ABOUT NE US SNOWSTORM THREAT LATE FEB 18 INTO FEB 19
STARDATE 201802.15 1030 est
The overnight and early Thursday morning models continue to go back in fourth with the potential snow event for Saturday night February 17 into the predawn hours of the 18.
The key to this forecast which is quite tricky but very interesting are the following factors
1 The timing of the event
2 The track of low pressure area itself
3 the rate of heavy precipitation -- if it comes down comes down hard enough it will keep low level temps JUST cold enough so the precip stays snow.
First here is the operational GFS model from early Thursday morning - the 0z run. As you can see the model keeps a low pressure area fairly weak and far to the south. This results in a lack of heavy precipitation so the low level temperatures near the surface go above 32° and precipitation falls as either rain or rain snow mix in the big cities of the northeast.
This is why the snowfall map is essentially well nonexistent on the 0z THURS operational GFS model. The GFS ensemble is also somewhat less snow here then is previous runs.
Next we have the operational Canadian model and because the LOW pressure area is somewhat stronger and further north ...and the model handles the cold air wedge in a little stronger ...the temperatures are just cold enough to keep the precipitation fall mostly as snow over the big cities of interstate 95 from Washington, DC to Boston. This would be a significant snowstorm for the big cities of the interstate 95 if the Canadian model or to verify. I am not a big fan of the Canadian model but in this case it's solution is viable and it cannot be dismissed.
Here is the early Thursday morning European model. The difference here is that it is somewhat warmer at the beginning of event which reduces the snow map fall amounts over nor the Virginia Washington, DC Baltimore. However the European model still shows a fairly well organized and developing area of low pressure especially once a leaves the coast. As a result the heavy precipitation brings down more snow from Philadelphia to Boston.
If you look at the Mid Atlantic region in more detail we can see that the 0z operational Thursday European model ...does have the rain snow line further to the north and as result only the northwest 25% of Virginia and Central Maryland see significant accumulating snow.
That being said there are significant differences between what the operational European models forecasting for snowfall and what the European in ensemble are forecasting. For example in Washington, DC the operational models forecasting snowfall up to 5 inches but the ensemble is forecasting 2 inches.
We see similar such discrepancies over much of northern Virginia and Maryland. For example Culpeper VA according to the operational European sees 6 inches of snow but the ensemble only shows 2 inches. The same sort of discrepancy exists for Staunton VA and Winchester.
In Philadelphia the operational European model has that city getting 7 inches of snow but the ensemble is under 3. In New York City operational model shows 6.5 inches but the ensemble mean is 2.75. And in Boston we see also a big discrepancy between the operational model and the in ensemble mean with respect to snowfall from this event
UPDATE ON FEB 18 SIGNIFICANT SNOWSTORM THREAT FOR I-95/ NE USA ... weather model come back to reality as I knew they would
First I will do a new snowstorm podcast at 4:00 PM when the European ensemble finishes running and all the other models have completed their cycles.
The various weather models have backed off significantly on the idea of a major or significant east coast snowstorm for the Northeast U.S. on Saturday night FEB 17 into Sunday FEB 18. Yes there is still the risk of these areas seeing some small snow accumulation[-- from northern and western portions of Virginia through the big cities of the Northeast into Boston and much of New England. But it does NOT look to be a major event and the snow may change over rain in the big cities of the northeast from Philadelphia to Boston.
Longtime followers of wxrisk.com know that one of the BIG rules I have is that when it comes to East coast snowstorms is this - sure weather models are nice to look at and very useful most of the time. But East coast snowstorms only occur when the upper atmospheric patterns are set in certain positions. Simply because a weather model is depicting LOW pressure area on the coast of Virginia or North Carolina does not "cause" the atmosphere to develop a snowstorm.
Weather models reflect the atmosphere or at least they try to reflected and simulate it. They don't "force" the atmosphere to "do" anything. So when you see these pathetically silly ass weenies claiming victory on a snowstorm forecast that is still 96 hours away from occurring.... Then you know you are talking to somebody who has no clue.
Interestingly because of the way the precipitation is going to break out and develop ...the Shenandoah Valley and the northern Piedmont including the Washington, DC and Central Maryland area could end up seeing the "best snow' out of this event.. This is because the precipitation is going to break out from SW to NE and the cold air will the deepest and entrenched over the Shenandoah Valley .... western Virginia into central Maryland.
By the time the precipitation pushes further north ...the large cold HIGH over New England will have moved off the coast and the low level surface winds will have become southeasterly--- bringing in mild air from the ocean and changing the snow over rain in places such coastal New Jersey New York City Long Island and southeastern New England.
The early Wednesday morning Canadian model still shows a significant snowstorm from Washington, DC Baltimore. However the more reliable and better weather models such as the 0z and 6z GFS and 0z wed European models-- have come down significantly on the snow amounts.
More importantly the GFS ensemble mean shows only a few inches of snow accumulating over the interior and shows the snow going over to rain with little accumulations in places such as central and southern New Jersey .New York City... Long Island and southeastern New England. It does show accumulating snow of 2-6" ver oportions of the Shenandoah Valley ...the northwest third of Virginia into the Washington, DC /Baltimore metro areas and up into Hagerstown and Philly.
The 0Z WED European operational model also has reduced the snow amounts significantly as you can see here. In addition the 0z WED has the Low staying further south ...which means the heavier precip stays to the south of NYC Long Island and southern New England. This in turn allows for low level temps to warm . Ideally in this kind of situation with marginal temps you want the precipitation to come in at Night and come down hard. Also
European ensemble mean and the operational European model are now in much better agreement.
Yesterday if you recall the 12z TUESDAY operational European model showed large snow amounts between 7 and 11 inches over many of the big cities from DC to Boston. But the European ensemble mean only showed snow amounts of 2-3". This sort of dichotomy between ensemble mean and the operational model is always a warning sign to forecasters to not accept the operational model at face value and also to realize that this is a very tricky and uncertain situation. Therefore the fact that the operational European and the ensemble on this Wednesday morning are now in close agreement is a sign that the model uncertainty may be ending.
Finally longer term is obvious that the blocking pattern that develops in the upper levels the atmosphere over the United Kingdom and Iceland ...is going to retrograde (move backwards) into Greenland by the end of the month. Obviously this is going to have a significant impact on the weather patterns over North America especially east of Mississippi River. It will knock down the persistent Southeast US ridge and probably allowed temperatures to return to near normal levels.
However the problems remain in the Pacific side of the equation. The mean trough remains on the West Coast which is not good if you want East coast winter weather. It is not necessarily horrible... but it is not ideal. In addition the PV gets pushed back into northwestern Canada in the 11 to 15 day and that also is not a good sign.
In summary I will hold open hold open the possibility that there might be something wintry for the Eastern CONUS in the first half of March 2018. But right now this is just a possibility and it is nothing to get excited about