photosynthesis

[ckcastillo@purdue_edu]

Hi! I'm tracing how photosynthesis is calculated in CLM3.5-DGVM. I've seen so far that total photosynthesis in CanopyFluxes is:

fpsn = psnsun*laisun + psnsha*laisha,

where fpsn is in umol/m2 patch area/s, right?

But, in SurfaceRadiation, laisun and laisha are calculated from elai, which is in turn calculated from tlai in DGVMEcosystemDyn, while tlai is based on lai_ind.
I have not yet found in the code any section where the average individual lai's are scaled up by multiplication by population density (nind) or something similar in order to get the total photosynthesis from all the individual pft's in a patch. If fpsn originates from individual leaf mass and lai, then how is it able to represent the total photosynthesis in the patch?

I ask because I've implemented a "deforestation run" in which the population density is thinned every year and no new saplings are allowed to estabish in the tropical band, yet despite initial decreases in elai/tlai, these steadily increase beyond those of the control scenario and the photosynthesis fpsn ends of higher in the long run (50-100 years). While it's true that the sunlit fraction of the canopy increases and grasses spread, I didn't think that these might be enough to compensate.

 

[slevis]

If I remember, we assumed that the lai of an individual in m2 leaf per m2 ground would be the same as the lai of the canopy. This assumption implies that there is no bare ground between individuals and that one individual never grows in the space of another. Grasses are treated as a single individual per gridcell, so their lai gets scaled by fpcgrid. Does that seem right? As always, feel free to make changes to the code according to your different assumptions.

Sam

 

[ckcastillo@purdue_edu]

Thanks for clarifying the assumptions!
Given these, how does the program deal with the addition of saplings then? If I understand right, nind, fpcgrid, total leaf mass (and total veg C pools in general) will increase but the average individual's leaf mass might decrease.

 

[ckcastillo@purdue_edu]

Also, in the Establishment and Light modules, the population density nind of trees is reduced if fpc exceeds a certain maximum. In these cases, nind and the litter pools are updated but the average individual leaf mass remains unchanged - understandably so since taking away individuals doesn't affect the characteristics if an average individual. But since the total leaf mass changes, shouldn't the photosynthesis be affected in some way which is not seen if it is computed based on average individual leaf mass and lai? Or, given the assumptions, would it be correct to say that photosynthesis per m2 remains the same though the pft now covers less area? I'm unsure if the latter is the case since the land diagnostics package that reported photosynthesis in PgC/yr in the line plots still indicated that my scenario with thinning of tree populations still had higher photosynthesis in the long run (20+ years).

 

[ckcastillo@purdue_edu]

Good day! I have another follow-up question to the assumptions stated above. As far as I can tell, nind and crownarea do not directly factor in the calculations of one or the other. But if it is implied that there is "no bare ground between individuals and that one individual never grows in the space of another", then wouldn't there necessarily be an inverse relationship between the two? For example, if nind is 1/m**2 then wouldn't the crownarea have to be 1m**2 so that it fulfilles the assumption that there is no space and no overlapping between individuals? And if, say, the nind decreases to 0.5/m**2, does this mean the crownarea increases to 2m**2?

I know that crownarea shouldn't be based on nind but on stem diam etc. as is done in the model, but I am confused by these assumptions.

Also, both nind and crownarea figure in the calculation for fpcgrid:
fpcgrid = crownarea*nind*fpc_ind
How do the above assumptions apply in this case? Fpc_ind is likewise a fraction with no units right?
I thought that fpcgrid would be (fpc_ind)*(number of individuals) but it's not
clear to me that (number of individuals) is what is achieved by multiplying the (area each individual tree takes) up by (the number of individuals per m**2 of the patch area).

Any light you can shed on these queries would be helpful. Many thanks!

 

[slevis]

To me it makes sense to have nind and crownarea independent of each other. The crownarea of a tree pft increases with increasing carbon. The number of individuals increases with establishment. As nind and/or crownarea increase, so does the pft area in the grid cell until it hits the max allowed limit:

fpcgrid = crownarea*nind*fpc_ind

In this, which is Eq. 28 from the dgvm technote, the units are:
(m^2 ground / individual) * (individuals / m^2 vegetated area) * (m^2 actually covered / m^2 ground) = m^2 actually covered / m^2 vegetated area

When fpcgrid cannot increase any more, we have competition for light (see Chapter 2.7 in the dgvm technote).

Sam

 

[ckcastillo@purdue_edu]

Thanks for clarifying the units for fpcgrid!

I do agree that it makes more sense that nind and crownarea are independent of each other in the way you describe - which is why I find the assumption stated above (for computing photosynthesis) a bit confusing. Having no bare ground or overlapping in between individuals seems to imply that nind and crownarea are linked.