Behind extreme variability: Unveiling causes using information theory beyond Shannon

Extreme events have a significant impact on nature, industry, agriculture, and society as a whole. From life-threatening heat waves and spring frosts that devastate crops in orchards and vineyards to other extremes such as epileptic seizures or financial market crashes, these phenomena remain a focus of intense scientific investigation.

The identification of causal relationships, specifically distinguishing cause from effect, is a rapidly advancing area of scientific research. Experts from various disciplines, including mathematics, physics, computer science, and others, are developing computational methods and algorithms to uncover causal links from experimental data.

Despite growing interest in these scientific fields, surprisingly few research teams integrate the study of causality with the analysis of extreme phenomena. Building on the information-theoretic generalization of Granger causality, Paluš et al. (2024) propose Rényi information transfer as a method for determining which of two or more potential causal variables gives rise to extreme values in an effect variable. Their study identifies the Siberian High as a key driver of increased probabilities of cold extremes in winter and spring surface air temperatures in Europe, while the North Atlantic Oscillation and blocking events are shown to induce shifts in the entire temperature probability distribution.

In this contribution we will employ Rényi information transfer to investigate the underlying causes of heat waves or warm extremes in summer surface air temperature in Europe. We will highlight the role of blocking events and examine the contribution of other relevant circulation phenomena, accounting for varying spatial and temporal scales as well as non-Gaussian probability distributions.

 

This research was supported by the Johannes Amos Comenius Programme (P JAC), project No. CZ.02.01.01/00/22_008/0004605, Natural and anthropogenic georisks, and by the Czech Academy of Sciences, Praemium Academiae awarded to M. Paluš.

 

Paluš, M., Chvosteková, M., & Manshour, P. (2024). Causes of extreme events revealed by Rényi information transfer. Science Advances, 10(30), eadn1721.