Update values of fates_hydro_solver and fates_hydro_vg_alpha_node

[jennykowalcz]

This PR changes the default values of the parameters fates_hydro_solver and fates_hydro_vg_alpha_node.

Description:

The default for hydro_solver is changed from 1 (Taylor 1D) to 2 (Picard 2D) to increase the hydro model stability and avoid having it crash under very dry conditions. See discussion 1187.

The default for fates_hydro_vg_alpha_node is changed from 0.005 to 0.12 MPa^-1. The previous small value led to unrealistic values of stem conductivity fraction over the Amazon Basin when using the van Genuchten hydraulic transfer function for plants (i.e. fates_hydro_htftype_node = 2 for all organs). See issue 1228. The new larger value was chosen to approximate the results from the default TFS hydraulic transfer function (fates_hydro_htftype_node = 1) with default hydro parameter values.

Disclaimer: I have not tested this is a global run, just regionally over tropical South America.

Collaborators:

@rgknox

Expectation of Answer Changes:

The change to fates_hydro_vg_alpha_node will change model output significantly if you are using the van Genuchten hydraulic transfer function.

Checklist

If this is your first time contributing, please read the CONTRIBUTING document.

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:

CTSM (or) E3SM (specify which) test hash-tag:

CTSM (or) E3SM (specify which) baseline hash-tag:

FATES baseline hash-tag:

Test Output:

[rgknox]

This is great @jennykowalcz , we will get some review going on this

[rgknox]

@jennykowalcz There is agreement to switch to Picard, and the updated VG parameters yield better results in the tropics. Have you tested VG and/or found if it's a stable alternative outside the tropics? I see that you also updated the VG parameters for the non-tropical PFTs as well. Are the changes for those also a stability improvement?

[jennykowalcz]

@jennykowalcz There is agreement to switch to Picard, and the updated VG parameters yield better results in the tropics. Have you tested VG and/or found if it's a stable alternative outside the tropics? I see that you also updated the VG parameters for the non-tropical PFTs as well. Are the changes for those also a stability improvement?

Hi @rgknox ! I have not done any testing outside the SA tropics. I can do a quick set of global low-res runs with the current and new parameter files. I updated the values for all PFTs because all of the other hydro parameters have the same values for all PFTs.

[jennykowalcz]

Some results from the global runs so far. These are in no-comp mode with compset IELMFATES with a 4x5 degree grid.

This is from 100 model years, so vegetation is not at equilibrium yet. Most of the biomass is in broadleaf tropical trees (evergreen and hydro-deciduous), followed by broadleaf evergreen extratropical trees.

Side note, it is very cool that FATES-Hydro is working globally out of the box! 😃

[glemieux]

@jennykowalcz just to clarify, "out of the box" refers to the TFS default, correct? Also, could you elaborate on the settings for each test regarding the organs?

Do you recommend going to vG as default and for which flux?

[jennykowalcz]

@jennykowalcz just to clarify, "out of the box" refers to the TFS default, correct? Also, could you elaborate on the settings for each test regarding the organs?

Do you recommend going to vG as default and for which flux?

Hi @glemieux! No, I think TFS should be kept as the default hydraulic transfer function (fates_hydro_htftype_node = 1, 1, 1, 1 ;).

By "out of the box" I meant with the default parameter file with all PFTs, and yes, TFS hydraulic transfer function.

For the test here I ran three simulations: 1) TFS hydraulic transfer function (for all hydro organs), 2) vG hydraulic transfer function (for all hydro organs) with vg_alpha = 0.12 for all PFTs ("vG, new") and 3) vG hydraulic transfer function (for all hydro organs) with vg_alpha as set in the previous default parameter file for all PFTs (the value is 0.0005 for PFTs 1-5 and 0.005 for the other PFTs.) The goal of the parameter update to vg_alpha is just to get results closer to what you'd get with TFS.