Commentary

Blowin’ in the Wind

This article originally appeared on American Spectator on August 1, 2005.
Five named storms before July 15! Obviously we’re headed for some kind of record year. Or are we? And what does all of this have to do with planetary warming?

The answers aren’t clear. Hurricane (and tropical storm) data are notoriously noisy from year-to-year, and the way that we gather these statistics hasn’t been constant. Prior to 1945, most hurricane information came from merchant shipping in the Atlantic. It’s likely that many storms were missed, especially those that didn’t cross major shipping lanes.

After World War II, we began to send out aircraft (called Hurricane Hunters), but these missions weren’t launched unless there was some evidence of a storm, i.e. weather reports from an island or a ship. Only since the mid-1960s has weather satellite technology been sufficient to detect all storms.

Since World War II the Hurricane Hunters have detected an apparent long-term decline in the average maximum winds measured in a given year. Apparently hurricanes are getting weaker.

Important word, that: “apparently.” It’s possible that the decline in average maximum winds is real. Some scientists (not this one) think global warming increases the frequency or intensity of El Nino, the big Pacific Ocean weathermaker. El Nino activity and hurricane intensity in the Atlantic are highly anticorrelated.

The same result — an apparent decline in maximum winds — would obtain if we were now naming more tropical storms than we used to. In that case, the apparent lack of any trend in long-term frequency would be wrong; the true picture would be a slight decline.

It’s complicated, okay?

Equally convoluted is the relationship between the number of early-season storms and overall annual activity.

One would think that it would be very straightforward. The peak of hurricane season is around September 12. If numbers tend to fall off smoothly on both sides of that peak, then a large number of early-season storms would signal a banner year.

But hurricanes aren’t evenly distributed at all throughout their season. Rather, they tend to “bloom,” with multiple storms occurring simultaneously (or nearly so). When that occurs, the storms in each pod tend to be on pretty similar tracks — which is why Florida got whacked four times last year in six weeks. A worse shellacking actually took place in northwestern Mexico in 1933, as six storms hit in fairly close proximity.

The relationship between early and entire-season hurricanes is non-existent. Only 3 percent of the year-to-year variation in total number of tropical cyclones per year (since 1945) is explained by the number observed before August 1. Statistically speaking, that number is indistinguishable from zero.

Further, there is no relationship between the number of severe hurricanes (“category 3” or higher) and early-season activity.

There is a pretty good reason for all of this, and it can be found in a study of recent Hurricane Dennis.

Dennis was an extremely threatening Category 4 hurricane on Sunday morning, July 10, a mere six hours prior to landfall. But, as it approached the Florida coast, it encountered the water disturbed by tropical storm Cindy just six days before (behold the “clumpiness” of hurricanes!). The rough seas mixed colder water with the hotter surface layer, reducing the heat energy available to maintain Dennis, who shriveled faster than a (provide inappropriate metaphor here).

Hurricanes require hot water. But in years where there’s a lot of early activity, there’s going to be less of that around, which means, in general, that later storms will tend to be a bit less frequent or weaker. At least that’s the theory. Reality is more complicated, as the relationship between water temperature and hurricane intensity is far rougher than indicated by simplistic computer models with warmer oceans.

Here’s the bottom line: Since World War II, there is no significant relationship between what happens in the entire hurricane season and what happened early in that season.

But then there’s 1933, the same year that Mexico got pounded. Five storms were detected before August 1, which ties the record (observed three times) in the postwar era. The entire season saw 21 tropical storms and hurricanes, a record that still stands today.