Model 'behaviour'? Big Apple snow bust blamed on weather model

In the ongoing weather model "battle" between the Americans and the Europeans, we may have come out on top for this week's blizzard — but meteorologists still went across the pond to get the forecast for New York City and environs.

New York City, much of New Jersey and Philadelphia got much less snow Tuesday than the National Weather Service had predicted. New York City officially received 7.8 inches at Central Park, quite a bit less than the 24 inches to 36 inches that the weather service forecast.

"The GFS (American) model did better on this storm than the European model," said National Weather Service Director Louis Uccellini, based on his early reading of e-mail traffic among meteorologists. The GFS (Global Forecast System) model is the main American model, which was recently upgraded.

The European model said the storm would be about 100 miles farther west than it actually was, which unfortunately is what the weather service went with in their prediction for New York City. The official name for the European model is the European Centre for Medium-range Weather Forecasts (ECMWF).

A few other models, including the GFS, the North American Model (NAM) and others said the worst of the snow would miss the city.

Predicting 30 inches for New York City, based on that one model, was "outlandish," tweeted private meteorologist Ryan Maue of Weather Bell Analytics.

"When one model is going gangubsters for NYC, but GFS, NAM ... and others suggest much less, don't go with worst-case," Weather Channel meteorologist Michael Palmer tweeted.

It should be noted that ALL of the models — including the ECMWF — were calling for a huge blizzard in New England, and that certainly verified.

The Weather Channel was one of the few media outlets late Monday that predicted a smaller total for the Big Apple: Roughly a foot. "Bottom line: this was a historic snowstorm and blizzard for parts of New England, as the Weather Channel predicted spot on. Never buy into just one model," tweeted Palmer.

One of the European model's shining moments was in 2012, when it correctly predicted the path of Hurricane Sandy well before the GFS came on board.

Uccellini, in a conference call Tuesday afternoon, said meteorologists need to improve how they present forecast uncertainty: "We need to make the uncertainties clear," he admitted.

The preparations in New York City from decision makers — which were based on weather service predictions for the storm — were "the right decision," Uccellini said, noting there were far fewer storm-related deaths, traffic accidents and traffic jams than in previous snowstorms. "It's best to plan for the worst and hope for the best," he said.

"People in areas where the snowfall was less than forecast woke up this morning to find the streets and sidewalks plowed, and their cities reopening; in effect, yesterday's decisions helped these cities recover faster than they otherwise would have," he said.