Time series analysis of characteristics of the Antarctic circumpolar wave through Symbolic transfer entropy

Abstract

Since 19th century industrialization, rapid global warming has been known to cause abnormal weather conditions. In order to continuously and accurately observe climate change, Antarctica and its vicinity, where human direct impacts are least, are the proper places to study. In this dissertation, we investigate the time series data of the ocean and atmosphere surrounding the Antarctic continent (sea surface temperature (SST), sea ice extent (SIE), sea level pressure (SLP), and meridional wind speed (MWS)), where the Antarctic circumpolar wave (ACW) appears and we verify the characteristics and existence of the phenomenon. First, we use the Empirical Orthogonal Function (EOF) analysis to obtain modes or patterns that can best represent the variance of data by decomposing the time-space data of the SST and SLP into functions of space and time, and reproduce the approximate spatiotemporal data according to a wave number of the pattern in order to examine a trace of the ACW. The ACW appeared strongly only from the early 1900s to the 1998, and thereafter it does not show a clear pattern, and it is confirmed that it exists only locally. Secondly, since the EOF method has limitations in revealing the existence of ACW, its wave propagation, which is the most important feature of the ACW, is analyzed by using symbol transfer entropy. As a result, it is observed that the information transmission to the east is biased only in the SST and SIE data. Instead, in the case of four variables, we confirm that the information flow eastward or westward is shown remarkably near the vicinity of East Pacific where an effect of El Nino Southern Oscillation (ENSO) comes directly. In conclusion, along with the results of the ENSO index and EOF analysis, the circumpolar nature of the ACW seems to be influenced by the Antarctic circumpolar current, and the global impact of ENSO is also found in the analysis of the information flow which is considered to be one of the methods to grasp the dynamical structure of complex system.