Robust analysis and improved understanding of the observational record is critical because climate models can contain significant uncertainties in their representation of atmospheric extremes. CASCADE researchers develop methods for characterizing changes in observed extremes and develop new extreme-focused data sets for climate model evaluation.
Traditional extreme value analyses are insufficient for the scale of the data sets over the spatial and temporal scales relevant to CASCADE research. State-of-the-art spatial statistical tools are utilized to increase the signal-to-noise ratio of observational analyses, characterize extreme climatology at the native spatial scales, and more appropriately characterize the pattern of extreme events over time. CASCADE’s primary focus is on precipitation and the extent to which natural variability affects the frequency and magnitude of precipitation extremes (both the upper and lower tails), for both short term extremes and the pattern of daily precipitation over the course of a season that leads to extreme seasonal precipitation totals.
Complementary to observational analyses is the generation of new data sets geared towards extremes in an effort to provide community tools for improved evaluation of climate model output. These new data sets focus on precipitation, temperature, mountain hydrology, snowpack, and precipitation attributable to various event types; their ultimate goal is to inform model evaluation, leading to improved understanding of how well climate models characterize the climatology of extremes. Furthermore, observational analyses identify meaningful modes of natural variability for extremes, which in turn can improve predictability in model output and projections.