Update to treatment of vertical passive tracers and a new pseudo-salt tracer by ashao · Pull Request #353 · mom-ocean/MOM6

ashao · 2016-09-13T20:25:39Z

The changes made here ensure that passive tracers are treated in the same way as the active tracers.
Overview of changes

tracer_vertdiff was moved from MOM_tracer_registry.F90 to MOM_tracer_diabatic.F90
The new subroutine applyTracerBoundaryFluxesInOut is an equivalent routine to applyBoundaryFluxesInOut. This is called during tracer_vertdiff if the appropriate options are passed into that routine.
This new subroutine also now allows passive tracers to be conserved when ALE is used
A new passive tracer, dubbed 'pseudo_salt' has been added that uses the same initial conditions and boundary fluxes as the thermodynamic salt tracer. The diagnostic 'pseudo_salt_diff' should be zero everywhere if active and passive tracers are treated the same.
A new run time flag has been added called 'TRACER_TRIDIAG' which uses tracer_vertdiff instead of tridiagTS to do vertical diffusion of T and S.
The fidelity of these changes was confirmed in a version of the Baltic test case configured for use in ALE z
The changes in this pull request should NOT change answers except for passive tracers

- Added ocean_SIS/OM4_025 (new basis for CM4 ocean component) - *Major fix for "interp from Z" - Added z-diagnostics to OM4/MOM6z

…ve tracers because of a layer thickness update.w Details: In the subroutine applyBoundaryFluxesInOut, the 'h' field is updated based on freshwater fluxes. T and S are treated in a way that is consistent with these mass fluxes. However for passive tracers, the change in layer thickness (particularly at the top) does not result in a change of tracer concentration. Thus, the 'hold' field that is passed into tracer_column_fns and eventually tracer_vertdiff is either thicker or thinner, but the tracer concentration in that cell has not changed resulting in a net change in tracer inventory. This fix modifies the thickness of the top layer by subtracting the freshwater flux (set in applyBoundaryFluxesInOut as ea(i,j,1)) before passing 'hold' into the tracer_column_fns. This change should only affect passive tracers and not change budgets of T, S, or mass.

…cknesses are guarded against

…Need to think about how to implement it most efficiently

… not getting stored correctly

… a passive tracer is next

…ngificant figures are identical except for the last 2 or 3, but it is different in the exponent by a factor of 1000. The factor of 1000 arises because write_energy divides the total salt stock by mass. However, differences do exist between pseudo_salt and salt on a grid cell. Unsure if this is a real difference or not.

…tween tridiagTS and tracer_vertdiff

…because of freshwater flux during ice formation/melting

…er_fix

…ed by comparing the passive pseudo-salt tracer to the active salt tracer used for thermodynamic calculations. A small difference in the placement of parentheses in tracer_vertdiff compared to tridiagTS lead to answers changing when T/S were diffused vertically using the passive tracer routines. In a Baltic ALE-z test case, the pseudo salt tracer tracks the active salt tracer perfectly.

ashao · 2016-09-13T20:36:23Z

Also note that this does NOT fix the passive tracer concentration for tracer which use MOM_generic_tracer. generic_tracer_vertdiff_G must be updated similarly to tracer_vertdiff in MOM_tracer_diabatic.F90.

…racer package

…Out is called explicitly in tracer column_physics routines

…irm that the COBALT tracers are conserved.

…the generic tracer package using CFCs and COBALT.

ashao · 2016-09-21T19:03:44Z

To validate this fix for the generic tracer framework, I used a OM4 0.5 degree configuration based on John Krasting's CFC/SF6 model run. I integrated it for a day, then used it as a restart for a short 5-day run where surface fluxes of CFCs were turned off. Total CFC-11 and CFC-12 stock were conserved to the last 4 decimal places which is the same level of accuracy as passive tracers using the MOM6 tracer framework.

jkrasting · 2016-09-22T15:45:11Z

@ashao -

@adcroft, @nikizadehgfdl, and I were looking at this update. Are any changes needed to ocean_shared for this to work?

ashao · 2016-09-22T15:51:08Z

@nikizadehgfdl , @adcroft , @jkrasting ,
No there should be no changes necessary to any of the ocean_shared routines.

Originally, I thought there might need to be, but I was able to organize it so that the fix is only implemented on the MOM6 side via MOM_generic_tracer.F90.

Andrew

jkrasting · 2016-09-22T15:52:12Z

Excellent. Thanks @ashao

Dr. John Krasting, Physical Scientist
NOAA/Geophysical Fluid Dynamics Laboratory
Biogeochemistry, Ecosystems, and Climate Group
201 Forrestal Road, Princeton, NJ 08540
P. (609) 452-5359 | F. (609) 987-5063

On Thu, Sep 22, 2016 at 11:51 AM, Andrew Shao notifications@github.com
wrote:

@nikizadehgfdl https://github.com/nikizadehgfdl , @adcroft
https://github.com/adcroft , @jkrasting https://github.com/jkrasting
,
No there should be no changes necessary to any of the ocean_shared
routines.

Originally, I thought there might need to be, but I was able to organize
it so that the fix is only implemented on the MOM6 side via
MOM_generic_tracer.F90.

Andrew

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
NOAA-GFDL#353 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AGSgc5khXuJlwUX5I-weonDSAmujIICPks5qsqPugaJpZM4J8Gs6
.

jkrasting · 2016-09-22T17:18:17Z

The fix from @ashao undoes the edits required for the generic_abiotic routines from commit 3b8ecaa. Thus, today's tag dev/master/2016.09.22 no longer works with the latest ocean_shared code. I will put back in the necessary changes, but once it is done, can we retest and update the 2016.09.22 tag?

adcroft · 2016-09-22T17:32:15Z

  use MOM_hor_index, only : hor_index_type
  use MOM_io, only : file_exists, read_data, slasher, vardesc, var_desc
  use MOM_restart, only : register_restart_field, query_initialized, MOM_restart_CS
-  use MOM_spatial_means, only : global_area_mean


@jkrasting is right, this commit did undo the changes from https://github.com/NOAA-GFDL/MOM6/pull/351/files . I'm guessing this happened because he was merging in changes from a file that is being copied in via the XML and isn't even on GitHub.

adcroft · 2016-09-22T18:45:04Z

This is fixed now. Thanks @ashao and @jkrasting .

* add geolat_u, geolat_v, geolon_u, and geolon_v coordinates to the ocean_geometry file.

jeffflick and others added 20 commits September 30, 2013 12:21

Initial commit

a111632

Merge tag 'dev/master/2014.06.23' into testing

456c21b

- Added ocean_SIS/OM4_025 (new basis for CM4 ocean component) - *Major fix for "interp from Z" - Added z-diagnostics to OM4/MOM6z

Merge branch 'user/z1l/bugfix' into testing

bf92d6f

Merge tag 'dev/master/2014.07.01' into testing

f69a5c6

Merge "testing" into master for tikal_201406 release

b4f2d78

Merge for Ulm release (21.0.0)

a8436ee

Changed so that ea(i,j,k=1) is set by dThickness so that negative thi…

5e8de22

…cknesses are guarded against

Continued work to fix conservation with passive tracer conservation. …

fa8e954

…Need to think about how to implement it most efficiently

Start testing conservation with new applyTracerBoundaryFluxesInOut

dfd22a2

Tracers still not conserving. Suspect that for some reason, h_orig is…

9e5dc3d

… not getting stored correctly

Passive tracers are now conserved. A test case treating salinity like…

145060a

… a passive tracer is next

Testing pseudo-salt tracer

5271591

Testing pseudo-salt tracer

17235bc

Differences between salt and pseudo-salt may be due to differences be…

36acb8a

…tween tridiagTS and tracer_vertdiff

Pseudo salt and salt have very small differences. Suspect difference …

982c373

…because of freshwater flux during ice formation/melting

Resolve conflicts with NOAA-GFDL dev/master

b9a0416

Merge branch 'master' of github.com:NOAA-GFDL/MOM6 into vertical_trac…

cd3a96d

…er_fix

Andrew Shao and others added 5 commits September 15, 2016 10:15

Move some code around to make it easier to incorporate into generic t…

6381e6b

…racer package

Moved some lines around to make sure that applyTracerBoundaryFluxesIn…

f38bd7b

…Out is called explicitly in tracer column_physics routines

Applied the fix for the generic tracer package, needs testing to conf…

e0ceede

…irm that the COBALT tracers are conserved.

Make sure everything is synced

8105945

Validated that applyTracerBoundaryFluxesInOut works as expected with …

9cf6b64

…the generic tracer package using CFCs and COBALT.

adcroft merged commit 9cf6b64 into mom-ocean:dev/master Sep 22, 2016

adcroft reviewed Sep 22, 2016

View reviewed changes

jkrasting mentioned this pull request Sep 22, 2016

Added back in abiotic mods #360

Merged

ashao deleted the vertical_tracer_fix branch February 15, 2017 22:13

iangrooms pushed a commit to iangrooms/MOM6 that referenced this pull request May 6, 2025

add u and v coordinates to the ocean_geometry file (mom-ocean#353)

3b78f09

* add geolat_u, geolat_v, geolon_u, and geolon_v coordinates to the ocean_geometry file.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update to treatment of vertical passive tracers and a new pseudo-salt tracer#353

Update to treatment of vertical passive tracers and a new pseudo-salt tracer#353
adcroft merged 25 commits into
mom-ocean:dev/masterfrom
ashao:vertical_tracer_fix

ashao commented Sep 13, 2016 •

edited

Loading

Uh oh!

ashao commented Sep 13, 2016

Uh oh!

ashao commented Sep 21, 2016 •

edited

Loading

Uh oh!

jkrasting commented Sep 22, 2016

Uh oh!

ashao commented Sep 22, 2016

Uh oh!

jkrasting commented Sep 22, 2016

Uh oh!

jkrasting commented Sep 22, 2016

Uh oh!

adcroft Sep 22, 2016

Uh oh!

adcroft commented Sep 22, 2016

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Conversation

ashao commented Sep 13, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ashao commented Sep 13, 2016

Uh oh!

ashao commented Sep 21, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jkrasting commented Sep 22, 2016

Uh oh!

ashao commented Sep 22, 2016

Uh oh!

jkrasting commented Sep 22, 2016

Uh oh!

jkrasting commented Sep 22, 2016

Uh oh!

adcroft Sep 22, 2016

Choose a reason for hiding this comment

Uh oh!

adcroft commented Sep 22, 2016

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

ashao commented Sep 13, 2016 •

edited

Loading

ashao commented Sep 21, 2016 •

edited

Loading