The availability of output from climate model ensembles,such as phases 3 and 5 of the Coupled Model Intercomparison Project(CMIP3 and CMIP5), has greatly expanded information about future projections,but there is no accepted blueprint for how this data should be utilized.The multi-model average is themost commonly cited single estimate of future conditions,but higher-order moments representing thevariance and skewness of the distribution of projections provide important information about uncertainty. We have analyzed a set of statistically downscaled climate model projections from the CMIP3 archive to assess extreme weather events at a level aimed to be appropriate for decisionmakers. Our analysis uses the distribution of 13 global climate model projections to derive the inter-model standard deviation, skewness,and percentile ranges for simulated changes in extreme heat,cold,and precipitationby the mid-21stcentury,based on the A1B emissions scenario.These metrics provide information on overall confidence across the entire range of projections(via the inter-model standard deviation),relative confidence in upper-end versus lower-end changes(via skewness),and quantitative uncertainty bounds(derived from bootstrapping). Over our analysis domain,which covers the northeastern United States and southeastern Canada,some primary findings include:(1)greater confidence in projections of less extreme cold than more extreme heat and intense precipitation,(2)greater confidence in relatively conservative projections of extreme heat,and(3)higher spatial variability in the confidence of projected increases in heavy precipitation. In addition,we describe how a simplified boot strapping approach can assist decisionmakers by estimating the probability of changes in extreme weather events based on user-defined percentile thresholds.
