The aim of the diploma thesis was to assess the Sokolov region landscape changes based on remote sensing data, to describe the trend of land use areas and to define the main causes that influenced trend.
The Landsat TM and Aster satellite images from the year of 1999 and 2005 were used for the proccession. The preproccessing of input data was the first step. There were three alternates created. Two alternates were various compositions of the bands and the third was build of the result of principal component analysis. Exctraction of the information from this preproccessed data was the next step. The automatic classification was used there, the metods of supervised (algorithm Isodata) and unsupervised (algorithms maximum likelhood, minimum distance and parallelepipeds). There were 8 classes classified: built-up areas, mine 3, mine 2, mine 1, bare soil, graseed areas, forest and water areas. The result of error matrix and the overall accurency assesment arbitrated which classified image of both year went to temporary analysis. There were the qualitative, spatial and quantitative changes detected. The ecological indicators (NDVI index, coefficient of ecological stability and coefficient of antropogenic influence) were determined.
The most important changes were between categories grassed areas, bare soil and mine 1-3 detected. This changes show the proccess of recultivation of this area. The water recultivation was detected on the mine Michal, there was grass converted to water. The changes showing the procedure of mining were detected around the mines Jiří and Družba. The category of mine 3 here moved on the west and on the border of mines the areas of forest were occupied by these mines. Quantification of results shows maximum increase of bare soil, it was more then half of original area. The most important increase was found in category mine 1 which converted from mine 2. Maximum decrease was recorded in category grassed areas.
The result of NDVI index shows increase of the value by 0.16. But the average change was 0.275, so the deviation of this average is equal -0.115 for the whole area. This value correlates to coeficient of ecological stability (decrease by 0.27) and coeficient of antropogenic influence results (increase by 0.13). However, detected changes of ecological indicators are only slight and the landscape of Sokolov region was valued as highly ecological stable with low intensity of usage and with middle gross vegetation during study time.
Next trend of landuse in this area will result of the proccess of the recultivation. There will increase areas of grass, forest and water. Ecological stability will be higher.