- Most companies importing certain products into the EU must comply with the European Union’s Regulation on Deforestation-free products (EUDR), which will go into application on Dec. 30, 2026.
- Satellite and other remote-sensing maps can guide both companies trying to comply with the regulation and government agencies verifying levels of deforestation risk attached to imports.
- But a recent review paper suggests that most of the available maps struggle to meet all of the requirements of the EUDR and could over- or underestimate the risk of deforestation for certain products.
- A key issue is the maps’ ability to differentiate forest from systems that look similar, such as agroforestry, commonly practiced by smallholder farmers producing cocoa, coffee and rubber.
A recent scientific review of forest maps used to ensure compliance with the European Union’s Regulation on Deforestation-free products, known as the EUDR, suggests that most may over- or underestimate forest areas, which could lead to inaccurate assessments of deforestation risk.
The authors write that those inconsistencies point to the need for EU companies to be discerning about which maps they use to ensure they comply with the regulation. The requirements will go into effect for most companies on Dec. 30, 2026, after a second postponement in as many years by the European Parliament.
Only two of the 21 data sets in the assessment met all of the indicators used to assess risk used by the EUDR.
The regulation will apply to seven commodities — cattle, cocoa, coffee, palm oil, rubber, soy and timber — as well as the products they’re used to make. Companies and government agencies are planning to use maps of satellite and other remote-sensing data to determine whether products entering EU countries are linked to deforestation after Dec. 31, 2020, the regulation’s cutoff date.
Importing companies and the EU countries’ government agencies tasked with screening imported goods for compliance will compare georeferenced plots for a commodity with historical maps of forest and tree cover to determine whether it was produced at the expense of recently cleared forest.
But many such maps exist, and the EUDR doesn’t specify the use of any one map. That means that a company using one map to verify compliance might come to a different conclusion about deforestation risk than an EU government agency that uses another map, said study co-author Melvin Lippe, a land system scientist at the Thünen Institute of Forestry in Hamburg, Germany.
Part of the rationale for allowing choice is that it affords companies and authorities the flexibility to use the publicly available maps that suit their needs. Some might offer better coverage in the geographic locations they operate in, for example, while others might be less accurate.
Lippe and his colleagues set out to determine which data sets work best for EUDR compliance. The team probed 21 global maps showing forest or tree cover for their alignment with EUDR requirements. They paid particular attention to the time frames, spatial resolution and forest definitions used by each. The researchers also compared maps with the U.N. Food and Agriculture Organization’s Global Forest Resource Assessment from 2020 and looked for discrepancies in forest cover. They found that only two maps measured all of the parameters used in the EUDR’s definition of forest.
Some maps actually missed areas of sparse forests if they relied on tree height as a key forest measure. More often, the study revealed that most maps tend to overestimate how much forest once existed in a given area. For example, a farmer might be growing cacao trees on land that had been cleared long before the EUDR’s 2020 cutoff date. But if the map that authorities select incorrectly suggests that the land had been deforested more recently, the raw beans or chocolate produced from them may not be allowed into EU countries.
The issue stems from the ability of maps to show the differences between various types of land cover and use. Crops like soy or pastureland for beef typically appear starkly different from forests, making them easier to spot on remote-sensing maps. But a lot of the world’s rubber, cocoa and coffee is cultivated under the leafy canopies that shade agroforestry plots, producing more subtle differences on the maps.
“Traditional coffee systems really look like a forest,” Lippe said. Coffee plants are often shade-grown, he added, which makes pinpointing deforestation more difficult.
Lippe and his colleagues are working on a broader research project that combines the analysis presented in this study with on-the-ground pilot projects in forested countries to examine how well the maps can guide compliance.
“We want to test data sources, data uncertainty, and then compar[e] it in different countries” across commodity types, Lippe said.
In Mexico, their initial results suggest that the available maps aren’t sufficient for ensuring EUDR compliance. The researchers looked at 600 coffee plots that covered 1-5 hectares (2.5-12 acres) of land. Many had been there for decades, meaning they couldn’t have been deforested since 2020 and, therefore, the beans produced should be legal to export to EU countries. But the maps suggested that three-quarters of these plots didn’t comply with the EUDR, Lippe said.
“Basically, the map says [it’s] forest, but in reality, it’s an agroforestry system, which looks like a forest from on top,” he added. “That’s the challenge.”
‘Collateral damage’
Meine van Noordwijk, a distinguished science fellow at the research organization World Agroforestry, who wasn’t involved in the research, said the paper “made a very useful contribution to the debate on the suitability of available global forest maps as reference tools for EUDR-compliant deforestation monitoring.”
In June, van Noordwijk and his colleagues published a paper that outlines other forms of evidence that can help differentiate agroforestry from forest, in addition to remote sensing. They suggest that remote-sensing data should be buffeted with approaches like “ground-truthing” plots to look for evidence that they existed prior to 2021, and the examination of official records of land-use registration within the production countries.
Van Noordwijk’s team has also warned that smallholders doing agroforestry could be “collateral damage” of the EUDR because the regulation’s concept of deforestation is oversimplified.
“In our view evidence beyond imperfect maps will be essential, and needs to become a bigger part of EUDR procedures and debates,” van Noordwijk told Mongabay in an email.
Lippe said an overreliance on maps could lead larger companies to shy away from working with smallholder farmers because of the risk their products would be incorrectly linked to deforestation — an outcome he called “perverse.” That could also impact supply chains that rely on small-scale producers. For example, some 85% of the world’s rubber comes from smallholders, according to the World Business Council for Sustainable Development. Lippe said he and his colleagues have shown mismatches where remote-sensing maps indicate that rubber was coming from deforested areas when that was not the case.
To grapple with that issue, his team’s project wants to make policymakers aware of the issues with remote-sensing data.
“You can’t take a silver-bullet approach for remote sensing,” he added. “You need to be more careful.”
Banner image: A cocoa farmer and member of a local cooperative inspects the state of bean fermentation in Lembe Yezoum, Cameroon. Image © FAO/Beloumou Olomo Daniel.
John Cannon is a staff features writer with Mongabay. Find him on Bluesky and LinkedIn.
‘Insignificant risk’ EUDR proposal threatens fight against deforestation, critics say
EUDR compliance costs to be minimal, report finds — but industry disagrees
Citations:
Freitas Beyer, J., Köthke, M., & Lippe, M. (2025). Assessing the suitability of available global forest maps as reference tools for EUDR-compliant deforestation monitoring. Remote Sensing, 17(17), 3012. doi:10.3390/rs17173012
Van Noordwijk, M., Dewi, S., Minang, P. A., Harrison, R. D., Leimona, B., Ekadinata, A., . . . Sayer, J. (2025). Beyond imperfect maps: Evidence for EUDR‐compliant agroforestry. People and Nature, 7(7), 1713-1723. doi:10.1002/pan3.70088
Van Noordwijk, M., Leimona, B., & Minang, P. A. (2025). The European deforestation-free trade regulation: Collateral damage to agroforesters? Current Opinion in Environmental Sustainability, 72, 101505. doi:10.1016/j.cosust.2024.101505
FEEDBACK: Use this form to send a message to the author of this post. If you want to post a public comment, you can do that at the bottom of the page.
Credits
