Create the input.nml file from global_control.nml.IN template #2733
Create the input.nml file from global_control.nml.IN template #2733dpsarmie wants to merge 7 commits into
Conversation
This represents the first pass of the parsing script. Tests show that it is working as intended. Next iterations will trim down the number of arguements as needed.
This commit reduced some of the redundant calls and settings that were default in the global_control.nml.IN. Changes were conservative to avoid future conflicts with variables that need to have dynamic values. Comments were also removed that were present in the previous commit.
Added a new function to compile the namelist after all environmental variables have been defined. Allows for different coupling cases to aggregate a finished namelist via atparse.
|
This is currently a work in progress. The UFS regression tests are currently being updated to work with GFSv17 and officially deprecating v16. This will probably cause changes to the global namelist template, so this will remain a draft until then. Once I'm done updating the RT side and have a finalized global_control.nml.IN file, I'll circle back to this and finish up the rest. The commit should be ~90% of the way there. Jun and I wanted to get this PR posted in case other people on the GW side wanted to look at the changes and to gather any input. The global_control.nml.IN file added here will be removed in the final commit (just here for testing). The 'unified' namelist will be on the UFS regression test side once the PR is complete. |
|
I've created a new branch since this one was too stale. I will open a new PR for this issue with the updated branch. Closing current PR. |
…arts for 4 tests after update to ESMF 8.8 (was NOAA-EMC#2796) (NOAA-EMC#2733) * UFSWM - Add DATM meshes which contain no land mask * UFSWM - Reinstate comparison of mediator restart files in 4 tests * CMEPS - Map DATM states bilinearly to OCN and ICE
1 participant
Description
This PR aims to bridge the GW ad regression test workflow. The regression tests have a new global_control.nml.IN for all global regression tests. The idea is to allow the GW framework to use this template to generate the input.nml that is used in the GW.
By pulling from a unified namelist, changes made on the regression test side can easily be added to the GW side and vice versa.
The parsing scripts in ./ush/ will be changed to use atparse to generate the input.nml file from the global_control.nml.IN namelist.
This PR will resolve #2731 and ufs-community/ufs-weather-model#1664.
Type of change
Change characteristics
How has this been tested?
Checklist