Weather

‘Nothing like a good graphic’: How the snow map became a fixture of winter weather forecasting

Evolving from magnets and markers to color-coordinated computer graphics, snowfall prediction maps have grown in both sophistication and importance.

By Hayden Bird

February 28, 2023

• Facebook
• Twitter
• Email
• Email

In the days leading up to a potential snowstorm, Bostonians pursue varying levels of preparation.

Whether it’s making sure that they have enough food in the house — “bread and milk” — or possibly readying their legally suspect “space savers” for action, the local population has become accustomed to periodic snow events (even if that has been conspicuously absent for much of the current winter).

But embedded in most residents’ pre-storm rituals, hidden in the plain sight of its seemingly utilitarian nature, is one of the most popular winter storm habits in existence: Checking the snow map.

The snowfall prediction map, or simply “snow map” as it’s popularly known, has many versions. Some are shaded yellow and orange, others more blue and purple. However they look, it’s incontrovertible that they have become a ubiquitous presence in winter forecasting over the last 30 years. And as is true on most news or weather outlets (including Boston.com), anything offering a snow map becomes heavily trafficked as readers scramble to get up to speed on the impending storm.

On the surface, explaining the widespread popularity of the snow map is obvious. Its appeal is almost an afterthought to actually describe: It offers viewers a localized glimpse at what they can expect. Beyond the specific context offered by a meteorologist’s full presentation, a snow map conveys the most distilled version of the winter storm forecast.

Yet it’s become such a breakthrough tool of meteorological communication that recalling an earlier time — before intricately color-coordinated snow maps informed the masses on a regular basis — is almost difficult to imagine.

The history of the subject, not unlike a snow map itself, traces the progression of something that was trivial and almost imperceptible in its origin, but has progressed and proliferated as it moved across time and space.

Magnets, markers, and the “Blizzard of ’78”

The usage of maps in weather forecasting goes back more than a century, as the ability to plot the progression of storms’ movements over time became vital knowledge.

“As a medium for establishing facts in the important department of meteorology, the world has never produced, nor even approximated, anything that would compare with the weather map,” noted science writer Isaac P. Noyes in his 1892 study of meteorology.

But prior to the advent of satellites and radar, tracking storms using maps was hardly an exact science.

“When you look at some of the higher impact weather events in the last 150 years, there’s almost a universal theme: really poor forecasts,” WBZ-TV chief meteorologist Eric Fisher explained.

Fisher, who studied some of the region’s classic storms in his 2021 book, “Mighty Storms of New England,” noted that there was a simple outcome from this.

“A generational weather event could come out of nowhere with a very poor forecast.”

The poor forecasting was hardly a reflection of meteorologists’ ability, as Fisher pointed out. It was largely the result of limited information. But even as technology evolved, there was still an occasional communication gap.

Few events illustrated this better than the infamous “Blizzard of ‘78,” when a record storm caught local residents off-guard that February and buried Southern New England under several feet of snow. The full explanation for why the storm was such a surprise requires historical context, as Fisher noted in his book.

“There had been a couple other warnings of snowstorms that season that had not lived up to their billing,” Fisher wrote. “Seeds of doubt were put into the public’s mind and recency bias can be extremely powerful. Though the Blizzard of ‘78 would feature much better forecasts, the perception was that it was overblown and would not live up to the hype. In the end, this is probably what led to the disaster that unfolded on roadways as snowflakes quickly piled up.”

“When I was dealing with the ‘Blizzard or ‘78,’ there was a hint that something would probably happen,” recalled longtime WCVB meteorologist Harvey Leonard, “but I had to make huge leaps of faith to actually forecast what I was forecasting.”

Despite Leonard (and other local forecasters) correctly predicting a major snow event, the message simply didn’t reach enough people. Looking back on Leonard’s television display, it’s easier to understand why a forecast might not have been properly heeded.

“When I first started on television, which was the early ‘70s in Providence, Rhode Island at WPRI Channel 12, we had magnetic stick-ons,” Leonard remembered of early weather maps.

“By the time I got up to Boston in 1977 at Channel 7 at that time, we still had the weather boards and what were like magic markers. So we would draw the weather maps.

“On the local map, you might be able to draw, say, a line from Boston to Providence, and maybe you’d say four to eight inches [of snow] to the left of the line, two to four inches to the right of the line,” Leonard explained. “But you weren’t coloring anything in or shading anything. There was no accounting for elevation differences. It would just be a line.”

Yet as it turned out, the late ‘70s would mark the end point of an era. The coming revolution in computer technology ensured that the time of hand-drawn maps and sunny magnets would soon pass into history.

“There was a kind of professionalization of TV weather.”

The real birthplace of the modern snow map can be traced back to the early ‘80s, when computer technology finally caught up with the public’s insatiable interest in specialized weather forecasting.

“Very quickly from the late ‘70s to the early ‘80s is when the transition happened,” said meteorologist and journalist Bob Henson. Author of “Weather on the Air: A History of Broadcast Meteorology,” Henson has studied this fascinating period in the development of forecasting.

“I know there were a couple stations that used Apple II computers in 1979. That was like this revolutionary advance, but really it was when some companies started helping TV stations develop graphics programs in the early to mid ‘80s,” said Henson. “That’s when it became much more feasible for a TV station to have, for example, a map with color shading showing snowfall accumulations.”

Locally, an Andover-based company, Weather Services International, became an early developer of graphics. Now known as “The Weather Company” (owned by IBM), it was bolstered initially by the creation of something that would have a profound impact on TV weather forecasting.

“The Weather Channel, I mean, they pioneered 24-7 weather coverage,” said Mike Convey, a product leader at The Weather Company.

When The Weather Channel first launched in 1982, it helped standardize many features that modern viewers have come to expect.

“They’ve been a user of our products since they went on air the first day,” Convey noted. “There were the data forecasting products, but the visualization products were important too. So we had a product called ‘Weather Director,’ which basically created snapshots of things. And then we created a product called ‘Show Effects’ that then allowed you to put things on a globe and be able to animate around the globe so you could show the different models that we were talking about forecasting.”

Suddenly, within less than a decade, meteorologists went from having to personally draw potentially imprecise lines on a board to being able to show satellite images as they moved across a computer map. It was a quantum leap in communication and forecasting capability.

“Keep in mind that things are happening side by side,” Leonard explained of the era. “Both the graphical presentation is coming along and advancing, but the computer models are also getting better over time.”

“Once you introduce computer graphics on a green screen, now you could show as many different things as you can creatively produce in the time you’re allowed on the air,” he added. “That’s when we started to do the snowfall accumulations with computer graphics.”

The newfound technological prowess not only improved forecasts, but helped build credibility in the eyes of the public.

“I would argue that it was the advent of computer graphics that helped advance the local weather-caster as a scientist, and as a professional,” said Henson. “There was a kind of professionalization of TV weather, especially in the ‘80s and ‘90s.”

“There isn’t one ‘best’ color bar, there are just a lot of opinions.”

“Nothing like a good graphic,” said Greg Carbin, succinctly summarizing the appeal of the snow map.

Carbin, the chief of forecast operations at National Weather Service’s Weather Prediction Center, marveled at the evolution in his industry since the early ’80s.

“It really is mind-boggling to think how far we’ve come in trying to understand winter storms over the last 30 years or so,” he said.

And beyond the improvement in forecasting, the snow maps themselves have grown in complexity and nuance.

Even seemingly mundane details are still hotly discussed among both local and national forecasters.

For something that is supposedly just a utilitarian device, the aesthetics of the snow map seem to matter a lot to meteorologists. Carbin gave an example: What color bar should be used to differentiate lines of snowfall on the map?

“There’s a lot of discussion about what makes the best display as far as your color selection, and how you show this information,” said Carbin. “I don’t think that you’re ever gonna get one answer, since everybody perceives this information slightly differently. But there’s a lot of talking out there in consultation about how best to display it.”

Recalling both The Weather Channel and National Weather Service color choices for the snow map forecasting, Carbin noted that he’s a graphics “hobbyist.”

“I’ve been been dabbling in maps for a long time,” he said. “I’ll change color bars once in a while to see, well, does this look any better?”

He knows it’s a debate that will probably endure as long as snow maps exist.

“There isn’t one ‘best’ color bar, there are just a lot of opinions,” Carbin said, chuckling.

One important point that Carbin and other meteorologists keep in mind — beyond the more subjective choices of simply trying to make the map look visually appealing — are viewers who are colorblind.

“I mean you don’t want red and green touching each other,” Convey said of the graphic design. “Somebody who’s colorblind will most likely have problems with red and green. There are other types of colorblindness, but that’s the most common.”

Ultimately, the most important factor meteorologists consider when making a snow map is that it’s accurate and precise. One of the newest iterations now even focuses on “probabilistic” predictions as opposed to the “deterministic” forecasts many are used to, offering one more option for those looking to best interpret varying possibilities.

The latest technology is also able to account for localized detail in a way that would’ve seemed impossible not that long ago.

“The way the computer model runs is it basically takes the world and chops it up into grids,” Fisher outlined. “Sometimes those grids are squares, sometimes they’re hexagons. And it basically takes every one of those little shapes and it says this is a point, and it fills that point with qualities of the atmosphere, and then it runs out over time.

“As we’ve gone through the last couple decades, those grids have gotten smaller and smaller and the shapes have changed to optimize how they run,” he added. “So we’re now at a point where we can have very tiny grid points filled with more information. The more information we have about the atmosphere to start, the better the forecast should be going forward.”

No matter how advanced the technology gets, however, predicting snowfall remains an inexact science. Winter weather, in particular, presents major challenges for meteorologists.

“Let’s say the rain-snow line, as best as you can tell in a storm, winds up five or 10 miles different,” Leonard said. “That’s not a large error meteorologically, but think about that happening right along I-95. If it’s a 10-mile difference from where the rain-snow line is now, maybe a million people who thought they were gonna get a foot of snow get an inch of rain, or the other way around.”

But as they continue their task of trying to accurately predict winter in a notoriously unpredictable region, forecasters will keep using the device that has proven so successful in communicating complicated weather patterns. The snow map is here to stay.

As for explaining its intrinsic appeal, Noyes — writing in his 1892 study — offered what might still be the simplest and yet most satisfying description.

“The weather map lifts our minds above the earth and shows us the mysterious factors that are moving over its surface, from the west towards the east, ever towards the rising sun.”