Policy makers are increasingly interested in identifying the potential costs and benefits of different climate change mitigation and adaptation policies. Ecosystem threshold responses can result in significant societal impacts and economic damages, and reducing the risk of such impacts through climate change mitigation can thus yield important benefits. Identifying and quantifying this “avoidance of risk” can be difficult. While current scientific information may provide insights into the nature of potential impacts and the ecological processes related to threshold phenomena, in many cases, little to no information is available regarding the probability of threshold dynamics occurring under different climate change scenarios. Explicitly addressing these issues is critical to the analysis of different policy proposals.

To begin to address some of these issues, the U.S. Environmental Protection Agency’s Climate Change Division within the Office of Air and Radiation commissioned a study that involved a synthesis of scientific literature for individual ecosystem threshold case studies and an “expert elicitation” process. The elicitation process was designed to provide the Agency with a sense of the scientific uncertainty associated with different ecosystem thresholds, addressing expert opinions about: 1) the likelihood of a given threshold occurring under future temperature increases; 2) the confidence of experts in their conclusions; and 3) key uncertainties. The ecosystem thresholds included in the analysis were loss of functioning coral reef ecosystems; local extinction of whitebark pines within the Greater Yellowstone Ecosystem; significant loss of snowpack in the Sierra Nevada mountain range in California; and loss of summer seasonal sea ice in the arctic. 
Results/Conclusions

Overall, the expert elicitation process revealed agreement among scientists that the likelihood of threshold impacts – and the certainty of scientists in their own judgments -  increases with increasing temperature levels.  The exception to this was the case study involving mountain pine beetle dynamics and impacts to whitebark pine, where expert uncertainty rises with temperature increases above 2-4°C because of changing insect population dynamics at higher temperatures. Overall, the loss of summer arctic sea ice was deemed most sensitive of all the case studies to small increases in temperature, and significant loss of snowpack in the Sierra Nevada was deemed the least sensitive to small temperature increases. While the elicitation process cannot provide probabilistic distributions of the likelihood of a given threshold phenomena occurring under different climate change scenarios, it can successfully convey information about key areas of expert agreement and uncertainty for policy makers.