ATL: Ex INVEST 99L - Discussion

[AJC3]

Tropical Weather Outlook
NWS National Hurricane Center Miami FL
200 AM EDT Mon Aug 25 2025

East the Windward Islands (AL99):
Showers and thunderstorms associated with a tropical wave over the
Windward Islands have decreased over the past 24 hours, and recent
satellite wind data indicate the winds associated with the system
have also decreased. The disturbance is moving into a less
favorable environment, and the chances of development are
diminishing. Although the system is weakening, gusty winds are
possible over the Windward Islands through this morning.
* Formation chance through 48 hours...low...10 percent.
* Formation chance through 7 days...low...10 percent.

[cycloneye]

Tropical Weather Outlook
NWS National Hurricane Center Miami FL
800 AM EDT Mon Aug 25 2025

For the North Atlantic...Caribbean Sea and the Gulf of America:

Active Systems:
The National Hurricane Center is issuing advisories on Tropical
Storm Fernand, located over the central subtropical Atlantic.

1. Eastern Caribbean Sea (AL99):
A tropical wave over the eastern Caribbean Sea is producing
disorganized showers and thunderstorms. This system has become
less organized since yesterday, and further development is not
anticipated as it moves quickly westward during the next few days.
* Formation chance through 48 hours...low...near 0 percent.
* Formation chance through 7 days...low...near 0 percent.



Forecaster Blake

[Woofde]

Tropical Weather Outlook
NWS National Hurricane Center Miami FL
800 AM EDT Mon Aug 25 2025

For the North Atlantic...Caribbean Sea and the Gulf of America:

Active Systems:
The National Hurricane Center is issuing advisories on Tropical
Storm Fernand, located over the central subtropical Atlantic.

1. Eastern Caribbean Sea (AL99):
A tropical wave over the eastern Caribbean Sea is producing
disorganized showers and thunderstorms. This system has become
less organized since yesterday, and further development is not
anticipated as it moves quickly westward during the next few days.
* Formation chance through 48 hours...low...near 0 percent.
* Formation chance through 7 days...low...near 0 percent.



Forecaster Blake

The classic Eastern Carribean effect, the graveyard adds another.

[REDHurricane]

-1 points for ICON

[TampaWxLurker]

It's dead, Jim.

(But it wouldn't be a bad idea to watch for where the remnant pieces of energy go down the road)

[BobHarlem]

It's dead, Jim.

(But it wouldn't be a bad idea to watch for where the remnant pieces of energy go down the road)

It'll show up in the East pacific later.

[LarryWx]

-1 points for ICON

ICON’s apparent poor performance reminds me of the Invest that failed to develop in the Gulf in July that mainly only the ICON was active with as it had a H or TS on many runs. It’s a very hit or miss model when it’s mainly on its own.

[Nimbus]

It's dead, Jim.

(But it wouldn't be a bad idea to watch for where the remnant pieces of energy go down the road)

Surface pressures were so broad and shallow yesterday when recon wasted their fuel investigating.
Would have been apex should be near 14.5 but would need a lot of persistent convection and that can't happen till at least Jamaica.

[LarryWx]

It's dead, Jim.

(But it wouldn't be a bad idea to watch for where the remnant pieces of energy go down the road)

Surface pressures were so broad and shallow yesterday when recon wasted their fuel investigating.
Would have been apex should be near 14.5 but would need a lot of persistent convection and that can't happen till at least Jamaica.

Do you know why the 0Z UKMET had the center way down at 10.7N?

TROPICAL STORM 99L ANALYSED POSITION : 10.7N 60.1W

ATCF IDENTIFIER : AL992025

LEAD CENTRAL MAXIMUM WIND
VERIFYING TIME TIME POSITION PRESSURE (MB) SPEED (KNOTS)
-------------- ---- -------- ------------- -------------
0000UTC 25.08.2025 0 10.7N 60.1W 1012 26
1200UTC 25.08.2025 12 CEASED TRACKING

[DorkyMcDorkface]

ICON is (rightfully) catching strays here but let's not forget several EPS members had an intensifying hurricane getting into the Western Caribbean/Gulf from this as well. This is from the 12z a mere two days ago:

[LarryWx]

ICON is (rightfully) catching strays here but let's not forget several EPS members had an intensifying hurricane getting into the Western Caribbean/Gulf from this as well. This is from the 12z a mere two days ago:
https://i.imgur.com/boudhRq.png

Yes, the 12Z EPS of Saturday, indeed, had several intense members. However:

-it was still just several out of 50 members (pretty small %)
-it was the only run like that

Thus, I feel the Icon deserves much more criticism than the EPS. I don’t think the EPS deserves much of any criticism considering my two points.

[TomballEd]

Some runs of the ICON during Hurricane Beryl were much more accurate that most of the other models. Not every run, but quite a few. That seemed to earn the ICON a reputation as the next best thing. Main benefit, it comes out earlier than the other globals

Perhaps the ICON is like the Canadian or NOGAPS of old and spin up more than their fair share of spurious cyclones.

The Canadian has been improved, I still trust the GFS/Euro more, but it no longer qualifies as the 'crazy uncle' model. GFS and Euro ensembles had members liking 99L, so no model/model family did perfectly.

The lack of surface convergence common with accelerating low level flow in the E Caribbean and the shear created not by strong upper level winds but rather strong low level winds, as I mentioned yesterday, is obvious. Just waiting for WxMan47 to sign the death certificate.

[wxman57]

This one is dead in the Atlantic. It'll cross Central America later this week and it will have a shot at EPAC development. No development in the Caribbean or Gulf, though.

[chaser1]

It's dead, Jim.

You beat me to it!

[USTropics]

Some runs of the ICON during Hurricane Beryl were much more accurate that most of the other models. Not every run, but quite a few. That seemed to earn the ICON a reputation as the next best thing. Main benefit, it comes out earlier than the other globals

Perhaps the ICON is like the Canadian or NOGAPS of old and spin up more than their fair share of spurious cyclones.

The Canadian has been improved, I still trust the GFS/Euro more, but it no longer qualifies as the 'crazy uncle' model. GFS and Euro ensembles had members liking 99L, so no model/model family did perfectly.

The lack of surface convergence common with accelerating low level flow in the E Caribbean and the shear created not by strong upper level winds but rather strong low level winds, as I mentioned yesterday, is obvious. Just waiting for WxMan47 to sign the death certificate.

https://i.imgur.com/UXbsTMm.gif

First, condolences to 99L. They were a fighter (and lasted much longer than I anticipated), but alas the Atlantic's graveyard came calling.

Your assessment above is solid, I'll add a few of my thoughts here. I lump the ICON with the UKMET model in terms of track and genesis. They both can be outliers, and if you understand a model's bias and tendency, they can be useful tools in that regard. The ICON tends to do well with weak systems and their initial development phase (it was the only model that accurately predicted Dorian's center relocation and track through the Caribbean). The UKMET also has some 'wins' (see Irma and Ivan). Just like the ICON on certain storms/situations, It also did poorly in certain seasons (Dennis and Emily from 2005 come to mind).

The 'big three' (or 2.5 ) right now imo is the ECMWF, GFS, and CMC. If those three operational models, and more importantly their ensemble suite back it up, you can almost be certain development will occur. The models and ensembles were on Erin early and often, whereas the models dropped 99L (and even 90L) for periods of time.

We use the NOGAPS/NAVGEM at the end of our weather discussions at FSU for entertainment only.

[LarryWx]

Some runs of the ICON during Hurricane Beryl were much more accurate that most of the other models. Not every run, but quite a few. That seemed to earn the ICON a reputation as the next best thing. Main benefit, it comes out earlier than the other globals

Perhaps the ICON is like the Canadian or NOGAPS of old and spin up more than their fair share of spurious cyclones.

The Canadian has been improved, I still trust the GFS/Euro more, but it no longer qualifies as the 'crazy uncle' model. GFS and Euro ensembles had members liking 99L, so no model/model family did perfectly.

The lack of surface convergence common with accelerating low level flow in the E Caribbean and the shear created not by strong upper level winds but rather strong low level winds, as I mentioned yesterday, is obvious. Just waiting for WxMan47 to sign the death certificate.

https://i.imgur.com/UXbsTMm.gif

First, condolences to 99L. They were a fighter (and lasted much longer than I anticipated), but alas the Atlantic's graveyard came calling.

Your assessment above is solid, I'll add a few of my thoughts here. I lump the ICON with the UKMET model in terms of track and genesis. They both can be outliers, and if you understand a model's bias and tendency, they can be useful tools in that regard. The ICON tends to do well with weak systems and their initial development phase (it was the only model that accurately predicted Dorian's center relocation and track through the Caribbean). The UKMET also has some 'wins' (see Irma and Ivan). Just like the ICON on certain storms/situations, It also did poorly in certain seasons (Dennis and Emily from 2005 come to mind).

The 'big three' (or 2.5 ) right now imo is the ECMWF, GFS, and CMC. If those three operational models, and more importantly their ensemble suite back it up, you can almost be certain development will occur. The models and ensembles were on Erin early and often, whereas the models dropped 99L (and even 90L) for periods of time.

We use the NOGAPS/NAVGEM at the end of our weather discussions at FSU for entertainment only.

Regarding the UKMET, don’t forget about its absolutely stellar performance for both landfalls of Ian. I still remember this clearly because I followed it very closely and documented it here in great detail. Going into W FL, it was well to the right of all of the major op global runs. And Icon did well, itself, as it was 2nd best. Euro was in the middle. The CMC and GFS were absolutely atrocious with Panhandle projections just a few days out!

But I’m not a UKMET “hack” as I’ll also agree with you and say that it has been way off for some others. So, like the Icon, it seems to be often be a hit or miss model. The secret may be to favor the UKMET only for “I” storms.

[USTropics]

Some runs of the ICON during Hurricane Beryl were much more accurate that most of the other models. Not every run, but quite a few. That seemed to earn the ICON a reputation as the next best thing. Main benefit, it comes out earlier than the other globals

Perhaps the ICON is like the Canadian or NOGAPS of old and spin up more than their fair share of spurious cyclones.

The Canadian has been improved, I still trust the GFS/Euro more, but it no longer qualifies as the 'crazy uncle' model. GFS and Euro ensembles had members liking 99L, so no model/model family did perfectly.

The lack of surface convergence common with accelerating low level flow in the E Caribbean and the shear created not by strong upper level winds but rather strong low level winds, as I mentioned yesterday, is obvious. Just waiting for WxMan47 to sign the death certificate.

https://i.imgur.com/UXbsTMm.gif

First, condolences to 99L. They were a fighter (and lasted much longer than I anticipated), but alas the Atlantic's graveyard came calling.

Your assessment above is solid, I'll add a few of my thoughts here. I lump the ICON with the UKMET model in terms of track and genesis. They both can be outliers, and if you understand a model's bias and tendency, they can be useful tools in that regard. The ICON tends to do well with weak systems and their initial development phase (it was the only model that accurately predicted Dorian's center relocation and track through the Caribbean). The UKMET also has some 'wins' (see Irma and Ivan). Just like the ICON on certain storms/situations, It also did poorly in certain seasons (Dennis and Emily from 2005 come to mind).

The 'big three' (or 2.5 ) right now imo is the ECMWF, GFS, and CMC. If those three operational models, and more importantly their ensemble suite back it up, you can almost be certain development will occur. The models and ensembles were on Erin early and often, whereas the models dropped 99L (and even 90L) for periods of time.

We use the NOGAPS/NAVGEM at the end of our weather discussions at FSU for entertainment only.

Regarding the UKMET, don’t forget about its absolutely stellar performance for both landfalls of Ian. I still remember this clearly because I followed it very closely and documented it here in great detail. Going into W FL, it was well to the right of all of the major op global runs. And Icon did well, itself, as it was 2nd best. Euro was in the middle. The CMC and GFS were absolutely atrocious with Panhandle projections just a few days out!

But I’m not a UKMET “hack” as I’ll also agree with you and say that it has been way off for some others. So, like the Icon, it seems to be often be a hit or miss model. The secret may be to favor the UKMET only for “I” storms.

Maybe UKMET

Totally forgot about Ian, it did really well there as well. I mean...if you're going to be right...be right on the 'I' storms It's going to get another go here soon.

[zhukm29]

I think 99L will be a good candidate for reanalysis post season, since I personally believe this was briefly a tropical storm around August 20-21. First, we have the ASCAT pass at 12z on the 21st that showed a fairly defined circulation and gale force winds. Many other systems were designated with worse, and 99L looked better than Erin when Erin was first designated (and much of Erin's early lifetime across the MDR):


At that time 99L also had T2.5 support for Dvorak, and plenty of sites online had estimates of 35kt, which also supports TS designation (here's one from 12z on the 21st, when that ASCAT pass was taken). There were also several tweets from other mets in this thread that supported the classification of 99L as a TC.

TXNT27 KNES 211215
TCSNTL

A. TROPICAL DISTURBANCE (99L)

B. 21/1200Z

C. 11.2N

D. 33.8W

E. THREE/GOES-E

F. T2.5/2.5

G. IR/EIR/VIS

H. REMARKS...5/10 BANDING RESULTS IN A DT OF 2.5. THE MET IS 1.5. THE
PT IS 2.0. THE FT IS BASED ON THE DT.

I. ADDL POSITIONS

NIL


...MONAGHAN

And of course, there was the microwave image of 99L that looked quite interesting:


The main thing going against 99L was that models kept predicting that it would go poof, so I can understand the hesitancy the NHC may have had in naming a storm only to have it die 6 hours later, especially since it was no threat to land. However, models completely underestimated how persistent 99L was, and while 99L did end up victim to the Caribbean graveyard, it lasted way longer than its initial prognosis of poofing on the 20th, and then the 21st, and then the 22nd... with the formation area on the NHC map continuously extended as each model run realized that 99L was not fully dead. Given that we've had many named systems before that were similar to 99L as it tracked through the MDR (Wilfred 2020 is a perfect example), I think it's only fair for consistency that this system be analyzed as a potential candidate as a tropical storm, even if it was not given a name operationally.

[tropicwatch]

Convection is kicking up again but vorticity not so much.

[ScottNAtlanta]

Convection is kicking up again but vorticity not so much.

Not so much, but it is still there, and convection is firing. If it can keep this up until the West Caribbean there could be something happening