15th Conference on Applied Climatology


Sea–surface temperature and tropical cyclones: Breaking the paradigm

Patrick J. Michaels, Univ. of Virginia, Charlottesville, VA and Cato Institute, Washington, DC; and P. C. Knappenberger and R. E. Davis

Computer simulations of enhanced tropical cyclones in a warmer world predict a gradually increasing frequency of powerful hurricanes. The primary reason that modeled future storms are more intense is that SSTs in the tropical cyclogenesis regions are higher in the high-CO2 scenario. However, the models ignore many factors such as changes in vertical wind shear, that act to inhibit tropical cyclone development. As such, the strength of the modeled storms more closely reflects “potential intensity,” that is, the theoretical maximum intensity that a tropical system may obtain absent influences other than SST and large-scale atmospheric temperature and moisture conditions. It is not the potential intensity, but the actual intensity that a storm attains that is relevant to the community of individuals who are impacted by tropical weather systems, and thus it is instructive to examine actual storm behavior under real-world conditions.

The relationship between observed SST and hurricanes is much weaker than in the models. Our hurricane data is National Hurricane Center's “Atlantic Tracks File” containing the 6-hourly (0000, 0600, 1200, 1800 UTC) center locations (latitude and longitude in tenths of degrees) and intensities (maximum 1-minute surface wind speeds in knots and minimum central pressures in millibars) for all tropical storms and hurricanes from 1851 through 2004 in the Atlantic Basin. We related the intensities to weekly sea surface temperatures from the 1º latitude by 1º longitude gridded data set of Reynolds and Smith for the period 1982 to 2002.

We find that the explained variance of the relationship between storm intensity and SST is almost five times lower in the observed vs. the modeled environment. While there are indications that SSTs play a role in determining the maximum intensity obtained by the strongest storms, this relationship weakens when all cyclones are considered. Furthermore, the data preliminarily indicate an optimal temperature of about 28ºC to 29ºC for the development of intense hurricanes. At temperatures either above 29ºC or below 28ºC, the frequency of strong cyclones (top winds equal to or exceeding 100kt) drops.

In addition to the weakened dependence of storm strength on SST in the real world vs. the modeled world, the real rate of carbon dioxide increase (which has been quite constant for the last three decades) is also much less than that incorporated into the climate models used to project stronger future hurricanes. When we factor in a more realistic scenario of future CO2 build-up, the prospective increases in modeled storm intensity are halved in model year 2080.

After accounting for these model shortcomings, we conclude that a detectable increase in Atlantic hurricane intensity in response to growing atmospheric greenhouse gas levels during this century is unlikely.

extended abstract  Extended Abstract (284K)

Uploaded Presentation File(s):

Session 2, Climate Trends and Variability
Monday, 20 June 2005, 1:30 PM-5:30 PM, North & Center Ballroom

Previous paper  Next paper

Browse or search entire meeting

AMS Home Page