Evaluation of the possibility of vegetation interpretation on thermal infrared satellite images, case of the Southern Urals and Kuznetsk Alatau

Abstract

The paper presents the results of the vegetation cover interpretation using multitemporal thermal satellite images of two mountain-steppe areas: in the Southern Urals (Abzelilovsky district of the Republic of Bashkortostan) and in Kuznetsk Alatau (Ust-Abakansky and Shirinsky districts of the Republic of Khakassia). These areas have a large amount of field data on vegetation, which allows for reliable verification of satellite data. On the basis of field data and images of high spatial resolution in the optical range, vegetation maps were compiled, which became the basis for further interpretation of thermal images—images of the TIRS sensor, Landsat 8 satellite, were used. Methods of controlled and uncontrolled classification were applied to multitemporal images. In the course of the study, it was possible to establish that, based on the results of vegetation interpretation using thermal satellite images for a site in the South Urals, it is possible to determine forest areas with good reliability (up to 50–70 %), and confidently draw the border between forest and treeless areas. With satisfactory accuracy (up to 44 %), petrophytic steppes are determined. The site in the Southern Urals is characterized by a small size of the territory, a low diversity of plant communities, and rather a large dependence of the intensity of thermal radiation on the exposure of slopes. The site in Kuznetsk Alatau showed more representative interpretation results. Larch and birch-larch forests (up to 70 %), fir and birch-fir forests (up to 56 %), dwarf birch and moss-lichen tundras (up to 49 %), and steppe vegetation (up to 45 %) are most confidently recognized.