Updates from mom6 driver #419
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Warn the GCM if the sum of the ice category fractions is not close to 1 (threshold is hard-coded to be within 1e-6 of 1)
Adding check_forcing() dropped throughput of a 1-mo C test case from 9 SYPD to 2.3 SYPD... so I think we want the default to be not to call this function.
Renamed a few data types: * marbl_surface_flux_output_type -> marbl_output_for_GCM_type * marbl_single_sfo_type -> marbl_single_output_type Also replaced "totalChl" with "total_surfChl" in the indices / constructor because we will use total_Chl for the 3D field. marbl_single_output_type now contains both forcing_field_0d and forcing_field_1d, because total_Chl will need to allocate memory with num_levels. I removed num_elements from the marbl_output_for_GCM_type because it was never used. There was lots of renaming variables to replace sfo or surface_flux_output with something more generic.
I think this is everything we need to pass 3D Chl back to the GCM
When copying from this%outputs_for_GCM to new_output before deallocating this%outputs_for_GCM / reassigning the pointer, I had forgotten to copy field_1d. This revealed itself as a problem in CI, because we check for unused variables and I had declared dim2_loc in anticipation of re-allocating field_1d
call_compute_subroutines test adds three more fields to history file: output_for_GCM_flux_co2 output_for_GCM_total_surfChl output_for_GCM_total_Chl This commit will fail CI because the baseline file does not contain these fields; next commit will contain updated baseline.
Baseline file now includes the three new fields added to the output
MOM6 needs to know the initial total_Chl field, so waiting until we call interior_tendency_compute() is not sufficient. So rather than getting totalChl in an analagous manner to how POP gets surface Chl, there is a new function on the interface [get_output_for_GCM()] that takes the tracer array and an integer index indicating what field to return as an intent(in). Note that this new function replaces negative tracer values with 0, but does not enforce consistency among the autotroph tracers (so Chl may be positive while C, P, or Fe are 0). Because of this, I expect this commit to fail CI but I'll update the baseline in the next commit and everything should pass again.
Also removed an unused variable declaration and added a fill value to the output_for_GCM_total_Chl field in the netCDF file since it is undefined below the ocean floor.
Rather than passing in an array of tracer values, users need to make sure this%tracers has the latest values (we were hoping that passing in an array of tracers would avoid a copy in the MOM6 driver, but the way tracers are stored in a derived type meant we were copying it anyway)
Right now we are only using get_output_for_GCM() to return 1D fields, but we may want to introduce an interface to allow it to return scalars as well
First version set num_levels = -1 for 2D fields, but 0 makes more sense
Removed the following: 1. use statement for a type that is not used in the module 2. return statement inside a function that doesn't need to be there 3. comment refering to a block of code that was removed earlier in this development Also cleaned up one last comment that refered to surface flux output instead of output for the GCM.
|
@klindsay28 and I reviewed this PR on March 8th; 0c46cd5, 1828803, and 75852f1 address all the small changes. The only outstanding item from the code review is that |
|
Per further discussion with @klindsay28, we're going to merge this as is and address the problems in |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The big feature is generalizing the
surface_flux_output_typeso MARBL can provide the GCM with fields frominterior_tendency_compute()as well assurface_flux_compute(); this new functionality is used to allow the option of returning the 3D chlorophyll field (POP only needed the surface chlorophyll, but MOM needs the 3D field).I also added an option to check consistency in the PAR forcing field, but it is turned off by default because there is a huge performance hit.