wthrmn654

Watch entire gfs run... it comes north after Bermuda area and seems to be deciding where to go.. north east should be watching in case this becomes a thing.

The curve to north near end of run Are increasing.

12z euro ai, nearly a cap cod landfall with a slow crawl north.

In defense of the ai models usage, it's better this year then last year from what memory remembers. . Idk how it has done with other storms, but with Erin the ai models track was so very much locked in from a very early start, from a rough aspect. As with this storm i saw the euro ai first to show a glancing blow to Carolinas then loops/meander off the coast... way before other non ai models started... Do they have more data at there resources versus conventional models? Not sure but there deft is some kinda difference, likely not obvious to the naked eye.

Cmc and euro ai want to bring it to the Mid Atlantic/ north east before all is said and done...... that's after wandering off the south east. Will have to keep an eye on that to see off that becomes a thing or not.

Ai models won big with Erin track days before everyone jumped on board... . Icon took a big win imo on that storm as well, which is crazy to say lol

This maybe a big sike storm.......

Euro ai shows low pressures for humberto at the 12z time so we'll see how it does.

Icon starts going up coast by florida then gets pulled towards humberto...

Gefs supports the no landfall the op showed. This time around there showing how strong humberto is with pressures in the 900s versus 1000+

18zgfs no landfall!!!!

Well that was a rough Summary from a professional meteorologist I see online before.

One thing that's clearly going throw a wrench in recent modeling..... humberto likely going be a Cat 3 or 4 tomorrow and all of a sudden.... may explain models now showing further east stall and some not coming ashore at all.........

Going be interesting to see if the AI models can be good like what they did for Erin, again.

Google models look to mainly stay offshore then go up coast in various ways to. https://www.weathernerds.org/models/genc.html?&inittype=zoom&initfield=SST_Anomaly_Surface&initcycle=determine&initfhour=120&initimdimx=1050&initimdimy=630&initrange=55.000000000000:230.000000000000:20.000000000000:300.000000000000&initcx1=NaN&initcy1=NaN&initcx2=NaN&initcy2=NaN&initcross=False&initsound=False&initsoundx=429&initsoundy=291&initloop=True&initoverlay=False&initlstart=000&initlend=360&initlint=6&initol1=SST_Anomaly_Surface&initol2=null&initol3=null&initol4=null&initol5=null&initcities=Off&inithgwys=Off&inittracks=On&initlows=On&initlatlon=Off

Extremely dry can lead to extreme run off until the ground gets moist enough to absorb rain.

Funny no one's talking about the 12z euro ai run...... Doesn't come to shore anywhere it lingers south until the high to north weakens and comes to the north east around 983 mb close to long island...

It will though be a wet storm so that's basically a given

Something I don't think anyone has talked about/ mentioned but this L has been tracking against guidance, nearly due west since at least 18z yesterday... i mean just look at the actual track versus modeling at 18z with each x. Being 00,06,12z actual placement...... this will likely dictate storms overall intensity the closer to Florida it gets along with all the islands is gotta cross. With 12z runs showing this likely a TS possibly barely a Cat 1 maximum strength wise

And possibly bring threats further north depending how long it stalls.

Found Google ai tropical cyclone models page, with plots and interactive for anyone interested. Make sure you click the menu menu button top left to turn on or off more settings pots etc. https://deepmind.google.com/science/weatherlab

For comparison, 6z and 12z for humberto Not going lie, there is deft looks like less of a curve towards Bermuda. If anything further west before it turns.

Gefs, 6z versus 12z

Oh, and the consensus tracks shifts way west at 12z. Of which will likely play a big part in the invest future plan of attack For humberto