A diagnostic framework for deep water formation and AMOC variability in selected CMIP6 models

The strength and variability of the Atlantic Meridional Overturning Circulation (AMOC) are closely linked to deep water formation (DWF) in three key regions: the Labrador Sea, the Irminger Sea, and the Greenland Sea. However, quantifying the relative contributions of these regions to the AMOC in climate models and how these contributions evolve under future climate scenarios remains challenging. While CMIP6 models consistently project a weakening of the AMOC, they show wide inter-model spread in the rate of decline. This highlights the need for robust metrics that enable more informative intercomparison. The commonly used mixed layer depth metric captures some aspects of convection, but does not directly quantify DWF. Here, we introduce a volume-conservation-based diagnostic that serves as an index for quantifying DWF, enabling robust comparison across models with differing resolution and complexity. It further quantifies the regional contributions of the Labrador, Irminger and Greenland Seas to the AMOC.  

When applied to EC-Earth3 at standard and high resolutions, the diagnostic suggests that DWF in the Labrador Sea is the main cause of the projected weakening of the AMOC. Meanwhile, the Irminger Sea emerges as the AMOC's largest overall contributor, experiencing only a modest decline and remaining essential for sustaining the circulation. At the same time, the contribution from the Arctic increases. We assess inter-model differences in DWF magnitude and examine their relationship to AMOC changes by extending the analysis to a suite of CMIP6 models. This allows us to evaluate the robustness of these processes across models. Overall, our results provide new insight into the factors underlying differences in AMOC projections among models and into the mechanisms that may influence the risk of an AMOC slowdown or tipping point.