The odds of a highly unusual, though not unheard of, period of October snow have increased for northern and western parts of the D.C. region on Saturday. Yesterday, Jason outlined what ingredients were needed to get accumulating snow in the Washington area during October . The majority of computer models are now showing more of those ingredients than yesterday.

The models appear to be converging towards the European computer model solution from two days ago with a storm tracking to our south and then rapidly deepening (developing) off the Delmarva coast. However, the models still differ enough on their tracks to provide lots of uncertainly about how much and intensely the precipitation will fall. They also differ on how cold the temperatures will be at the surface and the lowest several thousand feet making it difficult to make a definitive call on whether we will see snowflakes or not in the D.C. area.

Nevertheless, it is becoming increasingly likely that the D.C. area will get a period of at least moderate precipitation. The bigger question is whether the temperatures will be cold enough to support snow.

Inside the beltway and to the east, the majority of the event still is likely to be rain but with the rain possibly mixing with or changing to a little wet snow before ending.

However, there still is potential (albeit a low probability) for some accumulation west and north of the city, especially as you go up in elevation into the mountains. Any significant accumulation this time of year can cause tree damage because so many trees still have the majority of their leaves making them more susceptible to breaking. For that reason alone, this storm is worth monitoring closely over the next two days.


Last night’s European (Euro) computer models simulated snow in the Washington/Baltimore region and it suggests that the snow might fall heavily enough to coat the ground in the colder suburbs if the surface temperatures can fall into the mid-30s. The European model has a near perfect track of the 850 mb low (around 5000 ft) to support snow with the low centered to our southeast at 11:00 a.m. Saturday morning (left side below) allowing cold air to remain in place. The temperatures at that 5,000 ft level are below freezing (left map below), cold enough to support snow along and west of I-95.

Left: Temperatures at 5,000 feet (850 mb) at 11 a.m. Saturday (as simulated by Euro model). Right: Temperatures near the surface at 11 a.m. Saturday (as simulated by Euro model). (

Even though the surface temperatures are forecast to remain above freezing during the storm, the Wunderground snow accumulation maps from the Euro model are showing accumulating snow in the D.C .and Baltimore areas west of the city. But for this to materialize, the intensity of the snowfall will need to be heavy enough to offset the melting as any snow hits the ground. The Euro simulates a fairly dynamic storm capable of accomplishing this, but primarily in the far western and northern suburbs where there is higher terrain.

Supporting the Euro this morning’s GFS has a strong mid-level vorticity (spin) center passing far enough to our south to put us in an area of enhanced precipitation known as the deformation zone (or comma head). The UK Met model also simulates this scenario. Such a setup in late November would have me excited about snow everywhere in the region. However, the GFS’s forecast temperature at 11 a.m. around the peak of the storm is only just at freezing at 5,000 feet (850 mb) near Reagan National Airport (see below left, the blue line is the freezing line) - indicating how marginal temperatures are for supporting snow flakes inside the beltway.

By around 11 a.m would be the time when northern and western suburbs could mix and start changing to snow if the GFS or European models are correct . But near the city and areas east of I95, rain would still by falling.

Left: An image of the surface low off the Delmarva peninsula at 11 a.m. Saturday as simulated by the GFS model. The shaded areas indicate the 60-hour melted precipitation total. Right: NOAA snow algorithm valid at 2 p.m. Saturday

So most of the models suggest a decent chance of seeing flakes in areas well west of the city, with a chance of flakes mixing in at the tail of the storm closer to town. But a few model still simulate mostly rain or very little precipitation whatsoever. The main holdout now is the NAM model which has a more eastward track and lighter precipitation. It has limited support from some of the SREF (Short Range Ensemble Forecast model) simulations . The NAM and those SREF members are two of the factors for remaining cautious about forecasting snow - not to mention how marginal temperatures are for snow in any model (including the GFS and Euro).

In summary, most of us will start with a cold rain Friday night into Saturday morning. The rain could change to snow if the precipitation stays intense enough to allow the surface temperatures to cool into the mid-30s. Any changeover would occur first in the western suburbs, the same areas with the best chances of actually seeing snow whiten the ground. Accumulation - especially on grassy areas and trees - cannot be ruled out in the mountains above 1,000 feet.

Tomorrow, we’ll take a crack at developing some maps indicating what to expect where and will discuss any possible impacts on roads and trees - which - at this point - would appear to be most likely above 1,000 feet in the mountains.