Implement CCPPized check_energy_chng and check_energy_fix #145
Conversation
…s); namelist variable control for print_energy_errors
…ld check_energy_chng SDF
…mpatibility with check_energy_chng
|
Hi @nusbaume, I've added |
There was a problem hiding this comment.
Great work @jimmielin! I realize it appears that I have a bunch of requests, but hopefully most of them are fairly straight-forward to implement, or can be pushed off to a future PR. I also tried to only make a suggestion for a given standard name in a single comment, but could have certainly repeated myself.
Also, some of my standard name suggestions are different than what was in the spreadsheet. This is mostly motivated by what the standard name usage actually looks like now that we have an example, and by recent standard name rule discussions we have had with NOAA and others. Of course I am happy to receive push back on it if you or the scientists disagree.
Thanks again!
| character(len=64), intent(out) :: name ! parameterization name for fluxes | ||
| real(kind_phys), intent(out) :: flx_vap(:) ! boundary flux of vapor [kg m-2 s-1] | ||
| real(kind_phys), intent(out) :: flx_cnd(:) ! boundary flux of liquid+ice (precip?) [m s-1] | ||
| real(kind_phys), intent(out) :: flx_ice(:) ! boundary flux of ice (snow?) [m s-1] |
There was a problem hiding this comment.
remove "(snow?)"
There was a problem hiding this comment.
Thanks for resolving most of my concerns @jimmielin! I did have a few unresolved requests, but once those are resolved then I think this PR will be good to go. Of course please let me know if something is unresolved but you are unsure as to why (as I may have just missed it). Thanks!
jimmielin
force-pushed
the
hplin/check_energy
branch
from
428bc10 to
419bc0e
Compare
|
Regression tests using the latest commit in ESCOMP/CAM have passed (with expected existing failures): Filling in ChangeLog tag |
Tag name (The PR title should also include the tag name): `atmos_phys0_07_000` Originator(s): @jimmielin List all `development` PR URLs included in this PR and a short description of each: * #141 by @mattldawson @boulderdaze * #147 by @peverwhee * #144 by @mwaxmonsky * #151 by @mattldawson * #145 by @jimmielin List all test failures: N/A
Fixes #114
This PR fixes the following NCAR/atmospheric_physics Github issues: #114
Implements check_energy_chng. The routine computes total energies using physics and dycore formula and takes in boundary fluxes of vapor, liquid+ice, ice, sensible heat, and writes to log significant energy conservation errors.
In order to take in the scheme name and fluxes from the last calling physics scheme (usually zero) a check_energy_zero_fluxes has to be ran before the scheme, then the physics scheme to be checked, then check_energy_chng.
Implements check_energy_fix. The routine computes the heating rate required for global mean total energy conservation which is later applied by apply_heating_rate.
Supporting routines include the global mean calculator for total energy (check_energy_gmean) - stored in separate folder to prevent CAM from building it.
The global means are very useful for diagnosing model state (as the computed energy numbers include all model state incl. constituents) through a simple one-line printout. check_energy_gmean_diagnostics implements this.
At the end of the timestep check_energy_save_teout has to be ran in order to save total energies at the end of the timestep for use in the next timestep.
Diagnostics of intermediate quantities used in check_energy_diagnostics.
List all existing files that have been added (A), modified (M), or deleted (D),
and describe the changes:
List and Describe any test failures: N/A
Summarize any changes to answers: none