Has natural variability a lagged influence on global temperature? A multi-horizon Granger causality analysis

At present, the role of natural variability in influencing climate behaviour is widely discussed. The generally accepted view is that atmosphere-ocean coupled circulation patterns are able to amplify or reduce temperature increase from interannual to multidecadal time ranges, leaving the principal driving role to anthropogenic forcings. In this framework, the influence of these circulation patterns is considered synchronous with global temperature changes. Here, we would like to investigate if there exists a lagged influence of these indices on temperature. In doing so, an extension of the Granger causality technique, which permits to test both direct and indirect causal influences, is applied. A lagged influence of natural variability is not evident in our analysis, if we except weak influences of some peculiar circulation indices in specific periods.