ATL: IRMA - Models

[tolakram]

I'm curios how long does it take for the HWRF to do its complete run? It seems like some of the Models run fairly quickly and the images get shared, than some of the others are a little slower. Maybe it's just my nerves and impatience at this point. Thanks again for all the info being shared.

You can follow it here:

https://www.tropicaltidbits.com/analysis/models/?model=hwrf&region=11L&pkg=mslp_pcpn_frzn&runtime=2017090612&fh=6&xpos=0&ypos=0

[WxGuy1]

Topic is temporarily locked.

Euro runs at 1:45PM, HWRF is running now.

Too much back and forth in here, and frankly I don't care where you think it's going unless you are a model (not that kind of model ), or are interpreting a model. There's too much thinking without enough evidence in here and it's not helpful.

It will be unlocked shortly.

Think of all the effort and time that has been spent on Storm2K analyzing each model run out to 120+ hours down to nearly the mile, only for it to be rendered nearly useless in 6 hours when the next run of the model(s) come(s) in.

Frankly, we can guarantee that the current models will not be 100% correct. Who am I to tell anyone how to spend his/her time, but this thread currently has more than 8000 posts (!) and is more than 400 pages long. I mean absolutely no disrespect to any specific posters/participants here, but there are a lot of "We'll have to wait and see what X model says [at some future run]" and really overanalyzing the models. Of course "the next few runs" will tell us a lot, but that's almost always the case. Tomorrow, we'll see posts like "Wow, the next run will be extremely important" and "The atmosphere will be better sampled tomorrow, so tomorrow's runs will be the most important yet", which are fine but really obvious.

Now, for something more relevant to specific model output for Irma... Any brush with the mountainous terrain of Hispaniola or eastern Cuba should cause some detrimental impact on Irma. Although very, very bad for those areas (!), it would be of some benefit to the U.S. coastal communities. The wiggles and waggles that the models continue to show are normal in my experience, and I'd reiterate that it's really not worth the time right now to analyze each model run from each model down to the mile.

[chris_fit]

Much closer to SE FL

[Voltron]

Question:

In earlier runs, the models had Irma puling a Hugo and going up through Raleigh North. Not it is doing this bending where it hits between Savannah and Charleston and going NW into the states and up through Western NC. What is causing that and is there a component that would shift it more North once it hits in SC say?

[Steve]

Canadian 12Z

Close to the GFS Position at 96 hours (Sunday morning) and ends up hitting NC later.

HWRF looking big and bad at 920mb moving toward Miami at 87 hours. At 96 hours it's about 1/2 degree East and 1/4 degree South of Miami hedging closer. This will be the most watched hurricane in history if the new runs end up verifying. Remember, we're now within 5 days where the track accuracy markedly improves.

[SouthFLTropics]

Holy Smokes this is going to be close...Might not make landfall but the coast will get the eyewall

[Hurricane Andrew]

Hurricane Andrew's 12z Model Restrengthening Chart

Code: Select all

Model - Initialization Pressure - Max Pressure (D1-D5) - Min Pressure (D2-D5)
GFS - 923mb - 943mb - 892mb
HWRF - 909mb - 935mb - 918mb
HMON*** - 918mb - 918mb - 874mb
UKMET - 957mb - 958mb - 921mb
ECM - PENDING
CMC - 988mb - 988mb - 960mb
NAVGEM - PENDING
NAM - 961mb - 967mb - 941mb

*** This model has some really wonky stuff going on, take it for what it is worth


Take it as a ratio, rather than exact numbers; ALL models show weakening, then significant restrengthening. OHC values in the area are 100-150, SSTs are 30-32. Max potential intensity values are actually higher than where Irma is now. (CIMSS)

[ROCK]

Topic is temporarily locked.

Euro runs at 1:45PM, HWRF is running now.

Too much back and forth in here, and frankly I don't care where you think it's going unless you are a model (not that kind of model ), or are interpreting a model. There's too much thinking without enough evidence in here and it's not helpful.

It will be unlocked shortly.

Think of all the effort and time that has been spent on Storm2K analyzing each model run out to 120+ hours down to nearly the mile, only for it to be rendered nearly useful in 6 hours when the next run of the model(s) come(s) in.

Frankly, we can guarantee that the current models will not be 100% correct. Who am I to tell anyone how to spend his/her time, but this thread currently has more than 8000 posts (!) and is more than 400 pages long. I mean absolutely no disrespect to any specific posters/participants here, but there are a lot of "We'll have to wait and see what X model says [at some future run]" and really overanalyzing the models. Of course "the next few runs" will tell us a lot, but that's almost always the case. Tomorrow, we'll see posts like "Wow, the next run will be extremely important" and "The atmosphere will be better sampled tomorrow, so tomorrow's runs will be the most important yet", which are fine but really obvious.

Now, for something more relevant to specific model output for Irma... Any brush with the mountainous terrain of Hispaniola or eastern Cuba should cause some detrimental impact on Irma. Although very, very bad for those areas (!), it would be of some benefit to the U.S. coastal communities. The wiggles and waggles that the models continue to show are normal in my experience, and I'd reiterate that it's really not worth the time right now to analyze each model run from each model down to the mile.

you are right...we should all call it a day and hope for the best...

[AdamFirst]

The HMON is the successor to the GFDL, which was retired in June. Still needs works.

[SouthFLTropics]

Topic is temporarily locked.

Euro runs at 1:45PM, HWRF is running now.

Too much back and forth in here, and frankly I don't care where you think it's going unless you are a model (not that kind of model ), or are interpreting a model. There's too much thinking without enough evidence in here and it's not helpful.

It will be unlocked shortly.

Think of all the effort and time that has been spent on Storm2K analyzing each model run out to 120+ hours down to nearly the mile, only for it to be rendered nearly useful in 6 hours when the next run of the model(s) come(s) in.

Frankly, we can guarantee that the current models will not be 100% correct. Who am I to tell anyone how to spend his/her time, but this thread currently has more than 8000 posts (!) and is more than 400 pages long. I mean absolutely no disrespect to any specific posters/participants here, but there are a lot of "We'll have to wait and see what X model says [at some future run]" and really overanalyzing the models. Of course "the next few runs" will tell us a lot, but that's almost always the case. Tomorrow, we'll see posts like "Wow, the next run will be extremely important" and "The atmosphere will be better sampled tomorrow, so tomorrow's runs will be the most important yet", which are fine but really obvious.

Now, for something more relevant to specific model output for Irma... Any brush with the mountainous terrain of Hispaniola or eastern Cuba should cause some detrimental impact on Irma. Although very, very bad for those areas (!), it would be of some benefit to the U.S. coastal communities. The wiggles and waggles that the models continue to show are normal in my experience, and I'd reiterate that it's really not worth the time right now to analyze each model run from each model down to the mile.

You bring up some very valid points and the fact is that the public and everyone in general need to heed the advice of professionals such as yourselves and the fine staff at the NHC to interpret all of the model data and provide the best forecast based on that data. I think that the S2K model thread for amateurs, such as myself, is more like a reality TV show or TV drama where you just have to keep coming back to see what happens next. I certainly don't hang life or death decisions on each run of the models...Those decisions need to be based on the official forecast and advice from the NHC.

[xironman]

Topic is temporarily locked.

Euro runs at 1:45PM, HWRF is running now.

Too much back and forth in here, and frankly I don't care where you think it's going unless you are a model (not that kind of model ), or are interpreting a model. There's too much thinking without enough evidence in here and it's not helpful.

It will be unlocked shortly.

Think of all the effort and time that has been spent on Storm2K analyzing each model run out to 120+ hours down to nearly the mile, only for it to be rendered nearly useful in 6 hours when the next run of the model(s) come(s) in.

Frankly, we can guarantee that the current models will not be 100% correct. Who am I to tell anyone how to spend his/her time, but this thread currently has more than 8000 posts (!) and is more than 400 pages long. I mean absolutely no disrespect to any specific posters/participants here, but there are a lot of "We'll have to wait and see what X model says [at some future run]" and really overanalyzing the models. Of course "the next few runs" will tell us a lot, but that's almost always the case. Tomorrow, we'll see posts like "Wow, the next run will be extremely important" and "The atmosphere will be better sampled tomorrow, so tomorrow's runs will be the most important yet", which are fine but really obvious.

Now, for something more relevant to specific model output for Irma... Any brush with the mountainous terrain of Hispaniola or eastern Cuba should cause some detrimental impact on Irma. Although very, very bad for those areas (!), it would be of some benefit to the U.S. coastal communities. The wiggles and waggles that the models continue to show are normal in my experience, and I'd reiterate that it's really not worth the time right now to analyze each model run from each model down to the mile.

Would land interaction before the turn also increase the possibility of a larger wind field and hence surge?

[stormreader]

What trends? Which of the main models has this going beyond the Outer Banks?

I'd be very concerned in coastal Georgia & the Carolinas, for sure though...

The general eastward trend in the models since last night. Granted, that trend may reverse given time.

Many of the ensembles do -

http://www.weathernerds.org/tc_guidance ... ng?8695221

http://www.weathernerds.org/tc_guidance ... g?98796551

With all due respect, the eastward trend has halted with the 12z runs we are currently seeing. We anxiously await the 12z Euro which starts in about 30 minutes.

I agree, and am somewhat confident that a final consensus will come in about 36-48 hours on the uppermost keys. Don't know exactly the angle, may still have a slight west component which is why I said the extreme SW Fl coast, maybe due north then. But think that the uppermost keys is still a strong bet.

[Steve]

Question:

In earlier runs, the models had Irma puling a Hugo and going up through Raleigh North. Not it is doing this bending where it hits between Savannah and Charleston and going NW into the states and up through Western NC. What is causing that and is there a component that would shift it more North once it hits in SC say?

Look at Canada and the strong High Pressure coming down into Ontario and NY State or thereabouts. Looks to be a pretty strong high (though not all that big) that comes over top and forces it that way. It's a trend we started talking about a week ago or so and in my mind was the one thing that would force a landfall regardless of whether or not it ever hit Florida. Remember that the circulation around a high is clockwise, and the circulation of a low is counter clockwise. It can't realistically (though I'm sure it's happened a time or two) bounce off the south of that airmass and go East. It would have to follow the path of least resistance as fluids usually do (air is essentially a volume of fluids). Hope that helps. If not, run your models at the North American level rather than CONUS or W Atlantic and you can see how it unfolds. The trough lifts out and the high comes down in its place. It's one example of what they call a "block" or "blocking high."

[jasons2k]

I remember like it was yesterday when the GFDL was the darling new model after nailing Katrina's swing through South Florida...that didn't last long.

[shah83]

Really need some custom storm surge modeling for that track.

[gtalum]

Everything is converging on a SEFL landfall. Let's hope the track trends Eastward again...

[Palmcitycane]

I'm curios how long does it take for the HWRF to do its complete run? It seems like some of the Models run fairly quickly and the images get shared, than some of the others are a little slower. Maybe it's just my nerves and impatience at this point. Thanks again for all the info being shared.

Every model takes about 30-60 minutes to run, depending on how far out it extends. HWRF only goes through 5 days, so it's over faster. the Global models go out for weeks, and take around an hour, or a bit more, to finish. UKMET releases all their data at once.

Thank you, learning as I go. Also - thanks for the link Mark.

[tolakram]

[MatthewsRevenge]

Euro's coming: three minutes.

[Steve]

HWRF 102 hours - eyewall touching right in SE Miami touching the edge of Biscayne Bay and south Miami (Key Biscayne, Coral Gables, Coconut Grove, etc. Looks bad at 923.