The Washington PostDemocracy Dies in Darkness

Why was the Blizzard of 2016 so easy to predict? We can trace it to the tropics.

The ingredients coming together to make the storm one for the record books. (NOAA)
Placeholder while article actions load

It’s almost never the case.

Five days before the first flakes fell, forecasters warned that a big winter storm was brewing. “Potentially historic,” they said, all in agreement. And instead of disregarding the predictions like they sometimes do, people prepared, because meteorologists said they were more confident in this forecast than they had been in years.

Winter storms are among the most challenging things to forecast, not only because they are uncertain by nature, but because they can have an enormous impact on a large population. They are arguably the hardest to predict along the Northeast corridor, where the slightest variation in track and intensity can lead to dramatically different snowfall outcomes. Rapid elevation changes — from sea level to Appalachians in fewer than 200 miles — add even more uncertainty to already-complex storms.

It is along this corridor that forecast models are almost always in disagreement on timing, intensity and potential snow accumulation. But they were not for this storm, for reasons set in motion thousands of miles away more than a month ago.

[The forecast for Snowzilla was stellar in almost every way]

Interestingly, there was an overpowering connection between the Blizzard of 2016 and the weather in the tropical Pacific Ocean, says Michael Ventrice, an operations scientist at Weather Service International, the data and analytics arm of the Weather Co. As you can probably guess, El Niño played an important role. However, another lesser-known mechanism called the Madden-Julian Oscillation, or MJO, may have been the straw that broke the camel’s back.

For the vast majority of 2015, El Niño’s warmest waters, its center of gravity, was parked in the eastern Pacific Ocean. This is a classic El Niño setup — it kept all of El Niño’s strongest thunderstorms on the eastern side of the equatorial Pacific. But in mid-December, the center shifted west toward the dateline, and its strongest thunderstorm activity ended up around a longitude of 165 degrees west.

Around the same time, the MJO was trekking eastward to that same location. The MJO is a weather pattern that crosses the Indian and Pacific oceans from west to east over and over again. It has two phases: convective and suppressed; stormy and dry.

[This snowstorm was easy to predict. Here’s why that’s not normal.]

When the stormy phase of the MJO finally reached the region of El Niño’s warmest waters, it sparked “an explosion of thunderstorm activity near the dateline,” said Ventrice, that had a huge effect far into the North Pacific. It caused the jet stream to become unstable, developing large ridges and troughs which pushed warm air far north to the pole and set up the potential for cold air intrusions in the mid-latitudes.

Once the explosion of tropical thunderstorms had its effect on the jet stream, meteorologists knew the weather pattern was changing. Both the North Atlantic oscillation and the Arctic oscillation went strongly negative in mid-January, which signals the potential for big winter storms in the Northeast.

The pattern began to weaken just before the blizzard hit, says the Capital Weather Gang’s Matt Rogers, which meant we were nearly out of the woods for a big East Coast storm. “Both [oscillations] hit very strong intensities two weeks ago, and then started to break down,” Rogers said. “This storm was the ‘last chance’ in that pattern-type to do something.”

Eight days before the storm, the ingredients finally came together. Hints of a major coastal storm were showing up in the models. Five days before the storm hit, forecast models were converging on the idea that the storm was going to be a big one for the Washington, D.C., region. Four days out, the most likely forecast was “a crippling snowstorm.”

[The varying predictability of snowstorms]

Meteorologists were captivated by how confident the models were so far ahead of the blizzard. “The forecast models are in unusually close agreement for a storm that is more than four days in the future,” the Capital Weather Gang’s Wes Junker wrote Monday.

“This is rare, but tends to happen with the biggest storms,” Rogers said. “My simplified theory is that the bigger they are, the easier the models can find them and model them correctly.”

As for the tropical “bomb” that triggered it all, “I think the MJO hitting the warm El Niño waters was the powder keg,” said Rogers. “We have never seen an MJO this strong with an El Niño this strong.”

Ventrice says that forecast models don’t handle the tropics well, or maybe we would have seen the pattern shifting sooner. “Oh, the tropical-extratropical climate interface is so cool, and it’s so not well studied right now,” Ventrice said. “The area of research is really just in its infant stage in academia and the private sector.”

What the D.C. area looks like after the epic blizzard

Mike Mazza (R) and his son Gabriel (L) stand outside of their subdivision attempting to get plow service for snowy streets in Gaithersburg, Maryland January 26, 2016. Washington will need several more days to return to normal after a weekend blizzard dropped more than 2 feet (60 cm) of snow along the U.S. East Coast, likely causing billions of dollars in damage and killing more than 30 people. REUTERS/Gary Cameron (Gary Cameron/Reuters)