PSI - Issue 20

S.A. Tikhonova et al. / Procedia Structural Integrity 20 (2019) 230–235 S.A. Tikhonova et al. / Structural Integrity Procedia 00 (2019) 000 – 000

232

3

2. Materials and methods. Among the most common and used indices using quantitative estimates of vegetation cover is vegetation index NDVI (Normalized Difference Vegetation Index). NDVI index uses the contrast characteristics of two channels from multispectral raster dataset — absorption by pigment of chlorophyll in red channel and high reflectivity of plant materials in infrared channel (NIR). For each pixel, an index value from -1 to 1 will be calculated, which will characterize vegetation intensity of this pixel. Using this index, one can investigate time series of satellite information characterizing the state of vegetation. Monitoring of vegetation state cover will make it possible to determine changes occurring in the medium under the influence of pipeline, inhibition and change of vegetation types, initialization of activity or attenuation of geocryological processes. The formula for calculating NDVI index was proposed by Dubinin (2002): Multiple-time images from Landsat 5, 7, and 8 satellite were investigated. ENVI software package was used to process images and calculate NDVI vegetation index by Harris (2018). July month was chosen, since it is the month of the greatest flowering of plants. Measurement of relative change in vegetation state over time using satellite data should be compared with reference data (background), determined prior to pipeline construction or comparable areas of untouched vegetation and soil. To study the impact of pipeline on the environment, three sites were selected near the oil pipeline and three background sites. Construction of the first stage of 2694 km long pipeline on the territory of the Republic began in 2006 and was completed in 2009. The second phase of construction length of 2046 km, began in 2010 and finished in 2012. In 2015, the capacity was increased to 58 million tons per year as reported in Stroytransgas (2018). Therefore, time series of images includes 2006 as a background year, as well as to study changes in the state of vegetation in pipeline construction and operation years - 2009, 2011, 2015 and 2017. The main sites were selected next to the pipe at a distance of 50 m, and background areas were selected at a distance of 1 km from the pipe. A set of local reference pixels from undisturbed areas, with which you can compare pixel changes near the pipeline, can standardize time series estimates. Of course, climatic conditions affect the change in vegetation state. We used the weather data of the site “Weather in 243 countries of the world”. 3. Results and discussion. Based on vegetation types in the areas of pipeline, we carried out NDVI assessment of vegetation state separately for two groups of sites. Group 1 - sites № I and № II next to the city of Lensk and Vitim, group 2 - site № III next to the city of Neryungri. 3.1. Sites on the linear part of the pipeline • I -st site, located next to the town of Lensk; • II -nd site is located next to Vitim; • III -rd site is located next to the city of Neryungri. As an example, Figure 1 shows layout of the study area № I, II and III. On the map, number 1 indicates the main areas, and number 2 indicates the background areas. Multiple-time images from Landsat 5, 7 and 8 satellite were investigated. The time interval is July 2006, 2009, 2011, 2015 and 2017. To calculate NDVI, ENVI software package, NDVI tool and Vegetation Index Calculator were used. The results are shown in Table 1 and 2. The main sections were selected near the pipe at a distance of 50 m, and background areas were selected at a distance of 1 km from the pipe. NDVI= . (1)