The Atlantic Meridional Overturning Circulation (AMOC) is a key element of overturning in the global climate system. It is a large-scale system of ocean currents that transports heat and salt across the world's oceansAMOC collapse, defined as a transition from a strong and deep overturning current to a weak and shallow residual state, would have serious impacts on the climate in large parts of the world, particularly drastic changes would be felt in northwestern EuropeGiven these potentially serious global consequences, the current state of the AMOC is being closely monitored along several observing fields in the Atlantic Ocean.

However, current observations, such as the 20-year time series from RAPID-MOCHA at 26°N, are too short to separate natural variability from the long-term climate trend. AMOC reconstructions based on sea surface temperature (SST) suggest gradual weakening of the AMOC since 1900Some statistical early warning signals even suggest that the AMOC is approaching a tipping point. However, it is important to note that these signals are prone to false positives, and the reconstructions introduce significant uncertainty into estimates of the timing of collapse.

In an effort to develop a more reliable physical indicator of the impending AMOC collapse, the new study analyzed targeted climate model simulations. Previous indicators, such as freshwater transport at 34°N, have been shown to be limited under transient forcing conditions. The new method proposed in this work focuses on change in the sign of the surface buoyancy flux integrated over the area of rising isopynals in the North Atlantic (between 40°N and 65°N)This indicator is closely related to the adiabatic contribution to overturning and works successfully in quasi-equilibrium freshwater forcing, freshwater pulse forcing, and climate change scenarios for various climate models. Although there is also an indicator Dogs_adiab, which quantifies the isopycnal protrusion, requires 3D ocean fields, while the surface buoyancy flux (B_surface) is easier to use for multiple model analysis.

An analysis of 25 different climate models from the Coupled Model Intercomparison Project (CMIP6) suggests that AMOC could start to collapse in the near future:

• According to the transitional emissions scenario (SSP2-4.5), this could happen until 2063, with a confidence interval between 2026 and 2095.
• In the high emissions scenario (SSP5-8.5), collapse is estimated to until 2055, with an interval between 2023 and 2076.

These estimates are associated with a global warming anomaly of approximately +2.48 °C for SSP2-4.5 and +2.79 °C for SSP5-8.5, suggesting a need to reconsider previously estimated critical temperature thresholds for AMOC collapse. Of the 25 CMIP6 models analyzed, the majority (22 of 25 for SSP5-8.5 and 16 of 25 for SSP2-4.5) showed a sign change in the surface buoyancy flux indicator before 2100. Extended simulations confirm that models showing this sign change will indeed collapse when integrated after 2100.

When the AMOC collapses, the consequences are significant. The climate of northwestern Europe will change drastically. It is expected significant cooling, especially in the winter months, which partially compensates for global warming in high-emission scenarios. It is also projected increased storm activity over Europe and less precipitation. Tropical rain belts and the intertropical convergence zone will shift south. These AMOC-induced impacts are comparable to the collapse of the AMOC under constant pre-industrial conditions.

Study emphasizes that while AMOC strength alone is not always a reliable indicator of the proximity of a tipping point, the change in sign of surface buoyancy flux has proven to be a robust physical indicator of the onset of an AMOC tipping event. However, there are nuances; for example, in special cases where buoyancy flux does not consistently increase after a sign change, additional control with other indicators is needed. Future warming is also likely to increase the melting of the Greenland ice sheet, which would further weaken the AMOC, and this fact is not fully accounted for in most current climate models.

Overall, the results underline the urgency of climate actionTo limit the risk of AMOC weakening and potential collapse, it is essential that global society moves towards low-emission scenarios (e.g., SSP1-1.9 and SSP1-2.6). Improving the accuracy of climate models and extending simulation periods in future research are also key to better understanding and predicting the future of the AMOC. JRi