Forecaster Failure

Have our forecasters failed us?

We have just had Tropical Storm Watches posted for Cape Cod, the South Coast and the South Shore. It's Monday morning. Jose is coming Tuesday and Wednesday.

This isn't Harvey or Katrina. This storm isn't going to level your town, pour 50 inches of rain on us, smash a poodle's face with a brick or rip out your heart and show it to you. It should be like a big nor'easter, with a limited reach inland.

The forecasters did a crack job of calling it, for a while. While a loopty-loop isn't unheard of, it is impressive to call it and hit it. Then, the storm moved north.

If you listened to Boston weather forecasters, you would have believed that the storm was curving harmlessly out to sea. WBZ had a "No Way Jose" entry up at the top of their weather blog half the week, forecasting a damage-free drang nach osten for the storm that is instead about to impact us.

It could still miss us entirely, but that's beyond the point at the moment.

I recall them (and others) even chiding people for pointing at spaghetti plots and saying "A dangerous storm has a small chance of hitting us." The pros were contemptuous. "Those aren't forecasts, they are possible forecasts," they told us. "Don't believe that junk on the internet. I guarantee that there will be no Tropical Storm Watch."

Whoops!

I don't want to hold the Big L up at them or anything. I'd be a hypocrite. I stated more than once that I believed that this storm would turn out to sea, although I did get started in media as a sportswriter.

We used to bet on landfalls when I wrote for Cape Cod TODAY, and I would always bet on the sudden hook to the east at the last moment. It's the safe bet on Cape Cod, not quite akin to "the sun will set this evening," but still an in-the-black bet over the years.

"We know that too, Stacey," states Barry Burbank. "That's why we forecast them to go out to sea."

"It's impossible to predict what a hurricane will do five days out with our present technology," says Pamela Gardener.

Shiri Spear pipes in with "The National Hurricane Center said the same thing that we said, and they do nothing BUT study hurricanes. They also correctly called a loop, no mean feat."

"Tropical system forecasts have a 150 mile chance of error every day out," points out Bri Eggers.

Here comes Kevin Lemanowicz. "If we got jumpy about every tropical wave that formed in the Atlantic, we'd be in a non-stop panic from June to November."

"If we treated Spaghetti as accurate forecasts, we'd be predicting landfalls concurrently from Miami to Maine for the same storm," notes Eric Fisher.

"And people would tune us out, thus missing a forecast where we someday correctly call dangerous weather conditions," adds Cindy Fitzgibbons.

"Even God blows a weather forecast now and then," says Dick Albert (who might be dead, but I can't think of another local weatherhead name). "It's the nature of the beast."

I would agree with each of them.

I'd also respond with "A tropical storm was coming at us. You said it would curve out to sea well before New England. Instead, it is about to impact us."

Where did we fail?

TV weather forecasting is a touchy business, especially in Boston. We have 4 major channels, plus other weather sources on NECN and TWC. Each one runs the weather at about the same time. Each channel has a girl that a man could stare at reading the weather each morning, and each channel trots out their most believable weatherhead for the 6 O:Clock.

If you fail to get the ratings, you'll be dumped faster than a last-minute prom date. From there, you fall into the smaller markets, eventually broadcasting from a minor Idaho market where you stand in front of a blackboard with Idaho drawn in chalk, and you have a little sock puppet for storms that you have to hop across the screen yourself if there's a chance of rain. "Here comes Mister Thunderstorm!" says a newly relocated Shiri Spear in a falsetto voice, crouching so as not to appear on camera with the sock on her arm as she opens the puppet's "mouth" in a manner suggesting that it is about to devour Boise. That's a far cry from six figures in glamorous Boston, wearing $500 dresses every morning.

Those are the stakes.

With that in mind, you can understand how a weatherman would want to make to safe choice. If you believe that a storm is going to curve out to sea, if history tells you that storms always do that around here and if the NHC backs your play, there is nothing wrong with saying that the storm is going to be harmless.

Unless it isn't.

I see two solutions to this.

The first one involves a basic change to forecasting styles. We need to be warned if there is even a slight chance of a storm hitting us. I think it should be done so as to not incite unnecessary worry, but it should be done. "Some of the spaghetti has the storm hitting in our area, but that equates to about a .0001% chance at the moment." They can then worry more if the tracks merit it.

This involves retraining the public as well as the forecaster. Ideally, you want someone who can look at a spaghetti plot with a strand or two hitting Boston and just become aware of it. No panic, just a little mental notice to watch this storm as it progresses.

Whenever we post spaghetti plots on this site, I get grief from people who think that we are trying to cause a panic. I respond that the true panic is when a storm catches people off-guard.

We could instead have a well-informed and never-scared public, getting their info from weather experts who now have the license to free-lance a bit with the long-range forecast. If a hurricane sneaks up on you then, you only have yourself to blame.

The other solution is more fun, and I may do it myself if I think that money can be made consistently from it. We should rank our weathermen.

There should be a site somewhere that keeps track of what the local weather people forecast, and score it based on how the day works out. There should be a special category for events like hurricanes, tornadoes, heat waves, deep freezes and blizzards. How far off these events are called would factor into the score.

The scores would be based on a Boston forecast. However, just to keep them on their toes and to avoid a Ripton-like event of suburban/rural ignorance, I'd have the site throw a small town name up every morning for the weathermen to make a detailed call for.

In return, we'd name an MVP every year. The MVP would be the best forecaster in Boston.

I'd like to be involved, but I'd want the face of the company to be an ex-weatherman (or woman... I only say "weatherman" because it generally has a better cadence with whatever I'm writing), someone who got canned for being goofy or ugly or or fat or who worked with a bad anchorman and who has a serious axe to grind with the industry in general. I'd like to do nothing other than feed this bitter MFer jokes.

For weather, it would work like a charm. We'd have to be detailed, and we'd have to be fair. We'd be the only people on Earth doing what we're doing, and we'd have a model that could  be expanded into any market with more than one TV station.

This would both harm and help the local forecasters. The harm would come because someone would have to be at the bottom of the rankings, even if the difference between them and the MVP was "He said six inches of snow and we instead got nine."

The converse works, as well. "I was a three time MVP in the only town where someone ranks weathermen" is a nice to thing to have on your resume if you're applying to work for Good Morning America or to host Chronicle.

They'd also be better weathermen if a Sword Of Damocles were hanging over their heads. They'd work harder, that's for sure. They'd be stronger in a regional manner, because they'd never know which small town we'd be calling until 4 AM or so.

I think that it would also create a supervisor job for another forecaster, especially one with immense local knowledge who is a bit long in the tooth to be in front of the camera. If there's a chance that a site using real data will name you as the least accurate forecast in Boston, you're going to want your game to be airtight.

Either way, the present situation needs to change.

A storm has snuck up on us, in 2017 AD.

Our best minds with the best equipment told us it was harmless, right until it wasn't.

I doubt that anyone will be held accountable.

The public will forget soon enough.

The mistake will happen again.

It shouldn't happen again.

We need to fix the system.