Background . The 2022 hostilities in Kyiv's suburbs (Irpin, Horenka, Bucha) triggered substantial anthropogenic environmental change and the emergence of so-called belligerent-recreational systems (BRS), where pockets of war-related destruction coexist with recreational resources. Vegetation degradation and other alterations of land cover in these areas require detailed investigation and assessment to inform subsequent revitalization. The aim of this study was to detect spatial changes in the structure of land (vegetation) cover and to assess the degree of degradation within belligerent-recreational systems using Earth observation (remote sensing) methods by calculating spectral indices and their differenced forms. Methods . To monitor the spatial dynamics of BRS, Sentinel-2 and Landsat-8 imagery was used to construct pre- and post-hostilities snapshots within the study sites. We computed the NBR (Normalized Burn Ratio) and NDVI (Normalized Difference Vegetation Index) and, chiefly, their differenced forms – dNBR (Differenced Normalized Burn Ratio) and dNDVI (Differenced Normalized Difference Vegetation Index) – which reflect degrees of land-cover damage due to burning and other biomass losses. The resulting dNBR and dNDVI values were classified by damage severity (from critical to minimal). Validation of the proposed classification scales was carried out by calculating pairwise (Pearson) correlation coefficients. The results revealed a strong correlation between dNBR and dNDVI (r ≈ 0.87). Results . Monitoring of spatial changes in BRS land-cover structure showed that the most intense fighting resulted in large-scale fires – in the northern quarters of Irpin and Bucha, and the central and eastern parts of Horenka – captured by the computed dNBR. According to dNBR, 12 % of the study area fell into the critical damage class, 20.5 % into the severe class; roughly one-third of the territory exhibited moderate damage, about one-quarter minor damage, and 9 % was almost unaffected. Analysis of dNDVI indicates that more than 80 % of biomass was lost over 13 % of the area, 50–80 % over 32 %, while 39 % experienced moderate losses and 16 % minimal losses. Additionally, NDVI analysis helped identify sites with mechanically induced vegetation damage resulting from warfare. Conclusions . Calculating the dNBR and dNDVI spectral indices enabled a quantitative assessment and delineation of degradation levels in the land (vegetation) cover of BRS, as well as visualization of the spatial differentiation of damage by severity classes. Among these, areas with weak or moderate damage were observed to show rapid vegetation recovery, whereas zones of critical and severe damage require the implementation of robust reclamation and remediation measures. The high correlation between dNBR and dNDVI confirms the dominant role of fires in degradation; however, a combined analysis of these indices provides a more complete assessment that also captures mechanical damage. The quantitative and qualitative results of the monitoring of land-cover conditions across the study sites, together with their map-based modeling, can be used to design strategic measures for the ecological rehabilitation of areas affected by military actions.
Udovychenko et al. (Wed,) studied this question.