Tropical Moist Forest Accreditation Methodology Implementation

This repository contains an implementation of the PACT Tropical Moist Forest Accreditation Methodology.

From version 2.0.0 of the methodolgy, the git tags are synchronised with the corresponding versions of the methodology document (not necessarily the version number on the open-engage platform).

Development

This project relies on Python 3.10 or newer.

The easiest way to get setup is to use virtualenv then install the requirements.txt.

virtualenv venv
source venv/bin/activate
pip install -r requirements.txt

To typecheck the code run make type and to run the tests use make test.

System Requirements

Some stages of the pipeline require significant parallelisation to complete quickly which tends to also require plenty of memory too.

Structure

The code is broken into three main sections contained in the methods directory.

1. Inputs: the scripts in the inputs directory are used to download the necessary input data to run the methodology. For example, the land use class data from the JRC TMF dataset. Some of these scripts also do some light processing too, for example, to generate slopes from elevations maps.
2. Matching: the scripts in the matching directory are for performing the counterfactual pixel matching algorithms described in the methodology document. Broadly speaking, this first find potential candidates (referred to as the set M of pixels) and then finds actual pairs from pixels in the treatment area to pixels in the control area.
3. Outputs: the scripts in the outputs directory are used to generate the outputs from the methodology like the equivalent permanence and the additionality per year.

Additionally, there is a wrapper script called tmfpython.sh in the scripts/ folder. This contains all of the pipeline commands and uses the tmfpython3 magic wrapper.

Bugs

Should you find any bugs or issues with the code then please do open an issue on the github issue tracker.