DeepMind’s ForTy maps global forest types at 10-meter detail
Google DeepMind and Google Research, with WRI and IIASA, released Forest Typology (ForTy) 2020: a global 10-meter map that sorts land into six forest and land-use classes, plus an early yearly change series from 2017 to 2025—so users can see not only tree loss but what replaced natural forest.
DeepMind’s post argues that “tree cover” alone is not enough: old-growth rainforest, eucalyptus plantations, and oil-palm estates can look similarly green from space while differing sharply in carbon and biodiversity. Google DeepMind and Google Research, with the World Resources Institute (WRI) and IIASA, are releasing Forest Typology (ForTy) 2020—a global 10-by-10-meter map of six forest and land-use types framed as a key baseline for the EU deforestation regulation—alongside an early time series of yearly change from 2017 to 2025, including where natural forests were replaced by plantations or farms.
The six classes follow FAO definitions: primary forest, naturally regenerating forest, planted forest, plantation forest, tree crops/agroforestry, and other land. ForTy goes beyond last year’s Natural Forests of the World 2020 binary map by naming the forest type and what replaces it. Training used about 1.7 million labeled locations from expert surveys, national and regional datasets, and targeted reviews where the model struggled.
Delivery: DeepMind points readers to an Earth Engine app for 10-meter exploration. This article does not invent licensing or full API terms. The publisher date is October 22, 2026 (still Oct 22 in Asia/Shanghai); it is included as important official research tooling not yet covered on the site.
Sources
- Beyond tree cover: mapping the world’s forest types — Google DeepMind