Tracking mountain geohazards in Uzbekistan with application of remote sensing and advances in data analysis.

Climate change is affecting the environment of Central Asia, the high mountains and particularly the paraglacial and glacial environments. Trends shown to affect the state and dynamics of glaciation; a gradual decrease in both area and volume, and the increase in the area of moraine deposits can be observed. Meanwhile, the frequency of occurrence of hazardous geological phenomena's related to glacial dynamics, such as outbursts of glacial lakes (GLOF'S), avalanches, rockfalls, blockages of river channels, as well as mudflows, is increasing. Observations after them are complicated by the inaccessibility of areas due to lack of infrastructure, seasonal restrictions, the required number of people to conduct observations in the field and financial costs. Studies shown that remote sensing technologies allow observation of changes and proved to have sufficient spatial and temporal resolution to assess the transformations taking place. In our study, we propose a method for assessing statistically significant changes occurring in the catchment for each of the image elements (pixel) over time using nonparametric statistics. The novelty of the work is the application of an assessment of the significance of these changes that allows to reduce the influence of human error in the analysis of images thus producing higher quality results and reducing time required. Although the technology is known in the world - and is used to solve the problems of machine learning and deep vision. For the territory of Uzbekistan and in this research area, this method is innovative, and the statistical assessment of significance over time is used for the first time. Thus, the method makes it possible to identify areas where significant changes have occurred due to events mentioned above, and reduces the amount of time and costs required to search for these areas to prevent or reduce further damage.