• Hello, please take a minute to check out our awesome content, contributed by the wonderful members of our community. We hope you'll add your own thoughts and opinions by making a free account!

Wintry 1/9-12 Winter Potential Great Dane or Yorkie

Metwannabe said:
I'd like to see energy round base of the trough and see it go more neutral/negative, to @ILMRoss point with this positive tilt/flatter solution we compromise qpf. But this was an improvement for sure...... haha I haven't had measurable snow in 3 yrs and now I'm getting greedy
Click to expand...

QPF has been trending up on these AIFS runs, so hopefully this will continue going forward…still some time to adjust


Sent from my iPhone using Tapatalk
 
700mb RH with comma head on the Euro

VfPRJkq.gif
 
jrips27 said:
I’m scared for tomorrow lol cause today was about the best day of model trending I can remember (as someone who lives near Atlanta)
Click to expand...
The fight to get snow on the ground from Atlanta to Raleigh is real. The more west to east or ENE slider is the best way to do it. Keeps those nasty warm temps from thumping back to the northwest. Glad that the GFS/GEFS combo is the amped and warmer camp and not the EPS / Euro AI camp - the latter 2 are better scoring / more believable
 
griteater said:
bouncy - is this Euro Graphcast or GFS Graphcast run off Euro data?
Click to expand...
This is the Euro/ECMWF GraphCast.

Google Deepmind created GraphCast using ECMWF data in the paper and in the original training/fine-tuning. ECMWF started running it 2x per day last year.. eventually NCEP started running GraphCast using GFS initial conditions as initialization and called it GFS-GraphCast. I believe the current version that NCEP runs is even fine-tuned for GFS initial conditions. ECMWF runs ECMWF-GraphCast 2x per day and their own Ai model (AIFS) 4x per day. AIFS is based on GraphCast but is a bit more complex (and more accurate).

It doesn't matter a ton because the AI models only use the initial conditions. It's like a "starting point" for the model, and it generates the remainder of the forecast. So the only difference is the hour 0 GFS vs hour 0 ECMWF.. both of which use similar observations networks.
 
bouncycorn said:
This is the Euro/ECMWF GraphCast.

Google Deepmind created GraphCast using ECMWF data in the paper and in the original training/fine-tuning. ECMWF started running it 2x per day last year.. eventually NCEP started running GraphCast using GFS initial conditions as initialization and called it GFS-GraphCast. I believe the current version that NCEP runs is even fine-tuned for GFS initial conditions. ECMWF runs ECMWF-GraphCast 2x per day and their own Ai model (AIFS) 4x per day. AIFS is based on GraphCast but is a bit more complex (and more accurate).

It doesn't matter a ton because the AI models only use the initial conditions. It's like a "starting point" for the model, and it generates the remainder of the forecast. So the only difference is the hour 0 GFS vs hour 0 ECMWF.. both of which use similar observations networks.
Click to expand...

What are the snow totals on that


Sent from my iPhone using Tapatalk
 
The Euro has been steady as a rock during its past several runs. It continues the positive modeling mojo during the past 24 hours or so and things are looking good. I woke up and couldn't get back to sleep so I thought I'd see what happened. I'm going back to bed and hopefully when I wake tomorrow these positive model runs will continue and not just to be a dream when this storm gets here.
 
In the latest RDU discussion they mentioned that a better idea of what will probably happen when the energy that will be our storm arrives on the west coast tonight or early Wednesday morning. Then it can be sampled and included in the model data. RDU seems fairly certain that precipitation will begin as snow Friday night and from there it remains to be seen if and how long the CAD damming will hold out.
 
jrips27 said:
View attachment 160714
Click to expand...
Models are looking great, yes. Things are trending in the right direction and the window for big deviations is starting to close ..

I just can’t shake the feeling that this will turn out like so many other threats, with the r/s line ending up like 50 miles north of model projections.::
 
bouncycorn said:
Models are looking great, yes. Things are trending in the right direction and the window for big deviations is starting to close ..

I just can’t shake the feeling that this will turn out like so many other threats, with the r/s line ending up like 50 miles north of model projections.::
Click to expand...

I’m in Dallas as well. If it looks close I’m going to snag a hotel a little farther north. Calhoun, Ringgold, Blue Ridge, Chattanooga, Huntsville…
 
You must log in or register to reply here.
Back
Top