WRF 2 and 3 runs

Forum dedicated to older versions of EMS package (WRFEMS v3.2, v3.1 or older). Support is user-to-user based, so please help others if you can.
Post Reply
Laznik
Posts: 24
Joined: Sat Aug 29, 2009 8:38 am

WRF 2 and 3 runs

Post by Laznik » Tue Oct 13, 2009 2:24 pm

Hy to all

I have encountered some strange problem. I have been running model over Slovenia with resolution of 3km on wrf 2 version. Now i have created identical domain with all characteristics as older one. What i have found out is that times that takes to calculate both domains are significantly different. Wrf 3 calculates more than 10 hours longer, is that normal??? Should i change some conf files??

help please

jost

wettersat
Posts: 32
Joined: Fri Aug 21, 2009 9:30 pm

Re: wrf 2 and 3 runs

Post by wettersat » Tue Oct 13, 2009 6:35 pm

Jost, v3 uses 45 levels as default (v2 only 31 levels). Therefore I have about factor 1.5 in speed.
Gerald

Laznik
Posts: 24
Joined: Sat Aug 29, 2009 8:38 am

Re: wrf 2 and 3 runs

Post by Laznik » Tue Oct 13, 2009 6:59 pm

Thanks G.

I am very worried about because my factor is more than 2 :P

;)

meteoadriatic
Posts: 1519
Joined: Wed Aug 19, 2009 10:05 am

Re: wrf 2 and 3 runs

Post by meteoadriatic » Tue Oct 13, 2009 10:07 pm

This should not be the case. I found v3 much faster than v2, mostly because nesting is simultaneous. Are you really sure your domains are more or less same in size/number of grid points? Did you maybe swiched on DFI or something similar that could prolong calculation times? What physics options do you use? Is that NMM or ARW domain?

Laznik
Posts: 24
Joined: Sat Aug 29, 2009 8:38 am

Re: wrf 2 and 3 runs

Post by Laznik » Thu Oct 15, 2009 3:14 pm

hY

I think DFI is not activated, ... where can i check if it is or not, because i did not make any changes on conf files?
I use NMM core.

meteoadriatic
Posts: 1519
Joined: Wed Aug 19, 2009 10:05 am

Re: wrf 2 and 3 runs

Post by meteoadriatic » Thu Oct 15, 2009 3:39 pm

In domain conf directory, file run_dfi.conf but if you can' remember you changed that then it is OFF.

I don't know what could be, no doubt, v3 NMM should be faster than v2. I had problems with precise making domains/nests, are you sure you didn't end up with much larger domains than you think? Did you check for example in GrADS the complete, unrestricted (in respect to lat/lon) graphic output, IE start grads, don't specify 'set lat ...' and 'set lon ...' limits, load your domain ctl file, draw something and see how large your domains really are.

Laznik
Posts: 24
Joined: Sat Aug 29, 2009 8:38 am

Re: wrf 2 and 3 runs

Post by Laznik » Thu Oct 15, 2009 4:44 pm

I did that tests and domains are totally the same size and resolution.... i will do more tests in comming days.. thanks

meteoadriatic
Posts: 1519
Joined: Wed Aug 19, 2009 10:05 am

Re: wrf 2 and 3 runs

Post by meteoadriatic » Thu Oct 15, 2009 4:59 pm

Hm... another idea, are all your CPU cores are running?

Laznik
Posts: 24
Joined: Sat Aug 29, 2009 8:38 am

Re: wrf 2 and 3 runs

Post by Laznik » Fri Oct 16, 2009 5:08 am

Yes! ... .. I will try more runs today and report how did go.

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests