ATL: EARL - Models

[OuterBanker]

Of course the real problem is if they are wrong. It would cause most here to ignore Earl and go about their business and relax too much.

I guess what bothers me most about models is that we have become so reliant on computer models that we take meteorologist out of the picture. The question anymore is whether we rely on computer programmers more than meteorologist. We don’t even consider their “gut” feeling from many years of experience.

[TwisterFanatic]

GFS 24 hours...




48 hours....

[ConvergenceZone]

By looking at just how close these models are to the east coast, I would
be very nervous...
Just look at close those models re to the north eastern states, yikes!!!

http://www.wunderground.com/tropical/tracking/at201007_model.html

[TwisterFanatic]

60 hours....



Looks like its gonna be REALLY close.

[Duddy]

Keep posting that GFS output please. I don't know where to find them as they come out.

[TwisterFanatic]

72 hours...



84 hours...

[StormGuy]

Too close for comfort, that's for sure.

[TwisterFanatic]

102 hours...



Curving out to sea

[ConvergenceZone]

Nobody can say with certainty that this is not going to hit the east coast that's for sure.
All it would take is for the high pressure that's to the east of it to shift just
a little westward and we have potentially a cat 3 to cat 4 storm slamming into the northeast states.
Since it's forecasted to be THAT CLOSE to the states, nobody can call fish...

[hurricaneCW]

That trough looks very strong and has a negative tilt. Could it potentially try to suck the storm towards it rather than a weaker trough that allows for a basic NE motion. I assume that it have to be much further south and stronger.

[HurrMark]

GFS trend is looking better...seems it is depressed a bit further south than 12Z. Not saying the NE is out of the woods, but the westward trend has at least momentarily stopped...

[BigA]

NOGAPS has consistently been shifting west. 18Z run is only 30-50 miles from the Outer Banks

[Wthrman13]

Of course the real problem is if they are wrong. It would cause most here to ignore Earl and go about their business and relax too much.

I guess what bothers me most about models is that we have become so reliant on computer models that we take meteorologist out of the picture. The question anymore is whether we rely on computer programmers more than meteorologist. We don’t even consider their “gut” feeling from many years of experience.

I completely agree that this is certainly a danger, and while there are many forecasters who hug the models too much, I don't think the situation is as grim as you might think. Many of the operational forecasters I talk to and work with day in and day out are quite aware of the importance of the human element to any final forecast. Humans excel at pattern recognition and integrating many disparate sources of data together to come up with a final integrated forecast that consistently beats the models. Nevertheless, without the models, this forecast would suffer immensely, and thus ignoring or downplaying the importance of models in the forecast process is just as, if not more, dangerous as relying too much on them. It's safe to say that despite all their flaws, numerical models are the backbone of modern forecasting. However, it still takes a human (and will for the foreseeable future) to interpret and integrate all the output that they produce.

Despite the fantasies of some, there was no great forecasting "golden age" before the advent of computer models. The statistics clearly show that the models are forecasting better than ever before, with 3 day forecasts as good as 1 day forecasts were 20 years ago, and so on.

[thetruesms]

I completely agree that this is certainly a danger, and while there are many forecasters who hug the models too much, I don't think the situation is as grim as you might think. Many of the operational forecasters I talk to and work with day in and day out are quite aware of the importance of the human element to any final forecast. Humans excel at pattern recognition and integrating many disparate sources of data together to come up with a final integrated forecast that consistently beats the models. Nevertheless, without the models, this forecast would suffer immensely, and thus ignoring or downplaying the importance of models in the forecast process is just as, if not more, dangerous as relying too much on them. It's safe to say that despite all their flaws, numerical models are the backbone of modern forecasting. However, it still takes a human (and will for the foreseeable future) to interpret and integrate all the output that they produce.

Despite the fantasies of some, there was no great forecasting "golden age" before the advent of computer models. The statistics clearly show that the models are forecasting better than ever before, with 3 day forecasts as good as 1 day forecasts were 20 years ago, and so on.

Very eloquently put. I would take an experienced forecaster with an arsenal of models at his disposal than just an experienced forecaster any day.

Although there is one interesting perspective I heard from an HPC forecaster regarding not just pattern recognition and data synthesis with weather conditions, but in regards to the model. He was wondering if perhaps models were being updated too frequently. He remembers his earlier days when a model such as, say, the NGM would not necessarily be accurate, but he had become so familiar with how the model worked and how it was likely to be biased that he still had a good idea of what was going to happen. Now, he's not always sure how behavior will change from one model iteration to the next, since they're upgraded so rapidly. I think he was more musing about it than making any particular argument, but it was an interesting thought. The recent GFS upgrades brought that thought to mind again. However, it seems that forecasters were able to seize, almost instantaneously, on certain biases that I found it rather encouraging.

[Ivanhater]

HWRF possible landfall Outer banks

[TheDreamTraveler]

18z HWRF

Seems Earl has struck land

[BigA]

That's 2 HWRF landfalls-Cape Hatteras and Cape Cod. Actually the Hatteras may not be a technical landfall, but if the eyewall goes over you that's little consolation.

And given the last forecast point, probably 3, the last near Eastport, ME.

[Hybridstorm_November2001]

18z HWRF

Seems Earl has struck land

Does the Eye actually go across Hatteras, or just clip it? Wondering 'cause I honestly can't tell in that image, looks like maybe a clip to me...

[Annie Oakley]

HWRF possible landfall Outer banks

Thank you Ivan.

[dixiebreeze]

30/2345 UTC 19.6N 65.2W T6.5/6.5 EARL -- Atlantic