For some, it’s never too early to think about bread-and-milk weather; for others, the mention of snow is about as offensive as insulting their mother.
But, yes, there is a point where it’s too early to be specific about snow. That time is now: when it’s not really a part of the forecast.
What about the trends, odds, or wild guesses for the next few weeks or on through the winter season? As long as you understand that long-range ideas naturally have to be somewhat vague and non-specific to a particular place like a city or county or even group of counties.
If you say it can’t be very specific, then what about maps like this?
Those are two ‘panels’ from the GFS deterministic forecast for the next 384 hours starting at noon Tuesday, November 20th. Deterministic models often give you something like this…
The team lines up for a play that has only two ways to end, right? It’s either 3 points or no points…
That short field goal went back the other way for 6 points and sent Auburn – not Alabama – to the National Championship game. As crazy and rare as that kind of play is in football, it’s just as rare for one of these deterministic models to ‘know’ it’s going to snow in the South that many days away. They give us one solution, and most often it’s not even a really plausible solution that far out.
The best way to look at this is with ensemble models as well as some global-scale signals that might tip us off to a pattern that could produce some snow.
This week’s European ‘weeklies’ show some promise if you’re hoping for snow in the next few weeks, but they do NOT guarantee it.
Check this out: 35 of 50 members (70%) showing at least a trace of snow before January 1, 2019, and there are several that show some what we’ll call ‘healthy’ snowfall, too.
So what else can we look at?
There’s this: unusually warm water in the northeastern Pacific Ocean that usually signals a colder winter in the eastern United States. It helps build a ‘ridge’ over Alaska and the rest of western North America, and that usually means a persistent trough (colder weather) for the East.
And there’s also this: the Madden-Julian Oscillation. In phases 7, 8, and 1, colder air usually ends up in the eastern United States (maybe not exactly Alabama, but it’s close – and sometimes it’s just close enough).
Both the American and European projections take us through those three phases in the next 15 days:
Neither of those tell you much about snow, but they do set up the over-all pattern in which you might start looking for that chance. It does look cold, but the details of how cold and whether or not there’s precipitation involved just can’t be resolved this far out.
What are you getting at here?
Why go through all of that if you can’t really know if it’s going to snow or not?
It’s all about pattern recognition. If this pattern sets up over the next few weeks as it looks like it could, we will see a threat or two materialize in the region. That might not mean ‘Alabama and Tennessee,’ but it may be close.
Last years freak early December snow in Central Alabama and Georgia wasn’t more than a ‘hey, look at that, there might be some snow around in the first week or two of December.’
I doubt there will be a repeat of that anytime soon, but it shows what can happen when things line up just right.
If there’s a real chance of snow, we’ll let you know. Until then, be cautious of the maps flying around on social media promising major winter storms 10-15 days in the future. Most of those people are ‘kick-sixing’ every single time they open their mouths; it’s just that you rarely go back to them to see if they actually had any idea what they were talking about.