Possible non-linear responses of the climate system to anthropogenic forcing are gaining increasing scientific, political and public attention. While the early discussion about global warming and its impacts and consequences has focussed on "best guess" scenarios, i.e. the most probable outcome of greenhouse gas emissions, it is now increasingly recognised that "low probability / high impact" risks are an important and policy-relevant aspect of climate change. One of these risks is the possibility of a major ocean circulation change in the Atlantic. While this risk is much discussed, there is thus far very little scientific information on the impacts of such an event. The objective of this study is to investigate the risks of two types of ocean circulation change:

  • A substantial weakening of the North Atlantic Current and deep convection by 20-50%, as simulated by most climate models for the coming 50-100 years (Rahmstorf, 1999)

  • A (practically irreversible) complete shutdown of convection and the Atlantic thermohaline circulation, as simulated in some "pessimistic" model scenarios for the 22nd century (Rahmstorf and Ganopolski, 1999).