[MITgcm-support] odd diagnostics package output

[Jean-Michel Campin]

Hi Daniel,

Sorry, I did not realized that the shift was just 1 iteration (600.s).

And yes, there are some issues about 1 iteration shift (depending
on which variable it is, dumpFreq output & diagnostics snap-shot will
not necessary match).

Here is the choice (from diagnostics_write.F):
          IF ( freqSec.LT.0. ) THEN
C--     write snap-shot with suffix = myIter to be consistent with
C       time-average diagnostics (e.g., freq=-1 & freq=1):
c           wrIter = myIter
c           wrTime = myTime
C--     write snap-shot with suffix = myIter-1 to be consistent with
C       state-variable time-step:
            wrIter = myItM1
            wrTime = myTime - deltaTclock
          ELSE
            wrIter = myIter
            wrTime = myTime
          ENDIF

But if you look at the iteration number (in the surface.*nc file), it
should match with the state.*.nc file, at least for all state-variable
like ETAN (and with MDSIO, I don't get the time in the .meta file).
And for the time (T) in the netcdf file, I don't know why Ed did not
choose to use "wrTime" intead of "myTime" (like we use wrIter
instead of myIter).

Jean-Michel

On Sat, Sep 03, 2005 at 03:12:28PM -0400, Daniel Enderton wrote:
> Thanks for the response JM,
> 
> Despite setting
> 
> frequency(1) = -600.,
> timephase(1) = 0.,
> 
> I get the same behavior as originally described.  I looked at the 
> posts on the MIT-devel list mails, but this did not to help to 
> clarify the issue.
> 
> Daniel
>