Google DeepMind and the World Resources Institute have launched a new tool named ForTy. According to reporting by The Decoder, the project uses satellite data and artificial intelligence to map and classify forests and other land types across the globe into six distinct categories. These classifications include primary forests, tree crops, and plantations.
Tracking Deforestation Under EU Regulations
The announcement indicates that this global mapping initiative could potentially assist in tracking deforestation to comply with European Union rules. Satellite observation combined with machine learning models allows researchers and regulatory bodies to monitor land use changes at a scale that manual surveying cannot achieve. However, the release notes highlight notable accuracy limitations in the current data.
Accuracy Limitations and Unknowns
The mapping tool currently identifies only about half of all planted forests and tree-crop areas, alongside roughly 60 percent of traditional plantations. The initial announcement did not specify how Google DeepMind and the World Resources Institute plan to improve these detection rates or when subsequent updates to the ForTy map might be released. Further details regarding the underlying satellite datasets and specific AI architectures used were also absent from the disclosure.
Observers will need to watch how regulatory bodies respond to these accuracy margins as EU compliance deadlines approach.
Lead image: Radarsky 1984 / Pexels
