CCPP scheme simulator #378
Conversation
dustinswales
force-pushed
the
ccpp_scheme_sim
branch
from
1c899aa to
03b836b
Compare
|
Associated PR: NCAR/ccpp-physics#996 |
…p-scm into ccpp_scheme_simulator
scm/doc/TechGuide/chap_hsd.tex
Outdated
| Next, the physics namelist needs to be configured to: | ||
| \begin{enumerate} | ||
| \item Add data file, \execout{suite\_sim\_file} created in \ref{section:Creating_Custom_Data_for_Simulator} to the namelist. | ||
| \item Turn ``off'' all schemes except the cloud microphysics (see Fig. \ref{lst_css_nml_ex2}) |
There was a problem hiding this comment.
fix reference here - should point to listing 7.2
scm/doc/TechGuide/chap_hsd.tex
Outdated
| \end{itemize} | ||
| \end{itemize} | ||
|
|
||
| For example, in Listing \ref{lst_css_nml}, there are two active schemes, longwave and shortwave radiaiton, and five simulated schemes: PBL, gravity-wave drag, deep/shallow convection, and cloud microphysics. The radiation, gravity-wave drag and PBL schemes are all process-split, whereas convection and cloud microphysics are time-split. |
There was a problem hiding this comment.
It seems this is still pointing to the wrong Listing (looking at the most updated pdf)
| nfiles = 0 | ||
| fileSCM = [] | ||
| for nml in namelist: | ||
| fileSCM.append("../../../../scm/run/output_" + case_name + "_" + suite+"_"+nml+"/output.nc") |
There was a problem hiding this comment.
This does not work for my test runs as the 'nml' is not appended to my case directory.
For example, when executing this script, it looks for
../../../../scm/run/output_MAGIC_LEG15A_SCM_GFS_v17_p8_input_GFS_v17_p8/output.nc does not exist.
But it is named:
../../../run/output_MAGIC_LEG15A_SCM_GFS_v17_p8/output.nc
It looks like the 3d works in this way as well.
There was a problem hiding this comment.
@hertneky Good catch. It should be fixed now.
|
@dustinswales In input_GFS_v17_p8.nml, there is a parameter do_ccpp_suite_sim, but I don't see it documented. This looks like it is used by ccpp_suite_simulator.F90 and GFS_ccpp_suite_sim_pre.F90. Does this also need to exits and be turned on in the namelist? |
@hertneky Another good catch! |
|
@mkavulich This is ready to merge. Can you approve? |
There was a problem hiding this comment.
@dustinswales Sorry this review is so late. The code changes look good but I have some minor documentation corrections. Otherwise good to go.
Co-authored-by: Michael Kavulich <[email protected]>
Co-authored-by: Michael Kavulich <[email protected]>
Co-authored-by: Michael Kavulich <[email protected]>
Co-authored-by: Michael Kavulich <[email protected]>
Co-authored-by: Michael Kavulich <[email protected]>
Co-authored-by: Michael Kavulich <[email protected]>
Co-authored-by: Michael Kavulich <[email protected]>
This PR adds a new functionality, a physics process scheme simulator, to the CCPP-physics.
Data-driven tendencies can be used to drive process schemes (e.g. PBL, Deep-Convection, etc..). The control over which schemes are used are set in a new physics namelist, scm_data_nml. These data tendencies are read in during the CCPP initialization stage, and applied during the _run stage, if requested by the nml. There are new interstitial variables for the data tendencies.
To use this feature on the host-model side the Suite Definition File (SDF) needs to be modified to include the ccpp_scheme_simulator scheme, and GFS_ccpp_scheme_sim_pre.
There are examples included that exercise this new tool. Documentation will follow.