Analyzing Seized Arable Land In Ukraine For PEJ

PEJ began by reviewing official data on the areas planted with arable crops. These statistics served as a baseline to identify cropland in regions affected by the Russian invasion — Crimea, parts of Donetsk and Luhansk regions, and later, parts of Zaporizhzhia and Kherson regions ^{Grand Theft. Dark schemes Russia uses to plunder Ukraine’s resources in broad daylight. Project Expedite Justice (PEJ).}.

EOSDA enhanced PEJ findings by analyzing satellite images to detect crop growth, classify land cover, and verify whether the land was still being cultivated. The satellite-powered analysis helped refine the land classification and confirm the ongoing use of Ukraine’s farmland by Russian occupying authorities.

Our collaborative research sheds light on the alarming exploitation of Ukraine’s agricultural land by Russian occupying forces. Satellite imagery exposes the extent of exploitation and devastation in the occupied areas. Long-term data analysis further highlights how the loss of irrigation and widespread damage following the Kakhovka Dam destruction have led to a collapse of agriculture in the region. These are not just immediate yield losses, but lasting ecological and economic consequences.

Dmytro Zhylchenko

GIS Specialists Team Lead at EOS Data Analytics

Our scientific team examined farmland in Kherson and Zaporizhzhia between 2020 and 2024. Using Sentinel-2 satellite images, we’ve made a detailed field-by-field analysis to map which fields were being cultivated and which were left unused. This made it possible to see exactly where crops were growing, even in small plots. We also used satellite imagery to tell the difference between the types of crops planted in each area.

One of the key factors that made the project possible was EOSDA’s trained and calibrated machine learning model for detecting major crops. Such models require large datasets for training, including ground-truth data from fields with labeled crop types, which is impossible to gather at wartime. But since we had already trained the model using data from 2019 to 2022, we could analyze both Ukrainian-controlled frontline zones and, critically, the temporarily occupied zones.

Oleksii Neskorovnyi

Business Analyst at EOS Data Analytics

We then cross-referenced this field data with control maps to separate Ukrainian-controlled areas from those under Russian occupation. This helped reveal not just the scale of seized farmland but also confirmed that the Russian occupying authorities continued agricultural production on Ukraine’s land.

Satellite-based field analysis was validated through statistical cross-checks with both Ukrainian and Russian data sources. Where official Ukrainian statistics covered Ukrainian-controlled areas, reports from the Russian Ministry of Agriculture (available from 2023) helped confirm crop types, yields, and harvested hectares in occupied zones. These reports aligned closely with what we detected through satellite imagery.

The combination of satellite intelligence and official data leaves little doubt: agricultural production on Ukraine’s land continues under occupation, often at a significant scale and with little transparency. While some occupied farmland remains productive, other areas, such as those deprived of irrigation after the Kakhovka Dam collapse, have seen agricultural activity nearly disappear. What was once a national resource is now being used without consent, tracked only through analysis like this one.

Independent satellite verification not only exposes the extent of land seizure and exploitation but also offers a critical tool for truth and accountability. If your organization is looking to collaborate or requires tailored satellite analysis, reach out to us at sales@eosda.com.