legacyEmissions

Reporting-side legacy-clearing tail for land-clearing CO2. MAgPIE books the aboveground carbon of clearing (deforestation + other-land conversion) instantly; bookkeeping models (BLUE, OSCAR, Houghton & Castanheira) spread it over a slash/deadwood decay tail. This re-spreads the pulse over a first-order-decay (FOD) pool, as carbonLTS does for harvested wood products (HWP), and returns the net correction to the raw flux, its Storage/Release parts, and the pool carbon stock.

Usage

legacyEmissions(
  gdx,
  file = NULL,
  level = "cell",
  unit = "element",
  cumulative = FALSE,
  baseyear = 1995,
  priming = "hist",
  primingStart = 1850,
  harvestSlashFrac = 0,
  a0Biome = c(A = 0.25, B = 0.25, C = 0.2, D = 0.15, E = 0.15),
  halfLifeBiome = c(A = 4, B = 8, C = 10, D = 20, E = 25)
)

Arguments

gdx

GDX file

file

file name to write the output to with write.magpie

level

aggregation level: "cell", "reg", "glo", "regglo", or any superAggregateX level

unit

"element" (Mt C/yr) or "gas" (Mt CO2/yr)

cumulative

report annually (FALSE) or cumulatively from baseyear (TRUE)

baseyear

baseyear for cumulative emissions (default 1995)

priming

assumed pre-firstYear clearing history that sets the pool's initial stock: "hist" (default, the bookkeeping-ensemble LUC-CO2 shape), "ramp" (linear 0.3->1.0), "half" (flat 0.5x) or "peak" (flat 1x). The only non-mass-conserving element - it credits pre-firstYear clearing MAgPIE never modelled (cf. carbonLTS's 1970 seed).

primingStart

first year of the priming window (default 1850; results insensitive below ~1950).

harvestSlashFrac

fraction of the wood-harvest aboveground flux added as dead slash (default 0, OFF); a coarse sensitivity switch that does NOT net out the HWP carbon carbonLTS already defers.

a0Biome

immediate/combustion fraction per Koeppen main group (A tropical, B arid, C temperate, D boreal, E polar). Default A 0.25 / B 0.25 / C 0.20 / D 0.15 / E 0.15, from Houghton et al. 2012's clearing-year burn partition (doi:10.5194/bg-9-5125-2012).

halfLifeBiome

slash/deadwood decay half-life (years) per Koeppen group. Default A 4 / B 8 / C 10 / D 20 / E 25, from coarse-woody-debris decay data (Chambers et al. 2000, Harmon et al. 2020, Russell et al. 2014).

Value

MAgPIE object (region x year x name) with "legacy_net" (delta to add to Land-use Change), "legacy_storage" (<= 0, deferred emission), "legacy_release" (>= 0, decay outflow) and "legacy_stock" (the slash-pool carbon level, incl. the primed initial stock). legacy_net = legacy_storage + legacy_release = -diff(legacy_stock) by construction. Mt C or CO2 (per unit); cumulative fluxes are rebased to baseyear as in carbonLTS (legacy_stock, a level, is returned only for cumulative = FALSE).

Details

The pulse is aboveground (vegc+litc) clearing CO2 only - deforestation + other-land conversion; soil carbon and degradation are excluded. It is interpolated from MAgPIE's 5/10-yr steps to annual temporal resolution, split by Koeppen group (fast decay in the tropics, slow in the boreal) with a separate pool per group, then summed; the fraction a0 is released immediately and the rest (1-a0) decays with the group half-life (see the code for the recursion). y1995 has no reported clearing flux, so the pulse starts at the first reported year, anchored to the mean of the first few.

Caveats: (1) reporting reframe only - the model still emits instantaneously and the carbon price is unchanged; raw emissions stay available via legacyEmis = FALSE. (2) not mass-conserving: the priming credit is a one-off initial pool; being a linear decay it is proportional to the priming amplitude, so half = exactly 0.5x peak and the default hist sits near peak. (3) a residual scope gap to bookkeeping remains. The cumulative legacy series is the timestep-weighted integral of the yearly one (both at the model timesteps); carbon still in the pool at 2100 never enters the reported window.

Author

Florian Humpenoeder

Examples

if (FALSE) { # \dontrun{
x <- legacyEmissions(gdx)
} # }