Tuesday, November 27, 2007

Model Madness!

Here are a series of graphics using the latest GFS data in our Microcast computer model. This model doesn't distinguish between rain and snow. It's all green here. You have to look closely at the 32° line that is labelled as a white line. You can see the variations that we have to take into account when forecasting rain and snow. The first graphic is giving us no precipitation for Wednesday (but the others are fairly moist. We'll keep a chance of light rain or snow in the picture for Wednesday afternoon).


The second graphic shows our next little blip for Friday morning. With the 32° line to the south, we are expecting this to come in the form of light snow...a dusting at most.






Now we're onto the bigger system for the Saturday night/Sunday timeframe. Notice how far the 32° line goes during the overnight hours! If this model should verify the axis of heavy snow (probably in the order of 6") would stripe the ground from Kansas City through Quincy and into Central Indiana. We're still highlighted in the 0.50" accumulated precip however. Should that all come in the form of snow, we'd see around 5 inches.

3 comments:

HeyBuddy said...

So I guess when it comes to models, it's "who do you trust?" GFS shows a big storm; NAM shows that we really don't get a thing. How do you discern what really happens as time goes on and two different models show two different things?

Adam Painter said...

Some factors help you choose which model(s) to trust more than the others. One of the most important factors is recent performance. In the past month or so the NAM has struggled mightily with precipitation and high temperatures. Therefore, I would lean more on the GFS & ECMWF with this one.

For instance the NAM was completely bare of precipitation for today, while the GFS was indicating some areas of flurries and sprinkles. As of 10am there is a big batch of mixed precipitation stretching from Sheboygan, WI to the Wisconsin/Illinois border. If the NAM can't get the precipitation right in the next 12 hours, who says it has the correct solution to what will occur in 4-5 days.

Just my $0.02

tony said...

Basically as of today, if the GFS verifies, we are pretty much out of luck if we are a snow lover. But yet as in the december 1st storm last year, about 3 or 4 days before it hit, the models showed it going south of us, then 2 days before it hit, the models had it hitting us hard with snow, which is what happened. I will keep watching this blog since it is still 4 days out yet.