Global model runs discussion

[Ivanhater]

The long range Canadian is only for the 00z run

[KWT]

Ah ok that does make sense because I remembver in the winter when looking at that model it wouldn't go beyond a certain point on the 12z suite.

[hurricaneCW]

I don't why, but I hate looking at models for possible development. They are so inaccurate on the track, it's intensity, and other factors. These models don't have a clue. The only thing they are worth looking at is trends. Since many are starting to show storms developing, then it means the pattern is becoming more favorable for tropical activity. Still, nothing is set in stone and we still could be looking at nothing 10-15 days from now.

[KWT]

The models I think are useful but they certainly aren't the be all thats for sure. For example very few models developed Bonnie at first and it wasn't till it was actually developing.

[srainhoutx]

Interesting issue concerning the new GFS...

TROPICAL DISCUSSION - INTERNATIONAL DESKS
NWS HYDROMETEOROLOGICAL PREDICTION CENTER CAMP SPRINGS MD
227 PM EDT WED JUL 28 2010

NOTE TO ALL USERS. STARTING AT 12 UTC TODAY...A NEW GFS MODEL CAME
INTO EFFECT. ALTHOUGH THE MODEL SHOWS IMPROVEMENTS ON
PRECIPITATION FORECAST ACROSS MID LATITUDES...WE NOTED SOME
DEGRADATION ON THE WIND FORECAST ACROSS THE TROPICAL ATLANTIC AND
EASTERN CARIBBEAN... WHERE IT TENDS TO UNDER FORECAST INTENSITY OF
THE WINDS IN ASSOCIATION WITH TRADE WIND SURGES BELOW 700 HPA. THE
ERROR/ DIFFERENCE FROM ACTUAL OBSERVATIONS CAN BE AS HIGH AS 20
KNOTS. THIS IS A SYSTEMATIC PROBLEM WITH THE MODEL AND EXPECT IT
TO RECUR.

[KWT]

So infact now the model is probably a little worse in the tropics then the old run based on that...though it doesn't seem quite as keen on developing phantom lows.

[ROCK]

Interesting issue concerning the new GFS...

TROPICAL DISCUSSION - INTERNATIONAL DESKS
NWS HYDROMETEOROLOGICAL PREDICTION CENTER CAMP SPRINGS MD
227 PM EDT WED JUL 28 2010

NOTE TO ALL USERS. STARTING AT 12 UTC TODAY...A NEW GFS MODEL CAME
INTO EFFECT. ALTHOUGH THE MODEL SHOWS IMPROVEMENTS ON
PRECIPITATION FORECAST ACROSS MID LATITUDES...WE NOTED SOME
DEGRADATION ON THE WIND FORECAST ACROSS THE TROPICAL ATLANTIC AND
EASTERN CARIBBEAN... WHERE IT TENDS TO UNDER FORECAST INTENSITY OF
THE WINDS IN ASSOCIATION WITH TRADE WIND SURGES BELOW 700 HPA. THE
ERROR/ DIFFERENCE FROM ACTUAL OBSERVATIONS CAN BE AS HIGH AS 20
KNOTS. THIS IS A SYSTEMATIC PROBLEM WITH THE MODEL AND EXPECT IT
TO RECUR.

well I will be scratching out this model to look at the tropics going forward....

[srainhoutx]

Interesting issue concerning the new GFS...

TROPICAL DISCUSSION - INTERNATIONAL DESKS
NWS HYDROMETEOROLOGICAL PREDICTION CENTER CAMP SPRINGS MD
227 PM EDT WED JUL 28 2010

NOTE TO ALL USERS. STARTING AT 12 UTC TODAY...A NEW GFS MODEL CAME
INTO EFFECT. ALTHOUGH THE MODEL SHOWS IMPROVEMENTS ON
PRECIPITATION FORECAST ACROSS MID LATITUDES...WE NOTED SOME
DEGRADATION ON THE WIND FORECAST ACROSS THE TROPICAL ATLANTIC AND
EASTERN CARIBBEAN... WHERE IT TENDS TO UNDER FORECAST INTENSITY OF
THE WINDS IN ASSOCIATION WITH TRADE WIND SURGES BELOW 700 HPA. THE
ERROR/ DIFFERENCE FROM ACTUAL OBSERVATIONS CAN BE AS HIGH AS 20
KNOTS. THIS IS A SYSTEMATIC PROBLEM WITH THE MODEL AND EXPECT IT
TO RECUR.

well I will be scratching out this model to look at the tropics going forward....

It's not a deal breaker IMO, ROCK. Sort of like the Euro and the SW bias. Each of the models have their quirks so to speak.

[Weatherfreak000]

Exactly. Each model is set to different bias. Consensus being the key word there.


NOGAPS seeing a powerful system in the GOM sure is frightening.

[srainhoutx]

HPC:

ACROSS THE TROPICS...THERE ARE A QUARTET OF SYSTEMS WHICH WERE
COORDINATED WITH NHC AT 16Z...TWO IN THE ATLANTIC AND TWO IN THE
PACIFIC...WITH HPC AND NHC POINTS WITHIN TWO DEGREES...120
NAUTICAL MILES...IN ALL CASES WHICH LED TO A COMPROMISE BETWEEN
THE TWO SETS OF POINTS BEING AGREED UPON. IN GENERAL...THE
CANADIAN WAS THE MOST BULLISH...WITH THE ECMWF THE FARTHEST NORTH
WITH THE WESTERNMOST NORTHEAST PACIFIC SYSTEM.

[KWT]

I wouldn't be worried given the Nogaps is totally on its own with that solution, not exactly the strongest of the models...

[KWT]

The 12z ECM does show a MCS feature diving WSW into the Gulf briefly at 216hrs which is interesting:

http://raleighwx.easternuswx.com/models ... SLP216.gif

[Frank2]

I don't why, but I hate looking at models for possible development. They are so inaccurate on the track, it's intensity, and other factors. These models don't have a clue. The only thing they are worth looking at is trends. Since many are starting to show storms developing, then it means the pattern is becoming more favorable for tropical activity. Still, nothing is set in stone and we still could be looking at nothing 10-15 days from now.

Yes, looking at them can put a knot in a person's stomach (knot as in wind - lol), but then a few days later what was there has since disappeared (and with it the knot), so best to only look at them as needed...

In fact, and per what you mentioned, the NWS forecasters would (and probably still) look at them as just a point of reference for future trends (trending colder, warmer, drier or wetter, etc.) - but usually nothing more...

Once in awhile they might find something of interest many days in advance (the Superstorm of '93 for example), but that's rare...

Frank

[Stormcenter]

It sounds like the "new" GFS model is already broken.

Interesting issue concerning the new GFS...

TROPICAL DISCUSSION - INTERNATIONAL DESKS
NWS HYDROMETEOROLOGICAL PREDICTION CENTER CAMP SPRINGS MD
227 PM EDT WED JUL 28 2010

NOTE TO ALL USERS. STARTING AT 12 UTC TODAY...A NEW GFS MODEL CAME
INTO EFFECT. ALTHOUGH THE MODEL SHOWS IMPROVEMENTS ON
PRECIPITATION FORECAST ACROSS MID LATITUDES...WE NOTED SOME
DEGRADATION ON THE WIND FORECAST ACROSS THE TROPICAL ATLANTIC AND
EASTERN CARIBBEAN... WHERE IT TENDS TO UNDER FORECAST INTENSITY OF
THE WINDS IN ASSOCIATION WITH TRADE WIND SURGES BELOW 700 HPA. THE
ERROR/ DIFFERENCE FROM ACTUAL OBSERVATIONS CAN BE AS HIGH AS 20
KNOTS. THIS IS A SYSTEMATIC PROBLEM WITH THE MODEL AND EXPECT IT
TO RECUR.

well I will be scratching out this model to look at the tropics going forward....

[ROCK]

oh I agree not a deal breaker but still why change out a model for a model that doent work right. It like they are changing it for the sake of saying they did something.....way I see it anyway....


Weatherfreak- I always look at the consensus...

[Riptide]

The news about the para GFS is just sickening. NCEP should run the old model on the parallel site now.

[StormClouds63]

Before it ever emerged off the coast of Africa in 2007, people on this board were already speculating whether Dean was going to be a Texas, Florida, or mid-Atlantic storm. Several days before it formed, everyone knew about big bad Dean. It was one system the models latched onto very early. The ultimate landfall destination, however, varied with the early model runs. Looked like it might even be a fish in the early stages.

[KWT]

Yeah but looking back now Dean should have been a very easy call, we had an utterly monsterous upper high aloft, and whilst we've had a stronger upper high over the Azores its no where enar 2007 levels with Dean and Felix.

[Ivanhater]

18z GFS does show a couple of storms in the long range but quiet until then.

[cycloneye]

Second run of new GFS shows anything significant until day 16. At the rate this is going,the new one may not develop anything in August and September.