Add Johnson and Berry 2021 electron transport model option

[JessicaNeedham]

Description:

This PR adds the Johnson and Berry 2021 model of electron transport response to irradiance as an option. Julien Lamour et al. have found this formulation has a better goodness of fit to gas exchange than the original FvCB model. See issue #1337 for details and discussion.

Note this needs to come in after #1262 . Parameter file and namelist option changes came in with #1344 . This requires the parameter file to namelist migration on the HLM side. Right now I am testing with this branch which has the new electron transport model switch but none of the other new ones.

The current configuration compiles and runs with the JohnsonBerry2021 namelist option switched on, but I haven't evaluated it yet. I'm still working on adding a history variable to track photosynthesis limitation by Rubisco and RuBP.

Collaborators:

@alistairrogers @JulienLamour @rgknox @rosiealice @glemieux

Expectation of Answer Changes:

With the FvCB1980 namelist option this should be b4b.
With the new JohnsonBerry namelist option we expect lower photosynthesis at intermediate irradiance.

Checklist

All checklist items must be checked to enable merging this pull request:

Contributor

• The in-code documentation has been updated with descriptive comments
• The documentation has been assessed to determine if updates are necessary

Integrator

• FATES PASS/FAIL regression tests were run
• Evaluation of test results for answer changes was performed and results provided

Documentation

• Technical Note update:
• User's Guide update:

Test Results:

Compiles but no formal testing yet.

[rgknox]

This is based on branch leaf_funcunit_tests, here is the diff until the former is integrated:

rgknox/fates@leaf_funcunit_tests...JessicaNeedham:fates:jb-electron

[JessicaNeedham]

Chatting with @rgknox yesterday about how best to track how much photosynthesis is Rubisco or RuBP limited. It is possible but would require allocating memory to large arrays. An alternative is to use unit tests in python. This would call the photosynthesis code in isolation and would allow us to test under which conditions photosynthesis is Rubisco or RuBP limited using each electron transport model. @JulienLamour and @alistairrogers - do you think the unit test approach would work for you or is this something we should make a FATES history variable.

[JessicaNeedham]

This is the mean GPP over the last 10 years of a 20 year SP mode run on a 4x5 grid. Looks like GPP is ~5-10% lower in parts of the tropics with the new JB model.

Thanks @adrifoster for the SP mode parameter file! ☺️

[alistairrogers]

@JessicaNeedham this is great progress thanks! Is it possible to output global GPP (Gt yr-1) with the two model formulations?

[JessicaNeedham]

@JessicaNeedham this is great progress thanks! Is it possible to output global GPP (Gt yr-1) with the two model formulations?

I get 105 Pg C yr-1 for the FvCB model and 97 PgC yr-1 for the JB model. Note that this isn't the final calibrated parameter file so it might be worth focusing more on the relative difference.

[alistairrogers]

Thanks, yes that was my goal.