standard of evidence for considering security risks is different from the standards of evidence in fundamental science. Intelligence and security actions are often warranted to reduce the risks from events whose likelihoods are low or cannot be predicted with confidence because the phenomena are too complex or poorly understood or because of the importance of human agency in shaping the course of events. Many of the risks associated with climate change have these characteristics. Nonetheless, the relevant sciences sometimes can develop useful estimates of changes in the likelihoods of certain kinds of events and consequences. Security analysis needs to apply a risk-based analytic approach to recommending action, such as the one briefly described in Box 2-2.

There are several general approaches for implementing a risk-based climate–security analysis. One takes a forecasting approach: Analysts project the likelihoods of disruptive events and bring the high-risk forecasts to policy makers’ attention. Applied to climate change, this approach would need to involve the forecasting of climate events as well of as societal conditions that alter exposure and susceptibility to harm from those events and that affect the ways that governments, societies, and other social institutions respond when climate events create social disruptions. Forecasting requires an understanding of the key variables that need to be included in the forecast as well as a theoretical framework that specifies how current conditions and trends are linked to future outcomes of concern.

This approach potentially has the advantage of bringing to decision makers’ attention a range of scenarios worth worrying about. In particular, because risk is the product of likelihood and impact, this approach would likely bring both high-likelihood/medium-impact events and low-likelihood/high-impact events to policy makers’ attention. Both types of events may be relevant for assessing the impact of climate change on social and political stresses of security interest. An important drawback of this approach is that the ranking of risks—and thus the selection of the events that fall above and below the threshold for policy makers’ attention—may be highly sensitive to erroneous estimates of both likelihood and consequence in ways that are poorly understood or even unnoticed a priori. Such errors may occur in the forecasting of every relevant factor—in the forecasting of climate events, of exposure and susceptibility, and of reactions to disruptive events. The difficulties of forecasting all of these things given the current state of knowledge has led us to put a low priority on using this approach.

A second approach emphasizes early warning. Analysis can suggest early warning indicators of significant events in countries of interest. Analysts might, for instance, identify early indicators of climate conditions that are likely to lead to serious economic or social consequences for the affected populations or indicators of a lack of political or social capacity to cope with or respond to such consequences. They might also develop



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement