leafn_to_litter with biofuel harvest: Potential imbalance?

[samsrabin]

It looks like there are three ways that leafn_to_litter can be calculated in CNOffsetLitterfall().

If CNratio_floating is true:

CTSM/src/biogeochem/CNPhenologyMod.F90

Lines 3232 to 3238 in 1e2e2c3

	if (leafc(p) == 0.0_r8) then
	ntovr_leaf = 0.0_r8
	else
	ntovr_leaf = leafc_to_litter(p) * (leafn(p) / leafc(p))
	end if
	leafn_to_litter(p) = fr_leafn_to_litter * ntovr_leaf

If CNratio_floating is false and use_fun is true:

CTSM/src/biogeochem/CNPhenologyMod.F90

Lines 3188 to 3189 in 1e2e2c3

	leafn_to_litter(p) = leafc_to_litter(p)/leafcn_offset(p) - leafn_to_retransn(p)
	leafn_to_litter(p) = max(leafn_to_litter(p),0._r8)

If CNratio_floating is false and use_fun is false:

CTSM/src/biogeochem/CNPhenologyMod.F90

Line 3212 in 1e2e2c3

leafn_to_litter(p) = leafc_to_litter(p) / lflitcn(ivt(p))

I'm wondering if it's possible for there to be a discrepancy here with how much N is removed during biofuel harvest, which happens here (before leafn_to_litter is calculated):

CTSM/src/biogeochem/CNPhenologyMod.F90

Line 3126 in fdbb234

leafn_to_biofueln(p) = t1 * leafn(p) * biofuel_harvfrac(ivt(p))

Note that leafn is never reduced by leafn_to_biofueln, which I think would cause a problem in at least the CNratio_floating == .true. case.

Our testing might not be set up to catch such discrepancies and their resulting N imbalances. It's only the Smallville tests that have any biofuels, right?

Pinging @billsacks, as it looks like you did the final testing for the bioenergy crops PR #884.

[billsacks]

I haven't gotten my head into everything you are describing here, but it seems like your main point may be this one:

Note that leafn is never reduced by leafn_to_biofueln

and wanted to confirm that you saw this:

CTSM/src/biogeochem/CNNStateUpdate1Mod.F90

Line 284 in 1e2e2c3

ns_veg%leafn_patch(p) = ns_veg%leafn_patch(p) - nf_veg%leafn_to_biofueln_patch(p)*dt

But maybe you're saying that there's an issue even despite that? Let's talk more.

[samsrabin]

It actually seems like this is fine. leafn_to_litter is in all cases dependent on leafc_to_litter, which DOES have biofuel fraction removed. It's also dependent on leafc, which like leafn is not updated; since biofuel harvest affects leaf C and N equally, then leafc/leafn will be the same whether biofuel harvest is removed from both or neither.