ATL: MATTHEW - Models

[gatorcane]

12Z UKMET:

[La Sirena]

That certainly doesn't look good for SFLA or the Keys. It never crossed my mind that their model path was even an option considering the other trends. That's not a usual occurrence for the area this time of year.

[sma10]

12Z UKMET:

Besides the obvious, a track like this would likely lead to substantial strengthening once a bit more removed from the Cuban coast. The key is if the UKmet can produce two similar runs in a row.

[Dean4Storms]

And if this doesn't slow down it is going to be past its consensus area of the North turn which will indicate the Ridge was stronger and stretching further west than the models were forecasting!

[maxx9512]

This is why NHC should be following the dynamical models and NOT the LGE

TIME (HR) 0 6 12 18 24 36 48 60 72 84 96 108 120
V (KT) NO LAND 65 66 66 67 68 67 69 68 74 82 81 83 82
V (KT) LAND 65 66 66 67 68 67 69 68 74 82 81 65 66
V (KT) LGEM 65 67 68 68 67 64 62 61 63 67 72 61 65
Storm Type TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP

SHEAR (KT) 19 19 20 15 13 11 6 10 2 3 2 10 11
SHEAR ADJ (KT) 6 4 1 4 4 1 7 0 6 1 0 3 7

This is really only going to intensify by 10 kts while under < 10 kts of shear for 48 hours over the Caribbean? Really?

Stick with dynamical models. Statistical models tend to flop badly during rapid intensity changes

Is there a short answer as to the difference between the dynamical models and the statistical models?

[PTrackerLA]

12Z UKMET:

Besides the obvious, a track like this would likely lead to substantial strengthening once a bit more removed from the Cuban coast. The key is if the UKmet can produce two similar runs in a row.

Good luck with that. It's just slightly better than the CMC in my opinion, but some still swear by it.

Anywho, some very interesting trends with the midday runs. Everyone on the East coast needs to pay close attention this weekend.

[LarryWx]

The day 6 to day 7 12Z UKMET move is between WNW and NW and it gets to near 21.5N, 77W, at hour 168. So, not only is it the furthest west of the major 12Z operational dynamic models at the latitude of Cuba, it has a NWish heading meaning it would be even further west on later maps if this had gone further. Also, note that this is the slowest of all 12Z runs (furthest south) with the Euro being only a little slower.
This is obviously an ominous run for the CONUS. At the same time, the UKMET has been jumping around pretty wildly with the prior run much further east. So, it could be a bogus run? Who knows?

[Alyono]

This is why NHC should be following the dynamical models and NOT the LGE

TIME (HR) 0 6 12 18 24 36 48 60 72 84 96 108 120
V (KT) NO LAND 65 66 66 67 68 67 69 68 74 82 81 83 82
V (KT) LAND 65 66 66 67 68 67 69 68 74 82 81 65 66
V (KT) LGEM 65 67 68 68 67 64 62 61 63 67 72 61 65
Storm Type TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP TROP

SHEAR (KT) 19 19 20 15 13 11 6 10 2 3 2 10 11
SHEAR ADJ (KT) 6 4 1 4 4 1 7 0 6 1 0 3 7

This is really only going to intensify by 10 kts while under < 10 kts of shear for 48 hours over the Caribbean? Really?

Stick with dynamical models. Statistical models tend to flop badly during rapid intensity changes

Is there a short answer as to the difference between the dynamical models and the statistical models?

Statistical models simply use statistics to make their forecast. Dynamical models numerically approximate the equations of motion to arrive at a solution

[Stormcenter]

Just the East coast? I would say everyone should keep an eye on Matthew. IMO

12Z UKMET:

Besides the obvious, a track like this would likely lead to substantial strengthening once a bit more removed from the Cuban coast. The key is if the UKmet can produce two similar runs in a row.

Good luck with that. It's just slightly better than the CMC in my opinion, but some still swear by it.

Anywho, some very interesting trends with the midday runs. Everyone on the East coast needs to pay close attention this weekend.

[AtlanticWind]

Would expect shift a little west by nhc at 5 looking at models

[gatorcane]

12Z Euro ensembles at hour 168 still indicate not a lot of certainty especially as far as Florida impacts. It seems there are a number of ensembles showing a WNW turn north of Eastern Cuba around hour 144 generally more W of the Euro operational with others recurving like the GFS.

[caneseddy]

 https://twitter.com/watkinstrack/status/781575749689937924

[wxman57]

Is there a short answer as to the difference between the dynamical models and the statistical models?

Statistical models don't look at what's actually happening in advance of the storm.

[OntarioEggplant]

Still a significant number of Euro ensembles ending up in the Gulf

[hurricanedude]

What is the purpose of the xtrap model....always sraight line

[caneseddy]

Excerpt from 5:00 pm discussion

One should not focus on the details of the track at 72 to
120 hours due to the inherent forecast uncertainties. It should be
noted that the NOAA G-IV aircraft is currently doing a synoptic
surveillance mission to provide data in the environment of Matthew
for initializing the numerical models. It is hoped that these
additional data will improve the accuracy of tonight's model
runs.

[emeraldislenc]

Do you think they may be a west shift at 5?

[CrazyC83]

Is there any site available that shows projected steering currents at different levels? I know CIMSS has it for right now. That would help a lot in diagnosing potential differences (a big issue with Joaquin).

[gatorcane]

Animation of 12Z Euro ensembles from hour 48 to 192:

[ronyan]

What is the purpose of the xtrap model....always sraight line

It's not a model, just the extrapolation of current movement.