Physics‐Based Indicators for the Onset of an AMOC Collapse Under Climate Change
- Hakan Sener
- Oct 22
- 4 min read
New physics-based warning signal derived from surface ocean measurements successfully predicts AMOC tipping across multiple climate models—suggesting collapse onset by 2055 (2023-2076) under high emissions or 2063 (2026-2095) under intermediate scenarios.
A new study led by René van Westen (Utrecht University) develops a robust early-warning indicator for Atlantic circulation collapse based on surface buoyancy fluxes over the North Atlantic, successfully tested across different forcing scenarios and 25 CMIP6 climate models.
The research reveals that when area-averaged surface buoyancy flux between 40°N-65°N switches from negative to positive, it signals imminent AMOC collapse—providing prediction decades before full shutdown occurs, with critical implications for European climate adaptation.
Key Findings: A Reliable Tipping Point Predictor
Surface flux sign change precedes collapse by ~50 years
When the surface buoyancy flux (B_flux) between 40°N-65°N changes from negative (ocean losing buoyancy) to positive (ocean gaining buoyancy), it indicates the Atlantic can no longer sustain deep water formation pathways connecting both hemispheres.
In CESM simulations under climate change, this sign change occurs when AMOC strength remains 7-9 Sv (compared to historical 14-17 Sv), providing early warning well before circulation reaches critically weak states below 4 Sv.
Indicator performs across forcing scenarios and models
The B_flux indicator successfully predicts collapse timing in quasi-equilibrium hosing experiments, freshwater pulse experiments, and transient climate change scenarios—demonstrating robustness across different forcing rates and mechanisms.
Analysis of extended CMIP6 simulations (to 2300) confirms that models showing B_flux sign changes before 2100 subsequently experience AMOC collapse, validating the indicator's predictive skill beyond the CESM.
22 of 25 CMIP6 models show collapse signals by 2100
Under SSP5-8.5 high emissions, 88% of analyzed models (22/25) exhibit B_flux sign changes before 2100, indicating widespread collapse trajectory across the model ensemble. Under intermediate SSP2-4.5 emissions, 64% (16/25) show the warning signal.
The multimodel mean indicates collapse onset by 2055 (25th-75th percentiles: 2023-2076) under SSP5-8.5 and 2063 (2026-2095) under SSP2-4.5, occurring at ~2.5°C global warming—substantially lower than previously estimated 4°C threshold.
Why Surface Fluxes Matter
Thermal forcing triggers initial weakening
The B_flux increase under climate change is dominated by thermal components—reduced ocean heat loss to warming atmosphere—rather than direct freshwater forcing. Over the 21st century, B_flux changes are 40-45% smaller under intermediate emissions than high emissions.
This thermal-driven B_flux increase triggers AMOC weakening, which then activates the salt-advection feedback: as circulation weakens, less salt is transported northward, freshening surface waters and further inhibiting deep convection.
Adiabatic pathways must be maintained
A functioning AMOC requires shared isopycnals (density surfaces) between Northern and Southern hemispheres, creating "adiabatic pathways" for interhemispheric water mass exchange. The quantity Ψ_NADW measures water supply to these pathways.
When Ψ_NADW approaches zero, it indicates no shared density surfaces remain, meaning deep circulation cannot be sustained regardless of AMOC strength—explaining why moderately strong circulations (8-12 Sv) can still be on tipping trajectories.
Regional Climate Consequences Under Collapse
Northwestern Europe faces drastic cooling
In CESM simulations where AMOC collapses under intermediate emissions (≤2°C global warming), Northwestern European temperatures drop substantially below global means, with cities experiencing persistent cooling through winter months.
Under high emissions, regional cooling partially offsets greenhouse warming, preventing European temperature increases that would otherwise occur. All collapse scenarios feature reduced precipitation over Europe and southward shifts in tropical rainfall belts.
Temperature threshold revised downward
The multimodel B_flux analysis suggests AMOC collapse becomes likely around 2.5°C global warming, substantially lower than the previous 4°C (±4°C range) threshold established from CMIP5 models.
Several models show early tipping signals before 1900, suggesting the 2.5°C estimate represents a lower bound—particularly concerning as this threshold approaches Paris Agreement targets and current models exclude accelerating Greenland meltwater contributions.
Why It Matters for Climate Risk Assessment
Actionable early warning system: Unlike statistical indicators prone to false positives, the physics-based B_flux signal provides mechanistic understanding of why collapse occurs, enabling confident risk assessment decades before full shutdown manifests.
Observable from surface data: B_flux requires only surface ocean temperature, salinity, and heat/freshwater flux measurements—already collected by reanalysis products—making near-real-time monitoring feasible without requiring deep ocean observations.
Extended simulations urgently needed: Only 20 CMIP6 model runs extend beyond 2100, yet collapse timescales exceed 100 years. The study demonstrates most models ending at 2100 are already on collapse trajectories, making extensions to 2200+ critical for next CMIP phase.
Immediate policy relevance: With collapse onset potentially occurring within decades under current emission trajectories, and Northwestern European climate changes being severe, the findings underscore urgent need for rapid emissions reductions toward SSP1-2.6 pathways.
A Predictable Tipping Cascade
This study transforms AMOC collapse from abstract model behavior to quantifiable near-term risk with observable early-warning signals, demonstrating that surface buoyancy flux monitoring can provide decades of advance warning before circulation shutdown.
The finding that thermal forcing—not just freshwater—drives the initial tipping trigger, combined with the lower-than-expected 2.5°C temperature threshold, fundamentally revises AMOC risk assessment under climate change.
Most critically, the widespread B_flux sign changes across CMIP6 models before 2100 reveal that current century-long projections systematically miss impending collapses, as the full shutdown unfolds over 100+ years beyond initial tipping. This implies substantially higher near-term collapse risk than previously recognized, demanding immediate integration of B_flux monitoring into operational climate observation systems and fundamental reassessment of climate adaptation timelines for Atlantic-bordering regions.
Sign up for our newsletter or connect with us on social media to stay up-to-date with our latest posts and permaculture inspiration.
Explore our inspiring series and posts:
Love the post? Share it with your circle, inspire your people.