leaf biomass derivative and scaling

[rgknox]

We currently have an allometry function that assesses leaf biomass: see EDGrowthFunctionsMod.F90:Bleaf().

We also have an allometry function that returns the derivative of the leaf biomass function with respect to diameter (its actually the inverse of that derivative, change in diameter wrt change in biomass), see EDGrowthFunctionsMod.F90:dDbhdBl().

I noticed that in calling bleaf, we are scaling the result by this term "slascaler", which is the ratio of EDPftvarcon_inst%allom_d2bl_slascaler(cohort_in%pft)/EDPftvarcon_inst%slatop(cohort_in%pft).

However, we are not adding that term to the calculation of dDbhdBl(), as far as I can tell. I was expecting it to be there. The derivative of a function multiplied by a constant should maintain that constant.

[rosiealice]

Hmm. I'd thus be tempted to remove slascaler. It was meant to prevent the leaf allocation for high SLA plants being lower than low SLA plants, so that low SLA plants don't end up with a huge disadvantage, but now the pft specific allocation fractions override the need for that, since we can adjust the amount of leaf area externally if we so desire. I'd vote for getting rid of it. And good catch. :)

[rgknox]

Sounds good. When I remove it, I will modify the existing parameter file and adjust EDPftvarcon_inst%allom_d2bl1 to accomodate the loss of slascaler.

[rgknox]

Here is a quick test of the proposed fix. Results change a little bit, but not a big shift really.
bleaf-deriv-fix_plots.pdf

[rosiealice]

I guess I'd expect them to not be bfb. The solution is more correct and transparent though...

…

On Oct 3, 2017 7:33 PM, "Ryan Knox" ***@***.***> wrote: Here is a quick test of the proposed fix. Results change a little bit, but not a big shift really. bleaf-deriv-fix_plots.pdf <https://github.com/NGEET/fates/files/1354349/bleaf-deriv-fix_plots.pdf> — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#283 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AMWsQ6CpuFD32x6Q2yUD9WYO17kOb2RWks5souCBgaJpZM4PrbDt> .

[rgknox]

Same here, I was expecting the fix to not be b4b, and I'm happy to see that the results are not changed that much.

[ckoven]

fixed via #287. closing.