Winter Storms -- or -- Why I Have So Little Hair Left

Here is the latest snow forecast generated from the NWS North American Model this evening via a display system from Earl Barker. I want to use this to illustrate a point. People often perceive snow amount forecasts as not very accurate because what they get at their home may be somewhat (or a great deal) different than forecast.

Take a look at the gradient in snow amounts from the NAM model over central Kansas (purple ovals).

The lower left oval is over the City of Wichita. If the NAM is making a perfect forecast, the northwest part of the city will receive 10" while the southeast part of the city will get an inch or less. The gradient is even tighter along the northeast oval: Ten inches in 12 miles!

While I am confident there will be a major winter storm in central Kansas, I'm not confident the band of snow will lay out exactly as shown.

Meanwhile, the two gray areas (yellow arrows) are forecast by this model to get a foot and a half!

If the storm shifts just 50 miles north or south, it could be the difference between nothing and getting buried.

In the afternoon update (below), I depict a probabilistic way of looking at this. For now, it is the best approach.

Meanwhile, the green area in New Mexico is a blizzard watch and the navy blue is a winter storm watch, both from the NWS:

I'll another comprehensive update tomorrow morning.


UPDATE: 9am Sunday. The comprehensive update is posted, scroll up. One new model, the ARM, has 10" at my home on north Rock Rd. in Wichita. In far south Wichita, along Rock Rd. the forecast amount is zero! The forecast snowfall gradient is even tighter on that model. More hair is falling out as I try to make sense of this for blog readers.  :-)