Major Hurricane Melissa - 892mb - 185mph

[Amped]

1 hour ago, hawkeye_wx said:

Here's the latest Google model.

 

It seems like the less latitude it gains in the next 72hrs, the more it gets tucked back under the ridge and bombs in the following 72 hrs, like the Euro and Cmc are showing. 

The Icon and Gfs show it gaining more latitude in the first 72hrs, then interacting with land when it tries to bomb.

[Newman]

[WxWatcher007]

Tropical Storm Melissa Discussion Number   5
NWS National Hurricane Center Miami FL       AL132025
1100 AM EDT Wed Oct 22 2025
 
While Melissa continues to produce bursting convection on satellite 
imagery, with very cold cloud tops down to -95C, there is little 
evidence these convective bands are wrapping around the center. 
West-northwesterly shear of 20 to 25 kt is currently preventing the 
low-level and mid-level circulation centers from aligning. This 
downshear tilt with height has also been observed by the latest Air 
Force Reconnaissance mission, which reported significant winds at 
the surface in their most recent 850 mb center drop. The initial 
intensity for this advisory is held at 45 kt, taking a blend of 
recon observations which had a peak wind of 52 kt at 850 mb, and 
satellite intensity estimates from TAFB, ADT, and D-MINT. The wind 
field of Melissa also remains very asymmetric, as evidence of a 
recent 1040z RCM-1 Synthetic Aperture Radar pass showing the bulk of 
tropical-storm-force winds east of the center. 
 
The tropical storm has slowed down to a crawl this morning, with the 
estimated motion at 300/2 kt. The track forecast remains challenging 
today, and strongly related to the storm's structure in the 
short-term. For the next 12-24 hours, Melissa is forecast to move 
very slowly to the northwest as it attempts to move into a weakness 
produced by an broad upper-level trough centered over the Bahamas. 
Whether or not it turns more rightward into this weakness is related 
to the vertical depth the various model guidance is depicting. The 
06z GFS run, which is the model that has been on the right side of 
the guidance envelope for days, shows a more vertically coherent 
system over the next 24-48 hours, and causes Melissa to abruptly 
turn northeastward over the next 48 hours. While this appears to be 
an outlier solution, the Google DeepMind ensemble mean (GDMI) also 
shows this northeastward motion, and a significant fraction of its 
ensemble members show the Melissa making landfall on the Island of 
Hispaniola. However, the remainder of the track guidance, including 
the ECMWF and its ensemble mean, suggest that Melissa will remain 
misaligned and move more slowly north-northwest over the next couple 
of days. The NHC track solution slightly favors the latter 
scenarios, lying just east of the ECMWF mean, out of respect of the 
reliable GDMI aid. After about 60 h, mid-level ridging begins to 
build back in to the north, and most of the guidance responds by 
showing a sharp turn westward, albeit still at a snail's pace. The 
overall track guidance, GFS and GDMI excluded, has shifted left this 
cycle beyond 72 h, and the NHC track was shifted in that direction, 
but not as far south and west as the HAFS-A/B regional models. It 
goes without saying, this is a very challenging track forecast. 
 
The intensity forecast has its own share of challenges. At least in 
the short-term, vertical wind shear is expected to persist between 
20 to 30 kt, and will likely prevent Melissa from becoming fully 
aligned and taking advantage over the very warm 30-31C sea surface 
temperatures. After about 48 hours, the GFS and ECMWF-based SHIPS 
guidance shows the shear decreasing gradually, but the timing on 
when the system becomes more vertically coherent varies among both 
the global and hurricane-regional model guidance. Most of the 
guidance shows significant or even rapid intensification in days 
3-5, and the NHC intensity forecast was raised once again during 
this time frame, but not as high as the HCCA or hurricane-regional 
models due to continuity constraints to the previous forecast. The 
day 5 forecast now shows Melissa becoming a major hurricane, and 
further upward adjustments may be necessary in subsequent cycles. 
The intensity forecast is also more uncertain than usual, in large 
part related to the track uncertainty.

Key Messages:
 
1. Melissa is expected to produce heavy rainfall across portions of
the Dominican Republic, Haiti, and Jamaica through this weekend,
bringing a risk of significant flash flooding and landslides.
 
2. Hurricane conditions are possible in the southwestern peninsula
of Haiti where a Hurricane Watch is in effect.  A Tropical Storm
Watch is in effect for Jamaica.  Preparations to protect life and
property should be completed by Thursday.
 
3. There is significant uncertainty in the track and intensity
forecast of Melissa.  Interests elsewhere in Hispaniola and Cuba
should continue to monitor the latest forecasts.
 
 
FORECAST POSITIONS AND MAX WINDS
 
INIT  22/1500Z 14.4N  73.6W   45 KT  50 MPH
 12H  23/0000Z 14.8N  73.9W   45 KT  50 MPH
 24H  23/1200Z 15.2N  74.3W   50 KT  60 MPH
 36H  24/0000Z 15.6N  74.4W   55 KT  65 MPH
 48H  24/1200Z 16.0N  74.4W   55 KT  65 MPH
 60H  25/0000Z 16.3N  74.6W   60 KT  70 MPH
 72H  25/1200Z 16.5N  74.9W   70 KT  80 MPH
 96H  26/1200Z 16.6N  75.8W   90 KT 105 MPH
120H  27/1200Z 16.4N  77.4W  105 KT 120 MPH
 
$$
Forecaster Papin
  

[cptcatz]

Pressure rising and wobbling southwest. Time to take GFS out back and shoot it. With king euro reigning, it'll be very interesting to see how strong this gets in the western caribbean sea. 

[Amped]

The 12z Cmc is probably just as far off with the track at 120hrs as the GFS, just in the opposite direction. If you average them you might get an accurate position somewhere southwest of Jamaica.

[olafminesaw]

The Google models have been really consistent 

 

[Amped]

The 12 Cmc pulls a Sandy.

[batmanbrad]

8 minutes ago, Amped said:

The 12 Cmc pulls a Sandy.

given that it comes in around Halloween, it would seen that the CMC run is trying to look really... scary.  Presumably the 0Z run will be like Monty Python says ("And now for something completely different")

[hawkeye_wx]

The surface center has once again been spit out toward the west as the convection is unable to hold onto it.

[Chinook]

1 hour ago, Amped said:

The 12 Cmc pulls a Sandy.

That is pretty funny how it is similar to Sandy's track. I suppose we can say it's a good thing that Hurricane (non-Hurricane) Sandy was a 1 in a million shot.

[hawkeye_wx]

Latest Euro

The new HAFS is going sub-900.

[hawkeye_wx]

HAFS-B... when it's well sw of Jamaica.

[Amped]

35 minutes ago, hawkeye_wx said:

The surface center has once again been spit out toward the west as the convection is unable to hold onto it.

It seems almost like something related to the diurnal cycle.  A convective blob forms at night and pulls the center east. Then shear ramps up and rips it apart during the day and it drifts back west. 

[hawkeye_wx]

12z Google model.... more tracks missing Jamaica to the west

[NorthHillsWx]

11 minutes ago, hawkeye_wx said:

12z Google model.... more tracks missing Jamaica to the west

It’s crazy that we have a defined center and named storm and the ensemble outputs look like a 2 year olds drawing 

[jconsor]

On 10/20/2025 at 4:24 PM, jconsor said:

Notable left trend on FNV3 (Google DeepMind ensemble) from 12z run yesterday to today's 6z run - both in terms of the the members that show a sharp NE curve into Hispaniola being further west (Haiti vs. DR) and a lot more members turning west into the NW Caribbean

The westward trend continues.  Very few members going east of Cuba now, with a substantial increase in the number of members west of 80W by next Wed.

[cptcatz]

13 minutes ago, jconsor said:

The westward trend continues.  Very few members going east of Cuba now, with a substantial increase in the number of members west of 80W by next Wed.

How close to florida do you think we could see?

[olafminesaw]

Actually some consensus starting to emerge. Aside from the stubborn GFS of course

[olafminesaw]

Also worth noting the GFS is an outlier against it's own ensembles 

[andyhb]

Cayman Islands may also have a serious problem with this storm, especially if it gets as intense as a pretty significant chunk of guidance suggests it could.

[Newman]

Just a reminder of what the typical upper level setup looks like for New England landfalling hurricanes. Obviously this is a broad brushed composite, but there are a lot of similarities to the currently forecast upper-level pattern from the EPS. New England hurricanes are rare for a reason - even if the overall larger pattern supports the possibility, the smaller details, exact placement of trough/ridges, etc. will have a role in impact (i.e. trough interaction) vs no impact (less trough interaction)

[Newman]

Here's the 12z EPS for reference.

Right now, follow the ensembles. The clustering/highest probabilities keep it well off the coast. This is a complex setup and there will be lots of moving parts. The chance of a New England impact is still very low, but I do believe worth keeping an eye on. We've already seen this tropical season major shifts in guidance in the 7-10 day range vs what actually happened.

[MDSnow93]

Melissa already has more ACE than four named storms this season. Sure seems like it could get us above normal, which didn't seem likely last month.

[Floydbuster]

 

[TheDreamTraveler]

Actually amazing how useless the GFS has been for this storm lol 

[MJO812]

5 minutes ago, TheDreamTraveler said:

Actually amazing how useless the GFS has been for this storm lol 

Still way east of every model

Gefs has 2 camps 

[TheDreamTraveler]

5 hours ago, Chinook said:

That is pretty funny how it is similar to Sandy's track. I suppose we can say it's a good thing that Hurricane (non-Hurricane) Sandy was a 1 in a million shot.

 

1 hour ago, Newman said:

Here's the 12z EPS for reference.

Right now, follow the ensembles. The clustering/highest probabilities keep it well off the coast. This is a complex setup and there will be lots of moving parts. The chance of a New England impact is still very low, but I do believe worth keeping an eye on. We've already seen this tropical season major shifts in guidance in the 7-10 day range vs what actually happened.

Another thing to keep in mind too is late October is the only real window for these hybrid storms to form with the perfect conditions. The only two storms we have recorded Sandy and The Perfect Storm in 1991 both happened at the very end of October. So it's interesting given the time period we're in and that the overall setup is very similar. Maybe once the GFS stops being absolutely useless and figures things out we'll have a clearer picture of what will happen.

 

Definitely worth keeping an eye on. 

[Chinook]

4 hours ago, hawkeye_wx said:

Latest Euro

The new HAFS is going sub-900.

I really hope it doesn't get that bad, with stalling, for Jamaica.

[GaWx]

2 hours ago, Chinook said:

I really hope it doesn't get that bad, with stalling, for Jamaica.

 18Z Euro gives much of Jamaica near or just over 20” qpf with heavy rain still likely continuing for a few more hours with the storm still centered over SW Jamaica and SLP way down at 943 mb, which would apparently be the record low per what I read. Lowest on record 958 (Charlie of 1951) followed by 960 (Gilbert of 1988). It’s then moving slowly NNE while SSE of a strong upper trough that’s then over the E US.

[WxWatcher007]