[Wrf-users] Suggestions to the topic 1. wrf.exe error

KPAIKPAI Emil emilkpaikpai at yahoo.fr
Fri Aug 12 02:35:52 MDT 2016


Dear khaled,
  Afficher le message d’origine you need to reduce the time step because in your namelist, we can see that you have set the the interval_seconds = 10800 which is equal to hours meaning also that you might have using 3 hourly data if your running the model in real case acording to the wrf user-guide.

In the same user-guide, it was suggested that if DX is the horizontal resolution that you are using, because you have set you have set interval_seconds = 10800 (or because your are using 3 hourly data, your time step should be equal or less than 3xDX. if one is using 6 hourly data then the time step should be 6xDX.Here your horizontal resolution is DX=0.816 time step should be equal or less than 3 times DX (3xDX=<3 x 0.816) 3x0.816=2.448 and correspond to the fraction 306/125so, go to the namelist and choose any positive integer number less than 2.448 and set as the 2 examples bellow
&domains  
time_step                = 2, or 
&domains  
time_step                =1
If you wish to use a decimal number as time step, there are two options for that  (time_step_fract_num and time_step_fract_den), these options help to set the decimal number in form of fraction. In this case the option "time step" should be set to 0 as you can see in the example bellow:&domains                
time_step                = 0,
time_step_fract_num      = 306,      
time_step_fract_den      = 125,
Note that the first line of three seting always use the integer number. here we have switched off the first line to 0 (zero) and then used the fraction number from 2.448 useng the two option we have mentioned before.When you use the integer less than this fraction or a fraction number equal or less than 306/126 it should also work.
The second way to handle with your problem is to try using the command "ulimit -s unlimited" just after you have lunched wrf.exe
If none of the suggestions doesn't work, may you try setting in the namelist DX=DY=0.816 
BATEBANA


   

  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/wrf-users/attachments/20160812/50518156/attachment.html 


More information about the Wrf-users mailing list