Atlantic ocean current unlikely to collapse with climate change

The Atlantic meridional overturning circulation, commonly referred to as the “AMOC,” is a system of ocean currents confined to the Atlantic basin that plays a crucial role in regulating Earth’s climate by transporting heat from the Southern to the Northern Hemisphere. The AMOC also modulates regional weather, from the mild summers in Europe to the monsoon seasons in Africa and India. Climate models have long predicted that global warming will cause the AMOC to weaken, with some projecting substantial weakening amounting to a near-collapse relative to the AMOC’s strength today. Such a weakening would have far-reaching consequences, including changes in regional sea level rise and major shifts in regional climate, such as colder conditions in northern Europe and drier weather in parts of the Amazon and West Africa.

However, a new study from Caltech finds that although the AMOC will weaken under global warming, it is likely to do so to a much lesser extent than current projections suggest. The team developed a simplified physical model based on fundamental principles of ocean circulation—specifically, how density differences and the AMOC’s depth are related—that also incorporates real-world measurements of the ocean current’s strength, collected over 20 years through the use of monitoring arrays and other observationally constrained products of the Atlantic basin. The researchers found that the AMOC will weaken by around 18 to 43 percent at the end of the 21st century. While this does represent some weakening, it does not represent substantial weakening that the more extreme climate model projections suggest. This new understanding significantly narrows the range of future AMOC weakening, addressing a long-standing uncertainty in climate science.

The study is described in a paper appearing in the journal Nature Geoscience. The research was conducted in the laboratories of Tapio Schneider, the Theodore Y. Wu Professor of Environmental Science and Engineering; and Andrew Thompson, the John S. and Sherry Chen Professor of Environmental Science and Engineering, director of The Ronald and Maxine Linde Center for Global Environmental Science, and executive officer for Environmental Science and Engineering.

Paleoclimate records, like ocean sediments that record past climate conditions, indicate that the AMOC has experienced weakening in the past, such as during the Last Glacial Maximum (a period about 20,000 years ago), leading to major swings in the climate that affected North America and Europe. Contemporary climate models show wide variation in their 21st century projections of AMOC weakening: Some predict substantialAMOC weakening, while others predict only a small amount of weakening. The new study, led by former graduate student Dave Bonan (PhD ’25), aimed to better understand the physical mechanisms governing AMOC behavior in climate models, with the goal of reconciling these discrepancies. 

The research sheds light on a long-standing and previously unexplained feature of climate models: the link between the present-day and future strength of the AMOC. Climate models that simulate a stronger present-day AMOC tend to project greater weakening under climate change. The researchers found that this relationship stems from the depth of the AMOC. A stronger AMOC typically extends to greater depths and allows changes in surface water temperature and salinity properties—caused by global warming and freshwater input—to penetrate deeper into the ocean and drive greater weakening. In other words, a climate model with a stronger and deeper AMOC is less resilient to surface changes and experiences proportionally more AMOC weakening than one with a shallower current. Climate models with a shallower present-day AMOC still show weakening under climate change, but to a lesser extent than those with a deeper present-day AMOC.

The new study uses this understanding to constrain future AMOC projections by building a simplified physical model and incorporating real-world measurements of the ocean current’s strength. The results indicate that the AMOC will only experience limited weakening even in the highest emissions scenarios. The study suggests that much of the previous uncertainty and some of the more extreme AMOC weakening projections stemmed from biases in how climate models simulate the ocean's current state, particularly its density stratification. 

“Our results imply that, rather than a substantial decline, the AMOC is more likely to experience a limited decline over the 21st century—still some weakening, but less drastic than previous projections suggest,” Bonan says.

Bonan emphasizes the need to examine higher-resolution climate models that also include more sophisticated processes. Higher-resolution models might offer deeper insights into AMOC behavior and improve projections of its future changes. The study provides a framework to interrogate and evaluate more sophisticated models.

While conducting this research at Caltech, Bonan was funded by the National Science Foundation Graduate Research Fellowship Program (NSF-GRFP).

“The NSF-GFRP gave me the freedom to tinker and explore,” he says. “There is immense value in doing basic research — it can give us a better indication of what the future might look like, as our study shows.”

The paper is titled “Observational constraints imply limited future Atlantic meridional overturning circulation weakening.” In addition to Bonan, Schneider, and Thompson, co-authors are Laure Zanna of New York University, Kyle Armour of the University of Washington, and Shantong Sun of Laoshan Laboratory in Qingdao, China. Funding was provided by the NSF, the David and Lucile Packard Foundation, and Schmidt Sciences LLC.

END