Counting Hurricanes Is Like Pulling Rabbits From A Hat...

...You never know when another one is coming. U.S. upgrades storm Karen as hurricane season ends

U.S. weather experts posthumously upgraded Tropical Storm Karen to a hurricane as the 2007 Atlantic storm season drew to a close on Friday, making the year a near-average one for hurricane activity.

The U.S. National Hurricane Center, in a post-season analysis of Karen, said the storm briefly reached hurricane intensity on September 26, with winds of 65 knots, equal to 74.8 mph (120 kph) or just over the threshold at which tropical storms become hurricanes.

The upgrade of Karen took the 2007 season's hurricane toll to six, bang on the long-term average. The 14 named storms that formed exceeded the long-term average of around 10 for a six-month Atlantic hurricane season.

Boy, that sure is convenient for the forecasters who were way off in their predictions this year, isn't it? So, let me get this straight. We are flying hurricane hunter patrols, and have large weather service staffs whose only job is to monitor the tropics for exactly this type of weather activity, and we are only able to identify a storm as a hurricane two months after the fact?

I usually have a rule about things that look like bullshit and smell like bullshit. Chances are they're bullshit. The more these "re-evaluations" take place, and the more we see questions about the validity of what are being called named storms by the National Hurricane Center these days, you cannot help but get the feeling there is a lot of bullshit out there.

Now, there is two ways to look at this. One could say that this simply represents a difference in our data gathering methods which allows for a greater level of precision than was available in the past, or one could say that outside considerations, maybe of a political nature, are influencing the way data is categorized. I'll admit the former seems unlikely when we confine ourselves to the last 10-15 years. I don't see that there has been any dramatic upgrade in our capabilities over that time period. Certainly, the desire to have a homogeneous storm data set should have been enough to keep storm categorization criteria constant over time, but that doesn't seem to be the case anymore.

One is left with the impression that there are those at work who do not care that they are rending the data set useless for making valid comparisons over time. And, of course, there is a small cottage industry that has arisen simply to take advantage of these sorts of limitations in the data set to espouse their pet political projects, ala Kerry Emanuel.

Of course, these are the exact same people who insist that the pre-satellite era data is nearly 100% complete and accurate. Once again, how convenient.

For the time being, the only option I'm left with is to read this stuff and hold my nose.