For the AMOC, how fast we warm may matter more than how hot it gets

Researchers at Utrecht University report that the Atlantic Meridional Overturning Circulation, the current system that includes the Gulf Stream and moves heat into the Northern Hemisphere, can survive far more warming when it arrives slowly than when it arrives fast ^[1][2]. The finding, published Aug. 13 in Nature Climate Change, reframes a risk that current policy treats as a single temperature threshold ^[3][4].

In climate model runs, slow warming let the currents hold their strength beyond 5C above preindustrial levels, collapsing only near 5.5C ^[6][5]. Warming at a pace equal to or above today's emissions tipped the same system at just 2C ^[7]. Co-author Henk Dijkstra said that under slow warming the entire ocean has time to reorganise and adapt, while under faster warming it simply cannot keep up ^[8].

The mechanism is density. The AMOC runs on North Atlantic surface water sinking to form the bottom currents that drive the loop, and that water sinks only when it is colder and saltier, and so denser, than the layers below ^[9]. Warming surface water and Arctic meltwater diluting its salt are both making that sinking less reliable than before ^[10]. Given time, two feedbacks push back: more evaporation strips fresh water and concentrates the remaining salt, and a warmer world delivers less Arctic meltwater once most of the ice has already melted ^[11]. Both changes only help if they happen slowly ^[12].

The design isolated rate from endpoint. The team ran one slow CO2 ramp at 0.5 ppm per year and two fast ramps at 2.5 and 5 ppm per year ^[13]. Co-author Reyk Borner said they deliberately chose a scenario much slower than today's to isolate the effect of the warming rate alone, independent of how warm it eventually gets ^[14]. Atmospheric CO2 is currently rising by 2.4 to 2.5 ppm per year ^[15], which puts the present trajectory at the slower of the two fast ramps ^[16] and roughly five times the rate that left the currents intact ^[17].

Previous work puts the AMOC at its weakest in more than 1,000 years ^[18], with an estimated shutdown threshold near 4C above preindustrial levels ^[19]. A weakened or collapsed AMOC has been linked to freezing weather in parts of Europe, faster sea level rise along the U.S. East Coast, and drought around the equator ^[20].

What to watch is whether the near-term emissions rate, rather than the century-end target, becomes the operative variable in tipping-point risk assessments ^[4]. This is one study of one system in models, and the rescue mechanisms it credits at 5C depend on decades that today's trajectory does not obviously allow ^[12][17].