The Fine Art of Weather Prediction

Update: Don explains my mistake in the comments.

Is not enhanced by contradiction. Over on Chizumatic, there’s a multi-post discussion going on wherein NOAA’s Storm Prediction Center sounded the alarm for a day of massive tornadoes across the Plains and Midwest, aka “Tornado Alley.” Or didn’t… What I read didn’t sound very alarmist at all, in fact it sounded like a pretty ordinary day: “moderate” chance of thunderstorms, “slight” chance of tornadoes. etc. Here’s the map the SPC put out a few hours ago, labeled “Catagorical”:

Yawn

Not very exciting, eh? Below is a series of maps issued as a part of an “experimental” briefing by the SPC at 6am CDT this morning. I didn’t find it on NOAA’s site, instead I found it here. (Hat tip to Steven.) Look for the NOAA symbol over Norman, just south of Oklahoma City.

oooh, pretty colors!

Well, that’s a little more exciting, isn’t it? But going back to NOAA’s regular page, and clicking on “Probablistic” yeilds this page:

Now that’s a bit more like the experimental forecast. And here’s some text from The Weather Channel (highlights mine):

Developed by Dr. Greg Forbes, (Find him on Facebook) The Weather Channel’s severe weather expert, the TOR:CON index is an estimate of the likelihood of tornado activity within a given time period.

The TOR:CON values range from 0 to 10. A value of 4 means that there is about a 40% chance of a tornado within 50 miles of a location in the specified area of severe thunderstorm activity. This also means that there is a 60% chance that a tornado will NOT occur.

Areas listed below have an above-average threat of having severe thunderstorms with damaging winds, hail, and/or tornadoes for the specified days.

Saturday April 14
IA northwest 3 – 4
IA southwest – 5
IA south-central – 4
IA northeast night – 2
IA west night – 7
KS central – 9
KS east – 5
KS central, northeast night – 9
KS near Wichita night – 7
MN extreme southwest – 2
MN south night – 4
MO northwest, west-central – 4
MO northwest night – 7
NE south – central – 9
NE north – central – 5
NE central night – 8
NE east – 7
NE east – 6, night – 7
OK northwest – 9
OK southwest, central – 7
OK northeast – 5
SD southeast – 2
SD southeast night – 3
TX east panhandle – 7
TX northwest east of Midland to Wichita Falls – 4
TX northwest near San Angelo to Wichita Falls night – 5
WI west-central,southwest night – 2
Other areas – less than 2

Um… yikes? So much for Steven overstating things, which I thought was the case…I still don’t expect a “Super Outbreak” like we’ve had a time or two, but that’s not saying it will be an ordinary day.

On Chizumatic, I speculate that part of the problem is a difference in styles between NOAA’s Hurricane Center and their Storm Prediction Center. Obviously, living on the Gulf coast, I pay more attention to the Hurricane Center, and in recent years they’ve been getting a bit alarmist, with statements like [citizens] “in the path of the storm face certain death.” Not exactly mincing words, are they? In contrast, the SPC is saying today (Note where you see lower case, I’ve expanded their abbreviations for clarity):

…THERE IS A Moderate RISK OF Severe Thunderstorms OVER MUCH OF Central WI…AS WELL
AS EXTREME Southeastern MN AND Northeastern IA…

…THERE IS A Slight RISK OF Severe Thunderstorms FROM Eastern TX TO MI…

This is followed by:

THE PROGRESSIVE NATURE OF THIS TROUGH AND INTENSE WIND FIELDS SUGGEST SIGNIFICANT WIND DAMAGE POTENTIAL…AS WELL AS A FEW TORNADOES OVER THE UPPER MS VALLEY. …. TREMENDOUS WIND FIELDS WILL BE IN PLACE AND SHOULD EASILY RESULT IN SEVERE WINDS AT THE SURFACE ONCE STORMS GET ORGANIZED.AREAS OF ROTATION WITHIN THE LINE…AND PERHAPS TORNADOES…WILL BE POSSIBLE AS WELL…WITH SWATHS OF PARTICULARLY DAMAGING WINDS. HAIL WILL OCCUR WITH THE STRONGER CORES AS WELL. THERE IS A CHANCE THAT EARLY CONVECTION REMAINS DISCRETE ALONG THE LINE AS SUPERCELLS…WITH AN ENHANCED TORNADO THREAT…ALTHOUGH THIS IS LESS CERTAIN.

That’s it. Two minimal mentions of possible tornadoes, centered around Wisconsin, Iowa, and Minnesota. Most of the attention is on the potential for straight line wind damage. Yet they’ve already had four tornadoes in Kansas reported as I write this. Oh wait, add Oklahoma…

CNN thought a frat-boy pilgramage to an Indiana shrine and a UN observer mission to Syria were more important... to the US edition of their front page.

I’m still not convinced it’s the end of the world, Mayan calender or not, but it would be nice if NOAA could at least speak with one voice. One of the problems is that individual meteorologists are responsible for these forecasts, and some are more willing to beat the drum than others. And of course, when you’re dealing with commercial weather prediction (such as TWC or your local news media) every instance of rain has to be played up into Noah’s flood, in order to capture viewers and sell ad revenue. I’ve complained about this during hurricanes before (Rita and Eduoard in particular), alarmist behavior just results in people tuning you out: e.g. the boy who cried wolf.

One of these days, it’s going to cost us badly.