Devils Hole heats up debate over ice ages.

For the last 15 years, most climate researchers have looked to space for an explanation of the ice ages that have repeatedly gripped our planet in recent geologic times. The established theory, called the Milankovitch hypothesis, holds that wiggles and wobbles in Earth's orbit serve as a pacemaker that determines when the planet plunges into a glacial period and when it thaws out of one. But new evidence from a deep crack in the Nevada desert threatens to over-turn the Milankovitch theory and replace it with a more down-to-Earth solution.

Geoscientists report this week that the ice ages did not follow any pattern consistent with orbital variations. Rather, chaotic elements in Earth's own climate dictated when the planet slipped into and out of a deep freeze, the researchers suggest.

"We feel that the Milankovitch theory is incapable of explaining the climate shifts," says Isaac J. Winograd of the U.S. Geological Survey in Reston, Va. He and his colleagues discuss their findings in two papers in the Oct. 9 SCIENCE.

The climate information collected by Winograd's group comes from Devils Hole, an open fault zone in southwestern Nevada. The fissure is filled with mineral-rich water that has coated the rock walls with layer upon layer of calcite over the last 500,000 years. Divers equipped with scuba gear entered the fault and used a drill to cut a 36-centimeter-long cylinder out of the calcite coating.

By analyzing the ratio of two isotopes--oxygen-18 and oxygen-16 -- at hundreds of spots along the calcite core, Winograd and his colleagues identified changes in the temperature of the atmosphere when rain fell in the Devils Hole region. They dated these climate shifts by using the radioactive decay of uranium within the calcite as a clock. Previously, they had drilled a core that recorded information going back 250,000 years (SN: 12/3/88, p.356). The new core doubles the length of that record.

Scientists who study the glacial cycle have traditionally relied on climate records constructed by measuring the oxygen isotopes in seafloor sediments. But the dates of the climate events in these marine records are less certain than those in the Devils Hole core.

The idea that ice ages result from alterations in Earth's orbit goes back to the 19th century. However, the theory did not enjoy widespread support until 1976, when oceanographers analyzed two marine records and found that the glacial cycle closely matched changes in the shape of Earth's circular orbit, the tilt of the planet, and the way its axis wobbles.

The marine records led researchers to suggest that these orbital variations paced the climate changes by altering the amount of sunlight reaching the sub-Artic region during summer. When the summer sunlight weakened, the snow from winter could survive the warm months, building up year after year to form a great glacial cap over Canada and northern Eurasia. When the summer sunlight intensified, it would melt the ice sheets.

But the Devils Hole record, with its superior chronology, shows that the timing of specific events in the last 500,000 years does not match the predictions of the Milankovitch theory, according to Winograd and his co-workers. Three of the last four ice ages ended when the summer sunlight in the sub-Arctic was relatively weak. What's more, the ice ages lasted different lengths of time, varying from 80,000 to 130,000 years in duration. This suggests that the climate changes follow no regular pacemaker, the researchers contend.

They propose instead that ocean circulation patterns, greenhouse gases in the atmosphere, and the behavior of ice sheets all interact chaotically to cause the irregular glacial cycles. Orbital variations may play a part in this story, but they are by no means principal characters.

James D. Hays, a longtime champion of the Milankovitch theory, calls the Devils Hole data "a beautiful record" but argues that the new data do not sink the idea that orbital changes control the ice ages. Hays, a researcher at Columbia University's Lamont-Doherty Geological Observatory in Palisades, N.Y., points out that although the Devils Hole chronology has more accurate dates than other records, it still faces dating problems. Most important, he says, researchers don't know how many thousands of years rainwater takes to filter through the ground and reach Devils Hole -- a factor that could alter the timing of climatic events recorded there.

The Devils Hole record, like the marine ones, includes variations whose frequencies match those of the orbital cycles. This suggests to Hays that these astronomical variables do play an important role. While summer sunlight in the sub-Arctic might not control the growth of ice sheets, a different pattern of sunlight changes may, he says.