Published online 11 February 2010 | Nature | doi:10.1038/news.2010.59


Sea-level records challenged

High point 80,000 years ago may hint at flaws in ice-age theory.

Cave in SpainMinerals deposited in caves flooded by the Mediterranean show sea-levels 81,000 years ago.Tony Merino

Precise measurements of sea level from Mediterranean caves have revealed that about 81,000 years ago the seas stood much higher than previously thought — even higher than today's levels. The finding may force scientists to reconsider how Earth's large ice sheets wax and wane in response to changing climate.

Changes in global sea level are used to trace the fluctuations of ice sheets. As ice sheets advanced from a low-point around 125,000 years ago to a maximum around 20,000 years ago, sea level fell — although there was plenty of variation along the way. 125,000 years ago, seas stood roughly at today's heights, whereas at the last glacial maximum, sea levels were 130 metres lower than at present.

But a group led by Jeffrey Dorale, a geologist at the University of Iowa in Iowa City, now suggests that 81,000 years ago the sea level increased sharply to reach a high-point, roughly one metre above today's level1.

"At face value this is a very beautiful and solid piece of work," says Wallace Broecker, a geologist and climate scientist at Columbia University's Lamont-Doherty Earth Observatory in Palisades, New York. "It suggests that, for reasons unknown, the sea level which at the time should have been down some 15 metres instead went up one metre, and then went right back down again within just a couple of thousand years."

Sea up, ice down

If Dorale's measurements stand up to scrutiny, they also imply that glaciers and ice sheets receded dramatically over a very short period before starting to grow again at similar speed. What triggered the melting, and what made glaciers start growing again, is unclear. "It could all come down to an anomaly, caused by some complex feedbacks perhaps related to the global carbon cycle. But we really don't know," says Dorale.

Measurements of past sea levels have been plagued by uncertainties — such as the extent to which tectonic movement has deformed Earth. And sea level height around 80,000 years ago has been an issue of intense debate. Previous reconstructions, based on the growth of coral reefs off Haiti, Barbados and New Guinea, suggested that the sea level then was 7–30 metres below its present value.2

Dorale and his team dated layers of calcite, a mineral deposited in coastal caves on the Spanish island of Mallorca that were flooded when the sea level rose high enough. Because tidal activity in the Mediterranean Sea is extremely low, and because there has been little tectonic movement in the region, the mineral encrustations are thought to reflect the mean sea level with unique accuracy.

Experts acknowledge the robustness of the team's observations and dating methods. But they caution that the inferred rates of sea level change will not be the last word on the issue.

"This is an exciting study that will certainly spur further work," says Peter Huybers, a palaeoclimatologist at Harvard University in Cambridge, Massachusetts. "But it might be premature to draw definite conclusions on how rapidly the sea level has changed 80,000 years ago."

The rhythm of ice

Dorale says that the timing of the sea level changes raises questions about what kind of climate change could have caused it. Observations so far suggest that ice-sheet growth and decay in the past few million years happens on average to a 100,000-year beat. Temperatures and greenhouse-gas levels, measured from ice cores, also follow this long-term rhythm.

These cycles have been linked to the Milankovitch theory, which calculates how small changes in Earth's orbit and axial tilt bring about a regular succession of cold and warm periods because they alter the intensity of sunlight reaching the Northern Hemisphere in summer.3 Calculated sunlight variation comes in three rhythms — of 23,000; 41,000 and 100,000 years. Although the 100,000-year periodicity is weak, it seems to have dominated the most recent cycles, and has been used to explain observed sea-level and ice-sheet fluctuations.

The new discovery of rapid ice-sheet growth suggests that glacial cycles do not couple precisely with records of sunlight variation and greenhouse-gas concentrations, says Dorale. "The variations in solar power seem too small to cause the large changes in ice cover that we know have occurred over the past 700,000 years," he says. One of the shorter periods of Milankovitch sunlight variation may also need to be taken into account.


But Carl Wunsch, a physical oceanographer at the Massachusetts Institute of Technology in Cambridge, thinks that the result brings no fundamental conflict with established ideas about long-term glacial cycles — which he says have always hovered rather widely around the 100,000-year Milankovitch cycle.

Concluding that glacial cycles may arise because of climate fluctuations that are unrelated to long-term sunlight cycles would be a gross simplification, says Huybers. "It's tempting to conclude that the 100,000-year cycle must now be reconsidered," he says. "But people have been arguing about glacial cycles and their length for a long time." 

Commenting is now closed.