add utility to generate warm-start files for MOM6 and CICE6; modify cpld_gridgen as required by new ocnice_prep utility (was #945) #942
Conversation
DeniseWorthen
force-pushed
the
feature/ocnprep
branch
from
84c97f9
to
3463797
Compare
|
@GeorgeGayno-NOAA I've added documentation and unit test. The test checks for invalid settings in either the nml or csv file that the code uses. At this point I need to start thinking about how to add a regression test for this utility. Minsuk created the testing framework for cpld_gridgen for me---this isn't really my expertise. Do you have any suggestions for how to proceed w/ adding a regression testing framework? |
I can create the regression testing framework. Do you have the case(s) you want to run? Point me to the input data and a sample script. |
|
@GeorgeGayno-NOAA If all the weights and input data have been staged, do you want me to create baselines for this PR? |
The file whacker removed your weights files on WCOSS2 before I could host them. Sorry. Can you please regenerate? You can create your baselines on the RDHPCS machines. Thanks. |
|
@GeorgeGayno-NOAA Did you get the new weights files in place on jet? I'm getting a failure |
|
On WCOSS2, the new weights can be found at |
Sorry. Try it now. |
* use /work2 for cpld_grdgen and ocnprep on hercules. orion will use /work
|
@GeorgeGayno-NOAA One thing I just noticed is that the RT, at least for Here are the ocnprep baselines: jet: |
I hosted the baseline files in their official locations. |
|
@GeorgeGayno-NOAA Everything is ready except for Jet. I'll need to wait until the issues w/ lfs4 are fixed. |
* namelist is located in reg_tests/ocniceprep/parm
* remove MPI from CMakeLists.txt and hardwired outdir settings in cpld_gridgen and ocnice_prep RT scripts
|
@DeniseWorthen - one more thing before merging. The |
DESCRIPTION OF CHANGES:
Adds a new utility to down-scale MOM6 and CICE6 restarts from 1/4 deg tripole to a lower resolution, where they can be used as warm-start files.
TESTS CONDUCTED:
If there are changes to the build or source code, the tests below must be conducted. Contact a repository manager if you need assistance.
cpld_gridgenandocnice_prepconsistency tests locally on all Tier 1 machines. Done on Orion using d07c282. Done on Dogwood, Hera, Jet and Hercules using cd7ba2d.Describe any additional tests performed.
DEPENDENCIES:
A unit test testing the expected namelist and fields expected for ocean and ice has been added.
A simple regression test has been added which utilizes a staged copy of 1/4 deg ocean and ice ICs. These were obtained from the replay ICs staged on AWS
https://noaa-ufs-gefsv13replay-pds.s3.amazonaws.com/2021/03/2021032206. The ice restart is renamed generically asice.nc. The MOM restarts have been pre-processed using the following two NCO commandsDOCUMENTATION:
The code has been doxygenated and builds succesfully on hera. A basic description of the utility has also been added in the
sorc/ocnice_prep.fd/docs.Documentation for
cpld_gridgenhas also been updated to include a figure showing the tripole grid and to clarify the two types of weights produce: postitional weights (e.g. Cu->Ct) and mapping weights (e.g. tripole 1/4deg->rect 1/4 deg)ISSUE: