REMOTE SENSING STUDY OF VEGETATION COVER DURING CONSTRUCTION WORK (A CASE STUDY OF THE TERRITORY OF VOSTOCHNYI COSMODROME)
A. V. PUZANOV1, YU. M. TSIMBALEI1, I. A. ALEKSEEV2, N. M. KOVALEVSKAYA1, T. A. PLUTALOVA1
1Institute for Water and Environmental Problems, Siberian Branch, Russian Academy of Sciences, ul. Molodezhnaya, 1, Barnaul, 665038, Russia puzanov@iwep.ru 2Blagoveshchensk State Pedagogical University, ul. Lenina, 104, Blagoveshchensk, 675000, Russia igoralx20071@mail.ru
Keywords: стресс растительности, вегетационные индексы, типизация растительных комплексов, гло баль ная космокартографическая основа, снимки высокого разрешения, многоспектральная классификация, vegetation stress, vegetation indices, types of vegetation community, global digital basemap, remote sensing basemap, high-resolution images, multispectral classification
Abstract
A new technique for compiling vegetation maps on the basis of two-stage technology is presented. It includes a preliminary assessment of the state of vegetation, and a classification with training samples with the simultaneous use of the digital basemap (GlobCover, 2009) and multispectral images of very high spatial resolution (Alos and Rapideye). We examine the specific features of spectral reflection of vegetation and the vegetation indices for the analysis of vegetation viability. We developed the structuring of plant communities of the study area and the referencing of standard types to the GlobCover classes used as the basis. The possibilities of a multispectral classification have been studied on the basis of very high resolution images in the context of the study territory and the GlobCover classes. Land cover maps are obtained on the basis of training samples and the most plausible classification. The vegetation dynamics (2009-2012) has been investigated for some areas of Vostochnyi Cosmodrome occupied by facilities under construction. Specifically, updates were made concerning infrastructure development, the devastation of woody vegetation, etc. A new method of obtaining vegetation maps is proposed, based on using a global digital basemap and high spa tial resolution images for the desired local areas.
|