2017 Global Model Runs Discussion (Out to day 16)

[cycloneye]

Oh boy GFS.

 https://twitter.com/RyanMaue/status/866453529946357760

[Hurricaneman]

The GFS is embarrassing with it's verification and doesn't just need an upgrade but a complete overhaul, its as bad as the Canadian

[blp]

The problems with the GFS are unfortunate. What I like most about the GFS is how often it runs and how accessible the model is to the public. IMO has been at the second tier to Euro but still a reliable model.

We need to have at least two reliable models as we head into the start of the season 9 days away so this quite frustrating timing. Hopefully they do something quick to get this resolved.

[Hammy]

GFS verification there is beyond pitiful. I hope this isn't the updated model otherwise there are going to be major forecasting errors when we get into the heart of hurricane season.

[TheStormExpert]

Dang!!! That is almost rock bottom for the GFS, what a waste of upgrades it seems over the years. At this point I'd much rather have a dead season like 2013 just to keep us from going crazy.

[Kingarabian]

GFS verification there is beyond pitiful. I hope this isn't the updated model otherwise there are going to be major forecasting errors when we get into the heart of hurricane season.

The parallel GFS is supposed to replace it. But the NHC protested it citing it did worse than the OP GFS.

[Hurricaneman]

According to wheat I'm seeing the parallel version is better and has fewer phantom storms

[RL3AO]

The 12Z run from the 14th, which was the one that bottomed out was no longer available on Tropical Tidbits. So here is the 7 day forecast from the 0z GFS on the 15th and the 6h forecast for the 18z GFS today (e.g. both verifying the 22nd at 0z)

Just wanted to give an idea of what a really bad forecast looks like. It's probably not nearly as bad as you would think.

[Yellow Evan]

GFS verification there is beyond pitiful. I hope this isn't the updated model otherwise there are going to be major forecasting errors when we get into the heart of hurricane season.

The parallel GFS is supposed to replace it. But the NHC protested it citing it did worse than the OP GFS.

IIRC that's only in the EPAC.

[Kingarabian]

True. You know how in sports its known that stats dont paint the whole picture. The same applies for models I guess.

I mean the Euro has been flip flopping showing storms left and right in the Epac yet nothing materializing.

[Alyono]

MU going for the whole hog now. Trying to kill 1 million in Bangladesh with a cat 5 equivalent

http://www.tropicaltidbits.com/analysis ... 212&fh=192

[AutoPenalti]

MU going for the whole hog now. Trying to kill 1 million in Bangladesh with a cat 5 equivalent

http://www.tropicaltidbits.com/analysis ... 212&fh=192

that's horrible.

[cycloneye]

We had hoped that the GFS Parallel was better but is another fail for NCEP. Big upgrade or overhaul is needed.

 https://twitter.com/RyanMaue/status/866705082599059456

[RL3AO]

 https://twitter.com/TropicalTidbits/status/866711019174887425




Levi showing that even the GFS in week 1 has a bias in the SW Caribbean where surface pressures are too low.

[RL3AO]

12Z Run from the 17th verifying today is the worst skilled run since last July.
 https://twitter.com/RyanMaue/status/866722442680692736

[cycloneye]

And the failures keep comming for NCEP.

[RL3AO]

The GFS (and it some ways the ECMWF) is just terrible with handling the breakdown of mid/high latitude blocking patterns like we saw last week. It's a bad forecast in the polar regions, but when you look at the mid-latitudes and tropics, it still has the general synoptic pattern correct.

[RL3AO]

Where most people live, the GFS did fine. It generally has the positions of the mid-latitude troughs and ridges correct. It was in the high-latitudes that it struggled by wanting to keep the big blocking patterns in place. The numbers are bad, but when you actually look at the forecasts I think calling for the GFS to be thrown into the garbage is a little over the top.

[cycloneye]

NEW GFS UPGRADE ON JULY 12

Let's hope this big upgrade fix it in a definite way.The link says June 20 but it has been pushed back to July 12.Let's see if they dont have more pushbacks on the date.
------------------------------------------------------------------
http://www.emc.ncep.noaa.gov/gmb/noor/G ... FS2017.htm

 https://twitter.com/EricBlake12/status/867430996064522240

[Alyono]

why do we even take the MU seriously given how bad it verifies and its run to run consistency. See what it is doing in the BOB now. Absolutely no run to run consistency from it. It is absolutely worthless