ATL: ISAAC - Models

[TwisterFanatic]

6z NAM is approaching SE Louisiana at 36 hours.

[TwisterFanatic]

Makes landfall on the LA/MS coast at 48 hours.

[Hurrilurker]

Also, I doubt many of the general public are here looking at raw model runs. The vast majority of people probably see tracks from a) nothing, they don't even pay attention, b) television news -- not much you can do about that, or c) the NHC, which considers many factors including but not limited to all the models in their analysis. And most news outlets are probably going off the NHC anyway.

Nothing wrong with having model data available to the public. That's like saying let's not look at jobs reports or farm yields or housing trends or interest rates that might affect the economy, because people might misinterpret them.

[WxGuy1]

The models aren't ready for the public...they need to stop revealing the models like this....the models have been so back and forth and truly completely unreliable. It's been nothing short of a guessing game and it's not fair to do that to the generel public (especially the elderly). There are alot of people in the public who do not understand models and with how unreliable they are, I just don't think it is good for us to reveal their runs...we would do much better with forecasting by following the traditional methods of watching the barometric pressure. This storm has revealed all models to be a joke...they are only good for about 36 hours before any strike. Weather is complex and it is evident that it will take several more years before these models are meant to be taking seriously. As for now...they just don't have it. How many more years before we stop having weathermen simply interpreting these models to us over the television until we simply have a voice on our TV reading to us what the forecast is based on a computer analysis??? I miss Nash Roberts....now we are simply stuck with a bunch of weather people guessing as much as the computers because that is what they do...they follow what a computer says. Not good.

Sigh.... No human could ever consider the amount of data that the models do when forecasting. The good meteorologists, like many of the ones on this site, will sometimes add their own spice when discussing models, but the models are the best thing we have. Are they perfect? No, but I'll take them before I trust someone's 'gut'. It almost seems like you prefer climo over the computer models. I'll put the GFS and Euro up against climo any day of the week with regards to TC forecasting. We cant treat the public like children. Sorry, couldn't zip my lip. That's all I have to say about that.

Not only that, but skill has improved significantly over the past few decades, particularly when it comes to track forecast errors (reducing forecast errors in intensity has been more problematic). Certainly, better observing platforms and a better understanding of tropical cyclone evolution and structure has helped increase our ability to forecasts these systems, but improved numerical weather prediction has also played a significant role in reducing forecast errors.

Are computer forecast models perfect? Absolutely not. That's one of the purposes of meteorologists -- humans who are knowledgeable about how the atmosphere works take all available information, including data from forecast models, and come up with their best forecast.

For what it's worth, I'm not sure how much Joe Q Public knows about numerical weather prediction. Your local TV meteorologists certainly should be doing more than just throwing up a bunch of model plots, otherwise you have a bad TV met. As such, I don't think I'd say that computer models are intended for the general public anyway. Sure, the data are available for viewing, but the public is better off following forecasts from a good meteorologist, IMO.

[KWT]

Hmmm models really are agressive past 24hrs, they don't do much till then but the rate of strengthening after that is a little worrying, we don'y need a system starting RI within 24hrs of landfall as the ECM does for example.

[Dean4Storms]

Thru 36hrs the 06z GFS is ditto with the 00z, now just @ Mouth of MS River.

[Dean4Storms]

Then west along the coast of LA @ 42hrs........

http://www.instantweathermaps.com/GFS-p ... L&hour=042

[KWT]

That track would certainly rake coastal parts of LA, theres no doubting that!

[Dean4Storms]

06z NAM has it stalled pretty much over New Orleans from 48hrs thru at least 72 hrs!

http://www.instantweathermaps.com/NAM-p ... L&hour=072

[Janie2006]

We know the NAM is, shall we say, notorious about its tropical forecasts, but I'm wondering if we can extrapolate information about what the model is thinking based on the storm's position in addition to the 500 MB vorticity chart. For example, the model might consistently indicate a weakness at 500 MB, but keeps pushing landfall to the west. Does this indicate that Isaac finds the weakness further west than, say, even 24 hours ago?

On the face of it the answer seems obvious, but I don't really think so because of the disconnect between the NAM's very good synoptic capabilities and its rather shoddy TC forecasts.

OTOH, it's 6:00 am and I'm just having my first cuppa.

[CalmBeforeStorm]

http://raleighwx.americanwx.com/models/ ... fLoop.html

6z HWRF deepens Issac to 944 MB just before landfall just west of N.O.

[Dean4Storms]

http://raleighwx.americanwx.com/models/hwrf/06zhwrf500mbHGHTPMSLnesthwrfLoop.html

6z HWRF deepens Issac to 944 MB just before landfall just west of N.O.

But it is initialized a little too far west and starts out at 983mb.

[ronyan]

http://raleighwx.americanwx.com/models/hwrf/06zhwrf500mbHGHTPMSLnesthwrfLoop.html

6z HWRF deepens Issac to 944 MB just before landfall just west of N.O.

They should rename the HWRF and call it the catastrophe model.

Lurker here but I've been watching long enough to know that the HWRF often over does intensity.

[rainstorm]

http://raleighwx.americanwx.com/models/hwrf/06zhwrf500mbHGHTPMSLnesthwrfLoop.html

6z HWRF deepens Issac to 944 MB just before landfall just west of N.O.

But it is initialized a little too far west and starts out at 983mb.

i agree, and its intensity forecast has been rising since it showed 929 yesterday.

[Blown Away]

12z Bams - Big shift east.
12z TVCN - No much of a shift.

[Dean4Storms]

06z GFDL off to TX. This model is lost!

[Janie2006]

UKMET takes Isaac to Gulfport-Biloxi, more on the Biloxi side of things. Perhaps it will end up verifying after all. Some people really thought that model was out to lunch, but the west solution (as opposed to the Appalachicola solution) seems to be the correct one at this point.

[wxwatcher1999]

Anyone think the models will shift left today?

[rainstorm]

06z GFDL off to TX. This model is lost!

does it keep isaac weak and shallow? if so, thats most likely why.

[Ntxw]

This morning's individual GFS ensemble cluster