This repository contains the code used for the analyses in Understanding the reasons for over-crediting in REDD+ (Swinfield et al). The code will generate all of the figures and statistical outputs included in both the main text and the supplementary information.
This code is structured as separate R scripts relating to each figure or supplement. For scripts fig3, and fig4, compound rates extracted from the PACT parquet files will be written as .csvs to be read by fig.5.
The data can be downloaded at the following DOI: 10.5281/zenodo.14895067 Different data types are organised into folders relating to each file format. These folders should be placed in the same directory as the scripts.
Used in fig2, s3, and s8:
- ag_2022_avoided_amounts
- ag_2024_avoided_amounts
- tw_2020_avoided_amounts
- tw_2023_avoided_amounts
- tw_2024_avoided_amounts
- pact_2025_avoided_amounts
- certified_avoided_amounts
- project_metadata
Used in fig3, s4:
- evaluation_end_years
- certified_project_amounts
Used in fig 4, s7:
- evaluation_end_years
Used in fig5:
- acc_pact_control_rates
- acc_pact_project_rates
- certified_control_rates
- certified_project_rates
- acc_certified_control_rates
Used in fig3, s4:
- acc_project_area_parquets (folder of parquets)
Used in fig4, s7:
- acc_certified_control_parquets (folder of parquets)
- acc_pact_matching_parquets (folder of parquets)
Used in fig3, s4:
- project_area_geojsons (folder of geojsons)
Used in fig4:
- certified_control_area_geojsons (folder of geojsons)
- project_area_geojsons (folder of geojsons)
These scripts require R and several dependent libraries specified at the start of each script. It is important to note that fig4.R and s7.R may take a while to run due to calculating compound deforestation rates across multiple parquets.