FOX44 - Burlington / Plattsburgh News, Weather & Sports12.03.11 Don't Always Believe What You See

12.03.11 Don't Always Believe What You See

When I came into work today I did the usual. 

I took my old forecasts and updated them.  I had what I projected, wrote down what the ACTUAL highs and lows were, counted how many degrees off I was, then went on to see what's currently happening.  Nothing out of the ordinary, yet.

After looking at current conditions, levels of the atmosphere, and a couple of projections for where fronts and low pressure systems are headed this week, I loaded the midday update of the GFS model run (the name of one of the weather models meteorologists use).

Everything looked fine until Day 5.  Thursday.  Coastal storm? Storm gaining incredible strength and whacking New England with heavy rain and snow? What's this?!

Well then I had to look at a couple of other sources.  Luckily this weather model updates four times per day.  Luckily it's not the only one.  Luckily there are dozens of other weather models.

So I looked at the other three updates of this GFS model. None of them had the same projection.  One was close, but the others called for sunny and cool weather!  What's going on here.  I consulted a couple more, and they each had their own opinion.

Hopping onto the National Weather Service discussion, here's what I read (courtesy Burlington, VT office):

"PLEASE NOTE THAT PER HPC AND

INTER-OFFICE COORDINATION WE ARE DISCOUNTING THE OPERATIONAL 12Z

GFS SOLUTION WED NT/THU WHICH SUGGESTS DEEP INLAND CYCLOGENESIS AND

HEAVY RAINFALL ACROSS THE NORTHEAST. THIS SOLUTION HAS NO SUPPORT

FROM THE GEFS OR EURO/EURO ENSEMBLE RUNS OVER THE PAST 24-HOURS AND

IS CONSIDERED AN UNREALISTIC OUTLIER AT THIS TIME."

Alright, so here's a nice visual I put together of what I was looking at.  In the picture below you'll see four images.  Well four images within one picture.  They are all forecasts from the GFS model at the same time.  I stopped the model and took a picture of it at 1 pm Thursday, December 8.

From top left to bottom right, the updates are from (1) late Friday night, (2) early Saturday morning, (3) late Saturday morning, and (4) early Saturday night.  That is when these models were created, run, and published. 

When I came into work and loaded the run that I told you about earlier, that was number 3.  See what I mean?! Look at that thing.  The colors (green and blue) are estimated accumulated precipitation (rain or snow) in a 6-hour period.  The "L" is the projected placement of the surface low. 

Okay breaking this down you can see in the # 1 update high pressure is dominating the east coast with a fair and cool forecast.  The same goes for # 2.  However that changes in an instant with # 3, and # 4 is kind-of in the middle with a storm clipping the coast. 

As a forecaster I look for trends and consistencies.  When forecast models are saying, calm, calm, STORM, calm, I tend to think the outlier can't be trusted that well.  That's what the National Weather Service in Burlington commented on above.

Okay so let's take a look at a Canadian weather model, which is below:

This weather model is for 7 a.m. Thursday, December 8, so fairly close to the other predictions above.  This is similar to the # 4 of the GFS with a coastal storm.  The purples and greens again are accumulated precipitation.

Here's the European model run from Saturday:

This update puts the surface low "L" way out to sea, halfway between Bermuda and the east coast.

So there it is.  When I say 'chance' of snow, this is one of the scenarios I'm dealing with.  The best thing to do here is just wait until it gets closer in time and the forecaster can have a little more confidence.