Академическое издательство «Гео»

Keywords: geosystem, geom, аutomated interpretation, ervised classification, GIS technologies, landscape structure.


Geoinformational mapping of landscapes in the north-western part of Western Siberia using the Hansen mosaic dataset

E.I. Kuz’menko, A.A. Frolov, A.V. Silaev

V. B. Sochava Institute of Geography, Siberian Branch, Russian Academy of Sciences,
ul. Ulan-Batorskaya, 1, Irkutsk, 664033, Russia, kuzmenko48@mail.ru, f-vl1984@mail.ru, anton_s@bk.ru

Presented are the technique and results of studying the landscape structure of the north-west of Western Siberia on the basis of the Hansen mosaic consisting of elements covering the entire territory of the Earth. To study of this area used a synthesis of classical methods of thematic system mapping as described in the doctrine of geosystems of V. B. Sochava, and automated interpretation of remote sensing data and mapping using GIS technologies. The method of geoinformational mapping of the landscape structure with the use of MultiSpec and Quantum GIS software programs for the purpose of drawing up real-time raster landscape maps is shown. In drawing up the map and the legend, the principle of geomic system-hierarchical representation of the territory as a complex multi-level “bottom-up” structure from topological-level geosystems (groups and classes of landscape facies) to regional-level geosystems (geoms, groups and classes geoms) was used. Compilation of operational raster landscape maps of the north-west of Western Siberia involved multilevel automated interpretation of the Hansen mosaic and mapping of areas (clusters) on images with different brightness characteristics of the landscape structure of key areas and stationary physical and geographical profiles. A characteristic of this approach is also the use of a dynamic classification of geosystems based on the study of their variable states. Such an approach permits a monitoring of changes in landscape structure and its areas and the intensity of anthropogenic impacts on forest and marsh landscapes, and determination of permissible loads on geosystems in areas with high technogenic load of the oil and gas complex represented by the north-west of Siberia. An automated processing of data based on a supervised classification of territorial objects allows for the elimination of inaccuracies arising in visual image interpretation. Raster landscape maps can be used both for a further creation of vector maps and as an real-time information base for purposes of environmental management and protection.

Keywords: geosystem, geom, аutomated interpretation, supervised classification, GIS technologies, landscape structure.

DOI: 10.21782/GiPR0206-1619-2018-2(145-153)