How Stable are Ocean Currents?

05 Dec 2005

A team of scientists from nine climate research institutions in seven countries has completed a major intercomparison study of climate models, organised by the Potsdam Institute for Climate Impact Research in Germany (PIK). The study focused on the stability of the Atlantic ocean circulation, which could be at risk in future as a result of human-caused climate change. The study aimed to quantify a critical 'tipping point' of the climate system: a threshold where the so-called 'ocean conveyor' would break down. This ocean current contributes to Europe's mild climate. The main finding of the study is that such a critical threshold was found in all participating models, but a relatively large uncertainty remains about exactly how far from such a threshold the present climate is. This makes it difficult to assess the danger of future ocean circulation changes.

The possibility of ocean circulation changes has recently received renewed attention after British scientists reported a 30% slowdown in the deep flow of the Atlantic (Nature, 1 Dec. 2005). Scientists have been working for years to investigate the mechanisms of ocean changes and to assess the risks for the future. Many abrupt shifts in ocean circulation have been documented for the past, specifically for the last Ice Age. It is well-known that ocean currents can be changed by diluting the salty ocean water with freshwater from rain, river runoff and melting ice. A trend of declining salinity is observed in the northern Atlantic in recent decades.

The new study investigated the effects of freshwater input on the ocean in a systematic way: a slowly increasing amount of freshwater was added to the northern Atlantic in eleven different climate models, and the response of the ocean circulation and climate was analysed. All participating models agreed in one point: if enough freshwater is added, the Atlantic 'conveyor belt' circulation shuts down. This circulation contributes to the Gulf Stream flow and keeps Europe much warmer than it would otherwise be.

"The crucial question is: how much freshwater is needed to stop the flow?" says Stefan Rahmstorf, a professor of Physics of the Oceans in Potsdam, who initiated the study. "We designed a standardised experiment that we performed with different climate models to compare their response." In some models, a long-term sustained inflow of less than 0.1 million cubic meters per second was needed, while the most stable models required 5 times as much. A majority of models required amounts near 0.2 million cubic meters per second. For comparison: if the Greenland Ice Sheet were to melt over a time interval of 1,000 years (which some glaciologists think is a possible result of human-induced global warming), this alone would provide an average freshwater inflow of 0.1 million cubic meters per second into the Atlantic.

A complete shut-down of the 'Atlantic conveyor' is not considered a very likely outcome of global warming by most experts, but rather it is discussed as a low probability - high impact risk. "We're talking about a kind of accident in the climate system", says Rahmstorf. "The consequences are hard to foresee but could be severe, so it is best to minimise the risk."

The present study (to be published on December 6th in Geophysical Research Letters) helps to understand model differences and to develop better models with reduced uncertainty. The study is one of several international intercomparison projects looking at different aspects of the stability of ocean currents. In this way, scientists aim to provide society with a better foundation for assessing the risks involved in climate change.

The participating institutions:

  • Potsdam Institute for Climate Impact Research, Potsdam, Germany
  • Hadley Centre for Climate Prediction and Research, Met Office, Exeter, UK
  • Institut d'Astronomie et de Géophysique Georges Lemaître, Université Catholique de Louvain, Louvain-la-Neuve, Belgium

  • Joint Institute for the Study of the Atmosphere and the Oceans, University of Washington, Seattle, Washington, USA

  • National Center for Atmospheric Research, Boulder, Colorado, USA

  • Alfred Wegener Institute for Polar and Marine Research, Bremerhaven, Germany

  • Southampton Oceanography Centre, Southampton, UK

  • Department of Atmospheric and Oceanic Sciences, McGill University, Montreal, Quebec, Canada

  • School of Earth and Ocean Sciences, University of Victoria, Victoria, British Columbia, Canada

For further information please contact

Prof Dr Stefan Rahmstorf, Potsdam Institute for Climate Impact Research (PIK), Phone:+49 (0)331 2882688, Fax: +49 (0)331 2882600,

Dr Andrey Ganopolski, Potsdam Institute for Climate Impact Research (PIK), Phone:+49 (0)331 2882594, Fax: +49 (0)331 2882600,

Original article

S.Rahmstorf, M. Crucifix, A. Ganopolski, H. Goosse, I. V. Kamenkovich, R. Knutti, G. Lohmann, R. Marsh, L. A. Mysak, Z. Wang, and A. J. Weaver, 2005: Thermohaline circulation hysteresis: a model intercomparison. Geophysical Research Letters, 32, L23605, doi:10.1029/2005GL023655