[MITgcm-support] Re: EXF package: useExfYearlyFields=.FALSE.

Dimitris Menemenlis menemenlis at jpl.nasa.gov
Sat Jul 18 14:00:11 EDT 2009


Hi Madeline, a few things.

First, I don't think you need to change the forcing in order to
reproduce your problem.  For grid interpolation to work, you
would need to compile with
#define USE_EXF_INTERPOLATION
in EXF_OPTIONS.h

Second, as Constantinos points out, you can run the ocean model
with many different surface forcing configurations, depending on
what you specify in data,exf

Third, I had a look at /nobackup2f/madeline/MITgcm/run4_pal02
If that's where you are trying to run the code it won't work because
there is no ../../cube78_forcing/*
If you are trying to run in that directory you need:
cd /nobackup2f/madeline
ln -sf /nobackup2a/menemenl/forcing/cube78_forcing .
ln -sf /nobackup2a/menemenl/forcing/ncep_rgau .

Can you point me to directory where you tried to run and where
the error message lives.

D.

On Jul 17, 2009, at 9:15 PM, Madeline Miller wrote:

> I am trying to replicate my setup/the error with one of the simpler  
> verification experiments, but that has led to more questions.
>
> 1) Dimitri, what forcing are you using in the lab_sea experiment? If  
> I want to use cube78 forcing data do I need to worry about grid  
> interpolation when defining my lat_inc/lon_inc parameters in data.exf?
>
> 2) The EXF package code that downloads with the verification  
> experiments appears to be different than the current code; for  
> example it requires the use of wind stress data instead of wind  
> velocity data. Is there a reason for this?  If there are differences  
> between the verification EXF code and the current EXF code like  
> this, does testing with the verification experiments still make  
> sense?  Or, do you recommend compiling the current code with the  
> verification experiment parameter/input files?
>
>
> In the meantime I'm attaching my data.exf and data.cal files because  
> maybe my problem is something very simple that you will recognize  
> right away; Dimitri, you can also look in /nobackup2f/madeline/ 
> MITgcm/run4_pal02




More information about the MITgcm-support mailing list