The aim of the present study was to identify multi-decadal variability(MDV) relative to the current centennial global warming trend in available observation data. The centennial global warming trend was first identified in the global mean surface temperature(STgm) data. The MDV was identified based on three sets of climate variables, including sea surface temperature(SST), ocean temperature from the surface to 700 m, and the NCEP and ERA40 reanalysis datasets, respectively.All variables were detrended and low-pass filtered. Through three independent EOF analyses of the filtered variables, all results consistently showed two dominant modes, with their respective temporal variability resembling the Pacific Decadal Oscillation/Inter-decadal Pacific Oscillation(PDO/IPO) and the Atlantic Multi-decadal Oscillation(AMO). The spatial structure of the PDO-like oscillation is characterized by an ENSO-like structure and hemispheric symmetric features. The structure associated with the AMO-like oscillation exhibits hemispheric asymmetric features with anomalous warm air over Eurasia and warm SST in the Atlantic and Pacific basin north of 10°S, and cold SST over the southern oceans. The Pacific and Atlantic MDV in upper-ocean temperature suggest that they are mutually linked.We also found that the PDO-like and AMO-like oscillations are almost equally important in global-scale MDV by EOF analyses. In the period 1975–2005, the evolution of the two oscillations has given rise to strong temperature trends and has contributed almost half of the STgm warming. Hereon, in the next decade, the two oscillations are expected to slow down the global warming trends.
展开▼
