Hi Ufuk,
   If there is no error in time step for NaN error then this may be due to data. Sometimes while downloading, data may be corrupted. So, you try downloading data and run the model and see whether it is running without NaN error. I faced this problem in the middle of my run for some years.


Regards,
Rajalakshmi.D,
M.Sc(Ag.Met),
Junior Research Fellow,
ACRC,TNAU,
Cbe-3.
--- On Wed, 31/3/10, Ufuk Utku Turuncoglu <u.utku.turuncoglu@be.itu.edu.tr> wrote:

From: Ufuk Utku Turuncoglu <u.utku.turuncoglu@be.itu.edu.tr>
Subject: [RegCNET] NaN error at the beginning of simulation ...
To: "RegCM Users" <regcnet@lists.ictp.it>
Date: Wednesday, 31 March, 2010, 12:08 PM

Hi All,

I try to run RegCM using HadCM data but when i run the model it gives the NaN value at the beginning of the simulation.

BATS variables written at   1960010100  0.500000000000000
Writing rad fields at ktau =            1  1960010100
     at day =    0.0170, ktau =         50 :  1st, 2nd time deriv of ps =  NaN         NaN        ,  no. of points w/convection =       0
     at day =    0.0344, ktau =        100 :  1st, 2nd time deriv of ps =  NaN         NaN        ,  no. of points w/convection =       0
     at day =    0.0517, ktau =        150 :  1st, 2nd time deriv of ps =  NaN         NaN        ,  no. of points w/convection =       0

I know that normally this problem related with time step but in my case i use same time step value (60 sec) with other datasets such as  ECHAM5 and CCSM3 without any problem. I also check the created ICBC files but i could not see any problem. I think that the problem is in somewhere else. Can you give me the detailed information about calculation of this value? I check the source code and the variables ptntot and pt2bar related with this values. What is the definitions of these variables? What kind of possible sources (dataset ???) can cause this NaN error? Any comments can be helpful.

Best regards,

--ufuk


-- This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

_______________________________________________
RegCNET mailing list
RegCNET@lists.ictp.it
https://lists.ictp.it/mailman/listinfo/regcnet


The INTERNET now has a personality. YOURS! See your Yahoo! Homepage.