Why I Think The Forecast Of Snow Is A Bad One
Catchy headline right? Heck I may be the one totally wrong about all this but I like my odds…a lot.
Maybe I should actually rephrase the title to Why The Forecast Of Snow Is A Bad One (From 6-7 Days Out). I sort of like that title better…but it’s way too long.
So what am I talking about…last night while flipping and waiting for the late night jokes to happen on the talk shows I noticed that several of my colleagues including Mike, mentioned the potential of some snow associated with our next big weathermaker due in on Sunday. I saw the same thing in the modelling earlier in the day and just giggled to myself, thinking the odds of this verifying from 6-7 days out were slim at best and one look at the calendar and our weather history made the odds even slimmer.
Look I’m not saying it can’t snow on November 11th or 12th (SUN/MON) what I’m saying is that it’s highly UNlikely that it will snow on those dates. So if I’m working I don’t even mention the chance. Sure it will get a lot colder but typically when this happens the colder and drier air overwhelms the region and pushes any precipitation off towards the east before the atmosphere as chilled down enough to convert the rain to snow. The models were showing the “perfect” set-up for some snow to actually occur so hence the reasoning for the flakes int he forecast.
1st let’s play the odds game…I looked at weather records dating back to 1889 from the FOX 4 Weather Almanac page and of all the November 11 and 12ths out there, we’ve had measurable snow 3 times. On 11/12/86 we had 1/10″…on 11/11/1912 we had .2″ and finally on 11/12/1893 we had a whooping 2.5″. Now granted that doesn’t include the potential for a few flakes of snow that don’t amount to a trace or more (that don’t show up in the almanac) but still. So when you figure that all out that is almost 450 days (all the 11th-12ths) and 3 of those days have had measurable snow for an amazing 00.7% chance of snow on those day particular days. Like I said I like my odds.
The reason forecasters jumped on this was that both the EURO model, the American model (GFS) and with a nod from the Canadian model all showed the potential for something to happen somewhere from Manhattan, KS eastwards through Columbia, MO, depending on your choice of model. Fair enough, it wasn’t as if my colleagues were pulling things out of thin air and making it up. Here’s the thing, as we ALL know in the weather world, these models CAN’T be trusted on the specifics of weather from 6-8 days out. Heck while there are strong more reliable signals of a nice shot of cold air hitting on Sunday, there are timing differences, differences in the amount of cold air that will move in, and differences in the length of time that cold air will be here. We aren’t sure how much rain we’ll get from this transition, whether it will be a lot of a little. Yet some feel confident enough to insert snow in a day 6-7 forecast. More power to them.
That’s not my philosophy at all. It’s NOT that I wouldn’t insert snow in the forecast, but I would wait a couple of more days at least to do it because of the time of the year. Look I’ve forecasted snow on day’s 5-7 of the longranger numerous times in the past (maybe not last year though ) so that is not the issue. My problem with those forecasts is the HIGH likelihood that the models are wrong…and the time of the year we’re in. If this was mid JAN, I’d certainly be more in inclined to do something like that but not a week away in early November.
Lo’ and behold those same models today (overnight) are singing a different tune (shocking) about the snow chances. The GFS model says no chance at all. The EURO has pretty much thrown in the towel as well and the Canadian is still saying that snow will fall (and potentially accumulate). I even looked at the ensemble runs of the GFS and only one out of the 12 gave us the potential for some snow. Again I like my odds on this one.
So what is the bottom line about my forecasting style. I will never mention the potential of snow a week away in early November and I really don’t care what the models say because odds are they will be very wrong. Now IF I was working on THU or FRI night this week and it still looked interesting, then I would certainly be talking it up, but not a week away that’s for sure.
Now lets see if I have to eat my words on Sunday or Monday. I do think there is a decent chance that Sunday afternoon we’ll only be in the 30s compared to Saturday where 70s are possible (depending on clouds). If things happen as I think they will, Monday might be mostly sunny and highs in the 40s.