net terrestrial C uptake in GCAM about gcam-core HOT 3 CLOSED

The simple carbon cycle model within GCAM is documented at https://jgcri.github.io/gcam-doc/land.html
Note that this is different from the carbon cycle model that is part of Hector, documented at https://jgcri.github.io/gcam-doc/hector.html, which is only resolved at the global level, in which is not used for the reported terrestrial emissions or sequestration from the land system in a GCAM model run.
Despite the name "simple," this is one of the most difficult parts of the model for a user to modify in such a way as to achieve some desired result. The net reported emissions or sequestration, for any region and model time period, is dependent on all of the land use decisions that have been made up until that time period, considering the average vegetation and soil carbon densities of each modeled land-use type. Because the average carbon densities of each land-use type are assumed time-invariant (e.g., the maximum vegetation carbon density of a tropical rainforest is 120 tC/ha, or 12 kgC/m2), a user can't just increase assumed carbon densities going into the future in order to represent enhanced uptake.
Just as a final note here, the reported net emissions in sequestration make the most sense when reported by region, or alternatively by basin within a model region. The data are also written out by land-use type, the note there any reported changes over time reflect changes in the land area of the given land use type. For example, a scenario with an expansion of crop land into forest will report net emissions at the region or basin level, but will report net sequestration by the crop land land use types. This is just because the total stock of carbon in cropland is increasing as the land area increases.

from gcam-core.

Thank you for your kind and detailed reply. It really helps me.

from gcam-core.

The simple carbon cycle model within GCAM is documented at https://jgcri.github.io/gcam-doc/land.html Note that this is different from the carbon cycle model that is part of Hector, documented at https://jgcri.github.io/gcam-doc/hector.html, which is only resolved at the global level, in which is not used for the reported terrestrial emissions or sequestration from the land system in a GCAM model run. Despite the name "simple," this is one of the most difficult parts of the model for a user to modify in such a way as to achieve some desired result. The net reported emissions or sequestration, for any region and model time period, is dependent on all of the land use decisions that have been made up until that time period, considering the average vegetation and soil carbon densities of each modeled land-use type. Because the average carbon densities of each land-use type are assumed time-invariant (e.g., the maximum vegetation carbon density of a tropical rainforest is 120 tC/ha, or 12 kgC/m2), a user can't just increase assumed carbon densities going into the future in order to represent enhanced uptake. Just as a final note here, the reported net emissions in sequestration make the most sense when reported by region, or alternatively by basin within a model region. The data are also written out by land-use type, the note there any reported changes over time reflect changes in the land area of the given land use type. For example, a scenario with an expansion of crop land into forest will report net emissions at the region or basin level, but will report net sequestration by the crop land land use types. This is just because the total stock of carbon in cropland is increasing as the land area increases.

Based on my understanding of the GCAM and Hector documentation, the land_use_emission in GCAM corresponds to Hector's land_use_emission-land_use_uptake. This implies that the (CO2_emission_by_region+LUC emissions by region) in GCAM does not include NEP (Net Ecosystem Production) = NPP (Net Primary Production) - RH (Respiration of Heterotrophs). Therefore, to achieve changes in carbon sinks, it is necessary to modify NBP_constrain (NBP = NPP - RH - land_use_emission + land_use_uptake) in the Hector model.

If there are any inaccuracies, pls let me know!

from gcam-core.