Initial icepack integration #5958

eclare108213 · 2023-09-26T22:00:44Z

Installs Icepack in E3SM.

This PR includes MPAS-seaice changes necessary to run Icepack in equivalent physics configurations as the original column package code. The entire Icepack repository is installed as a submodule linked from the main branch of E3SM’s Icepack fork.

Some minor changes were made to the original column package code for consistency with Icepack during testing. Icepack replaces colpkg as the default column physics package in this PR.

Fully coupled E3SM v3alpha trigrid with Icepack physics is statistically compatible with the 100-year control at the 99% confidence interval.

Icepack is thoroughly and promptly documented in its git repository (the source code is in icepack/doc/).

[NML]
[CC]

…eclare108213/E3SM into eclare108213/seaice/icepack-integration

…ntegration

Add icepack as new submodule Add build of icepack to MPAS-Seaice build system Add extra variables to Registry new to icepack Added mpas_seaice_icepack module as replacement for mpas_seaice_column module. This is initially a straight copy of mpas_seaice_column Replaced all calls to mpas_seaice_column module with calls to mpas_seaice_icepack. Except radiation (as described below) mpas_seaice_icepack still calls the column package Replace column radiation calls from ice_colpkg to icepack_intfc in mpas_seaice_icepack Add initialization of icepack parameters/tracers to mpas_seaice_icepack Add initialization of icepack shortwave data structures to mpas_seaice_icepack Renamed exposed subroutines in mpas_seaice_icepack from _column_ to _icepack_ Replaced few remaining references to ice_colpkg from files other than mpas_seaice_column.F and mpas_seaice_icepack.F with extra subroutines in those files Note: icepack_warnings, icepack_warnings_getall needs to be public Conflicts: .gitmodules components/mpas-seaice/src/icepack components/mpas-seaice/src/shared/mpas_seaice_constants.F

…amelist

… set to be on

…car_ad calls

Bring in changes from AKT icepack-integration branch Merging now in order to cleanly update to the latest E3SM with better testing options.

Update to E3SM master

…ial-integration branch

updating icepack to cice-consortium/E3SM-icepack-initial-integration branch

updating icepack to hash 8fad768 on main

eclare108213 · 2023-10-14T03:07:13Z

This PR is ready for review. Additional information is in https://acme-climate.atlassian.net/wiki/spaces/ICE/pages/3952476924/Overview+of+Icepack+physics+merge+into+MPAS-SI, including links to complete testing results.

proteanplanet

Based on my testing in the E3SM v3alph04-trigrid B-case configuration, where this branch was substituted for master and MPAS-SI was run with config_column_physics_type = 'icepack' with an otherwise identical physical configuration, I approve this PR.

For the case cited above, the climate between the two simulations after 100 years was comparable for the E3SM WCYCL1850 ne30pg2_r05_EC30to60E2r2 configuration. The version with Icepack maintains a stable climate analogous to the control. The sea ice climate between the two simulations is statistically identical at the 99% compatibility interval for key metrics of global sea ice extent, volume, and snow-on-sea ice volume, as indicated in the sea ice lemnisc below.

The above figure summarizes daily sea ice phase average (solid) and interquartile range (gray). Dashed red lines indicate the null hypothesis for days of the year that MPAS-SI with Icepack is different from the standard version of MPAS-SI at the 99% compatibility interval. Traces start on January 1, passing the northward equinox (•), northern solstice (◦), southward equinox (•), and southern solstice (◦) to the end of December (→)

erinethomas

I approve this PR based on my own ongoing testing using B, G and D cases (testing with and without waves). These tests compare icepack to the original column physics and my conclusion is as follows: Icepack performs very similar to the original column package. No results being posted here as @proteanplanet has many nice figures which summarize all of my conclusions from my own testing well.

njeffery

Approved based on thorough review of code changes and the fully coupled simulations performed by @proteanplanet. Climate changes are small. Largest impacts appear to be in the Arctic with similar extents but thicker sea ice in the new codebase.

darincomeau

Approved based on documentation and extensive testing by the integration group over the past several months, including the recent coupled simulations by @proteanplanet .

jonbob · 2023-10-19T20:31:34Z

@eclare108213 - we have a process with some semi-automated scripts to make the E3SM bld files match changes in Registry files, but it looks like this has been done by hand for this PR. I ran the scripts and they found issues in four different files:

#	modified:   components/mpas-seaice/bld/build-namelist
#	modified:   components/mpas-seaice/bld/build-namelist-group-list
#	modified:   components/mpas-seaice/bld/build-namelist-section
#	modified:   components/mpas-seaice/bld/namelist_files/namelist_definition_mpassi.xml

How do you want to handle the differences? I am hesitant to overwrite any of the changes you made, mostly because it looks like there are a lot of changes to the namelist definitions. Those are usually extracted from Registry files, so we may want to change Registry to get some of those descriptions?

proteanplanet · 2023-10-20T04:07:37Z

@eclare108213 @jonbob @wlin7 : A new PR is awaiting review and merge into this branch to fix F-case simulations with Icepack.

Fix icepack prescribed ice mode for F-case simulations

eclare108213 · 2023-10-20T15:50:50Z

How do you want to handle the differences? I am hesitant to overwrite any of the changes you made, mostly because it looks like there are a lot of changes to the namelist definitions. Those are usually extracted from Registry files, so we may want to change Registry to get some of those descriptions?

@jonbob I thought Registry was only used for standalone MPAS runs, the other files for coupled runs, and they weren't expected to be the same. This is extremely confusing and error-prone! If you need them to be the same, then keep the data in the non-Registry files and change Registry. Thanks -

Also, I'm not sure what to do about the conflicting .gitmodules.

jonbob · 2023-10-20T16:06:42Z

@eclare108213 - it is a complicated relationship between the standalone models and what E3SM uses, but Registry is included in both and we have scripts to take what's in Registry and make the E3SM bld files consistent with it. I'll work on making the descriptions in Registry match what you've added to the namelist definitions and we can sync up next week.

I can handle the conflict in .gitmodules when I merge this PR, so don't worry about that

wlin7 · 2023-10-23T16:12:08Z

F2010 simulation using the icepack in prescribed seaice mode shows the impact of switching to icepack is largely negligible.

e3sm_diags for model vs model comparison with v3alpha04 baseline

wlin7

Also verified icepack is used by default.

Move comments to Registry and update bld files to match

jonbob

Approved based on developer testing and long run by the coupled model group

…5958) Initial icepack integration Installs Icepack in E3SM. This PR includes MPAS-seaice changes necessary to run Icepack in equivalent physics configurations as the original column package code. The entire Icepack repository is installed as a submodule linked from the main branch of E3SM’s Icepack fork. Some minor changes were made to the original column package code for consistency with Icepack during testing. Icepack replaces colpkg as the default column physics package in this PR. Fully coupled E3SM v3alpha trigrid with Icepack physics is statistically compatible with the 100-year control at the 99% confidence interval. Icepack is thoroughly and promptly documented in its git repository (the source code is in icepack/doc/). [NML] [CC]

jonbob · 2023-10-30T22:09:40Z

Passes sanity testing during merge with expected results, merged to next

jonbob · 2023-10-31T16:32:25Z

merged to master and expected NML and regular DIFFs will be blessed with PR #5999, which was merged at the same time

eclare108213 and others added 30 commits December 5, 2022 10:31

add icepack submodule

d922306

Merge branch 'eclare108213/seaice/icepack-integration' of github.com:…

a11e238

…eclare108213/E3SM into eclare108213/seaice/icepack-integration

remove extraneous Icepack entry from .gitmodules

98cd0af

compile icepack with cmake

9af2f7c

Merge branch 'E3SM-Project:master' into eclare108213/seaice/icepack-i…

8a922a1

…ntegration

Merge branch 'E3SM-Project:master' into eclare108213/seaice/icepack-i…

d240501

…ntegration

Changes needed for BFB icepack/column pckg shortwave

0df4339

Add namelist option to run with either icepack or column package

afe58d2

Update icepack interface for newer icepack commit

73f41c5

Changes for e3sm build

80bb443

Minor changes for existing E3SM consistency

5b1c863

Removing duplicate use calls

4ddee4a

Adding new config_column_physics_type namelist option to E3SM build n…

2b5a9d7

…amelist

Additional changes for E3SM build namelist

bfab10f

Correcting erroneous 'use' in new namelist option

4fb0fe0

Correcting new namelist option type

d1de208

Adding ifdefs to turn off icepack_step_radiation even when icepack is…

85146bf

… set to be on

Changing 'gregorian_noleap' to 'noleap'

e62effe

Changes to config_shortwave_type for icepack

a6a5afe

Changing tmp config shortwave_type to pointer

068b476

Adding local use_snow logical

c36c445

Adding missing config_use_snicar_ad pool call

2e219f2

Fixing location of use_snow local logical, removing unused config_sni…

807b992

…car_ad calls

Reverting addition of ifdef statements

85e3174

Merge pull request #1 from darincomeau/cherry-pick-from-akt

2ad4b45

Bring in changes from AKT icepack-integration branch Merging now in order to cleanly update to the latest E3SM with better testing options.

update from E3SM master

5dfc119

Merge pull request #2 from eclare108213/update_E3SM_2

33d6cb0

Update to E3SM master

updating icepack to latest version, cice-consortium/E3SM-icepack-init…

9669e94

…ial-integration branch

Merge pull request #3 from eclare108213/update_icepack_1

6bf020e

updating icepack to cice-consortium/E3SM-icepack-initial-integration branch

eclare108213 added 2 commits October 13, 2023 10:27

updating icepack to hash 8fad768 on main

79cc95c

Merge pull request #43 from eclare108213/move_submodule

dc5947c

updating icepack to hash 8fad768 on main

eclare108213 marked this pull request as ready for review October 14, 2023 03:05

proteanplanet approved these changes Oct 14, 2023

View reviewed changes

erinethomas approved these changes Oct 16, 2023

View reviewed changes

njeffery approved these changes Oct 16, 2023

View reviewed changes

proteanplanet added CC PR is climate changing Coupled Model and removed CC PR is climate changing labels Oct 16, 2023

darincomeau approved these changes Oct 17, 2023

View reviewed changes

Fix to prescribed ice mode for F-case simulations

08a8df3

Merge pull request #44 from eclare108213/proteanplanet/seaice/prescribed

b6a0f2c

Fix icepack prescribed ice mode for F-case simulations

wlin7 approved these changes Oct 23, 2023

View reviewed changes

jonbob and others added 3 commits October 25, 2023 16:29

Move some description changes from bld files to Registry

fc4d146

Update bld files to match Registry changes, using automated scripts

e0d08e9

Merge pull request #45 from eclare108213/jonbob/seaice/update-bld-files

52b0551

Move comments to Registry and update bld files to match

jonbob approved these changes Oct 30, 2023

View reviewed changes

jonbob merged commit d18c6e9 into E3SM-Project:master Oct 31, 2023
1 check passed

jonbob mentioned this pull request Oct 31, 2023

Fixes PMSL coupling between ocean and atmosphere #5999

Merged

proteanplanet mentioned this pull request Feb 9, 2024

Add cryo terms to coupler budget E3SM-Ocean-Discussion/E3SM#74

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Initial icepack integration #5958

Initial icepack integration #5958

eclare108213 commented Sep 26, 2023 •

edited by jonbob

Loading

eclare108213 commented Oct 14, 2023

proteanplanet left a comment

erinethomas left a comment

njeffery left a comment

darincomeau left a comment

jonbob commented Oct 19, 2023

proteanplanet commented Oct 20, 2023

eclare108213 commented Oct 20, 2023

jonbob commented Oct 20, 2023

wlin7 commented Oct 23, 2023

wlin7 left a comment

jonbob left a comment

jonbob commented Oct 30, 2023

jonbob commented Oct 31, 2023

Initial icepack integration #5958

Initial icepack integration #5958

Conversation

eclare108213 commented Sep 26, 2023 • edited by jonbob Loading

eclare108213 commented Oct 14, 2023

proteanplanet left a comment

Choose a reason for hiding this comment

erinethomas left a comment

Choose a reason for hiding this comment

njeffery left a comment

Choose a reason for hiding this comment

darincomeau left a comment

Choose a reason for hiding this comment

jonbob commented Oct 19, 2023

proteanplanet commented Oct 20, 2023

eclare108213 commented Oct 20, 2023

jonbob commented Oct 20, 2023

wlin7 commented Oct 23, 2023

wlin7 left a comment

Choose a reason for hiding this comment

jonbob left a comment

Choose a reason for hiding this comment

jonbob commented Oct 30, 2023

jonbob commented Oct 31, 2023

eclare108213 commented Sep 26, 2023 •

edited by jonbob

Loading