Warm band may have girdled snowball Earth.

In the era before multicelled animals appeared some 550 million years ago, extreme cold may have clutched Earth in a death grip for millions of years. According to the controversial snowball Earth theory, the planet swung between deep freeze and feverish heat several times during a 200-million-year period preceding the multicellular explosion.

Offering an alternative to the snowball theory's ice-engulfed planet, a team of researchers from Texas and Canada now demonstrates that a swath of liquid ocean may have hugged the planet's midriff even during the chilliest episodes.

"That life survived through this episode seems to favor this open-ocean solution," says William T. Hyde of Texas A&M University in College Station. He and his colleagues describe their computer simulations in the May 25 NATURE.

By also linking possible snowball conditions to detailed dynamics of ice sheets for the first time, the researchers add "important new twists" to current understanding of the icy era, comments Mark A. Chandler of NASA's Goddard Institute for Space Studies in New York.

For more than 35 years, scientists have suspected that between 800 million and 600 million years ago, glaciers extended all the way to the equator. Two years ago, after discovering widely fluctuating carbon isotope ratios in rocks from that so-called Neoproterozoic era, Paul F. Hoffman of Harvard University and his colleagues proposed a radical hypothesis to explain the glaciation.

Building on notions of earlier investigators, Hoffman's team proposed extreme climatic yo-yoing. Weak sunlight, the researchers said, had combined with low atmospheric concentrations of the greenhouse gas carbon dioxide and a runaway feedback effect to freeze Earth's surface for some 10 million years (SN: 8/29/98, p. 137). Carbon emissions from volcanoes reversed the pattern, but the planetwide ice returned, perhaps several times, they argued.

Sparing only a few cold-dwelling organisms or those at hot vents in the deep ocean, the catastrophe may have opened niches for new life-forms, the researchers suggested. It may have even spurred a tremendous bloom of evolutionary innovation known as the Ediacaran fauna (SN: 5/18/96, p. 308).

Earlier computer simulations of Neoproterozoic conditions had indicated that if global carbon dioxide concentrations dropped low enough, the planet could indeed have snowballed into disaster. Hyde and his coworkers at Texas A&M and the University of Toronto wondered if adding the realism of detailed ice-sheet behavior would melt the snowball theory.

To find out, they ran simulations using a so-called energy-balance model. They combined it with specialized software to model such details as how ice flows and responds to changing temperature and precipitation. In these and further tests employing a general-circulation model, they tried various settings for atmospheric carbon dioxide concentration and other climate parameters. For all the computer runs, they assumed the sun would be 6 percent dimmer than today--a widely accepted value for that era.

Many of their simulations yielded a snowball Earth. To their surprise, however, Hyde says, they found conditions in which about a fifth of the global oceans remained unfrozen even while glaciers extended to the equator.

This partial snowball may instead be a stage of thawing in the rebound from a deep freeze, Hoffman argues--perhaps a stage that persisted for millions of years. "It's potentially quite interesting biologically," he says, because unfrozen seas may have sheltered multicelled animals and seaweeds.

Chandler and his coworkers, on the other hand, plan to report in a future issue of the JOURNAL OF GEOPHYSICAL RESEARCH--ATMOSPHERES simulations that generate both open waters and tropical glaciation with much less carbon dioxide than in Hyde's models.