: Guide To 'Model Run' Times -

Jump to content

Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

Guide To 'Model Run' Times

#1 User is offline   Uskys 

  • Group: Forum Managers
  • Posts: 26329
  • Joined: 22-November 02
  • LocationBrecon Beacons, 330m ASL

Posted --

Model 'Run' - Types & Times

Dan G
Posted 22/3/2008 22:54 (#314689 - in reply to #305492)
Subject: Re: Model Chat Thread - Looking at a wintry Easter?
Quote Reply Alert Edit Delete
Online

Location: Caversham, Reading

GFS initialises four times a day, 0000z, 0600z, 1200z and 1800z. Products start coming out about 3.5 hours after initial time, early forecasts first. Forecasts run to T+384 hours.

I think the Met Office Unified Model runs for times a day, but they only publish data from the 0000z and 1200z runs (I think the intermediate runs are shorter). Run to 144 hours. The "fax" charts are based on this data but are tweaked by human forecasters, often using input from other models. Four fax analysis charts go out a day, and 12 hourly charts go out to +120 hours. The production order is a little odd, mind you.

ECMWF runs twice day too, out to 240 hours. Rather limited data available only, and runs come out a good while after initialisation, around 7-8 hours later. Worth waiting for though as it operates on a 25 km grid, nearly twice the resolution of the GFS and UKMET (both 40ish km).

ECM and UKMET both use slightly superior analysis techniques to the GFS, but the latter did get a major upgrade last year that closed the gap considerably. The initial analysis is important because if a model doesn't have a good understanding of what's happening now, it's forecasts aren't going to be much cop...

Rough schedule (BSL is Meteoblue now) (all times zulu)

0400 GFS 00z
0530 Fax 00z an, 24, 36, 48
0600 UKMET 00z
0630 Fax 00z+60, 72
0800 ECM 00z
0800 BSL 00z
0930 Fax 06z an, 00z+84
1000 GFS 06z
1500 Fax 12z An
1600 GFS 12z
1630 Fax 12z+24
1700 Fax 12z+36, 48
1800 UKMET 12z
1830 Fax 12z+60, 72
2000 ECM 12z
2000 BSL 12z
2100 Fax 18z an
2200 GFS 18z
2130 Fax 12z+96, 120

-----
ESTOFEX • Report severe wx • Storm forecasting guide


0

#2 User is offline   Uskys 

  • Group: Forum Managers
  • Posts: 26329
  • Joined: 22-November 02
  • LocationBrecon Beacons, 330m ASL

Posted --

Captain Caveman
Posted 22/3/2008 23:51 (#314699 - in reply to #314696)
Subject: Re: Model Chat Thread - Looking at a wintry Easter?
Quote Reply Alert Edit Delete
Online




Location: UK
Uskys - 22/3/2008 23:43
Captain Caveman - 22/3/2008 23:28

 

Uskys - 22/3/2008 21:39 OK. thanks Dave - Its just I thought the meto fax was taken from GFS. I could really do with an idiots guide to the model runs!!

 

 

GFS!?!?!

Perhaps I should take the time to explain the process of producing the fax charts at some point - when I am not on a run of shifts.

 

For dollops like me, yes (please) ........ but it can wait - whats a few more weeks? it has waited 6 years so far!! .. btw- I didn't mean I understood the fax charts were GFS - but that the 18z GFS had a part to play in them... but I'm just digging myself a whole- I either need to get myself off to reading for a few years or shut up :-)

 

As a very brief summary:

The 'fax' charts are generated using all available data, operational runs and ensembles - the initial NWP field will be NAE or GM, but is then modified to varying degrees using a tool called metmorph to provide the most probable outcome. The modified fields from T+36 are generated by the Deputy Chief, the ASSX and FSXX (ASXX obviously incorporating actual observations) by the Chief.

I'll try to remember to add more detail later - and add a little to the NWP outputs times and chart availability times when I have more time. 



Edited by Captain Caveman 23/3/2008 00:02


(Message has 1 revisions)


-----
Steve.
Lowest temp of winter 2007/8 so far -4.8C.
Rockbeare Weather

0

#3 User is offline   Dave W 

  • Group: Executive
  • Posts: 5348
  • Joined: 17-October 03
  • LocationBrighton

Posted --

Also available and of use sometimes, if carefully used, from KNMI is the ensemble forecast of all the ECMWF runs for one location, De Bilt (Holland). It can give a general overview of what each member was suggesting for the location for a number of parameters out to 15 days ahead: http://www.knmi.nl/e...agse/index.html

Other models are also available regularly but usually they are less reliable than those above as there resolution is lower, though the German model (GME) out to HH132 is not at all bad. These include NOGAPS (USA Navy), GEM (Canadian) and JMC (Japanese).. all run twice a day except NOGAPS (4 times a day now).

See Top Karten on Wettercentrale or Meteociel for access to these.. http://www.meteociel.com/index.php then 'Models' on left hand side menus

0

Share this topic:


Page 1 of 1
  • You cannot start a new topic
  • You cannot reply to this topic

1 User(s) are reading this topic
0 members, 1 guests, 0 anonymous users