Gonzalo Barr

Hurricane season starts today, which means that we can look forward to another round of dumb reporters blown away live on television, like this –

It also means that the National Oceanic and Atmospheric Administration Climate Prediction Center (the “Center”) has released their prediction of the number of named storms, hurricanes, and major hurricanes they expect to form in the North Atlantic before the conclusion of this season on November 30, 2009.In 2009, the Center is predicting 9-14 named storms, with 4-7 becoming hurricanes, and 1-3 of these becoming major hurricanes.

We won’t be able to tell how accurate this prediction is until the end of the year, once the season is over. But if past predictions are any indication, I choose to remain skeptical.  Let’s look at the past five years, from 2004 through 2008, inclusive.

2008

In 2008, the Center predicted 12-16 named storms, with 6-9 becoming hurricanes, and 2-5 of these becoming major hurricanes.

The records for 2008 show that there were 16 named storms, with 8 becoming hurricanes, and 5 of these (Bertha, Gustav, Ike, Omar, Paloma) becoming major hurricanes.

All three categories (named storms, hurricanes, and major hurricanes) came within the margin of predicted storms, even if they did so at the upper end of that margin. OK, three out of three. Not bad.

2007

In 2007, the Center predicted “a very high likelihood of an above-normal hurricane season,” with 13-17 named storms, 7-10 becoming hurricanes, and 3-5 of these becoming major hurricanes.

The records for 2007 show that there were 15 named storms, with 6 becoming hurricanes, and 2 of these (Dean, Felix) becoming major hurricanes.

While the number of named storms came perfectly within the margin, the number of hurricanes was less than the lowest number predicted, as was the number of major hurricanes. One out of three is not good.

2006

In 2006, the Center predicted a “very active 2006 season,” with 13-16 named storms, 8-10 of these becoming hurricanes, and 4-6 of these becoming major hurricanes.

The records for 2006 show that there were 9 “named storms” (10 tropical storms), with 5 becoming hurricanes, and 2 of these (Gordon, Helene) becoming major hurricanes.

Here, the Center over-predicted the number of all three categories by a substantial margin. Zero out of three is even worse.

2005

In 2005, the Center predicted 12-15 tropical storms, with 7-9 becoming hurricanes, and 3-5 of these becoming major hurricanes.

The records for 2005 show that there were 27 tropical storms, with 15 becoming hurricanes, and 6 of these (Dennis, Emily, Katrina, Maria, Rita, Wilma) becoming major hurricanes.

It’s pretty obvious that 2005 was likely the Center’s worst year for the accuracy of their predictions. The number of tropical storms was 180% of the predicted number, the number of hurricanes 167%, and the number of major hurricanes 120%. Zero out of three.

2004

In 2004, the Center predicted 12-15 tropical storms, with 6-8 becoming hurricanes, and 2-4 of these becoming major hurricanes.

The records for 2004 show that there were 14 tropical storms, with 9 becoming hurricanes, and 4 of these (Charley, Frances, Ivan, Karl) becoming major hurricanes.

Not a bad year, especially when compared with 2005, but not stellar either. While the number of tropical storms was within the margin predicted, the number of hurricanes exceeded the predicted number and the number of major hurricanes fell in the margin at the upper limit. Two out of three.

What do we make of these predictions? They would be no more than a curiosity for the average person, except for the fact that insurance companies use these predictions to set premiums for any area that could be affected by a North Atlantic hurricane.

Image:  Hurricane Wilma crosses Florida, GOES satellite (October 2005); Sources:  all data obtained from the National Oceanic and Atmospheric Administration Climate Prediction Center website and the National Hurricane Center website (accessed June 1, 2009)