Category Five Hurricane Lee

[GaWx]

27 minutes ago, wthrmn654 said:

Mama Mia nams boat loads west

The NAM is a horrible model to use for the tropics. That being said, the JMA may not be all that much better and I still post about it.

[wthrmn654]

2 minutes ago, GaWx said:

The NAM is a horrible model to use for the tropics. That being said, the JMA may not be all that much better and I still post about it.

Icon looks to be a little west.

[LakeNormanStormin]

Maine landfall.

Biggest rainfall and wind/surge impacts.

Squally weather for rest of NE

[Floydbuster]

Video Update on Lee

https://www.youtube.com/watch?v=_fiR9IOZnBE

 

For how bad Lee looks on satellite, recon still shows this is a very powerful hurricane. Easy 100 mph and large.

[wthrmn654]

Left lean? 

[GaWx]

0Z run landfalls:

-Arpege: SW NS early 9/16 968 mb (faster and E shift vs 12Z's NB)

-ICON: NB early 9/17 981 mb (similar to 12Z)

-GFS: NB early 9/17 976 mb (similar to 12Z)

-CMC: SW NS late 9/16 980 mb (slight W shift vs 12Z)

-UKMET: SW NS late 9/16 970 mb (slight E shift  vs 12Z)

-JMA (inferior): C ME early 9/17 973 mb (NE shift vs 12Z, which crossed CC)

-Euro: NB early 9/17 976 mb (similar to 12Z)

-KMA: SW NS midday 9/16 ~960 mb (big E shift vs 12Z's E ME)

[wthrmn654]

00z versus 06z

Clearly more to the west/ left lean 

 

[GaWx]

 -Out of twelve 0Z models including the 4 tropical models and excluding the NAM, the only one showing a US landfall is the lowly JMA (central ME). And the JMA shifted away from Cape Cod since 12Z. How far off does this 0Z JMA, even after its significant NE shift, appear to be? It is to the left of all 51 members of the last three EPS runs ending with today's 6Z run:

 

 -Recent EPS Runs' US Landfalls: going down

9/14 6Z: 6 (12%) 6 ME

9/14 0Z: 7 (14%) 7 ME

9/13 18Z: 10 (20%) 10 ME

9/13 12Z: 14 (27%) 13 ME, 1 MA

9/13 6Z: 35 (69%) 27 ME, 6 MA, 2 RI

9/12 18Z: 36 (71%) 21 ME, 9 MA, 3 RI, 3 NY (LI)

[NorthHillsWx]

Other than FL winds, nothing else points to this being a cat 2 anymore. Max SFMR values of 70-75 kts two flights in a row 

[eyewall]

52 minutes ago, NorthHillsWx said:

Other than FL winds, nothing else points to this being a cat 2 anymore. Max SFMR values of 70-75 kts two flights in a row 

Yeah it will basically be a gale center soon enough.

[Wannabehippie]

10 minutes ago, eyewall said:

Yeah it will basically be a gale center soon enough.

Is Lee already losing its tropical characteristics? I didn't think this would happen until it got further north.

[WaPo]

It’s starting to transition. 

[WxWatcher007]

Should still be plenty impactful for Maine and NS.

[NJwx85]

If anything the transition will spread out impacts even further. You will see tropical storm force winds and hurricane force gusts hundreds of miles away from the center.

[MattPetrulli]

19 minutes ago, NJwx85 said:

If anything the transition will spread out impacts even further. You will see tropical storm force winds and hurricane force gusts hundreds of miles away from the center.

Accidentally posted in another thread but

NOAA buoy 41048 located 
about 160 miles (260 km) northwest of the center of Lee has recently 
reported a peak one-minute sustained wind of 51 mph (83 km/h) and a 
gust to 60 mph (96 km/h).

[NJwx85]

2 minutes ago, MattPetrulli said:

Accidentally posted in another thread but

NOAA buoy 41048 located 
about 160 miles (260 km) northwest of the center of Lee has recently 
reported a peak one-minute sustained wind of 51 mph (83 km/h) and a 
gust to 60 mph (96 km/h).

Based on the latest guidance it looks like the stronger winds will stay just offshore the New England coast but it wouldn't take much of a last minute shift to change that.

[GaWx]

12 hours ago, GaWx said:

 -Out of twelve 0Z models including the 4 tropical models and excluding the NAM, the only one showing a US landfall is the lowly JMA (central ME). And the JMA shifted away from Cape Cod since 12Z. How far off does this 0Z JMA, even after its significant NE shift, appear to be? It is to the left of all 51 members of the last three EPS runs ending with today's 6Z run:

 

 -Recent EPS Runs' US Landfalls: going down

9/14 6Z: 6 (12%) 6 ME

9/14 0Z: 7 (14%) 7 ME

9/13 18Z: 10 (20%) 10 ME

9/13 12Z: 14 (27%) 13 ME, 1 MA

9/13 6Z: 35 (69%) 27 ME, 6 MA, 2 RI

9/12 18Z: 36 (71%) 21 ME, 9 MA, 3 RI, 3 NY (LI)

9/14 12Z: 11 (22%) 10 ME, 1 MA (CC)

So, the 12Z EPS had an increase in ME along with one MA outlier, but this still suggests a high chance (78%) for no US landfall. So, I'm going with no US landfall.

[wthrmn654]

Gefs and gfs at 18z went west maybe 50 miles was on the 65W line hard

 

[NorthHillsWx]

Did no expect the core to completely disappear this soon. 

[yoda]

Hurricane Lee Advisory Number  39
NWS National Hurricane Center Miami FL       AL132023
1100 PM AST Thu Sep 14 2023

...TROPICAL STORM WARNINGS ISSUED FOR PORTIONS OF ATLANTIC CANADA...
...DANGEROUS SURF AND RIP CURRENT CONDITIONS AFFECTING MUCH OF THE
EAST COAST OF THE UNITED STATES...


SUMMARY OF 1100 PM AST...0300 UTC...INFORMATION
-----------------------------------------------
LOCATION...32.9N 67.6W
ABOUT 170 MI...270 KM WNW OF BERMUDA
ABOUT 590 MI...945 KM SSE OF NANTUCKET MASSACHUSETTS
MAXIMUM SUSTAINED WINDS...85 MPH...140 KM/H
PRESENT MOVEMENT...N OR 10 DEGREES AT 14 MPH...22 KM/H
MINIMUM CENTRAL PRESSURE...957 MB...28.26 INCHES


WATCHES AND WARNINGS
--------------------
CHANGES WITH THIS ADVISORY:

The Canadian Hurricane Center has issued a Tropical Storm Warning
for New Brunswick from the U.S./Canada border to Fort Lawrence,
including Grand Manan Island, and a Tropical Storm Warning for the
coast of Nova Scotia from Fort Lawrence to Point Tupper.

SUMMARY OF WATCHES AND WARNINGS IN EFFECT:

A Hurricane Watch is in effect for...
* Stonington, Maine to the U.S./Canada border
* New Brunswick from the U.S./Canada border to Point Lepreau,
including Grand Manan Island
* Nova Scotia from Digby to Medway Harbour

A Tropical Storm Warning is in effect for...
* Bermuda
* Westport Massachusetts northward to the U.S./Canada border
* Martha's Vineyard
* Nantucket
* New Brunswick from the U.S./Canada border to Fort Lawrence,
including Grand Manan Island
* Nova Scotia from Fort Lawrence to Point Tupper

 

[GaWx]

7 hours ago, GaWx said:

9/14 12Z: 11 (22%) 10 ME, 1 MA (CC)

So, the 12Z EPS had an increase in ME along with one MA outlier, but this still suggests a high chance (78%) for no US landfall. So, I'm going with no US landfall.

9/14 18Z EPS had a mere two (4%) US landfalling members and they were on the far E end of ME. It's looking very good for no US landfall.

9/15 0Z UKMET into SW NS PM of 9/16

[bluewave]

The GFS did the best with the track of Lee.  Maybe this was just the GFS right of track bias at work. Like the old broken clock is right twice a day metaphor.

 

 

[wthrmn654]

Gfs on 2021 was ranked best for hurricanes,  not sure how it ranked on 2022, but I'd assume similar ranking as well. 

[NorthHillsWx]

Recon is… unimpressive 

[olafminesaw]

[40/70 Benchmark]

On 9/9/2023 at 7:14 PM, 40/70 Benchmark said:

I think RI is done for Lee.

 

On 9/9/2023 at 8:24 PM, 40/70 Benchmark said:

There is certainly a chance this still hits as far back as the cape...just because I don't favor that, it is a viable outcome. I'm not pushing any agenda. Not sure what that was all about.

Fairly certain that Lee did not end up undergoing RI again, nor will it landfall in the US as a Hurricane, so this is further evidence that the BS about me having some sort of agenda last weekend was a complete fabrication.

No changes to my thoughts from last week (Sep 7).

Hurricane Lee Minimal Threat to US | Eastern Mass Weather

Happy tracking.

[MANDA]

So much for the AI modeling.  Looks like they have more "learning" to do.  Was interesting to watch though.

[bluewave]

44 minutes ago, MANDA said:

So much for the AI modeling.  Looks like they have more "learning" to do.  Was interesting to watch though.

It’s funny how the OP Euro and machine learning models were both showing landfalls in SE New England in earlier runs before correcting further east toward the consensus. As to why this was the case, we can leave to the modelers writing the code. It may come down to the OP Euro being over amped at times for some reason. But this bias at times can work like in Sandy when the pattern was super amped up. So I guess some storms play to certain model biases depending on the situation. If they could train the AI to learn how and which biases to correct on a situational basis, then they will really have something.

https://www.ecmwf.int/en/about/media-centre/news/2023/how-ai-models-are-transforming-weather-forecasting-showcase-data

In a recent ECMWF blog and a Newsletter article, we’ve been highlighting the rapid rise of data-driven forecasts. These are produced by machine-learning-based forecasting models, created by innovators such as NVIDIA, Huawei and Google DeepMind.

The ERA5 reanalysis, produced by the EU-funded Copernicus Climate Change Service (C3S) implemented by ECMWF, is the dataset used to train these models. They can make accurate 10-day forecasts when provided with good initial conditions.

This summer, we focused on analysing Huawei’s Pangu-Weather to understand its strengths and weaknesses, with results collated in a submitted manuscript. At this early stage in the technology, some results are already displaying comparable skill to ECMWF’s Integrated Forecasting System (IFS), which is an exciting achievement.

[SENC]

34 minutes ago, bluewave said:

It’s funny how the OP Euro and machine learning models were both showing landfalls in SE New England in earlier runs before correcting further east toward the consensus. As to why this was the case, we can leave to the modelers writing the code. It may come down to the OP Euro being over amped at times for some reason. But this bias at times can work like in Sandy when the pattern was super amped up. So I guess some storms play to certain model biases depending on the situation. If they  could train the AI to learn how and which biases to correct on a situational basis the they will really have something.

https://www.ecmwf.int/en/about/media-centre/news/2023/how-ai-models-are-transforming-weather-forecasting-showcase-data

In a recent ECMWF blog and a Newsletter article, we’ve been highlighting the rapid rise of data-driven forecasts. These are produced by machine-learning-based forecasting models, created by innovators such as NVIDIA, Huawei and Google DeepMind.

The ERA5 reanalysis, produced by the EU-funded Copernicus Climate Change Service (C3S) implemented by ECMWF, is the dataset used to train these models. They can make accurate 10-day forecasts when provided with good initial conditions.

This summer, we focused on analysing Huawei’s Pangu-Weather to understand its strengths and weaknesses, with results collated in a submitted manuscript. At this early stage in the technology, some results are already displaying comparable skill to ECMWF’s Integrated Forecasting System (IFS), which is an exciting achievement.

Remember that OLD term? 

A far as *coding & Modeling*..  

Garbage In,, Garbage Out..  Equals a Pre-fabricated (NOT Necessarily TRUE ), Result.. 

[bluewave]

2 minutes ago, SENC said:

Remember that OLD term? 

A far as *coding & Modeling*..  

Garbage In,, Garbage Out..  Equals a Pre-fabricated (NOT Necessarily TRUE ), Result.. 

On the whole, modeling has taken a big leap forward from the 60s and 70s. Now we can use a super ensemble to get a good track on a hurricane at even 5 days out. Plus ensembles can now identify general day 6-10 day patterns fairly well. But we have still been pushing up against the 15 day limit.