Mathematical Models for Predicting Extreme Temperature Events and Minimizing Their Consequences
Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì · 22.06.2026 · 19 · 307-315
DOI: 10.23939/sisn2026.19.307 ↗Kurz gefasst
Fachartikel in der Zeitschrift Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì, online erschienen am 17.05.2026. Verfasst von Andrii Fedosieienko. Die Arbeit nennt 30 Literaturangaben.
Abstract (Originalsprache)
The article considers the problem of forecasting extreme temperature phenomena as one of the key components of ensuring the stability of the functioning of modern natural-technogenic and socio-economic systems in the face of climate change. The relevance of the study is due to the increase in the frequency and intensity of abnormally high and low temperatures, which cause significant risks to energy infrastructure, transport systems, the agro-industrial complex and public safety. From the standpoint of system analysis, temperature processes are interpreted as complex non-stationary stochastic systems with the presence of rare but critically significant extrema that cannot be adequately described exclusively by traditional statistical or deterministic models. The aim of the work is to develop and study mathematical models for forecasting extreme temperature phenomena, focused on quantitative assessment of the probability of exceeding critical temperature thresholds and the formation of scientifically sound scenarios forminimizing their negative consequences. The study uses amultilevel approach that combines extreme value theory methods for analyzing the tails of temperature distributions, classical time models for reproducing trends and seasonality, and neural network algorithms for modeling nonlinear dependencies in the short and medium term. Hybrid models are proposed that integrate physically based parameters of temperature dynamics with adaptive learning algorithms, ensuring increased accuracy and stability of forecasts in non-stationarity conditions. The paper performs a comparative analysis of the effectiveness of different classes of models in terms of RMSE and MAE indicators, and also performs a quantitative assessment of the risk of extreme temperatures in terms of the probability of exceeding specified thresholds and the expected frequency of such events. It is shown that the integration of forecasting results into decision support circuits allows moving from reactive response to the consequences of temperature anomalies to preventive management of climate risks. The results obtained can be used in the development of intelligent monitoring and early warning systems, as well as in the tasks of strategic planning of adaptation measures in various application areas.
Zitieren
APA
Fedosieienko, A. (2026). Mathematical Models for Predicting Extreme Temperature Events and Minimizing Their Consequences. Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì, 19, 307-315. https://doi.org/10.23939/sisn2026.19.307
Vancouver
Fedosieienko A. Mathematical Models for Predicting Extreme Temperature Events and Minimizing Their Consequences. Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì. 2026;19:307-315. doi:10.23939/sisn2026.19.307
BibTeX
@article{Fedosieienko2026,
author = {Fedosieienko, Andrii},
title = {Mathematical Models for Predicting Extreme Temperature Events and Minimizing Their Consequences},
journal = {Vìsnik Nacìonalʹnogo unìversitetu "Lʹvìvsʹka polìtehnìka". Serìâ Ìnformacìjnì sistemi ta merežì},
year = {2026},
volume = {19},
pages = {307--315},
doi = {10.23939/sisn2026.19.307}
}
EUVIMED