Add soil tillage for crops

[samsrabin]

Description of changes

This PR brings in the tillage code written by Sam Levis and Michael Graham and used in Graham et al. (2021, ERL, doi:10.1088/1748-9326/abe6c6). Low- and high-intensity tillage here work by increasing the decomposition rate of different soil carbon pools. These "decomposition multipliers" vary based on soil pool and how long it's been since the crop was planted; they are set with new paramfile variables till_decompk_multipliers and mimics_till_decompk_multipliers.

NOTE: This PR prevents tillage from actually being enabled. The reasoning is that we want to get the new parameters into the paramfile soon, but scientific testing will take a while.

Specific notes

Contributors other than yourself, if any: Michael Graham (@mwgraham), Sam Levis (@slevis-lmwg), Danica Lombardozzi (@danicalombardozzi)

CTSM Issues Fixed:

• Resolves Add "tillage" to bgc-crop model #112.

Are answers expected to change? No

Any User Interface Changes (namelist or namelist defaults changes)?

• Adds parameter tillage_mode with options off (default), low, and high.
• Adds parameter use_original_tillage_phases (false by default; see "Substantive changes" below).
• Adds parameter max_tillage_depth.

Testing performed, if any:

• Unit tests pass.
• aux_clm tests OK (NLFAIL) except for expected failures. Baseline saved as tillage.d43c6178c.

---

Substantive changes from original @slevis-lmwg / @mwgraham code

• I left out the dependence of decomposition multipliers on GDP. Code for distinguishing more- vs. less-developed countries was in the original branch, but the "less developed country" option was hard-coded to always be used. (Note that what I brought in reflects the description in Graham et al. (2021).)
• The original code calculates but ignores "days past planting" (idpp), using instead a calculation that seems incorrect for growing seasons that cross over into a new calendar year. This version uses idpp instead. However, the old version can be used by setting use_original_tillage_phases to true.
• Default max_tillage_depth is 26 cm. This was the intention of the original code, but it accidentally used 32 cm.

Limitations

• The effects of tillage on soil physical properties and albedo are not considered.
• Tillage is either enabled or disabled globally for all CFTs. There is no option to limit it to only some CFTs and/or gridcells.

Questions

Substantive

• @wwieder: Should I extend this to work with MIMICS decomposition? Doing so would require some scientific thinking and probably literature review. Yes; done.
• Should I extend this to work with FATES? I think the only reason I disabled that is because I'm unsure how/whether coarse woody debris, which FATES handles, should be affected by tillage. Yes.
• Should I bring in (an option to enable) GDP dependence, as was in the original code but unused? Unfortunately the "developed country" version uses hard-coded dates to determine decomposition multiplier; we'd surely want to generalize that but it'd require a discussion. Not doing this for now, as it seems to have only been an initial attempt.

Technical

• Should I move the 54 decomposition multipliers (currently hard-coded) to the parameters file? Yes; done.
• Is there a simpler way than this to get the active patch associated with a given crop column? Seemingly no.
• Should I add a test in testlist_clm.xml that uses the till testdef? And/or, should I edit one of the aux_clm tests to use it instead of crop?

Remaining work

• aux_clm
• Replace paramfile paths
• Scientific testing
• Add Tech Note documentation in Decomposition (link from Crop/irrigation)
• Add User Guide documentation in "Running special cases"
• Add till testdef.

[samsrabin]

@slevis-lmwg, can I look on your calendar to schedule a re-review?

[slevis-lmwg]

Yes, please do. My calendar is very busy, so I would prefer no sooner than Thursday afternoon.

…

On Mon, Oct 2, 2023, 12:02 PM Sam Rabin ***@***.***> wrote: @slevis-lmwg <https://github.com/slevis-lmwg>, can I look on your calendar to schedule a re-review? — Reply to this email directly, view it on GitHub <#2040 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AINPYFCJXEEQ3W3WMKZXZPTX5MFWBAVCNFSM6AAAAAAZQY7D26VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTONBTGU4TMNZRGI> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[samsrabin]

Tasks from re-review with @slevis-lmwg:

• Delete "Test identicality when using get_prev_days_per_year()" todo c8d6754
• Correct comment about "uses get_prev_calday() instead of get_curr_calday()". 7510431
• Do not till generic crops. d565176
• Warn instead of error when trying to enable tillage, allowing override. b86402e
• Delete add_tillage_to_paramsfile.py. 1a3a8b7

Additional tasks:

• Use ntill_intensities_max; set to private. ae56384, 693750c

[samsrabin]

I've figured out the reason my runs showed such a small tillage effect relative to Mike's! Turns out, I had incorrectly ordered the dimensions of the *_till_decomp_multipliers variables on the netCDF—they need to be reversed from how they are in the code.

Now that this is fixed (v2), the tillage effect is about as strong as it was with Mike's code.

[slevis-lmwg]

@samsrabin and I went over updates since I last reviewed this PR. Sam R will change a couple of existing "crop" tests to "till" tests and beyond that this PR is ready.