News FocusClimate Change

Atlantic Climate Pacemaker for Millennia Past, Decades Hence?

See allHide authors and affiliations

Science  01 Jul 2005:
Vol. 309, Issue 5731, pp. 41-43
DOI: 10.1126/science.309.5731.41

An unsteady ocean conveyor delivering heat to the far North Atlantic has been abetting everything from rising temperatures to surging hurricanes, but look for a turnaround soon

Benjamin Franklin knew about the warm Gulf Stream that flows north and east off the North American coast, ferrying more than a petawatt of heating power to the chilly far North Atlantic. But he could have had little inkling of the role that this ponderous ocean circulation has had in the climatic vicissitudes of the greater Atlantic region and even the globe.

With a longer view of climate history and long-running climate models, today's researchers are tying decades-long oscillations in the Gulf Stream and the rest of the ocean conveyor to long-recognized fluctuations in Atlantic sea-surface temperatures. These fluctuations, in turn, seem to have helped drive the recent revival of Atlantic hurricanes, the drying of the Sahel in the 1970s and '80s, and the global warming of the past few decades, among other climate trends.

The ocean conveyor “is an important source of climate variability,” says meteorologist James Hurrell of the National Center for Atmospheric Research in Boulder, Colorado. “There's increasing evidence of the important role oceans have played in climate change.” And there are growing signs that the conveyor may well begin to slow on its own within a decade or two, temporarily cooling the Atlantic and possibly reversing many recent climate effects. Greenhouse warming will prevail globally in both the short and long terms, but sorting out just what the coming decades of climate change will be like in your neighborhood could be a daunting challenge.

Wobbly ocean.

North Atlantic temperatures have wavered up and down at a roughly 60- to 80-year pace.


Researchers agree that the North Atlantic climate machine has been revving up and down lately (Science, 16 June 2000, p. 1984). From recorded temperatures and climate proxies such as tree rings, researchers could see that temperatures around the North Atlantic had risen and fallen in a roughly 60- to 80-year cycle over the past few centuries. This climate variability was dubbed the Atlantic Multidecadal Oscillation (AMO). Ocean observations suggested that a weakening of the ocean conveyor could have cooled the Atlantic region and even the entire Northern Hemisphere in the 1950s and '60s, and a subsequent strengthening could have helped warm it in the 1980s and '90s. But the records were too short to prove that the AMO is a permanent fixture of the climate system or that variations in the conveyor drive the AMO.

Taking the long view, climate modeler Jeff Knight of the Hadley Centre for Climate Prediction and Research in Exeter, U.K., and colleagues analyzed a 1400-year-long simulation on the Hadley Centre's HadCM3 model, one of the world's leading climate models. The simulations included no changes in climate drivers such as greenhouse gases that could force climate change. Any changes that appeared had to represent natural variations of the model's climate system.

At April's meeting of the European Geosciences Union in Vienna, Austria, Knight and colleagues reported that the Hadley Centre model produces a rather realistic AMO with a period of 70 to 120 years. And the model AMO persists throughout the 1400-year run, they note, suggesting that the real-world AMO goes back much further than the past century of observations does. The model AMO also tends to be in step with oscillations in the strength of the model's conveyor flow, implying that real-world conveyor variability does indeed drive the AMO.

Such strong similarities between a model and reality “suggest to me it's quite likely” that the actual Atlantic Ocean works much the same way as the model's does, says climate modeler Peter Stott of the Hadley Centre unit in Reading, who did not participate in the analysis. Hadley model simulations also support the AMO's involvement in prominent regional climate events, such as recurrent drought in North East Brazil and in the Sahel region of northern Africa, as well as variations in the formation of tropical Atlantic hurricanes, including the resurgence of such hurricanes in the 1990s.

On page 115, climate modelers Rowan Sutton and Daniel Hodson of the University of Reading, U.K., report that they could simulate the way relatively warm, dry summers in the central United States in the 1930s through the 1960s became cooler and wetter in the 1960s through 1980s. All that was needed was to insert the AMO pattern of sea-surface temperature into the Hadley atmospheric model. That implies that the AMO contributed to the multidecadal seesawing of summertime climate in the region.

Bad warmth.

The AMO's warm years favor more U.S. hurricanes (right).


If the Hadley model's AMO works as well as it seems to, Knight and his colleagues argue, it should serve as some guide to the future. For example, if North Atlantic temperatures track the conveyor's flow as well in the real world as they do in the model, then the conveyor has been accelerating during the past 35 years—not beginning to slow, as some signs had hinted (Science, 16 April 2004, p. 371). That acceleration could account for about 10% to 25% of the global warming seen since the mid-1970s, they calculate, meaning that rising greenhouse gases haven't been warming the world quite as fast as was thought.

Judging by the 1400-year simulation's AMO, Knight and colleagues predict that the conveyor will begin to slow within a decade or so. Subsequent slowing would offset—although only temporarily—a “fairly small fraction” of the greenhouse warming expected in the Northern Hemisphere in the next 30 years. Likewise, Sutton and Hodson predict more drought-prone summers in the central United States in the next few decades.

But don't bet on any of this just yet. The AMO “is not as regular as clockwork,” says Knight; it's quasi-periodic, not strictly periodic. And no one knows what effect the strengthening greenhouse might have on the AMO, adds Sutton. Most helpful would be an understanding of the AMO's ultimate pacemaker. In the Hadley Centre model, report modelers Michael Vellinga and Peili Wu of the Hadley Centre in Exeter in the December Journal of Climate, the pulsations of the conveyor are timed by the slow wheeling of water around the North Atlantic. It takes about 50 years for fresher-than-normal water created in the tropics by the strengthened conveyor to reach the far north. There, the fresher waters, being less dense, are less inclined to sink and slide back south. The sinking—and therefore the conveyor—slows down, cooling the North Atlantic and reversing the cycle.

That may be how the Hadley AMO works, says oceanographer Jochem Marotzke of the Max Planck Institute for Meteorology in Hamburg, Germany, but it doesn't settle the mechanism question. How a model generates multidecadal Atlantic variability “seems to be dependent on the model you choose,” he says. Before even tentative forecasts of future AMO behavior are taken seriously, other leading models will have to weigh in, too.

Navigate This Article