Assessing causal dependencies in climatic indices

Abstract

We evaluate causal dependencies between thirteen indices that represent large-scale climate patterns (El Nino/Southern Oscillation, the North Atlantic Oscillation, the Pacific Decadal Oscillation, etc.) using a recently proposed approach based on a linear approximation of the transfer entropy. We demonstrate that this methodology identifies causal relations that are well-known, as well as it uncovers some relations which, to the best of our knowledge, have not yet been reported in the literature. We also identify significant changes in causal dependencies that have occurred in the last three decades.