People swear by farmers’ almanacs. Every autumn these publications arrive with their eclectic mix of miscellany and lore. They also deliver a region-by-region weather forecast for an entire year, notably the coming winter, which everyone from farmers to city dwellers can plan their businesses and lives around. There are actually two competing almanacs: the Farmers’ Almanac and the Old Farmer’s Almanac. Both have been published for more than 200 years, and both highly protect their secret forecasting formulas, which they say have stood the test of time.
Do the almanacs perform better than weather companies and agencies that provide nationwide forecasts that are built on the latest atmospheric science and supercomputer models? And why do people trust the almanacs so deeply?
Among farmers, the almanacs are highly valued and respected because of their long history, says Steve Hu, an agricultural climatologist at the University of Nebraska–Lincoln. “Farmers, over time, have developed a certain kind of trust with their product and information,” he says.
Today people want long-range forecasts for a whole host of purposes, however: to plant commercial crops, of course, but also to plan for energy usage, guide the insurance industry, schedule weddings and vacations, and more. The publications have met that demand, says Sandi Duncan, an editor of the Farmers’ Almanac, which began printing in 1818. And unlike meteorologists, the almanacs are willing to stand by specific days-long predictions made some 18 months in advance.
“We don’t guarantee the forecast 100 percent, but I don’t know who does,” Duncan says. “I think people appreciate being prepared. Even when we’re off, they give us a little more leeway.”
Although the winter weather outlooks are perhaps the almanacs’ highest-profile product, Duncan’s competitor is clear about where its allegiance lies. “We don’t think of ourselves as a weather company; we think about ourselves as for the farmer,” says Carol Connare, editor of the Old Farmer’s Almanac, which was first published in 1792.
But the distinction is perhaps becoming fuzzier as the almanacs face the rise of seasonal forecasts produced directly by weather forecasting companies and governmental agencies. These organizations, aware of the constraints of their predictions, used to shy away from forecasting more than a week or two in advance. Forecasts have improved so steadily that they are now producing 30-day, 90-day and even one-year outlooks that offer a sense of how temperature and precipitation over time may compare with norms.
The almanacs have noticed. “Maybe 10 years ago, people would be like, ‘Well, how can you make long-term predictions like this?’ And yet now there are so many copycats out there that make their long-range forecasts as well,” Duncan says. “Even the Weather Channel now goes out, and they do it. They never did it before.”
The widespread interest in forecasts that stretch ever farther into the future is particularly evident as the temperatures in North America start to drop and visions of the coming winter begin to dance in people’s head. Whether we are dreaming of a white Christmas, dreading another season of shoveling snow or just hoping for a break from a scorching southern summer, few of us manage to be apathetic about what winter will bring.
For the coming winter, the Farmers’ Almanac predicts that cooler temperatures and higher snowfalls will return to the U.S. after last year’s strange warm winter. Its long-standing competitor, the Old Farmer’s Almanac, is also all-in on snow and cold across much of the U.S. But the National Oceanic and Atmospheric Administration says that the northern U.S. will be warmer than usual. And for the southern swath of the country, the agency offers even odds of cooler, average or warmer temperatures.
It’s a confusing spread. “When you have so many things out there, and some of them are in conflict or totally opposite, which ones do you use?” Hu says.
Both almanacs claim authority from the long heritage of the formulas they use. The formulas, both brands say, are still, at heart, the work of each publication’s founder despite some changes over the past two centuries.
The Old Farmer’s Almanac claims an overall accuracy of about 80 percent for its winter forecasts; its staff calculates this based on whether they accurately predict the direction of departure from normal for precipitation and temperature in one city for each region they evaluate. (Last winter, they say, their precipitation forecasts were right on, yet their temperature forecasts were all over the board.) The Farmers’ Almanac doesn’t publish a similar statistic but reviews its predictions each year to highlight its forecasting victories and defeats.
Although neither almanac will share the details of its formula, both are open about what goes into their predictions.
The Farmers’ Almanac doesn’t identify its weather forecaster, whom it has dubbed “Caleb Weatherbee”; the current Weatherbee has served for about 30 years. The formula Weatherbee uses is based primarily on the solar activity cycle and the motion of the moon, as well as tidal action and certain winds in the stratosphere over the equator, Duncan says. “Right now he’s primarily looking at the motion of the moon and the lunar cycles and how they lined up with weather patterns from years ago,” she adds.
Scientists say that solar activity and lunar motion are poor weather indicators. “Those things all have tiny, very small effects on regular weather variations,” says Adam Scaife, head of long-range forecasting at the Met Office, the U.K.’s national meteorological service.
This year’s prediction—as anyone’s would be—is highly influenced by the current El Niño, a climate state that occurs when the tropical waters of the eastern Pacific are warmer than usual, with consequences felt around the globe. Duncan says Weatherbee does incorporate insight from the U.S. federal government’s predictors. “The newest advance would be that we will look at outlooks by NOAA and the Climatic Prediction Center just to make tweaks,” she says.
The forecast is ready about a year in advance, and compiling it doesn’t involve computers. “It’s quite time-consuming and cumbersome,” Duncan says of the formula. “We use the computer to type it all out, but it pretty much is still a formula that takes a lot of human intervention.”
The Old Farmer’s Almanac has a somewhat more modern approach, although it still advertises its long centuries in the weather prediction game. Rather than working with an in-house forecaster, it contracts out to AccuWeather for its long-range predictions, Connare says. And although this almanac doesn’t cite the moon’s influence on weather, it does rely on solar activity, as well as climatology and meteorology. The Old Farmer’s Almanac is particularly interested in long-distance and long-term atmospheric patterns called teleconnections. The best-known of these manifests as the El Niño and its opposite, the cooler La Niña.
Like those of its counterpart, the Old Farmer’s Almanac’s forecasts look back through history to see the weather that accompanied upcoming combinations of factors. “That’s really how it’s done, is looking across big data sets for patterns,” Connare says. “Now, of course, it’s all done by computer, but it’s the same idea. It’s pattern recognition.”
These predictions are what scientists call statistical weather forecasts, which rely on previously observed connections between variables. Although statistical forecasts are easy to run, they are limited by the inherent chaos of the weather system. “That prediction is based on some statistical relationship,” Hu says. “But that relationship may not hold for this year.”
As the climate crisis wreaks havoc on our weather, those relationships will become less helpful, says Sarah Kapnick, chief scientist at NOAA. “Climate change has complicated just being able to rely on statistical models alone because the past isn’t going to always be a predictor of the future for us now,” she says. “New types of events are happening that we don’t have records for, and so that is where observations are no longer helpful.”
More modern approaches can take existing conditions and understand how they will evolve according to fundamental science rather than past experience. “These computer models are based on basically five equations you could write on a T-shirt, but they can generate all of the weather from those equations,” Scaife says of his office’s forecast. “All of that stuff emerges spontaneously from those five equations if you run them forward in a computer.”
These forecasts improve with a better understanding of the physical processes underlying the weather. They also do so with better resolution and more individual runs of the model, which make the predictions expensive but powerful.
Overall it’s a golden age for weather prediction, Kapnick says. “I don’t think people are really aware of how much the skill has improved.”
Even the most accurate forecast using the strongest prediction technique has little influence, however, if people don’t read it or think it’s just as likely to be wrong as any other weather outlook they can find. “The science and the forecasts and the predictions only have value if they’re used,” Kapnick says.