Since the 1970s, ocean surface temperatures around the globe have been on the rise--from one half to one degree Fahrenheit, depending on the region. Last summer, two studies linked this temperature rise to stronger and more frequent hurricanes. Skeptics called other factors into account, such as natural variability, but a new statistical analysis shows that only this sea surface temperature increase explains this trend.

Climate researcher Judith Curry and her colleagues at the Georgia Institute of Technology looked at the hurricane records for storms between 1970 and 2004 in all of the world's ocean basins, yielding a total sample of 210 seasons over the six regions. They subjected the records to a mathematical test derived from information theory--so-called mutual information, which measures the amount of information two variables share, so that if they do not overlap at all this measure would be zero.

The researchers then looked at sea surface temperature, specific humidity, wind shear and wind variation over longitude to see what, if anything, these variables shared with the increasing number of strong storms the world over. According to the analysis appearing online today in Science, this trend only depends on sea surface temperature. "If you examine the intensification of a single storm, or even the statistics on intensification for a particular season, factors like wind shear can play an important role," Curry says. "However, there is no global trend in wind shear or the other factors over the 35-year period."

The link between rising ocean temperatures and overall climate change remains murky because of the overlap between natural cycles and any global warming. "But if you buy the argument that global warming is causing the increase in sea surface temperatures--and everybody seems to be buying this--then it's a pretty small leap to say global warming is causing this increase [in hurricane frequency]," Curry says. Her team will now focus on clarifying the mechanisms at work in the North Atlantic by separating out the 75-year natural cycle and climate change. "The last peak was in 1950, the next is in 2025," she adds. "We're only halfway up [the cycle] and we're already 50 percent worse [in terms of storms]. To me, that's a compelling issue that needs to be confronted."

2 Comments

1. 1. rroto1 07:44 AM 1/12/08

   This study is based on less than accurate data, and thus it is pure conjecture.

   Reply | Report Abuse | Link to this

2. 2. MendelPotok 10:48 AM 7/27/11

   Yes, you wouldn't happen to have a link to their paper of <a href="http://www.kellystatisticalconsulting.com">statistical analysis</a>? I think we would all like to see just how much of a stretch the authors are making.

   Reply | Report Abuse | Link to this