In the whimsical world of weather prediction, the ordinary is inexplicable, the uncommon is commonplace.

East of the Rockies, it has been the coldest fall and winter since the 1880s. In most of the West, the weather has been warmer than usual. It has snowed in Florida, while the sun shone in Alaska. Swimmers frolicked in the Pacific while, in New York, two elederly men froze to death.

Amid these erratic fluctuations, Donald L. Gilman, chief of the National Weather Service's long-rang prediction group, has maintained the outward calm of a man who knows his limits.

"There are always going to be these swings from one season to another." says Gilman, who has worked for 18 years among piles of squiggly-lined weather maps and computerized temperature charts.

"Nineteen seventy-six is a standout year, but there have been other freaky years. Each year is freaky in its own way."

The 45-year-old meterologist who operates out of the World Weather Building, a glass-enclosed high-rise in Marlow Heights, has become a celebrity of late, beseiged by television networks and news magazines.

All are clamoring to know why it is so cold here and so warm there, and whether it will get colder or warmer next week, next month and next year.

To such questions, Gilman smiles and says, "If there are any ultimate explanations, we don't know what they are."

Meteorologists can, however, observe what is happening. Since mid-September, the Westerlies - high speed winds that eternally swirl around the Northern Hemisphere miles above our heads - have maintained an unusual pattern.

Rather than blowing steadily across the northern section of the continent with periodic northward and southward undulations, the winds swung into a winter pattern early this fall and have remained virtually fixed for four months.

They are carrying warm Pacific air up to Alaska and Canada, where they pick up frigid Arctic cold and head southward through the Midwest and down the East Coast. Continuing out into the Atlantic, they push warm air from the Gulf Stream up to Iceland and Greenland, creating unusually mild weather there.

This steady pattern, which might normally occur for only a few weeks during the winter, is partly responsible for the Western drought that is drying up ski slopes and threatening agriculture.

Instead of blowing Pacific storms into the West Coast mountains and the Rockies - where snow would then accumulate, eventually melting into spring irrigation water - storms are breaking in the ocean off Southern Alaska. Likewise, the winds have created unusual storminess over the Atlantic.

When it comes to explaining why this pattern has occurred, all the sophisticated computers and mathematical models of the weather service are of little help.

Not that science is totally at the mercy of capricious wind-goals. Meteorologists know that wind patterns are influenced by variations in sea temperature, cloudiness, rainfall, air pressure, sunshine and even dust and chemicals spewed into the air by man's factories.

But exactly how all these elements join together at a given moment is not always understood. "The atmosphere is a mysterious, complicated physical system, Gilman said. "Causation is hard to analyze because everything that happens influences everything else."

Most of the Weather Service's 6,000 employees are engaged in short-term prediction, monitoring day-to-day weather and forecasting two or three days ahead. Only eight people, headed by Gilman, are involved in "long-range" prediction - trying to forecast five days ahead, a month ahead and three months ahead.

The weather service's five-day predictions are correct roughly 80 per cent of the time. When it comes to monthly and seasonal predictions, the odds are about 60-40. (However, Gilman's group did correctly forecast the current cold wave, in testimony two months ago before a congressional subcommittee.)

"Long-range predicrion is in the twilight zone of forecasting," Gilman said. Although a cold fall frequently indicates a cold winter, winter weather gives few hints about spring. February and the rest of the year, could be warmer or colder than usual depending on complicated and unpreditable atmospheric physics.

"There's no statistical model of the atmosphere to guide us," Gilman said. "We're always groping around for a few meterological clues to put together into a consistent image. Basically, it's a lot of head-scratching."

If this winter's weather tells little about the spring and summer, it reveals even less about future climatological change. A dispute is raging in the scientific world over whether the earth is becoming gradually warmer or colder.

"I don't think you can make any connection between this winter and the recent public argument that we're descending into a new ice age," Gilman said. "This isn't part of any long-term warming or cooling trend. It is a very regional thing with the extremes confined to only half of the Northern hemisphere."

As far as Gilman is concerned, other scientists may speak in lofty terms about decades and centuries hence. But the Weather Service is busy with the impredictable present.