What will the hurricane season be like this year?

[DrCloud]

As far as I know there is no real way to forecast something as complex and broad as a whole "hurricane season" with any certainty.

 

This is correct, particularly given the distinction we make between "predict" and what we (try to) do with climate, which is to "project" it. Seasonal hurricane patterns and trends are definitely "projected."

 

This is done using methods pioneered by Prof. Bill Gray of Colorado State University. His technique involves, essentially, analogs from the past. 'Round about each April, his group looks at the current state of the climate and then looks for years in the past that were similar. Then they look at the hurricane seasons in those similar years and use various kinds of factor analysis to come up with a best estimate (including uncertainty estimates) of the coming hurricane season.

 

Their methods have been shown to have "skill", meaning that they're better than either random guess or pure extrapolation from last year. Further, the skill level is high enough that the National Hurricane Center has adopted some of the methodology for their official seasonal outlook.

 

Now, it's interesting that all of the people on the "natural cycles" side of the current hurricane discussion (the other side being the "global warming" crowd) are in the groups that use these projection techniques. They need the natural cycles explanation to be right for their seasonal projection techniques to be credible, so they're not entirely objective about the whole thing. (And, to be fair, the folks on the global warming side tend to be people who have a reputation for attention-seeking. They just love the limelight, and their current stand puts them smack in the middle of it.)

 

I don't mean to suggest that the research is invalid on either side, because it has all undergone considerable scrutiny by outsiders in the peer review process. But such biases have ways of creeping in to everything. HPH

[SkepticLance]

Dr. Cloud.

Well done. You just put up two very good postings.

 

My view has been that intial conditions in hurricane formation are the result of winds that are, in turn, the result of temperature gradients. I think we agree on that.

 

I appreciate that later development of a hurricane is influenced by many other factors. However, I feel that it is likely that initial conditions must be one factor at least, and those initial conditions are influenced by temperature gradients. Indeed, it appears reasonable that a lowered temperature gradient should reduce the intensity of the winds that lead to hurricanes, and should thus generate fewer hurricanes. That is, of course, deduction, and could well be wrong.

[D H]

To get back on topic, both the National Hurricane Center and the forecasters at Colorado State have recently downgraded their expectations for the 2006 Atlantic hurricane season. See this news release, for example.

[DrCloud]

^ Thanks for pointing that out -- I've been on vacation, offline, mostly, and hadn't seen it.

 

At this point, of course, they're sort of shutting the barn door after the horses have left, because the official season is about 40% over. To do a real climate projection of any kind, you need to make your call before what you're calling begins to happen (or doesn't). After the fact "predictions" or projections are cheating.

 

It's appropriate for them to lower their expectations, however, because this projection effort is more than an academic exercise. Its effect on the public's behavior is relevant, and they need to do whatever mea culpas are appropriate and try to stay as credible as possible.

 

And as far as I'm concerned, a less active season than expected is just fine indeed. HPH

[Phi for All]

As a side note, I can tell you that post-Katrina insurance money came in a bit too late for much of the reconstruction that was supposed to happen on the US Gulf Coast. Now developers are holding out to see just how bad this season is before rebuilding to any great extent. If this season doesn't have any major property damagers, look to see one of the most ambitious building booms ever in Louisiana, Mississippi, Alabama and the Florida panhandle.

[herpguy]

My official prediction:

10-11 named storms

7 Hurricanes

4-5 Major Hurricanes

 

An El Nino is forming which should lower the number of named storms. However, with the warm water in the Carribean, I think many of the systems that develop will be catastrophic.

[herpguy]

Tropical Depression Four just formed and it looks like it will intensify into Tropical Storm Debby. However, conditions are unfavorable for it to become a hurricane, which lowers the chances of my prediction to be correct.

 

http://www.nhc.noaa.gov/text/refresh/MIATCPAT4+shtml/212033.shtml

 

and it's also

[herpguy]

The National Hurricane Center now says there will only be five hurricanes this year.

 

http://news.yahoo.com/s/afp/20060901/sc_afp/usweatheryear