A weakened AMOC is not a plateau: Brazil sediments record two abrupt surges

Marine sediments off Brazil's northeast coast record two abrupt intensifications of the Atlantic Meridional Overturning Circulation during a stretch of the last deglaciation when the current was, for the most part, much weaker than it is now, according to a Brazilian-German study published in Nature Communications ^[3][4][5][2]. That takes away a planning convenience: the assumption, supported by earlier work, that a weakened AMOC settles into a relatively stable low state lasting thousands of years ^[8].

The window in question is Heinrich Stadial 1, 17,800 to 14,800 years ago ^[4], about 3,000 years wide ^[2]. Inside it, the team led by Cristiano Mazur Chiessi of the University of Sao Paulo and Stefan Mulitza of the University of Bremen dates one intensification from 16,500 to 15,800 years ago, roughly 700 years ^[2][6][1], and a second of about 100 years around 15,400 years ago, during which the circulation's intensity exceeded its present level ^[7]. Roughly 400 years separated the end of the first from the second ^[3], and the second was about a seventh as long as the first ^[4]. Chiessi says this is the first time the AMOC has been shown to produce bursts of strengthening while in a weakened state ^[10], and reports that one reviewer wrote that the finding completely changes how the circulation is understood ^[11]. The authors' operational reading is that society should prepare not for one abrupt change but for a series of them ^[9].

The physical setup is not in dispute. The circulation is driven mainly by cold, salty water sinking off Greenland, flowing south at depth, upwelling largely in the Antarctic Circumpolar Current under strong winds, and returning north at the surface ^[13]. Greenland ice melt, a warming Arctic Ocean and increased rainfall are all reducing the salinity and density of surface water, which makes sinking harder ^[14]. What that transport does when it changes is broad: it moderates temperatures in Europe and parts of North America and influences rainfall in intertropical Africa and South America ^[12].

The forecast side remains weak. Researchers do not know when a sudden significant weakening would occur or how intense it would be ^[15], and until recently even the best models assessed by the IPCC could not predict the AMOC's evolution with sufficient accuracy ^[16]. In April, a University of Bordeaux group published revised predictions in Science Advances using new observational data, putting the weakening at 43% to 59% by 2100 even if countries meet their existing emissions commitments ^[17][18]. That range brackets a halving of the circulation ^[5]. A weakening of that scale has not happened since the end of the last Ice Age ^[19], when ice more than 3,000 meters thick covered much of Eurasia and North America and reached as far south as Chicago ^[20].

For anyone whose assets are sensitive to rainfall or North Atlantic temperature, the useful change here is statistical, not directional. A century-scale swing inside an already-weak regime ^[7] is shorter than the design life of a reservoir or a port, so the planning variable is variance rather than a single step down.

What to watch: whether the two events turn up in independent proxies elsewhere in the Atlantic, and whether any model that struggled to reproduce the AMOC's trajectory ^[16] can generate a 100-year burst from a weakened base ^[7]. Also worth watching is whether the Bordeaux range ^[18] is treated as a plateau or as a midpoint of a noisier band.