Simulation of emergencies using almost periodic analysis of images of typhoon structure

A. A. Paramonov ; A. V. Kalach

Vestnik Ûžno-Uralʹskogo gosudarstvennogo universiteta. Seriâ, Matematika, mehanika, fizika, Tome 16 (2024) no. 4, pp. 67-74

Résumé

The paper considers the method of transforming images from rectangular coordinate system to polar coordinate system for further near-periodic analysis of typhoon structure based on generalized shear function. The basic method of near-periodic analysis involves dividing the data into high frequency oscillatory components and low-frequency trends based on the proportion theory, which can identify near-periodic characteristics. The paper shows that near-periodic analysis is applicable to both radius-vector and angle image data in a polar coordinate system. The practical component of the work is demonstrated by analyzing an image of a tropical cyclone. Experiments performed using horizontal and vertical sections of the image revealed significant near-periods corresponding to certain pixel values, such as 600 pixels in the horizontal section and 900 pixels in the vertical section. An angular analysis was performed for the polar coordinate system, which revealed additional almost-periods that accounted for the periodicity of the angular coordinates. Extending the angular interval to 720 and 1080 degrees confirmed the robustness of the almost-periodic analysis results, indicating that a twofold interval is sufficient for qualitative analysis. Thus, the proposed method can analyze image data more accurately and efficiently by identifying significant near-periods, which opens new perspectives for prediction and data processing in various scientific and applied problems of data analysis with ordered argument.

Détail

Keywords: data analysis methods, ordered argument data, trend, nonlinear fluctuations, near period, image analysis, emergencies, typhoons
Mots-clés : coordinate transformation.