Understanding, Modeling and Predicting Extremes

Weather and climate extremes have profound impacts on both human society and the natural environment. The potential societal payoffs from gains in understanding, predicting and projecting extreme events and their changes are enormous, ranging from improved early warning of extreme events to providing information that is essential for longer-term adaptation decisions (e.g. Findings from the NOAA Science Challenge Workshop, 2011). However, understanding, modeling and predicting extremes remains a great scientific challenge. It involves improved understanding of what causes extremes and their inherent predictability based on suitable datasets (climate reanalysis is often use in such analyses) and improved modeling capabilities with higher resolution and more accurate models. For example, it is recognized that climate processes in the Intra-Americas Sea (IAS) region can affect climate extremes over North America, including tropical cyclone activity and the frequency of droughts and floods (http://www.eol.ucar.edu/projects/iasclip). IAS-related phenomena of relevance to these extremes suggest a potential for predictability on intra-seasonal to decadal time scales. Yet, state-of-the-art global climate models have large mean biases and erroneous variability over the IAS and neighboring regions of North America that limit the capability to use models to fully understand these predictive linkages and improve climate model prediction.