have been ice near the equator.

So how did the frozen Earth thaw out again? The computer models showing an eternal deep freeze, Kirschvink realized, had left out something important: greenhouse gases. Nowadays many scientists worry that heat-trapping gases like carbon dioxide (CO2) are raising Earth's temperature. Hundreds of millions of years ago, of course, there were no cars or factories releasing CO2 into the air, but there were volcanoes. Volcanoes perpetually belch carbon dioxide gas. Usually, rainwater washes volcanic CO2 out of the atmosphere. But on a frozen Earth, there would be no rain; instead of evaporating to form clouds, water would stay on the ground, cold and rock-solid. On a rainless Earth, CO2 from volcanoes would simply build up in the atmosphere until it trapped enough heat to melt the ice.

In 1992, Kirschvink wrote a paper outlining his ideas and coined the term "Snowball Earth" to describe that biggest of chills. Other scientists attacked the notion. Even in the depths of the last ice age, which ended only 15,000 years ago, ice sheets had barely spread beyond northern Europe and the middle U.S. To suggest that the oceans themselves could freeze over-and do so right across the equator-seemed far-fetched. Kirschvink didn't feel like arguing, so he let the matter drop and turned to other research.

Then Paul Hoffman came along. Hoffman, a geologist, was studying rocks in Namibia, in southwestern Africa. Everywhere he looked, he saw evidence of ice. "I was fascinated by Kirschvink's idea," he says. And after years of work, he found more evidence to support it.

The evidence came from ancient rocks called carbonates, which are made when carbon dioxide dissolves in water, combines with other chemicals, and precipitates out. Hoffman was particularly interested in the element carbon. In seawater, carbon comes in two forms: a light variety called carbon-12 and rarer, heavier carbon-13. Ocean-dwelling plants, which need carbon to live, take both kinds out of seawater, but they prefer carbon-12. That means that when an ocean is filled with plant life, carbonates that form there contain more heavy carbon than usual. Heavy carbon tells geologists that carbonate rocks formed in a living ocean; light carbon, that they formed in a dead one.