"Since the 1980s, scientists have been discussing and studying the possibility of instabilities in this major ocean circulation," Matthias Hofmann told environmentalresearchweb. "Such instabilities were very likely the cause of at least twenty abrupt climate shifts during the last Ice Age, recorded for example in the Greenland ice cores."
The AMOC transports around 1015 W of heat from the tropics toward high northern latitudes, bringing Europe a relatively warm and mild climate. So any slowdown or switch-off of the currents could have a dramatic effect. Today, the circulation brings water from the tip of South Africa right up into the Arctic Ocean. It then sinks in deep-water formation regions in the Nordic and Labrador Seas and returns south at depths of 2000–3000 m.
The more vigorous hydrological cycle predicted under climate change is likely to increase rainfall over the northern Atlantic and boost the amount of freshwater entering the ocean via rivers. Melting of the Greenland Ice Sheet will also add freshwater to the mix. These factors will reduce salinity and, together with higher sea surface temperatures, this will reduce seawater density and inhibit deep-water formation.
"The North Atlantic as the major region of deep water formation is a highly vulnerable area in times of climate change," said Hofmann. "It has to be regarded as a crucial tipping element in the Earth's climate system – or even its 'Achilles Heel', as the famous US ocean scientist Wally Broecker once put it."
According to Hofmann, over the last few decades numerous model studies have revealed that global warming and North Atlantic freshening could provoke a weakening or even a shutdown of the circulation and hence trigger an abrupt and dangerous climate change. The last IPCC report concluded there is an up to 10% risk of "a large and abrupt transition" of the Atlantic Ocean circulation occurring this century.
"To understand the stability of the AMOC, scientists usually study how it responds in models to different amounts of freshwater influx, as would occur when Greenland's ice continues to melt," said Hofmann. "We used an improved model which eliminates spurious mixing that previous studies had suggested might affect the stability of ocean currents in models."
Hofmann and colleague Stefan Rahmstorf, also at the Potsdam Institute for Climate Impact Research, Germany, found that the basic AMOC instability mechanism is a very robust feature that also occurs in this low-mixing model, where the AMOC is predominantly wind-driven. "However, the question of how close to a dangerous threshold we are remains unresolved," he said.
The pair also reviewed recent findings by Dutch colleagues, who found strong evidence that most current climate models could have a systematic bias towards the AMOC being far too stable, compared to the real ocean. "So we may still be underestimating the risk, based on the current climate models," said Hofmann. "These models are good at computing simple things like global mean temperature, but they are still unreliable when it comes to highly non-linear threshold responses like an instability in ocean currents."
Now the pair plan to continue their research by trying to derive a simple measure of the stability of the AMOC, employing a concept recently proposed by their Dutch colleagues (de Vries and Weber, GRL, 2005; Dijkstra, Tellus A, 2007). "The goal is to find better ways to check the stability of ocean currents in models with data from the real ocean, to test and improve the reliability of our models," said Hofmann.
Hofmann and Rahmstorf reported their work in PNAS.