A Data-Driven Storm Surge Analysis for the U.S. Gulf Coast

Abstract

This dissertation provides the first empirical storm surge analysis for the U.S. Gulf Coast. Data are provided by SURGEDAT, a comprehensive storm surge database. A global storm surge literature review provided more than 700 observations in six ocean basins. The most severe storm surges have occurred in the Bay of Bengal, and the most frequent low-magnitude surges have occurred in East Asia. The U.S. Gulf Coast experiences the second highest frequency of low- and high- magnitude storm surges. Two Gulf Coast studies revealed that storm surge heights correlate better with pre-landfall tropical cyclone conditions, such as maximum wind speed and size, than cyclonic conditions at landfall. Surges correlated best with maximum wind speeds and storm size 18 hours before landfall. Logarithmic plotting provided the best statistical method for estimating storm surge return levels in the region. These levels showed considerable geographic variability, as the highest 100-yr level was 7.95 m at Bay St. Louis/ Pass Christian, Mississippi, and the lowest 100-yr level was 2.53 m at Cedar Key, Florida. Along the Northern Gulf Coast, surge levels were relatively low near Morgan City, however rapid sea-level rise threatens this area. The 100-year storm surge along the Gulf Coast would inundate 72% of the oil refineries and 63% of the power plants in the coastal zone, if not for local flood protection. Southeast Texas contains a dense network of vulnerable energy infrastructure, as the 100-year flood threatens to inundate 92% (12 of 13) of the refineries in the Galveston-Baytown-Sabine Pass region, and 18 power plants near Baytown. These results will be valuable to planners, emergency management personnel, professionals in the energy and insurance industry, as well as coastal scientists and storm surge modelers.