Premium
Unveiling the multiscale teleconnection between Pacific Decadal Oscillation and global surface temperature using time‐dependent intrinsic correlation analysis
Author(s) -
Sankaran Adarsh
Publication year - 2017
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.4713
Subject(s) - teleconnection , pacific decadal oscillation , anomaly (physics) , climatology , sea surface temperature , oscillation (cell signaling) , correlation , series (stratigraphy) , hilbert–huang transform , scale (ratio) , statistical physics , geology , mathematics , physics , el niño southern oscillation , statistics , chemistry , paleontology , biochemistry , geometry , white noise , quantum mechanics , condensed matter physics
Global surface temperature is significantly influenced by different climate forcings operating at specific time scales. This study investigates the association between global surface temperature and the Pacific Decadal Oscillation ( PDO ) in a multiscaling framework in terms of both time scale of variability and non‐stationarity. First, the ensemble empirical mode decomposition ( EEMD ) is used for multiscale disintegration of the Global Surface Temperature Anomaly ( GSTA ) and PDO datasets. A close matching of the periodicity of different modes of PDO and GSTA is noticed, and the subsequent cross‐correlation analysis of the modes showed that their linear association is the most perceptible at the slowly varying trend component. The correlation between the different modes is further analyzed using a multiscale dynamic correlation method namely, time‐dependent intrinsic correlation ( TDIC ). This study found a strong long‐range positive correlation between the time series pairs in decadal and inter‐decadal modes exceeding 20‐year periodicity. Further it is found that, the multiscale teleconnection between PDO and GSTA is not always of unique character but associated with localized reversals in the nature of correlation in the time domain. The study further observed a similar pattern of correlation for both cold phases of the 20th century (1901–1924 and 1947–1976); whereas the pattern of correlation is different for the warm phases of PDO (1925–1946 and 1977–1995) in different process scales.