Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update marbl #75

Merged
merged 9 commits into from
May 11, 2023
Merged

Update marbl #75

merged 9 commits into from
May 11, 2023

Conversation

mnlevy1981
Copy link
Collaborator

Description of changes:

Update MARBL to use latest version (after marbl-ecosys/MARBL#423 is accepted / merged / tagged)

Testing:

Test case/suite: none yet; will run aux_pop and aux_pop_MARBL and comment on results below
Test status: [bit for bit, roundoff, climate changing]

Fixes [POP2 Github issue #] No tagged issues, though there are some MARBL-related issues that might be easy to fix that could be included

User interface (namelist or namelist defaults) changes? latest+4p2z is a new valid value for OCN_BGC_CONFIG

mnlevy1981 added 2 commits May 2, 2023 10:36
MARBL was updated to be able to pass the 3D total chlorophyll field in addition
to just the surface chlorophyll; support for that feature changed the name of
several underlying datatypes and variables (e.g. replacing "surface flux
outputs" with the more general "outputs for GCM" because total chlorophyll
comes from the interior tendency computation rather than the surface flux
computation).
Kristen has a MARBL PR that provides settings_latest+4p2z.yaml to let users run
the tuned 4p2z model without passing around user_nl_marbl. This commit contains
the changes allow OCN_BGC_CONFIG="latest+4p2z" and to provide an
init_ecosys_init_file that contains all the necessary tracers.

The initial condition file was generated by
1. Autotroph updates:
   -- create coccoC, coccoChl, coccoP, and coccoFe tracers that are 7% of the
      original sp values
   -- renaming spCaCO3 to coccoCaCO3 (and updating the long_name)
   -- reducing spC, spChl, spP, and spFe tracers to 93% of original value
2. Zooplankton updates:
   -- split zooC into microzooC (56% of zooC) and mesozooC (44% of zooC), and
      update long_name accordingly
@mnlevy1981
Copy link
Collaborator Author

I still need to update Externals_POP.cfg once there is a MARBL tag.

Should we run 4p2z by default? I either need to add a 4p2z test to aux_pop_MARBL or make the default 4p2z and add a 3p1z test to aux_pop_MARBL depending on the answer to that question.

mnlevy1981 added 7 commits May 9, 2023 22:11
This has updates to the SPECTRA model, which requires an update to
user_nl_marbl in the spectra test (and also finally updates testlist_pop.xml
for NUOPC)
This has updates to surface_forcing_output (generalizing to "outputs_for_GCM"
to allow MOM6 to receive the full 3D chlorophyll field).
Instead of creating a new IC file, use io_read_fallback for microzooC and
mesozooC, then we can read the 4p1z IC file
These came from a different local branch
@mnlevy1981
Copy link
Collaborator Author

Testing results: I ran aux_pop_MARBL on cheyenne (with both intel and gnu). All tests fail NLCOMP due to additional variables that were added to marbl_in in recent MARBL updates (we are jumping from marbl0.40.3 to marbl0.43.0, so there's a lot to catch up on). Further, all tests are failing BASELINE because those recent MARBL commits also added some new diagnostics for FEISTY. There are two tests that are actually changing answers, and failing the BASELINE due to RMS differences detected by cprnc:

  • ecosys_add_cocco: One of the MARBL changes is to cap picpoc at 1.0, which I believe only affects explicit calcifiers. Since the default 3p1z configuration has an implicit calcifier, the majority of tests do not run through that portion of the code. Coccolithophores are an explicit calcifier, though, so the 4p1z configuration in the add_cocco test does compute / use this picpoc term.
  • ecosys_spectra_pfts: One of the MARBL changes was an update to how zoo_loss is routed; it was brought in as a recent update to SPECTRA. Not only does that specific code modification change answers, but I also provided a more recent SPECTRA-fied version of user_nl_marbl for this test to keep the test in line with how the code has most recently been run.

@mnlevy1981 mnlevy1981 merged commit 3d150f4 into ESCOMP:master May 11, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant