NATL: MELISSA - Models

[CrazyC83]

This is just my opinion based on the model runs over the last few days. If Melissa gets close enough to slightly hit or barely scrape by South Florida, a direct hit on the northeast is pretty much guaranteed.

Melissa is not a threat to the northeast. Most evidence points to it going out to sea after the Cuba hit.

They said that when Sandy first formed. Then look what happened.

Also 12z CMC just hates me to the fact that’s literally on Halloween and under a week before the gubernatorial election here in NJ, just like how Sandy was in 2012 with the presidential election. The fact some 12z EURO AIFS members start bending it back west towards the U.S. is not helping.

If anything, the atmosphere next week looks even more chaotic than in 2012. You're going to have not one, but at least two opportunities to draw it in as it looks like at least two troughs will come through. The big question is whether a ridge can build in between the two as the first trough lifts - that would be the difference. Or if this can outrace the first trough (less likely).

[xironman]

That is not Melissa on the GFS

At this distance the models can break things apart but still end up at the right solution

[KirbyDude25]

Absolutely nutso run from HAFS-B; 884 mbar, 173.4 knots at 126 hours and it looks like higher at 132 (which I don't have access to):

https://i.imgur.com/sIAb01V.png

Amazingly, the experimental v2.2.1-B version of the model is even more absurd, with a peak of 873 mb and 186 knots. Again, definitely overcooked, but the western Caribbean (especially Jamaica and Cuba) should be keeping a very close eye on this storm

[CrazyC83]

Absolutely nutso run from HAFS-B; 884 mbar, 173.4 knots at 126 hours and it looks like higher at 132 (which I don't have access to):

https://i.imgur.com/sIAb01V.png

Amazingly, the experimental v2.2.1-B version of the model is even more absurd, with a peak of 873 mb and 186 knots. Again, definitely overcooked, but the western Caribbean (especially Jamaica and Cuba) should be keeping a very close eye on this storm

It was 10 years ago this week that Patricia happened. Let's not count the chickens before they hatch. Not saying it will happen, or that even being the most likely, but nothing can be ruled out.

[Spacecoast]

A few ensemble members from GenCast / FNV3 approach, or cross SFL. Maybe 5 members out of 360 (only 1.5%).

[Iceresistance]

Absolutely nutso run from HAFS-B; 884 mbar, 173.4 knots at 126 hours and it looks like higher at 132 (which I don't have access to):

https://i.imgur.com/sIAb01V.png

Amazingly, the experimental v2.2.1-B version of the model is even more absurd, with a peak of 873 mb and 186 knots. Again, definitely overcooked, but the western Caribbean (especially Jamaica and Cuba) should be keeping a very close eye on this storm

Someone managed to get this one, and it shows 199 knots and 861 mb, absolutely insane

https://s12.gifyu.com/images/b3vfk.png
The image is after the data got plotted on Python

[Travorum]

Here's the 12z experimental HAFSV2.1.1B, which went absolutely insane, the peak winds were in the sounding above at 199kts but the frame after the pressure dropped more to 859mb. I'm not sure what all's different from the operational or how well it has verified in the past (obviously it seems quite overcooked here) but it's been extremely bullish on Melissa for days.

[MarioProtVI]

Melissa is not a threat to the northeast. Most evidence points to it going out to sea after the Cuba hit.

They said that when Sandy first formed. Then look what happened.

Also 12z CMC just hates me to the fact that’s literally on Halloween and under a week before the gubernatorial election here in NJ, just like how Sandy was in 2012 with the presidential election. The fact some 12z EURO AIFS members start bending it back west towards the U.S. is not helping.

If anything, the atmosphere next week looks even more chaotic than in 2012. You're going to have not one, but at least two opportunities to draw it in as it looks like at least two troughs will come through. The big question is whether a ridge can build in between the two as the first trough lifts - that would be the difference. Or if this can outrace the first trough (less likely).

Amazing

[Teban54]

As a throwback, I found some HAFS-B runs for Milton in its model thread and the Crazy Model Runs thread, and compared them to its actual intensity.

I focused on HAFS-B solely for simplicity and considering that it's the most aggressive model for Melissa now. Note that in some of these 2024 runs, HAFS-A actually showed a stronger Milton. Also, not all runs are available (especially because the models thread was focused on landfalls at the time rather than peak intensity).

Note that these lead times are short (<48 hrs), as I didn't look through earlier runs beyond the first mention in Crazy Model Runs (and Milton's potential to even form at all wasn't apparent until a handful days prior). Their generalizability to 126-hr HAFS-B runs may thus be in question.

• HAFS-B, 00z 10/06/24: 887 mb / 166 kt
• HAFS-B, 06z 10/06/24: 899 mb / 160 kt
• HAFS-B, 12z 10/06/24: 901 mb / 157 kt
• HAFS-B, 00z 10/07/24: 901 mb / 164 kt
• HAFS-B, 12z 10/07/24: 884 mb / 172 kt
• Actual, 20z 10/07/24: 895 mb / 155 kt

For reference, here are the HAFS-B Melissa runs, with a much longer lead time:

• HAFS-B, 00z 10/22/25: 889 mb / 164 kt
• HAFS-B, 06z 10/22/25: 900 mb / 144 kt
• HAFS-B, 12z 10/22/25: 884 mb / 173 kt

[Hypercane_Kyle]

Pretty clear at this point the GFS is in lalaland. Not even really worth looking at.

[Iceresistance]

You are seeing this correctly, the EPS has a member that goes to 877 mb, I don't remember the last time the EPS had something like this before

https://s12.gifyu.com/images/b3vrY.png

[ronjon]

A few ensemble members from GenCast / FNV3 approach, or cross SFL. Maybe 5 members out of 360 (only 1.5%).

[url]https://i.ibb.co/M5k4XZHM/grd1.jpg [/url]

Yes unlikely at this time with these particular ensemble runs. However the trough and NE turn is 6 days away and things in the upper air pattern can still change - especially as we transition into a fall progressive long wave pattern. In the tropics, timing is everything. Where the storm is located, it's forward speed, it's intensity, timing of the trough, does the trough slow down and dig deeper, does a deeper storm pump up any ridging - just a few of many variables still unresolved at this point. Case in point - just look at the difference in GFS and Euro in the next 24-48 hrs.

[SFLcane]

Hmmm…

[Soluna16]

Hmmm…

https://i.postimg.cc/sx3sWsLQ/IMG-1417.jpg

https://i.postimg.cc/fLNN4MDY/IMG-1418.jpg

No sir I am not a fan

[LarryWx]

For the record: The last 2 12Z JMA runs (they go out to 192) have shifted greatly to the E after earlier being pretty close to the UKMET W solutions:

12Z 10/21: very slow throughout and end up just W of Jamaica after curling back to the NNE

12Z 10/22: curls back to the NE over Jamaica at 156 and then to E tip of Cuba followed by Turks/Caicos

[ChrisH-UK]

Google Cyclone ML has the intensity modelling with 9 been Tropical storm, 4 Cat 1, 5 Cat2, 7 Cat 3, 7 Cat 4 and 14 been Cat 5 with the highest Speed been 179 knots and 897mb. Now with the lack of tropical cyclone activity this year in the Caribbean there is so much energy in the waters I wouldn't be surprised if it went nuclear like some models are showing. Where ever she goes it's going to apocalyptic.

Google Weather Lab - Tropical Cyclone Model

[aspen]

GFS once again refuses to back down on its quick Hispaniola scenario.

The ICON stalls over Haiti after a Cat 2/3 landfall, which would be horrifically devastating if it occurred. Fortunately it probably won’t.

[ScottNAtlanta]

Google Cyclone ML has the intensity modelling with 9 been Tropical storm, 4 Cat 1, 5 Cat2, 7 Cat 3, 7 Cat 4 and 14 been Cat 5 with the highest Speed been 179 knots and 897mb. Now with the lack of tropical cyclone activity this year in the Caribbean there is so much energy in the waters I wouldn't be surprised if it went nuclear like some models are showing. Where ever she goes it's going to apocalyptic.

Google Weather Lab - Tropical Cyclone Model

[url]https://imagizer.imageshack.com/v2/800x600q70/924/Ork9sn.png [/url]

Might as well give a 3 year old some coloured crayons and a map to scribble on. Would look close to the same

[chaser1]

GFS once again refuses to back down on its quick Hispaniola scenario.

The ICON stalls over Haiti after a Cat 2/3 landfall, which would be horrifically devastating if it occurred. Fortunately it probably won’t.

I'm guessing that Melissa still has nearly 2 days to impress the judges Still time to deepen and align its LLC & MLC. Maybe unlikely after what we've been seeing for the last day or so but also within the realm of possibility. All model suggest the New England trough to remain in place for nearly 2 more days. Short of that occuring though, we'll have ICON & GFS at the door with mud on their faces sometime Friday, asking to come out of the cold and join their friends on the Western Caribbean bandwagon.

[SouthFLTropics]

18z HAFS-B has so far backed off significantly from the earlier insane 12z run.