Richard L. Smith

University of North Carolina at Chapel Hill

 

Assessing the risk of extreme weather events is a daunting task even without considering climate change. I illustrate this point with two examples. The first is the extreme precipitation associated with Hurricane Harvey. Several analyses of that event agree that it was well beyond the 100-year return level that is standard in risk assessment, but exactly how extreme it was, and how to assess the uncertainty, is a subject of active debate. A second example is the assessment of storm surge risks – I shall highlight recent work at UNC and SAMSI on statistical methods for that problem. These problems are even more complicated if one brings climate change into the discussion. That requires data from climate models and consideration of how best to combine model data with observations. Recent advances in the field known as detection and attribution suggest statistical ways forward on that problem. Much of the talk will be based on research with various collaborators during the current climate program at SAMSI.