We are still getting our arms around all of the data to offer an accurate Christmas forecast. But due to a major airmass change that is set to happen around Christmas Day, it is making it difficult.
That airmass change is the reason that you see a lot of different stuff floating around social media or even hear from friends saying different things. Things like “A White Christmas in Mississippi? Check out this snowfall forecast!” or “Tornadoes possible on Christmas in Mississippi!” are the two I’ve heard about most often.
I’ve even seen some major news organizations promoting the possibilities. This isn’t the first time there has been a chance for snow on Christmas in the weahter models this far out. You only have to look back to 2014 to find it.
Neither scenarios – snow or tornadoes on Christmas – are likely. Both scenarios are possible.
So what’s the hang up?
Forecasting, a lot like life, is all about timing. And the timing for about two dozen different parameters will dictate pretty much everything about your Christmas Eve, Christmas Day and Day After Christmas forecast. On the shortlist of most important things to nail down:
Timing of the warm air from the Gulf vs. Timing of the cold air behind the front vs. Timing of the area of Low pressure
If you caught the news on WDAM last night, you saw Patrick Bigbie showing the difference between our two most reliable models last night for Christmas. One showed a temps around 35, the other showed temps around 75.
That is pretty incredible. Rarely do the models differ that much along the Gulf Coast.
But, in this case, it makes sense. As each model tries to resolve (fancy meteorology word for “figure out”) all of the details, it can come up with drastically different solutions.
I think the National Weather Service in Kansas City explained it pretty well:
Remember the info below when looking up the forecast for Christmas…
At this point, we aren’t saying it won’t be a white Christmas, but we aren’t saying it will be either… pic.twitter.com/mDlsvuH0MB— NWS Kansas City (@NWSKansasCity) December 15, 2017
And a student at Mississippi State showed just how far – and how fast – some of these computer models can get things moving in the wrong direction:
Forecasting a #WhiteChristmas or a long range storm looking at models? The circles indicate possibilities. The red indicates model guidance. Notice how many other possibilities exist especially in longer range forecasts? It’s irresponsible to hype “storms” so far away. #Christmas pic.twitter.com/midbLH1Q5q
— Craig Ceecee (@CC_StormWatch) December 14, 2017
Not trying to rain on anyone’s parade, but I am trying to keep people’s expectations in check.
As a quick aside, the above explanations are part of the reason the forecast was the way it was for the last snowfall event. Calling for 0.5″ to 1.5″ was the most reasonable and accurate forecast that could’ve been done given the available data. Was it accurate in the end? No. Did we find out where we missed? Absolutely. Are we factoring that in this time? Totally. But maybe not like you’d think.
The statistics of a White Christmas
But, let’s come at this next chance for snow from a different angle: Match.
Because the record books for snow in south Mississippi aren’t great, I’m having to use semi-anecdotal info. I’ve asked around “What years do you remember seeing an inch of snow?” to people in South Mississippi. Based on anecdotal evidence, the likelihood that we got 1″ of snow in south Mississippi was quite low. Anecdotally, it had a 0.25 recurrence rate. Or in other words, a once-in-4-year event. That equates to a 25-percent of occurring on any given year.
Christmas Eve, Christmas Day, and Day After Christmas are just three days out of 365. Or just 2.4 percent of the time of the year that it snows in South Mississippi (December 1st through March 30th).
So, there is a 25 percent chance that it snows on any given year. And if you want it to snow on just 2.4% of those days, you need to look at 25 percent of 2.4.
That gives you about a 0.6 percent chance it snows 1″ on either Christmas Eve, Christmas Day or the Day After Christmas.
Then, let’s look at what it takes to get a snowfall of that magnitude, in the atmospheric setup that was unfolding? Around 10-percent.
So, that means, without looking at any forecast data, there is a 0.6% chance we get 1″ of snow on Christmas. And that isn’t a ratio of 0.6/1.0, meaning 60% nor is it a 6% chance.
It is a under one percent chance. A zero-point-six percent chance. A once-in-every-160 year event. Climatologically speaking, the chance is very, very low.
What about the last time it snowed?
The area got – officially – more than 4″ of snow. That has happened three times in recorded history, back to 1885. So, four times in 132 years. While that would mean a once-in-every-33-year event, the last time it snowed that much – officially – was in the 1970s. Before then it was the 1880s. And back then, it did it twice!
So, what are the chances that a once-in-every-160-year event (Christmas snow) happens on the same year that we have a once-in-every-33-year event (4″ of snow)?
Again, without looking at the forecast data, the climatological chances are quite low.
Cutting through the junk is what I try to do when we start to look at the forecast data. So, with forecasts like this one (and the last event), we ask ourselves given the available forecast data and the climatological data, what are the chances it snows? Again in the same year? More than one-inch? On Christmas?
You can imagine how quickly that number gets smaller and smaller.
Even looking at the forecast data available, it calls for a 15-percent chance for precipitation on Christmas morning. And a 13-percent chance it is snow. And 13-percent of 15-percent leaves only about a 2-percent chance of snow on Christmas morning.
Speaking of the forecast
As for this forecast in particular, it looks like rain and storms will be around on Friday and Saturday, followed by lingering rain on Sunday – perhaps even into Christmas morning – before clearing out in the afternoon.
Right now, it looks like temperatures will be in the 70s on Friday. The 60s Saturday. The 50s on Sunday. Then into the 40s and 50s by Christmas.
There may be severe weather on Friday, but not around Christmas. There may be rain around Christmas, but snow isn’t likely. Again, just a 2-percent chance, based on the available data.
Also of note, this forecast is likely oging to change a bit between now and then. Because there is a big shfit coming and the timing isn’t completely locked down, we can’t offer a forecast with high confidence at this time.
A brief note about standard weather apps
Looking at the standard app on your cell phone, or any other nationally produced weather app, the forecast for South Mississippi the forecast will change wildly through the day. Because these apps pull blended model data, when a new model is rendered (this happens every 6 hours) a new forecast is kicked to these apps.
But every new rendering of a model isn’t always done with better information. And – as pointed out above – because the models can really end up in la-la land once they get to 10 days out, you – the end user (or a news organization with no meteorologists on staff) – can end up with wildly different forecasts. Some that show snow, others that show severe storms.
I can hear people saying, “But it got it right last time!”
No doubt. The models did a good job. A better job that humans in some cases. But models accurately depicting the timing and amount of all of the available parameters that far in advance happens so infrequently, that it isn’t worth trusting for high-impact weather events. As noted above, the same apps/models called for a white Christmas in 2014. And, well, that didn’t happen.
If you want to use an app or a site for weather forecasts I highly recommend finding one that is run by a human. Or a local organization. Find someone that is willing to invest in taking the time and energy necessary to do more than just “grip and rip” the model numbers and maps.
