Greenland meltwater could cut the Atlantic Meridional Overturning Circulation (AMOC) by an extra 10% to 20% by 2100, and by about 40% by 2300, according to new modeling. The big twist: in this setup, that weakening does not turn into a permanent, irreversible shutdown. The circulation weakened strongly, yes, but it recovered when the team switched off the meltwater or stopped greenhouse gas emissions.

That matters because AMOC is not just an ocean current. It is a climate system with a job description: moving heat northward and returning cold water southward. The study frames Greenland meltwater as a “freshwater pulse” that makes the North Atlantic harder to run in its usual mode. When fresh water and warmer conditions prevent deep currents from forming, the AMOC loses strength. But the model’s recovery results are the counterweight to the worst-case narratives floating around climate risk.

To understand why executives, boards, and planners should care, zoom out: AMOC regulates climate patterns that show up in insurance pricing, infrastructure design, energy demand, agriculture planning, and coastal risk. The paper ties Arctic and Greenland melt to a backdrop where many scientists already expect AMOC to be at its weakest in more than 1,000 years due to Arctic ice melt and soaring ocean temperatures. In that context, the study adds another layer: Greenland runoff may not be as instantly disruptive on shorter timescales as Arctic melt, but it ramps in later centuries. Von Hardenberg put it plainly, saying Greenland starts melting strongly only after the year 2100, and the meltwater becomes extremely strong as the simulations go farther, out to 2300.

Here is how the researchers ran the experiment, and why the methodology itself is part of the story. They used a global climate model called EC-Earth3, which does not include the Greenland ice sheet. To explore Greenland’s role, they added the ice sheet from CESM2 to EC-Earth3. Crucially, they ran comparisons: one set without Greenland and another with Greenland. That “with versus without” approach is what lets them clearly reveal the ice sheet’s added weakening effect on AMOC.

The mechanism is also central to the conclusion. Meltwater from the Arctic and Greenland slows the AMOC by preventing the formation of deep currents in the North Atlantic that drive the circulation south toward Antarctica. Deep currents form when masses of cold, salty water sink to the seabed. Fresh water works against that sink by diluting and warming surface waters, making it harder for those dense water masses to form. In other words, the Atlantic is not “stopping” because of a single switch. It is being starved of the conditions that keep the loop running.

The study’s most consequential claim is the one the title promises to deliver: strong weakening without a shutdown. Unlike other studies that have suggested the AMOC could collapse irreversibly under climate change, these simulations recovered once meltwater was turned off or greenhouse gas emissions were reduced or stopped. That recovery implies the AMOC is more stable than previously thought, at least within this modeling framework. Von Hardenberg described the result as “consistent with a strong weakening, but not a shutdown,” and the paper points to the same theme: the circulation does not suddenly flip from active to collapsed in this version of the scenario.

For decision-makers, this should not read like a free pass. The model still points to ongoing consequences. It indicates AMOC weakening continues for decades, with potential knock-on effects including freezing weather in Northwest Europe, additional sea-level rise along the U.S. East Coast, and droughts around the equator. Those are not abstract effects. They map to real-world systems: ports, power grids, coastal defenses, water systems, and food supply chains. Even if the current loop does not irreversibly collapse, partial weakening can still mean prolonged stress, higher variance, and expensive adaptation.

The study also lands in a policy and modeling reality that matters to regulators and risk teams: it highlights uncertainty drivers. Greenland meltwater is a freshwater input that some climate models currently do not take into account, partly because representing Greenland melt correctly requires a specific, dynamic model for the Greenland ice sheet, which is resource-intensive. In this work, the authors and von Hardenberg argue that adding Greenland improves constraint on future AMOC uncertainty. They plan to include this component in the next version of EC-Earth3 to better narrow how much AMOC might decrease across different scenarios.

Outside experts agree the integration step is valuable, while warning against over-generalizing a single model outcome. Jonathan Baker, a senior climate scientist at the Met Office in the U.K., said the study is particularly interesting because it directly tests a process highlighted as an important source of uncertainty in future AMOC projections. Baker also emphasized that whether the system could collapse remains uncertain since the team used only one model. Nicholas Foukal of the University of Georgia’s Skidaway Institute of Oceanography said he finds the result robust in that it comes from a high-resolution and realistic coupled climate model, and that it aligns with other modeling results requiring immense amounts of fresh water to trigger irreversible change. But Sybren Drijfhout, University of Southampton, pushed back on the idea that the outcome should be treated as exemplary for the whole model ensemble, saying the parameters make the AMOC more resilient than in other models, and calling for replication across a wider range of models.

This is where the board-level implication shows up. If you manage long-term assets or advise on climate risk disclosures, your biggest enemy is not just catastrophic worst-case events. It is decision paralysis caused by uncertainty you cannot bound. This study offers a narrower bound in one direction: in at least one modeling framework, Greenland-driven weakening does not necessarily mean an irreversible shutdown. But it simultaneously raises the bar for rigor: repeating the experiments across multiple models is described as important to assess robustness. Translation: the market will reward teams that can treat AMOC risk as scenario-based and continuously updated, not as a single fate event.

In the near term, the strategic stake is simple. Researchers should include the Greenland ice sheet component to better constrain uncertainty, and other experts are urging replication to determine how robust the recovery behavior is. For executives in adjacent risk worlds, that means planning, financing, and adaptation strategies should be resilient to continued weakening through decades, while also tracking which parts of the uncertainty range appear to shrink as models improve.