Causal networks for quantifying the links of boreal winter atmospheric variability with Mediterranean climate on multiple temporal scales

Large-scale atmospheric circulation is the major driver of near surface climatic variability and extremes, with teleconnection patterns being a significant component by connecting climates in remote locations. The recently developed powerful concept of causal effect networks (CENs) enables the detection of causal relationships among a set of actors by removing the confounding effects of autocorrelation, indirect links, and common drivers, retaining eventually only the actual causal links.

In this study, we apply the causal discovery algorithm to analyze the causal links among teleconnection patterns and other circulation features of the North Hemisphere and their influence on Mediterranean winter climate variability. We employ different sets of actors and multiple-scale temporal resolution reanalysis datasets to examine the consistency of the CENs across different timescales and to uncover the underlying mechanisms of their links. By investigating the strength of the different remote drivers compared to local drivers, this analysis contributes to a better understanding of the mechanisms controlling the intraseasonal variations in boreal winter circulation over the Mediterranean. In addition, we find that the strength of the causal links is affected by interannual and multidecadal variability, suggesting the potential involvement of external physical mechanisms.