Forecasting Snow

By Bill Korbel

If there is one thing that all meteorologists in the NY/Long Island area would agree on, it's that forecasting how much snow will fall is one of the most difficult, if not the most difficult thing we do.  Why? Let me explain.

To come up with an accurate snow forecast we must get many things as close to perfect as possible.  Obviously, we need to know when the snow will start and when it will stop.  That's not so easy on Long Island where we are often on the dividing line between snow and rain.  We need to know the temperature at the surface and up through 10,000 feet very accurately.  We have to factor in the temperature of the ground.  Is it frozen?  If so, the snow will start accumulating sooner then if the ground is not frozen. How much water is available in the atmosphere to turn into snow flakes.  A good average is 1" of liquid water equals 10" of snow (a 10 to 1 ratio).  But if the temperature is very cold, the ratio can be 15, 20, even 30 to 1.  The difference between  1" and 4" of snow can be as little as 1/3 to 1/10" of an inch of water. 

On the other hand, If temperatures are warmer, the ratio can be as little as 5 or 6 to 1.  That same 1" of water will produce only 5 or 6" of snow.  We also need to calculate the intensity of the snow and the rate of accumulation per hour based on all of the above. 

In a perfect world, our computer models would calculate all of the numbers very precisely. Alas, they don't.  We have many computer models and they all are based on different assumptions and equations.  Rarely do they all agree.  We try to find which one is handling the situation the best and go with those numbers.  We also monitor the storm to see if it is following or deviating from what we expected. 

Let's assume that we have perfect predictions of start and stop time, intensity, moisture and temperature.  That still won't produce a perfect snow forecast because small things further complicate how much really falls.  Heavy snow tends to form in bands that are embedded in lighter snow. If one of these bands of heavy snow sits over your house for an hour or two, that could make a huge difference in how much snow falls.

What it comes down to is this. We make the very best forecasts we can using the tools available, but when dealing with snow, any forecast, because of the many variables involved, is just an approximation.  In warm weather, if we say 1/4" of rain will fall but it turns out to be 1/2", no one notices.  It rained. It's wet. There are puddles. It was a good forecast.  But, If we say 1-3" of snow will fall and we actually get 4-6", the forecast, to many people, is a bust, even though it was just as accurate as the summer rain forecast. 

Hopefully this will give you a bit of insight into the ever humbling world of meteorology.  Thank goodness spring is only 56 days away.

Bill Korbel

Comments

[John]

Those darn computer models!

Heh. I remember being taught about those paste-on low and high systems that meteorology used to be about, and how to interpret those now-fossilized NOAA maps. I think of myself as pretty jaded, but even I have to be impressed, sometimes, at the difference a couple of decades makes, knowing that forecasters like yourself are now using the equivalent of a building full of mathematicians to predict a few days out.

Of course, then that darn butterfly in Japan flaps his wings again...and there's probably an occasional round-off error in the software...