Tree planting decision-making under uncertainty in GB

Replication code for paper on tree planting uncertainty under the ADVANCES project intended for public access.

Inputs (stored in /data)

Will need to contact author for a copy of the directory because this is over 10GB

1. Binary files: full matrices of output from NEV suite Monte Carlo simulations from the distribution of uncertain parameters, for each climate-economy realisation (CER)
2. a1_landuse_chg.mat: Matlab struct describing hectares of landuse change under a tree planting scenario
3. high_level_lcs.mat: landuse cover in GB
4. param_table.csv: a table describing parameter specifications for each of the CERs

Outputs

• Tables showing optimal parcels of tree planting (decision) and NPV outcomes (returns) across climate-economy realisations for reaching a 12MtCO2e carbon sequestration target excluding displaced agricultural emissions, ran under the following settings
  • Assuming each of the three "focus" climate-economy realisations (P-NH, P-ME, and P-HE), leading to 3 planting strategies with 4,000 possible NPV outcomes
  • Assuming each of all of the climate-economy realisations (4,000), leading to 4,000 planting strategies
  • Optimising the "average" NPV outcome (EV) and optimising a risk measure of NPV outcomes (conditional value-at-risk) for only using tree-planting to reach 12MtCO2e
  • Optimising EV and CVaR in the presence of a "risk-free" carbon removal target, described by a "net cost" (difference between GGR cost and social cost of carbon)
• Figures (relying on the tables above)
  • Figure 1: Line graph showing changes in climate and economic variables for the three "focus" CERs, and bump chart showing the change in natural capital value for planting under the three CERs, and the realisation of each respective CER
  • Figure 2: Shows (1) the distribution of outcomes across 4,000 CERs for planting under NH, ME and HE realisations, (2) distribution of outcomes (in jitter plots) for maximising EV, (3) distribution of outcomes (in jitter plots) for minimising CVaR, (4) planting maps showing location of planting for P-EV and P-RA, and (5) correlation matrix
  • Figure 3: Shows change in (1) area planted, (2) annual carbon removal distribution and (3) natural capital value distribution for increasing levels of net costs (GGR cost - carbon price) for an alternative risk-free carbon removal technology

Steps in analysis, with portions replicable in this repo in asterisks

Because replication of the entire analysis requires full access to NEV where sensitive data are used as inputs, portions that depend on NEV will not be publicly replicable but private access could be granted upon reasonable request.

1. Drawing of parameters in climate-economy scenarios from statistical distributions
2. Running the NEV forestry and agriculture models for each of the climate-economy realisation
3. Optimising land-use decisions *
4. Constructing figures and diagrams in the paper *