[Dart-dev] CESM modelers only - anyone with an opinion, please weigh in

[Kevin Raeder]

See embedded opinions...

On 10/22/13 12:59 PM, Tim Hoar wrote:
>
> Functionally, I would like to propose that the shell_scripts directories for each model component have (something like) the following:
>
> CESM1_1_1_setup_inital           (pull model states from all over)
> CESM1_1_1_setup_pmo              (advance a single model state)
> CESM1_1_1_setup_hybrid           (stage an ensemble from a reference case and carry on)
> perfect_model_obs.csh
> assimilate.csh
Is CESM1_1_1_config missing from this list?  Or has it been absorbed 
into something else?
It's good to have the model version in the name.  If it's useful, 
unchanged, for future model versions,
we can just have  note in each file saying so.
I like Nancy's proposal for perfect_model{_obs}.csh.
Will st_archive.sh be in here too?
Nitpick; 'initial' instead of 'inital'

>
> CAM may need additional scripts to provide backward compatibility.
> CESM1_1_1_setup_startup     (run in perpetual startup mode ...)
> assimilate.csh              (to manipulate the value of CONTINUE_RUN ...)
>
> Does anyone have an improvement on that? I need to rename the scripts in the CESM directories to be more consistent. This will help the learning curve as well as the documentation.
>
> Does anyone have an opinion about making the names a bit more specific?
>
> perfect_model_obs.csh  --> [cam,clm,pop]_perfect_model_obs.csh
> assimilate.csh         --> [cam,clm,pop]_assimilate.csh
I'm in favor of either this, but I'd listen to arguments in favor of 
keeping the generic names,
but copying them to new, local names when they might be confused with 
sibling scripts for
other components, like when doing multi-component, or coupled assimilation.
Will people be doing assimilations outside of the CESM framework, where 
the generic name
would be slightly more convenient?
>
> Can we call the observations from a perfect model experiment  obs_seq.xxxx.perfect instead of
> obs_seq.out?  Can we reserve the obs_seq.out for 'real' observations?
That would be an improvement.
We could also consider 'obs_seq.real' for real obs files that haven't 
been assimilated yet.
'Obs_seq.final' seems fine for the output from assimilations.
>
> Now is the time to provide your opinions.
>
> Thanks -- Tim
>
> Tim Hoar, Associate Scientist
> National Center for Atmospheric Research
> thoar at ucar.edu
> 303.497.1708
>
> _______________________________________________
> Dart-dev mailing list
> Dart-dev at mailman.ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/dart-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ucar.edu/pipermail/dart-dev/attachments/20131022/23e42df1/attachment.html