DeepMind and partners release 10-meter map of the world's forest types

Why tree cover alone falls short

From space, an old-growth rainforest, a eucalyptus plantation and an oil palm estate can look like the same patch of green, yet they hold very different amounts of carbon and biodiversity. According to the researchers, a primary forest that has never been cleared holds far more carbon and biodiversity than a single-species timber plantation harvested every decade or two. Most global maps only show whether trees are present.

That gap matters for regulation. The EU regulation on deforestation-free products (EUDR) requires that goods such as coffee, cocoa or palm oil do not come from land that was forest in 2020 and has since been converted to agriculture. Wood products such as furniture must also be free of forest degradation, meaning they cannot come from primary forests converted into plantation or planted forests, or from naturally regenerating forests converted into plantation forests. Companies and regulators therefore need to know what kind of forest they are dealing with, and what replaces it after a harvest.

ForTy builds on Natural Forests of the World 2020, a 10-meter map released last year that separated natural forests from all other tree cover. That map answered whether a patch was natural forest in 2020 or not. ForTy adds the type of forest and, through the time series, its yearly change.

Six categories and how the model was trained

ForTy uses definitions from the UN Food and Agriculture Organization (FAO) and assigns land to one of six classes:

• Primary forest: native forest with no visible signs of human disturbance.
• Naturally regenerating forest: forest that grew back on its own, for example after logging or fire.
• Planted forest: forest that people planted or seeded and managed over long cycles.
• Plantation forest: fast-growing, intensively managed tree farms, often of one species.
• Tree crops and agroforestry: farms growing things other than wood, such as oil palm, rubber, cocoa or fruit.
• Other land: everything else, from grassland and cropland to cities and water.

The categories depend partly on land use, not only appearance, which made training difficult. The team built a collection of forest-type examples with about 1.7 million labeled locations, drawn from expert surveys, national and regional datasets, and targeted reviews of places where the model struggled. The model uses a custom architecture for remote sensing data and takes AlphaEarth Foundations embeddings as a key input. These are compact summaries of years of satellite observations for each 10-meter pixel, which help the model detect planting patterns, growth cycles and harvest rhythms.

Accuracy and adjustable thresholds

With IIASA and WRI, the team validated ForTy against more than 9,000 randomly chosen sites worldwide. At each site, two trained experts studied high-resolution imagery and years of satellite records to decide what land type was present. According to Google DeepMind, ForTy picked the right forest type 90% of the time and separated forest from non-forest land 95% of the time, outperforming other global forest maps.

Instead of a yes-or-no label, ForTy gives a probability score for each class. A company assessing EUDR risk might set a strict threshold and accept only pixels the model is highly certain about, while a researcher estimating total forest extent could use the full range of probabilities.

A time series reveals hidden change

The early-release time series applies a similar approach to every year from 2017 to 2025. It shows where forests disappear and regrow, and what kind of change occurred, such as a natural forest turned into a timber plantation or abandoned land regenerating into forest. This is hard to catch with standard tree-cover maps, because a logged and replanted area soon looks green again.

Availability and what it means

The Forest Typology 2020 map, described in Scientific Data, and the time series are freely available on Google Earth Engine under an open license. The authors say they hope governments, companies, conservation groups and scientists will use them to build deforestation-free supply chains, direct forest finance where it is needed and protect the natural forests that matter most. They plan to keep improving the time series with partners and users over the coming months.

ForTy is part of Google's AI for Nature research and the latest addition to Google Earth AI, its collection of geospatial models and datasets.