[[TOC]]

Reclaim land is a function used to create new stands from natural stands and to ensure consistency in fractions and pools. It’s straightforward in the general idea, but has a few caveats technically.

The routine reclaim_land creates a copy of an existing natural vegetation stand.
All vegetation of the to-be-cleared natural stand fraction (which becomes a new stand here) will be removed. Technically, it is not even established, but we use the vegetation carbon pools of the stand that we made the copy from to distribute it to the various pools. Soil variables are copied via source:trunk/src/soil/copysoil.c, but vegetation variables are non-existent on the new stand, i.e. the pftlist is empty. The code therefore uses the vegetation variables of the natural vegetation stand. However, these must not be changed, because that would affect the carbon pools on the remaining natural vegetation stand, which should not be altered.

In the simplest case, all vegetation carbon is simply allocated to the litter pools via litter_update.
In the coupling with IMAGE, but also for any timber harvesting in LPJmL-stand-alone runs, clearing forests can also contribute to [timber harvest](timber harvest) and carbon emissions from [[slash and burn]].
All vegetation-specific variables must not be altered here, because they are part of the natural vegetation stand that is not deforested. Therefore there is a local variable nind as copy of pft->nind that is used to account for the carbon adjustments in the subsequent carbon distribution steps:

• At first, tree carbon is allocated to the product pools via timber_harvest() if image_data->timber_frac is larger zero.
  To account for the removed carbon, the local variable nind is adjusted.

• Secondly, a fraction of the remaining vegetation carbon (non-harvested tree biomass) is burnt, as specified by the variable image_data->fburnt, which is passed on from IMAGE.
  Technically, source:trunk/src/landuse/timber_burn.c actually burns a fraction of the litter carbon that corresponds to the fraction of the vegetation carbon. The burnt litter carbon (which is virtual vegetation carbon if you want) is filled by the litter_update call just after that. That’s why nind is not adjusted here. I’m not really sure why we didn’t work with adjusting nind here as well, but there’s probably a good reason for this…

After allocation of carbon to the product pools or deforestation emissions, all remaining carbon is allocated to the litter pools via litter_update, also for grasses. If nothing has been harvested or burnt, all carbon goes to the litter pools.

So far, this only works in the coupling with IMAGE, but technically this can be used for any LPJmL application with timber harvest or [[slash and burn]].

Sibyll Schaphoff, Christoph Müller, Elke Stehfest

land-use change, IMAGE coupling, stand, grid cells