Extend the MedlynConductance slope parameter to support SpaceVaryingInput

[imreddyTeja]

Purpose

Extend MedlynConductanceParameters to support a slope parameter that is
<: Union{FT, ClimaCore.Fields.Field}
Closes #756 -- this will automatically close issue 756 on PR merge

To-do

• The clm_data file for vegetation properties and the script that generates it should be changed to reflect
  the actual units.

Content

• MedlynConductanceParameters struct change to support g1 as Field
• Constructor for MedlynConductanceParameters change to reflect struct changes
• Add for loop for tests that check g1 as Float and Field
• unrelated: change way the vcmix kwarg is passed to the FarquharParameters constructor to unify code style
• Change g1 in land.jl to use the map in clm_data (see map below)

Previous map image (has mask over ocean and no unit conversion)

Re-made map with no mask and converted to sqrt(Pa)

---

• I have read and checked the items on the review checklist.

[Sbozzolo]

Thank you! Please squash commits that need to be squashed before merging

[kmdeck]

Thank you! This looks great

[kmdeck]

Would it be possible to update the map of g1 here to plot in units of sqrt(Pa)? Why is value over the ocean nonzero/non NaN?

[imreddyTeja]

Would it be possible to update the map of g1 here to plot in units of sqrt(Pa)? Why is value over the ocean nonzero/non NaN?

I added an re-made map under the original in the pull request. The original map in the PR had an ocean mask applied to it. vegetation_properties_map.nc lists the_FillValue as NaN, but there are no NaNs in the data. It also seems like there are no zero values in the data. I believe the value over the ocean is 284.60498 $\sqrt{Pa}$ because the median and maximum of the data is also 284.60498 $\sqrt{Pa}$

[kmdeck]

Would it be possible to update the map of g1 here to plot in units of sqrt(Pa)? Why is value over the ocean nonzero/non NaN?

I added an re-made map under the original in the pull request. The original map in the PR had an ocean mask applied to it. vegetation_properties_map.nc lists the_FillValue as NaN, but there are no NaNs in the data. It also seems like there are no zero values in the data. I believe the value over the ocean is 284.60498 P a because the median and maximum of the data is also 284.60498 P a

@braghiere It is not a big deal, but does it make sense to you why the ocean has a nonzero/nonNan value for medlyn slope?

@imreddyTeja Thanks for remaking the plot! I think we can merge.

[braghiere]

@kmdeck It was just the way it was constructed. Bare soil and water were assigned values in the map but those should be ignored.

@imreddyTeja The values look wrong now. Medlyn g1 is in sqrt(kPa) which can be anywhere from ~1.5 to ~6. The values were "right" but unit names were wrong.

[imreddyTeja]

Could I have performed the unit conversion incorrectly? I converted to sqrt(Pa) by multiplying by 10^(3/2). This is what the map looks like with no unit conversion or mask:

The data without unit conversion ranges from 1.62 to 9.0

[braghiere]

Could I have performed the unit conversion incorrectly? I converted to sqrt(Pa) by multiplying by 10^(3/2). This is what the map looks like with no unit conversion or mask:

The data without unit conversion ranges from 1.62 to 9.0

Got it! I see you converted from sqrt(kPa) ) to sqrt(Pa) to be in the same units as VPD. That makes sense! Thanks.