Hurricane tracks, changes in hurricane clustering, and other notes from Tucson

I'm in Tucson for the American Meteorological Society's 29th Conference on Hurricanes and Tropical Meteorology. This is the premier scientific conference on hurricanes, and is held only once every two years, so pretty much all of the world's greatest hurricane experts are here. Below are some quick snapshots of four of the talks I attended yesterday; I hope to more more snapshots each day this week.

Angela Colbert of the University of Miami/RSMAS showed how different weather and climate patterns affect the Azores-Bermuda High, and thus the tracks of Atlantic hurricanes. She divided storms into straight-moving storms that move straight west-northwest through the Caribbean, recurving landfalling hurricane that hit the east coast of the U.S., and recurving ocean storms that miss land. Roughly 1/3 of all hurricanes between 1950 - 2009 fell into each of these three categories. These proportions stayed pretty constant during La Niña and neutral years, but El Niño caused a weakening of the high, resulting in far fewer hurricanes hitting the U.S. East Coast. These storms instead recurved out to sea.

Jim Kossin of the University of Wisconsin separated all Atlantic storms from 1950 - 2007 into 4 clusters, based on genesis location. Two of the clusters were more northerly-forming storms that tended to be less tropical in nature--Gulf of Mexico storms, and storms off the U.S. East Coast that tended to recurve. The other two clusters were more southerly tropical-origin systems--ones that tended to form in the Caribbean, and storms that form near the Cape Verde Island region off the coast of Africa. The more tropical Cape Verde and Caribbean storms dominated major hurricane frequency by a factor of four. In mid-1980s, there was an abrupt shift to more of these more dangerous tropical type storms--ten years prior to the active hurricane period that began in 1995. It is unknown what caused this shift. The shift is unlikely to be a result of measurement error, since we had good satellite imagery then. Independent of any trends in frequency, this shift caused an increase in intensity metrics of Atlantic hurricanes. A doubling of these tropical systems has also occurred since 1950. Interestingly, there has been no change in the number of Gulf of Mexico storms, and a slight increase in storms forming off of the U.S. East Coast. Since slight changes in track can make a big difference in what SSTs and atmospheric environment a storm sees, there is a lot of natural "noise" in the system that will make it difficult to get a clear sense of when climate change is having a substantial impact on hurricane intensity.

Bin Wang of the University of Miami studied the global number of storm days from 1965 - 2008, which should be a less sensitive quantity to data problems than the number of storms or their intensity. Storm days were defined as any day when a tropical cyclone of tropical depression strength or greater existed. However, there are still some data problems, as evidenced by a sharp drop in storm days observed in the North Indian Ocean beginning in 1978. Dr. Wang found that there was no global trend in storm days. The Atlantic was the only individual basin that showed an increase in storm days.

Greg Holland of NCAR looked at the distribution of the strongest hurricanes over time by using a mathematical description of the historical hurricane data. His analysis showed that during the period 1995 - 2008, we probably had about a 30% increase in Category 5 storms in the Atlantic, and an 18% increase in Category 4 hurricanes. Using a climate model, he predicted that by the years 2045 - 2055, we should see a 60% increase in Cat 5s, 32% increase in Cat 4s, and 16% increase in Cat 3s in the Atlantic.

Jeff Masters