In CABLE, when land-use change occurs, for example from forest to grass, the forest is completely cleared and the harvested wood is placed into wood product pools. However it’s possible to remove wood from a forest by selective logging that does not involve a complete transition to grass. We have an implementation of this in CABLE, which we call “thinning”. Despite this being developed before CMIP6, the testing of it was incomplete and the impending CMIP6 deadlines resulted in it being disabled in those simulations. Since then, it has been largely forgotten. It is however potentially very useful. Considering the fact that ACCESS’ land-use change flux is too low (and consequently NBP), additional wood thinning may help.
I’m currently in the process of trying to understand the LUH2 wood thinning inputs and the outputs of the model. The purpose of this post is to share the progress and information about wood thinning in CABLE as I work on it.
1. Implementation
1.1 Land-use change in CABLE coupled to ACCESS
CABLE within ACCESS does not distinguish between primary and secondary forests and land-use change only accounts for net transitions in land cover type. At the start of the year when land-use change occurs (forest to crops for example), the logged biomass is allocated evenly to three anthropogenic wood product pools. These pools then decay to the atmosphere at exponential decay rates of 1, 10 and 100 years^-1, representing paper, buildings and other long-lived wood products. The leaf and root biomass are reallocated to the metabolic and structural litter pools of the new vegetation type based on the C/N ratios. Wood from tundra and shrubs is allocated to the Coarse woody debris litter pool of the new vegetation type.
1.2 Thinning implementation
The biomass allocation for thinning is much the same as for land-use change above, however, it only occurs for forest vegetation types and leaf and root biomass are allocated to the litter pools of that forest type.
Thinning occurs once a year during the initialisation of CASA. The forest thinning (the removal of biomass from the plant pools) is done in the subroutine casa_reinit_pk in the module casa_um_inout_mod (file casa_um_inout.F90) immediately after land-use change occurs. There is also a subroutine newlitter_thin, located in the landuse_mod module (file casa_landuse.F90), which reallocates the leaf and root biomass to the litter pools.
The thinning ratio applied to the CNP pools uses the convention that a value of 1 means no thinning and 0 means all biomass is removed.
1.3 Procedure
- Calculate CNP pools after thinning.
- Add the removed wood to the wood_flux variable.
- Add the wood harvest from thinning to the 3 wood product pools.
- Redistribute the harvested plant mass to litter pools.
- Find the difference between the plant biomass pools before and after thinning.
- Ignore wood biomass (because it’s already transferred to wood harvest).
- Calculate the plant to litter ratio matrix based on C/N.
- Transfer biomass to litter pools.
The existing implementation of product pools is then used to calculate the resulting land-use change emissions. That is, the wood product pools respire at decay rates of 1, 10 and 100 years^-1.
Code changes are documented in the CABLE github issue and pull request.
2. Forcing data
The forcing data for thinning ratio are derived from the LUH2 transition variables:
primf_harv: Primary forest area harvest fractionsecmf_harv: Secondary mature forest area harvest fractionsecyf_harv: Secondary young forest area harvest fraction
These variables are added together and remapped to ACCESS resolution using CDO’s second order area conserving remapping. Missing values on land points from the remapping are filled with 0’s. The thinning fraction is converted to CABLE’s convention of 1 being no thinning by multiplying by -1 and adding 1. The thinning fractions are then broadcast to CABLE’s forest vegetation types by multiplying by the land cover fractions.
Scripts used to process the forcing data are here:
The forcing data are stored on gadi here: /g/data/p66/tfl561/thinning/harvest_frac
The forcing data are ingested through the ACCESS’s atmosphere restart files. The restart file stashvar variable that contains the thinning fields for a single year has code 916. ACCESS is typically resubmitted to the queue once a year. For each year, a new map of wood thinning data is injected in the restart dump before the start of the simulation.
At runtime the restart file is loaded and the thinning data is then passed through the UM to CABLE via arguments to subroutine calls as follows.
ATM_Step
Atmos_Physics2
Ni_bl_ctl
LB_INTCT
BDY_LAYR
SF_EXPL
SF_EXCH
cable_explicit_driver
interface_UM_data
init_casacnp
casa_init_pk
casa_reinit_pk
newlitter_thin
The wood flux (biomass removed from land-use change clearing and thinning) is orders of magnitude larger than the standard historical simulation (blue=thinning, orange=historical).
