Extraction of snow-covered area of Sabalan Mountain using Landsat satellite images by object-oriented classification method

Snow and ice cover is a vital resource and is very sensitive to climate change. Accurate measurement of the spatial distribution of snow and ice at regular intervals is essential for water resources management. In this research, the snow cover of Sabalan Mountain is extracted using Landsat 5 satellite images for 2000 and Landsat 8 for 2020 using object-oriented classification and also using NDSI and NDSINW spectral indices. The results of the classification maps extracted from the object-oriented method showed that the area covered by snow for the years 2000 and 2020 is 2500 and 1954 square kilometers, respectively, with a kappa coefficient and an overall accuracy of 99%. And 0/99 for both years. Then the snow covered area was extracted using NDSI spectral indices and NDSINW index. Both of these indicators are able to extract the snow cover surface with appropriate relative accuracy in cold and winter regions. However, because the NDSI index does not distinguish between snow and water, it may produce different results in areas with waterlogging associated with snow extraction. The NDSINW index, on the other hand, is more accurate in eliminating waterborne backgrounds. In relation to the present study, the NDSI index has extracted 2610 and 2577 square kilometers of snow cover and the NDSINW index has extracted 2577 and 1937 square kilometers for 2000 and 2020, respectively. The area obtained from the NDSINW index has been extracted by the object-oriented classification is closer.