Uncertain weather is for certain: Our view

For New England and eastern Long Island, this week's blizzard represented a forecasting triumph. The National Weather Service predicted a crippling storm, and nature delivered. Parts of Massachusetts experienced hurricane-force winds, coastal flooding and record snowfall exceeding 30 inches.

For New York City, however, the story was different. A National Weather Service discussion issued late Monday afternoon warned that the storm "has the potential to produce the largest single-storm snowfall in NYC history," breaking the 2006 record of 26.9 inches, and "should be a raging blizzard" by 6 a.m. Tuesday.

Oops. New Yorkers awoke Tuesday to a moderate snowstorm, hardly a historic one. Central Park ended up with 9.8 inches of snow.

What happened?

The storm intensified farther east than forecasters, relying on their preferred computer models, predicted it would. That left New York City on the western edge of the system that clobbered New England.

Weather service forecasters in New York and Philadelphia apologized for the overdone predictions — a refreshing acceptance of accountability in an era that too often lacks it, but the mea culpas were somewhat beside the point.

The weather service's real mistake was a failure to communicate adequately, to the public and political decision-makers, the uncertainties associated with the forecast. That prompted news media outlets to hype the worst-case scenario, New York Mayor Bill de Blasio to warn "this could be the biggest snowstorm in the history of the city," and New York Gov. Andrew Cuomo to announce the first snow-related shutdown of the city's subway system in 110 years.

It's hard to blame political leaders for erring on the side of caution, but "cry wolf" episodes such as this have the unfortunate consequence of undermining public confidence in official warnings. That lesson reverberates well beyond the Big Apple.

As the storm came together Monday, TV meteorologists such as Chad Myers on CNN and Carl Parker on the Weather Channel showed that there are more nuanced ways to convey information. They gave viewers a heads-up about computer models suggesting that New York City would escape the worst of the storm. (Ironically, weather service forecasters put more faith in the usually reliable European model than in recently upgraded U.S. models that did a better job with the endgame of this storm.)

The public seems capable of understanding that, despite remarkable advances, forecasting remains an inexact science.

The National Hurricane Center uses familiar "cones of uncertainty" to show the potential tracks of tropical systems. At The Washington Post's Capital Weather Gang blog, forecasts come with informative confidence levels (low-medium-high) and "boom" and "bust" scenarios.

The National Weather Service has been slow to adopt these sorts of techniques more broadly. Louis Uccellini, the weather service director and an authority on East Coast snowstorms, says "we all need to improve how we communicate forecast uncertainty."

With another significant winter storm set to move across the USA from the southwest to the northeast, this weekend would be a good time to start trying some new approaches.

USA TODAY's editorial opinions are decided by its Editorial Board, separate from the news staff. Most editorials are coupled with an opposing view — a unique USA TODAY feature.

To read more editorials, go to the Opinion front page or sign up for the daily Opinion e-mail newsletter.