Many deforestation-free checks run today were done once: a batch of plots was assessed against satellite data up to a certain year, and the result was filed. The plot was “clear as of 2023”, and nothing happened afterwards. Under the EUDR, that is not enough, and it is worth understanding why.
What does the EUDR actually ask about deforestation?
Article 3 of Regulation (EU) 2023/1115 bars products that are not deforestation-free. “Deforestation-free” means the commodities were produced on land not subject to deforestation after 31 December 2020 (Article 2(13)). Article 9(1)(d) adds that any deforestation or forest degradation on a plot automatically disqualifies all commodities from that plot.
Three features of the law follow:
- The window is open-ended. It starts at the cut-off date and runs up to the production of the commodity being placed on the market. It does not stop at the year of the first assessment.
- The obligation is per product. Each due diligence statement covers specific products, with the date or time range of their production. The evidence has to cover that production.
- The obligation continues. Operators who learn of new information indicating a risk must inform authorities and their customers (Article 4(5)), and must review their risk assessments at least once a year (Article 10(4)).
Why is a frozen assessment a problem?
Consider a cocoa plot assessed in early 2024 against data up to 2023. It showed no forest loss since 2020. Suppose part of the plot, or a forest patch at its edge, was cleared in 2025. Cocoa from the 2026 harvest is then shipped under a statement relying on the 2024 assessment.
The statement declares negligible risk. The evidence behind it does not cover the period when the clearing happened. If a competent authority checks the plot, it will see the clearing, and the operator will not be able to show how its due diligence addressed it.
Authorities have the means to find such gaps. They may use “any technical and scientific means adequate to determine whether the relevant products are deforestation-free, including Earth observation data such as from the Copernicus programme” (Article 18(2)(d), as amended by Regulation (EU) 2025/2650). Anyone can submit substantiated concerns (Article 31). An assessment that stops at one year leaves a gap that grows every month.
What does assessing “up to today” involve?
Assessing a plot up to today means combining two kinds of evidence.
A forest baseline at the cut-off date
Was the plot, or part of it, forest on 31 December 2020? Several public global and national maps answer this, with different definitions and resolutions. The EU Observatory on deforestation and forest degradation, run by the Commission’s Joint Research Centre, publishes a global forest cover map for 2020. It is a useful reference, but operators may use other sources, and no single map is right everywhere.
Remember the EUDR definitions: forest is land of more than 0.5 hectares with trees higher than 5 metres and canopy cover of more than 10%, excluding land predominantly under agricultural use. Agricultural plantations, including agroforestry, are not forest. Baseline maps sometimes confuse shaded cocoa or coffee with forest, which produces false alarms.
Change since the cut-off date, up to now
Did anything change on the plot after 31 December 2020? Public forest-change and disturbance alert systems, based on optical and radar satellite imagery, detect forest loss at intervals of days to weeks. Radar sees through cloud, which matters in the humid tropics where cocoa, coffee, rubber and palm oil are grown.
Combined, the two answer the legal question: was this land forest at the cut-off, and has it been converted since?
Why use several sources rather than one?
Every dataset has blind spots: cloud cover, resolution, definitions of forest, detection thresholds. A single source produces both false positives (a cocoa farm flagged as cleared forest) and false negatives (small-scale clearing missed). Combining several sources and looking for convergence gives more reliable results, and makes it clear when a plot needs a closer look.
This is the principle behind Whisp (“What is in that plot?”), the free, open-source solution developed by FAO. It combines multiple public datasets using a convergence-of-evidence approach and classifies plots by risk, including a “more information needed” category for cases where evidence is unclear.
Should buyers be able to choose the method?
Yes. The EUDR does not prescribe a method. It prescribes an outcome (no deforestation after the cut-off) and a process (information, risk assessment, mitigation). Operators are responsible for their statements, so they need to be able to explain and defend the method used.
In practice, buyers differ:
- some apply a public reference method such as Whisp;
- some have their own internal methodology, thresholds and datasets;
- some need to show consistency with a certification standard or a sector approach.
A good assessment service supports all three. The plot data stays the same, the method is a choice, and the result records which method, which sources and which date were used. When two buyers source from the same farmer, each can apply its own method to the same verified plot, without asking the farmer to be mapped twice.
What should an up-to-date deforestation record contain?
For each plot and each assessment:
- the plot geometry used, and its version;
- the baseline sources and the change sources used;
- the method and thresholds;
- the date of the assessment and the period covered, from 31 December 2020 to the latest available data;
- the result, per source and combined;
- any follow-up: imagery review, field verification, farmer explanation;
- the decision, and who made it.
This record is part of the due diligence report that sits behind each statement.
How does Farmer1st assess deforestation?
Our in-house geo service validates plot polygons and runs a multi-source deforestation assessment up to today, not frozen at a past year. Buyers can apply their own calculation method or reference methods such as Whisp (FAO), on the same verified plot data. Results feed due diligence reports and TRACES-ready statements, and new signals on a plot can be connected to fraud detection. See Geo & polygons, EUDR compliance and the EUDR guide.
Frequently asked questions
How often should plots be re-assessed?
At least before each due diligence statement that relies on them, since the statement covers a specific harvest. Operators must also review their risk assessments at least once a year. Between those points, monitoring for new alerts lets you react when something changes rather than find out at the next check.
Is a forest map for 2020 enough on its own?
No. A 2020 baseline tells you where forest was at the cut-off date. It does not tell you what happened afterwards. You need both: a baseline to know whether the plot was forest, and change data to know whether it was converted after 31 December 2020.
What if two methods disagree about a plot?
Treat disagreement as a reason to look closer, not to average the results. Check the plot against high-resolution imagery, the farmer's history of the plot and field verification. Record the evidence and the decision, because that reasoning is part of the due diligence the operator must be able to show.
Updated ·
