ATL: DORIAN - Models

[HDGator]

The most important take away from the latest ICON is that it is finally giving up on its idea of landfall in Dade County and is trending towards the GFS and Euro. That model may not be so good after all, I guess it got lucky when it forecasted Dorian to miss P.R.

Jupiter to Dade is what 50 miles? Not a huge variation over three days of track.

Maybe closer to 100

Jupiter to Dade County line is 65 miles. Jupiter to Miami city center is 80 miles.

[MrJames]

GFS is in line with 18z so far at 18hrs out.

[MrJames]

Legacy GFS is starting to move NW at 18hrs.

[HurricaneFrances04]

Coming in further NE through 42 hours. May be a betterish run for the Bahamas.

[sma10]

GFS picking up quite a bit of latitude tomorrow and tomorrow evening

[BobHarlem]

Looks like the GFS 0z wasn't initialized with the latest recon data (948mb) vs what's there in the model

[Aric Dunn]

well GFS just gives up on any west motion.. things will get complicated if dorian keeps going west.. lol

[beachman80]

Feel GFS is going to steer clear of Florida altogether this run. Already going with higher latitude. Guess that high isn't so impressive.

[SouthDadeFish]

The trough over the eastern US has gotten deeper at the 2-day period over the last few GFS runs. This allows the weakness for Dorian to gain latitude.

[Blinhart]

quick question...how accurate is the ICON in general. i know its a newer model, but how good is it in terms of tropical meteorology

It does not seem like it's been performing well with Dorian, it's been well to the south of the other models, at one point going straight west towards Cuba.

Can't really say that about Dorian that much until after we see where he goes. I would love to see the average difference overall for the German Model (ICON) (gonna start calling it the German Model because the NHC uses a different ICON).

Since it is a new model, I can't really tell you how accurate it is, just know it is one of the most consistent ones.

[HurricaneFrances04]

0Z GFS may be very close to the 18Z NAVGEM.

[Meteorcane]

Looks like the GFS 0z wasn't initialized with the latest recon data (948mb) vs what's there in the model

That's not unusual since the GFS is a global-scale (e.g. lower resolution model). Dorian is very small with a very sharp pressure gradient which cannot be resolved by a model like the GFS. This is a benefit some of the higher-res (e.g. HWRF) have, although those models can run into many other issues with Tropical Cyclones.

[NDG]

Looks like that on this latest GFS run it will miss FL to the east.

[AutoPenalti]

GFS says no west motion.

[AtlanticWind]

The trough over the eastern US has gotten deeper at the 2-day period over the last few GFS runs. This allows the weakness for Dorian to gain latitude.

I am feeling a little better in S Florida (not quite out of the woods)
Still worried for up the coast

[BobHarlem]

0z GFS is much more kind to the Bahamas.

[MrJames]

Legacy GFS moving due north 66-72hrs.

[HurricaneFrances04]

What a huge shift in 24 hours.

[beachman80]

The east trend seems to be continuing here. This new GFS run isn't even going to sniff Florida.

[birddogsc]

Where did that huge L over the upper midwest come from at 78 hrs on the GFS?