There is potential for snow in the Washington area late Wednesday night into early Thursday morning even though less than week ago I posted that the current pattern was not a good one for snow. Making predictions certainly is humbling. Sometimes it can snow when the overall pattern isn’t favorable. This looks to be one of those cases perhaps.

Like most of our threats, this one still is cloaked in uncertainty as the models differ on how quickly the cold air shifts into the area and on how much precipitation will accompany the system. Small changes in the storm’s track can still make differences in how quickly the cold air gets into the area and how much precipitation we get.

What the computer models are saying, and what to make of them...

The latest GFS and NAM models both forecast rain Wednesday night but then indicate that it will probably transition to snow sometime between 1 and 4 a.m. Thursday.

The biggest question mark about accumulations will be temperatures at the surface as it may well be too warm for snow to stick.

The European model shows heavy precipitation (between 0.75-1” liquid equivalent) between 7 p.m. Wednesday night and 1 a.m. Thursday morning. Temperatures at 5,000 feet are cold enough for some of this to fall as snow but surface temperatures are above freezing except in the mountains, reducing the likelihood of significant accumulation. (

This morning’s European model came in stronger and wetter than last night and now looks more like the NAM albeit a little warmer at the surface. The greatest potential for accumulating snow will be areas north and west of the city especially in areas where there is elevation.

Tomorrow, we’ll try to provide a little more detail about where accumulations might be expected. For now, the only things that can be said about the Wednesday night and Thursday is there is potential to see snow but this storm probably will be a minor one. It is still too early to make any definitive calls about amounts especially for areas north and west of the city.

Technical model analysis

The biggest differences in the models are not so much in the storm’s forecast track. All the the operational models take a weak surface low to our south bringing a slug light to moderate precipitation into the area. The more significant differences are in how quickly the cold air gets into the area and how much precipitation will fall once it is cold enough to snow.

The maps below illustrate the differences in how the GFS and NAM handle the cold air and changeover to snow across the area. This morning European model was more similar to the NAM than the GFS in terms of precipitation.

This morning’s GFS (left) is slower pulling the cold air eastward and also predicts less precipitation than the NAM (right). The latter predicts as much as .38” of the precipitation will fall in the form of snow or around 3 or 4 inches provided the surface temperatures fall below freezing.

GFS model shown on left and NAM model on right for 4 a.m. Thursday morning. The NAM simulates slightly heavier precipitation and the snow a bit farther east (as it’s colder).

The probability of temperatures falling below freezing during the storm are less than 20% according to the SREF model above, making accumulation a dubious prospect. (NOAA)

Finally, you can get a sense for the difficulty in forecasting the timing of the changeover and any dip in temperatures towards freezing (that might allow accumulations) by looking at the time cross section of temperatures from the 1200 UTC (7 a.m.) run of the NAM (to the right below).

NAM model cross section of temperatures and how they cool with time. Tightly packed green lines where arrow is pointing indicate temperatures are changing quickly. Slight shift in how the timing can significantly affect the forecast.

In summary, it looks like rain to start, probably changing to snow early Thursday morning. If it pans out, this will probably be a minor snow event in D.C. but, in D.C.,even minor events are worth watching.