Release notes summary
EUDR v1.13 strengthens deforestation and legality assessments through expanded reference-map coverage, refined risk logic, and new country-specific regulatory checks. Compared with v1.12, it completes the rollout of current deforestation data across the remaining countries, adds three Brazil-specific legality tests, and aligns several data-quality tests with the EU Information System (EUDR IS / TRACES) and ICE CoT requirements.
Key changes include:
Updated deforestation maps: DEF04 commercial maps are refreshed for 10 origins with 2026 data, and a new core timber plantation deforestation class is added for all supported countries. Updated public (DEF01) maps are also available in the Visual Investigation Tool.
New Brazil legality checks: LGL04, LGL05 and LGL06 add environmental-embargo, non-designated public land and forced-labour checks. The EUDR profile grows from 37 to 40 tests.
Stricter EUDR IS / TRACES validation: GET11 now truncates coordinates to six decimals (as the EUDR IS does) and scores duplicated vertices as Critical, because the EUDR IS rejects these geometries on submission.
ICE CoT-aligned country-boundary scoring: GER01 now uses two bands: inside or touching the boundary is Low, anything outside or crossing it is Critical.
More lenient farm-count thresholds: ATP01 moves its Medium, High and Critical bands upward, so farmers with many plots are flagged less often.
Commodity-specific farmer distance check: GEC09 now applies only to cocoa and coffee datasets.
Expanded protected-area context: LGL01 adds updated protected-area data and legal context for Kenya and Paraguay.
More accurate area and distance calculations: country-specific Coordinate Reference Systems (CRS) are updated.
Older maps and previous behaviour remain available by running an older EUDR Test Profile version.
Change log
Test | Test-Version | Changes | Background |
ATP01 | v1.03 | Medium, High and Critical thresholds for the number of farm plots per farmer are raised. | Farmer has implausibly high number of farms. The previous thresholds produced false-positive classifications for farmers with legitimately larger numbers of plots. Only the cut points move; the input and logic are unchanged. |
DEF04 | v1.04 | Commercial deforestation maps updated to 2026 versions for Brazil, Costa Rica, El Salvador, Guatemala, Honduras, Indonesia, Malaysia, Nicaragua, Panama and Papua New Guinea. A new core timber plantation deforestation class is included for all supported countries. | Farm plot overlaps with deforestation area (commercial map). The newest Space Intelligence maps (Q3 2026) identify conversion of timber plantations into other relevant commodities, which is deforestation under the EUDR. Severity thresholds are unchanged. |
DEF02 | v1.05 | The reference layer follows the updated DEF04 implementation. | One farm plot with deforestation has contaminated all farm plots of this farmer ID. Keeps contamination results consistent with the new DEF04 maps. Severity logic is unchanged. |
GER10 | v2.01 | The buffered geopoint assessment uses the updated DEF04 reference layer. | Geopoint is too close to deforestation area. Keeps geopoint results consistent with the new DEF04 maps. Buffer method and thresholds are unchanged. |
GEC09 | v1.05 | Applies to cocoa and coffee datasets only; skipped for all other commodities. | Farm is distant from farms with same farmer ID. Distance between plots of the same farmer is only a reliable anomaly signal in smallholder-heavy supply chains. Thresholds and country exceptions are unchanged. |
GET11 | v1.02 | (1) Coordinates are truncated to six decimals instead of rounded. (2) Duplicated vertices are now scored as Critical. | Farm plot needs review before uploading to EUDR IS / TRACES. The EUDR IS truncates coordinates before storing them and rejects geometries with duplicate coordinates. GET11 v1.01 already detected duplicated vertices but did not include them in the risk score. |
GER01 | v1.02 | Risk thresholds simplified to Low and Critical. Medium and High are no longer used. | Farm is outside of country boundary. Aligns with ICE CoT requirements: reference-based tests can only assess a farm plot that lies fully within the country borders. |
LGL01 | v2.02 | Updated protected-area data and legal context for Kenya and Paraguay. | Farm overlaps with protected areas. Integrates IBAT updates (August 2026) and granular legal context on farming permissions. All other origins keep their existing layers. |
LGL04 | v2.00 | Enabled in the EUDR Test Profile for Brazil. | Producer listed on Brazilian environmental embargo register (IBAMA). Requires farmer tax IDs. |
LGL05 | v2.00 | Enabled in the EUDR Test Profile for Brazil. | Farm plot overlaps with public non-designated lands in Brazil (Brazilian Forest Service, CNFP). |
LGL06 | v2.00 | Enabled in the EUDR Test Profile for Brazil. | Producer listed on Brazilian forced labor register (“Dirty List”). Requires farmer tax IDs. |
Before / after by test
GET11 — EUDR IS / TRACES validation
Farm plot needs review before uploading to EUDR IS / TRACES
This is the change most likely to affect existing datasets. Plots that showed Low risk under v1.12 can show Critical under v1.13, even though the data itself has not changed.
| Before (v1.12) | Now (v1.13) |
Coordinate processing | Rounded (half-up) to six decimals | Truncated to six decimals, as the EUDR IS does |
Duplicated vertices | Detected and listed in the test output, but not scored → Low | Critical |
Severity bands | 0 issues = Low; ≥1 high-risk issue = High; ≥1 critical issue = Critical | Unchanged |
A duplicated vertex is the same boundary point recorded two or more times in a row with identical coordinates. It usually happens during field mapping, for example when a point is saved twice at a corner of the plot. It does not change the shape or area of the plot, but it creates a boundary segment with zero length.
💡 Why this is Critical now: the European Commission GeoJSON documentation states that the EUDR IS truncates coordinates to six decimals before storing them, and lists duplicate coordinates among the cases the system rejects. Under v1.12 these plots would have been rejected at submission while showing a risk score of 0.
💡 Why your GIS tool may show no issue: desktop GIS tools such as QGIS check geometry on a flat (planar) model, where a repeated point leaves the polygon valid. The EUDR IS applies a stricter check and still rejects it. A clean result in QGIS does not guarantee the plot will be accepted in TRACES.
How to fix it: remove the repeated point in each affected plot and re-upload the dataset. The location of each duplicated vertex is shown in the GET11 test output (see Find and fix GET11 issues below).
Tests & Resolutions → GET11 details: “Duplicated vertices count as critical from version 1.02 onward”
LGL04, LGL05 & LGL06 — New Brazil legality checks
Three compliance tests are now part of the EUDR Test Profile for Brazilian datasets, bringing the profile from 37 to 40 tests.
Test | What it checks | Input | Low | Critical |
LGL04 | Producer listed on the federal IBAMA environmental embargo register | Farmer tax ID (CPF or CNPJ) | Producer not found | Producer found |
LGL05 | Farm plot overlaps with public non-designated lands (Brazilian Forest Service, CNFP) | Geopoint or polygon | No overlap | Any overlap |
LGL06 | Producer listed on the federal forced labour register (“Dirty List”) | Farmer tax ID (CPF or CNPJ) | Producer not found | Producer found |
💡 Add farmer tax IDs to your Brazil datasets. Map your tax ID column to the new Tax ID attribute in step 5 (Map attributes) of the dataset wizard. LGL04 and LGL06 are skipped if no tax IDs are provided, and return NA when a value is not a valid CPF or CNPJ. Medium and High are not used by these tests.
DEF04 — Commercial deforestation maps
Data change only. Severity thresholds are unchanged from v1.12.
Country | Before (v1.12) | Now (v1.13) |
Brazil | v5 | 2026: v6 |
Costa Rica | v3 | 2026: v4 |
El Salvador | v2 | 2026: v3 |
Guatemala | v3 | 2026: v4 |
Honduras | v2 | 2026: v3 |
Indonesia | v3 | 2026: v4 |
Malaysia | v2 | 2026: v3 |
Nicaragua | v2 | 2026: v3 |
Panama | v1 | 2026: v2 |
Papua New Guinea | v2 | 2026: v3 |
All other DEF04 origins keep their existing maps.
| Before (v1.12) | Now (v1.13) |
Pixels counted as deforestation | Core deforestation | Core deforestation + core timber plantation deforestation |
Applies to | — | All countries supported by DEF04 |
💡 Some plots may move from Low to Medium or Critical even when their geometry has not changed, because conversion of timber plantations into other relevant commodities is now part of the deforestation reference. DEF02 and GER10 follow this change automatically.
GER01 — Country boundary thresholds
Farm is outside of country boundary
Situation | Before (v1.12) | Now (v1.13) |
Fully inside the country | Low | Low |
Touching the boundary | — | Low |
Crossing the boundary | Medium | Critical |
Fully outside, 1–500 m from border | High | Critical |
Fully outside, ≥500 m from border | Critical | Critical |
💡 Medium and High are no longer used. A farm plot that is not fully inside (or touching) the country boundary cannot be assessed correctly by reference-based tests such as deforestation and protected areas, so it is flagged as Critical. The national boundary layers for Brazil (IBGE) and Uganda (Uganda Bureau of Statistics) have also been adjusted.
ATP01 — Farm-count thresholds
Farmer has implausibly high number of farms
Farm plots per farmer | Before (v1.12) | Now (v1.13) |
Low | <10 | <10 (unchanged) |
Medium | 10–19 | 10–99 |
High | 20–99 | 100–999 |
Critical | ≥100 | ≥1,000 |
💡 This is a one-way loosening. Every farmer with 20–999 plots scores lower than before, and no farmer scores higher. Plot geometries do not change, but the aggregated Data Quality score can still move on datasets with many plots per farmer.
GEC09 — Commodity scope
Farm is distant from farms with same farmer ID
| Before (v1.12) | Now (v1.13) |
Cocoa and coffee datasets | Test applied | Test applied |
All other commodities | Test applied | Test skipped |
Thresholds are unchanged: ≤10 km Low, 10–20 km Medium, 20–50 km High, ≥50 km Critical. The exception bands for Nigeria (cocoa) and Brazil (≤50 / 50–100 / 100–200 / ≥200 km) also remain.
LGL01 — Protected areas, Kenya and Paraguay
Reference data change. Applies to Kenya and Paraguay only; all other origins keep their existing protected-area layers.
| Before (v1.12) | Now (v1.13) |
Source | IBAT / WDPA | IBAT / WDPA |
Last updated | February 2026 | August 2026 |
Legal context | Standard protected-area classification | Granular legal context on whether agriculture is prohibited, conditionally permitted or generally permitted in each protected area |
Coordinate Reference Systems (CRS)
Country-specific CRS are updated to improve the accuracy and consistency of area calculations and spatial assessments. No severity thresholds change. Source GPS/GNSS data (EPSG:4326) is reprojected to the applicable working CRS before each test runs.
Test type | Working CRS in v1.13 |
Area tests | Country-specific projected CRS in metres. Countries spanning more than one UTM zone without a suitable national CRS (e.g. Brazil, Indonesia, Kenya, Mexico, Peru) use a custom equal-area projection. |
Distance tests | Country-specific projected CRS in metres; EPSG:7030 (WGS 84) for countries using a custom area projection. |
Shape tests | EPSG:3395 (WGS 84 / World Mercator) for all countries. |
💡 Some results and aggregated risk scores may change even when your data has not. More accurate projections improve area and distance measurements, and some geometries that previously could not be assessed can now be completed. The full list of projections per country is in the methodology document.
Working with v1.13 in the Verify Portal
1. Run a dataset with EUDR v1.13
Go to Datasets and click + to add a dataset.
In step 2, Select test profiles, choose EUDR and select version 1.13 (marked New release, 40 tests).
To compare with previous behaviour, run the same data with version 1.12.
Dataset wizard, step 2: EUDR version 1.13 selected
2. Map farmer tax IDs for Brazil datasets
In step 5, Map attributes, open the Tax ID dropdown.
Select the column in your file that contains the farmer CPF or CNPJ.
Continue to Run tests. LGL04 and LGL06 will use these IDs.
Dataset wizard, step 5: Tax ID mapped to the tax ID column
3. Check which version a dataset was run with
Open Results and select the dataset. The Test profile badge in the dataset summary shows the version, e.g. EUDR V1.13.
Results → Summary: Test profile badge “EUDR V1.13”
4. Find and fix GET11 issues
On the Results → Summary page, switch to Top data quality issues.
Find the GET11 row and click View farm plots to see the affected plots, or Info & Resolutions for guidance.
In the farm plot export, the GET11 test output lists the issue and its location, e.g. duplicated vertices at a longitude/latitude.
Remove the repeated point from the plot boundary and re-upload the dataset.
Results → Top data quality issues: GET11 row
GET11 Info & Resolutions: severity thresholds and resolution protocol
5. See all tests and the full methodology
Go to Tests & resolutions, select EUDR and version 1.13. You can open each test’s details, or download the complete methodology with Open methodology PDF.
Tests & Resolutions: EUDR 1.13 selected, Open methodology PDF
What did not change in v1.13
DEF01 severity thresholds: no overlap = Low, any overlap = Critical; geopoints ≥300 m from the nearest deforestation area = Low.
DEF04 severity thresholds are unchanged from v1.12; only the maps and the deforestation class change.
GER10 buffer method (2× reported farm size, capped at 5 ha) and thresholds are unchanged; only the reference layer follows DEF04.
GEC09 distance thresholds and country exceptions are unchanged; only the commodity scope changes.
GET11 severity bands are unchanged; what changes is the coordinate processing and which issues count as Critical.
LGL03 IPLC risk categories are unchanged.
GEC01–GEC08, GEP01–GEP02, GER05, GER06, other GET tests, and ATT / ATC attribute tests are unchanged.