| We study global climate networks constructed at various time-scales by means of ordinal time series analysis of monthly-averaged surface air temperature (SAT) anomalies. To quantify climate interdependencies, we compute the mutual information from the anomaly values and from its symbolic, ordinal-based representation. The ordinal analysis allows to identify changes in the topology of the network when varying the pattern, covering a short, intra- season time scale (e.g., of a few months) to a longer, inter-annual time scale (e.g., of a few years). We report changes in the network topology with the various time-scales and present evidence of correlations between geographical regions that occur at certain time scales only. |
|