Incorrect litter calculation?

[bahaelaila7]

I am looking at how the calculations for woody and nonwoody dead pools.
It seems that the pools are passed fractions rather than the actual dead biomass

Extension-Biomass-Succession/src/PlugIn.cs

Lines 280 to 281 in 7802869

	ForestFloor.AddWoody(woodInput, cohort.Species, site);
	ForestFloor.AddLitter(foliarInput, cohort.Species, site);

A few lines earlier, foliarInput and woodInput are computed by multiplying nonwoody and woody fractions by partialMortality

Extension-Biomass-Succession/src/PlugIn.cs

Lines 226 to 231 in 7802869

	}
	double nonWoodyFraction = (double) cohort.ComputeNonWoodyBiomass(site) / (double) cohort.Data.Biomass;
	double woodyFraction = 1.0 - nonWoodyFraction;
	float foliarInput = (float) (partialMortality * nonWoodyFraction);
	float woodInput = (float)(partialMortality * woodyFraction);

And partialMortality comes from the MortalityEventArgs.Reduction, which itself is a fraction.

Extension-Biomass-Succession/src/PlugIn.cs

Line 221 in 7802869

double partialMortality = (double)eventArgs.Reduction;

Eg, here's the event emitted from RemoveCohort in UniversalCohorts' (reduction=1, ie 100% mortality upon cohort removal)
https://github.com/LANDIS-II-Foundation/Library-Universal-Cohort/blob/f19b04a6ca8fd6022eaaedf3896a87dd560b3a10/SpeciesCohorts.cs#L349-L367

foliarInput and woodInput are modified in cases where fire and/or harvest disturbances occur, and they do not seem to be modifications on fractions but rather modifications on supposed mass.
Harvest effects

Extension-Biomass-Succession/src/PlugIn.cs

Lines 245 to 246 in 7802869

	woodInput -= woodInput * (float)HarvestEffects.GetCohortWoodRemoval(site);
	foliarInput -= foliarInput * (float)HarvestEffects.GetCohortLeafRemoval(site);

Fire effects

Extension-Biomass-Succession/src/PlugIn.cs

Lines 264 to 268 in 7802869

	double woodFireConsumption = woodInput * (float)FireEffects.ReductionsTable[(int)SiteVars.FireSeverity[site]].CoarseLitterReduction;
	double foliarFireConsumption = foliarInput * (float)FireEffects.ReductionsTable[(int)SiteVars.FireSeverity[site]].FineLitterReduction;
	woodInput -= (float)woodFireConsumption;
	foliarInput -= (float)foliarFireConsumption;

I double checked ForestFloor; the AddWoody and AddLitter methods directly add the input without actually multiplying by cohort biomass

Extension-Biomass-Succession/src/ForestFloor.cs

Lines 19 to 27 in 7802869

	public static void AddWoody(double woodyBiomass,
	ISpecies species,
	ActiveSite site)
	{
	double currentWoodyDebris = SiteVars.WoodyDebris[site].Mass;
	SiteVars.WoodyDebris[site].AddMass(woodyBiomass, SpeciesData.WoodyDebrisDecay[species]);
	//PlugIn.ModelCore.UI.WriteLine(" BIOMASS SUCCESSION: Former Wood = {0}, New Wood = {1}", currentWoodyDebris, SiteVars.WoodyDebris[site].Mass);
	}

Extension-Biomass-Succession/src/ForestFloor.cs

Lines 33 to 55 in 7802869

	public static void AddLitter(double nonWoodyBiomass,
	ISpecies species,
	ActiveSite site)
	{
	IEcoregion ecoregion = PlugIn.ModelCore.Ecoregion[site];
	double siteAET = (double)EcoregionData.AET[ecoregion];
	//Calculation of decomposition rate for species litter cohort
	// Decay rate from Meentemeyer 1978. Ecology 59: 465-472.
	double leafKReg = (-0.5365 + (0.00241 * siteAET)) - (((-0.01586 + (0.000056 * siteAET)) * SpeciesData.LeafLignin[species] * 100));
	// From Fan et al. 1998 Ecological Applications 8: 734-737:
	//double leafKReg = ((0.10015 * siteAET - 3.44618) - (0.01341 + 0.00147 * siteAET) *
	//SpeciesData.LeafLignin[species]) / 100;
	//PlugIn.ModelCore.UI.WriteLine("Decay rate for {0} within {1} = {2}. LL = {3}.", species.Name, ecoregion.Name, leafKReg, SpeciesData.LeafLignin[species]);
	double decayValue = leafKReg;
	SiteVars.Litter[site].AddMass(nonWoodyBiomass, decayValue);
	}

Pool.AddMass is also expecting raw Mass

Extension-Biomass-Succession/src/Pool.cs

Lines 96 to 112 in 7802869

	public void AddMass(
	double inputMass,
	double inputDecayValue)
	{
	double totalBiomass = (Mass + inputMass);
	if (totalBiomass == 0)
	{
	DecayValue = 0;
	}
	else
	{
	DecayValue = ((Mass * DecayValue) + (inputMass * inputDecayValue)) /
	totalBiomass;
	}
	Mass = totalBiomass;
	}

Therefore I think partialMortality needs to be reduction multiplied by cohort.Data.Biomass. Am I missing something?

[rmscheller]

B,

Another good catch. When we migrated to UniversalCohortLibrary, we shifted to the interface standard from LeafBiomassCohorts, whereas Biomass Succession was originally designed around the BiomassCohorts standard. This change was overlooked during the migration. I committed 7.1.2 now.

Could you give it a test run?

Thanks!

R

[bahaelaila7]

Thank you for the swift reply!
Actually there seems to be some complication stemming from mismatch in Universal-Cohort. In few words: reduction means a ratio when it's by RemoveCohort but an absolute biomass value when it's by ReduceCohort, and there is no sure way to distinguish them by distrubanceType.

I opened an issue there:
LANDIS-II-Foundation/Library-Universal-Cohort#4

[bahaelaila7]

As expected
reduction from disturbances comes back as raw biomass. (When the event is emitted by ReduceCohort)
So multiplying with cohort.Data.Biomass will make litter calculations balloon

  Disturb Reduction: abiebals, 121 yrs, 80 Mg/ha, reduction=4, partial=True
  cohort reduced: abiebals, 121 yrs, 80 Mg/ha (disturbance:harvest), reduction=4
   BIOMASS SUCCESSION MORTALITY I: species=abiebals, age=121, disturbance=disturbance:harvest.
   BIOMASS SUCCESSION MORTALITY I: eventReduction=4.0, new_cohort_biomass=80, biomass*eventReduction=320.0.
   BIOMASS SUCCESSION MORTALITY II: species=abiebals, age=121, woodInput=208, foliarInputs=112.

Whereas for senescence (emitted by RemoveCohort) its reduction is just 1

  cohort removed: abiebals, 123 yrs, 1 Mg/ha (senescence), reduction=1
   BIOMASS SUCCESSION MORTALITY I: species=abiebals, age=123, disturbance=.
   BIOMASS SUCCESSION MORTALITY I: eventReduction=1.0, new_cohort_biomass=1, biomass*eventReduction=1.0.
   BIOMASS SUCCESSION MORTALITY III: ForestFloorInputs: Foliar=0.00, Wood=1.0.

(the new_cohort_biomass is just coincidentally 1 here)
Here is its last succession

Yr=3.    Spp=abiebals, Age=123.
Yr=3.    ANPPact=1.0, Mtotal=5.0, litter=5.00.
Yr=3.    DeltaB=-4.0, CohortB=5, Bsite=23960

[rmscheller]

Please test 7.1.2: https://github.com/LANDIS-II-Foundation/Extension-Biomass-Succession/blob/master/deploy/installer/LANDIS-II-V8%20Biomass%20Succession%207.1.2-setup.exe

Thanks!

R

[bahaelaila7]

Added comment:

09be3bd#r169815254

[rmscheller]

Yes, I see your point, thanks. I added a quick fix as suggested; presumably a reduction of 1.0 will be quite rare. At least this move us forward until the universal cohort library is fixed. New installer 7.1.2

[rmscheller]

B,

We followed your suggestion and revised the Universal Cohort Library so that it consistently returns fraction biomass mortality.

I subsequently altered Biomass Succession and did preliminary testing: https://github.com/LANDIS-II-Foundation/Extension-Biomass-Succession/blob/PartialReduction/deploy/installer/LANDIS-II-V8%20Biomass%20Succession%207.2.PartialReduction-setup.exe

(This installer includes the revised UCL dll)

Could you give it a test?

Thanks!!

Rob