ATL: MATTHEW - Models

[Bocadude85]

The 18z GFS continues the ever so gradual westward trends from over the past few GFS runs.

The gap between the Euro and GFS is slowly closing. The timing is still a little different but the distance gap has definitely been closing with each new run.

Wouldn't be surprised to see the 00z Euro overnight trend even more eastward while the 00z GFS trends more westward.

I'm not sure the 12Z EURO really shifted east...more north then east of the 0Z run. Tonight's 0Z run could very well shift east however. It will be interesting to see if the 0Z GFS continues the trend of small westward shifts. I'd say the threat to Jamaica/Cuba/Bahamas is quickly increasing.

[p1nheadlarry]

Yikes at the Euro ensembles! Most in the Gulf and showing major hurricanes...

How accurate are ensembles?

Depends how clustered they are over time. Ensembles basically run with slight tweaks to data for the massive equations that are used, which yields different results in track and intensity and etcetera of the atmosphere (not just our low pressure system, but the jet, bermuda high, and so on). The more clustered they are, the less likely that these tweaks change the outcome of the systems involved and is a more confident forecast. Catch-22 is, these models write the equations for these systems slightly different from each other, which is why you see them interpret our weather differently.

[TheProfessor]

Question is if the GFS predicts a major Hurricane, of course this far out is very flexible but I am rather confused. On one hand, the MB is around 950 upon landfall, Cat 3. On other GFS models, it's 930ish, Cat 4.

So I think you're asking why on some images the GFS is showing a 950 mb storm and on other images it shows a 930is mb storm. If this is your question and your were looking at the same hours for each pressure read you likely see this difference because the map of the 930ish storm is a higher resolution than the map of the 950is storm.

[ConvergenceZone]

Wow just hooks it right into the northeast. I'm practically laughing at what the GFS is showing

Why?

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

Nonetheless, it's the trend westward that shouldn't be ignored. First, the Euro showing it dangerously close to Miami, and now this? I'd say we better start paying attention just in case

[meriland23]

Interesting that this run is almost right up into NY as a Maj. I mean, I didn't expect that big of a shift from the GFS in six hours. That is impressively scary. All that being said, we shall see what 00z GFS shows, if it is still in agreement with 18z or more-so, then I would be a little worried given the fact it is shifting from a fish into a direct hit the more time goes on. It is worth a close eye on.

[CalmBeforeStorm]

Wow just hooks it right into the northeast. I'm practically laughing at what the GFS is showing

Why?

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

I agree that anything out that far is fantasy land but I don't agree that a synoptic setup at that range is impossible. Why is such a setup laughable synoptic ally? What's laughable about it?

[p1nheadlarry]

Why?

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

Nonetheless, it's the trend westward that shouldn't be ignored. First, the Euro showing it dangerously close to Miami, and now this? I'd say we better start paying attention just in case

Though a ~920 mb hurricane hitting near OBX/VA Beach is quite crazy. It's just another model run without a closed center to initialize far out in time. The mesoscale/hurricane models especially have trouble interpreting the disturbance and nearby conditions of our ocean/atmosphere without this feat accomplished yet. Remember the HWRF with Hermine? It did fairly well, once it was a TD/TS that is.

[sma10]

Okay. Regarding the 18Z GFS...I don't even know where to begin. Prior to it slamming into NYC at 933mb, it spends 3.5 days lumbering up the coast from the Bahamas at 915-925mb, basically 3 days of Cat 4/5 strength. The storm surge would be unimaginable, the destruction beyond catastrophic.

Will this happen? Not a chance.

But it would make a HELL of a great film.

[toad strangler]

Why?

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

I agree that anything out that far is fantasy land but I don't agree that a synoptic setup at that range is impossible. Why is such a setup laughable synoptic ally? What's laughable about it?

This gigantic cyclone basically goes POOF after that crazy hook through Manhattan. I honestly chuckled. The entire cyclone just evaporates on the double.

#bogus

[meriland23]

Wow just hooks it right into the northeast. I'm practically laughing at what the GFS is showing

It's not funny though, and there is a reason it is hooking NW in this run. There have been many instances where the GFS/EURO were initially correct 200+ hours out before changing it into the wrong outcome. So despite how unusual the run is, pay attention to the fact it did change in a way that somehow will likely affect us.

[CalmBeforeStorm]

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

Nonetheless, it's the trend westward that shouldn't be ignored. First, the Euro showing it dangerously close to Miami, and now this? I'd say we better start paying attention just in case

Though a ~920 mb hurricane hitting near OBX/VA Beach is quite crazy. It's just another model run without a closed center to initialize far out in time. The mesoscale/hurricane models especially have trouble interpreting the disturbance and nearby conditions of our ocean/atmosphere without this feat accomplished yet. Remember the HWRF with Hermine? It did fairly well, once it was a TD/TS that is.

I really don't think the models have much trouble with such well defined invests as 97l.

[meriland23]

Okay. Regarding the 18Z GFS...I don't even know where to begin. Prior to it slamming into NYC at 933mb, it spends 3.5 days lumbering up the coast from the Bahamas at 915-925mb, basically 3 days of Cat 4/5 strength. The storm surge would be unimaginable, the destruction beyond catastrophic.

Will this happen? Not a chance.

But it would make a HELL of a great film.

I am curious as to what makes you so sure it wont happen? Like I would like to know why without a doubt you 'know' it wont.

[Blown Away]

The consistency of the models showing a major hurricane run after run and day after day is unbelievable given what we have seen for years... Wow, today has been a very scary modeling day... Fortunately the focus changes daily, yesterday I thought Puerto Rico was in play...

[CalmBeforeStorm]

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

I agree that anything out that far is fantasy land but I don't agree that a synoptic setup at that range is impossible. Why is such a setup laughable synoptic ally? What's laughable about it?

This gigantic cyclone basically goes POOF after that crazy hook through Manhattan. I honestly chuckled. The entire cyclone just evaporates on the double.

#bogus

It's over land, moving quickly, colder temperatures. Many reasons for quick deintensification. Bottom line is I don't discount any model solutions because they are unusual because the records are full of unusual behaving storms.

[p1nheadlarry]

Nonetheless, it's the trend westward that shouldn't be ignored. First, the Euro showing it dangerously close to Miami, and now this? I'd say we better start paying attention just in case

Though a ~920 mb hurricane hitting near OBX/VA Beach is quite crazy. It's just another model run without a closed center to initialize far out in time. The mesoscale/hurricane models especially have trouble interpreting the disturbance and nearby conditions of our ocean/atmosphere without this feat accomplished yet. Remember the HWRF with Hermine? It did fairly well, once it was a TD/TS that is.

I really don't think the models have much trouble with such well defined invests as 97l.

They sure don't in the Pacific. I guess Gaston was forecast pretty well too.

[TheDreamTraveler]

Wow just hooks it right into the northeast. I'm practically laughing at what the GFS is showing

It's not funny though, and there is a reason it is hooking NW in this run. There have been many instances where the GFS/EURO were initially correct 200+ hours out before changing it into the wrong outcome. So despite how unusual the run is, pay attention to the fact it did change in a way that somehow will likely affect us.

Oh don't get me wrong, if that run were to verify I'd be afraid if a cat 3/4 was barreling into the northeast. It's just one run out of many that actually showed this. And there's way too many variables on the table that could change the outcome. Even just a few days from now the entire solution might be way different. Say if it plows into Florida like the Euro is alluding to. Now if I start seeing this on the GFS on the 0z and 6z runs? Then I'll definitely take it a little more seriously.

[wxmann_91]

Because it's complete nonsense, almost like the GFS wants to win a popularity contest by showing something dramatic out in fantasy land. The GFS skill at that range is a little better than a blindfolded kid throwing darts at a map.

I agree that anything out that far is fantasy land but I don't agree that a synoptic setup at that range is impossible. Why is such a setup laughable synoptic ally? What's laughable about it?

This gigantic cyclone basically goes POOF after that crazy hook through Manhattan. I honestly chuckled. The entire cyclone just evaporates on the double.

#bogus

It actually phases with the midlatitude cyclone and then occludes.

Not bogus physically per say (besides the whole staying at Cat 3/4 strength up the coast), but this solution is incredibly low-probability.

[gatorcane]

GFS ensembles continue to trend toward the ECMWF with even more turning WNW into Florida in the Bahamas:

[sma10]

Okay. Regarding the 18Z GFS...I don't even know where to begin. Prior to it slamming into NYC at 933mb, it spends 3.5 days lumbering up the coast from the Bahamas at 915-925mb, basically 3 days of Cat 4/5 strength. The storm surge would be unimaginable, the destruction beyond catastrophic.

Will this happen? Not a chance.

But it would make a HELL of a great film.

I am curious as to what makes you so sure it wont happen? Like I would like to know why without a doubt you 'know' it wont.

Well, I may have overstated my case with "not a chance". Obviously, there is no one in the world who knows what will happen with 97L, myself included. Like many on this board, I have seen literally hundreds of modeled catastrophes over the years from 200+ hours out, and they just do not play out as modeled. But this particular one made me chuckle because it is perhaps THE VERY WORST case scenario I have ever personally seen in a model. In fact, it's so outrageous, that it looks like a Hollywood script for the worst disaster ever.

Hey, who knows - maybe this is the one. I suppose there will come a day that NYC will be hit directly by a Category 4 behemoth, that has been driving Category 5 whipped waves for 3 days ahead of it. I would just wait another 5-6 days yet before committing to this solution.

[wxmann_91]

Okay. Regarding the 18Z GFS...I don't even know where to begin. Prior to it slamming into NYC at 933mb, it spends 3.5 days lumbering up the coast from the Bahamas at 915-925mb, basically 3 days of Cat 4/5 strength. The storm surge would be unimaginable, the destruction beyond catastrophic.

Will this happen? Not a chance.

But it would make a HELL of a great film.

I am curious as to what makes you so sure it wont happen? Like I would like to know why without a doubt you 'know' it wont.

Well, I may have overstated my case with "not a chance". Obviously, there is no one in the world who knows what will happen with 97L, myself included. Like many on this board, I have seen literally hundreds of modeled catastrophes over the years from 200+ hours out, and they just do not play out as modeled. But this particular one made me chuckle because it is perhaps THE VERY WORST case scenario I have ever personally seen in a model. In fact, it's so outrageous, that it looks like a Hollywood script for the worst disaster ever.

Hey, who knows - maybe this is the one. I suppose there will come a day that NYC will be hit directly by a Category 4 behemoth, that has been driving Category 5 whipped waves for 3 days ahead of it. I would just wait another 5-6 days yet before committing to this solution.

It's pretty crazy how this is almost *literally* the worst possible scenario for NYC that is modeled.

But forget about 5-6 days, this solution won't be there six hours from now.