Update to gfdl 20200108 dev/master commit #14
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
This patch addresses the inconsistency of signed zero in the minimum and maximum values used in checksum report. The behavior of the Fortran intrinsic min() and our MPI library's implementation of MPI_Reduce with MPI_MIN can give different results for different values of signed zero, e.g. min(0,-0) vs min(-0,0). Additionally, the MPI_Reduce result appears to not consistenty follow these rules in more complex MPI calculations. Due to these issues, we add the result to positive zero to ensure that any negative zero results are reported as positive.
Added a new optional scale argument to calculate_density, calculate_spec_vel calculate_density_derivs, calculate_density_second_derivs, and calculate_specific_vol_derivs, to rescale the densities or related variables. All answers are bitwise identical, but there are new optional arguments to public interfaces.
Rescale bulkmixedlayer densities and their derivatives via the calls to calculate_density and calculate_density_derivs. All answers are bitwise identical.
Rescaled density units in MOM_entrain_diffusive for dimensional consistency testing. All answers are bitwise identical.
Changed the units of GV%Rlay from [kg m-3] to [R] for dimensional consistency testing. This required the addition of unit_scale_type arguments to several interfaces. All answers are bitwise identical, but new arguments have been added to several public interfaces.
Moved rescaling of Rlay to [R] into the various set_coord routines. This required the addition of unit_scale_type arguments to two interfaces. All answers are bitwise identical, but new arguments have been added to two public interfaces.
Changed the units of GV%Rho0 from [kg m-3] to [R] for dimensional consistency testing. This required the addition of unit_scale_type arguments to several interfaces. All answers are bitwise identical, but new arguments have been added to several public interfaces and the units of an element in a public type have changed.
Rescaled density units in MOM_regularize_layers for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_set_viscosity for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_set_diffusivity for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_kappa_shear for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_internal_tide_input for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in diagnoseMLDbyDensityDifference in MOM_diabatic_aux for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_tidal_mixing for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_geothermal for dimensional consistency testing. This required adding a unit_scale_type argument to geothermal_init. All answers are bitwise identical, but a public interface has a new argument.
Corrected dimensions in comments. All answers are bitwise identical.
Partially rescaled the units of itide%TKE_itidal_input for dimensional consistency testing. All answers are bitwise identical, but the units of an element of a transparent public type have changed.
Rescaled density units in MOM_energetic_PBL for dimensional consistency testing. All answers are bitwise identical.
Rescaled the density units of the cTKE or TKE_forced variables passed to energetic_PBL and applyBoundaryFluxesInOut for dimensional consistency testing. All answers are bitwise identical, but the units of an argument to two public interfaces have changed.
Rescaled the specific volume (density) units of the dSV_dT and dSV_dS variables passed to energetic_PBL, applyBoundaryFluxesInOut, and absorbRemainingSW for dimensional consistency testing. Also rescaled the dimensions of TKE returned from absorbRemainingSW. All answers are bitwise identical, but the units of a arguments to 3 public interfaces have changed.
Used GV%H_to_RZ to simplify rescalings in applyBoundaryFluxesInOut and absorbRemainingSW. All answer are bitwise identical.
Rescaled the units of the mixed layer densities passed to apply_sponge and set_up_sponge_ML_density to [R] for dimensional consistency testing. This required adding a unit_scale_type argument to RGC_initalize_sponges. All answers are bitwise identical, but the units of two arguments to public interfaces have changed.
Rescaled units of dRhodT in applyBoundaryFluxesInOut for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_thickness_diffuse for dimensional consistency testing. All answers are bitwise identical.
Rescaled units of diagnostic FrictWork variables in MOM_hor_visc.F90 for dimensional consistency testing. All answers are bitwise identical.
Rescaled density units in MOM_mixed_layer_restrat for dimensional consistency testing. All answers are bitwise identical.
Rescaled the units of variables set in or passed to set_int_tide_input for dimensional consistency testing. All answers are bitwise identical, but the units of several arguments to public interfaces have been changed.
Added an optional tmp_scale argument to global_i_mean and global_j_mean to specify an internal rescaling of variables being averaged before the reproducing sum. All answers are bitwise identical, but there are new optional arguments to two public interfaces.
Use tmp_scale when taking the i-mean interface heights for i-mean sponges, to give a greatly expanded range of dimensional consistency testing. All answers are bitwise identical.
MOM6: +(*)Dimensional consistency completion
Refactored how time-averaging of fluxes in forcing types that span multiple timesteps and flux diagnostics are handled, and rescaled the units of fluxes%dt_buoy_accum from [s] to [T]. This involved changing the arguments to fluxes_accumulate, forcing_accumulate, mech_forcing_diags and forcing_diagnostics, but because of the differing types of the arguments, an incompatible mix of code will not compile. Also changed the units of dt as passed to accumulate_net_input, and made a minor change to extractFluxes1d to avoid the possibilty of a division by zero. All answers are bitwise identical, but there are public interface changes, including changes that impact the mct and nuopc driver codes.
Added the new runtime parameters KAPPA_SHEAR_ITER_BUG and KD_TRUNC_KAPPA_SHEAR to permit correction of a dimensionally inconsistent expression in the Newton's method solver code of kappa_shear, and to allow the value of shear mixing that is neglected compared with the background mixing to be set at run-time instead of being hard-coded. By default, all answers are bitwise identical, but there are two new runtime parameters and the MOM_parameter_doc files change.
Added the new runtime parameter VERT_FRICTION_2018_ANSWERS that avoids the use of the hard-coded maximum viscous mixing length per timestep in the vertical viscosity code, and added h_neglect in the denominators of several terms in the viscosity code. All answers in the MOM6-examples test cases are bitwise identical, but the answers will change if ANGSTROM is set to 0, and there is a new entry in the MOM_parameter_doc files.
…AA/MOM6 into Hallberg-NOAA-revise_vert_friction
…OAA/MOM6 into Hallberg-NOAA-simplify_forcing_time
about mixed complex and real variables
Corrected arguments to iceberg_forces and iceberg_fluxes in the mct version of update_ocean_model. This should correct the recently introduced problems with compiling MOM6 with the mct coupler.
|
Jiande, think it would be good to try to start documenting at even a basic level the amount of change we see when we update to a new MOM6 from GFDL. Even a figure showing the difference in SST after two days would allow us to get an idea of how much the model changes. Could you add a couple of figures showing this? |
|
Good suggestion, will do
On Fri, Jan 17, 2020 at 2:31 PM Denise Worthen ***@***.***> wrote:
Jiande,
think it would be good to try to start documenting at even a basic level
the amount of change we see when we update to a new MOM6 from GFDL. Even a
figure showing the difference in SST after two days would allow us to get
an idea of how much the model changes. Could you add a couple of figures
showing this?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#14?email_source=notifications&email_token=ADMGAS2V53ZQQDHGDYBSRQDQ6IBQXA5CNFSM4KIKVZHKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJIXLUI#issuecomment-575763921>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ADMGAS73GKS3DD63TW3DOSDQ6IBQXANCNFSM4KIKVZHA>
.
--
Jiande Wang
IMSG at NOAA/NWS/NCEP/EMC
cubic 2088
jiande.wang@noaa.gov
phone:3016833725
|
|
I studied the run log of compset runs and compared with current master runs, the failure in "report.txt"(MOM6_RESTART/MOM.res_1.nc: bit-for-bit MISMATCH) is due to the fact that there is one more variable being added to MOM6 restart file (MOM.res_1.nc, new variable name: kg_m3_to_R) in this commit of MOM6. On the other hand, MOM6 output remain unchanged. |
DeniseWorthen
approved these changes
Jan 19, 2020
There was a problem hiding this comment.
It's great to be able to document that this changes baseline only because of added fields in MOM6 restarts.
I'm comfortable merging to develop and wrapping the MOM6 update into the other baseline changing updates we have upcoming for s2s (fv3 bug fix, remove merge in nems).
|
this resolve issue #13 |
Closed
jiandewang
pushed a commit
to jiandewang/MOM6
that referenced
this pull request
Jun 17, 2021
Merge in latest dev/gfdl updates
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.
this resolves issue #13 , to update NOAA-EMC/MOM6 dev/emc to GFDL's newly committed dev/master (20200108, hash # 9f31539). This commit will change baseline results.