Remote research methods for assessing the impact of russia's armed aggression on the ecological security of natural reserved areas

Abstract

The article is devoted to solving a pressing scientific and practical problem, which consists in developing and implementing information technology for remote and geoinformation monitoring of nature reserves in Ukraine in conditions of military operations. The research is aimed at increasing the reliability of assessing environmental losses, detecting mechanical damage to landscapes, temperature anomalies, fire centers, hydrological disturbances and structural changes in the earth's surface. The work uses optical and radar satellite data, index analysis, multi-temporal composites, digital terrain models and machine learning algorithms. The proposed conceptual model of an integrated system allows for spatio-temporal analysis, generating analytical maps and supporting decision-making on the preservation and post-war restoration of natural ecosystems. The results obtained confirm the significant scale of degradation of protected areas and the need for systematic satellite monitoring as a tool for the evidence base of environmental losses. The work uses a set of satellite data of different spatial and spectral resolutions. Optical images of Sentinel-2, Landsat-8/9, PlanetScope and WorldView allowed to perform an analysis of vegetation cover, to detect traces of mechanical damage and burns. Radar data of Sentinel-1 and ICEYE provided monitoring regardless of weather conditions.
The mathematical apparatus of the study involves the use of classification algorithms of machine learning for automated selection of damaged areas. Spatio-temporal dynamics were analyzed by comparing multi-temporal composites.