Americans who are concerned about climate change have long found themselves in an unenviable position: They have to debate about the existence of a debate.

For about two decades, the vast majority of climate scientists have agreed that human industrial activity is forcing the planet to warm. For about as long, some doubters have argued that this consensus is nonexistent or premature—and that, despite repeated studies identifying it, media attempts to report on the consensus constitute so much liberal bias.

These fights will likely be recapitulated this month. Scott Pruitt, the attorney general of Oklahoma and President-elect Donald Trump’s nominee to lead the EPA, has invested a lot of time in fighting the Obama administration’s climate and environmental regulations. He has not, however, said very much on the record about climate change.

One of his only quotes on the matter appeared in a National Review editorial last year. “Scientists continue to disagree about the degree and extent of global warming and its connection to the actions of mankind,” he wrote, in an article co-authored with Alabama attorney general Luther Strange.

The problem is: Not all of this sentence is true. While scientists continue to explore the consequences of climate change, there is essentially no debate among scientists about global warming’s “connection to the actions of mankind.”

Nor has there been a debate for years. Since at least 1995, the balance of evidence in climate science has indicated that human-caused greenhouse-gas emissions are behind the planet’s warming. Agreement on this question has only strengthened since. By 2012, an international panel of leading researchers in the field said there was at least a 95 percent chance that human activity has caused global warming since 1950.

There are active discussions in climate science—they’re just not about this. So before we all have to talk about a topic on which there is near total scientific agreement, I thought it might be fascinating to examine a real area of dispute in the field. And one of the most consequential disagreements is about something called the Atlantic Meridional Overturning Circulation, or AMOC.

* * *

In 1769, Ben Franklin became the first person to map the Gulf Stream when he made the above chart. (NOAA)

Many Americans know AMOC as the Gulf Stream: the warm, surface-level current in the Atlantic Ocean that hugs the East Coast. You may have seen it in the old map by Ben Franklin, pictured above: It flows up the Carolinas, passes by New England and Nova Scotia, and then veers toward Europe. Eventually it arrives near the British isles and northwestern Europe.

The Gulf Stream is part of a much larger system, however. As that warm water flows northeast, it gradually cools, and in cooling, compresses and sinks. Eventually, in the Labrador and Greenland Seas, it becomes dense enough that it plunges down thousands of meters into the deep ocean. There it becomes a new current, running back south. It can remain in this deep-ocean current for many years until it eventually upwells at the equator or in the Southern Ocean.

This global conveyor belt of water is AMOC, and it is critical to the world’s climate. (Most scientists pronounce it as AY-mock.)

When AMOC is strong, it sends millions of cubic meters of ocean water north every day. A strong AMOC seems to shape the entire planet’s climate systems. It moderates the intensity of Atlantic hurricanes, lessens the risk of drought in North America, and assures the health of monsoons in India. AMOC also ferries warm weather from the equator to Western Europe, where it helps bring the region unusually mild winters. (Consider that temperate Berlin is about as far from the equator as the snowy Chilean city of Punta Arenas.)

Crucially, the entire AMOC system depends on cool, dense water “overturning” in the northwest Atlantic Ocean. Without cooled water plunging into the deep ocean near Greenland, and turning back south, the entire conveyor belt will stop.

About 30 years ago, climate researchers became concerned that AMOC could suddenly shut down as a result of anthropogenic climate change. The “paleoclimatic record”—that is, what the planet’s geology and fossil record reveal of previous global climates—showed that the AMOC has rapidly collapsed in the past. “Rapidly” here means “within the span of a human lifetime.”

The crumpling of AMOC could potentially cause big problems for the global economy. AMOC’s disappearance would quickly worsen sea-level rise on the U.S. East Coast and subject the Southeast to unusually intense tropical storms. It could upheave agriculture in India, Europe, and the African Sahel.

But as climate models improved, those fears dissipated. “No current comprehensive climate model projects that the AMOC will abruptly weaken or collapse in the 21st century,” wrote a team of NOAA researchers in 2008. “We therefore conclude that such an event is very unlikely.”

Thomas Delworth was the lead author of that report. Delworth is a researcher at the NOAA Geophysical Fluid Dynamics Laboratory and a professor of atmospheric and oceanic science at Princeton University. He says that scientists are now re-examining those old conclusions.

“Some recent work now is challenging that consensus. It suggests that the real climate system may be less stable than [the models] think,” Delworth told me.

The most attention-getting of this work: a paper last year by James Hansen and 18 other scientists that argued the AMOC’s collapse could threaten global civilization this century. The paper built on older work showing that huge injections of freshwater have historically destabilized AMOC, essentially by flooding the Atlantic with cold water and screwing up its finely tuned density cycle. Hansen and his colleagues argued that as the Greenland ice sheet melts, it would be able to provide exactly such a pulse—and that, crucially, climate models failed to account for this physical process.

The paper made headlines around the world. Though he now is a professor at Columbia University, Hansen led the NASA Goddard Institute for Space Studies for more than three decades. In 1988, he became one of the first scientists to warn Congress of the dangers of global warming.

This week, the consensus on AMOC was challenged again. A team of researchers have showed in Science Advances that a popularly used climate model may significantly overestimate the stability of AMOC. Once you account for this bias, AMOC proves much more likely to collapse, they argue. And this collapse could happen without any freshwater injection from Greenland.

In other words, they show that the stress of global warming can push AMOC into collapse all by itself in at least one model. Freshwater doesn’t need to pour in from Greenland for AMOC to fall apart; simply increasing the temperature of the ocean can do it.

That’s because climate models make AMOC more stable than it actually is in nature, said Wei Liu, an oceanography researcher at Yale University and one of the authors of the study. “In a stable routine, if you increase the CO2, then AMOC only weakens. But in an unstable routine, if you add global warming, then AMOC will collapse by itself,” he told me.

He argues that field observations of the Atlantic Ocean suggest that AMOC is in fact unstable. Between mid-2009 and mid-2010, AMOC appeared to weaken, with the current carrying only two-thirds of its usual volume of water. At the same time, sea-level rise on the East Coast accelerated and Europe experienced an unusually frigid winter.

In their study, Liu and his colleagues tried to make their model more unstable. Most models, they say, do a bad job of representing AMOC. They don’t have enough salty water entering the Atlantic at the equator, and they also don’t have enough freshwater leaving it in the deep ocean.

In their experiment, they fixed this extremely crudely. Instead of fixing the underlying physics, they told the model to add much more saltwater and freshwater to the simulation. Then they doubled the amount of carbon dioxide in the simulated atmosphere, stepped back, and watched to see what would happen.

What happened is that, between year 200 and 300 of their adjusted model, AMOC rapidly collapsed.

Delworth said that even though their experiment was crude, it was revealing. “It’s a very interesting and provocative work,” he told me. “I think they are opening up this topic and saying our models may be too stable.”

“In this new study, they’ve just put a band-aid on [this stability]. They’ve said, if we alter these characteristics, the model is much less stable. But sometimes it’s really good to have these simple ad hoc techniques to address, ‘What’s the sensitivity of our models?’” he said.

The paper alone didn’t overthrow the consensus, he added, but it did suggest it should be re-examined.

Hansen, on the other hand, was more dismissive of the study’s approach. “You can’t fix the climate model simulation via ‘bias removal’—you should fix what is wrong with the model physics,” he said in an email. “They are doubling CO2, letting that change the temperature, rainfall, etc. and seeing what that does to the AMOC in their model. It’s been more than 35 million years since we had that much CO2 in the air, and sea level was more than 200 feet higher then. If we (humanity) are so stupid as to double CO2, you can count on the AMOC to shut down much faster than 300 years.”

Other climatologists, especially those who study Earth’s past, were much more positive about the paper, describing it as a necessary improvement to how we understand current climate models.

“This is an important step forward,” said Jean Lynch-Stieglitz, a professor of Earth and atmospheric sciences at Georgia Tech. “This study identifies a specific property of the climate models that would tend to make the AMOC in the models more stable than in reality.”

“Importantly, it reminds us that even if most climate model projections agree on their projections for future AMOC changes, that doesn’t necessarily mean that the projections are correct,” she added.

People visit the beach at Slea Had in Ventry, Ireland, on December 27. Ireland benefits from a strong AMOC. (Clodagh Kilcoyne / Reuters)

The instability of AMOC is one of the great open questions remaining in our understanding of climate change, one of the ongoing explorations into global warming’s “degree and extent.”

It is, in other words, an active debate in climate science.

But I want to highlight how it does and doesn’t look like a political debate. To my eye, it looks more like an open investigation: Researchers share their results, compare the models to the field observations, make tentative corrections to the software and underlying assumptions, and move chaotically together toward a deeper understanding of how the planet works.

Sometimes, they may disagree about how best to proceed or about the validity of any one study. But they do not disagree about the underlying chemistry and physics of their enterprise—all of which show that people are warming the planet through their industrial greenhouse-gas emissions.  

Martha Buckley, a research professor of oceanography at George Mason University, may have put it best.

“It is certainly a possibility that the AMOC is too stable in current [global climate models],” she said in an email.“The most obvious weakness of the paper is that the experiment is done for a single model.” She called the authors’ methods of fixing it “relatively crude.”

But then she went on. From what we know right now, the possibility that AMOC will shut down remains a “potential impact of climate change with significant consequences.”

“Yet other impacts are much more certain” to result from climate change, she said, listing “increased surface temperatures, sea level rise, and ice melt.” While some harms, like those of a collapsing AMOC, are still up for debate, it is almost completely certain that climate change will bring serious consequences for us in our time. Rampant drought, drier rivers, and vanished coasts are all ours to inherit.