Wires in the Irminger Sea

The Forecast on Trial

Eli: July 2023. Two Copenhagen researchers, Peter and Susanne Ditlevsen, publish a paper in Nature Communications. It says the Atlantic's great overturning circulation, the AMOC, is likely to pass a tipping point this century. A tipping point here is the moment when slow change in the ocean flips into an abrupt, self-reinforcing one. Their window opens in the late 2030s and closes just past 2100. The likeliest moment: around mid-century.

Mara: A listener put the question to us plainly: what would happen if the Gulf Stream shut down? Today's angle is an ocean forecast on trial.

Eli: So we follow that date. Into the physics of what's actually moving. Down the wires of moorings at sea, where people like the hydrographer Sarah Nickford work. And into what models say a collapse would do. Whether the date holds stays open.

Mara: Mid-century was the headline. Underneath, the estimates shifted with each choice of data.

Eli: So the deadline everyone quoted, Mara, sits inside a spread of a thousand years and more. That spread is what's on trial.

The Surface Fingerprint

Mara: Start with what's actually moving. The Gulf Stream is pushed mostly by wind over the North Atlantic. Earth's rotation keeps the biggest currents pressed to the western side of the ocean, and in the Atlantic they combine into the Gulf Stream. It carries more water than the overturning does.

Eli: So the picture of the Gulf Stream just stopping doesn't fit. It's expected to keep flowing even if the overturning collapses.

Mara: Right. The overturning is a different machine. It carries warm surface water north. Up there the water cools, gets denser, sinks, and heads back south as cold deep water. Much of that runs on thermohaline circulation: flow driven by differences in temperature and salt.

Eli: So cold, salty water is heavy, it sinks, and the sinking keeps the loop turning.

Mara: And it works in reverse. Warming brings more rain and meltwater from Greenland and the Arctic. Warmer, fresher water is lighter, so it doesn't sink as deep. A scientist at Britain's National Centre for Atmospheric Science put the result simply: sinking takes longer or happens less, and the overturning slows. Winds feed in too, so it's not only heat and salt.

Eli: So a current can keep flowing while the larger system delivers much less warmth north.

Mara: And that larger system is what the 2023 paper tried to time. The Ditlevsens worked from a fingerprint: sea-surface temperature in the subpolar North Atlantic, with twice the global average warming subtracted out.

Eli: That's a surface thermometer standing in for a loop that reaches the sea floor.

Mara: Then they read how it behaves. A system nearing a tipping point gets slower to recover from every knock. That's critical slowing down. It shows up as bigger swings, and as each year looking more like the one before.

Eli: And those warning signs did show up in a full climate model fed steadily more freshwater.

Mara: Here's the piece to hold onto. To turn those signs into a date, the study fitted the fingerprint to a curve that assumes the circulation is heading for a tipping point. A review on the RealClimate blog called those assumptions very plausible for a first-order estimate. So the date rests on a surface fingerprint, read through a curve that already expects the tip.

A Snapped Float

Eli: Now the other way of knowing. During the pandemic, science parties on research cruises were cut in half. Sarah Nickford became an alternate for an OSNAP cruise in the subpolar Atlantic. She did fourteen days of isolation. Then she was asked to isolate again and join the next one, in the Irminger Sea, as a hydrographer.

Mara: Fond memories of an earlier month crossing the North Atlantic helped her say yes.

Eli: Aboard, the first science job was a shallow test cast, so she could practise water sampling with the chief scientist. A cast lowers a probe that profiles temperature, salinity and oxygen. Cruises use those profiles to calibrate the moored instruments and check them for drift.

Mara: So the moorings get checked against the water itself, not against a proxy.

Eli: In December 2024, on another cruise, a crew is hauling in a mooring called WB2: a line held up by floats, instruments along its length. During recovery, WB2 parted at the surface, between its top two floats. The crew grappled it again at the second float and carried on.

Mara: Lines like that make up RAPID, strung across 26.5 degrees north, between the Bahamas and the Canary Islands, measuring continuously since 2004. Sensors bolted to wires stretch thousands of metres down, recording current, pressure, temperature and conductivity.

Eli: Still, those are readings at points. The overturning is a whole ocean.

Mara: So they get stitched together. Density differences across the line give the deep flow. Wind data adds the surface push. And a NOAA cable in the Florida Straits tracks the Gulf Stream through its voltages.

Eli: Nickford's array covers the north. OSNAP started in 2014, running from Canada to Greenland and Greenland to Scotland.

Mara: Then the unit. Oceanographers count in sverdrups, and one sverdrup is a million cubic metres of water every second. At RAPID the average is about seventeen, and three other arrays land close by. The heat riding north between Florida and the Canaries comes to about 1.2 petawatts, a petawatt being a million billion watts. Ben Moat, who leads RAPID at the UK's National Oceanography Centre, says that's “equivalent to a million power stations.”

Eli: And it all rests on lines like the one that parted at the surface that December.

Mara: Twenty years of that number now exist, Eli. That's the record the forecast's clock has to answer to.

Waiting for 2034

Eli: September 2025. The RAPID team releases its latest time series. Ben Moat reported the weakening had held steady since the last update. About 0.9 sverdrups a decade, give or take 0.7, over twenty years of observations.

Mara: Out of about seventeen. And the margin is nearly as wide as the trend.

Eli: The record turned up something nobody expected, too. The overturning swings over days, over years, up to a decade, driven largely by wind. The old picture was a slow, steady conveyor run by heat and salt.

Mara: Which is the noise any long-term trend has to rise above.

Eli: That's Moat's point. Direct measuring started in 2004, and that's too short to separate human influence from natural swings. He's called for observations through at least 2034. Whether the slowdown is us or the ocean's own rhythm is still unknown.

Mara: So the record has a deadline of its own: 2034.

Eli: Then, in April 2026, a team pooled four mooring arrays down the western edge of the Atlantic, from about 16 degrees north to about 42. They used changes in pressure on the sea floor to estimate the flow below about a thousand metres. That's the deep western boundary current, the narrow deep road cold water takes back south. The team reported a consistent decline across all four.

Mara: But the same work flags the eastern side moving the opposite way. Whether that pattern holds across the whole Atlantic, and why, is unclear.

Eli: So it's a measured decline in one branch, not a measured collapse of the whole. And that's where the readings part. Peter and Susanne Ditlevsen, with researchers like Niklas Boers and René van Westen, read early-warning signals as the circulation losing resilience, with a tip possible within decades.

Mara: The other reading belongs to Ben Moat, Laura Jackson, Raffaele Ferrari and the IPCC. They see strong variability and no proof of a collapse under way. The IPCC's latest assessment says the overturning will very likely weaken this century in every scenario. Under low emissions, by about 24 percent by 2100. Under very high emissions, about 39.

Eli: Weaker in both. Neither one comes with a date.

Mara: And in August 2026 a modelling study moved the question. René van Westen told the Dutch broadcaster NOS, “The collapse occurs due to the rate of warming rather than reaching a specific temperature.” In their runs, slow forcing let the circulation hold on, and fast forcing tipped it. Femke de Jong put it bluntly in the same report: there's no temperature below which the risk can be ruled out.

Eli: So testing the forecast's clock is one question. What a collapse would actually do, if it came, is another, and that's where the models get specific.

A Shift in the Rain

Mara: Start with one modelled world: about two degrees of warming, plus a full collapse. That's beyond the weakening the IPCC projects. North-west Europe gets brutal winters. One-in-ten-year cold snaps approach minus twenty in London. One scientist said those winters would be like living in an ice age, while summer extremes barely change. But this is the climate two hundred years on, once everything has settled. Another scientist noted it doesn't tell us how a collapse would hit immediately.

Eli: In the four-degree run, warmer seas keep sea ice from spreading south, so less cooling.

Mara: And the heat doesn't vanish. Without the overturning, more of it stays in the tropics and the Southern Hemisphere, warming them.

Eli: Then the coasts. Sea level isn't flat like a bathtub. Currents tilt it, and that's dynamic sea-level rise: water piling up or dropping because of how the ocean moves. In the Netherlands, researchers warn of higher seas and more storm surges.

Mara: And there's a measured glimpse. Over 2009 and 2010, the overturning dipped by about thirty percent. A 2015 study named that as one of two factors behind a jump of about thirteen centimetres north of New York City.

Eli: A dip, and one of two factors. A full collapse reaches further. The IPCC says it would very likely push the tropical rain belt south. It could weaken the African and Asian monsoons and dry Europe, bringing drought to some places and wetter weather to others. Some of those regions rely on monsoons for up to ninety percent of their yearly rain, and on that rain, food for millions.

Mara: So the stakes run from London winters to harvests half a world away.

Eli: And it has happened before. Ice cores show the overturning stopping and restarting during the ice ages. Near the end of the last one, meltwater entered the North Atlantic. In the Younger Dryas that followed, the Atlantic's heat pump stopped for about a thousand years.

Mara: The ancient record supplies a precedent for disruption, not a deadline for this century.

The Clock Dissolves

Eli: Back to July 2023, and that window. In 2026, a team of researchers re-examined the forecast. They concluded current observations and simplified statistical models can't reliably predict when the overturning might tip. Specific timelines, they said, shouldn't be treated as reliable or actionable. With a corrected fingerprint, the signal no longer passed the 2023 paper's own test for unusually high variance and year-to-year memory. And other statistical checks found that changing the data, or dropping the built-in assumption that a tip was coming, blew the window open to millennia, or infinity.

Mara: That doesn't show the real circulation can't collapse. Its timing is what's unresolved.

Eli: The reanalysis challenges the clock, not the existence of the climate risk. The wind-driven Gulf Stream can't simply shut down while the Earth turns. A collapse of the deeper overturning is the real worry, and models say it would chill north-west European winters, lift Atlantic coastal seas and push tropical rains south.

Mara: So the next tests are at sea. Moored arrays are, for now, the only way to know the overturning's state, and to check whether climate models get its swings right. In the deep-mixing seas, the Labrador, Irminger and Nordic, mixing has trended down over the last five to ten years. That could be an early warning, but natural variability still dominates it. And the four-latitude team says the eastern boundary has to keep being watched to capture the long-term trend.

Eli: So, three signals. A record past 2034 that shows whether the weakening outlasts the swings. Timing estimates that hold across datasets and assumptions, including models that allow no tip, checked against the moorings. And eastern-boundary data showing whether it offsets the western decline. None of them promises a date.

Mara: Whether the ocean keeps to that 2023 window will be read off moorings, from the cruises Sarah Nickford sailed on to the ones still to come.

Eli: Listeners can read into the AMOC tipping point.

Mara: Follow climate in more stories as researchers track how ocean currents change.

Eli: That's our story for today. I'm Eli.

Mara: And I'm Mara. Thanks for listening to Angle, and until next time.