Extended Abstract (396K)J3.1
Global warming, U.S. hurricanes, and insured losses
James B. Elsner, Florida State University, Tallahassee, FL; and T. H. Jagger
Increases in Atlantic hurricane activity during the past decade are related to increases in tropical Atlantic warmth. A debate concerns the attribution of these increases with some suggesting a natural atmospheric/oceanic cycle and others suggesting climate change. Atmospheric scientists in academia and government are found on both sides of the debate. The large correlation between late summer/early fall Atlantic sea surface temperature (SST) and global near-surface air temperature together with lagged values of global temperature predicting SST but not the other way around argues in favor of the climate change hypothesis. However the positive influence of climate change on Atlantic hurricane activity is limited to the connection with Atlantic SST possibly as a consequence of increased atmospheric thermodynamic stability and/or wind shear from greater warmth.
A parametric statistical model is built that shows the return period of a Katrina-like hurricane is 21 years for the Gulf coast of the United States (Texas through Alabama) and 14 years for the entire coast. Hurricane Katrina might be a harbinger of things to come in a warmer world as the observed and modeled consequence of climate change on hurricane intensity appears to start at Katrina's observed near-coastal intensity of 71 m/s. The annual probabilities for hurricanes weaker than Katrina do not change between globally warm and globally cool years. However, for hurricanes stronger than Katrina the increase in the 100-year return intensity from cold to warm years is 11% consistent with results from recent numerical simulations. A similar parameteric model is used to forecast the expected and maximum annual aggregated insured loss in the United States prior to the start of the hurricane season. From year to year the effect of climate change on insured losses will be minor relative to El Niņo, the North Atlantic Oscillation, and random SST fluctuations, but averaged over several years the effect could be significant.
Supplementary URL: http://garnet.fsu.edu/~jelsner/www
Joint Session 3, Tropical Cyclones and Probability/Statistics 1
Tuesday, 22 January 2008, 8:30 AM-9:45 AM, R02-R03
Next paper
Recorded presentation