Entertaining but unreliable.
Like the blaze of color that has brightened North American woods, the appearance of the latest edition of The Old Farmer's Almanac is a familiar sign of autumn.
It also symbolizes the season for long-range forecasts of the winter. As with the Almanac itself, these will have some use but should be taken with a pinch of salt.
The ''Old'' in the title refers to the publication, not the farmer. First issued in 1792, it is the oldest continuously published periodical in the United States. It retains its old-time flavor with an amalgam of useful astronomical tables, folksy articles (such as one on how to catch and cook catfish), and ads for various home remedies and gadgets - all seasoned with pseudo-rustic wit. But , for its anticipated 4 million-plus buyers, the core of the Almanac is its weather forecast, made for 16 regions of the US for each of the next 12 months.
Editor Jud Hale, publisher Rob Trowbridge, and their consultants - using the collective pseudonym Abe Weatherwise - claim 80 percent accuracy for Almanac forecasts, according to the Associated Press. That's as may be, for there is precious little scientific evidence to back up the claim. Certainly the regional forecasts that predict monthly average temperatures through October 1985 to within half a degree F. reflect a presently unattainable precision.
As Donald L. Gilman, chief of long-range forecasting at the US National Weather Service, notes, forecasts of average regional temperatures and precipitation a month or season in advance have marginal skill. Anything beyond that is speculation.
Forecasting scientists do see the possibility of useful longer-range forecasts. Research over the next five years could bring significant improvement. But the most that has been scientifically demonstrated so far is the marginal skill at forecasting a month or season in advance. And Gilman cautions that even this is ''more like cooking than ... science. Everybody has his own recipe.''
Given this perspective, the Almanac's general winter forecast (in contrast to the specific regional projections through next October) is probably as good as anybody's at the time it was issued, although the publication makes a mystery of its recipe. This is said to consist of ''a secret weather formula devised by the founder of this almanac in 1792 and ... the most modern calculations based on solar activity.''
It yields a forecast for November through March in which the region east of the Rockies is expected to be colder and dryer than normal, with below average total snowfall. This will be despite ''relatively mild weather'' during much of November, December, and March and ''well above average'' snowfall in February and March. There will be variations from the general forecast within the region. Warning is given for ''cold snaps'' extending well into the South during November and February and into Florida in January.
''Normal precipitation'' and ''below normal temperatures'' are forecast for the mountain region west of the Rockies. The Far West and Southwest should have ''a mild winter with relatively dry conditions in the Northwest and wetter to the South.''
These forecasts should be taken with several reservations. First, any winter forecast made this early is suspect; Gilman's official National Weather Service (NWS) forecast for December through February won't be issued for another month. This point was demonstrated dramatically two years ago. Several early forecasts, including that of the Almanac, called for a cold winter. Gilman's forecast in late November envisioned mildness. Some of the other forecasters had, by then, revised their projections also. The Almanac, already issued, was stuck with its original projections. The season turned out to be the second mildest winter on record.
Second, the use of solar activity to forecast weather trends is tricky, as well as scientifically controversial. Hurd Willett of the Massachusetts Institute of Technology points out that it involves forecasting solar activity as well as weather. If solar activity changes unexpectedly, weather forecasts based on it will be invalid. This happened for the 1982-83 winter season, Willett says.
Third, the Almanac forecast extends from late fall into early spring, rather than covering just three winter months. Weather is changing rapidly at both ends of that period, so the forecast is not strictly comparable to those the NWS and some private or university meteorologists will issue.
Familiar, entertaining, and sometimes useful, The Old Farmer's Almanac is a welcome harbinger of autumn. But as for the 80 percent reliability unofficially claimed for its weather forecasts, readers should note the official disclaimer printed in the Almanac: ''Neither we nor anyone else has as yet gained sufficient insight into the mysteries of the universe to predict weather with anything resembling total accuracy.