Adds icepack BGC #6457

njeffery · 2024-05-31T17:15:13Z

Changes are non-BFB with biogeochemistry and/or sea ice aerosols active.

Tested in ocean-sea ice BGC active simulations.

Needs to be tested and merged with main icepack-integration branch BFB in the physics.

Changed back to original formulation.

-bugfix in carbon conservation - Fluxes scaled by aicen not aicen_init -puts snow aerosols into ice rather than ocean for small snow volumes -Changes vertical z-tracer flags so that aerosols and bgc are turned on separately: 1) config_use_column_biogeochemistry indicates biogeochemical tracers are on 2) config_use_zaerosols indicates dust and black carbon are on BFB in the default configuration. nonBFB with zaerosols or column_biogeochemistry

This column_package subroutine was accidentally added during the merge. BFB

Most of the updates are for use in init_zbgc which now initializes many of the configs, indices and size declarations for bgc Contains updates to column package to make puny and accuracy consistent with carbon conservation check BFB with bgc and aerosols off

Removed zsalinity references BFB

…ated-icepack-bgc This version uses a branch of Icepack, not main (see .gitmodules) Modifications bring changes to enable Icepack bgc into master BFB with physics only.

Added flag to turn off BGC coupling between the ocean and sea ice: config_couple_bgc_fields=false Added black carbon conservation analysis member -track ocean black carbon flux Added carbon conservation check options in interfaces for both icepack and column Bugfixes to column package version: 1. Bugfixes to lateral melt BGC snow flux and ice flux 2. Bugfix to add_new_ice_bgc -replaces adjust_tracer_profile with simpler and more accurate version update_vertical_bio_tracers Define carbon molar mass in mpas constants Commented out problematic omp directives in icepack interface Corrections to some bgc Registry descriptions (still more to do) NOTE: Changed submodule path for Icepack to a branch that works with the BGC interface. nonBFB with BGC

-Cleans up some missing bgc info in Registry (more to do) -Adds some additional info to carbon conservation analysis member -Saves bgc ocean fluxes weighted by ice area (for carbon conservation) -Comments out $omp statements around step therm1 (code was crashing -with seg fault) -Removes and cleans up unnecessary write statements for carbon debugging Tested with E3SM-Project/Icepack.git branch njeffery/merge-fixes-to-bgc commit c090b4b20e4ec0317d2c7d742c5ab291ec7a2240 BFB except correcting a seg fault

Now consistent with E3SM master BFB

Small error introduced in last commit - broke compile BFB

Merge remote-tracking branch 'remotes/origin/njeffery/seaice/updated-icepack-bgc' into origin/master

bfb

-First uses init_parameters to initialize bgc namelist parameters -Second (icepack_init_zbgc_tracer_indices) calls bgc indexing and array definitions. -Third defines parameter arrays for bgc (init_zbgc) bfb (round-off level in bgc)

Adds 2 namelist fields: config_use_atm_dust_file - true, use monthly dust climatology config_use_iron_solubility_file -- true, use dust-iron solubility map from file (in icepack: use_atm_dust_sol = .not.config_use_iron_solubility_file) New fields track atmospheric wet and dry dust separately from the coupler to compute the iron contribution. Fraction of iron in dust is also generalized and read from a file. Dust/iron file is identical to the one used by mpas-o. Corrects some units in Registry. BFB with new namelist parameters = false and for all runs without zaerosols or without iron.

This field oceanDustIronFlux will be used by the ocean for dust-iron consistency across the coupled system Also corrects a bug in the calculation of iron from dust introduced in the last commit Works with E3SM-Project/Icepack.git origin/fixes-to-zaerosols-take3 BFB with zaerosols off

-adds subroutine in the icepack interface, init_zbgc_tracer_indices, to replace icepack_init_zbgc_tracer_indices. -moves icepack_init_bgc_trcr from icepack to init_bgc_tracer_indices in mpas_seaice_icepack Tested in GCASE with ice-ocean bgc and aerosols active works with njeffery/Icepack.git branch: fixes-to-bgc-indices

For BGC

github-actions · 2024-05-31T18:05:41Z

PR Preview Action v1.4.7
🚀 Deployed preview to https://E3SM-Project.github.io/E3SM/pr-preview/pr-6457/
on branch `gh-pages` at 2024-06-22 17:35 UTC

eclare108213

Initial comments: I'd like to run some D tests on the code. Here are some things I noticed, just reading through the code diffs. Comments on mpas-seaice code are direct. For Icepack:

Where is this new submodule hash located (which fork/branch)? Is it different from the submodule listed in E3SM/.gitsubmodules?
Do the changes in icepack_aerosol.F90 change the answers for ‘old’ aerosol runs? e.g. is subroutine update_snow_bgc called when tr_aero=T?
icepack_parameters.F90: Is grid_o_t used in icepack? I had removed it from the parameters list because it wasn't used.
icepack_therm_itd.F90: The first initialization of vicen_init (originally =0, now =vicen(:)) in lateral_melt could be removed completely, since it’s set later. Or initialize both vicen_init and vsnon_init here and remove the later ones.
icedrv_init_column.F90: There’s a new cpp, NEWINITZBGC. Should that be here? I don't think it's set anywhere in the icepack driver.

FYI @apcraig

eclare108213 · 2024-06-01T22:05:35Z