The North Atlantic cycle

Discover, January 31, 1995

The ghost of George Callendar haunts today’s global warming experts. In 1938 Callendar, a British coal engineer by trade and a climatologist by avocation, demonstrated that global temperatures had been rising for the previous 50 years. “He realized that it was most likely due to the burning of fossil fuels, and he predicted global warming from there on out,” explains Michael Schlesinger, a climatologist at the University of Illinois. “And then the warming ceased; there was a cooling period from about 1940 until the 1970s.”

Since then the climate has begun to warm again, and many climatologists now believe that Callendar was fundamentally right. “But the question becomes, ‘Well, if the warming trend up to 1940 was the greenhouse effect, then why didn’t we continue to see it?'” says Schlesinger. Last February he reported that he may have found the reason. The North Atlantic, it appears, acts as a tremendous heat pump that cyclically warms and cools the atmosphere over the course of decades, sometimes reinforcing and sometimes counteracting the greenhouse effect.

Schlesinger and his student Navin Ramankutty modeled the warming created by the emission of carbon dioxide and other greenhouse gases over the past 130 years, as well as the cooling created by sulfates, which are also released by burning fossil fuels. When they subtracted their results from the actual record of atmospheric temperatures, only natural fluctuations remained. But far from being random variations, they formed a 70-year cycle. And when the researchers broke the global data down into 11 different regions, says Schlesinger, “we found the change occurs overwhelmingly in the North Atlantic and the adjacent continental regions. In the North Atlantic, it is huge”–swinging 1.2 degrees Fahrenheit over the course of 70 years.

A simulation of Atlantic Ocean currents done recently by Thomas Delworth and his colleagues at the Geophysical Fluid Dynamics Laboratory in Princeton may explain how the cycle is created. Warm water flowing north from the equator–in the Gulf Stream, for example–cools and then sinks into the deep sea near Greenland, pulling more warm water north to take its place, like a conveyor belt. By warming the air above the North Atlantic, the conveyor belt keeps northern Europe relatively warm.

The GFDL simulation shows, though, that the conveyor doesn’t run smoothly. The phenomenon has to do with salt, which, like cold temperatures, makes water dense. The east Atlantic is salty because the shallow Mediterranean flows into it, while the west is relatively fresh. As the conveyor belt pushes warm water into the North Atlantic, it creates the marine equivalent of a high-pressure region, and the clockwise flow out of that region tends to draw in fresh western water. This relatively light water doesn’t sink as easily, and so it slows the belt down. With the northward flow of heat reduced, though, the water in the North Atlantic cools enough to start sinking again–and with that the region becomes a low-pressure zone whose counterclockwise flow draws salty water in from the east. That water sinks easily, speeding the conveyor belt back up again. According to the GFDL simulation, this cycle takes anywhere from 40 to 60 years.

That’s a little shorter than the cycle Schlesinger found. It’s still not clear whether the researchers are looking at the same thing–or whether the 70-year rhythm Schlesinger thinks he can see in only 130 years of data is real. (Recent work supports him; Natesan Mahasenan of Carnegie Mellon says he’s found the cycle in ice cores and coral reef cores that go back 300 years.) “The good news is that if what we have found is what Delworth at GFDL has found, then it brings some climate predictability with it,” says Schlesinger. “We may be able to forecast what it’s going to be like in the next century.” That would finally put Callendar’s ghost to rest.