The US National Research Council defines abrupt climate change as a change of state that is sufficiently rapid and sufficiently widespread in its effects that economies are unprepared or incapable of adapting. This may be too restrictive a definition, but abrupt climate change does have implications for the choice between the main response options: mitigation (which reduces the risks of climate change) and adaptation (which reduces the costs of climate change). The paper argues that by
(i) increasing the costs of change and the potential growth of consumption, and
(ii) reducing the time to change,
abrupt climate change favours mitigation over adaptation.
Furthermore, because the implications of change are fundamentally uncertain and potentially very high, it favours a precautionary approach in which mitigation buys time for learning. Adaptation–oriented decision tools, such as scenario planning, are inappropriate in these circumstances. Hence learning implies the use of probabilistic models that include socioeconomic feedbacks.