GFS Users - BEWARE!

[wxman57]

A note from NCEP:

THE FOLLOWING PBLM WAS APPARENTLY SOLVED EARLIER THURS AFTN
BUT HAS SINCE RECURRED...

*** THERE HAS BEEN NO NESDIS SATELLITE RADIANCES AND SOUNDING
PRODUCTS FROM NOAA 14/15/16/17...POLAR ORBITERS. ADDITIONALLY
...THERE HAS BEEN NO QUIKSCAT DATA SINCE THE NCEP NETWORK CAME
BACK ON LINE YESTERDAY AFTERNOON. <b>THIS LOSS OF SATELLITE DATA WOULD HAVE A MINIMAL IMPACT ON THE ETA...BUT COULD HAVE AN EFFECT ON THE GFS...ESPECIALLY IN THE TROPICS AND SOUTHERN HEMISPHERE SINCE THIS IS PRIMARY DATA SOURCE FOR THOSE DATA SPARSE REGIONS. THERE COULD BE SOME ANOMALIES. *** </b>

[ameriwx2003]

Hmmmmmmmmmm.. thanks for the heads up on that wxman57:):)

[Stormchaser16]

Theres explanations for the models reversing tonight(well the GFS and the Ec at least)

The GFS did not stall Fabian off the coast of Florida, mainly in fact that it doesnt have the front there that it has had in ALL other runs. Somehow it just disspeared. So when the trough comes Fabian is still moving and then is picked up by the trough, it doesnt undercut it this time. Pretty wacked idea, and this is the stuff u have to look for when interpretting models, hence the posts about GFS turns Fabian out to sea should be reconsidered. Since people just seem to post a qpf output and dont ever care to look any deeper then that.

As for the EC, it too is out of wack, mind you it only runs once a day, therefore when it initialized it MISSED the W track that Fabian has developed throughout the day(in otherwords its initialization with Fabian is off). Thus it takes it out to sea.

[Stormsfury]

Theres explanations for the models reversing tonight(well the GFS and the Ec at least)

The GFS did not stall Fabian off the coast of Florida, mainly in fact that it doesnt have the front there that it has had in ALL other runs. Somehow it just disspeared. So when the trough comes Fabian is still moving and then is picked up by the trough, it doesnt undercut it this time. Pretty wacked idea, and this is the stuff u have to look for when interpretting models, hence the posts about GFS turns Fabian out to sea should be reconsidered. Since people just seem to post a qpf output and dont ever care to look any deeper then that.

As for the EC, it too is out of wack, mind you it only runs once a day, therefore when it initialized it MISSED the W track that Fabian has developed throughout the day(in otherwords its initialization with Fabian is off). Thus it takes it out to sea.

Not to mention, the EC was initialized as a tropical depression at 12z.

SF

[Steve H.]

Even if there were no problems with the GFS, there would still be flip - flopping of the models this far out. One should not react to "SE US", "Fish", SE US, fish, etc. That being said the ridging is still the key to the forecast for the long range. We're still a week away from the potential event. BTW, did any other models reaact to the lack of data...and was the Florida then up the coast solution run with bad data??? Thoughts. Cheers!!!

[Stormchaser16]

I doubt the Florida then up coast was run with bad data..... as there are other models that are indicating the same thing.

Fury... that is correct, and makes a huge difference in the potential output.

[Steve H.]

OK, thanks for the response. So the key here really is for the stall to allow the trough moving through the NE and Mid-Atlantic to bypass Fabian, then have the ridge re-build. Wonder what tomorrows runs will show.

[wxman57]

I doubt the Florida then up coast was run with bad data..... as there are other models that are indicating the same thing.

Fury... that is correct, and makes a huge difference in the potential output.

Actually, those runs were made before the recent computer shut-down (due to overheating). The Problem began yesterday morning.