Temperature extremes can have major impacts on both human society and the natural world. As the climate warms as a result of human-induced climate change, changes in the frequency and magnitude of temperature extremes are likely. Increases in warm extremes and decreases in cold extremes have already been observed, and further changes are predicted by climate model simulations by the end of the 21st century. A number of studies have related climate extremes to low-frequency modes of natural climate variability, raising the possibility that future changes in such modes of variability will affect the frequency and magnitude of extreme temperature events. The objectives of this proposal are to understand the extent to which extreme values of observed daily temperature in North America are influenced by the amplitude and phase of major modes of climate variability and to determine if changes in such modes of variability are an important influence on extreme values of daily temperature in projections of future climate change. The proposed research will test the following hypotheses.

1. Temperature extremes during the latter part of the 20th century over North America show regional associations with the state of large-scale modes of natural climate variability
2. Extreme temperatures in winter will be more closely associated with the state of the major modes of natural climate variability than they are in summer
3. Any systematic changes in the major modes of climate variability resulting from future climate change will cause changes in temperature extremes to be of a larger or smaller magnitude than would be expected from a change in mean temperature.

The first phase of the proposed research will use observed data from second half of the 20th century to test the first two hypotheses by seeking relationships between the state of large-scale modes of natural climate variability and extreme values of daily temperature. The second phase of the project will test the third hypothesis by examining the extent to which changes in extreme temperatures in climate model projections of future climate are influenced by changes in these modes of variability. Future climate projections using current generation coupled atmosphere-ocean models will be obtained from the publicly accessible CMIP5 database. The proposed research addresses the fundamental goals of DOE climate science and, more specifically, one of the core areas of the Regional and Global Climate Modeling Program: "Simulation of Climate Extremes under a Changing Climate." The project will use recently developed observational data sets and results from current generation climate models to further our understanding of extreme temperatures and their relationship to modes of natural climate variability. Because many of the impacts of global warming will result from changes in temperature extremes, an improved understanding of the physical processes that govern changes in temperature extremes will help policy makers determine safe levels of greenhouse gases in the atmosphere.