FARGO — It happens every year: The warm days of summer and fall end and a crispness makes its way into the air. Winter is here.

At the local grocery store, you might spot a familiar yellow book with big red letters, "Winter weather forecast," sparking your curiosity.

Legend has it in 1815, The Old Farmer’s Almanac's founding editor, Robert B. Thomas, was interrupted by a boy wondering what to include for the weather forecast for July 1816. A distracted Thomas answered, and the entry for July 1816 was supposedly “rain, sleet and snow.”

As wild as a prediction of July snow might sound, the forecast came true — albeit with a little help from the eruption of Mount Tambora in Indonesia blocking the sun and sending temperatures plummeting throughout Europe and North America, killing crops and creating snow and frost through much of June, July and August. That wild weather helped secure the reputation of the almanac.

Fast forward to present day, The Old Farmer’s Almanac, now in its 277th edition, still includes a prediction for a dreaded season — winter — though how anyone comes up with long-term meteorological outlooks months in advance has changed quite a bit since that weird July 1816.

Early start

Forecasting for this winter began long ago in 2017. The Old Farmer’s Almanac predicts the weather for 18 U.S. regions no less than 18 months in advance for each calendar year to be able to produce an edition of the book each fall.

In the Midwest, the almanac predicts above-average temperatures to rule over the winter, with precipitation being below-normal.

In layman’s terms, the winter season for the Midwest, which includes all or part of Michigan, Minnesota, North Dakota, South Dakota and Wisconsin, will be slightly milder and drier than normal, with snowfall near to below-normal, according to the almanac.

But what is shaping the weather? The answer lies in the sun.

The almanac states its weather forecasts are derived from a secret formula devised by its founder, who believed that weather on Earth was influenced by magnetic storms on the surface of the sun known as sunspots. Using a combination of climatology, meteorology and solar science, The Old Farmer’s Almanac predicts weather trends and events by comparing solar patterns and historical weather conditions with current solar activity.

Currently, the secret formula is locked away in a black tin box at the almanac offices in Dublin, N.H., and the almanac claims to be 80 percent accurate with its forecasting — even with the 18-month lead time. Meteorologists aren’t buying it.

'Fun reading'

In order to gain the most accurate information, meteorologists constantly monitor weather patterns and radars and interpret the data for their communities

John Wheeler, chief meteorologist for WDAY, says predicting weather for the end of a week is difficult to get 100 percent accurate because weather can change in an instant.

“The forecast itself is a continuum," Wheeler says. "It’s always changing. Because the weather is very complicated and our ability to forecast the weather using computers and other means is never perfect, we are always trying to refine the forecast and get as close as we can be.”

Daily weather forecasts are produced by a combination of computer modeling and humans that interpret the data. When it comes to The Old Farmer’s Almanac, Wheeler says knowing your sources is important.

“Pay attention to what you are reading,” he says. “The Old Farmer’s Almanac has been around for hundreds of years, but there is so much information today. They go by a formula that was made up over 200 years ago. Something that has been around for 200 years, that may look good if you’re just reading it, but why should one believe it?”

The almanac and its competitors keep weather predictions just vague enough to apply to any area of the country, much like a horoscope that can be applied in any situation, he says.

“It’s fun reading,” Wheeler says. “There’s a lot of cultural nature in that magazine and games and fun articles… But it’s not a serious forecast. That’s how they sell a magazine by saying, 'Here’s your winter forecast.'”

Best guess

Seasons in the weather world do not follow the traditional calendar beginnings and ends and instead are broken up into three-month segments, with fall running September through November, winter from December through February, spring from March through May and summer taking place June through August.

Mark Ewens, a retired meteorologist formally trained through the military and National Weather Service, says while forecasting for winter can be done, it isn’t entirely exact.

“What weather forecasters and meteorologists and climate forecasters do is take all the computer models generated and take an average,” he says. “Basically when you see an extended forecast from the National Weather Service that says this winter and has a big red blob over the country, that is essentially the best guess. It’s a scientific guess, but it’s a statistical guess.”

Climatologists break their predictions down into three categories based on the average gained by computer models: above-average, below-average and neutral.

“The computer model bends the forecast in one direction or another — colder, warmer or normal,” Ewens says. “That’s how most professional weather agencies do their forecasting: computer models and a little human interaction.”

Trends also play a role in determining future forecasts. Ewens says the general trend across our region during the winter has been warmer, with a little less snow and a few more days with rain or freezing rain events. Significant weather phenomena — like El Nino, La Nina or polar vortexes — also impact what the weather will hold because oceanic temperatures can create weather patterns even far inland.

“At this stage of the game, my personal take is a normal overall winter pattern with a lot of variations,” Ewens said in the fall before winter started here.

Whether the weather predictions come from a 200-year-old tabloid sold in the checkout line or from scientific evidence and cutting-edge computer modeling, one thing is certain, according to Wheeler.

“If anybody really knows what the weather is going to be three to six months from now, they could sell that information to the energy companies and retire,” he says. “A forecast like that would be worth that much money.”