ANALYSIS OF THE DYNAMICS OF DROUGHT IN A SEMI-DESERT ZONE USING THE CLUSTERING METHOD OF MULTISPECTRAL IMAGES

Abstract

Drought is a serious natural threat that causes significant damage to the environment. In recent decades, humanity around the world has experienced a drought with serious socio-economic and environmental consequences. According to the results of the work of the authors who studied the problem of drought, it is expected that by the end of the century the severity and frequency of droughts will increase. The purpose of this work is to identify the dynamics of drought from 2015 to 2023 in the territory of the Uygur district of the Almaty region of Kazakhstan, based on meteorological indices and space monitoring data. A review of the physical and geographical location of the monitoring object, a potentially vulnerable drought zone of the semi-desert zone of the temperate climate zone, was conducted. The Earth remote sensing data obtained in various ranges of the optical spectrum using Sentinel-2 satellite images were thematically processed and the multispectral image clustering method was applied. The NDVI values were taken as parameters for clustering satellite images by differentiation into non-overlapping spatial territories. The spatial and temporal dynamics of the NDVI values of vegetation cover are analyzed. The NDVI values were analyzed based on satellite imagery data obtained from the Sentinel-2 satellite on the EOS digital platform, https://crop-monitoring.eos  for 2015-2023 years.