WINTER STORM BUST
If 3-6" is forecast and 1-3" is observed, it is a bust. Discount the fact I did get northern Wisconsin to 3-4" before the snow started. Generally unhappy with my work this weekend, but southern Wisconsin turned out okay.
The forecast for Sunday 1/14 to 1/16 was in good shape until the Saturday 18Z GFS run. Agreement and censuses was great, but the 18Z run was when everything started shifting. I waited until the 00Z-12Z SUN run to make changes and I never base predictions on the 06-12Z runs. I already felt the frustration Saturday PM, dreading the need to make major changes to the forecast; I told myself to "forget what it was and start over". There was still hope for a northward trend which is common at that time period.
If your forecast map was mostly accurate before Saturday PM; you're either god, have the biggest pair known to man, extremely biased, or very inexperienced. There is no gray area. The truth is nobody had it right.
The northward return trend did not occur. This was one of two personal biases I was happy to break in the past couple days. Northern Wisconsin became the tough call. With the NWS issuing advisories galore, it was 3-4" for northern Wisconsin on the final map. I felt locked in by the previous map, old ideologies, and didn't want to give the audience the impression that we are just throwing maps around. I can't quantify the time and effort that goes into this for the sole purpose of being accurate. It's never been about making pretty maps for click bait or attention. I need these to be right!
STEP #1: DO NOT POST SPECIFIC PREDICTIONS BEFORE 24HR
To improve my future forecasts I'm making big changes to the 24-76 hr forecast period. Nobody gets it right the first time. I learned this weekend that model agreement and consistency mean very little, even within three days. We've demonstrated that 24hrs out you can have shifts of 100mi or more. This event was definitely on the extreme end and most will not wobble like that. There was a list of ten other things I wrote down that will help for next time.
TRACKING THE NEXT ONE
With all the issues of this last system it seems hypocritical that I would even bring up the topic of next week. As a passionate storm chaser this is never going to change.
Finally the pattern we've been waiting for - literally the entire winter
- is setting up to the west. Our upper level pattern will favor troughing in the west with negative PNA
arriving in significant fashion for the first time since mid-November. This long stretch of pos PNA
was associated with below average temperatures across the eastern United States (this is not a causal relationship). The upcoming pattern change will set the stage for one, possibly two major storms to cross the country. If you don't get the first one, you'll probably catch the next. Happens every winter. The question for Wisconsin is precipitation type. All snow, snow freezing rain, sleet, or mostly rain? Dry slot? Will it hit 100% of Wisconsin or 40%? I could say it will impact "Wisconsin" right now and have great odds of being technically correct, but I'll default to traditional conservative approach for now. If I wanted to stir up hype I wouldn't write this in a member's only-limited access blog. I trust that you (the reader) understand the risk of forecasting a snowstorm 7 days in advance? You obviously are taking time to understand if you've made it this far into the post.
Just because we have the western trough doesn't make the storm bulletproof. Flow is almost zonal, not massively amplified like we saw on earlier runs. Especially the Euro with it's warm long range bias. More amplitude was making this system go polar (south -> north). Flatter flow would force the storm on a flatter trajectory, almost straight west-east through the plains. I think phasing will be complicating factor because right now we assume the storm is full bore before it arrives. If phasing is slower, then the system could have issues relating to things not being where they should be, which would lower confidence. At the end of January we should have enough cold air hanging around to work with. The 12Z- 1/15 GFS end result is pretty emphatic with 12" > snow potential.
Forecast ensemble and operational simulations are spread out, far from consensus so expect this system to keep dancing around the next few days. Relative to GFS prog above, the 12Z Canadian is weaker/further north. It's 00Z ensembles were split with half favoring the operational/control, the other half resembling the GFS. The 00Z Euro was stronger/further north and has been since the beginning of time. CIPS is suggests weaker snow for southern Wisconsin, still noteworthy. The GFS looks weird when you compare it to the top analog, showing a weaker 300mb wind pattern relatively speaking. The GFS forecast has a stronger low level cyclone.
You could call the GFS an outlier at this stage. 12Z-18Z runs reinforcing GFS forecast.
The Climate Prediction Center has been on top of this one for awhile. The 6-10 day precipitation trends map is highlighting the great lakes with high likelihood of above average precipitation. This doesn't mean snow, it could be rain or thunderstorm! But there will be precipitation.
In conclusion, I am expecting *potentially a high impact* storm this weekend into early next week. Whether it trends more snow or rain, we'll see. Our experimental in-house forecast model has been suggesting a cold shot behind this storm. I wouldn't take it to the bank but I think both the cold and snow compliment each other right now. It is early and things can change.
Become a member or sign in to unlock this content!