Registration Number
0225U004321,
(0124U003753)
, R & D reports
Title
Geospatial monitoring system for the war impact on the agriculture of Ukraine based on satellite data
popup.stage_title
Створення геопросторової системи моніторингу впливу війни на сільське господарство у хмарному середовищі
Head
Shelestov Andrii Yu., Доктор технічних наук
Registration Date
01-12-2025
Organization
National Technscal University of Ukraine "Kiev Polytechnic Institute".
popup.description1
The goal of the project is the development and implementation of a geospatial system for monitoring the impact of war on the state of agricultural land in Ukraine for regular assessment of direct losses of agricultural production, forecasting the yield of major crops, substantiating state support measures for the agricultural sector and improving the effectiveness of management decisions in the agricultural sector based on geo-information technologies and earth observation data.
popup.description2
This project produced the first cloud-based geospatial system in Ukraine designed for regular detection and quantitative assessment of war-induced damage to agricultural lands. The system integrates multi-year Sentinel-2 time series, advanced spectral analysis, and ensemble machine-learning algorithms capable of distinguishing conflict-related degradation from natural seasonal dynamics. A comparative assessment of major satellite platforms identified Sentinel-2 as the most balanced source for nationwide monitoring due to its temporal frequency, spectral richness, and open access.
An extended feature set was developed, including vegetation indices (NDVI, GCI), spectral bands B2–B3, local statistical homogeneity metrics, temporal signatures, and adaptive anomaly detectors for identifying small-scale objects such as craters, burn scars, or machinery traces. The damage-detection methodology was enhanced through a combination of spectral anomaly analysis, structural descriptors, and machine-learning techniques, ensuring robust performance under diverse regional and phenological conditions.
Using Google Earth Engine and CREODIAS, the project implemented a complete processing pipeline: automatic ingestion of satellite data, cloud filtering, predictor generation, model execution, and production of cartographic layers with a bi-weekly update cycle. The resulting geoportal provides interactive access to damage maps, statistical indicators, temporal profiles, and spatial analytics tools, enabling practical adoption by governmental bodies, international organizations, and analytical institutions.
The scientific significance of the system is demonstrated by its practical adoption by the World Bank and the European Commission.
Product Description
popup.authors
Andrii Y. Shelestov
Kolotii Аndrii V.
Volodymyr Lukin
Yailymova Hanna O.
popup.nrat_date
2025-12-01
