First "Bust" of the Season and Lessons Learned

Only one subject to cover tonight, and that is the Wednesday Night/Thursday Morning storm, which is rapidly becoming an early disappointment. While the forecast models, if anything, have shifted a bit to the east from yesterday, two other overriding factors lead me to tone down the accumulation forecast. The first factor is that the cold air just isn't going to come in quickly enough for any significant snow to fall. Obviously, it's early in the meteorological winter so cold air will usually be a problem, but I admit that I never expected it to be this much of a problem. The other factor is that surface observations support the low to track a bit further to the west of where it is being forecasted by the models. This as well does not help snow accumulations. Therefore, I am toning down forecasted accumulations with this storm to around 1".

Obviously, this is disappointing, but the weather is inherently unpredictable and it's a given that most forecasts will have to be adjusted as we get closer to the event. If one looks back through earlier blog entries, one will find many other times when my original accumulation forecast had to be adjusted up or down as the event drew closer (although somehow...only the downward adjustments seem to get noticed :P).

Still, there are lessons to be learned from every busted forecast, and here are the ones I offer up:

1) Expect the westward shift to be present again this winter. A theme of the last few winters has been forecast models ejecting an upper low out of the southwest too quickly, resulting in an original storm track of further east than what actually happens. Based on this storm, that trend may be occurring again this winter and I will have to watch out for that and take that into account when forecasting 3-7 days out.

2) Forecast models usually overdo cold air, especially early in the season. Somehow, cold air always ends up being one of the main problems for almost all winter storms, especially early in the season. Yet forecast models usually don't see this problem until 2-3 days before the storm. After a long break from forecasting, I admittedly underestimated the cold air problem and hope to better account for the cold air problem in future forecasts.

So although this storm won't have much to offer up in the way of snow, it offers me a chance to learn from my mistakes and improve my forecasting as the season goes on. And I would still look for snow to be falling Thursday morning, just not as much as previously expected.

Finally, I would like to close by offering up a new "grading system" of snow day probabilities for Lake Orion. I will still be doing yes/no predictions the night before any significant storm, but for the medium/long range I will offer a quick look at the possibilities of a snow day with a grading scale.

This idea was proposed to me by Vinnie Warren, and I thank him for suggesting it, as it is a good idea.

A = Lock it in, snow day has a great shot at happening
B = Above average chance for a snow day, but there's still things that could go wrong
C = Average chance for a snow day, things could break either way as the storm draws closer
D = Below average chance for a snow day, but it's not impossible
E = Forget about it, move on to the next storm if you only read this for snow day chances

For the Thursday Morning Storm, the current snow day grade is an E.