[MITgcm-support] abnormal end

[Martin Losch]

Maxiaohui,

I think that the error message is quite clear:
> (PID.TID 0000.0001) *** ERROR *** S/R INI_THETA: theta = 0  
> identically. If this is intentionalyou will need to edit  
> ini_theta.F to avoid this safety check.
Your theta is 0 (somewhere). Please have a look at "S/R INI_THETA" to  
see what's going on in detail. How do you find this subroutine? As a  
rule of thumb, the file name is the same as the subroutine (S/R)  
name, but that does not always work. So I usually change into the  
"build"-directory and type "grep -ni ini_theta *.F", which will give  
you the files with extension .F in which "ini_theta" appears (see  
"man grep" for more details, "grep" is very useful!!!!). One of them  
will be the one where the subroutine is defined. Then open this file,  
find the line with the error message, and you'll see, why the routine  
stops (because theta(i,j,k,bi,bj)=0. at at least on grid point).
Now, why is theta=0? Because, when you interpolated the temperature  
field, that you have the model read, you missed at least one point;  
this is a common mistake, that others (including myself) often make,  
see other postings on this email list. To recover, make sure that all  
your "wet" grid points have valid temperatures (getting exactly 0. is  
very unlikely from "real" data), or if you want theta=0 in some  
places, which can actually be necessary in some instances, then you  
have to comment out the stop statement in ini_theta.F (and also in  
ini_salt.F, where the model will stop, after you'll have fixed the  
ini_theta-problem, I predict) and recompile.

I encourage you to have a look at the documentation, email lists,  
*and* the code, as it never hurts to familiarize yourself with the  
tools that you are using.

Martin

On 16 Jun 2008, at 09:01, Maxiaohui wrote:

> hi, Martin
> Thank you for your help.
>  I have run the 1X1 experiment based on the  
> tutorial_global_latlon,but I met an abnormal end: S/R INI_THETA,  
> the STDERR.0000 file shows that
> (PID.TID 0000.0001) *** ERROR *** S/R INI_THETA: theta = 0  
> identically. If this is intentionalyou will need to edit  
> ini_theta.F to avoid this safety check.
> I want to know what's wrong with my experiment,how to solve this  
> problem? Thank you very much!
> Maxiaohui
>
> 雅虎邮箱，您的终生邮箱！
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support