Models on Groundhog Day Storm, Southeast Snow

Our first official map is out; please read the Headlines at right for further updates. As I pointed out earlier with quotes from our meteorologists, the models trended eastward overnight, but in the Morning Operation Meeting it was said that our mets aren't buying into that trend yet because of what the models have done so far this season.

That said, there is a lot of buzz about the latest GFS model run (12Z), which says that the low will be much further off the coast, possibly negating the forecast map that we made above by bringing lower snowfall over all, and shifting the heaviest accumulation to the coastal cities. Frankly, I'm surprised to see this much of a swing in the track of the storm, given the model agreement yesterday. We shall see in the next few runs if that is a trend or not.

Briefly before I show some model probability maps, here are the headlines:

1. There will probably be a major (newsworthy) winter storm next week in the East. 2. Severe weather and flooding will also be a threat for the East. 3. Where the heaviest snow will fall is impossible to determine at this point. 4. Parts of the Southeast probably will see some snow. 5. An extended period of extreme cold (at least from a wind chill standpoint) will follow for the East. 6. In areas where snow was heavy, there will be blowing and drifting for at least a day afterwards.

The storm track is by no means set in stone and there's still plenty that can change. To prove this, I'll show several SREF Ensemble graphics below. This model, avaiable on our Pro site, shows different forecasts (ensemble members) from several different models run at the same time with slightly different conditions. The model only forecasts out to 4 days, so we're just getting the beginning of the snow on the end of the model. As such, we'll look in the Southeast, where the question is: Where will it snow?

One way of looking at that is: What is the probability of snow at a given point on the map Monday night?

North Carolina mountains? Looks good, 65% chance. But Atlanta, although it could snow there, the probability is only about 25% as far as this model goes. An alternate way to look at the data is looking at the map below, called a "Mean/Spread Combination."

The black line is the mean, or average of the forecasts; the color is the spread, or difference in the forecasts. Here you see a 24-hour snowfall of 1.5" of snow in Tennessee, but a spread of 1.5", meaning that they could get between 0 and 3 inches. You don't like to see those yellows and oranges because that means that the model is not very confident in itself (the ensemble members are varying widely).