APPLICATION OF MULTIVARIATE STATISTICAL METHODS FOR ANALYSIS OF CLIMATE PROJECTIONS

Abstract

The article presents the results of a comprehensive statistical analysis of temperature time series under RCP4.5 and RCP8.5 scenarios for the Zhytomyr city territorial community during 1981–2100. The research relevance is determined by the growing impact of climate change on environmental security and sustainable development of urbanized territories. In the context of global warming, understanding local temperature trends is particularly important for developing effective strategies for adaptation and mitigation of transformational changes, including climatic ones. Using parametric and non-parametric statistical methods, statistically significant (p < 0.05) relationships between temperature parameters were identified. Strong direct correlations were established between absolute temperature values and RCP4.5 (r = 0.92) and RCP8.5 (r = 0.87) scenarios, indicating significant warming in the long term. Spearman’s rank correlation confirmed the presence of strong monotonic dependencies (| r | > 0.7), indicating the stability of the identified temperature trends. Box-plot analysis showed significant temperature fluctuations (from -5.0186 °C to 1.8157 °C), emphasizing the need to consider temperature extremes when developing urban environmental security strategies. The obtained results provide a scientific basis for improving regional climate models and developing comprehensive climate change adaptation programs. This is particularly important for ensuring sustainable development of urbanized territories, biodiversity conservation, and enhancing urban ecosystem resilience to climate change. The practical significance of the results is confirmed by their implementation in environmental organizations’ activities and their use in developing local climate adaptation programs.