Dud of a forecast explained

IT WAS THE 2015 snowstorm that wasn't.

Beginning Sunday, a cacophony of ominous forecasts by Philly TV meteorologists - who relied on Doppler radar, weather models and years of expertise - warned that a whopper of a system was headed our way.

The forecasts resulted in a state of emergency, citywide school closings and a dearth of sliced bread.

By now, we know the feared superstorm failed to arrive. In its place was a light dusting of 1.2 inches of snow sprinkled over Philadelphia.

So how did so many professional forecasters miss the boat?

For one, the system appeared almost out of nowhere on Saturday, said meteorologist John Bolaris, founder and president of WeatherSavior.com. Usually, systems start to appear about five days before they hit.

"When all of a sudden you go from no storm to an intense bomb," the confidence level is low in forecasting, Bolaris said yesterday. "It's strange and it sounds off alarm bells."

NBC10 meteorologist Sheena Parveen said she had her own set of challenges with the storm system known as Juno.

"The challenge was forecasting a storm before it actually developed offshore. Many computer models positioned the storm forming near the New Jersey coastline, and many had it forming close enough to impact part of our area with heavy snow," Parveen wrote in an email.

"Hindsight is 20-20 [and] we now know the position of the storm and the amount of moisture was wrong. An area of dry air over our region with the storm moving away kept the heavy moisture away," she said.

Also, Bolaris said, "The models were correct in the intensity of the storm, but the models' [trends] were all far apart." Normally, disparate weather models begin to converge and "agree" on the outcome 24 to 48 hours before the storm hits.

In Juno's case, that never happened, Bolaris said.

The European computer model, which many forecasters rely on and considered the most reliable, initially showed 14 inches and later downgraded it to 10 inches, Bolaris said. The Global Forecasting System initially forecast a snow dump on the Philly region, but on Sunday the model downgraded its prediction to 3 or 4 inches, he said.

Parveen said she initially forecast 1 to 6 inches Saturday night, then increased that to 2 to 15 inches. By Monday night, Parveen had downgraded the range to 2 to 8 inches, she said.

Also on Monday, Parveen said, she "didn't like how the bands of snow were moving into Ocean County [N.J.], then falling apart. They were not progressing west into the area."

"That was a red flag, so we started lowering totals," she said.

Bolaris, however, said he's glad the region planned for the storm because it was a "ticking time bomb" that could have veered more to the west and dumped more snow in the region. "God forbid we went for lower amounts and then kids are in school and school buses are on the ground. . . . It could have happened."

"Bottom line is, it will never be an exact science," he said.

Parveen and Bolaris weren't the only ones reassessing their forecasts. MyFoxNY.com reported that the National Weather Service in Mount Holly, N.J., will evaluate its storm models after it predicted that New Jersey and the Philadelphia area would receive a foot or more of snow.

The office's meteorologist-in-charge, Gary Szatkowski, apologized on Twitter for the forecast.

"My deepest apologies to many key decision makers and so many members of the general public," he wrote.

In a second tweet, he wrote: "You made a lot of tough decisions expecting us to get it right, and we didn't. Once again, I'm sorry."

6ABC's Cecily Tynan, who was unavailable for comment yesterday, also took to Twitter to discuss her forecast and knee status: "Monday started with a busted ACL diagnosis & ended with a busted forecast. #badday."

Last night, Fox 29 aired a special segment in which meteorologists Scott Williams and Sue Serio apologized to passers-by on the street for getting the forecast wrong.

"We understand people are frustrated, but no one is more frustrated than the ones forecasting the weather," Parveen said. "We don't come to work early and leave late and come in on weekends to intentionally make a bad forecast."

- Staff writer Vinny Vella

contributed to this report.