The natural vegetation (natveg) land module is one of the land modules in MAgPIE (see also the other land modules: 30_crop, 31_past, 32_forestry, 34_urban). It calculates land and carbon stocks, as well as the biodiversity value, of natural vegetation, which consists of primary forest, secondary forest and other natural land. The module determines the availability of natural vegetation for land conversion.
| Description | Unit | A | |
|---|---|---|---|
| fm_bii_coeff (bii_class44, potnatveg) |
bii coeff | \(unitless\) | x |
| fm_carbon_density (t_all, j, land, c_pools) |
LPJmL carbon density for land and carbon pools | \(tC/ha\) | x |
| fm_luh2_side_layers (j, luh2_side_layers10) |
luh2 side layers | \(grid cell share\) | x |
| im_plantedclass_ac (j, ac) |
Raw Distribution of ageclass in secondary forest as a proxy for planted forest | \(10^6 ha\) | x |
| pcm_land (j, land) |
Land area in previous time step | \(10^6 ha\) | x |
| pm_carbon_density_ac (t_all, j, ac, ag_pools) |
Above ground natveg carbon density for age classes and carbon pools | \(tC/ha\) | x |
| pm_land_start (j, land) |
Land initialization area | \(10^6 ha\) | x |
| pm_timber_yield (t, j, ac, forest_land) |
Forest growing stock | \(tDM/ha/yr\) | x |
| pm_timber_yield_initial (j, ac, forest_land) |
Initial Forest yield | \(tDM/ha/yr\) | x |
| sm_fix_SSP2 | year until which all parameters are fixed to SSP2 values | \(year\) | x |
| sm_wood_density | Representative wood density based on IPCC | \(tDM/m3\) | x |
| vm_bv (j, landcover44, potnatveg) |
biodiversity value for all land cover classes (unweighted) | \(Mha\) | x |
| vm_carbon_stock (j, land, c_pools) |
Carbon stock in vegetation soil and litter for different land types | \(10^6 tC\) | x |
| vm_land (j, land) |
Land area of the different land types | \(10^6 ha\) | x |
| Description | Unit | |
|---|---|---|
| vm_cost_hvarea_natveg (i) |
Cost of harvesting natural vegetation | \(10^6 USD\) |
| vm_landdiff_natveg | Aggregated difference in natveg land compared to previous timestep | \(10^6 ha\) |
| vm_prod_natveg (j, land_natveg, kforestry) |
Production of woody biomass from natural vegetation | \(10^6 tDM/yr\) |
In the dynamic_may18 realization, land and carbon stock dynamics of natural vegetation are modeled endogenously. The initial spatial distribution of the sub-land-types primary forest, secondary forest and other natural land is based on the LUH2 data set (Hurtt et al. 2018). The realization includes 2 different kinds of land protection. a) Land protection based on the World Database on Protected Areas (WDPA) maintained by the International Union for Conservation of Nature (IUCN) and b) land protection based on national policies implemented (NPI) and nationally determined contributions to the Paris agreement (NDC) taken from individual country reports. Land protection based on WDPA is static over time, while the NPI/NDC polices ramp up until 2030 and are assumed constant thereafter. For WDPA the level of land protection is based on IUCN catI+II, which reflect areas currently under protection (e.g. strict nature reserves and national parks), and distributed equally across all sub-land-types (primary forest, secondary forest and other natural land). On top of the IUCN catI+II land protection, there are options to protect different conservation priority areas such as biodiversity hotspots (BH), centers of plant diversity (CBD), frontier forests (FF) and last of the wild (LW). NPI/NDC land protection polices are be applied on forest and other land, depending on individual country reports. Additionally, this module provides the ability to harvest natural vegetation for timber. Both wood and woodfuel can be produceed from primary and secondary forest but other land is only allowed to be harvested for woodfuel.
The interface vm_land provides aggregated natveg land pools (ac) to other modules.
\[\begin{multline*} vm\_land(j2,"secdforest") = \sum_{ac} v35\_secdforest(j2,ac) \end{multline*}\]
\[\begin{multline*} vm\_land(j2,"other") = \sum_{ac} v35\_other(j2,ac) \end{multline*}\]
Carbon stocks for primary forest, secondary forest or other natural land are calculated as the product of respective area and carbon density. Carbon stocks decline if the area decreases (e.g. due to cropland expansion into forests). In case of abandoned agricultural land (increase of other natural land), natural succession, represented by age-class growth, results in increasing carbon stocks.
\[\begin{multline*} vm\_carbon\_stock(j2,"primforest",ag\_pools) = vm\_land(j2,"primforest") \cdot \sum_{ct} fm\_carbon\_density(ct,j2,"primforest",ag\_pools) \end{multline*}\]
\[\begin{multline*} vm\_carbon\_stock(j2,"secdforest",ag\_pools) = \sum_{ac}\left( v35\_secdforest(j2,ac) \cdot \sum\left(ct, pm\_carbon\_density\_ac(ct,j2,ac,ag\_pools)\right)\right) \end{multline*}\]
\[\begin{multline*} vm\_carbon\_stock(j2,"other",ag\_pools) = \sum_{ac}\left( v35\_other(j2,ac) \cdot \sum\left(ct, pm\_carbon\_density\_ac(ct,j2,ac,ag\_pools)\right)\right) \end{multline*}\]
The biodiversity value (BV) of primary forest, secondary forest and other land is computed by multiplying their respective land area with bii coefficients, which depend on the age class and whether the potential natural vegetation forest or non-forest (luh2 side layers).
\[\begin{multline*} vm\_bv(j2,"primforest",potnatveg) = vm\_land(j2,"primforest") \cdot fm\_bii\_coeff("primary",potnatveg) \cdot fm\_luh2\_side\_layers(j2,potnatveg) \end{multline*}\]
\[\begin{multline*} vm\_bv(j2,"secdforest",potnatveg) = \sum_{bii\_class\_secd}\left( \sum\left(ac\_to\_bii\_class\_secd(ac,bii\_class\_secd), v35\_secdforest(j2,ac)\right) \cdot fm\_bii\_coeff(bii\_class\_secd,potnatveg)\right) \cdot fm\_luh2\_side\_layers(j2,potnatveg) \end{multline*}\]
\[\begin{multline*} vm\_bv(j2,"other",potnatveg) = \sum_{bii\_class\_secd}\left( \sum\left(ac\_to\_bii\_class\_secd(ac,bii\_class\_secd), v35\_other(j2,ac)\right) \cdot fm\_bii\_coeff(bii\_class\_secd,potnatveg)\right) \cdot fm\_luh2\_side\_layers(j2,potnatveg) \end{multline*}\]
NPI/NDC land protection policies are implemented as minium forest land and other land stock.
\[\begin{multline*} vm\_land(j2,"primforest") + vm\_land(j2,"secdforest") \geq \sum_{ct} p35\_min\_forest(ct,j2) \end{multline*}\]
\[\begin{multline*} vm\_land(j2,"other") \geq \sum_{ct} p35\_min\_other(ct,j2) \end{multline*}\]
The following technical calculations are needed for reducing differences in land-use patterns between time steps. The gross change in natural vegetation is calculated based on land expansion and land contraction of other land, and land reduction of primary and secondary forest. This information is then passed to the land module (10_land):
\[\begin{multline*} vm\_landdiff\_natveg = \sum_{j2,ac}\left( v35\_other\_expansion(j2,ac) + v35\_other\_reduction(j2,ac) + v35\_secdforest\_expansion(j2,ac) + v35\_secdforest\_reduction(j2,ac) + v35\_primforest\_reduction(j2)\right) \end{multline*}\]
\[\begin{multline*} v35\_other\_expansion(j2,ac\_est) = v35\_other(j2,ac\_est) - pc35\_other(j2,ac\_est) \end{multline*}\]
\[\begin{multline*} v35\_other\_reduction(j2,ac\_sub) = pc35\_other(j2,ac\_sub) - v35\_other(j2,ac\_sub) \end{multline*}\]
\[\begin{multline*} v35\_secdforest\_expansion(j2,ac\_est) = v35\_secdforest(j2,ac\_est) - pc35\_secdforest(j2,ac\_est) \end{multline*}\]
\[\begin{multline*} v35\_secdforest\_reduction(j2,ac\_sub) = pc35\_secdforest(j2,ac\_sub) - v35\_secdforest(j2,ac\_sub) \end{multline*}\]
\[\begin{multline*} v35\_primforest\_reduction(j2) = pcm\_land(j2,"primforest") - vm\_land(j2,"primforest") \end{multline*}\]
For natural forest, per-hectare harvesting costs are positive to make sure that older forest with higher growing stock is preferred over younger forest. To mimic the difficulties in accessing primary forest, per-hectare harvesting costs for primary forest are higher than for secondary forest. Harvesting costs are paid everytime natural vegetation is harvested. The “real” harvested area are received from the timber module 73_timber.
\[\begin{multline*} vm\_cost\_hvarea\_natveg(i2) = \sum_{ct,cell(i2,j2),ac\_sub} v35\_hvarea\_secdforest(j2,ac\_sub) \cdot s35\_timber\_harvest\_cost\_secdforest + \sum_{ct,cell(i2,j2),ac\_sub}\left( v35\_hvarea\_other\left(j2, ac\_sub\right)\right) \cdot s35\_timber\_harvest\_cost\_other + \sum_{ct,cell(i2,j2)} v35\_hvarea\_primforest(j2) \cdot s35\_timber\_harvest\_cost\_primforest \end{multline*}\]
Woody biomass production from secondary forests is calculated by multiplying the area under production with corresponding yields of secondary forests, divided by the timestep length.
\[\begin{multline*} \sum_{kforestry} vm\_prod\_natveg(j2,"secdforest",kforestry) = \frac{ \sum_{ac\_sub}\left( v35\_hvarea\_secdforest(j2,ac\_sub) \cdot \sum\left(ct,pm\_timber\_yield(ct,j2,ac\_sub,"secdforest")\right)\right) }{ m\_timestep\_length\_forestry} \end{multline*}\]
Woody biomass production from primary forests is calculated by multiplying the area under production with corresponding yields of primary forests, divided by the timestep length.
\[\begin{multline*} \sum_{kforestry} vm\_prod\_natveg(j2,"primforest",kforestry) = v35\_hvarea\_primforest(j2) \cdot \frac{ \sum_{ct} pm\_timber\_yield(ct,j2,"acx","primforest") }{ m\_timestep\_length\_forestry} \end{multline*}\]
Wood-fuel production from other land is calculated by multiplying the area under production with corresponding yields of other land, divided by the timestep length. Wood production from other landis not allowed.
\[\begin{multline*} vm\_prod\_natveg(j2,"other","woodfuel") = \frac{ \sum_{ac\_sub}\left( v35\_hvarea\_other(j2,ac\_sub) \cdot \sum\left(ct, pm\_timber\_yield(ct,j2,ac\_sub,"other")\right)\right) }{ m\_timestep\_length\_forestry } \end{multline*}\]
Following equations show the harvested area from natural vegetation i.e. primary forests, secondary forests and other land. Important to note here that no wood production should be realized from other land. Harvested area for production purposes can be lower oe equal than land reduction in natural vegetation as not all lost area is (or should be) used for production.
\[\begin{multline*} v35\_hvarea\_secdforest(j2,ac\_sub) \leq v35\_secdforest\_reduction(j2,ac\_sub) \end{multline*}\]
\[\begin{multline*} v35\_hvarea\_primforest(j2) \leq v35\_primforest\_reduction(j2) \end{multline*}\]
\[\begin{multline*} v35\_hvarea\_other(j2,ac\_sub) \leq v35\_other\_reduction(j2,ac\_sub) \end{multline*}\]
Harvested secondary forest is still considered secondary forests due to restrictive NPI definitions. Also primary forest harvested will be considered to be secondary forest i.e., harvested primary forest gets reclassified as secondary forest and ends up in the youngest age-class (and follows regrowth)
\[\begin{multline*} \sum_{ac\_est} v35\_secdforest(j2,ac\_est) = \sum_{ac\_sub}v35\_hvarea\_secdforest(j2,ac\_sub) + v35\_hvarea\_primforest(j2) \end{multline*}\]
Harvested other land is still considered other land
\[\begin{multline*} \sum_{ac\_est} v35\_other(j2,ac\_est) \geq \sum_{ac\_sub}v35\_hvarea\_other(j2,ac\_sub) \end{multline*}\]
The following two constraints distribute additions to secdforest and other land over ac_est, which depends on the time step length (e.g. ac0 and ac5 for a 10 year time step).
\[\begin{multline*} v35\_secdforest(j2,ac\_est) = \frac{ \sum_{ac\_est2} v35\_secdforest(j2,ac\_est2)}{card(ac\_est2)} \end{multline*}\]
\[\begin{multline*} v35\_other(j2,ac\_est) = \frac{ \sum_{ac\_est2} v35\_other(j2,ac\_est2)}{card(ac\_est2)} \end{multline*}\]
If the vegetation carbon density in a simulation unit due to regrowth exceeds a threshold of 20 tC/ha the respective area is shifted from other natural land to secondary forest.
p35_recovered_forest(t,j,ac)$(not sameas(ac,"acx")) =
p35_other(t,j,ac)$(pm_carbon_density_ac(t,j,ac,"vegc") > 20);
p35_other(t,j,ac) = p35_other(t,j,ac) - p35_recovered_forest(t,j,ac);
p35_secdforest(t,j,ac) =
p35_secdforest(t,j,ac) + p35_recovered_forest(t,j,ac);Limitations Initialization of both primary and secondary forest in highest age class or equal distrivution of such areas in all age classes. Data exists on a more emperically obtained distribution in different age classes based on satellite data but this results is highly negative land-use change emissions. Inclusion of this data in MAgPIE remains work in progess and is not available for release yet. Additionally, in this module realizatuion, harvested secondary forest stays secondary forest and harvested primary forest is reclassified as secondary forest.
| Description | Unit | A | |
|---|---|---|---|
| f35_forest_disturbance_share (i) |
Share of area damanged by forest disturbances | \(1\) | x |
| f35_forest_lost_share (i, driver_source) |
Share of area damanged by forest fires | \(1\) | x |
| f35_gs_relativetarget (i) |
Relative growing stock target in each region | \(m3/ha\) | x |
| f35_land_iso (t_ini10, iso, land) |
Land area for different land pools at ISO level | \(10^6 ha\) | x |
| f35_min_land_stock (t_all, j, pol35, pol_stock35) |
Land protection policies [minimum land stock] | \(Mha\) | x |
| f35_protect_area (j, prot_type) |
Conservation priority areas | \(10^6 ha\) | x |
| f35_protection_fader (t_all, prot_target35) |
Protection scenario fader | \(1\) | x |
| i35_land_iso (iso) |
Total land area at ISO level | \(10^6 ha\) | x |
| i35_other (j, ac) |
Initial other land | \(10^6 ha\) | x |
| i35_plantedclass_ac (j, ac) |
Area of age-classes in secondary forest | \(1\) | x |
| i35_secdforest (j, ac) |
Initial secdforest | \(10^6 ha\) | x |
| p35_carbon_density_other (t, j, ac, ag_pools) |
Carbon density other land | \(tC/ha\) | x |
| p35_carbon_density_secdforest (t, j, ac, ag_pools) |
Carbon density secdforest | \(tC/ha\) | x |
| p35_country_dummy (iso) |
Dummy parameter indicating whether country is affected by selected land protection policy | \(1\) | x |
| p35_disturbance_loss_primf (t, j) |
Loss due to disturbances in primary forest | \(10^6 ha\) | x |
| p35_disturbance_loss_secdf (t, j, ac) |
Loss due to disturbances in secondary forest | \(10^6 ha\) | x |
| p35_gs_scaling_reg (i) |
Regional calibration factors for natural vegetation yields | \(1\) | x |
| p35_land (t, j, land_natveg, ac) |
Natural vegetation area | \(10^6 ha\) | x |
| p35_land_start_ac (j, ac, land_natveg) |
Initial Natural vegetation area | \(10^6 ha\) | x |
| p35_min_forest (t, j) |
Minimum forest stock [land protection policies] | \(Mha\) | x |
| p35_min_other (t, j) |
Minimum other land stock [land protection policies] | \(Mha\) | x |
| p35_observed_gs_reg (i) |
Observed growing stock in natural forests before calibration | \(m3/ha\) | x |
| p35_other (t, j, ac) |
Other land per age class | \(10^6 ha\) | x |
| p35_poulter_dist (j, ac) |
Share of age-classes in secondary forest | \(1\) | x |
| p35_protect_shr (t, j, prot_type_all, land_natveg) |
Land protection share for primforest secdforest and other land | \(1\) | x |
| p35_protect_shr_ini (j, prot_type_all) |
Land protection share for primforest secdforest and other land | \(1\) | x |
| p35_recovered_forest (t, j, ac) |
Recovered forest | \(10^6 ha\) | x |
| p35_region_prot_shr (i) |
Land protection share of the region | \(1\) | x |
| p35_save_dist (j, ac) |
Distribution of secondary forest saving | \(1\) | x |
| p35_save_natveg (t, j, land_natveg) |
Land protection for primforest secdforest and other land | \(10^6 ha\) | x |
| p35_secdforest (t, j, ac) |
Secdforest per age class | \(10^6 ha\) | x |
| p35_updated_gs_natfor (t, i) |
Updated growing stock in natural forests after calibration | \(m3/ha\) | x |
| pc35_other (j, ac) |
Other land per age class in current time step | \(10^6 ha\) | x |
| pc35_secdforest (j, ac) |
Secdforest per age class in current time step | \(10^6 ha\) | x |
| q35_bv_other (j, potnatveg) |
Biodiversity value of other land | \(Mha\) | x |
| q35_bv_primforest (j, potnatveg) |
Biodiversity value of primary forest | \(Mha\) | x |
| q35_bv_secdforest (j, potnatveg) |
Biodiversity value of secondary forest | \(Mha\) | x |
| q35_carbon_other (j, ag_pools) |
Other land carbon stock calculation | \(10^6 tC\) | x |
| q35_carbon_primforest (j, ag_pools) |
Primforest carbon stock calculation | \(10^6 tC\) | x |
| q35_carbon_secdforest (j, ag_pools) |
Secdforest carbon stock calculation | \(10^6 tC\) | x |
| q35_cost_hvarea (i) |
Cost of harvesting natural vegetation | \(10^6 USD\) | x |
| q35_hvarea_other (j, ac) |
Harvested area other land | \(10^6 ha\) | x |
| q35_hvarea_primforest (j) |
Harvested area primforest | \(10^6 ha\) | x |
| q35_hvarea_secdforest (j, ac) |
Harvested area secdforest | \(10^6 ha\) | x |
| q35_land_other (j) |
Other land pool calculation | \(10^6 ha\) | x |
| q35_land_secdforest (j) |
Secdforest land pool calculation | \(10^6 ha\) | x |
| q35_landdiff | Difference in natveg land | \(10^6 ha\) | x |
| q35_min_forest (j) |
Minimum forest land constraint | \(10^6 ha\) | x |
| q35_min_other (j) |
Minimum other land constraint | \(10^6 ha\) | x |
| q35_other_conversion (j) |
Other land remains secondary forest | \(10^6 ha\) | x |
| q35_other_est (j, ac) |
Distribution of other land additions over ac_est | \(10^6 ha\) | x |
| q35_other_expansion (j, ac) |
Other land expansion | \(10^6 ha\) | x |
| q35_other_reduction (j, ac) |
Other land reduction | \(10^6 ha\) | x |
| q35_primforest_reduction (j) |
Primforest reduction | \(10^6 ha\) | x |
| q35_prod_other (j) |
Production of woody biomass from other land | \(10^6 tDM/yr\) | x |
| q35_prod_primforest (j) |
Production of woody biomass from primary forests | \(10^6 tDM/yr\) | x |
| q35_prod_secdforest (j) |
Production of woody biomass from secondary forests | \(10^6 tDM/yr\) | x |
| q35_secdforest_conversion (j) |
Secondary forest remains secondary forest | \(10^6 ha\) | x |
| q35_secdforest_est (j, ac) |
Distribution of secdforest additions over ac_est | \(10^6 ha\) | x |
| q35_secdforest_expansion (j, ac) |
Secdforest reduction | \(10^6 ha\) | x |
| q35_secdforest_reduction (j, ac) |
Secdforest reduction | \(10^6 ha\) | x |
| s35_forest_damage | Damage simulation in forests | \(0=none 1=shifting agriculture 2= Damage from shifting agriculture is faded out by c35_forest_damage_end\) | x |
| s35_hvarea | Flag for harvested area | \(0=zero 1=exognous 2=endogneous\) | x |
| s35_hvarea_other | annual other land harvest rate for s35_hvarea equals 1 | \(percent/year\) | x |
| s35_hvarea_primforest | annual primforest harvest rate for s35_hvarea equals 1 | \(percent/year\) | x |
| s35_hvarea_secdforest | annual secdforest harvest rate for s35_hvarea equals 1 | \(percent/year\) | x |
| s35_natveg_harvest_shr | Constrains the allowed wood harvest from natural vegetation (1=unconstrained) | \(1\) | x |
| s35_secdf_distribution | Flag for secdf initialization (0=all secondary forest in highest age class 1=Equal distribution among all age classes 2=Poulter distribution from MODIS satellite data) | \(1\) | x |
| s35_shift | number of 5-year age-classes corresponding to current time step length | \(1\) | x |
| s35_timber_harvest_cost_other | Cost for harvesting from other land | \(USD/ha\) | x |
| s35_timber_harvest_cost_primforest | Cost for harvesting from primary forest | \(USD/ha\) | x |
| s35_timber_harvest_cost_secdforest | Cost for harvesting from secondary forest | \(USD/ha\) | x |
| v35_hvarea_other (j, ac) |
Harvested area from other land | \(10^6 ha\) | x |
| v35_hvarea_primforest (j) |
Harvested area from primary forest | \(10^6 ha\) | x |
| v35_hvarea_secdforest (j, ac) |
Harvested area from secondary forest | \(10^6 ha\) | x |
| v35_other (j, ac) |
Detailed stock of other land | \(10^6 ha\) | x |
| v35_other_expansion (j, ac) |
Other land expansion compared to previous timestep | \(10^6 ha\) | x |
| v35_other_reduction (j, ac) |
Other land reduction compared to previous timestep | \(10^6 ha\) | x |
| v35_primforest_reduction (j) |
Primforest reduction compared to previous timestep | \(10^6 ha\) | x |
| v35_secdforest (j, ac) |
Detailed stock of secdforest | \(10^6 ha\) | x |
| v35_secdforest_expansion (j, ac) |
Secdforest reduction compared to previous timestep | \(10^6 ha\) | x |
| v35_secdforest_reduction (j, ac) |
Secdforest reduction compared to previous timestep | \(10^6 ha\) | x |
| description | |
|---|---|
| ac | Age classes |
| ac_est(ac) | Dynamic subset of age classes for establishment |
| ac_planted(ac) | Young age classes which probably belong to managed forest |
| ac_sub(ac) | Dynamic subset of age classes excluding establishment |
| ac_to_bii_class_secd(ac, bii_class_secd) | Mapping between forest ageclasses and bii coefficent land cover classes |
| ag_pools(c_pools) | Above ground carbon pools |
| age | Population age groups |
| bii_class_secd(bii_class44) | bii coefficent land cover classes secondary vegetation |
| bii_class44 | bii coefficent land cover classes |
| c_pools | Carbon pools |
| cell(i, j) | number of LPJ cells per region i |
| combined_loss(driver_source) | Combined loss from fire plus agriculture |
| ct(t) | Current time period |
| driver_source | Source of deforestation drivers |
| forest_land(land) | land from which timber can be taken away |
| i | all economic regions |
| i_to_iso(i, iso) | mapping regions to iso countries |
| i2(i) | World regions (dynamic set) |
| iso | list of iso countries |
| j | number of LPJ cells |
| j2(j) | Spatial Clusters (dynamic set) |
| kforestry(k) | forestry products |
| land | Land pools |
| land_natveg(forest_land) | Natural vegetation land pools |
| landcover44 | land cover classes used in bii calculation |
| luh2_side_layers10 | side layers from LUH2 |
| pol_stock35 | Land types for land protection policies |
| pol35 | Land protection policy |
| policy_countries35(iso) | countries to be affected by land protection policy |
| potnatveg(luh2_side_layers10) | potentially forested biomes |
| prot_target35 | Target year for nature protection |
| prot_type_all | Conservation priority areas |
| prot_type(prot_type_all) | Conservation priority areas |
| t_all(t_ext) | 5-year time periods |
| t_ini10 | Time periods with land initialization data |
| t_past(t_all) | Timesteps with observed data |
| t(t_all) | Simulated time periods |
| Target | year for set aside policy |
| type | GAMS variable attribute used for the output |
Florian Humpenöder, Abhijeet Mishra
09_drivers, 10_land, 11_costs, 14_yields, 29_ageclass, 44_biodiversity, 52_carbon, 73_timber
Hurtt, George C, Louise P Chini, Ritvik Sahajpal, Steve E Frolking, Benjamin Bodirsky, Katherine V Calvin, Jonathan C Doelman, et al. 2018. “LUH2: Harmonization of Global Land-Use Scenarios for the Period 850-2100.” AGUFM 2018: GC13A–01.