The changelog below will highlight what has changed between this new version and the scoring model we used previously.
Table of Contents:
Release notes summary
As information about the European Commission’s EUDR Information System (IS) data requirements have become clearer and following a review of our scores and thresholds, several tests were either updated or created. We will be closely monitoring how data requirements evolve and industry adoption to adapt our methodology. In this release, the tests GET03 (“the farm plot has self-intersecting boundary segments”), GET07 (“farm plot polygon contains holes”) were both modified to produce a “Critical” risk score for conditions that would prevent upload to the EUDR Information System. Several tests scoring thresholds were capped to avoid assigning “Critical” risk where the EUDR requirements do not specify that level of data quality: ATC03 (“inconsistent names for farmer id”), ATT01 (“farmer name incorrect or illegible”), GET02 (“Farm plot has extremely short boundary segments”), GET04 (“Farm plot has duplicate boundary vertices”), GET01 (“Farm plot has extremely few boundary vertices”). Additionally, Meridia internal research on the operations and logistics of specific country and commodity supply chains informed new thresholds to be developed that better represent those agriculture sectors: GER03 (“Farm plot is far away from the aggregator”), GEP03 (“Farm plot is extremely large”).
Changelog
Test | Test Version | Changes | Background |
ATC03 | v1.01 | Critical and high risk outcomes removed, maximum risk is now capped at medium risk. | Farmer name is not required by EUDR. This test can generate additional insight but will not result in a high or critical risk outcome. |
ATT01 | v1.01 |
| As farmer name is not required by EUDR, this test does not return a high or critical risk. Additionally, this test is now automated and does not require manual intervention. |
GER03 | v1.01 | Threshold exceptions for Brazil (for cocoa, coffee, palm, and soy commodity):
Cocoa:
Coffee:
Soy:
Palm oil:
| Brazil can have slightly different supply chains, depending on commodities. A commodity specific threshold set has been set up to accommodate various situations in the field. |
GEP03 | v1.01 | Threshold exceptions for Brazil (all commodities):
Threshold exceptions for Ecuador (all commodities):
| Plantations in Brazil and Ecuador are larger than in many other countries where production lies mainly in smallholder farmers. Note: Planned for next release is differentiation between different areas within Brazil. |
GET07 | v1.00 | New test: identifies holes in a farm plot.
| Aligning with requirements for the EUDR Information System (see EUDR - Geolocation File Description). |
GET01 | v1.01 | Counting unique vertices only, not duplicated ones. New thresholds:
Farm plot polygons with <3 unique vertices are invalid and scored as NA. | Duplicated vertices are detected and reported by test GET04. |
GET03 | v1.01 | New thresholds in accordance with EUDR Information System:
| Aligning with requirements for the EUDR Information System (see EUDR - Geolocation File Description). |
GET04 | v1.01 | Critical risk outcome removed, maximum risk is now capped at high risk. | Duplicated vertices can be easily fixed, and are not detrimental to data quality. Therefore, this test does not generate a critical risk anymore. |
GET02 | v1.01 | Critical risk outcome removed, maximum risk is now capped at high risk. | Extremely short boundary segments are not detrimental to data quality. Therefore, this test does not generate a critical risk anymore. |
Changes to scoring model
The scoring model was also modified to produce a “Data Quality” score, a “Compliance” score, and an “Overall” score for each farm plot. This scoring model is simple to interpret - the maximum score from a test within each group is assigned as the group score.