ATL: HARVEY - Models

[Alyono]

new EC is saying Harvery will be a repeat of...


Harvey. Track very similar to the 2011 version

[tolakram]

I took the post to mean MDR, or the deep tropics, not no development at all. Both the big models have shown MDR development only to back off, dissipate, and then pick up development clues days later. IMO.

[Alyono]

dissipated

I stand by my statement of nothing imminent in the tropical Atlantic

Key word...imminent. Careful about dismissing this area of disturbed weather completely. Ironically today marks the Anniversary of a twave that trekked across the Atlantic and Caribbean only to quickly organize on August 13, 1932 near the Yucatan and make landfall on August 14th at Freeport, TX as a Category 4 Major Hurricane. I will go with climatology over any one run of a dynamical computer model any day.

I only meant Atlantic. This could develop in the Caribbean or the BOC if it gains enough latitude to make the BOC

[tolakram]

[srainhoutx]

I took the post to mean MDR, or the deep tropics, not no development at all. Both the big models have shown MDR development only to back off, dissipate, and then pick up development clues days later. IMO.

As did I. But you and I both know many folks live and die by the models... This disturbance should not be dismissed or written off in my humble opinion.

[RL3AO]

I'm curious if we get a weakening trend with the EPS.

[tolakram]

216h is well inland on a due west path. Interesting change, to say the least.

What did the GFS pick up on that now the euro sees?

[Hurricaneman]

216h is well inland on a due west path. Interesting change, to say the least.

What did the GFS pick up on that now the euro sees?

I'm thinking SAL intrusion

[drezee]

No model is Gospel...48 hours ago Euro had Gert as a Tropical Depression right now...

[otowntiger]

dissipated

I stand by my statement of nothing imminent in the tropical Atlantic

Key word...imminent. Careful about dismissing this area of disturbed weather completely. Ironically today marks the Anniversary of a twave that trekked across the Atlantic and Caribbean only to quickly organize on August 13, 1932 near the Yucatan and make landfall on August 14th at Freeport, TX as a Category 4 Major Hurricane. I will go with climatology over any one run of a dynamical computer model any day.

I only meant Atlantic. This could develop in the Caribbean or the BOC if it gains enough latitude to make the BOC

So it sounds like you think the ridge will stay pretty solid and make this a low rider? wxman57 said earlier that he didn't see the ridge staying place that long and said that he thought it would be turning toward the N. Gulf Coast or possible threaten SE US coast. Difference of opinion or are you guys agreeing with different models or both?

[Michele B]

First of all, I'm NO expert, even tho having lived in FL most of my life...

I've said it before, I'll say it again, "Garbage in, garbage out" applies here.

The models are ONLY as good as those folks who program them! Always keep in mind, even with all their "knowledge," they only program in the elements, or trends THEY think are relevant.

I remember two or three years in a row, on the WC, it was Dave Schwartz - gone now - who kept repeating, "This storm is going to turn north....out to sea...." Even as we were watching it move ever closer to the FL shoreline or further up the Eastern seaboard!! We laughed at him!

All he was doing was repeating the "party line" that had been fed the pro-mets at WC - all based on info that had been FED into a computer somewhere so it would "predict" what the storm would do!!

The point is, I take the "models" with a grain of salt based on my *own* experiences with them over the years. Don't misunderstand! I'm grateful for the advances that have been made in predicting storms and their intensity and their paths, but always remember that the computers that spit out the models have been programmed with SOMEONE's bias.

What Dave Shwartz was echoing was the beliefs from the last decade or two before him (80's or 90's, I think) in which the Bermuda high had receded and ALL storms approaching the Eastern coastline *DID* in fact, curve back out to sea! But then close to the millenium,it seemed that the large, powerful Bermuda high returned and kept storms from tracking northward so early on. So a LOT of predictions of storms heading out to sea turned out to be wrong, but they kept at it for several years...I want to say at least 3 or 4, before they had to admit the storms were coming ashore, or at least were not missing the mainland at all.

Anyway, long post to simply say, I don't think it's worth arguing about any models' predictions when I'm not sure whose bias has been used to write the programming that is doing the modeling. That's kind of where I am in "trusting" one model over another.

My own opinions on the matter, not trying to be argumentative.

[Kingarabian]

First of all, I'm NO expert, even tho having lived in FL most of my life...

I've said it before, I'll say it again, "Garbage in, garbage out" applies here.

The models are ONLY as good as those folks who program them! Always keep in mind, even with all their "knowledge," they only program in the elements, or trends THEY think are relevant.

I remember two or three years in a row, on the WC, it was Dave Schwartz - gone now - who kept repeating, "This storm is going to turn north....out to sea...." Even as we were watching it move ever closer to the FL shoreline or further up the Eastern seaboard!! We laughed at him!

All he was doing was repeating the "party line" that had been fed the pro-mets at WC - all based on info that had been FED into a computer somewhere so it would "predict" what the storm would do!!

The point is, I take the "models" with a grain of salt based on my *own* experiences with them over the years. Don't misunderstand! I'm grateful for the advances that have been made in predicting storms and their intensity and their paths, but always remember that the computers that spit out the models have been programmed with SOMEONE's bias.

What Dave Shwartz was echoing was the beliefs from the last decade or two before him (80's or 90's, I think) in which the Bermuda high had receded and ALL storms approaching the Eastern coastline *DID* in fact, curve back out to sea! But then close to the millenium,it seemed that the large, powerful Bermuda high returned and kept storms from tracking northward so early on. So a LOT of predictions of storms heading out to sea turned out to be wrong, but they kept at it for several years...I want to say at least 3 or 4, before they had to admit the storms were coming ashore, or at least were not missing the mainland at all.

Anyway, long post to simply say, I don't think it's worth arguing about any models' predictions when I'm not sure whose bias has been used to write the programming that is doing the modeling. That's kind of where I am in "trusting" one model over another.

My own opinions on the matter, not trying to be argumentative.

I respect your opinion but that is simply not how computer models work.

[tolakram]

First of all

...

My own opinions on the matter, not trying to be argumentative.

They are physics models and so complex that it's not easy to understand biases, I suspect even if you are on the team that programs them. I do think your statement is valid, as far as not understanding biases, as long as the poor programmers are left out of it.

Anyway, this is the models thread and most of us know what to expect. We look for trends and past performance .. like the fact that both the GFS and Euro predicted the location of Gert fairly well, they just missed the development until a few days ago, the euro picking up on it first but only after dropping it almost completely.

Every year is a little different and sometimes model weaknesses are known until we experience a particularly challenging year. IMO anyway.

[Michele B]

Just a quick reply to say, "I respect YOUR opinion, too!"




If you know how computer models are built into a computer program, I would be glad to be educated! i was just ASSUMING it was based on collections of previous data programmed in....

[Alyono]

EC has this as a TS in the PACIFIC in 10 days

[Kingarabian]

EC has this as a TS in the PACIFIC in 10 days

That's what the EPS control was showing in the 00z run.

[gatorcane]

The ECMWF builds the Bermuda High and expands it west over Florida and the Gulf in tandem with this invest moving through the Caribbean and into Central America (animated image of 500MB heights from hour 96 to 192 below). Just a couple of days ago the long-range models showed no Bermuda High in the long-range. Particularly this year with the upgraded GFS and ECMWF, I really am not giving the models much weight beyond about 5 days and certainly not beyond a week.

[AtlanticWind]

From what I've seen over the years, models when it comes to genesis is only as good as their last run.

Just less than two days ago the gfs had this as a depression and intensifying by this morning. It was wrong.
Now it shows no development of this system at all. May be right or may be wrong.

The Trend in the models are definitely away from development and now we just wait and see what the 91L does
and if the trend in the models continue.

Also wouldn't put much stock in model genesis past 3 to 4 days. We saw that with Gert already. JMO.

[Kingarabian]

For the 12z EPS run through day 5, the number is down to 15 in regards to members showing development compared to the 00z run which had 31 members show development.

[AtlanticWind]

https://www.tropicaltidbits.com/analysi ... =0&ypos=88

GFS from 2 days ago showing this afternoon , Do we see Gert?
Hard to trust models on storm genesis.