Mapping of forests in icings zones of the Selenga River Basin

Icings are cryospheric features that exert a significant influence on natural ecosystems and engineering infrastructure. Particular interest is drawn to icings located within forested areas, as trees can respond to their impacts, recording information about the dynamics of icing processes in their growth rings. This study focuses on mapping forest stands within icing-affected zones to identify promising areas for future dendrochronological investigations of icing processes in the transboundary Selenga River Basin. For the purpose of icing mapping, the study employed medium-resolution satellite remote sensing data (Landsat-8, Sentinel-2, Kanopus-V). To assess the presence of forest vegetation within icing zones, three global canopy height products were utilized: ETH (based on Sentinel-2 and GEDI), GFCH (based on Landsat and GEDI), and GCHM (an AI-based model derived from satellite imagery). The analysis revealed that approximately 65 % of the mapped icings are covered by forest vegetation. These forested icings are predominantly located in the Russian part of the basin, particularly on mountain slopes, where larch and cedar-fir forests are dominant. In contrast, forested icings are significantly less common in the Mongolian part of the basin and are mainly represented by sparse larch stands. At a key site a high-resolution digital elevation model with 0.1 m accuracy was created based on UAV-based aerial photography. This dataset was used to validate the global canopy height models. Comparative analysis at the test site demonstrated that global models tend to overestimate tree heights when compared to values derived from UAV. As expected, the model with the higher spatial resolution more accurately represented the horizontal structure of the forest canopy, though it still exhibited substantial height inaccuracies. Nevertheless, these global models proved useful for identifying areas containing trees that are potentially suitable for dendrochronological analysis. Thus, the spatial database of forested icing zones developed in this study can serve as a foundation for planning the collection of tree-ring samples and conducting further dendrochronological research on icing processes throughout the Selenga River Basin.