Sunday, December 21, 2014

The Weather Buffet



Who doesn't love a buffet?  All the different choices of food and the big sign that says, "all you can eat".  I can do some real damage!  I'll have a little bit of everything and a whole lot of a few things.  Then to cap it off, there seems to be an endless selection of desserts too.  I'll leave 10 pounds heavier with a loosened belt.

Much can be said about weather forecasting.  Hear me out on this one.  You walk up to the computer model buffet and here's your selection...

European
GFS
Experimental GFS
NAM
WRF
HRRR
RPM
CMC
JMA
UKMET
ADONIS
CMC
CFSv2
DGEX
NOGAPS
BAM
GFDL
RAP
SREF

I know I probably forgot a few there, but there's almost information overload.  Much like that fantastic all you can eat buffet, there's the all you can forecast model buffet.  However, if you want to have longevity in life, or in forecasting, you must choose wisely.  Selecting only one food or model just because you like what it says or how it tastes isn't good for you.   You can have a little fried food, but you really need to eat your veggies too, even if they aren't your favorite.

Out of all those models, the GFS was the only one I could find showing snow for much of central and northern Arkansas Tuesday evening into Wednesday morning.  This is why I was extremely hesitant about using that guidance for my forecast.  Remember, there's no guarantee in weather until after it happens,

I made the call Saturday to only call for a brief change to a few wet snowflakes for northern Arkansas Tuesday evening into Wednesday morning and ignore the extreme solution of the GFS.  This morning, the GFS is backing off BIG TIME.  Yes, there could be snow, but it shows the far north.  While this hasn't even happened yet, I'm glad I did what I did.  No one wants snow more than me, especially around Christmas, but I would much rather get it right.

The following maps are from weatherbell.com and shows you the progression of the GFS over the past few runs.

The Saturday morning 12Z run of the GFS really cranked up the surface low and the model was hitting on a strong piece of energy wrapping into the trough to produce widespread snow on the back side of the system for much of central and northern Arkansas.

Then the 18Z run came out on Saturday and continued to show basically the same thing Tuesday night into Wednesday morning.

The 00Z run Saturday night went crazy for snow


Then the 06Z run came in and severerly backed off on the threat in central Arkansas

And the 12Z Sunday morning run shows just a few flakes up north as it ends Wednesday morning and I'm not even sure that will happen.
In summary, you can clearly see the trend with the only model forecasting snow Tuesday night into Wednesday.  Yes, I'm glad I never went on the air with the forecast calling for all that snow.

I have also learned to "never say never to Arkansas weather" and while I don't think the large majority of us will get any snow, it's still worth watching each run of all the models and I'll keep you updated.  I love snow, but I like getting it right better.

No comments: