2023 Global Model Runs Discussion (Out thru day 16)

[Teban54]

 https://twitter.com/dmorris9661/status/1694162657375399952



 https://twitter.com/dmorris9661/status/1694162658939879451

[Blown Away]

06z GFS

[ChrisH-UK]

Surprised no one has posted the 00z CMC. Sends a 979mb storm right up Tampa Bay in 168 hours. Looks to develop from the EPac crossover.

This is just 7 days out, 12z run had something near here too, just not anywhere near as organized. GFS does have something move over central america but it remains disorganized too. 12Z euro also had something around here in a week. I guess this will be the area to watch over the next few days for damn sure (besides Franklin and Bermuda) Models really depend on how/where this pops out at on the Atlantic side. West Caribbean here in the Canadian. GFS destroys it over Central America and never really recovers it, Euro is somewhere in between. (18z has it in over Belize)

I.e. watch this on Friday.

https://i.imgur.com/AH28yTo.png

Well if you look at the hurricane parent models they show a tropical storm brewing up near Cuba that's heading to Florida.

[SFLcane]

Surprised no one has posted the 00z CMC. Sends a 979mb storm right up Tampa Bay in 168 hours. Looks to develop from the EPac crossover.

This is just 7 days out, 12z run had something near here too, just not anywhere near as organized. GFS does have something move over central america but it remains disorganized too. 12Z euro also had something around here in a week. I guess this will be the area to watch over the next few days for damn sure (besides Franklin and Bermuda) Models really depend on how/where this pops out at on the Atlantic side. West Caribbean here in the Canadian. GFS destroys it over Central America and never really recovers it, Euro is somewhere in between. (18z has it in over Belize)

I.e. watch this on Friday.

https://i.imgur.com/AH28yTo.png

Well if you look at the hurricane parent models they show a tropical storm brewing up near Cuba that's heading to Florida.

[url]https://imagizer.imageshack.com/v2/800x600q70/922/0F0GQU.png [/url]

I would not rely on that unless other global models are showing it.

[SFLcane]

Just saying...

It's really something to see the models show an Mdr hurricane and next run it’s gone. Glad I didn’t renew my weathermodels subscription I can watch that on tidbits certainly isn’t worth my $. Those waves on the 06z gfs already moving nw midway in the Mdr. Bet it’s gone at 12z. 0 consistency this year with anything. Surely things can still turn around but for now thats what i see.

[toad strangler]

Jus saying...

It's really something to see the models show an Mdr hurricane and next run it’s gone. Glad I didn’t renew my weathermodels subscription I can watch that on tidbits certainly isn’t worth my $. Those waves on the 06z gfs already moving nw midway in the Mdr. Bet it’s gone at 12z. 0 consistency this year with anything.

And some say modeling has come a long way the past decade or so I don't think so. What has come a long way are the humans at the NHC and elsewhere around the pro weather campus.

[Steve]

Jus saying...

It's really something to see the models show an Mdr hurricane and next run it’s gone. Glad I didn’t renew my weathermodels subscription I can watch that on tidbits certainly isn’t worth my $. Those waves on the 06z gfs already moving nw midway in the Mdr. Bet it’s gone at 12z. 0 consistency this year with anything.

And some say modeling has come a long way the past decade or so I don't think so. What has come a long way are the humans at the NHC and elsewhere around the pro weather campus.

While I agree that people are better, models have gotten much better in the last 10 years and even way better than that going backwards. I read something a while back that said that a 5 day forecast was about as good as tomorrow's was in 1980. Also hurricane track guidance and intensity have greatly improved. We have more data, better tools and more super computers. That's all I'm saying. I still don't think the models are good enough with genesis or anything and often miss things that seem obvious to observers.

Yesterday's JMA (scroll over JMAN) skirted the circle at 1-2-3. https://www.cpc.ncep.noaa.gov/products/ ... r_wh.shtml

Its' the only model really doing that as many of the others cut across the circle and head back for 6-7 which are not favorable Western Atlantic phases.

[jasons2k]

Friendly reminder that model verification is easy to numerically verify so be careful with opinions about how well they have or haven’t performed. If you’re wrong, someone will pull the data and show it.

[toad strangler]

Friendly reminder that model verification is easy to numerically verify so be careful with opinions about how well they have or haven’t performed. If you’re wrong, someone will pull the data and show it.

That's fair, but in regards to tropical genesis I think modeling has a long way to go. A LONG way. Once storms are actually on the table my comment would easily be refuted for sure.

[cheezyWXguy]

Jus saying...

It's really something to see the models show an Mdr hurricane and next run it’s gone. Glad I didn’t renew my weathermodels subscription I can watch that on tidbits certainly isn’t worth my $. Those waves on the 06z gfs already moving nw midway in the Mdr. Bet it’s gone at 12z. 0 consistency this year with anything.

And some say modeling has come a long way the past decade or so I don't think so. What has come a long way are the humans at the NHC and elsewhere around the pro weather campus.

While I agree that people are better, models have gotten much better in the last 10 years and even way better than that going backwards. I read something a while back that said that a 5 day forecast was about as good as tomorrow's was in 1980. Also hurricane track guidance and intensity have greatly improved. We have more data, better tools and more super computers. That's all I'm saying. I still don't think the models are good enough with genesis or anything and often miss things that seem obvious to observers.

Yesterday's JMA (scroll over JMAN) skirted the circle at 1-2-3. https://www.cpc.ncep.noaa.gov/products/ ... r_wh.shtml

Its' the only model really doing that as many of the others cut across the circle and head back for 6-7 which are not favorable Western Atlantic phases.

This is a great point. We have a ton more data being fed into these models than we did 10 years ago, more variables being accounted for, and all at ever-increasing resolution. This increases the potential for accuracy and precision, but also increases the number of opportunities for compounding errors. The models have made great strides since I first started paying attention to them 15 years ago or so, but we are still a long way away from reliable human-free forecasting in a lot of contexts

[Steve]

And some say modeling has come a long way the past decade or so I don't think so. What has come a long way are the humans at the NHC and elsewhere around the pro weather campus.

While I agree that people are better, models have gotten much better in the last 10 years and even way better than that going backwards. I read something a while back that said that a 5 day forecast was about as good as tomorrow's was in 1980. Also hurricane track guidance and intensity have greatly improved. We have more data, better tools and more super computers. That's all I'm saying. I still don't think the models are good enough with genesis or anything and often miss things that seem obvious to observers.

Yesterday's JMA (scroll over JMAN) skirted the circle at 1-2-3. https://www.cpc.ncep.noaa.gov/products/ ... r_wh.shtml

Its' the only model really doing that as many of the others cut across the circle and head back for 6-7 which are not favorable Western Atlantic phases.

This is a great point. We have a ton more data being fed into these models than we did 10 years ago, more variables being accounted for, and all at ever-increasing resolution. This increases the potential for accuracy and precision, but also increases the number of opportunities for compounding errors. The models have made great strides since I first started paying attention to them 15 years ago or so, but we are still a long way away from reliable human-free forecasting in a lot of contexts

Yeah. They still use the "clipper" model because it has some statistical value over blind luck. But it's pretty bad. The NAVGEM which I think was an upgrade from the old Navy Nogaps model is also pretty bad. I barely even look at it anymore when running models I want to see (EC, UK, CMC, GFS, RAP, FV3, CFS, etc.). I'm on about 25-30 years of watching them, and though we aren't where we want to be (particularly with genesis as Toad mentioned), it's a long way from the WRF, MRF, AVN (think maybe became the NAM?) and some of the ones that got replaced or upgraded over the years.

[jasons2k]

Oh wow I haven’t seen some of those acronyms in so long.
I remember awing at GFDL handling Katrina moving into then WSW across the Everglades - while strengthening - and thinking it was the model of the future.

Yes, we have come a long ways. Even with genesis. Models are much better now at sniffing-out medium-long range legitimate threats than they used to be. But you still have to reconcile the storms with the overall pattern and question if it even makes sense to start with. That aspect never changes…

[cheezyWXguy]

While I agree that people are better, models have gotten much better in the last 10 years and even way better than that going backwards. I read something a while back that said that a 5 day forecast was about as good as tomorrow's was in 1980. Also hurricane track guidance and intensity have greatly improved. We have more data, better tools and more super computers. That's all I'm saying. I still don't think the models are good enough with genesis or anything and often miss things that seem obvious to observers.

Yesterday's JMA (scroll over JMAN) skirted the circle at 1-2-3. https://www.cpc.ncep.noaa.gov/products/ ... r_wh.shtml

Its' the only model really doing that as many of the others cut across the circle and head back for 6-7 which are not favorable Western Atlantic phases.

This is a great point. We have a ton more data being fed into these models than we did 10 years ago, more variables being accounted for, and all at ever-increasing resolution. This increases the potential for accuracy and precision, but also increases the number of opportunities for compounding errors. The models have made great strides since I first started paying attention to them 15 years ago or so, but we are still a long way away from reliable human-free forecasting in a lot of contexts

Yeah. They still use the "clipper" model because it has some statistical value over blind luck. But it's pretty bad. The NAVGEM which I think was an upgrade from the old Navy Nogaps model is also pretty bad. I barely even look at it anymore when running models I want to see (EC, UK, CMC, GFS, RAP, FV3, CFS, etc.). I'm on about 25-30 years of watching them, and though we aren't where we want to be (particularly with genesis as Toad mentioned), it's a long way from the WRF, MRF, AVN (think maybe became the NAM?) and some of the ones that got replaced or upgraded over the years.

Ha, I still remember the days of either wxman or one of the other early pro Mets referring to the NGM as the “No-Good Model”

[SouthFLTropics]

Meanwhile, the 12z Icon and CMC once again show development approaching the SW Florida coast by early next week. I think the area around Belize and the Yucatan is going to need a bear watch over the next few days.

[jasons2k]

Meanwhile, the 12z Icon and CMC once again show development approaching the SW Florida coast by early next week. I think the area around Belize and the Yucatan is going to need a bear watch over the next few days.

Levi Cowan talked about this in his video update yesterday.

[jfk08c]

Meanwhile, the 12z Icon and CMC once again show development approaching the SW Florida coast by early next week. I think the area around Belize and the Yucatan is going to need a bear watch over the next few days.

Same location and time frame, definitely something to keep an eye on. Seems to be a good amount of shear in that area which should keep it from intensifying too much but that's always a sneaky area for development

[SFLcane]

Meanwhile, the 12z Icon and CMC once again show development approaching the SW Florida coast by early next week. I think the area around Belize and the Yucatan is going to need a bear watch over the next few days.

Decent signal for sure

06z eps

[cycloneye]

12z CMC has a MDR rider (Not 92L) and also has the Yucatán area.

[Ianswfl]

Meanwhile, the 12z Icon and CMC once again show development approaching the SW Florida coast by early next week. I think the area around Belize and the Yucatan is going to need a bear watch over the next few days.

Same location and time frame, definitely something to keep an eye on. Seems to be a good amount of shear in that area which should keep it from intensifying too much but that's always a sneaky area for development

Yea. Models had it up to a cat2 a couple days ago. It will be a small system too but look how Charley ramped up and the waters are warming back up again now after cooling a bit. IMO we see mid range cat1 just due to the warm waters but wouldn't be shocked to see a cat2 from Tampa on South. However even just a cat1 or 2 will set back recovery efforts even just 6 foot of surge over those areas.

Could be as early as Tue this hits so not a lot of time. You would think NHC highlights that area any time now since it's well within 7 days if it starts to form.

[SouthFLTropics]

Euro is jumping onboard the Yucatan disturbance. Think we may have a player for Florida. And unfortunately, it will most likely get the dreaded "I" name. Hopefully the curse of that letter storm can be broken, and this just stay weak.