Joe’s Weather Blog: The odds of a White Christmas (FRI-12/15)

Good morning…I’m hearing and starting to see a LOT of shared posts with all sorts of terrible forecasts for Christmas. Many come from weather enthusiasts who are just posting terribly flawed model forecasts of tons of snow that really are not based in reality. Others are coming from trained meteorologists. I can only control what I can control. The “sexy” colors of a bizzilion inches of snow covering areas are widely shared…the images a few days later of little to no snowcover are shared a fraction of the times. As I see more and more TV folks go down this road…I’m starting to give up the ghost in hoping others could just wait a few more days before going so all in. Going all in then backing away then going all in and backing away is not the way to approach forecasting in my opinion but that is just me. It’s when that same person who changes a forecast 10 times or has 17 different posts about 17 different outcomes then says “see I was right”! That still ticks my off in my “Joe Weather World” but again I can control only what I can control.

The reason for that mini-rant was some posts I was seeing today from supposedly more reliable sources. I don’t want to start going down that path on-the-air at least. I will try to provide some context in the blog though for you because when I decided to start this blog more than 15 years ago that was what I envisioned for it. A place where I can talk about stuff that I wouldn’t necessarily talk about on the air. When I do go down a particular road “on the air”, either for the set-up of snow or for why we may see little snow at all…It’s done when I feel it should be done…when I feel that I won’t be changing the forecast 50 times. You may notice that on-the-air I haven’t really talked much about Christmas…I have mentioned some general trends though…colder…but in terms of ice…snow…blizzards…sleet…stuff like that…it’s just wait and see. If that’s not “sexy” enough for you…I’m sorry. Some of you are kind enough to tweet me or FB me some of these wilder forecasts…asking for my opinion…that is GREAT. All I ask is to think back over the last few years…think about all the times you saw something (particularly winter related) that blitzed our area with all sorts of terrible conditions…ice…snow…whatever…and think what really happened.

Not much right? Less than 10″ of snow and an ice storm that never really came together in the last 2 years isn’t exactly verifying those wild forecasts that you might have seen on social media. I rest my case. 

I can control only what I can control.

I still think posting forecasts with numbers on them from 10-15 days out…just taking the model verbatim (mainly when it comes to snow) is crazy…but then again when I was a kid and IF I had access to all this GREAT information that they have access to now…and considering what a weather nut I was growing up…I could EASILY see me doing this. So I do want to be careful in my criticism for mainly younger weather enthusiasts. I just wish it wasn’t shared so much.

Whew…now I feel better…that was cheaper than seeing a psychologist! Hope it made sense!

How about something graphically somewhat representing what I just wrote 500 words about…this is from the NWS in Dallas.

Onwards…



Forecast:

Today: Mostly sunny and pleasant for mid December. Highs well into the 40s

Tonight: Fair skies and cool…but really not too bad…lows in the 30s

Saturday: The better of the two weekend days…by far. Windy and warmer. Highs in the 60-65° range. Winds gusting to 35 MPH…perhaps higher in the afternoon. Mostly sunny skies.

Sunday: WE may actually have at least some rain in the area, especially from the 36 highway corridor southwards. The best chance is through mid day…then drier weather in the afternoon. A few light showers are again possible Sunday evening/night. Highs in the mid 40s



Discussion:

Not that I got all that off my chest…let me post some long range snow forecasts! j/k

Actually I may start dabbling more into that towards Saturday night or Sunday regarding Christmas week. For those who watch each run of the models…last night wasn’t kind for snow lovers really. It’s no secret the models have really be flipping and flopping. That is what they do…most of you weather enthusiasts know this. Trust me…I’m up till 1AM or longer each night when something is even remotely possible during the winter. Last night though…after looking at the EURO model come in…I should’ve gone to bed early.

Does that mean it’s right…does that mean the GFS model which flipped again is right in predicting little snow heading into Christmas is right? Not really. I do feel, like I’ve written and stated for awhile…that several things are in our favor. The most important…it will be getting colder for at least some of the time. That, just in itself, is somewhat of a reversal in a month that has been running warm overall. I do have questions though about just how cold it gets with a lack of snow on the ground from here northwards and how long that cold air can sustain itself.

I’ve also written in the past about the use of what we refer to as “deterministic” model runs…these are the more easily available pieces of data that many solely look at to get excited or crushed when the weather they want shows up or disappears. From this far out though…that is NOT necessarily the right approach. My gentle suggestion is to look at the ensemble runs of these “deterministic” models.

This is a great site for that information…click on models and then look to the far right side…the GEFS represents the GFS model ensemble and the EPS is the EURO model ensemble.

Why are ensembles a better way to go…well as I’ve mentioned before…the models are only as good as the data that they ingest. The model output is not perfect as we all know…especially farther out in time. My colleagues at the NWS in Springfield, MO put this graphic out…and really perhaps the best way of looking at ensembles is it MAY give you a level of confidence in trying to figure out what MAY happen in about 5-10 days from today.

For the GFS modes…there are about 20 other “perturbations” run with different physics or different initial conditions…for the EURO model…more than 50 other perturbations are run. You then average all this other runs into an “ensemble” forecast. So does the ensemble forecast somewhat agree with the deterministic forecast that is more readily relied on? If so…you get a bit more confidence in the possibility. IF the ensemble run is vastly different…confidence decreases.

In the case of the ensembles for what MAY happen towards Christmas…there is still enough doubt in them that some snow is certainly still on the table as far as a forecast goes around Christmas…despite the “deterministic” saying little to no snow chances. There is still energy hanging back in the SW that could be ejected into the cold air through the Plains. I’m not jumping up and down about this possibility at this point…

So for snow lovers…don’t give up the ghost quite yet. I’m not. By the same token though…don’t rush out for that new snow blower either.

Simply the science of weather models is just NOT there yet. Heck a couple of days ago…no snow was forecast for the NE part of the country…now several inches of fluffy snow may fall there. When models flip back and forth like they do…there is still a lot of potential…it’s just a waiting game really. That’s why I STRESS NEVER TO GET TOO EXCITED about snow depictions until at least 4 days before the event itself.

Several days ago…there was no rain in the forecast for Sunday…now we may end up with <1/4″ or so…again just highlighting that IF the models from 2-3 days out are wrong…then the models from 10-15 days out can be VERY wrong. Specifics are always the hardest part of the weather forecast when it comes to changeable weather.

Finally the title of the blog…the odds of a White Christmas. Historically this is what the stats say…

Graphic via Climate Central

More specifics…

Now my forecast chance of at least 1″ of snow on the ground on Christmas Day:

20%

Our feature photo comes from Ron Tilsworth wayyyy down in Harlingen TX when they had snow last week…hey if it can snow there…it can snow here…right? I think? 

Joe

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s

13 comments

  • Rockdoc

    Excellent blog post! And don’t worry about your rants, because it’s right on. I figure that if the models start converging towards the same solution by next week Wednesday into Thursday then we’ll have a higher level of confidence as well. Cheers, and have a Great Weekend Joe!

  • Mike

    Joe-

    Thanks for some sanity in the midst of all the hype. I love snow and hope that we get buried in it weekly all winter long. But, I have to be patient and realistic: I live in NE Kansas, NOT Buffalo, NY.

    That being said, I look forward to another winter of hoping for snow and reading your blogs,even the rambling ones full of stuff I don’t understand much of!

    Happy Holidays!

  • Patrick Trudel (@sedsinkc)

    Bravo, Joe. Bravo. Unbelievable how irresponsible some who should know better are about sounding the alarm way too early about winter weather, then flip-flopping a day later when the model data 12 days out shifts markedly. They should know better as scientists, but clicks and ratings are what it’s about for some, above all else.

  • Patrick Trudel (@sedsinkc)

    As for a weather enthusiast, I have in the past sometimes been guilty of sharing model data more than 5-6 days out, but I try to make sure and put a caveat in my post that this data is highly suspect and should not be viewed as accurate so far out. I may possibly have omitted this caveat on occasion, not sure about that. I do try to observe the general rule that model data more than 5 to 6 days out can not be trusted many times.

    • Brandon

      Yeah, I’m the same way. When I first started following weather models, I’d see snow 10 days out and get excited. Nowadays, I don’t even get bothered one way or the other until about 3 days out.

      • Joe Lauria

        All weather enthusiasts love eye candy model forecasts…me included during the winter at least. Trust me I’m not staying up till 1AM looking at the EURO model for my health! For the last couple of winters though…there hasn’t been much of a need to stay up to look at anything considering how boring it’s been!

    • Joe Lauria

      Hey I don’t mind showing something really long range in the blog…and talking about it as you know. When you’re in a bupkis weather pattern for 2 months…it’s a struggle to come up with stuff as you know! Context though is so important!

  • Richard

    Chance of a white Christmas “Odds do not look good.”
    Thanks for the refreshing, honest opinion on air Joe !
    It’s better than changing it every day with percentages saying 10, then 51, then 45, then 55, then back down to 50%.
    Honestly it is refeshing to hear you and read this blog.

    • Joe Lauria

      Hi Richard…thx…and as my blog readers know…my favorite forecasting challenge is snow…I do enjoy nailing that forecast moreso a severe weather event. So much at play. Welcome to the blog. One of these days it’ll switch…right?

    • Heat Miser

      Not to mention it’s cheap. 50 percent, 55 percent, 51 percent..of snow ten days out? First of all, we cant predict snow in KC ten days out at all, and neither can the models. Secondly, if you are going to go their in spite of that, and give predictions…man up. Are you predicting snow or are you predicting no snow. Fifty percent isnt a forecast, its playing it safe and saying “who knows”, its a coin flip.