[MITgcm-devel] bling

[Jean-Michel Campin]

Hi Matt,

You may want to have the things in 2 places, the "bling" pkg
(that you can put for now in MITgcm/MITgcm_contrib/bling,
 to be moved later in MITgcm/MITgcm/pkg)
and the verification experiment that you could already put 
in MITgcm/MITgcm_contrib/verification_other/
In this case, I would suggest not to pick the same name as
tutorial_global_oce_biogeo, but a different one, for instance: 
"global_oce_biogeo_bling" (not yet a tutorial at this point).

The only difference with what you propose is that, once pkg/bling
is added to the main MITgcm code, we can start to test occasionally
 MITgcm_contrib/verification_other/global_oce_biogeo_bling
(as we do for few additional experiment).
But the drawback is, until that time, things will be at 2 places.

Cheers,
Jean-Michel

On Tue, May 20, 2014 at 11:14:02AM -0700, Matthew Mazloff wrote:
> Hello
> 
> Ariane Verdy, Brendan Carter, and I have implemented the Biogeochemistry with Light, Iron, Nutrients and Gases (BLING) model: https://sites.google.com/site/blingmodel/ into the MITgcm copying as closely as possible the conventions used for dic. Files are:
> login2$ ls MITgcm/pkg/bling
> bling_ad_check_lev1_dir.h  bling_atmos.F                          bling_diagnostics_init.F  bling_main.F                         bling_set_control.F
> bling_ad_check_lev2_dir.h  BLING_ATMOS.h                    bling_fields_load.F          BLING_OPTIONS.h              bling_store_fluxco2.F
> bling_ad_check_lev3_dir.h  bling_calcium_carbonate.F  bling_ini_forcing.F           bling_production.F               bling_surfforcing_init.F
> bling_ad_check_lev4_dir.h  bling_carbon_chem.F            bling_init_fixed.F             bling_read_co2_pickup.F   BLING_VARS.h
> bling_ad_diff.list                      bling_cost.F                             bling_init_varia.F             bling_readparms.F               bling_write_pickup.F
> bling_ad.flow                           BLING_COST.h                       bling_insol.F                     bling_read_pickup.F
> bling_airseaflux.F                   BLING_CTRL.h                       BLING_LOAD.h                bling_remineralization.F
> 
> 
> we would like to check in this package. Is this OK? We need advice on best approach to implement this package.
> 
> How about this idea for a starter:
> 
> MITgcm/MITgcm_contrib/bling/pkg/
> MITgcm/MITgcm_contrib/bling/README
> MITgcm/MITgcm_contrib/bling/verification/tutorial_global_oce_biogeo/code_bling
> MITgcm/MITgcm_contrib/bling/verification/tutorial_global_oce_biogeo/code_ad_bling
> MITgcm/MITgcm_contrib/bling/verification/tutorial_global_oce_biogeo/input
> 
> Or is there a better tutorial with which to start from?
> 
> code_bling would only have additional needed changes to gchem:
> pkg/gchem/GCHEM.h
> pkg/gchem/gchem_forcing_sep.F
> pkg/gchem/gchem_init_vari.F
> pkg/gchem/gchem_readparms.F
> 
> code_ad_bling would have necessary mods to 
> pkg/autodiff/checkpoint_lev?_directives.h  
> and thus includes added to 
> model/src/forward_step.F
> model/src/the_main_loop.F
> model/src/main_do_loop.F
> and  also the necessary mods for optimization:
> pkg/profiles/*
> pkg/cost/*
> pkg/ctrl/*
> 
> we can add folders 
> MITgcm/MITgcm_contrib/bling/verification/tutorial_global_oce_biogeo/build
> MITgcm/MITgcm_contrib/bling/verification/tutorial_global_oce_biogeo/results
> MITgcm/MITgcm_contrib/bling/verification/tutorial_global_oce_biogeo/run
> for consistency…
> 
> Thanks for guidance
> Matt
> 
> 
> 
> 
> 
> 
> 
> 

> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel