Conclusions: Computer models that came in between 9 p.m. and midnight forecast very substantial to historic snowfall amounts for the D.C. area Friday into Saturday night or Sunday morning.
The models suggest snow is most likely to begin between late morning and early afternoon Friday although a late afternoon onset isn’t out of the question.
Snow is likely to be steady and heavy at times by Friday evening, continuing overnight and into Saturday. Models differ on how long the snow will last, ranging from Saturday evening to Sunday morning.
Accumulations shown by all models are very substantial, as high as 30 inches in some cases. We think more plausible totals may be in 12 to 20 inch range, although amounts above two feet cannot be ruled out, especially in northwest Virginia and in the Blue Ridge.
From the District and to the southeast, we cannot rule that snow will mix with sleet and/or rain, reducing snow totals some.
Particularly on Saturday, winds may become strong posing a power outage risk given the heavy snow.
Overview: The ingredients are in place for a heavy, possibly historic, snow event in D.C. Friday into very early Sunday. But will they gel?
Computer models, without exception, are forecasting over 10 inches of snow in the D.C. area with a few throwing around 20 inch plus numbers. We think there is at least a 50 percent chance of a foot of snow in the immediate area.
Lots of questions remain about this storm. When will the storm start? When will the conditions be worst? Will a perceived southward shift in the storm track continue?
Follow along as we share and comment on the latest model information coming in this evening through around midnight. At the end of the evening we will post some updated, concluding thoughts on the storm potential.
12:05 a.m. update: The latest GFS ensemble mean (average of group of GFS model simulations with initial conditions tweaked) simulates a more reasonable, though very high, 18-25 inches for the D.C. area. See map below.
11:47 p.m. update: As promised in the 11:35 p.m. update, here’s the Canadian model snow output, which is on the order of 20-30 inches around the region.
These numbers seem too high to be real, and may well be. But given how we’re repeatedly seeing multiple models spit out these kind of amounts, it is certainly suggestive something extreme may be in the works – even if not of the simulated magnitude.
11:35 p.m. update: The Canadian model is in and is an impressive snow hit for the D.C. area but not as extreme as the GFS. It starts the snow around the same time, between 10 a.m. and 1 p.m Friday. The heaviest snow – and it’s very heavy – falls between about 10 p.m. and 4 a.m. Friday night into Saturday morning.
Snow continues Saturday, in varying intensity, tapering off between 7 and 10 p.m. Saturday night. Its duration is a good 12 hours less than the GFS. I’ll have a snow map from it shortly.
11:26 p.m. update: While I’m still digesting the GFS run, a couple notes. 1) The model run does bring the temperature in DC a little above freezing Saturday, which might lower amounts a hair (due to possibility of melting and/or mixing with sleet). 2) The storm stall and snow duration on the GFS hasn’t been shown by other models. It’s not to say it’s necessarily wrong, but maybe not likely.
11:09 p.m. update: Hold your breath folks and remember this is just one model run, 3 days away from the actual storm. But here is the astonishing GFS model snowfall forecast:
This would rank among the top three storms in D.C. history and utterly cripple the region. It is unlikely reality will be this extreme, but this model run conveys the possibility of a historic event.
11:04 p.m. update: Remarkable. The GFS still simulates snow in D.C. at 7 a.m. Sunday morning, suggesting it will snow for over 40 hours given projected noon start time on Friday.
10:58 p.m. update: The GFS run is a very heavy snow hit for the entire metro region, with excessive snow amounts in NW Virginia. It snows, heavy at times, for a very long duration as the storm stalls off the Mid-Atlantic coast. Maps soon.
10:55 p.m. update: Got plans Friday afternoon and evening? The GFS model shows moderate to heavy snow between 1 and 7 p.m.
10:52 p.m. update: The GFS is suggesting a snow start time between 10 a.m. and 1 p.m. Friday (consistent with the SREF below). Map below is simulation of accumulated precipitation through 1 p.m. Friday.
10:45 p.m. update: While we await the GFS (it’s getting there), I see the National Weather Service has issued a snowfall forecast through 7 p.m. Friday night. It calls for 3-5 inches in the D.C. area by that time. Of course, the bulk of the storm is still to come beyond that time. This may be a 36-hour snow event in parts of the region.
10:33 p.m. update: The GFS is starting to come in. This is an important run, but it will be a few minutes before it’s far enough out into the future to start showing our storm.
10:28 p.m. update: The SREF model brings in the snow much faster than NAM, with an onset time between 10 a.m. and 1 p.m., which is more consistent with other modeling.
10:20 p.m. update: The SREF model, which stands for the Short Range Ensemble Forecast, is now out. Like the NAM, it only goes out through 7 a.m Saturday morning and it simulates a hefty 8-12 inches by then, with potentially plenty more snow to go before the storm winds down. Parts of the Blue Ridge are already into the 15-20 inch range by Saturday morning in this model.
10:10 p.m. update: Below is the NAM model’s snowfall forecast through 7 a.m. Saturday, when its simulation ends. It suggests rough 6-10 inches over the region by then, which might be considered around the storm’s midpoint. I want to stress this model is not very good beyond 48 hours into the future, so take it with a grain of salt.
10:00 p.m. update: The NAM model, taken literally, introduces the possibility of a changeover to sleet and rain along and southeast of I-95. While it simulates extremely heavy precipitation, it tracks low pressure farther west than most other model runs today. Such a track draws in enough mild air to increase temperatures above freezing at high altitudes.
Notice the blue line cutting through the District in the map below, valid at 7 a.m. Saturday morning. That’s the freezing line at 5,000 feet aloft and often a good approximation of the rain-snow line. If the rain-snow line works this far inland, it would cut snow totals near and east of the I-95 corridor.
As a caveat, the NAM model is not particularly good beyond 48 hours and the above simulation is 84 hours into the future. But it’s plausible this storm could take the kind of track it’s suggesting and result in mixed precipitation from the District and southeast.
9:50 p.m. update: The NAM model run is now completely finished. It indicates snow will begin between 4 and 8 p.m. Friday from southwest to northeast in the D.C. area, which is later than most of the other models, except for the European. If it’s right, it would suggest schools could potentially open Friday.
The image below shows accumulated precipitation through 7 p.m. Snow is just moving into the District at that time according to this model.
By 10 p.m., it is snowing everywhere in the metro region and picking up in intensity quickly:
9:42 p.m. update: CWG winter weather expert Wes Junker: “Someone is going to get 30 inches [in the Mid-Atlantic]. I just don’t know where.”
9:38 p.m. update: The NAM model is streaming in as we speak. Its forecast is out through 1 p.m. Friday and snow has not yet started according to its simulation. Might this mean school on time Friday???
9:30 p.m. update: The National Weather Service forecast office in Sterling, Va. just posted the following discussion about the storm potential, with some cautionary notes about the snowfall forecasts and possibility of model shifts:
THE COMBINATION OF A SLOW MOVING STORM SYSTEM...COLD AIR AHEAD OF THE STORM SYSTEM...AND PLENTY OF MOISTURE FROM THE GULF OF MEXICO AND ATLANTIC CONTINUES TO SUGGEST THAT THERE IS A POTENTIAL FOR A MAJOR WINTER STORM TO AFFECT THE AREA... BUT THERE IS STILL HIGHER UNCERTAINTY GIVEN THE FACT THAT THIS THREE TO FOUR DAYS OUT ... ALMOST AS IF TO REMIND US OF THIS...THE [MORNING EUROPEAN MODEL RUN] SHIFTED SEVERAL HUNDRED MILES SOUTH FROM ITS PREVIOUS RUN AND GREATLY REDUCED ITS FORECAST SNOWFALL ACROSS OUR NORTHWESTERN ZONES...A SCENARIO WHICH DOES NOT APPEAR LIKELY BUT IS DEFINITELY WITHIN THE REALM OF POSSIBILITIES.
9:25 p.m. update: Through 7 p.m. Friday, when the storm will just be getting going, the National Weather Service predicts there is about a 40-70 percent chance of 4 inches of snow on the ground. This is as far out as this particular product goes.
9:20 p.m. update: For fans of CWG’s winter weather expert Wes Junker and Dave Tolleris of WxRisk fame, listen as they (and a couple others) discuss the storm in live internet radio broadcast starting at 9:30 p.m. as you follow our live blog: American Weather Radio show.