Update Carbon isotope data to CMIP7

[ekluzek]

Description of changes

Start bringing in updates for the CMIP7 Carbon isoptope data.

Specific notes

Contributors other than yourself, if any: @slevis-lmwg

CTSM Issues Fixed (include github issue #):
Work on #3346
Fixes #3528

Are answers expected to change (and if so in what way)? When it's turned on

Any User Interface Changes (namelist or namelist defaults changes)? Default for C13/C14 data will be updated

Does this create a need to change or add documentation? Did you do so? No No

Testing performed, if any: None yet

Definition of Done:

• Change units of time expected on file from years to days
• Handle the opposite direction that the latitude sectors are for CMIP6 and CMIP7
• Do enough testing of a hardcoded version to CMIP7 that science experiements can run
• Make a branch tag of it for those purposes
• Fix the latitude problem Preliminary CMIP7 C14 Carbon isotope dataset is upside down! #3528

[ekluzek]

Note, NEON/PLUMBER2 cases use a SSP scenario, which means they will point to CMIP6 isotope data. We'll have to do some reworking for them to point to the CMIP7 datasets.

[ekluzek]

Not fully tested, but seems to have the minimum required to get this to work now.

[slevis-lmwg]

Thank you @ekluzek

1. I see your point, we may not need to preprocess the c13 and c14 files, which is good I think. But we may want to add a comment stating that the new code is not backwards compatible with CMIP6.
2. I looked at your code changes. In this review I expressed a couple of concerns based on my inspection of the files. Let me know if you want to follow up.

[ekluzek]

I realized there's another part where it's hardcoded to the old dataset. Here in PhotosynthesisMod

               ! determine latitute sector for radiocarbon bomb spike inputs
               if ( grc%latdeg(g) .ge. 30._r8 ) then
                  sector_c14 = 1
               else if ( grc%latdeg(g) .ge. -30._r8 ) then            
                  sector_c14 = 2
               else
                  sector_c14 = 3
               endif

So that needs to be handled.

[ekluzek]

The history variables for this are default off, but are: RC14_CANAIR, RC13_CANAIR

[ekluzek]

OK, and it looks like RC14_CANAIR needs to be added as a history output variable.

[ekluzek]

Here's a conversion to Delta using ncap2, I'm trying out:

 ncap2 -s 'c14ratio = 1.e-12' -s 'PDB = 0.0112372' -s 'preind_atm_del13c = -6.' \
-s 'preind_atm_ratio = PDB + (preind_atm_del13c * PDB)/1000.0' -s 'c13ratio = preind_atm_ratio/(1.0+preind_atm_ratio)' \
-s 'Deltac13 = (RC13_CANAIR/PDB - 1.0)/1.0e-3' -s 'Deltac14 = (RC14_CANAIR/c14ratio - 1.0)/1.0e-3'  infile.nc outfile.nc

[ekluzek]

Results are looking good so far. But, I'm running a test for a few years to make sure time is correct.

[ekluzek]

Ok. It looks like there's something wrong with the time interpolation. So I'm looking into this more.

[ekluzek]

I added some logging for time, and saw there were problems with both c13 and c14 time. So I've modified it so that the time is in units of CE year, and can then be directly compared to dateyear for the current time. This should resolve the issues with time. Running a test for this now.

[wwieder]

Thanks for working on this and providing updates, Erik.

[ekluzek]

Tag to make:
branch_tags/CMIP7_ciso.n01.alpha-ctsm5.4.CMIP7.11.ctsm5.3.075

[wwieder]

I'm spinning this up in an ne30 bgc-crop test. If we're OK with how these results look, I suggest we don't worry about maintaining the flexibility of using CMIP6 vs. CMIP7 datasets for carbon isotopes.

Other input data are easier to maintain with streams files (e.g., #3471 to switch between CMIP6 and CMIP7 data), but given the brittle nature of how we handle C isotopes I suggest we only support CMIP7 13C and 14C datasets moving forward. It's not not worth the overhead costs of maintaining code to ingest both files at this point. Happy to have further discussion on this if it's helpful.

[ekluzek]

@slevis-lmwg and I went over this, and I'll bring this into the ctsm5.4 alpha branch largely as is at this point, even though it has problems. It's already been used to run simulations so we want to preserve that version. And the work to do this with streams should be an independent PR.

[slevis-lmwg]

Ok, not spending too much time here, since the code has been tested, and we are aware of problems if I am not mistaken.

I am approving with Erik's idea in mind that we keep a record of this code because it was used in specific simulations. However, I want to also introduce another idea: What if we revert entirely back to the cmip6 code here? Doing so allows backwards compatibility with the cmip6 isotope data, now that we are proceeding with streams for cmip7.

[ekluzek]

A new problem I realized was that we have CMIP6 SSP datasets, which don't work in this setup. The way forward we decided on is to turn Carbon isotopes off for SSP scenarios even for clm6_0. And we might as well turn carbon isotopes off for some compsets such as single point cases.

I'm going to make a tag of this branch, and then close without merging once we have the streams way forward up and working.

[ekluzek]

I tagged the fixed version as branch_tags/CMIP7_ciso.n02.alpha-ctsm5.4.CMIP7.11.ctsm5.3.075, so the fix is preserved in a tag in case we need it. This will also allow us to compare the new answer with streams to this version to validate the streams method.