Snow Forecast, Pick Your Pony

December 28th, 2012 at 10:04 pm by under Weather


When it comes to forecasting snow, the forecast models have come a long way.  I remember back when I was in college, the models only forecasted the liquid amount and the meteorologist had to calculate how much snow that would be.  Now, we have numerous models to look at and each model may have several solutions or outputs.  The same model may have a completely different snow forecast, depending on the snow to liquid ratio that was occurring during the snow event.  Is it 10 to 1 or is it 15 to 1?   Meaning, 10 inches of snow for every inch water.

Here is an example of what I am taking about.  Take a look at this model:

18Z GFS Snow Forecast, courtesy

Now take a look at this model:

18Z NAM Snow Forecast, courtesy

The first map is the GFS model and it has at least 3″, up to 4″ for Fort Wayne for this Saturday morning.   Compare this to the second model, which is the NAM and it is only forecasting 1″ up to 2″ for Fort Wayne.  Now, this is not that big of a difference, but which model is right?   Well, that’s my job as a meteorologist to determine that and know which model is handling the current weather solution better.  Want to see a huge difference?

Check out these two computer forecasts for the Day after Christmas snow storm.  Here is the GFS:

Total Snow Forecast from GFS Model, for the 26th of December 2012, courtesy

Here is the NAM:

Total Snow Forecast from NAM Model, for the 26th of December 2012, courtesy

Can you see a difference?  Even though both models had about the same amount for Fort Wayne (about 5-6″ and we really picked up 3-5″), check out the difference in the heavy snow band from Greensburg to Dayton!   The NAM is out to lunch with over 16″ and the GFS was almost spot on with the heaviest snow at about 8″ from Indy to Lima Ohio.

Can you see what we deal with when we forecast?  Now imagine, we look at about 6 models a day and each model may be updated 4 times a day.  Factor in how many different solutions you can have with each model based on just changing the air temperature a few degrees and you have dozens of solutions.  This was pretty much the case with the same system but over Oklahoma on Christmas day.  One model was forecasting over a foot of snow between OKC and Tulsa and that stretch on I-44 was lucky to see an inch, if that.  Actually, it was funny watching some of the OKC broadcast meteorologists handle the forecast and get tricked into the heavier snow forecast.  Speaking of forecasting snow, I need to turn my attention to the snow right now that is moving across the WANE viewing area.

Comments are closed.