Collapse of AMOC ocean current may already be locked in
Get the Energy newsletter
Daily energy & climate — solar, EVs, oil, the policy fights and tech bets shaping the transition. Free.
- Holden, Lenton and colleagues ran 21 computer simulations of the AMOC and estimated a 10 to 23 percent chance that the collapse of the Atlantic Meridional Overturning Circulation is already committed, depending on Greenland ice melt and emissions assumptions.
- Under conservative assumptions — emissions peaking in 2025 and Greenland ice adding 54 mm to sea levels by 2100 — the model puts the chance of locked-in collapse at 10 percent; under less conservative projections of 274 mm of melt, the probability rises to 23 percent.
- If the world delays net zero until 2100, the model predicts an 80 percent chance of committed AMOC collapse, demonstrating how rapidly the risk escalates with slower emissions cuts.
- Even when collapse is committed, the simulations show an average 84-year delay before it actually occurs, with the earliest collapse arriving around 2060; delaying net zero by 10 extra years past the commitment point would shorten that delay to 57 years on average.
- The model used a coarse 5° grid resolution, a deliberate choice due to limited computing power; Met Office scientist Jonathan Baker cautioned that this low resolution may make the model less sensitive than higher-resolution models, potentially distorting the risk estimates.
- Tim Lenton at the University of Exeter argued the findings make the case for reaching net zero as fast as possible to hold probability at roughly 10 percent, while Till Wagner at the University of Wisconsin-Madison said the evidence for weakening is solid but the large-scale dynamical outcome remains uncertain.
- Recent research also hints that AMOC slowdown may be reversible if carbon dioxide emissions come down enough, a finding Lenton said chimes with the new model's call for rapid action.
- The study was published on EarthArXiv (DOI: 10.31223/X5N48Q).
Why it matters: Europe's climate stability depends on the AMOC, and this model suggests the window for preventing its collapse may already be partly closed — the probability of locked-in collapse jumps from 10 percent under a 2025 emissions peak to 80 percent if net zero is delayed until 2100. Even so, the 84-year average lag between commitment and actual collapse means rapid emissions cuts could still slow the onset, giving policymakers a concrete urgency argument: every decade of delay compounds irreversible risk.
