Sick of GFS only Mets

[Valkhorn]

OK. I am getting sick and tired of local mets and regional mets who only use the GFS for guidance. I am seeing some 10 day forecasts in regions that are so way off it's not even funny, and all they are doing are using the GFS, and the GFS alone.

Past 84 hours, the GFS is clueless.

[powderfreak]

I think the GFS actually isn't that bad up to a week out. I've noticed with east coast storms, it often picks up on them in the D+8 and 7 range but loses them in the 3-5 days out term, only to pick back up on them 48 to 72 hrs out.

Think about it, but I know I've seen it happen numerous times.

I'll totally agree with you that in the long range and out past 7 days it is totally worthless...although for some reason the smart guys at TWC and Accuweather use the GFS for all of their long range needs.

[Valkhorn]

I think the GFS actually isn't that bad up to a week out. I've noticed with east coast storms, it often picks up on them in the D+8 and 7 range but loses them in the 3-5 days out term, only to pick back up on them 48 to 72 hrs out.

It's allright on the level of storminess or not, to a point. However the temperatures are generally way off. If we believed the GFS back around in early December there would have been a ridge in the Northeast instead of a MECS. Also if we believed the GFS we wouldn't have had near 0 and sub zero readings in NYC and Boston tonight.

I'll totally agree with you that in the long range and out past 7 days it is totally worthless...although for some reason the smart guys at TWC and Accuweather use the GFS for all of their long range needs.

The smart guys at TWC and Accuweather... hehe actually they don't. JB rarely uses the GFS for anything except if there's nothing else to fish for. The forecasts are generated by the GFS, which is why they swing around so much from time to time. Actually all the local forecasts there are automated, so I wouldn't take their long range forecasts with any grain of sincerity.

Many of the OCM's at TWC read off of the script and maps they are given, and rarely do they mention anything out of the ordinary that their maps aren't showing. The experts like Steve Lyons and Paul Kocin do jump out of the way of the sheep though, so I do give them many kudos for that.

[stormraiser]

I think it may vary by region. NWS Buffalo uses it quite often and generally have been righ.

[powderfreak]

Yes, I know JB rarely uses the GFS for anything, and that Kocin is the man to turn to on TWC with anything related to winter, but JB and Kocin do not make the forecasts for Accuweather and TWC. I'm not sure how TWC runs things, but I know JB does have imput into what happens at Accuweather. Yet, whenever you look at their graphics (the cute little maps they put out explaining to the public exactly what is going on) they almost always follow the GFS. If you want a storm track map, the GFS goes up. Rain snow line? GFS.

Now, how about the NWS? Read their long range discussions and no matter how outta wack the GFS is, they always either blend it with something, or go solely with it. Never have I heard them say we are disregarding the GFS and going with a Euro/Ukmet blend. It's always GFS blended with something or solely the GFS.

[Valkhorn]

Now, how about the NWS? Read their long range discussions and no matter how outta wack the GFS is, they always either blend it with something, or go solely with it. Never have I heard them say we are disregarding the GFS and going with a Euro/Ukmet blend. It's always GFS blended with something or solely the GFS.

I hate to be the conspiracy theorist but that could be because the GFS is the american model, and the NWS is a government agency. Put those together.

[powderfreak]

I realize this as well...but wouldn't you think that regardless of where their funding comes from they would pick the superior model? Oh wait, what am I talking about....it is the US Gov't we are talking about...I guess they want money or they want to play up their own model instead of a quality forecast.

[FLguy]

the ECMWF normally has much higher skill in the MR than the GFS. and there are several times when the ECMWF and ETA are in perfect agreement as one can reasonably expect within 3 days of an event and the GFS completely clueless.

quite frequently the operational GFS will not only be at war with the other operational MR and SR models (ETA...GGEM...UKMET...and so on) but in complete disagreement with its own ensembles.

there can even be times where the operational run is completely different than any of the other ensembles. this is why the GFS operational runs (at 0z and 12z) should be checked against its ensembles before used in making any forecasts.

the same goes for the canadian operational model and ETA (check against the ensembles to see what if any support the operational runs have). if the ensembles show one thing and the operational runs show something completely different...the operational run is most likely incorrect.