A winter storm could develop Monday night and Tuesday in the Washington region. But a lot of ingredients must come together just right for significant snow to materialize.

The storm setup is notoriously difficult for predicting snowfall amounts and, in the past, has often disappointed snow lovers.

A disturbance diving south from Canada into the Mid-Atlantic will attempt to merge with an area of low pressure off the Mid-Atlantic coast to generate a storm. Where this merger occurs, and how rapidly, will determine whether the Washington region sees heavy snow, light snow, a mix of rain and snow, or very little precipitation.

We see three most likely scenarios:

1) Major storm (35 percent chance): The merger between a Canada disturbance and the coastal disturbance occurs close to the North Carolina coast, and a powerful storm develops. It draws heavy snow into the D.C. area — with a slight chance it pulls in enough warm air that precipitation could mix with sleet and rain along and east of Interstate 95.


GFS model simulation of storm Monday night and Tuesday.

Thursday’s GFS and Canadian model predict this scenario, and the potential for major snow accumulation in Washington.

2) Minor storm (35 percent chance): The merger occurs a little farther north and/or offshore than in the first scenario — closer to the Delmarva Coast, so that the heaviest precipitation focuses north of the D.C. metro region and turns into a big snowstorm from Philadelphia to Boston. The D.C. area could get a period of accumulating snow in this scenario, but the possibility of a dry slot in our vicinity would limit overall snowfall potential.


European model simulation of precipitation at 8 a.m. Tuesday. (WeatherBell.com)

Wednesday night’s European model hinted at such a scenario.

3) Storm out to sea (30 percent chance): The merger occurs so far off the coast that the D.C. area might just get a few snow showers from the Canada disturbance with little or no accumulation.


Last night’s Canadian model simulation of precipitation at 2 p.m. Tuesday.

Wednesday night’s Canadian model forecast presented a scenario like this.

It’s usually wise to be skeptical about any individual model forecast days in advance in these types of systems as there are more things that can go wrong to mess with the storm’s potential than can go right. That said, we cannot dismiss the possibility of a significant winter storm for this area given the overall pattern.

Large groups of simulations from both the GFS and European ensembles indicate about a 60-70 percent chance of at least an inch of snow and a non-trivial 20-30 percent chance of at least six inches.


European model ensemble likelihood of at least one inch of snow Monday night to Tuesday night. (WeatherBell.com)

We should gain greater confidence in the forecast as time wears on, and the forecast bears watching.

Model discussion

Winter storms with two merging streams of energy that require almost perfect timing for significant snow to fall in Washington are often the most difficult to predict.

Wednesday night’s models illustrate how sensitive small changes in the timing of the merging streams can impact the forecast.  They offer clues as to how the storm could lead to significant snowfall and also how it could mostly miss the region.

A comparison of Wednesday night’s European model and the overnight run of the GFS reveal two important differences in the weather pattern evolution that could have major implications for the forecast.

In the image below, note how the green shading in the European model fails to dip nearly as far south as it does on the GFS. This reveals that the GFS model forecasts a much sharper dip in the jet stream.


Comparison of the jet stream configurations predicted Tuesday morning from the European model (top) and GFS model (bottom from Wednesday night simulations.)

If the European model predicted setup turns out to be right and the jet stream is flatter, it is exactly the kind of setup that would probably lead to a weaker storm or the storm missing us entirely. In this scenario, one area of low pressure tracks to West Virginia and weakens while a second area of low pressure forms too far off the coast to give us much snow. In such a storm evolution, the D.C. area frequently ends up in a dry slot in which we only see light snow or miss out altogether.

But the overnight GFS run (as well as Thursday’s), with its stronger, sharper jet stream, ends up with the secondary low closer to the coast and farther south, and would offer us a significant snowstorm.

Wednesday night’s Canadian model shows another way for us to miss the snowstorm (image below).


Canadian model forecast jet stream forecast for Tuesday morning.

It has a strong initial surface low driving into West Virginia that pushes the frontal zone near the coast farther offshore so any secondary low forms too far east to give us any accumulating snow. The green area that helps depict the jet stream is almost as far south as the GFS but is much rounder and makes it a little less dynamic. The Canadian model eventually does develop a strong low over the ocean but too far east to supply snow to our region.

However, note that Thursday’s Canadian model backs off from this scenario and actually forecasts a crushing snowstorm for the D.C. area.

As a good synthesis of the uncertainty and range of possible outcomes Monday night and Tuesday, let’s look at precipitation simulations from the GFS model ensemble (known as the Global Ensemble Forecast System or GEFS):


GEFS model showing precipitation possibilities Monday night into Tuesday.

Each line represents a model forecast with only slight tweaks to the data going into them. Blue lines indicate snow and green represents rain. When the lines are flat, no precipitation is falling. When they are rising, that indicates it is snow or raining. Some of the simulations display a major snowstorm while others only show a little light snow. There are even a couple of simulations that show rain because the primary low is strong and far enough north to draw above freezing air into the region.

The bottom line is this is a very difficult forecast. Phasing streams of flow can produce major snowstorms but also can cause lots of heartburn for forecasters and snow levers. There are more ways for the storm to not quite make it than there are to achieve a snow lover’s dream storm.