Google DeepMind and partners release 10-meter global forest typology map to track deforestation and forest conversion
A new high-resolution map categorizes every 10-by-10-meter patch of Earth’s land into six forest and land-use types, with annual time-series data from 2017 to 2025 to support EU deforestation regulation and conservation efforts.
1 source · cross-referenced
- A new global map, Forest Typology (ForTy) 2020, classifies every 10-by-10-meter patch of land into six forest and land-use types using UN FAO definitions.
Google DeepMind, Google Research, the World Resources Institute (WRI), and the International Institute for Applied Systems Analysis (IIASA) released Forest Typology (ForTy) 2020, a global map that categorizes every 10-by-10-meter patch of land on Earth into six distinct forest and land-use types. The map uses UN Food and Agriculture Organization (FAO) definitions to classify land into: primary forest, naturally regenerating forest, planted forest, plantation forest, tree crops and agroforestry, and other land.
Alongside the map, the partners released an early time series tracking annual changes in forest types from 2017 to 2025. This time series allows users to see not only where forest loss occurred, but what replaced it—such as timber plantations, tree-crop farms, or other land uses. The data is intended to support the European Union’s deforestation regulation (EUDR), which requires products like coffee, cocoa, or palm oil to be deforestation-free.
The initiative builds on a 2020 map released last year, Natural Forests of the World 2020, which distinguished natural forests from other tree cover. ForTy expands this by specifying the type of forest or land use, providing a more granular baseline for regulators and companies to implement deforestation-free commitments and track forest degradation over time.
- Oct 21, 2026 · arXiv cs.AI
Researchers propose LLM-and-RL system for autonomous construction planning and progress updates
Trust79 - Oct 21, 2026 · arXiv cs.AI
Multi-objective reinforcement learning method improves Pareto-optimal RTL circuit generation with LLMs
Trust79 - Oct 19, 2026 · arXiv cs.CL
Few prompt tokens drive most F1 loss when reducing experts in MoE prefill
Trust79