Tom Cardy said:
Agasaurus Tex said:
Is GFS reliable for cold weather forecasts? I realize this far out forecasts can be sketchy but for tropical storm forecasts one model run may have landfall in Florida and and the next run in Mexico.
Much like hurricane season, the best ideas come from using ensembles to see the possibilities and look for trends in the long range of 5+ days. Operational/deterministic models (GFS, euro, icon, cmc, etc) are usually better in the mid-short range. Think 5 days and in, but with better accuracy as you get closer. There's another set of models that deal exclusively in short range that have higher resolution and typically are pretty good with precipitation.
For winter specifically, no model is always the "best" for Texas weather. The air masses and related systems that give chances for precipitation are pretty complex and timing varying by a couple of hours can make a big difference.
Well said! That's why I like Thursday, possibly Wednesday as a day to start talking about sensible weather in regards to temps. As for precip, I look at the hi-res models late week to get a good idea of what they want to do.
NAM, HRRR, WRF, RAP (RAP over did it the most) did a pretty decent job with the North Texas snow last week and that was a much more complex set up.
These true Arctic setups are a little easier to sniff out potential winter weather setups and you can tell right away in the mid range and short range if the output is legitimate or not