Groundhogs and Scientists

As we've reported, the groundhog saw his shadow this morning, foreshadowing six more weeks of winter.

Don't be surprised if the groundhog tries to update that forecast before the end of the month.

This is the season to be skeptical of any long-term outlook. This has not been the best of times for the humans who trying their hands and minds at long-range forecasting.

Recall that the early consensus among the meteorologists was that the winter would get off to a cold start with a rocking December.

In Philadelphia, it was the fifth-warmest December on record, and even warmer elsewhere.

Accu-Weather pulled back on that December chill idea and issued an update on Dec. 1, but it still called for a "brutal winter," with Chicago and Minneapolis in the thick of it.

In January, Chicago's temperatures average 6.6 above normal. By comparison, as warm as it was in Philly, the departure from normal was a mere 3.7.

WSI Corp., in Massachusetts, released an update Jan. 24 that sees a cold February and March. Around here, February got off to a late-April start, and no truly cold air is in the extended outlooks.

Perhaps this will all change, but so far hints of a major pattern shift have been misleading.

One thing is certain: The scientists will keep trying to crack this case. Never have more forecasts been available, and never has the speculation been more detailed and fascinating.

Sometimes, they even get it right. One company, Atmospheric and Environmental Research, actually did call for a mild winter in the East.

But the obstacles to reliable long-term forecasting are formidable. The atmosphere is what the scientists call a "non-linear chaotic system."

Numerical models have wrought amazing progress in shorter-term forecasting, and even they are far from perfect. Forecasting a month or two or three ahead is a whole other discipline.

The shorter-term models rely on worldwide observations of the "initial condition" of the atmosphere and then try to calculate how conditions will change over the next few days.

The atmosphere, however, is imprecisely measured because of various data gaps, and errors that creep into the forecasts get worse over time.

In the view of Mike Halpert, the lead long-range forecaster at the government's Climate Prediction Center, print journalism may be obsolete before the seasonal foercasting puzzles are solved.

The government doesn't issue forecasts per se, but probabilistic "outlooks" that list the likelihood that given areas of the country will experience, above, below or just plain normal temperatures and precipitation.

Making a detailed forecast for a three-month period, he said, is "almost impossible."

A predecessor at the climate center, Jim Laver, once remarked that he used to believe the atmosphere was guarding certain secrets and that once scientists figured them out, they'd be able to nail these outlooks consistently.

He later concluded that the atmosphere had no such secrets, that nature may well be making it up as it goes along.

The groundhog was unavailable for comment.