On the whole, our forecast was pretty good for the Monday (1/22) snow storm. There were a few areas that slipped through the cracks. This post is a discussion of some of the things that went wrong with the primary goal of making my future forecasts better. I believe that if I write it out it will have a better chance of sticking in memory. This post would be useful for someone interesting in making snowfall maps or trying to understand how they work.
The main snow band setup across northwest Wisconsin into the UP. For this event I wanted to test a strict snow accumulation bracket. 1-3,3-6,6-9,9-13 and no in-betweens. Couldn't help myself over the UP. We went 6-9" in this band highlighting southeast Minnesota and Marquette(MI) with 9-13". We were definitely on the conservative side. I'd say 70 percent of the 6-9" range was accurate. I would rather a few spots over than a bunch under. We are missing data points in northeast Wisconsin. The color scheme does not line up between maps but I will fix that.
Most Proud Moment - Northern Wisconsin
On Monday morning I saw the slight northward trend and stronger dry slot forecast for later. I trimmed totals in the Stevens Point area. Short range models and GFS were cutting even deeper but I didn't. I lined it up with where I thought the main snow band would be. 13.5" was observed in Phillips and I'm okay with that. There will always be spots that go over in the main band. I might have been too tight in the Wausau area trying to illustrate sharp rain/snow cut off. I think we can change the snow ranges to make a better fit too.
Here are the problem areas:
#1. Southeast Minnesota
On the first forecast I had the 9-13" further south. I moved it north after I saw the 00 and 12Z NAM closer to Minneapolis. We were too low in this area with the thought that NAM would overforecast QPF, and it did. I saw similar QPF on the EC. The difference was snow-to-liquid ratios. We highlighted the correct place but didn't go far enough because of QPF. Good lesson to be learned here that under certain situations, certain models do better than others. With Kuchera amounts of 15"+, that should tell us something! Snow banding is a very efficient method of snow accumulation.
#2. Far Northwest Wisconsin (north of main snowband)
Too stingy on Duluth. I like to increase totals up there just because(snow ratio/climatology) but I didn't this time. I would consider this a fairly minor miss.
#3. Upper Michigan
This was actually a successful forecast. I circled because we didn't go far enough. I predicted this area to under-perform with models decreasing QPF from west into the UP. Some predictions from others were much higher, so we were on the right side of things. We contemplated bumping up to 9-13" but confidence was too low. Needed to be more confident in the models here.
#4. Southeast Wisconsin
This was the biggest miss in my opinion. The second craziest thing that's happened this winter. 9" of snow in southeast WI is more than most places in the snowstorm up north! Saw it for about 20-30 consecutive runs bouncing around southeast Wisconsin. I saw it as minor/small scale wrap around, weird surface low relocation thing. Bigger snows to the north took priority so I pushed this off into it's own separate bucket planning to deal with it but never did. Never thought of it as a 4-9"er. When MKX issued the advisory I thought they were just doing what they always do lol, but it was a great move on their part. Just fell through the cracks.
Overall, I give this forecast a B-. I was too tight on the southern and northern snowfall extent in the 1-6" range. Conservative is good, but maybe too conservative. Given character and strength of the system it was okay to go bigger. Didn't go far enough with heavy snow in southeast Minnesota. Didn't emphasize snow banding enough in northern Wisconsin. And missed the boat in southeast Wisconsin. The good news is that all of these issues are correctable! It's not model limitation; it's "me" limitation. Just don't make the same mistake twice and we should be good to go!
Final observed snow map
Become a member or sign in to unlock this content!