Smack dab between eastern Canada’s Misery Point and Greenland’s Cape Desolation is a place where the thrashing of the Atlantic Ocean’s churn sounds about as friendly as the nearby place names. This stretch of water, the Labrador Sea, has long been considered a critical junction in the global circulatory system of the world’s oceans. By pumping warm water north and cool water south, the system regulates the planet’s climate.

For decades, scientists have turned to the Labrador Sea to understand how ocean processes there may be affecting the strength of a massive oceanic conveyor belt known as the Atlantic Meridional Overturning Circulation. Many researchers have found that the system is weak, which could spell trouble for changing climate conditions in the future. But data emerging from new suites of ocean-monitoring instruments suggests this narrative is headed for a twist.

The idea of a faltering AMOC entered public consciousness years ago with the popular but exaggerated catastrophe flick The Day After Tomorrow. In it, the AMOC grinds to a halt, causing superstorms to ravage entire cities and mega hurricanes to suck frozen air down from space.

That’s not going to happen. But if the AMOC continues to lose strength, or even temporarily shuts down (as some studies have suggested), it could mean cooler winters and summers in Europe and other regions around the North Atlantic coast. Great Britain might see crop production plummet, according to one study, and the ocean could end up sequestering less carbon, which would leave more heat-trapping CO2 in the atmosphere and trigger faster warming elsewhere.

Given the direct impacts the AMOC can have on our climate, scientists have been trying to assess its circulatory strength mainly through the use of computer models, a generally reliable tool in climate science. But with the AMOC specifically, models often don’t work as well. They fall short, scientists say, when trying to simulate important but small-scale ocean processes. For example, ocean eddies–swirling currents that can bring deep waters to the surface and strengthen the system’s overturning (the “O” in AMOC)–are difficult to simulate accurately. Martha Buckley, a climate scientist at George Mason University, says it’s mainly a problem of scale. “Certain ocean processes cannot be explicitly represented by models due to resolution constraints,” she says.

An underlying issue is that the ability to measure what’s happening in the ocean–not simulate it–has been extremely limited. “Observations of the AMOC are sparse, and scientific knowledge is mainly based on model simulations,” wrote Monika Rhein, a scientist at the University of Bremen, in Germany, in the journal Science last year.

That’s starting to change.

Scientists are increasingly looking to sensor-based instruments that monitor the ocean at various depths to take the pulse of the AMOC. And in some cases, these instruments’ observations are serving up dramatic twists to the reigning AMOC narrative. “We’re seeing signs through direct ocean observations that the AMOC is not as weak as some scientists are suggesting,” says Igor Yashayaev, a marine scientist at the Bedford Institute of Oceanography.

As part of an ongoing ocean monitoring program, he and his colleagues observed that during a series of colder-than-average winters between 2010 and 2018, surface waters in the Labrador Sea became denser and more voluminous as they surrendered their heat to the atmosphere. These heavier surface waters then sank deeper into the ocean, pushing up less-dense waters from below. It was as if the ocean was turning upside down. As a result, the conveyor belt’s throttle was notched higher.

Yashayaev and his team based these observations on data collected with instruments known as Argo floats. The floats, which look like human-sized yellow syringes, autonomously nose-dive thousands of meters into the abyss and then resurface, measuring water temperature and salinity along the way. The measurements provide snapshots in time that allow him and his team see how far surface waters end up sinking, “and thus infer overturning strength,” he says.