A Weakening Atlantic Current Could Let the Planet Hold More Heat

Think of the North Atlantic as a drain in the floor of a very large bath. When the Atlantic Meridional Overturning Circulation, or AMOC, is running strong, that drain is wide open: heat gathered across the tropics is hauled north and dumped into the sky over Greenland and Iceland, and the planet as a whole cools. Slow the current down and the drain narrows. A new modelling study finds that when the AMOC weakens, the heat does not vanish or shuffle south, as textbooks have long implied. It stays, banking up inside the ocean, so the planet’s total heat actually climbs even as the North Atlantic itself turns colder.

That’s a counterintuitive result, and it upends the tidy story most of us learned. The published account of AMOC weakening has always been a regional one: Europe, Greenland, the eastern seaboard of North America plunged into cold while, supposedly, the southern hemisphere warmed to compensate.

What the Old Seesaw Left Out

For decades that compensation had a name, the thermal bipolar seesaw, and a mechanism: heat carried across the equator by the current, tipping from one hemisphere to the other like weight on a plank, as a Nature analysis of the idea describes it. It is an elegant picture. The trouble, according to the team behind the new work, is that it books the heat as merely moved rather than counting whether the planet gained or lost any. Christo Buizert, a paleoclimatologist at Oregon State University who led the study, and his colleagues went back to the global energy budget instead, tracking every joule in and out at the top of the atmosphere.

Their reframing is almost plumbing. “The AMOC works like a heat valve that controls the energy budget of the planet,” Buizert says. Open the valve and heat drains from the ocean and radiates away to space; close it and the heat is trapped below.

To test the idea the researchers leaned on three separate climate models (CCSM4, MIROC4m and COSMOS) that each produce, spontaneously, the abrupt ice-age lurches known as Dansgaard-Oeschger events. These are the closest thing the palaeoclimate record has to a tipping point, widely treated as an archetypal example of abrupt climate change, sudden flips between cold and warm states that recurred throughout the ice ages stretching from about 11,700 years ago back some 2.7 million years. In all three models, a strong current drained heat from the whole ocean; a weak one let it pool.

“It’s as if the whole ocean acts as a giant bucket of heat,” Buizert says, with the AMOC as the spigot deciding how much pours out.

Where the Heat Actually Goes

The mechanism, in the simulations, runs like this. Ocean water soaks up sunlight mostly in the tropics. A vigorous AMOC ferries that warmth to the subpolar North Atlantic, where wintertime deep convection vents it to the atmosphere and, ultimately, out through the top of the atmosphere as the region warms and radiates harder. Weaken the current and the venting slows. Only a thin skin of North Atlantic surface water cools; underneath, and across the rest of the ocean basins, heat accumulates, a pattern separate modelling work ties to stronger stratification and weaker deep convection in the subpolar gyre. The models tie this to the ice cores neatly: Antarctic temperature tracks the ocean’s total heat content, while Greenland temperature tracks the rate at which the North Atlantic is shedding heat. That is why the two poles seem to move in a coupled, offset rhythm without any need for heat to physically slosh between them.

There is a caveat worth flagging, and the team is upfront about it. One of the three models, COSMOS, fails to reproduce realistic temperature swings over Antarctica, so the Antarctic side of the story rests on the other two.

Everything here is also a model result under ice-age conditions, not a measurement of the world we live in now. The authors are careful to say the glacial past, with no strong human forcing in play, is a simpler and imperfect stand-in for the modern case.

What It Means for a Warming World

Still, the implication for the future is hard to shrug off. Many models expect the AMOC to weaken as the climate warms, and if the valve framework holds, a throttled-down current would mean the planet banks more heat rather than less. Buizert offers a yardstick for the scale involved: he reckons a weak-AMOC event during the last ice age produced warming equivalent to roughly 25 parts per million of carbon dioxide today, about ten years of human emissions. That figure comes from him rather than from the paper itself, so treat it as a way to picture the magnitude rather than a precise ledger entry.

There is a silver lining tucked in the CO2 experiments, though. Across a run of eight simulations spanning a range of ice-age carbon dioxide levels, a warmer world made the AMOC steadier, hinting that the violent Dansgaard-Oeschger flips might not return, and that a future weakening could recover rather than collapse for good. “An irreversible collapse of the AMOC might not occur,” Buizert says, before adding the line every honest scientist adds: more research is needed to pin down the current’s future stability. Either way the bath is filling from below, and the valve we have been watching for signs of regional cold may be the thing to watch for global heat.

FAQ