On the characterization of compound flood hazard at the coastal-riverine interface

Floods are the most common and most destructive natural disasters, have already been exacerbated by climate change, and promise to worsen further in the future. In this dissertation I address the unique threat of coastal compound floods: those floods caused by combinations of coastal surge and high river discharge occurring together.

The first part of this dissertation introduces the issue and challenge of compound floods and presents a review of existing methods for their analysis. The second part describes my analysis of coastal floods from an unusual perspective, beginning with flood seasons rather than flood heights. This analysis relies on circular statistical models to predict when coastal floods are most likely to occur. The third part describes my analysis of river flood seasons and the clusters of flood likelihood they reveal across the United States. Next, I combine the coastal and riverine flood hazards by testing whether coastal high water and heavy rainfall are more likely to occur together. If true, this would undermine several of the most common methods in use today.

Finally, I connect conclusions from the preceding sections to propose a new analysis method for coastal compound flood risk. The findings described here are useful to the public and managers, providing a clearer description of the risk coastal residents may bear without their knowledge. Understanding this risk can allow better planning of infrastructure and communities, improve the efficiency and transparency of financial markets, and allow better preparation for and recovery from disasters.