Here it is a perfectly seasonable October and already the winter of 1982-83 has people shivering.

The folk of folklore say frog spittle is thick, the hornets have nested high in the trees and woolly bear caterpillar coats are tellingly thick and dark. Meteorologists have arrived at similar conclusions by studying volcanic dust, jet stream conditions and sunspot cycles. Even the publisher of the Old Farmer's Almanac, with the fabled black box of secret formulas, predicts this winter will bring the eastern half of the United States "an incredible amount of sustained cold."

Over the last month there has been a plethora of apocalyptic talk about the upcoming winter weather and a rare consensus among academic weather forecasters that parts of the country are headed for the "worst winter of the century."

But if the subject is weather, a consensus can be announced when all of two people agree. Traditional forecasters have turned a cold shoulder to reports of a frigid winter, all but relegating them to the category of prognostic techniques such as the one based on the size of the pancreas in freshly butchered pigs.

"It's all baloney," says a testy Joel Myers, a meteorologist whose commercial weather forecasting company, ACCU-Weather, has been beseiged by callers wanting to know what the winter holds. "The truth is there is no scientific way to forecast the weather beyond 10 days."

For their part, the scientists of the National Weather Service also disclaim dire reports of a deep freeze exceeding the record marks set in 1977-78 and 1917-18.

"There is absolutely no way they can reliably forecast an extreme winter," says Robert Livezey, deputy chief in the prediction branch of the Weather Service. "They could do it 100 times a century and they would be right once . . . . We don't attach any credibility to these forecasts and the public ought to be wary of them."

Cautiously cleaving to statistical probabilities, the Weather Service offers a 90-day forecast that ventures only to suggest temperatures through December have a better than even chance of being below their seasonal averages. As for January and February, weather watchers will have to wait until the annual Weather Service post-Thanksgiving press conference on The Winter.

But the handful of academic meteorologists whose extreme predictions have sent a shiver of dread through the East, and caused more than a few people to gaze up at the sky wondering what malevolent natural forces are mustering there, stand by their forecasts.

"In certain sections, such as the southeast part of the United States, it could be the worst winter of the century," says Hurd Willett, a professor emeritus at the Massachusetts Institute of Technology who bases his forecasts on sunspot cycles. "It will be 5 degrees below normal in the Northeast, 6 to 8 degrees below normal in the Southeast. The coldest months will be January and February."

Another forecaster, Douglas Paine of Cornell University, is calling for the winter to start early with temperatures in the eastern United States sinking as low or lower than readings during the winter of 1976-77. That winter, the 90-day average for December, January and February in Washington was 5.8 degrees below the seasonal mean of 36 dgrees. Paine is calling for temperatures 6 degrees below normal for the Washington area, and predicts the mercury will plunge from 8 to 11 degrees below normal in parts of the Midwest -- a phenomenal drop that would spell record cold and boost heating fuel demand as much as 15 percent. Meanwhile, Paine expects the West will enjoy a mild winter.

"We think we're on to something that could be a breakthrough in ultralong-range prediction," said Paine, who has been making national weather predictions for four years and claims a high rate of accuracy, using a technique that evaluates the effect of sunspots and ultraviolet radiation on the stability of the upper stratosphere.

And for what it's worth, the Old Farmer's Almanac, which has been forecasting the weather for 191 years, is also calling for extreme cold weather east of the Rocky Mountains. Throughout the East, temperatures are supposed to be 5 to 7 degrees below normal from January into April, owing to a "weakening in the jetstream," according to publisher Rob Trowbridge.

A black box on Trowbridge's desk contains a "secret forumla" passed down from the Almanac's founder that, supplemented with solar data and weather information collected since 1641, enables Almanac forecasters to predict weather trends.

However, as they made rounds on the television circuit this month, Trowbridge and Almanac editor Jud Hale kept a pair of woolly bear caterpillars with them as "props." If the fate of those two woolly bears is any indication, an Ice Age is at hand: Hale carried his around for three days before he realized it was dead, and Trowbridge last saw his woolly bear, which he housed in a jar, whisked off the window ledge of his 14th story hotel room by a chilly gust that may or may not be a harbinger of things to come.