noresm3_0_042_cam6_4_121: override CO2 surface source to send to land/ocean#294
Conversation
|
@tjiputra - I think I've implemented the behavior you wanted in NorESMhub/NorESM#825. |
|
@mvertens - thank you. Should i test this on beta22 or other specific branch? Using pr-noresm3_0_beta22 branch, I got this error during case.build: |
|
@tjiputra - yes - please test this on beta22. However, this code is still in a PR status #294. For you to test this - you need to do the following edit .gitmodules to have the following Then you can run the following That should give you the CAM code you need. Please let me know if this is working for you. |
|
@mvertens - But the run crashes: (shr_stream_verifyTCoord) ERROR: calendar dates must be increasing Case directory (Olivia): |
|
@tjiputra - I had not tested the 1pctCO2 and I have you the wrong namelist: If this does not work - I'll try to duplicate the problem and resolve it. |
|
@mvertens - the same error occurs again. It seems that it tries to read and cycle the first year even if I set stream_co2_surface_source_year_first = 0 |
|
@tjiputra - that's very odd. I'll track this down. Sorry you ran into this. |
|
@tjiputra - the problem is not in the code - it's in the dataset. I generated a time coordinate for this dataset and it seems to have duplicated the first year: |
|
@tjiputra - I have fixed this file on olivia and tested that the following user_nl_cam works: |
gold2718
left a comment
There was a problem hiding this comment.
I have several clean-up and documentation requests as well as some bug fixes and safety features.
|
@mvertens - works and runs well. I will keep this running longer and make sure the behavior of land/ocn/atm is as expected. Could you also create a stream co2 input file for historical: Thank you. |
|
@tjiputra - I have created the new dataset - here is the new namelist (assuming you want to start the data at 1850) |
|
@mvertens - Could you correct the the unit of CO2 sent to both land and ocean, which expect [ppm] unit but receive [mol/mol]. Simply multiply them by 1e6. |
|
@tjiputra - thanks for catching this! I'll put the fix in. |
|
@tjiputra - I have fixed the unit conversion and pushed it back to my branch. |
|
@mvertens - being inexperience with git, how do i get the updated code? remote: Enumerating objects: 5, done. |
|
Try this: |
gold2718
left a comment
There was a problem hiding this comment.
This looks mostly good. I will fix merge conflict and a couple of missing items.
|
@tjiputra - if you do a clean checkout of pr-noresm3_0_beta22 you should get this tag. |
Summary: New ability to have co2 sent to land/ocean be different than used in radiation/chemistry
Purpose of changes: address ability to run the experiments outlined in NorESMhub/NorESM#825 and #296.
Changes made to build system: None
Changes made to the namelist: New namelist group co2_surface_source_stream_nl
To activate this behavior edit user_nl_clm to have the following:
If you want to send 1pct CO2 to land/ocean:
NOTE: a new namelist was also introduced for nitrogen deposition -
stream_ndep_taxmodeChanges to the defaults for the boundary datasets: None
Substantial timing or memory changes: None
Detailed description of changes:
The above custom namelists must be set in user_nl_cam - there are no out of the box defaults.
The 1850 and 1pct CO2 files in
$DIN_LOC_ROOT/atm/waccm/lb/cmip7_ghg_version20260607/only had [lat,time] dimensions - and CDEPS cannot work with such files. As a result - 20deg longitude datasets were created for the above scenarios.Verified that using the following compset 1
at
ne16pg3_tn14- the co2 diag field sent to the ocean had the expected behavior that was different than the co2 being used internally for the radiation in CAM.Testing:
fixes #296