APPLICATION OF SHIFT FUNCTION DISTANCE METRICS FOR NEAR PERIODIC ANALYSIS OF SATELLITE IMAGES

To this day, the world community faces the problems of emergencies, response to them, as well as modeling and forecasting the development of such processes. Processing data on the causes of emergencies is critical to ensure the safety of the population, as well as to minimize the consequences of devastating events caused by the sources of dangerous situations. Analyzing accumulated historical data on emergencies and their causes makes it possible to identify and study systematic factors in the dynamics of emergency development. One of the main sources causing emergencies is tropical cyclones. These natural phenomena are characterized by strong winds, significant precipitation and, in coastal regions, storm surges. At present, the observation of tropical cyclones is mainly carried out by satellite remote sensing methods, which poses a challenge to researchers to implement and apply methods of media data processing and modeling. Most of the methods and studies are aimed at analyzing the motion of a tropical cyclone, i.e., the dynamics of its development in terms of speed and direction of motion. This paper discusses the approach of analyzing the dynamics of typhoon structure development based on near periodic analysis. This paper considers the application of distance metrics underlying shift functions - the main tool for almost periodic analysis of data with an ordered argument. Some of the distance metrics for the shear function are presented and images of tropical cyclone Nanmandol are analyzed under different distance estimation metrics. The results showed the independence of the obtained almost periods from the type of applied distance metric in the shear function.