ATL: HARVEY - Models

[Steve]

Sort of. I edited the post when I got the rest of the run. Essentially coastal effects from Savannah to the Outer Banks at the 13-15 day timeframe. I think the general low resolution of those runs showing 960s and 970s would be too high. A system down there in the 20s of August is more likely 950s or 940s. That's got Cat 3/4 written all over it. Luckily other than the Western Basin probably being ripe, that's a long way off. Unluckily is that the GFS is biased East that far out. Anything can happen, but any farther West it's hello Florida or the Gulf Coast. I'm hoping for a recurve obviously.

Oh and right directly along the coast where Joaquin was probably a couple hundred miles farther east.

[RL3AO]

I've started added model run GIFs to the first post of the thread. We'll try to update it once or twice per day during notable systems. Hopefully people can use it to catch up on model trends.

[gatorcane]

That 18Z GFS run is eye-popping looking at the long-wave pattern in the medium to long-range. I'd say that looks more like an Andrew-seque setup with the west to WSW turn east of Bahamas. Yes it turns north right before impacting Florida but that is like 300 hours out. Just yesterday the GFS had a huge trough over Eastern North America instead now look below..not to mention Andrew in 1992 was about in that location on that date. At least we are pretty sure the GFS is probably not getting the initial genesis correct which would really influence the track downstream and the long-range is basically a crap-shoot.

[gatorcane]

Looks like the NAVGEM has shifted more north:

[Steve]

Big block on that 500mb run Gator. It could go north but GFS was almost looking like it wanted to get this caught at the end of the run by another reinforcing high coming down. That's a hook and back west. Again, no chance this far out. But it's a run that should wake people up. It's not every day EC, GFS and CMC are so close together with a system that hasnt developed from this time frame. No teleconnections in the WPAC to get hints from as the only system out that way moves N-NNW before recurving. It's close to Wake Island so very far to the east of Asia. It sort of argues for a recurve, but there is high pressure to the west and northwest of there.

[Digital-TC-Chaser]

GFS has a tendency to over amplify troughs, ridges, hpa. Its just par for the course
with the model.

[Kingarabian]

Now let's see if the Euro holds on to it @ 00z. Past 2-3 waves it would show development for 2 runs and then drop them.

[RL3AO]

The EPS from Ventrice. Seems to be a split of western wave (Carribean) vs eastern wave (N of Caribbean). It's going to be an interesting week or two.

[RL3AO]

I'm wondering/suspecting the NHC will call 91L whatever wave becomes dominant.

[cycloneye]

I'm wondering/suspecting the NHC will call 91L whatever wave becomes dominant.

Yes.They put the TWO x more to the west now SW of CV islands.

[Sciencerocks]

If I remember right the Gfs has done exactly this with Don, TD4 and was doing it with Gert. All westward threats to the United states at one time and very powerful hurricanes.

I no longer trust the Gfs and I've watched it since it was the AVN in the late 1990's. What in the good god did they do to this model?

[RL3AO]

If I remember right the Gfs has done exactly this with Don, TD4 and was doing it with Gert. All westward threats to the United states at one time and very powerful hurricanes.

I no longer trust the Gfs and I've watched it since it was the AVN in the late 1990's. What in the good god did they do to this model?

They did well to eliminate some of the model-canes. However, something is seriously off with how it creates convection in the deep tropics in high RH environments. I'm not sure if it's an Atlantic only issue (probably not). I don't have much trust in the GFS genesis forecasts right now. We'll see how it does once a system forms in the Atlantic (I suspect it will do fine).

[Sciencerocks]

This will probably in all llikelyhood have faster development compared to Gert but same general path out to sea.

[Steve]

It's heavily weighted Gulf bound there. Hopefully like the GFS/GPS is biased for Eastern troughs, EPS like ECMWF is biased toward the south. That's not good news for the US or Mexico.

[blp]

The 18z HWRF from Gert is showing a very similar solution to the Euro with the SW vorticity winning out.

http://moe.met.fsu.edu/cgi-bin/hwrftc2.cgi?time=2017081318-eight08l&field=Sea+Level+Pressure&hour=Animation

The Hmon is similar.
http://moe.met.fsu.edu/cgi-bin/hmontc2.cgi?time=2017081318-eight08l&field=Sea+Level+Pressure&hour=Animation

Looks like GFS is on its own with the NE vorticity.

[cycloneye]

I'm wondering/suspecting the NHC will call 91L whatever wave becomes dominant.

Another sign they will stick with 91L is the 00z Best Track is delayed and I suspect they are adjusting it more west.

[blp]

I agree I think they will go with the adjustment. I think we should shut this thread down and continue posting on the 91l thread since the content is going to get strung out and confusing between the two threads.

[otowntiger]

The EPS from Ventrice. Seems to be a split of western wave (Carribean) vs eastern wave (N of Caribbean). It's going to be an interesting week or two.

Looks like if it takes the northern route, 90% of those models recurve it harmlessly out to sea.

[Alyono]

The 18z HWRF from Gert is showing a very similar solution to the Euro with the SW vorticity winning out.

http://moe.met.fsu.edu/cgi-bin/hwrftc2.cgi?time=2017081318-eight08l&field=Sea+Level+Pressure&hour=Animation

The Hmon is similar.
http://moe.met.fsu.edu/cgi-bin/hmontc2.cgi?time=2017081318-eight08l&field=Sea+Level+Pressure&hour=Animation

Looks like GFS is on its own with the NE vorticity.

cannot use the parent domain of a nested grid model. The parameterization gets beyond messed up

[tolakram]

If they do we can merge this thread in with 91L.