Skip to main content
ABC News
D.C., The Snow Is Coming For You

During the record-setting warm days of a month ago, I didn’t hear a whole lot of people talking about snow. Now, there is a blizzard watch for the Washington, D.C., metropolitan area, and words like “historic” are being thrown around. How the heck is that possible given the warmth last month, and how much of this is hype?

There may be some hype involved, but this storm is very real and could be very powerful. It may deliver record amounts of snow to the Washington area, though there is still some doubt about how it will affect areas farther north.

A principal reason for the widely diverging weather patterns in the last few weeks is a teleconnection1 called the North Atlantic Oscillation (NAO). As we discussed last month, the NAO was raging positive, which meant low pressure over Greenland and high temperatures in the East. Now, it’s negative (meaning high pressure in the same location), which often gives us colder air. Additionally, it’s forecast to go positive again, and rising NAOs are often associated with big-time snowstorms in the East.positive Pacific North American teleconnection compared with a negative one last month. That means a ridge in the West, which is often a key factor in Mid-Atlantic snowstorms.


At the same time, we’re in an El Niño pattern. This means a very active southern jet stream that allows a lot of moisture to work its way up the East Coast. When you combine the colder air and more moisture, it allows for a lot of snow. It’s really a matter of timing, as I pointed out last month. As my good friend Ian Livingston at the Washington Post’s Capital Weather Gang has written, many of the largest snowstorms in the Mid-Atlantic have occurred during El Niño periods, including three 2009-10 snowstorms that were among Washington’s top 25 on record.

Of course, you may have the pattern to make snow and not actually get it. In this case, though, we’ll have a low-pressure system developing over the southeastern United States by Thursday night with cold air in place. This low has the potential to produce snow from western North Carolina through Virginia and West Virginia and on up to an area that includes Washington, Philadelphia and New York City.

The amount of moisture forecast in almost all the models does suggest the potential for historic snow totals. For example, the Global Forecast System spits out 2.8 inches of liquid precipitation, the Global Deterministic Prediction System (also known as the Canadian model) has more than 2 inches and the European model has right around 2 inches for Washington. Usually, we convert liquid to snow on a 1:10 ratio, so 2 inches of liquid equals 20 inches of snow. Moreover, the ensembles (models re-run with slightly different initial conditions because we can’t measure initial conditions perfectly) are generally agreeing with the operational model runs.

Indeed, there is not a lot of doubt that Washington is going to get slammed by a foot or more of snow starting late in the day on Friday and lasting through Saturday night. When you combine that snow with strong winds, you see the reason for the hype and the blizzard watch. Still, it should be noted that things can change, which is why the Capital Weather Gang still has D.C. with a 25 percent chance of receiving 8 inches or less.

Unfortunately for forecasters, the predictions become a lot more difficult up the coast. It looks at this point as though there’s going to be a sharp cutoff of precipitation somewhere around New York City, but the forecast is especially difficult for lower-resolution3 ensembles to nail down.

That’s visible in the latest run of the Canadian model (below),4 which has about 40 millimeters (1.6 inches) of liquid precipitation falling as snow in the immediate New York City metropolitan area (where snow isn’t forecast to start until Saturday) and very little in the northern two-thirds of Connecticut.


It wouldn’t take much at all for that cutoff to shift a little farther north or a little farther south. The latter possibility would spare New York City from much of the snow in the Canadian model.

There are a couple of reasons for that, including a confluence zone in northern New England and questions about where exactly the low-pressure system responsible for this storm exits the coast. Slight changes in these factors can make a big difference to New York City and the surrounding area, and they are difficult to forecast.

For now, we can say only that a lot of people (probably including the politicians of Washington) are going to get hit hard by this storm.


  1. A teleconnection is essentially a situation where a weather pattern in one place can greatly influence the weather in another.

  2. We also now have a positive Pacific North American teleconnection compared with a negative one last month. That means a ridge in the West, which is often a key factor in Mid-Atlantic snowstorms.

  3. This essentially means the ensembles forecasting at fewer specific points than the operational models we speak about here. This allows the models to produce more information faster but also means they sometimes miss some of the finer points of the forecast.

  4. A number of other models also have or hint at the possibility of such a cutoff.

Harry Enten was a senior political writer and analyst for FiveThirtyEight.