A Scalable and Automated Recurrence Plot Method for Detecting Climate Change: The Case of the UTCI Bioclimatic Index in Central Europe

Recent decades have brought increasing effects of climate change and variation. Climate changes, indicated by various bioclimatic indices, influence the functioning of living organisms, including humans. Previous studies have documented multiyear fluctuations in such indices. In this study, we employ the Universal Thermal Climate Index (UTCI), a recently developed bioclimatic indicator designed to condense complex thermodynamic and meteorological conditions into a single value representing specific categories of thermal stress. The average annual UTCI values for the period 1826–2021 in Kraków, Poland, increased at a rate of $0.27~^{\circ }$ C per decade, with the most pronounced rise observed during the winter season, as documented in prior studies. However, when considering the more recent period of 1951–2023, the trend intensifies significantly, reaching $0.61~^{\circ }$ C per decade, indicating a clear acceleration in bioclimatic warming. In the present study, we apply an automatic recurrence plot (RP) analysis method to investigate these changes in human bioclimate. We introduce a novel approach by automating RP interpretation with a genetic algorithm and statistical deterministic analysis, thereby providing readers with ready-to-use algorithms. The research focuses on a univariate time series of UTCI recorded from 1951 to 2023 in three locations in Poland–a key region of Central Europe where air masses from different parts of the continent interact. The main conclusion from the analyzed data is the identification of a date range (1980–1990), making a phase-transition-type passage between two periods that differ in statistical behaviors of UTCI. This bioclimatic change led to an increase in the frequency of very strong heat stress events and a decrease in the occurrence of very strong cold stress days. The RP analysis demonstrates that the data from 1990 differ most from data from the preceding period 1952–1965. This provides details of the findings reported in previous research regarding climate change in Poland. This discovery of a distinct transitional period marking a shift in the climatic regime in Poland, based on RP analysis, enriches existing knowledge on regional climate change with new empirical evidence.