"NOT A FORECAST," read the first line of the Facebook post. Below it: An image projecting Irma, a "strong hurricane," making landfall near Washington, D.C., on Sept. 11.
The postoriginated from central Pennsylvania-based S&S Storm Chasing and Forecasting team. And it was shared more than 9,000 times.
The page alsoposted several similar imageslast weekend, usingmodels projecting more than a week in the future.
More:How will Hurricane Irma impact Pennsylvania? (Sunny skies this weekend, for one thing)
Chilling footage:Caribbean webcam destroyed as Irma, category 5, strikes
Had it happened, the event shown in the image would have been catastrophic for central Pennsylvania and surrounding states.
We now know the early model was wrong.
Ask a meteorologist, such as Penn State's Dr. Jon Nese, and he'll likely be critical of such long-range projections.
Ask S&S, and it will provide a different perspective: Prepare for the worst.
Long-range forecasts aren't reliable
There's little disagreement on this point: Long-term forecasts are not known to produce trustworthy results when it comes to hurricanes.
"You can’t forecast anything nine, 10 days out," acknowledgedScott Pennewill, S&S lead storm chaser. That's why S&S iscareful to label such posts "not a forecast."
Nese provided more specifics, sayingthe science simply isn't there to reliablyforecast the potential impact from a hurricane to a specific location more than a week in the future.
In a recentepisode of Penn State's "Weather World" show, Nese explained that hurricane landfall forecasts in the 10-day range can vary by hundredsof miles.
More:Hurricane Irma: Where is the Category 5 storm now and where is it headed next?
One day a model can predict landfall in Florida –the next, it might projectNew England.
That's why meteorologists run computer models as many as 51 times so they can see the different possibilities, Nese said.
He also cited a National Weather Servicetweet posted last weekend warningagainst "fake" forecasts. It said theHurricane Center's official forecasts only go out five days.
Included with that tweet was a five-day projection that proved to reliably track Irma's path.
'At least you're prepared'
Although itacknowledges long-rangeforecasts are unreliable, S&S makes judgment calls about when to post snapshots of them, according to Pennewill.
The group, with a following of more than180,000 on Facebook, ismade up of nearly all volunteers living along the East Coast.
Pennewill said S&S willcollaborate when a big weather event is approaching, especiallyseeking input from the meteorologist in the group: New York-based Joe Cioffi, who claims more than three decades of forecasting experience.
When the group's instincts say the public should prepare for the worst, they turn to long-range forecasts.
“We get no joy ... out of hype," said Pennewill.
More:Hurricane Irma a Category 6? That's fake news.
It's not about providing a bulletproof forecast, and it's not about clicks, he claims.
It's about raising the public's awareness –something bad might be on the horizon and you may need to plan ahead.
It's about being prepared.
Maybe some people will make back-up plans for their vacations. Maybe they will buy some non-perishable food supplies.
What's the harm in that, he wondered.
Crying wolf
Nese sees it a different way.
While not commenting on S&S specifically, he saidpromoting or publicizing an unreliable long-range forecast"hurts the credibility of the entire weather forecasting enterprise – too many ‘cry wolf’ situations and people start to tune out.”
He went sofar as to call those types of projections"irresponsible," saying they do more harm than good.
"There are lots of computer solutions out there," he said. "By picking the one that looks scariest for your location, you're really doing a disservice ... particularly, beyond about seven days."
WGAL-TV meteorologist Matt Moore agreed that weather information should be disseminated cautiously so as not to cause unnecessary fear.
"I don't think there's anything wrong with informing people of any possibilities that might be," he said. "The line you have to make sure you don't cross is panicking the public, especially for something very far out in the long range."
Instead, Nese suggested more cautious approaches. Hecited meteorologists who –a week out from Irma's U.S. landfall –advised planning ahead without speculating where the hurricane might land.
Pennewill sympathizes with Nese's concerns: “I can see where he’s coming from … you don’t want to send the wrong message, obviously.”
But in the end, he said,S&S errs on the side of publishing speculative projections when it'sbelieve to bein the public's interests, based on S&S'sreadings of the models,knowledge of theaudience and their historical experiences with extreme weather.
Different roles
S&S isn't critical of mainstream meteorologists' caution in itsown forecasts.
“We respect every degreed meteorologist … they are above us,"saidPennewill."We just want to follow in their footstepsand just one day be as good as they are."
He says S&S's followers understand that the group serves a different role than more official and professional weather organizations.
If people want to know the most accurate information about severe weather, they'll go to more professional sources.
If people want a more raw, engaging outlook, S&S is the place to go.
Engagement is key, Pennewill said.
The group prides itself on talking directly with its social followers. Although, as the page has noted multiple times, it has strict rules against "bashing or being combative."
Engagement extends past social media, according to Pennewill. He says the group is regularly involved in events ranging from relief efforts to supporting local charitable events.
Talk to Nese about the role ofPenn State's Department of Meteorology and Atmospheric Science, and you'll likely hear a different emphasis: They strive to be a trustedsource.
Nese saidthat means going beyondforecasting toeducate the public about the limitations of weather forecasts. He hopes this will let themdifferentiate between real and fake or misleadingweather news.
