Initial addition for energy fixer for the MPAS dycore #23

MiCurry · 2021-06-08T16:07:41Z

I'm just creating this PR to easily see @PeterHjortLauritzen energy fixer changes in dp_coupling.F90. We can delete it if need be.

…or MPAS for adiabatic setup (so it follows the approach for other dycores)

Merge Lauritzen's mpas_fixes_phl branch with Truesdale's maps_fixes branch

…or MPAS for adiabatic setup (so it follows the approach for other dycores)

…en/CAM into mpas_fixes_phl

…90, regression tested for QPC6

…inate

…re side

…mpas_fixes_phl

…umbers added to similar error messages to aid debug, ChangeLog updates, remove unused variable and code in geopotential.F90, Make MPAS external non-optional again

cam6_3_020: MPAS bug fixes, performance enhancements, allow initial files writes, refactor/cleanup PR mods: . Update CIME external to include NUOPC compliance for MPAS . Update MPAS external: NUOPC compliance, timer output, new sponge layer diffusion . Remove MPAS Streams generation (unused) . dyn_in and dyn_out refactored to point to the same structure. . Fixes to allow NAG compilation . Fix to allow running on 1 pe. . Refactor for weak scaling. . bug fixed in MPAS sub stepping. . fixed a small discrepancy in the diagnosed pressure of one of the simple model cases. . Update default MPASA grids, initial files, and boundary data for newly supported resolutions . New defaults for MPAS dycore timestep. Changed the 3rd-order scaler transport upwind coefficient from 600s to 900s. . Add functionality to output initial condition files. . Add new sponge layer diffusion options for model stability 1) Rayleigh 2) diffusion similar to SE Fixes ESCOMP#385

MiCurry · 2021-06-08T16:34:45Z

src/dynamics/mpas/dp_coupling.F90

+   compute_energy_diags=&
+        (hist_fld_active('SE_dDP').or.hist_fld_active('SE_dDM').or.hist_fld_active('SE_dPD').or.&
+         hist_fld_active('KE_dDP').or.hist_fld_active('KE_dDM').or.hist_fld_active('KE_dPD').or.&
+         hist_fld_active('WV_dDP').or.hist_fld_active('WV_dDM').or.hist_fld_active('WV_dPD'))


I see that this is setting a flag compute_energy_diags to turn on the energy fixer down below, but I'm unfamiliar with the fields listed, SE_dDp, KE_dDP etc. Can you explain, and perhaps in a comment above this statement, on why we turn on the energy fixer with these variables?

If anything, a comment would help other non-scientists understand when we need to turn on the energy fixer.

They are explained in dyn_comp.F90. Should I repeat description here or we could set the logical in dyn_comp.F90?

Also: these are energy diagnostics and not the energy fixer. The energy fixer is in physics.

Also: these are energy diagnostics and not the energy fixer. The energy fixer is in physics.

Ah yes, thanks for the clarifications!

They are explained in dyn_comp.F90. Should I repeat description here or we could set the logical in dyn_comp.F90?

Hmm yes, I see that now. I think it might make most sense to set the logical in dyn_comp.F90. However I'm not sure if dyn_init in dyn_comp.F90 is the best to define these diagnostics fields. @jtruesdal do you know if these diagnostics should be defined in dyn_init? It looks like there's some fields that are defined in stepon.F90, but I'm not familiar with the purpose of that module.

Most of the history fields specific to a parameterization are defined in the initialization routine. It makes sense to me that would be dyn_init for the dycores. We should probably move the addfld calls taking place in stepon_init to dyn_init.

@jtruesdal moving the addfld calls into dyn_init sounds good to me, if you think it is good. I'm not familar with the stepon module. Do you know what its purpose is?

@PeterHjortLauritzen after some thought, I think setting compute_energy_diags in dyn_init makes the most sense. As, I'm assuming, that compute_energy_diags wont change throughout a run. i.e. the energy fixer will either be on or off. Is that assumption correct?

Can you also remind me in what cases the energy fixer will be on/off? Or will it always be on?

Will move to dyn_init

This is a logical for diagnostics and it will not impact energy fixer in physics.

MiCurry · 2021-06-08T16:47:50Z

src/dynamics/mpas/dp_coupling.F90

+  real(r8), dimension(qsize,nVertLevels, nCells), intent(in) :: q       ! tracer array
+  real(r8), dimension(nVertLevels, nCells),       intent(in) :: ux      ! A-grid zonal velocity component
+  real(r8), dimension(nVertLevels, nCells),       intent(in) :: uy      ! A-grid meridional velocity component
+  character*(*),                                  intent(in) :: outfld_name_suffix ! suffix for "outfld" names


Other character input arguments in the MPAS dycore are defined as character(len=*), intent(in) ::. Can we update this outfld_name_suffix to do the same for consistency:

character(len=*), intent(in) :: outfld_name_suffix ! suffix for "outfld" names

character(len=*) is the common syntax in CESM/CAM as well.

I'll update on next push.

MiCurry · 2021-06-08T17:00:31Z

src/dynamics/mpas/dp_coupling.F90

+  name_out3 = 'WV_'   //trim(outfld_name_suffix)
+  name_out4 = 'WL_'   //trim(outfld_name_suffix)
+  name_out5 = 'WI_'   //trim(outfld_name_suffix)
+  name_out6 = 'TT_'   //trim(outfld_name_suffix)   


I suggest renaming the name_outx to be more specific. For instance, name_out_se, name_out_ke, etc. if that is appropriate.

This will help the writes on: 872, 873, 874, 888, 902 be clearer to readers and will help ensure that the correct variable is going out to the correct field.

I agree and will rename.

Sounds good. Since the same code exists in physics we should change there too:

subroutine calc_te_and_aam_budgets in src/physics/cam/check_energy.F90

MiCurry · 2021-06-08T17:13:55Z

src/dynamics/mpas/dp_coupling.F90

+  name_out6 = 'TT_'   //trim(outfld_name_suffix)   
+
+  if ( hist_fld_active(name_out1).or.hist_fld_active(name_out2).or.hist_fld_active(name_out3).or.&
+       hist_fld_active(name_out4).or.hist_fld_active(name_out5).or.hist_fld_active(name_out6)) then


Is this if statement necessary? Is compute_energy_diags preforming the functionality in each of the calls to tot_enegry?

The if-statements are in there so that we do not compute the vertical integrals if we are not going to output them. The way the subroutine is used in dp_coupling makes the logicals not necessary (as you are pointing out).

There is a similar subroutine in physics (calc_te_and_aam_budgets) and the calls to that subroutine do not have logicals around them in which case we would be computing several vertical integrals even if the energy diagnostics are not active. That is where I got the code from.

I am fine with either approach

Unless the internal if statement provides some other functionality, I don't think it necessary. I think using compute_energy_diags on the calling code is more explicit to whether the integrals are being computed or not. Rather than having a call to tot_energy which potentially does nothing. That would make the calling code clearer as to when they are being computed.

src/dynamics/mpas/dp_coupling.F90

MiCurry · 2021-06-08T17:44:00Z

src/dynamics/mpas/dp_coupling.F90

   ! CAM physics output redistributed to blocks.
   real(r8), allocatable :: t_tend(:,:)
-   real(r8), allocatable :: qv_tend(:,:)
+   real(r8), allocatable :: q_tend(:,:,:)


Now that this variable contains all scalars, and not just qv, would scalars_tend be a better name? MPAS names its scalars array as such.

The variable only contains thermodynamically active species and not all advected species (in a standard F2000 AMIP setup the thermodynamic active tracers are water vapor and condensates; for WACCM-x it would include components of dry air). I am fine with renaming if you wish.

I think tracer would be more standard. That name is used throughout CAM including the other dycores (and I see it used a lot in MPAS that way as well). It used to be called the q array in CAM but it is being standardized as tracer. If that's OK with everyone I will rename to tracer_tend

tracer_tend sounds good to me

It is only the thermodynamically active species but that would make the name really long: thermodynamically_active_tracer_tend

I am fine with what you choose.

MiCurry · 2021-06-10T16:07:10Z

@PeterHjortLauritzen did you want me to review other files you changed in dynamics/mpas?

src/dynamics/mpas/dyn_grid.F90

… in MPAS

…ixer does not have liq and ice species. Fix for SPCAM

…into cam_mpas_dev_pel

Fix TSKIN field to correct value

PeterHjortLauritzen added 30 commits January 15, 2021 09:05

fix energy diagnostic for MPAS, clarify logic, turn on energy fixer f…

1f79cac

…or MPAS for adiabatic setup (so it follows the approach for other dycores)

1st pass of some of the total energy diags

c6e1b73

current "consistent" formulation

d5d4188

clean-up; science verified for temp tendency - looking into moisture

889998f

forgot to add dt in energy diagnostics

1044b83

accidently left debug code - revert

d32da86

Merge pull request #3 from jtruesdal/mpas_fixes

c05a960

Merge Lauritzen's mpas_fixes_phl branch with Truesdale's maps_fixes branch

fix energy diagnostic for MPAS, clarify logic, turn on energy fixer f…

087cab3

…or MPAS for adiabatic setup (so it follows the approach for other dycores)

1st pass of some of the total energy diags

284bfff

current "consistent" formulation

d4b03ab

clean-up; science verified for temp tendency - looking into moisture

b50e96d

forgot to add dt in energy diagnostics

552abbe

accidently left debug code - revert

c5ab165

fix bug for WACCCM-x (all dycores) and SE

f316fde

Merge branch 'mpas_fixes_phl' of https://github.com/PeterHjortLauritz…

b451b82

…en/CAM into mpas_fixes_phl

add support for moist q in subroutine get_cp

17f61a8

minor: FV3 mods + dp name change in prim_advance_mod

6e5feeb

infrastructure for 2 energies, move energy computation of physconst.F…

cd1b105

…90, regression tested for QPC6

infrastructure for vc_dycore/vc_physics,store initial temperature

c9e0a71

save initial values of zm and T, compute cv for height vertical coord…

979685a

…inate

use new energy routines in diagnostic code

1068f3c

update diagnostic description

690e77d

update energy diagnostics (and cleaner implementation)

92912ec

1st working version with dycore energy and diagnostics

73259b2

array size bug fix

a0c3821

shift energy fixer to use dycore energy formula, clean-up

96480f8

SE science bug

a010e4e

first total energy consistent working version with tendencies on dyco…

bfaff8a

…re side

Merge branch 'mpas_fixes' of https://github.com/jtruesdal/CAM-1 into …

e3adc0c

…mpas_fixes_phl

minor bug

7ecb422

jtruesdal and others added 13 commits June 3, 2021 13:34

better error coverage for user errors specifying decomposions, line n…

02e503f

…umbers added to similar error messages to aid debug, ChangeLog updates, remove unused variable and code in geopotential.F90, Make MPAS external non-optional again

Bringing mpas_fixes up to cam_development for PR

0e87e44

update ChangeLog and remove added space from mod

3852251

update date

610c8cf

update Changelog for merge/commit

462c073

energy consistent physics-dynamics coupling, some WACCM-x code

b5b093f

update pel development to cam6_3_020

fb5f501

change Rayleigh friction depth, activate dry air mass scaling

356afcf

name change to be consistent with physics energy diags naming convention

561c5f4

NAG fixes, new default for dry mass scaling

8645e09

jt updates to mpas dev pel, nag fixes, up to cam6_3_020

2e2a8d1

revert CESM2.2 bug fix changes that were accidently put in this branch

eb77cac

MiCurry commented Jun 8, 2021

View reviewed changes

PeterHjortLauritzen added 2 commits June 9, 2021 11:25

add thermodynamic initialization in FV

c6cccc8

add diagnostic calls on simple physics

51fd051

jtruesdal reviewed Jun 11, 2021

View reviewed changes

src/dynamics/mpas/dyn_grid.F90 Outdated Show resolved Hide resolved

PeterHjortLauritzen and others added 11 commits June 11, 2021 10:45

mod to make FV3 bit-for-bit

476656d

remove accidental commit of research code

4aa15ee

remove do from accidental merge

828bc7e

re-introduce CESM2.2 bug so that SE passes baseline tests

d1236fc

change computation of pmiddry to be consistent with equation of state…

fad14d3

… in MPAS

remove duplicate code

92b26cc

fix bug

f0bf038

fix BFB issue; if only Q is in water_species_in_air then the energy f…

69bf975

…ixer does not have liq and ice species. Fix for SPCAM

remove old tendency coupling code (#ifdef)

dbcafa7

update for use of optional parameter

a7aa452

Merge branch 'cam_mpas_dev_pel' of https://github.com/jtruesdal/CAM-1 …

9d5a70a

…into cam_mpas_dev_pel

jtruesdal pushed a commit that referenced this pull request Feb 8, 2024

Merge pull request #23 from jimmielin/hplin/fix_tskin

217721d

Fix TSKIN field to correct value

Initial addition for energy fixer for the MPAS dycore #23

Are you sure you want to change the base?

Initial addition for energy fixer for the MPAS dycore #23

Uh oh!

Conversation

MiCurry commented Jun 8, 2021

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

MiCurry Jun 10, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

PeterHjortLauritzen Jun 11, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

MiCurry Jun 10, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

MiCurry commented Jun 10, 2021

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

MiCurry Jun 10, 2021 •

edited

Loading

PeterHjortLauritzen Jun 11, 2021 •

edited

Loading

MiCurry Jun 10, 2021 •

edited

Loading