ISSN 0536-101X (print)
ISSN 2618-7299 (online)
8 (499) 261-82-86
English version

Картографирование природно-территориальных комплексов острова Котельный по разновременным радиолокационным снимкам Sentinel-1

Е.А. Балдина, К.А. Трошко

Страницы: 162-172

Аннотация. Исследованы возможности использования разносезонных радиолокационных снимков Sentinel-1A для картографирования природно-территориальных комплексов арктических районов на примере острова Котельный. Предложены способы создания производных радиолокационных изображений, пригодных для визуального и автоматизированного дешифрирования. Пространственные и временные различия коэффициента обратного рассеяния позволили выделить границы природно-территориальных комплексов острова, характеризующихся разным сочетанием рельефа, растительного покрова и грунтов. Географическая интерпретация выделов и составление тематической карты выполнены с опорой на известные закономерности взаимодействия радиоволнового излучения с земной поверхностью и дополнительные источники.

Ключевые слова: географическое дешифрирование, динамика коэффициента обратного рассеяния, космические радиолокационные снимки, многолетняя мерзлота, обработка изображений, сезонные изменения, тематическое картографирование, Sentinel-1, backscatter dynamics, geographical

Mapping of the natural landscapes of Kotelny island using multi-temporal Sentinel-1 radar images

E.A. Baldina, K.A. Troshko

Pages: 162-172

Summary.  Satellite synthetic aperture radar (SAR) remote sensing is the most suitable technique for studying and mapping of the vast uninhabited polar landscapes due to its capability to acquire the data regardless of daylight or cloud coverage. SAR data provide information not only about location of an object but also about its surface characteristic. The rapid increase of the open radar data volume after the launch of Sentinel-1 satellites suggests the relevance of a research aimed at developing methods of radar data use for thematic mapping. The research presents results of a new approach for mapping of the Arctic natural landscapes based on multi-temporal SAR data processing and interpretation. One of the Russian Arctic islands – Kotelny island (the New Siberian Islands archipelago) was chosen as an area of interest due to its rather high landscape diversity. 19 multi-seasonal Sentinel-1A SAR images, selected for the study, span the period from October 2015 to September 2016 with acquisitions every 12 days. So the images represent different ground conditions in each season of the year, covering all the seasons. Methods for creating modified radar images adjusted for visual and automated interpretation are proposed. The approach exploits the benefits of both the unsupervised classification applied to a large number of multi-temporal radar images, and the visual interpretation of modified preprocessed radar images of selected seasons as color composites. Spatial and temporal variations of backscattering coefficient gave the support for semi-automatic distinguishing the boundaries of landscapes which are characterized by different combination of the relief, vegetation and soils. Geographical interpretation of the outlined polygons and creation of the thematic map is based on the well-known principles of radio waves interaction with the Earth surface, as well as on supplementary geoinformation sources, such as optical satellite images (Landsat-8/OLI), weather data, and cartographic data available. The main result of the research is a map of natural landscapes of Kotelny island at a scale 1:750 000. The map contains information on landscapes features derived from seasonal backscatter behavior.

Скачать статью