Guyana has some of the most pristine old growth tropical rainforest in the world. Preserving natural landscapes in these high forest, low deforestation (HFLD) countries is critical for enhancing biodiversity, and mitigating climate change.

We built a measurement, reporting and verification for system in Guyana to assess deforestation at the national scale. We built a robust, scientifically sound program to map deforestation and disaggregate it according to driver such as fire, mining or agriculture. In doing so I built capacity within the local staff at the Guyana Forestry Commission, with the future goal of conducting and reporting on this mapping independent of international consultants. We initially built an ArcGIS desktop-based system that used freely available Landsat imagery. After the first few years we sought higher resolution imagery to detect forest degradation as well as deforestation. We purchased 5m resolution RapidEye imagery over Guyana, enabling us to map forest degradation directly.

The six GFC analysts were all taught the basics of geospatial and remote sensing science. Mapping the country wall-to-wall was labour intensive, requiring skill and concentration. I later published a study which looked at the utility of a sampling based approach, including the relevant statistical tradeoffs and cost of reference imagery.

An independent team of specialists from Durham University performed an accuracy assessment on our work. They chartered the acquisition of 0.25m aerial orthophotography as higher resolution reference imagery. Our maps were found to be ~99% accurate. The entire system, from mapping to accuracy assessment, was then subject to external audit by a team employed through the Norwegian funding agency and always received excellent results.

Given the advancements in GIS and remote sensing science, We can now build national scale deforestation mapping programs from scratch. We can also teach your team the necessary skills to conduct these sorts of assessments using entirely free software.