The NAM has been bullish on a Mid-South snow event for Friday afternoon-evening with a range of solutions from the extreme case of a 4-8″ snowstorm (Tuesday 18Z) to a rain followed by a couple of inches of snow (12Z Wednesday). On the other hand, every other operational model from the NGM to the AVN to the UKMET to the CMC to the European to the GFS show a light rain event that could possibly end as flurries. Remember those old commercials that said “Four out of five dentists agree that chewing Trident gum prevents cavities”? Would you go to the one dentist that doesn’t agree?? Me neither.
Let’s take a detailed look at the solutions from the GFS and NAM to see what could possibly lead to this discrepancy. Starting with the 12Z GFS at 48 hrs (12Z Friday morning), you can see a weak Great Lakes clipper over upstate NY, an arctic air mass centered over the northern Plains, and a cut-off low over Arizona. Typically, the only way the Mid-South can get accumulating snowfall from an Alberta clipper is when a large upper level trough is centered on the East Coast and the vort max from the Clipper follows the trough from IA into central IL and over northern KY right along a baroclinic zone. In that type of setup it is common to get a 2-6″ snowfall right under the vort max. Usually, the Bluegrass is the big winner from this type of clipper.
The current setup, is nothing like this at all. With a clipper cutting west-to-east across the Lakes, there is not any dynamic forcing to provide lift over the Mid-South. At 850 mb, you can see that there is no flow out of the Gulf at all, as winds at the low levels are primarily zonal. So with no lift and no low-level moisture, I would expect this arctic front to struggle to produce more than 0.10″ of moisture and for any precipitation to be light and spotty. This is what all models except the NAM show. The 66 hr GFS (06Z Saturday) shows that the moisture is gone before the cold air (shown by the 540 dm thickness line) arrives.
So why does the NAM show a snowstorm? Let’s start with the oldest run from 0Z Wednesday. The 60hr NAM from 0Z (12Z Friday morning) shows a much stronger clipper over upstate NY (1004 mb vs. the 1012 mb low from the GFS) and a cut-off low over NM rather than AZ. By 78hr (06Z Saturday) the NAM has the cut-off low over NM interacting with the arctic front from the clipper and developing a surface low (1016 mb) over LA. This stalls out the arctic front and allows for Gulf moisture to overrun the Arctic air mass. Notice that QPF over the Mid-South is between 0.25-0.50″ and is well within the cold air (540 dm line right on KY-TN border).
So why do I think the NAM is incorrect? Consider a time series of NAM solutions for Saturday 06Z starting with the 0Z run and progressing through the recent 18Z run. 0Z run (78hr), 06Z run (72 hr), 12Z run (66 hr), 18Z run (60 hr).
Note that as we get closer to the event, the NAM is losing the idea of the surface low over LA and is getting warmer with less precip on each run. The most striking difference is in the placement of the Southwest cut-off on the last run compared to the first three runs. Finally, now compare the model solutions of the 12Z GFS at 66 hr and the 18Z NAM at 60 hr. Looks like that one dentist out of five now thinks that maybe there really is something to that whole Trident gum thing.
Great uncertainty remains for the Sunday storm. Models vary from the European and CMC, which takes the storm into Lake Erie by 120 hrs to the Navy NOGAPS which takes it up the coast. The GFS ensembles are mainly in between. Current HPC thinking is that the storm will cut up west of the Appalachians, which will bring heavy rain to the Mid-South with a chance of a brief changeover to snow across northern KY.