European modeling system forecast of a large area of high pressure over Greenland, known as the Greenland Block, in eight to 10 days. This feature is often present during snowy weather patterns in Washington. (

There are certain things for which weather forecasters look when it comes to favorable pattern for snow in the Mid-Atlantic and Northeast. One that gets snow lovers particularly excited is the “Greenland Block.”

In early March, a particularly strong Greenland Block is predicted to form, increasing the potential for cold, stormy weather in the Mid-Atlantic and Northeast. The block does not guarantee snow, but it does increase the odds, particularly when it shows up in the configuration portrayed by the latest weather models. (Caveat: The models could be wrong.)

During the winter of 2009-2010, Washington’s snowiest on record, a massive Greenland Block developed and persisted, setting the stage for big snow events.

The Greenland Block, which is a strong area of high pressure over the massive island, is a common feature of a pattern known as the negative phase of the North Atlantic Oscillation (NAO). During a negative NAO, the air pressure rises in the high latitudes of the North Atlantic. This causes the jet stream to sink south in eastern North America and cold air to infiltrate the Mid-Atlantic and Northeast. In this pattern, coastal storms are prone to forming in the transition zone between the cold air over land and milder air over the ocean.

Weather models now predict the North Atlantic Oscillation to tank to historically extreme negative levels by around March 1, and forecast maps of the overall weather pattern in early March now feature a huge Greenland Block.

European model forecast for the North Atlantic Oscillation. The big circles outlined in black portray the latest forecast. The other color outlines indicate forecasts from previous dates. (

In the left panel of the image below, which shows the forecast from the European modeling system 10 days from now (around March 3), the Block is portrayed by the orange and red shades over Greenland (indicative of much higher than normal air pressures). The Block persists but weakens some four days later (around March 7), as shown in the right panel.

But it is the middle panel that should pique the curiosity of those who like snow. It shows the average pattern associated with Washington’s 15 snowiest March days. The pattern very closely resembles what the European modeling system is forecasting 10 days (or 240 hours) from now.

One of the past events that featured a very similar pattern to the European model forecast was the great storm of March 18 to 23, 1958. While Washington saw only a few inches, 10 inches or more fell in its northern suburbs and 30 to 50 inches were reported near the Mason-Dixon Line. There were also two more snow events that month that occurred in the same general pattern, giving Washington more than 10 inches of snow in total.

Big snow events have occurred in such a pattern in other years, as well, but there is perhaps a cautionary tale in these past events. Despite the favorable pattern, many storms unloaded most of their snow at higher elevations, while the city received either minor or no accumulation.

The latest patterns projected by European modeling system for early March show some potential for accumulating snow based on analysis of the dates from the past they most resemble, known as analogs. In the figure below, the patterns predicted at six to 10 days (left panel) and 11 to 15 days (right panel) show the most similarity to the 10 past analog dates shown at the bottom right in each panel. After the dates presented in each panel, an average of about two to three inches of snow fell.

The average pattern predicted by the European modeling system 6 to 10, left, and 11 to 15, right, days from present and, at the bottom right of each panel, the dates from the past which most resembled the pattern shown. (

While that doesn’t sound like a lot, it is more than Washington’s March average snowfall of 1.3 inches. In several of the cases, more than five inches fell. At least a little measurable snow fell after 70 percent of the dates shown. So there’s certainly some snow signal present in this pattern.

The pattern unfolding bears characteristics of those associated with Washington’s greatest snowstorms, regardless of month, as shown in the graphic below:

(CWG adaptation of ESRL/NOAA data)

So, with the pattern potentially setting the stage for snowfall in March, what does recent history tell us about our chances?

First, March is very hit-or-miss when it comes to snow. About 5 percent of Marches in D.C.’s history have seen no snow at all. Another 32 percent have seen only flakes and no accumulation. On the flip side, about 6 in 10 have at least some snow accumulation. Many have a little; some have a lot.

The month averages 1.3 inches in the city, and 2.8 inches over at Dulles International Airport. In other words, a range of about one to three inches is a reasonable general expectation.

It is a month known for its volatility as the seasons begin to change. As alluded to above, big March snows are much more common at high elevations. But we have only to look back to 2014 for a snowy March in the city, when more than a foot fell from multiple events even at low-elevation Reagan National Airport.

Since it will be March, snow lovers will need more than usual to bend in their favor. The odds begin to stack up against snow, and sustained springtime is around the corner. We can get the red and blue blobs in generally the right spot, and still come away with not a lot to shovel. Either way, the pattern does suggest it should tend stormy, and we need storms for snow.

Jason Samenow contributed to this report.