ATL: IRMA - Models

[MatthewsRevenge]

[Kingarabian]

18z GFS hour 72-102

[xironman]

The upper low by NE is a _lot_ slower moving out.

[AutoPenalti]

If anyone could post ensembles that would great, I don't know how I feel about the GFS this run. Ridge just breaks down too quick.

[SouthFLTropics]

that north turn is key and all it has to do is move west for a couple hours and its inland.. very precarious place to do a north turn like that..

Agreed...angle of approach means everything in this case. With the GFS' propensity for breaking down ridges too soon it doesn't give me a warm and fuzzy feeling.

[Steve]

I wish you guys well in Homestead. Kinda seeming like ground zero or very close to it based on the 12Z model aggregate. Hang in there!

As a Homestead resident, I am not happy about the model trend/consensus this afternoon. Potentially scary days ahead.

Wisely so. It's the place of the worst U.S. hurricane destruction I've ever seen. I'm behind so with the GFS 18z and other 18z models now running, we are 6 hours closer to landfall. Maybe we can get some more tightening so that all of you will be able to plan accordingly within the 90-100 hour window you have. And don't forget, traveling gets tough as the bands start coming, winds increase to TS and Gale force in and bridges close and such. If I'm reading the models correctly, effects appear to begin Saturday (possibly some outer band action Friday night). So the true window for action is probably 72 hours or less.

[GlennOBX]

I kept seeing the "comparison" graphics posts, that showed previous and current runs, comparing what the model was thinking then and is now. Then I noticed the "hours out" numbers were different in the different frames, and couldn't figure out why.

When I finally DID figure it out, I pictured Agent Gibbs from NCIS "Gibbs-slapping" me for that brain cramp.

Admins, feel free to delete this post. Tangentially related to models at best, but I thought it was worth mentioning.

[tolakram]

If anyone could post ensembles that would great, I don't know how I feel about the GFS this run.

They won't be out until after the run is completed.

[Kingarabian]

18z GFS hour 120, landfall:

[OntarioEggplant]

Yeah the eastern bias of the GFS has been concerning considering its performance over Irma's lifetime. I would still lean with the Euro moreso. But no one in Florida should feel out of the woods with just one model run.

[MatthewsRevenge]

[Vdogg]

18z GFS hour 120, landfall:

That's the 4th consecutive model run with a landfall in SC. Hopefully they are paying attention to this.

[Blown Away]

@30 miles E at 120 hours... Not a big change in the model world, but amazing how 30 miles can be the difference...Still NW inland just E of GA/SC border then NW towards Western Carolinas...

[xironman]

The extent of the winds on the NE quadrant at landfall continues to be impressive. Should have storm surge over a large area

[Kingarabian]

18z GFS hours 126->138:

[Vdogg]

Looking at Satellite there is definitely an increased Northerly component to Irma's motion. The eye is passing well north of Puerto Rico. Perhaps this is why GFS was slightly east on this run.

[txwatcher91]

Here’s the thing with Irma. Where that north turn happens will be critical, and despite how good the Euro and other models may be, they don’t have the accuracy at this range to nail down something that 20-40 miles is the difference between landfall or the east coast of Florida. Either way, it looks like Miami and surrounding areas will see winds gusting well over 100mph and I hope people take it seriously. We probably won’t know about the north turn until it actually happens...

[Ken711]

Here’s the thing with Irma. Where that north turn happens will be critical, and despite how good the Euro and other models may be, they don’t have the accuracy at this range to nail down something that 20-40 miles is the difference between landfall or the east coast of Florida. Either way, it looks like Miami and surrounding areas will see winds gusting well over 100mph and I hope people take it seriously. We probably won’t know about the north turn until it actually happens...

For someone like me with no meteorological knowledge, what is the feature that makes the models so confident a turn will occur within that area? Thanks

[Bocadude85]

The GFS has CONSISTENTLY been to far north every run, it's easy to see this by going through today's 6z,12z and 18z initializations and comparing the previous runs initial forecast point.

[Hybridstorm_November2001]

Here’s the thing with Irma. Where that north turn happens will be critical, and despite how good the Euro and other models may be, they don’t have the accuracy at this range to nail down something that 20-40 miles is the difference between landfall or the east coast of Florida. Either way, it looks like Miami and surrounding areas will see winds gusting well over 100mph and I hope people take it seriously. We probably won’t know about the north turn until it actually happens...

For someone like me with no meteorological knowledge, what is the feature that makes the models so confident a turn will occur within that area? Thanks

Part of an upper-level trough currently centered near Minnesota.