Global Model Runs for W.Caribbean Early Season Development

[Category 5]

Never fails, it's always the GFS or the Canadian. Call me when a REAL model tries to develop something.

SEASON CANCEL

...and we don't need extra preparedness for the -removed- that is routinely encouraged for every thunderstorm, low level trade cumulus, or SHRA in the Atlantic basin.

And amen to that. The scary thing is it's only May.

[Vortex]

5/11 6Z GFS back on with carribean development..Of course this far out "Entertainment Only"

http://www.nco.ncep.noaa.gov/pmb/nwprod ... n_384l.gif

[MiamiensisWx]

Actually, the operational GFS and its ensembles are one of the better models, in addition to the ECMWF (Euro). The problems and these types of threads occur when people search for development beyond five or six days. Additionally, you should never expect tropical cyclogenesis to definitely take place in May, unless it's an Ana type situation; the latter scenario is the only plausible possibility, since it is a subtropical North Atlantic system. Most of the time, you should never anticipate purely tropical development in May.

[gatorcane]

384 hours out is entertaining but way too far out to draw any conclusions. What it is suggesting, however, is that the start of the hurricane season is nearing with global models starting to already develop areas in the Western Caribbean, albeit with nothing at all we need to worry about at the moment.

[Sanibel]

Watching models in early May?

[curtadams]

It would be interesting to see if any of the models are biased in predicting cyclogenesis. If they were fair, then the chance of a cyclone being predicted in a given 384 hour period should be equal to the actual historical chance over that period (perhaps adjusted for the recent high activity levels). Even then you'd get a lot of false alarms because you get a new set of predictions every 6 hours and while there's some correlation you'd still see more storms predicted than actually happen (made up for by the fact that phantom storms aren't predicted every time the model is run). Even accounting for the multiple run effect I get the impression the GFS is biased towards -removed- (I KNOW the CMC is!). But it would be nice to know for sure.

[SETXHurricane]

It has been extremely breezy here in SETX. We have had constant winds every day. Some recent gusts up to 35 mph. Anyone know why - normally it is hot and muggy here - in May.

[KWT]

To be fair to the GFS whilst it still does create phantom storms (though often it develops non-tropical systems that people assume are tropical just because of those phase diagrams...) its not as bad as it once was with this problem, indeed at one point it was as bad as the CMC with regards to such systems.

[flwxwatcher]

To be fair to the GFS whilst it still does create phantom storms (though often it develops non-tropical systems that people assume are tropical just because of those phase diagrams...) its not as bad as it once was with this problem, indeed at one point it was as bad as the CMC with regards to such systems.

Your right.. the GFS a few years back was as bad as the Canadian with it's phantom storms. Since they fixed a few problems it's done a lot better. We will see what this year brings.

[Ed Mahmoud]

Beyond 180 hours, the resolution of the GFS is halved to save computing time, meaning the accuracy of the model suffers in a big way. The GFS beyond 7 days is useless for anything but possible trends.

[Vortex]

6z 5/13 GFS-Gotta love it


http://www.nco.ncep.noaa.gov/pmb/nwprod ... n_384l.gif

[KWT]

Yep that looks like a classic phantom GFS storm right at the end of the run, signs that the atmosphere is becoming progressivly more condusive for systems in the SW Caribbean however, even if the individual storms aren't gonna happen.

[Dean4Storms]

Are we already in Bear mode?

[Eyewall]

It's possible the gfs may be on to something there have been many stormes that have developed in the first week of the hurricanes season. Looking back at the last couple of years.

[Ad Novoxium]

Only 05 and 06 really had strong first-weeks activity. In 07, both Andrea and the disturbance that ultimately became Barry formed in May (with Barry lasting only about a day or two), and Chantal didn't form until July.However, if we ignore the dark horses (Alex, Alberto of 00, and Arthur), it is noticeable that June activity seems to be coming to a head of sorts.

Speaking of, does anyone know if the 04 and 00 seasons had El Nino? (Though I doubt it.) I know 02 did and 06 had a late one.

[boca]

The GFS model also picks up on lower pressure down there thats why the model develops phantom storms this time of year.The law of averages comes into play because the GFS is bound to be right sooner or later.

[Eyewall]

What about the historic Tropical strom Allison in 2001 that formed at the begining of June in the Gulf of Mexico.

[flwxwatcher]

The GFS model also picks up on lower pressure down there thats why the model develops phantom storms this time of year.The law of averages comes into play because the GFS is bound to be right sooner or later.

Exactly, The GFS picks up on the lower pressures down in the SW Caribbean and thats why it tries to develop in the longer range these storms this time of year. It's interesting to watch but this time of year I am believe it when I see it mode in regards to any tropical development.

[KWT]

Well the SW Carribean/gulf has historically been the place to look, esp in recent years. As others have said the fact that there tends to be lower pressures in that area probably helps to explain this.

[boca]

The funny thing is when I look down in that area and don't see any clouds within 500 to 1000 miles of the SW Caribbean and the GFS model is picking something up thats definetely a phantom storm.If their were convection developing down there then it would be believable. Its getting close though.