SMINN_TO_PLANT_FUN units incorrect in history

[slevis-lmwg]

Brief summary of bug

A user suggests that SMINN_TO_PLANT_FUN should have units gN/m3/s in history rather than gN/m2/s.
From a quick look with @wwieder and @olyson, they agree.
The discussion is here:
https://bb.cgd.ucar.edu/cesm/threads/output-variables-from-fun-model.9016/post-52292
Before correcting, I would like a careful look to confirm, unless others are sure beyond doubt.

General bug information

CTSM version you are using:
ctsm5.1.dev153

Does this bug cause significantly incorrect results in the model's science?
NO

Details of bug

biogeochem/CNVegNitrogenFluxType.F90: call hist_addfld1d (fname='SMINN_TO_PLANT_FUN', units='gN/m^2/s',&

[olyson]

Actually, in looking at this further, I think the SMINN_TO_PLANT_FUN is correct (gN/m^2/s) because it is not dimensioned by soil layers. The user was outputting SMINN_TO_PLANT_FUN_NO3 which I assume corresponds to sminn_to_plant_fun_no3_vr, which has a vertical dimension. In the code, the variable is described as being gN/m^2/s

real(r8), pointer :: sminn_to_plant_fun_no3_vr_patch (:,:) ! Total layer no3 uptake of FUN (gN/m2/s)

but it should be gN/m^3/2, right?

@wwieder @slevis-lmwg