NOAA [JessePedia] released their 2007 Hurricane Forecast today, and I'm a little perplexed. Perhaps it's just too late in the day and some of you readers can help explain this to me. I'm also interested to hear what AccuWeather.com Professional's Joe Bastardi [BIO] (PRO) has to say about this... if I know Joe, I feel a rant coming on. I'll post it here when it happens.

But for now, my beefs with the forecast. First, they state, and I quote: "Experts at the NOAA Climate Prediction Center are projecting a 75 percent chance that the Atlantic Hurricane Season will be above normal this year... NOAA scientists predict 13 to 17 named storms... an average Atlantic hurricane season brings 11 named storms." Below is a handy pie chart showing the probabilities.

First, saying 13-17 is quite a wide range and runs the gamut of all forecasts issued this year by five different groups. But wait, there's more. A prediction of 13-17 is a 100% chance of being above normal, so why do they state the likelihood at only 75%? If there's a 5% chance we'll be below normal and a 20% chance we'll be at normal, doesn't that mean they're predicting 10 - 17? And if their lower number is not for sure, could we apply the same "fudge factor" to the upper limit and end up with a forecast of 10 - 20 named storms?

To me, that's too big of a range to be useful. Most Atlantic hurricane seasons are probably in that range, so their forecast has a high chance of verifying. If you calculate the percentages of their forecasts vs. average, that's like saying the predicting temperature tomorrow for State College will be between 60 and 90. That doesn't help you plan your picnic, if you get my drift. Joe Bastardi will tell you, forecasting is not about verifying, it's about sticking your neck out based on the best data you have.

Second, these maps, which were stuck in with no explanation, and are supposed to compare active to inactive 24-year periods (why 24, I have no idea), are a bit questionable.

Scientifically, I think these results are questionable because the map on the left uses two 12-year periods which are separated in time by the 24-year period on the right. You'd think you'd want to use consecutive years if looking for climatological trends.

Whether you look at a graph of hurricanes per decade, storms by year, or U.S. landfalls by decade, or the 1960's were not impressive years, in fact they were the 3rd least active decade for hurricanes since 1850, so why did they pick that decade? If you want to talk about active years, let's go back to the 1950's.

And what is the conclusion printed on the graphic?

Well, duh, since one defines the other, that would be true. But finally, their real point, which could have been better illustrated (and didn't have to involve the 1960's at all):

That statement IS useful information and it's a shame that it got "clouded" by the other unnecessary information in the graphic. I wonder however why this study was limited to "major hurricanes forming in the main development region," since the forecast doesn't address why this region is important in 2007, and hurricane victims certainly don't care where the storms formed.

Remember, our tax money went to pay for this forecast, so we have the right to question it.

Report a Typo