Several weather models are hinting at the potential for an early-season snowfall across parts of the northeast this weekend. Weather models forecasting snow more than five days in advance are like predictions of the rapture—the odds are against it, but get ready just in case.

Social media is lighting up as amateur meteorologists (read: high schoolers) post outlandish model images predicting a foot or more snow to their inexplicably-large Facebook and Twitter followers. Tens of thousands of people are seeing these forecasts and taking them as meteorological gospel.

Let's say this now: there very well could be snow in the northeast between Friday and Sunday, but it's a long way off and a lot can change. Before talking about what could happen, let's take a look at the points for and against a potential snowstorm this weekend.

Points Against a Snowstorm:

1. It's early.

It's pretty early in the year, and while snow in late October and early November certainly isn't impossible, it's not exactly common, either. The interior of the northeast has the best climatological chance of seeing an early-season snowfall. Burlington, Vermont, for example, has only measured >1.0" of snow during the month of October five times since 1943. However, Burlington recorded 7.7" of snow on November 1, 1993. Not likely, but not unheard of.

2. The models are in la la land.

The models are predicting this snow five to six days in advance of any possible flakes falling. Weather models start to drift into fantasy land after about five days, meaning that their accuracy drops and they start spinning off storms that may or may not happen. They're especially notorious for creating snowstorms that never come to fruition.

Take model forecasts more than a few days in advance with a huge grain of rock salt.

3. Temperatures, temperatures, temperatures.

Temperatures are everything. It's likely going to get really cold across much of the eastern United States later next week as a sharp trough digs down through the country and allows cold air to flood in from the north. The models are spinning up some kind of storm to coincide with this cold air, which is why they're predicting snow.

The big question, if this coastal storm comes to fruition, is whether or not the air at the surface will be cold enough to support any significant snowfall accumulations.

Point in Favor of a Snowstorm:

1. More than one model is predicting it.

When you forecast the weather, you can't just look at one model. When one model shows an enormous hurricane slamming into the coast but all the others show sunny skies, you should stay alert but don't start boarding up just yet. When more than one credible weather model begins to show something for more than one run, that's when you should take it seriously.

While five and six days out is a relatively long time in model land, what gives this scenario a huge boost in credibility is that more than one model is showing it. The potential snow is being shown by both the GFS (American) and ECMWF (European) models. However, the models dramatically differ on how much snow will fall and where.

The GFS model is the one that's been the most bullish about the potential storm, showing it for several days now. This morning's run of the GFS shows a nice little blanket of snow covering much of New York state and western parts of Connecticut and Massachusetts. The European model, on the other hand, is showing the snowfall staying much farther inland, with the best snowfall accumulations spanning from the Finger Lakes northeast through Vermont, New Hampshire, and through most of Maine.

They both show snow around the same time, but where and how much is still very much up in the air.

2. It's going to be cold.

These are the forecast highs on Saturday, from the National Weather Service:

Lows will be around or below freezing for many of the colder locations in the northeast. It will be cold, but 35-37 degrees is the upper limit of warmth when it comes to snowflakes reaching the ground before they melt.

3. It's not unheard of.

Just as the "climatological odds bet against the potential for snow" argument plays well for the no snow argument, it goes just the same for the pro-snow point of view. It has snowed this early in the year before—Hurricane Sandy aside—and the calendar date doesn't dictate the weather. One of the worst snowstorms in Buffalo's history, for example, was an unprecedented lake effect snow event in the middle of October 2006.

And again, Burlington measured more than half a foot of snow back on November 1, 1993.

Keeping both sides in mind, here's what will probably happen:

1) It will probably snow this weekend somewhere in the northeast. If The Vane took bets, we'd place the odds at 50.1% in favor. If it does snow, flakes are most likely sometime between Friday night and Sunday night.

2) We do not know how much or exactly where the snow will fall. The most likely scenario would be a coating of snow measurable in inches across some interior parts of the region.

3) At this point, it's too early to say if any major cities along the coast will see any snow at all, including flurries. It's possible, but unlikely at this point (>80% against).

Again, all of this information can and probably will change as the we draw closer to the weekend and the models get a better handle on what the atmosphere will do. For what it's worth, the National Weather Service in Burlington casually mentions that their forecast area will probably see the season's first snowfall this weekend.

Always get your weather information from scientific weather forecasts created by trained professionals, and stay away from any unsubstantiated weather model images you see posted on social media. This post intentionally does not include any weather model imagery because of the high potential for it to be taken out of context. At this point, they only serve to inflame rather than inform.

It could snow this weekend, but it's not set in stone and it certainly won't be the blockbuster blizzard that some would like you to believe.

[Images: AP, NWS]

You can follow the author on Twitter or send him an email.