Mild or wild? Forecasts look toward winter
Winter-storm watches have been posted as nearby as Pottsville. A cold and nasty nor'easter is poised to punish the Philadelphia region for the next two days, followed by another during the weekend.
The solstice is still better than two months away - and, for that matter, most folks probably haven't yet made any hard decisions about Halloween costumes.
But some weather experts are saying that what the region will get in the next few days is a foretaste of the winter to come.
For the Phillies and their fans, Sunday might feel more like a bitter aftertaste, with conditions reminiscent of those of Game 5 of the 2008 World Series. (Note to pitchers: At least the wind should be blowing in.)
Yesterday, AccuWeather Inc. called for the coldest winter in several years around here, with snowfall well above normal, maybe as much as 30 inches. "I do think there are going to be a couple of big-ticket storms," said Joe Bastardi, the commercial service's long-range forecaster.
Another private service, Commodity Weather Group in Washington, has predicted the coldest season in a decade, warning that the atmospheric pattern bears similarities to that of the 1976-77 season, which set all-time standards for chill.
Do not evacuate quite yet, however.
The government will weigh in today, and its outlook is likely to be far more circumspect, and at least one private forecast has come down on the mild side.
And then consider the loneliness of the long-range forecaster, who has to gaze into one of nature's most opaque crystal balls.
The clues are cleverly hidden all over the planet, from Siberia to the International Date Line to Philadelphia International Airport.
"We haven't had the advances in the seasonal forecasting that we've had in weather forecasting," said Judah Cohen, a long-range forecaster himself.
The short-term forecasts, those calling for back-to-back coastal storms, are based on computer models that solve complex equations. Small errors mess up the works, though, and the predictions become less reliable with time. Things deteriorate over weeks and months, Cohen said, as "the errors just build and build and build."
For seasonal outlooks, the computers aren't much help. So meteorologists have to look at things such as overall atmospheric patterns and how they compare with those of past seasons, long-term temperature trends, and changes in the oceans, which are slow to add and lose heat.
With so much uncertainty, the forecasts tend to be generalized. If they were too specific, they would "often be incorrect," said Todd Miner, a meteorologist at Pennsylvania State University. That would violate one of the cardinal rules for weather forecasters: "We try to look like we know what we're talking about," he said.
The outlooks of the government's Climate Prediction Center, for example, confine themselves to assigning probabilities of above- or below-normal temperatures and precipitation for broad regions of the nation.
By contrast, Bastardi's outlook is bold. In addition to a cold and snowy winter in the Northeast, he said that Vancouver, British Columbia, might be snow-deprived in February, when it will host the Winter Olympics. He said Philadelphia would average two degrees colder than normal, which would make it the coolest year since 2002-03, the year of the 20-inch snow on Presidents Day weekend.
Two winters ago, AccuWeather predicted a mild one, and it was. Last year, Bastardi correctly saw it would be the chilliest in five years. However, Bastardi called for "bookends of cold" in December and February, interrupted by a January thaw. It turned out that December and February were milder than normal, and January was quite cold.
"The devil is in the details," Miner said.
Bastardi said that El Niño, the unusual warming of the tropical Pacific from the International Date Line to the South American coast, would be a factor this winter, although he said it eventually would fade.
El Niño affects the west-to-east jet-stream winds that deliver weather to North America. Strong ones can flood the atmosphere with mild air, but that probably won't happen this time.
"We are in a relatively weak El Niño," Miner said, "and a lot of our more interesting, exciting winters tend to happen in that regime." (By that, he means stormy, by the way.)
Cohen, a scientist at AER Inc., a forecast service in Massachusetts, isn't buying into the cold scenario, at least not yet. Cohen is an expert on the Siberian snow cover, which is now below normal.
His research shows that Siberian snow in October is a key indicator of weather throughout the Northern Hemisphere in winter. The snow has a complicated relationship with the atmosphere, and it can take a few months for its effects to take hold.
"We do have a preliminary warm forecast," Cohen said, but he warned that could change by the end of the month - ironically, because of worldwide warming. With warming, snow blitzes have become more common, he said. "You get these advances in snow cover that are really quick. We just haven't seen that before," he said.
Closer to home, another indicator to watch this month is the thermometer at Philadelphia International.
During El Niño years, below-normal temperatures in October all but assure that the ensuing winter won't be mild, based on an analysis by Tony Gigi, a meteorologist at the National Weather Service in Philadelphia. He found only one exception in 17 cases.
So far, October has been just about normal, but that will change during the next several days, and temperatures might not get out of the 40s until Sunday.
Computer models and those who rely on them have problems seeing too far beyond that, but that won't stop the efforts to look farther ahead.
"There's a market for wanting to know the future," Miner said. "That's why there's palm readers out there."
Contact staff writer Anthony R. Wood at 610-313-8210 or twood@phillynews.com.




