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.

Sign up for GitHub

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

Jump to bottom

Candidate dev master 2017 07 05 #550

Merged
merged 106 commits into from
Jul 21, 2017
Merged

Candidate dev master 2017 07 05 #550

merged 106 commits into from
Jul 21, 2017

Conversation

adcroft
Copy link
Collaborator

@adcroft adcroft commented Jul 7, 2017

This PR will bring dev/master up to date with dev/gfdl (as of 2017 July 5th).

This PR is not to be merged until we have had reviews from NCAR. There are 137 files changed so "reviewing" really means testing (unless @gustavo-marques really want to read through the 106 commits since May 9).

This is the code that was tagged dev/master/2017.07.05 from NOAA-GFDL/MOM6-examples even though we have not merged onto dev/master yet. It is the where dev/gfdl was on July 5th. I should have tagged it dev/gfdl/... but that would have caught others off guard.

This is a test of the new procedures for updating dev/master and master. We need to get @kshedstrom added to the reviewer list so the ESMG fork is part of the process for updating dev/master and master.

Andrew Shao and others added 30 commits May 9, 2017 15:05
Clean up some changes that were made for development purposes
d2af9df
Changes made so that regridding onto pure isopycnal kind of works. Ne…
f4e0063 
…ed to

figure out why the redistribution of residual does not fully conserve.
Add option to skip convective readjustment when reading offline fields
ab0575d 
from z inputs
To aid in the debugging of tracer routines, a new subroutine is created
ff690d1 
in MOM_tracer_registry.F90: MOM_tracer_chkinv. This routine calculates
the global inventories of all tracers in the registry. Some of the
debugging checksums in the offline tracer routines are updated to use this
new function.
Tracer now conserves when remapping input fields onto a different grid.
af73ea8 
This primarily resulted from an error in the indexing when the fields were
being updated. Now to test in OM4_05
Added option to intersperse ALE step between horizontal advection
7ddc722 
iterations.
Fixed conflicts with upstream dev/gfdl
53d03a7
@sdbachman
Added modified Leith viscosity.
48fe2c7 
Added a term proportional to the gradient of divergence to the
Leith viscosity, which turns regular 2D Leith into Modified Leith.
@adcroft
Pipeline: force cleanup before git pull
ebeb5fa 
- When regression tests fail, the working dir is left dirty and
  needs to be reset before pulling the latest regression answers.
@nichannah
Don't use j wind stress in set_viscous_ML u interpolation. Closes #516
f700ca4
@Hallberg-NOAA
(*)Moved set_viscous_BBL out of the dynamics code
0576b2b 
  Moved the calls to set_viscous_BBL out of the dynamics code and up into
step_MOM, eliminating 4-way redundant calls and making the various algorithms
more obviously similar.  The answers are bitwise identical with 3 of the 4
dynamic solvers, but with MOM_dynamics_unsplit the time level of thicknesses
used to calculate the bottom boundary layer properties are replaced with the
initial thicknesses (like the other schemes), rather than the average of the
initial thicknesses and the predicted thicknesses from the first (of perhaps
several) step of the dynamics.  All existing test cases are bitwise identical,
but if SPLIT=FALSE and USE_RK2=FALSE, answers will change.
@Hallberg-NOAA
+Eliminated calc_bbl from the vertvisc_type
e1aed13 
  Eliminated calc_bbl and bbl_calc_time_interval from the vertvisc_type.  Also
renamed pass_kd_kv_turb to pass_kv_turb and added local variables inside of
step_MOM to take the place of the variables that were eliminated, and added a
new clock to step_MOM to record the time taken by calls to set_viscous_BBL. The
unused module use statements for set_viscous_BBL were also eliminated from the
MOM_dynamics modules.  All answers are bitwise identical.
@Hallberg-NOAA
Merge branch 'dev/gfdl' of github.com:NOAA-GFDL/MOM6 into dev/gfdl
144daae
@Hallberg-NOAA
Merge branch '516-transform-fail-uhtot-uses-tauy' of https://github.c…
13e43be 
…om/nicjhan/MOM6 into nicjhan-516-transform-fail-uhtot-uses-tauy
@Hallberg-NOAA
Merge branch 'nicjhan-516-transform-fail-uhtot-uses-tauy' into dev/gfdl
269f630
@Hallberg-NOAA
Merge branch 'dev/gfdl' of github.com:NOAA-GFDL/MOM6 into dev/gfdl
b80fc8b
@adcroft
Merge pull request #519 from Hallberg-NOAA/dev/gfdl
2b8670f 
Moving set_viscous_BBL out of dynamics code

Passes: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/pipelines/1947
@wfcooke
Bugfix to reintroduce group halo pass check removed in 282a160
833dd84 
Runs fail with
FATAL from PE   107: MPP_RESET_GROUP_UPDATE_FIELD_3D_: group%reset_index_s > group%nscalar
etc.

" .and. (.not.CS%thickness_diffuse_first .or. CS%t_dyn_rel_adv == 0)) " was removed in a previous update.
@Hallberg-NOAA
(*)Corrected the barotropic mass source compensation
db3a5b5 
  Corrected a bug when DIABATIC_FIRST=True in how the compensation for the
difference between the dynamic sea surface height used by the barotropic solver
and the height that the layers would have if net mass sources were applied
gradually instead of abruptly at the start of the thermodynamic time step. This
would change answers if both SPLIT and DIABATIC_FIRST are true and the value of
BT_MASS_SOURCE_LIMIT > 0.0, but as there are currently no test cases that match
these conditions, the answers are unchanged in all of the test cases. In cases
where the answers do change, they are unambiguously better based on weaker high
frequency gravity waves arising from coupling shocks.
@kshedstrom
+BT_OBC%H_u fix.
ab9e05e
@wfcooke
Merge pull request #520 from NOAA-GFDL/user/wfcooke/grouphalopass_fix
1e8c2b4 
Bugfix to reintroduce group halo pass check removed in 282a160
@Hallberg-NOAA
Merge branch 'dev/gfdl' of github.com:NOAA-GFDL/MOM6 into dev/gfdl
e8b7f0e
@Hallberg-NOAA
Corrected typos in dOxyGen comments
5daf60f 
  Corrected spelling errors and content errors in several of the comments
describing subroutine arguments.  All answers are bitwise identical.
@Hallberg-NOAA
Replaced group_pass of h with pass_var of h
a8565c7 
  Reverted to the simpler pass_var routines for the single 3-d variable h.  This
avoids the issue that triggered the recent pull request #520, and it should have
no performance impacts.  (In fact, by letting each call have its own halo size,
it might lead to performance improvements in subsequent commits.)  All answers
are bitwise identical.
@CarolineCardinale
added Doxygen comments to MOM_checksum_packages.F90
e22cdeb 
added Doxygen comments to all of the arguments to the subroutines in
MOM_checksum_packages.F90. All the answers are bitwise identical.
@adcroft
Merge pull request #521 from Hallberg-NOAA/dev/gfdl
d20e897 
Change to pass_var for h and correct barotropic mass source

...and it passed (no surprise): https://gitlab.gfdl.noaa.gov/ogrp/MOM6/pipelines/1956
@CarolineCardinale
Merge remote-tracking branch 'refs/remotes/origin/dev/gfdl' into dev/…
76a8484 
…gfdl
@CarolineCardinale
added Doxygen comments to MOM_checksum_packages.F90
73380f5 
added Doxygen comments to all of the subroutines in MOM_checksum_packages.F90.
All the answers are bitwise identical.
@CarolineCardinale
Added Doxygen comments to MOM_dynamics_legacy_split.F90
a06258f 
Added Doxygen comments to all of the subroutines in the
MOM_dynamics_legacy_split.F90. All the answers are bitwise identical.
@CarolineCardinale
Added Doxygen comments to MOM_dynamics_unsplit.F90
e216c50 
Added Doxygen comments to all of the subroutines in the
MOM_dynamics_unsplit.F90. All the answers are bitwise identical.
Hallberg-NOAA and others added 24 commits June 29, 2017 11:18
@Hallberg-NOAA
+Always call initialization routines to get_params
e6b7df2 
  Modified MOM_state_initialization and all of the thickness, temperature, and
velocity initialization routines it calls so that the relevant routines are
always called, regardless of whether this run segment is for a new run or
started from a restart file, so that all valid parameters are read.  This will
enable detection of misspelled or invalid runtime parameters without requiring
that the MOM input files be changed between run segments. Several
MOM_parameter_doc files have been changed by the addition of missing log_version
calls for some initialization routines. All answers are bitwise identical, but
there are new optional arguments to numerous intialization routines.
This commit addresses and should close NOAA-GFDL/MOM6-examples Issue #135.
@Hallberg-NOAA
+Made OS argument to ocean_model_flux_init optional
530fea2 
  Made the ocean_state_type argument to ocean_model_flux_init optional, so that
this routine can be used more versatilely from the coupler.  All answers are
bitwise identical.
@Hallberg-NOAA
Removed trailing white space
5f0fec7 
  Removed trailing white space that had recently been added on two lines.  All
answers are bitwise identical.
@Hallberg-NOAA
Merge pull request #536 from adcroft/fix-gnu-openmp-490
83f606a 
Renames "mod" variable to avoid using imod() fn
@Hallberg-NOAA
Merge branch 'dev/gfdl' of github.com:NOAA-GFDL/MOM6 into dev/gfdl
ebbce13
@adcroft
Merge branch 'dev/gfdl' of https://github.com/Hallberg-NOAA/MOM6 into…
74bcc77 
… Hallberg-NOAA-dev/gfdl
@adcroft
Merge branch 'Hallberg-NOAA-dev/gfdl' into dev/gfdl
249e9f6
@Hallberg-NOAA
Merge branch 'dev/gfdl' of github.com:NOAA-GFDL/MOM6 into dev/gfdl
f6bade7
@Hallberg-NOAA
Renamed variables "mod" to "mdl"
fec16b9 
  The name "mod" had been used for various module name variables throughout the
MOM6 code, but this can cause name-space conflicts with the mod intrinsic
function, so "mod" has been renamed "mdl" throughout the MOM6 code.  All answers
are bitwise identical.
@adcroft
Merge pull request #538 from Hallberg-NOAA/dev/gfdl
a1ac57f 
Renamed variables "mod" to "mdl"

Passed: https://gitlab.gfdl.noaa.gov/ogrp/MOM6/pipelines/2054

(Initial fail seems to have been a stale file issue).
@adcroft
Re-factor of MOM_lateral_mixing_coeffs.F90 (VarMix)
3e50273 
- VarMix was originally for a crude implementation of a Visbeck et al.
  diffusivity formula. It has become a catchall for resolution functions,
  stored slopes and length-scales. It infact has never contained an actual
  diffusivity. The meaning of the parameter USE_VARIABLE_MIXING was original
  "use Visbeck" but has really meant "allocate and calculate everything in
  this module even if not being used". This has now been cleaned up...
- USE_VARIABLE_MIXING still exists but does nothing.
  - It will be made obsolete shortly.
- Visbeck related fields (VarMix@L2v, ...) are allocated only if
  KhTh_Slope_Cff>0 or KhTr_Slope_Cff>0.
- The Eady growth rates are now allocated and calculated if
  either of Visbeck or MEKE schemes are in use.
- The field VarMix%Rd_dx_h is allocated and calculated if either the
  diagnostic is requested or the field needed "Passivity" or the
  resolution functions.
- The wave speed VarMix%cg1 is calculated only if needed but which
  includes resolution fns, Rd_dx, and FGNV stream fn.
  - Closes #522.
- Added logic in tracer_hordiff() to avoid referencing VarMix%L2v or
  VarMix%SN_v when VarMix%KhTh_Slope_Cff==0.
- Added logic in tracer_hordiff() to avoid referencing VarMix%L2v or
  VarMix%SN_v when VarMix%KhTr_Slope_Cff==0.

- This commit affects available_diags and MOM_parameter_doc for some
  experiments but does not changes answers.
@adcroft
Fix VarMix cleanup with openMP
a110df9
@adcroft
Implemented scale-aware option for MLE
f733ff0 
- The third paper, in the series of Fox-Kemper et al., proposed upscaling
  the buoyancy gradients from the frontal scale to the grid-scale. To date
  this has been implemented as a globally constant upscaling indicated by
  the FOX_KEMPER_ML_RESTRAT parameter. This commit implements the upscaling
  as a function of grid-spacing, 1st-baroclinic deformation radius and a
  frontal length scale parameter, MLE_FRONT_LENGTH.
- VarMix is now passed to mixedlayer_restrat since we need VarMix%Rd_dx.
- Updated MLE doxumentation.
@adcroft
Fixed argument name to match declaration gen_trcr_col_physics()
6e4eaa4 
- For some reason we had "MLD" in the argument list but "Hml" in the
  subroutine code and declarations. This was breaking the CM4 compile.
@adcroft
Replaced "sum" with "res" for variable name
3eecc26 
- This fixes another occurrence of a keyword being used as a
  variable name. It is allowed but is rather poor practice and
  makes searching for the use of keywords and constructs somewhat
  harder.
@adcroft
Replaced array-syntax with loops
0f28767 
- The forcing diagnostics were using array syntax than meant calculations
  were being made for points that could be invalid (uninitialized in halos)
@adcroft
Avoid allocating unused arrays in diabatic_driver
04116e8 
- In order to call calculateBuoyancyFlux2d() along with the ePBL
  the KPP fields are all allocated but not all are needed. This
  allocates only those used.
@adcroft
Added "skip_diags" argument to extractFluxes1d()
b56e593 
- The resetting/zeroing and accumulation of diagnostic arrays within
  extractFluxes1d() assumes the routine is called only once. With the
  need for a buoyancy flux in ePBL, it is currently being called twice.
  - This was leading to the "hdfs" diagnostic accumulating too much.
- I have added a flag to indicate to not calculate the diagnostic fields.
  - A better solution will be to only call once but calculate a buoyancy
    flux in extractFluxes1d() but this will require more coding since
    the latter is dependent on the equation of state and needs a different
    interpretation of mass fluxes.
- I checked that this changes "hdfs" but have not checked whether other
  diagnostics were affected.
- Closes NOAA-GFDL/MOM6-examples#128 but I opened #543 to make a note that
  we have more tidying up to do.
@Hallberg-NOAA
Merge pull request #539 from adcroft/cleanup-varmix
7d401bc 
Re-factor of MOM_lateral_mixing_coeffs.F90 (VarMix)
@Hallberg-NOAA
Merge pull request #540 from adcroft/mle-length-scale
13541f4 
Mle length scale
@Hallberg-NOAA
Merge pull request #541 from adcroft/fix-generic-tracer-MLD-arg
e05d8ae 
Fixed argument name to match declaration gen_trcr_col_physics()
@Hallberg-NOAA
Merge pull request #542 from adcroft/replace-keyword-variable-name
e9f146d 
Replace keyword variable name
@Hallberg-NOAA
Merge pull request #544 from adcroft/fix-hdfs-diagnostic
0192c3c 
Fix hdfs diagnostic
@adcroft
Fix openMP directive in MOM_forcing_type.F90
6db42af 
- A new local variable was missing from the openmp directives.
@kshedstrom
Copy link
Collaborator

A checkout of gfdl/candidate-dev-master-2017-07-05
passes all my tests.

@gustavo-marques
Copy link
Collaborator

gustavo-marques commented Jul 21, 2017 via email

@adcroft adcroft merged commit 6db42af into dev/master Jul 21, 2017
@adcroft
Copy link
Collaborator Author

adcroft commented Jul 21, 2017

Thanks all - dev/master on all repos have been rolled forward.

MOM6-examples dev/master now points to the commit that we tagged dev/master/2017.07.05 and which uses MOM6 dev/master implied by that tag. I did this by doing fast-forwards - I had placed a dev/master tag on dev/gfdl branch. From here on, we'll use dev/gfdl/yyyy.mm.dd tags on dev/gfdl.

@adcroft adcroft deleted the candidate-dev-master-2017-07-05 branch July 21, 2017 17:33
@adcroft adcroft mentioned this pull request Aug 17, 2017
Merged
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
9 participants
@adcroft @kshedstrom @gustavo-marques @Hallberg-NOAA @sdbachman @nichannah @wfcooke @CarolineCardinale @awallcraft