Yet another snowstorm...

There has been quite a bit of chatter about a possible snowstorm to affect the Lawrenceburg area toward the end of this coming week. I've even heard of one account where a fellow heard that "16 inches" of snow is coming. I thought I would post a few thoughts about this potential storm and how it looks to affect us in a few days.

First, a little background on why "every weatherman says something different." There are various computer models that are used to predict the weather. It's sort of like how there are various brands of cars. They are all designed by different groups of engineers and they all have their own idiosyncrasies, but at the end of the day they are all still just a car that gets you from point A to point B. However, some may be better suited for certain uses than others.

The weather models are like this, too. They are all programmed by different meteorological agencies or even nations, and some do better with certain weather patterns than others.

Typically, meteorologists will begin by looking at all the models to see how each handles a storm in question, and then they might average the results to make a "safe" forecast, or they might lean on their own experience to give a little more bias to a particular model that is known to handle the type of storm they are dealing with best. This is why sometimes different forecasters will see things differently.

The most widely used models around here are the North American Model (NAM) and the Global Forecasting System (GFS). The NAM model is slightly higher resolution, but only forecasts out to about four and a half days.  The GFS is a little lower resolution, but it can forecast out to 14 days (not very accurately past about eight days however). Think of resolution as how specific a model can be.  There is also the European model, called the "Euro" for short, the Canadian model, the Japanese model, and a couple others.

I sometimes look through the Euro model runs and some of the others for support, but most of the time I concentrate on the NAM and GFS. They are run every 6 hours, so there are four new runs of these models each day. Two of those runs, at 0:00 GMT and 12:00 GMT, are better runs because they use actual upper air data that is sampled and ingested by the model to produce a more accurate forecast. 

If we look at what the NAM and GFS are showing for the end of this week, this is what we see: The GFS model shows a slightly weaker system that lays down a swath of 3-5" of snow here, while the NAM shows a stronger and more compact storm that drops 5-7" of snow. 

Here in Kentucky, we usually see a 10:1 ratio. This means that for every inch of rain, you have the equivalent of 10 inches of snow.  So we look at the total precipitation predicted by a model for a particular storm, and then multiply by 10. That gives you a ballpark idea of how much snow to expect for the typical snowfall here.

Since temperatures look to be close to freezing, the 10:1 rule should work for this storm. The GFS model is showing a storm total of .25 to .50 inches of precipitation.  Multiplied by 10, this is how we come to the rough estimate of 3 to 5 inches of snowfall.

I'd look for snow to break out across our area on Thursday evening.  It likely will not accumulate rapidly at first but as we go into early Friday morning the snow should pick up in intensity.

The GFS model has done a little better with totals in our recent storms, so I would go with a 3-5" snowfall by Friday afternoon (the 21st) at this point. This will be a significant storm, but not a paralyzing one.

This storm does have a good amount of potential to be our biggest storm this season, but don't let that scare you. It's best 3 or 4 days out to be cautious and just observe trends in the model data. If the model data comes in with more moisture and therefore more snowfall in later runs, I'll update in the comments section below.