SUMMARY OF MODEL DATA FROM THURS 12/21 FOR DEC 28 & DEC 30 MID ATLANTIC
STARDATE 201712.21 2300
I probably should of posted of this earlier on Thursday but it has s been a super busy day. This is a the summary of Mode data with regard to 2 winter potential events for the Mid Atlantic region -- a relatively minor to moderate event on December 28 and a much bigger threat December 30. The main focus of this discussion will be the areas central NJ/ SE third of PA into the northern third of NC and d including WVA eastern KY/ eastern TN. This is what I usually mean what I refer to the Middle Atlantic region
As I stated before when it comes East coast winter storms one of my techniques is to ignore the operational GFS model beyond 72 or 84 hours. This is because most of the significant winter storms which affect this region are Southern systems that come up from the Lower Plains/ Delta/ Gulf of Mexico and because of well know n flaws in the GFS models it " loses" Low pressure areas in the southern jet stream. By that I mean the upper level energy that drives the surface LOW w pressure area gets lost or weakens over time. The European and the Canadian models do not have this problem and as result they usually do a lot better when it comes to southern stream winter storm scenarios.
But once you get within 72 to 84 hours the GFS is a lot better model. Let me show you what I mean. This image shows a comparison of the GFS model the DEC 25 possible NYC north NJ / New England snowstorm . The image on the LEFT is from the early Thursday morning GFS model and the image on the RIGHT is from the afternoon Thursday GFS model. As you can clearly see these two images are nothing alike. However what is important to realize that the GFS model suddenly got a clue when the DEC25 Low moved from being 96 hours away to 84 hours away. It is this sort of bad model performance which is why I simply avoid the model when it comes to forecasting in the extended range for any East coast winter storms. I know lots of other people LOVE to post GFS snow maps and having sexual fantasies over them. I simply don't do that because the models is too unreliable. To be sure other weather models such as the European Ukmet and the Canadian have their flaws and miss a significant aspects of weather as well. But they also have much better consistency.
Let me emphasize again that the key to both events for the Middle Atlantic region on December 28 and December 30 is not what any 1 weather models depicting but what the atmosphere is actually doing. The key feature is the development of a large UPPER LOW over southeastern Canada. In fact the Christmas Day LOW which will bring snow to New England moves up into southeastern Canada and becomes part of this enormous system. It is the position of this UPPER LOW over southeastern Canada which keeps both systems on December 28 and the one on December 30 staying to the south-- and if that is correct these will be mostly snow and or ice winter storms and not rain events.
Of course it is possible that the weather models are wrong with regard to this Upper Low . If this large Upper Low in southeastern Canada is not there on December 27 and 28 .then either one of these systems will likely start off as snow and quickly go over to sleet then rain for the Middle Atlantic region. I do not think that is likely to happen but it could.
The operational Euro is fairly bullish on the system DEC 28 showing 2-6" of snow over much of VA and ICE over Northwest and north central NC. The Euro ensemble is weaker but the GFS ensemble model shows a light to moderate snow event over VA MD DEL WVA... and the 12z Thursday CMC
Here is another example - This is the GFS model with regard to the system for December 30. These 3 images are from early Thursday morning .... 6:00 AM Thursday morning ...and then at 12PM Thursday. Notice how the Low position shifts from Chicago to KY to off the NC in the course of 3 Model runs. This kind of shift is well stunning. Some people might even say it is funny. The point is I am hoping that you understand WHY I simply don't rely on this model when it comes to forecasting East coast winter storms beyond 84 hours. The 12z Thurs GFS Model for DEC 30 event buries western and central NC with heavy snow and Ice and major snowstorm for eastern / southern portions of WVA all of VA much of MD DEL into southern NJ and Maybe Philly
Even as good as the European model is or the Canadian for that matter using the operational or deterministic run beyond 5 days becomes a real problem. Much more important than the operational run is the ENSEMBLE mean .
MORE IMPORTANTLY this image shows the GFS ensemble mean for DEC 30th. And we can see considerable precipitation along the frontal boundary running more less in W-E direction bringing snow or snow and ice to the Middle Atlantic region.
The operational European for DEC 30 is warmer -- has a MAJOR ICE STORM over southwest and central VA with heavy snow over WVA northwest VA most of DEL MD PA NJ.