2012 Long Range Model Discusssion (> 7 Days Out)

[Jeff Duda]

Seems like the GFS is beginning to show a more consistent trend for a pattern shift towards the end of May here. Today's 12Z control run and the ensemble members are all showing multiple bouts of troughiness showing up across the Rockies starting early next week. The ECMWF is also starting to show a trend towards a trough, but the character is way different. The trough in the ECMWF has less curvature, larger wavelength, and seems more strung out. However, the height gradient would indicate pretty high-speed flow. FIM graphics have conveniently stopped coming out past hour 168 or so.

 

[Warren Faidley]

If (a big if) the 12z GFS verifies, or comes close, Memorial Weekend will be wild.

W.

 

[Michael Snyder]

If (a big if) the 12z GFS verifies, or comes close, Memorial Weekend will be wild.

W.

The GFS sure is all over that huge deep trough, it would be nice to split the difference between the Euro and the GFS at this point.

 

[John Sickels]

I hereby predict that a week from now everyone will be gushing over what will look like a multiple-day outbreak in the May 25-30 range.

I'm not a trained met and even after years of study I barely understand this stuff, but I refuse to get depressed (or excited) over anything the models predict more than 4-5 days out, especially this time of year.

 

[Kelly Sugden]

I'm predicting that the 12Z GFS is showing a too quick solution with the UL trough/pv anomaly. So take whatever day looks best and move it back by one day. Prefer the 12Z ec slower solution given the NH teleconnection pattern. Regardless, yeah, the general trend for a more active period towards the end of the month is there. Hopefully the 500 hPa pattern won't dampen with time (as suggested by old ec runs for this weekend which now the GFS has latched on to). MJO forecast is to move from the eastern hemisphere to the western hemisphere which might also reinvigorate the STJ. Who knows though, I'm guessing much like previous posts as this point. Just because the GFS shows western troughing isn't a reason to all hopeful.

 

[Michael Snyder]

The 18z gfs really took a turn towards the euro here.
Maybe we will get some run to run continuity going now....

 

[Derek Weston]

The 18z gfs really took a turn towards the euro here.
Maybe we will get some run to run continuity going now....

The 18z GFS is the most unstable thing known to man. It can never be trusted. It performs worse than random. Sure, it's synced up right now with the euro, but earlier in the week it was in sync with the breeding cycles of south american bumble bees and beef jerky production in Singapore. So sure, maybe it will lead you to tornadoes... but maybe it'll lead you to bumble-jerky bees. And you don't want pictures of that. Trust me.

 

[Michael Snyder]

Well the 00z GFS verifies same as the 18z, timing is very close to the Euro.
Waiting for the 00z euro before go ahead and make some plans for the 23rd.

Sucks to have to plan out so far on this chasecation this year.

 

[Kelly Sugden]

The off synoptic time models (06Z/18Z) also do not have any raob data to ingest. As Derek mentioned, its best to stay away from them, particularly in the long range.

 

[Michael Snyder]

The off synoptic time models (06Z/18Z) also do not have any raob data to ingest. As Derek mentioned, its best to stay away from them, particularly in the long range.

Which is why it was helpfull that the 00z run verified.

Euro slowed down the trough and now exits it into the plains later than the GFS.

Its what makes model watching half the fun.

 

[Kelly Sugden]

Which is why it was helpfull that the 00z run verified.

Euro slowed down the trough and now exits it into the plains later than the GFS.

Its what makes model watching half the fun.

How do you know last nights 00Z run verified? I personally would have to wait for the end of the model run and in the future and compare 500 geopotential heights with those measured from raobs. Look at the skill scores recently of the GFS tanking to around 0.60 within the last two months. Yes, I look at this stuff and is why I constantly insist that the GFS is a piece of garbage. I refuse to get excited about one model perturbation which could be an outlier from its ensemble members.

 

[Michael Snyder]

How do you know last nights 00Z run verified? I personally would have to wait for the end of the model run and in the future and compare 500 geopotential heights with those measured from raobs. Look at the skill scores recently of the GFS tanking to around 0.60 within the last two months. Yes, I look at this stuff and is why I constantly insist that the GFS is a piece of garbage. I refuse to get excited about one model perturbation which could be an outlier from its ensemble members.

You are acting like the Euro isnt showing the same thing, and the NAM and canadian line up well as far out as they go.
I explained that I have to plan my chase almost a week out and the 00z GFS verified what the 18z GFS said, thats it.

I understand fully that implications of looking at GFS data 6 days out, but we are discussing long range forecasts here.

The Euro and GFS are all on board with the trough and pattern change.

Its definitely worth discussing.

 

[Kelly Sugden]

You are acting like the Euro isnt showing the same thing, and the NAM and canadian line up well as far out as they go.
I explained that I have to plan my chase almost a week out and the 00z GFS verified what the 18z GFS said, thats it.

I understand fully that implications of looking at GFS data 6 days out, but we are discussing long range forecasts here.

The Euro and GFS are all on board with the trough and pattern change.

Its definitely worth discussing.

Yes, it is worth looking at.
No, a model does not verify because it comes into consensus with another model. Verification is based on reality. Also the ecmwf is much slower.

 

[Michael Snyder]

Yes, it is worth looking at.
No, a model does not verify because it comes into consensus with another model. Verification is based on reality. Also the ecmwf is much slower.

The 00z verified what the 18z (the run without all the raob data) said. And there is nothing wrong with verifying one run of The GFS with another to check continuity.

I was talking about the GFS, not consensus between differing models.

You dont check to verify the continuity of the model runs? Sure you do.

The ECMWF is slower yea, but the pattern change is coming. I just hope im in the right place!

 

[Michael Snyder]

Yes, it is worth looking at.
No, a model does not verify because it comes into consensus with another model. Verification is based on reality. Also the ecmwf is much slower.

Maybe ill use the word compare instead of verify from now on.