ATL: MATTHEW - Models

[KWT]

Yes the 12z GFS does look quite a bit further north than previous runs. Given we now have a center we will get a much better handle on where it is going.

With that being said, lets wait and see what the rest of the run does!

[chris_fit]

Like like a Southern component at 42 hours

[OntarioEggplant]

Uh, SW dip through 42 hours?

[ronyan]

It's very slightly further north guys through 36 hrs.

[chaser1]

12z GFS running now

Is it just me or does it appear that the GFS initialized a hair too far north. Might be splitting hairs here but just looks that way to me.

I hadn't taken a close look to see yet so couldn't say but a couple times where i've seen that happen for a given model run, that same input also seemed to represent an incorrectly initialized forward motion as well. No doubt that accurate direction, speed of forward motion, and initialized COC might together impact that particular run's overall accuracy. That all said, it's possible that this 12Z GFS run might simply appear as a continuation of its past couple of runs. Just means that the 18Z or 0Z tonight might begin to appear slightly different, and possibly more accurately initialized?

[SouthFLTropics]

Yes the 12z GFS does look quite a bit further north than previous runs. Given we now have a center we will get a much better handle on where it is going.

With that being said, lets wait and see what the rest of the run does!

I believe the 12z would have been initialized using the best track and not the actual center that recon found. They wouldn't have had time to use the recon finding I don't think.

[HurrMark]

I don't see this initialized too far north at all. It is virtually initialized at the same spot as 6Z.

[gatorcane]

12Z GFS back to same 06Z position at 48 hours because it takes a SW dip.

[KWT]

Yeah GFS did start too far north, I'd estimate about 1 degree too far north, about the right speed though.

[Hamanard]

12Z GFS back to same 06Z position at 48 hours because it takes a SW dip.

Stronger as well

[OntarioEggplant]

Yes the 12z GFS does look quite a bit further north than previous runs. Given we now have a center we will get a much better handle on where it is going.

With that being said, lets wait and see what the rest of the run does!

I believe the 12z would have been initialized using the best track and not the actual center that recon found. They wouldn't have had time to use the recon finding I don't think.

The correct recon center should be input for the 18Z run.

[tarheelprogrammer]

Slightly further NE of 06z run at hour 54.

See here:
http://www.tropicaltidbits.com/analysis/models/?model=gfs&region=atl&pkg=z500a&runtime=2016092812&fh=54&xpos=0&ypos=200

[Alyono]

I don't see this initialized too far north at all. It is virtually initialized at the same spot as 6Z.

There have been recent studies that have shown relocating a model vortex to the NHC initial position DEGRADES the forecast. It is one of the reasons why Joaquin busted by the MU last year. Had they not relocated the vortex, it would have shown an out to sea track

[SouthFLTropics]

The track differences on 12z versus 06z so far are probably negligible. I'm more interested in watching the 500mb pattern evolving. At 60 hours it looks like the ridge to the north may be expanding to the west a bit more than it did on 06z. Time will tell. Look for the turn around the 90 to 100 hour time frame.

[srva80]

looks a tad slower this run as well

[Dean4Storms]

This run might catch Haiti, looks like the turn will be a shade further East.

[MississippiWx]

GFS a good bit north and east of previous run. Last two runs have done this...eastward trend so far today from GFS. Doesn't mean the end result will be east, but certainly implies it.

[PTrackerLA]

Yep looks like the GFS west trend might have ended. Wouldn't be surprised to see this run miss the EC we will know soon.

[OntarioEggplant]

GFS a good bit north and east of previous run. Last two runs have done this...eastward trend so far today from GFS. Doesn't mean the end result will be east, but certainly implies it.

The "eastward trend" is both due to an incorrect initialization too far north and the slower movement. Otherwise this would be nearly identical to 00z.

[Alyono]

GFS a good bit north and east of previous run. Last two runs have done this...eastward trend so far today from GFS. Doesn't mean the end result will be east, but certainly implies it.

The "eastward trend" is both due to an incorrect initialization too far north and the slower movement. Otherwise this would be nearly identical to 00z.

see my post before. Do NOT touch the vortex position, you only mess everything up if you shift the vortex to the best track position