Media Muckraker
November 12, 2002

The surprise 3.1-inch snowfall last Monday, Dec. 2, resulted in more than 100 Lansing area accidents. Little did I know, as I chugged my car on U.S. 127 that morning, that over 20 of those accidents were taking place at the I-96 exchange just up around the bend. Fortunately, before I arrived at the ice-slick, my instincts got the better of me and I averted a possible accident by turning off I-127 early.

No thanks to the weathermen of WILX-10 (who share double duty as the forecasters for the Lansing State Journal). They had forecast snow, but had never said how much, hinting at just an inch or so.

Then it happened again. On Tuesday, Rockcole, Provenzano and Drummond predicted a low temperature “near 10.” In fact the mercury fell to 18 degrees below zero, the day’s lowest temperature since 1869!

How could the weathermen be so wrong? I decided to do a little weather muckraking.

In Britain, earlier this year, Ben Magoo wondered about the accuracy of the BBC’s weather reporting after the sunny vacation day they predicted for him turned out soggy. “Is the super computer in the [BBC] office accurately modeling the world’s climate, or is it resting its brain and picking out sun and rain symbols at random? We will find the answer!” Magoo developed a computer program to automatically analyze their weather data at 10 sites, including York, the Tower of London and Cambridge. Here’s what he found at Cambridge:

Incredibly, the chance of the next day’s forecast being right was just 55 percent. Note that Magoo ignored the same-day predictions, making “the assumption that predicting today’s weather is dead simple, so the BBC couldn’t possibly get this wrong.” Really now?

Turning to Lansing, I analyzed 14 days of WILX-LSJ forecasts between Nov. 24 and Dec. 7. I determined a forecast to be in error if at least one of the following occurred: 1) the predicted temperature was incorrect by 5 degrees or more (for either the high or low); 2) precipitation did not occur as predicted (e.g… they predicted snow, but there was none, or the converse), or 3) the precipitation prediction was off by 100% or more (e.g.,. they predicted 1 inch of snow, but it snowed 3 inches, a 200 percent difference).

Remarkably, my analysis demonstrated that the WILX-LSJ forecasters were unable to predict the day’s weather – for the same day – a full seven of 14 days (50 percent)! The British chap had evidently presumed way too much. Distant predictions tended to be about 50/50, with fifth day a poor 20 percent.

You’d figure that predicting the weather a few hours hence would be a breeze. But they missed 3.1 inches of snow on Dec. 2 and were off by 28 degrees on Dec. 3. On Nov 29, the LSJ predicted that day’s weather would have a high in “the upper 30s,” which was significantly lower than the actual high of 46. And on Dec. 4, the LSJ predicted a low temp in the “low teens,” which was a far cry (for the freezing news carriers delivering the newspaper to your doorstep) from the actual low of 4 degrees below zero.

All tolled, of 60 days forecast, the accuracy rate was just 43 percent. Don’t believe it? Check it out for yourself, the evidence is in the library (the other TV weathermen do not have evidence so accessible). Lansing’s numbers are remarkably close to the Cambridge study, suggesting that this level of miscalculation might be consistent over the entire year.

One moral is to not rely on the forecasts to plan time off work.

At the very least, weathermen should humbly state the truth; there is a 50/50 chance that our forecasts will be wrong in at least one important area.Incompetence? Arrogance? It goes much deeper than that.

In Oscar Wilde’s “The Importance of Being Earnest,” Jack comments on the weather thus, “Charming day it has been, Miss Fairfax.” To which Gwendolen Fairfax replies, “Pray don’t talk to me about the weather, Mr. Worthing. Whenever people talk to me about the weather, I always feel quite certain that they mean something else. And that makes me so nervous.”

It’s true. Weather forecasts are less about the weather than about cementing social relations – telling you who has authority. While weather seems so bloody innocuous, in fact, culturally speaking, the weather forecast is a covert agent of social control.

It doesn’t matter to the mainstream media bosses that weathermen are wrong most of the time (if they even know it). What’s important is that weathermen exude an aura of certainty (precision numbers) while expressing an undercurrent of fear (of the possible storm). Just like the IRS, the traffic cop or your boss, no matter how wrong, he’s the person in charge – with certainty. There’s no way out. That’s one hidden message.

The good news is that they’re wrong!

Here’s what needs to be done. Lose the “Stormtracker” and hire a muckraker. Don’t circumvent serious issues like the amount of PCBs in the morning’s snowfall, or the amount of soot in a Lansing fog. And tell the viewers/readers where the historic danger spots are (like I-127& I-96) before the next snowstorm.

Here’s my forecast. Under the current corporate structure, they’ll never do it.

Alex Peter Zenger is the pen name for the Media Muckraker. It is inspired by the work of John Peter Zenger, one of the founding fighters for press freedom in the United States.