Klaus Keller is analyzing two interrelated questions. First, how can we understand, detect, and predict potential anthropogenic threshold responses of the climate system? Second, how can we use this scientific information to design sound risk-management strategies? One potential climate threshold response involves an abrupt and/or persistent weakening of the North Atlantic meridional overturning circulation (MOC), an ocean circulation system that transports heat from low latitudes to the North Atlantic basin and the surrounding regions. The geologic record and model simulations suggest that the MOC may weaken or even collapse in response to climatic forcing. Such a threshold response could be associated with considerable ecological and economic impacts. The current predictions about the future fate of the MOC are, however, deeply uncertain. Reducing this uncertainty poses nontrivial scientific and operational challenges. Overcoming these challenges can provide potentially large economic value.

Keller et al. show that the currently implemented MOC observation system might well succeed in detecting anthropogenic MOC changes within a few decades. However, prediction of future threshold crossing – not just detection of past anthropogenic changes – is the relevant task for many decision-making frameworks. A confident prediction of a potential MOC threshold response requires, however, considerably longer observations or much higher signal-to-noise ratios than confident detection. This raises the distinct possibility that the current MOC observation system would provide a confident prediction of a MOC threshold response only after the forcing threshold has been crossed. Keller et al. also show how the MOC observation system might be considerably improved by optimizing the spatial design of the current observation system and by adding high-resolution oceanic oxygen observations.