Public Release: 

Sea-level clue to climate change

University of Illinois at Chicago

It sounds like the plot for a disaster film: rising temperatures melt polar ice, causing a flood of freshwater to rapidly enter the salty North Atlantic. As the fresh and salty water mixes, density changes, altering the Gulf Stream ocean currents that moderate the North Atlantic climate. In just a few years, average temperatures plummet, ushering in a deep freeze that lasts a century or more before fresh and salty water is back in balance, ocean currents adjust and temperatures return to normal.

Science fiction? Not to a growing number of geologists and climatologists who've studied facts showing a precipitous 6-degree Centigrade drop in Greenland's average temperature some 8,200 years ago as the Earth was exiting the last ice age and polar ice sheets were melting in retreat. Many scientists believe a catastrophic flood of freshwater entering the North Atlantic clipped the flow of the Gulf Stream current. The suspected source of the floodwater was a glacial reservoir called Lake Agassiz. A popular theory suggests that Lake Agassiz's huge volume of freshwater -- more than twice that of today's Caspian Sea -- may have breached an ice dam or tunneled under Hudson Bay's ice sheets, then gushed into the North Atlantic, perhaps in a period lasting only months. Scientists call it the largest mega-flood of the last 100,000 years.

Torbjörn Törnqvist, an assistant professor of earth and environmental sciences at the University of Illinois at Chicago, reports in the Dec. 11 online issue of Geophysical Research Letters about a new set of 8,200-year-old core samples that indicate an abrupt sea-level rise. The finding adds credence to the theory that a catastrophic freshwater flood into the North Atlantic triggered the great chill around that date.

"Few would argue it's the most dramatic climate change in the last 10,000 years," said Törnqvist. "We're now able to show the first sea-level record that corresponds to that event."

The discovery came by coincidence. Törnqvist and his graduate students are conducting ongoing studies into sea-level changes along the Gulf of Mexico, using core samples of peat retrieved from the swamps and marshes in the Mississippi River delta in Louisiana. Samples gathered in 2003 from a saltwater marsh in an area known as Bayou Sale held the clue.

As sea levels rise, peat deposits are formed. These deposits can be accurately radiocarbon dated. They also contain organic debris that can suggest whether water was salty or fresh at the time of deposition, based on plant salt tolerance. Analyzing his samples, Törnqvist discovered them to be around 8,200 years old and found evidence that a saltwater marsh was abruptly flooded and turned into a lagoon, indicating a sudden sea-level rise at the time.

Törnqvist said that if comparable sea-level readings can be taken from other coastal areas on Earth, it could add evidence that a catastrophic freshwater flood into the North Atlantic 8,200 years ago did cause ocean current disruption and the consequent abrupt climate change.

"We happened to sample along the Gulf of Mexico, but there's no reason you can't study this in, say, China or New Zealand as well," said Törnqvist. "The oceans are all connected. If we can measure the amount of sea level rise that occurred 8,200 years ago, we will be able to convert that back into a measurable amount of freshwater. With our first data, we now know the amount of sea-level rise was probably less than 1.2 meters - which is less than several previously published estimates. In the future, we hope to come up with a more accurate number. Climatologists urgently need this type of information to run their climate models in order to understand the conditions that can produce such an abrupt climate change."


The research was supported by grants from the National Science Foundation, the National Geographic Society, the Geological Society of America and the Gulf Coast Association of Geological Societies. Radiocarbon dating was done at the Robert J. Van de Graaff Laboratory at Utrecht University in the Netherlands.

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.