Decouple monthly litter from number of resource groups assigned to a vegetation type

[dschlaep]

• Issue was identified with Update to SOILWAT2 v8.3.0 #586, see Update to SOILWAT2 v8.3.0 #586 (comment)

• The current approach to estimating SOILWAT2 litter is looping over resource groups and adding up monthly litter with v->veg[k].litter[m] += vegTypeBiomass[k] * SXWResources->_prod_litter[g][m];

• See

  STEPWAT2/sxw_soilwat.c

  Line 365 in 1a53494

  v->veg[k].litter[m] += vegTypeBiomass[k] * SXWResources->_prod_litter[g][m];

• This approach was implemented following Decouple monthly litter calculations from monthly biomass #447 which replaced v->veg[k].litter[m] = v->veg[k].biomass[m] * SXWResources->_prod_litter[m]; for good reasons.

• The problem with the current approach is that it adds the amount vegTypeBiomass[k] * SXWResources->_prod_litter[g][m] to litter for each resource group that is assigned to the vegetation type. Thus, litter of a vegetation type increases with each additional resource groups that is assigned to that vegetation type (even if that resource group currently has zero biomass). Even if a resource group has no biomass in a year, the mere assignment to a vegetation type causes that vegetation type to have additional litter.

• For instance, the values for the example run are as follows for the first month of the first year for vegetation type k=4 "C3-grass"

  • With assigned resource groups g=5 "a.cool.grass" and g=6 "p.cool.grass" (default: "p.warm.grass" assigned to the new "C4-grass" type): litter[k=4][0] = 1.095 = 1.5 * 0.23 + 1.5 * 0.5 with vegTypeBiomass[k=4] = 1.5, _prod_litter[g=5][0] = 0.23, _prod_litter[g=6][0] = 0.5 (note that bmass[g=5] = 1.5 and bmass[g=6] = 0.)
  • With assigned resource groups g=5 "a.cool.grass", g=6 "p.cool.grass", and g=7 "p.warm.grass" (reverting to previous assignment groups): litter[k=4][0] = 1.845 = 1.5 * 0.23 + 1.5 * 0.5 + 1.5 * 0.5 with vegTypeBiomass[k=4] = 1.5, _prod_litter[g=5][0] = 0.23, _prod_litter[g=6][0] = 0.5 and _prod_litter[g=7][0] = 0.5 (note that bmass[g=5] = 1.5, bmass[g=6] = 0. and bmass[g=7] = 0.)

One option is to calculate litter components based on the resource group's biomass instead of the vegetation type's, i.e., v->veg[k].litter[m] += bmass[k] * SXWResources->_prod_litter[g][m];, then following the example from above
- (new veg types): litter[k=4][0] = 0.345 = 1.5 * 0.23 + 0 * 0.5
- (previous veg types): litter[k=4][0] = 0.345 = 1.5 * 0.23 + 0 * 0.5 + 0 * 0.5

[kpalmqui]

@Huntermd

Daniel's proposoal above makes sense. We should plan to replace the calculation of :

v->veg[k].litter[m] += vegTypeBiomass[k] * SXWResources->_prod_litter[g][m];

with:

v->veg[k].litter[m] += bmass[k] * SXWResources->_prod_litter[g][m];

We will want to do this in a new branch to address this issue and also issue #588

[Huntermd]

Updated the calculation for litter as suggested to bmass from vegTypeBiomass .