11:00 p.m. update: No changes to our forecast after crunching this evening’s model information. It’s now almost certain the entire area will receive a ton of precipitation Tuesday night through Wednesday.
The tricky part here remains how cold the air above will be, and whether precipitation will be heavy enough for long enough to bring the cold air down to the surface and support the snow accumulations in our first call accumulation map. But confidence continues to grow in a moderate to major snowstorm for much of the area..
Taking into account the balance of today’s model data, and our experience with previous storms of a similar nature, we still believe it will be just cold enough for 3-8” of snow for D.C./I-95 and east toward the bay — our confidence level for this area is in the medium range.
Thanks to colder temperatures, confidence is higher for our 6-12” forecast starting just west of D.C./I-95 and stretching west into Frederick, eastern Loudoun and eastern Fauquier counties, and for our 8-15”+ forecast further west along the I-81 corridor.
While this is shaping up to be a significant storm in and around D.C. with the potential for some power outages due to heavy wet snow, areas out west toward Winchester, Martinsburg and the Shenandoah valley are looking at a very serious storm with immobilizing snow, wind (especially at high elevation), and possibly widespread power outages.
Note that while precipitation probably begins as light rain or mix Tuesday evening (probably after 8 p.m.), it may fairly quickly transition to snow especially along and west of I-95, with some snow accumulation possible by dawn. Precipitation looks to end Wednesday evening.
(Dan Stillman and Jason Samenow created this forecast update)
Overview: Winter storm watches are up across the D.C. area as the storm we’ve dubbed Snowquester closes in. Odds are presently quite good for a heavy late-season snowstorm to paste the region with anywhere from 3-8” along and southeast of I-95, to as much as 15”+ over far western (and higher elevation) portions of our area.
While tonight’s model results roll out, we’ll keep you updated on the latest data they show. The entries should not be interpreted as our forecast. Instead, we aim to provide insight into the process of model watching. Once we get past the main early set of model data (around 11:00 p.m.), we’ll provide a short update to our forecast.
8:35 p.m.: If you weren’t around last night and are new to model watching, first up is the 21z Short Range Ensemble Forecast (SREF). This model guidance will be out soon.
In the meantime, the image above shows infrared satellite with an overlay of the 500 mb (i.e., mid-level) pattern and features in it. I chose this graphic to start because the big red “circle” you see over the Dakotas is the main piece our storm as it drops southeast, ultimately to the coast. You can see it on an earlier run of the GFS, and where it is forecast to end up near us. We’ll probably return to this feature at least once or twice tonight.
8:50 p.m. update: SREF looks pretty similar to prior runs. Precip is plentiful across the area when it comes to the storm, with near 2” of liquid equivalent precipitation for everyone. Good run to run continuity is also seen with the surface low placement (see the near-sunrise location: current run | last run). Later in the run (current | last) it’s a little west, which may mean it tries to push the heaviest snow a bit west. Precipitation type appears tricky on this run, as in it might be a little warmer given the features.
The North American Model (NAM) has also started spitting out maps. Sometimes SREF “leads the way” for the NAM.
9:05 p.m. update: By Tuesday evening, the NAM simulated radar shows some of the beginnings of the storm approaching the area, probably as light rain or light mix. The bulk is still to come after midnight and into Wednesday. Very few changes are seen so far from the last run (current | previous). One might be that 850 mb temps (around 5,000 feet — the colored lines) are a little cooler initially, note the 0c blue line (freezing) a smidge south. We look at this (as one feature) for precip type. May be meaningless in the end...
Briefly back to the SREF, earlier “plume guidance” for DC had a mean of 12” of snow. The mean is somewhat skewed by high outliers though, and the snow ratios are likely to be lower than used to calculate (roughly 1” of rain to 10” of snow). Still, an interesting look at the guidance (21z not yet available but can be found eventually by clicking the 21 button).
9:12 p.m. update: Winter storm warnings are now up for the I-81 corridor starting at 6 p.m. Tuesday. Accumulations of 8-12” are expected by the National Weather Service in that area. The NWS forecast map is similar to ours.
9:20 p.m. update: The new NAM appears to be very similar to the last run as the low reaches the coast to our southeast. The last run was pretty big for snow, favoring areas to the west of D.C., overall. The low location early in the morning on Wednesday is also right where the SREF has it. We’ll have to examine the layers more to see about precip type across the area, but it looks similarly cold, especially to the west of D.C.
9:36 p.m. update: A little wonky, but good news for snow lovers. The NAM seems like it’s largely a snow profile in D.C. itself. Basically, all the lines are to the left of the 0C (freezing line) most of the time, except very near the surface. This run is pretty cold and very wet. A snow bomb for most of the area. If you want more on reading a sounding: primer.
9:50 p.m. update:In the lull between runs — the GFS comes out around 10:30 — want to go ahead and share snow probabilities from the HPC (soon to be WPC, Weather Prediction Center). We think these are a bit conservative. Probability maps: 4”+ | 8”+ | 12”+
10:00 p.m. update: CWG’s Steve Tracton offers further/better insight to the SREF plumes I mentioned below in the 9:05 p.m. update. Worth reposting in full here (it is also out for 21z with slightly lower totals overall). THIS IS NOT A FORECAST:
When appraising the plume diagrams, it’s important to pay attention to clusters and cluster means, not just to the overall ensemble mean. By clusters I’m referring to members that group close together. The more members in a given cluster, the more likely it encompasses the the best forecast.
Thus, in the plume linked to above (http://tinyurl.com/bstoj4k) the cluster encompassing snowfall from about 12-18” appears dominant. The mean of the cluster is close to 16”. Additionally, note the cluster has a greater mix of perturbation (model differences) and therefore less likely subject to bias of any given model (e.g, high and low extremes)
Contrary to above the ensemble mean is biased toward the high side not so much because of the extreme values on top but by the cluster just above the overall ensemble mean.
10:25 p.m. update: The Global Forecast System (GFS) is about to get going. While we wait, how’s it going to be cold enough to snow?
Simply, temperatures aloft should be plenty cold in many spots during the height of the storm. Plus, we’ll be starting at night so temperatures won’t rise much with the sun. Heavy precipitation will help draw down these colder readings, bringing spots nearer freezing during the storm.
10:38 p.m. update: The GFS runs fast. It looks like another big hit.As with most guidance, and our forecast, it will favor the western half of the area with the heaviest snow. Temperatures may be a little more marginal (warmer) than the NAM, but they also appear colder than the last GFS run. Have to see the different layers. Without question, the western suburbs get hammered.
10:50 p.m. update: CWG’s Dan Stillman sums it up nicely regarding the GFS: “This was a pretty darn good GFS run if you like snow.”