Key finding
The study is the most precise global quantification of tree cover loss and its drivers ever conducted. Using 3 to 10 meter resolution satellite imagery from multiple commercial and public sources, the team analyzed the fate of approximately 277,000 square kilometers of tree cover loss globally in the year 2018. The results show that nearly one third (29.8%) of this loss was permanent, meaning the land was converted to another use with no expectation of forest regeneration.
The geographic distribution is striking. The tropics accounted for nearly two thirds of global tree cover loss. Latin America had the largest share at 26.6%, followed by South and Southeast Asia at 21.6%, and Africa at 18.2%. The top three countries for tree cover loss were Brazil, Russia, and Indonesia. Permanent conversion was driven primarily by cattle pasture expansion (15% of global loss), cropland expansion (6.4%), and oil palm and rubber plantations (3.8%). Mechanical clearing was responsible for 84% of global loss, while fire accounted for 12.4%.
A critical finding concerns how existing global maps misidentify drivers. Current maps at 1 to 10 kilometer resolution and 20-plus-year temporal scales confuse plantation harvesting cycles with permanent deforestation and mistake shifting cultivation for long-term agricultural conversion. The authors caution that when maps get the drivers of tree cover loss wrong, it can shift the focus of deforestation efforts to the wrong regions. The study also found that roughly 17% of shifting-agriculture-related loss in the tropics expanded into primary forests rather than previously disturbed areas, with hotspots in the Democratic Republic of Congo and Borneo.
What the research asks
Global tree cover loss figures, derived primarily from Landsat at 30-meter resolution, are widely used to track deforestation trends and inform climate policy. But these figures aggregate permanent forest conversion, plantation harvesting cycles, and shifting cultivation into a single number, making it impossible to distinguish reversible loss from permanent deforestation. This study asked: what is the precise global extent of permanent tree cover loss versus temporary loss, using 3 to 10 meter resolution satellite imagery from multiple commercial and public sources, and what are the direct land-use drivers of permanent conversion?
What it finds
The study is the most precise global quantification of tree cover loss and its drivers ever conducted. Using 3 to 10 meter resolution satellite imagery from multiple commercial and public sources, the team analyzed the fate of approximately 277,000 square kilometers of tree cover loss globally in the year 2018. The results show that nearly one third (29.8%) of this loss was permanent, meaning the land was converted to another use with no expectation of forest regeneration.
The geographic distribution is striking. The tropics accounted for nearly two thirds of global tree cover loss. Latin America had the largest share at 26.6%, followed by South and Southeast Asia at 21.6%, and Africa at 18.2%. The top three countries for tree cover loss were Brazil, Russia, and Indonesia. Permanent conversion was driven primarily by cattle pasture expansion (15% of global loss), cropland expansion (6.4%), and oil palm and rubber plantations (3.8%). Mechanical clearing was responsible for 84% of global loss, while fire accounted for 12.4%.
A critical finding concerns how existing global maps misidentify drivers. Current maps at 1 to 10 kilometer resolution and 20-plus-year temporal scales confuse plantation harvesting cycles with permanent deforestation and mistake shifting cultivation for long-term agricultural conversion. The authors caution that when maps get the drivers of tree cover loss wrong, it can shift the focus of deforestation efforts to the wrong regions. The study also found that roughly 17% of shifting-agriculture-related loss in the tropics expanded into primary forests rather than previously disturbed areas, with hotspots in the Democratic Republic of Congo and Borneo.
Why it matters
This is a landmark paper for anyone working on forest monitoring, REDD+, supply chain due diligence, or climate policy. The finding that nearly 30% of tree cover loss is permanent conversion makes clear that deforestation is not declining as fast as headline numbers suggest. The two-thirds that is temporary may recover, but the one-third that is permanent locks in carbon emissions and biodiversity loss.
The methodological advance is equally important. The use of 3 to 10 meter resolution imagery, including Planet and other commercial data, demonstrates that monitoring systems have moved beyond the era when Landsat's 30-meter resolution was the best available. For EUDR compliance systems, corporate deforestation commitments, and national MRV frameworks, this study provides the methodological template: combine high-resolution imagery with temporal analysis to distinguish permanent conversion from rotational harvesting or shifting cultivation.
The study's caution about driver misclassification is a direct call to action. If a monitoring system cannot tell the difference between a plantation harvest cycle and primary forest conversion, it cannot target interventions effectively. Earth observation systems used for compliance must therefore include driver attribution, not just change detection. The paper also sets a baseline that future assessments can build on, enabling trend analysis as the climate and global economy shift the drivers of forest loss over time.