Machine Learning for Ukraine’s Forest Cover Damage Assessment based on Satellite Data

H. Yailymova, B. Yailymov, A. Shelestov and N. Kussul, "Machine Learning for Ukraine’s Forest Cover Damage Assessment based on Satellite Data," 2025 IEEE 13th International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS), Gliwice, Poland, 2025, pp. 1241-1245, doi: 10.1109/IDAACS68557.2025.11322046.

Abstract: This research presents a comprehensive machine learning approach for assessing forest cover damage in Ukraine based on satellite data from 2016 to 2024. Using Sentinel-1 and Sentinel-2 imagery processed through Google Earth Engine (GEE), we developed a Random Forest classification model to generate annual land cover maps with 10-meter spatial resolution. Our analysis reveals significant forest losses across Ukraine, totaling 190 thousand hectares by 2024 compared to 65 thousand hectares in 2021 (relative to 2016). The most substantial forest cover damage occurred within protected areas of the Emerald Network, particularly in eastern and southern regions affected by military operations. This research provides critical insights into the environmental impact of conflict on Ukraine’s forest ecosystems and establishes a methodology for ongoing monitoring of these valuable natural resources.