Gfs Accuracy?

[johnholmes]

hi

Bored with the weather at the moment? Its funny what one does to fill in the time.

I was browsing around the very extensive Extra site on Net Weather, it really is worth at least a look for a month, you will then be hooked. Okay end of commercial and honest I don't get a penny from Paul.

I came across this

http://www.net-weather.co.uk/secure/cgi-bi...0ab2d090b1b99ad

I used to do a Week Ahead Forecast for Net Weather. This is probably the last one, although I'm not sure.

Done on 22/12/05. No White Xmas but I did point the way to what would happen soon after Xmas. I take no credit but its GFS, which at the time was streets ahead of any other model, for 90% of the time.

Those of you on the forum in 2004 may remember me being foolish enough to suggest what Xmas 2004 might be like SIXTEEN days before the day. Again thanks to GFS.

Anyway the Wetterzentrale charts for the actual charts shows just how close GFS was 5 days out to what happened.

Quite what has happened to it since is a mystery. Blocked situations are not its forte, obviously.

Enjoy the read and memories.

here are the Wz charts for around that time.

27/12/05 850 mb chart

27/12/05 surface chart

John

[summer blizzard]

Now that the major blocking has ended, it appears that the GFS is back to its old ways, it has picked up on a major storm and northerly thereafter and stuck with it since 288 hours out, there is still time for change, however i am very impressed.

[johnholmes]

time will tell SB, a common enough comment from me. The work that TWS is doing should show interesting variations over even the relatively short time he has been doing it. Mind you I would be surprised if TWS did not feel he had been doing it a very long time.

Oh for a new pc and the ability to see those Excel data sets. One day I will!

John

[Thundery wintry showers]

The GFS was a poor third to ECMWF and UKMO during March, but has improved since. In recent months there has been little to choose between GFS and ECMWF with UKMO third.