Kuchma T. Landscape diversity indication using remote sensing

The thesis is focused on the theoretical study and improvement of methods for landscape diversity monitoring using remote sensing for conservation of agricultural landscapes biotic potential. The efficiency of 10 landscape metrics for landscape diversity evaluation is demonstrated, in particular: the largest fragment index, the edge density index, core area index, Shannon diversity index, Simpson's evenness index, contact index, the circle index, patch density index, the proximity index and the cohesion index. The landscape diversity index interpretion method is developed based on the measure of arable land and natural land proportion in the agricultural ladscape structure. The proposed metodology of optical Landsat-8 and SAR Sentinel-1data fusion make it possible to enhance the spatial resolution of input data (to 8.25 m). CORINE landcover classification is detailed by adding the fourth level for the category of agricultural land with allocation of 20 classes for more efficient landscape diversity monitoring including crop structure and rotation. These results confirm the efficacy of landscape diversity index as a tool for landscape diversity monitoring, as well as to quantify the projected or simulated changes in land use patterns and their impact on the diversity of agricultural landscapes.