The Winter Storm Bust

This past Sunday, the computer models were unanimously forecasting heavy snow would fall over southern Kansas and northern Oklahoma beginning this afternoon and continuing into Saturday night. The Governor of Oklahoma (prematurely, in my opinion) declared an official "state of emergency." Food flew off store shelves.

Now it is Friday afternoon and the storm...

Ain't gonna happen.

This is a historic bust and I want to spend a moment discussing what happened and why.

It is has been a very long time since the computer models, in unison, were so very wrong. I hope meteorologists at the National Severe Storms Laboratory, the University of Oklahoma or some other research institution in the area affected by the bust will take hard look at what went wrong and use it to create more robust models. Meteorologists have a bad habit of presenting scientific papers about successes but often do not focus on the busts. This one demands attention.

Next is the role of television consultants and news directors. We have a station here in Wichita that forecasts weather conditions and -- down to the single degree -- temperatures out to ten days. We have no reliable skill with ten day forecasts. But, because meteorologists put them on the air, they are expected to be dependable and of good quality. If it were up to me, the forecasts for the public would go out no more than five days. In the central United States, I wouldn't try to forecast winter storms for the public beyond 3.5 days (commercial meteorologists would continue to make five-day winter storm forecasts for highway departments, utilities and others that need more notice).

Finally, there is the role of the public. Meteorologists don't recommend running to the grocery store for bread, milk and eggs five days before the onset of a storm (the "french toast warning" as Dave Barry puts it). Some get far too emotional about these forecasts.

From an OKC TV meteorologists' Facebook.
H/T James Spann

This is crazy talk. Just ten days ago, meteorologists correctly forecast the blizzard that occurred the Sunday after Thanksgiving. This forecast saved innumerable lives and allowed people -- if they chose -- to change their plans and prevent a major inconvenience.

This past Sunday, meteorologists did an incredible job with an out-of-season tornado outbreak in Illinois.

So, yes, this was a huge bust in Kansas and Oklahoma. But, the level of venom in the above tweets is too much. Meteorologists don't, and won't, get every storm warning correct.

My advice: Unless you need forecasts for commercial purposes, don't worry about Great Plains winter storms more than about three days into the future (for example if heavy snow is forecast for Friday, don't worry about it until about Tuesday evening). 

Addition: The Capital Weather Gang has written a similar,  and very insightful, piece on the same topic. It is here and I recommend it. 

There is one other thing I should have added to my piece above. And, that is the some storms are much more "forecastable" than others. This fact has nothing to do with climate change or anything else. Meteorologists were lamenting this fact in the 1970's and, probably, long before.