| description abstract | NSO-like variability is examined using a set of univariate indices based on unfiltered monthly global sea surface temperature (SST), sea level pressure (SLP), outgoing longwave radiation (OLR), sea level, and the three-dimensional ocean temperature (OT) fields. These indices, many of which correspond to the leading principal components (PCs) of the respective global fields, are highly correlated with each other. In combination with their spatial regression patterns, they provide a comprehensive description of ENSO-like variability in the atmosphere and ocean across time scales ranging from months to decades, from 1950 onward. The SLP and SST indices are highly correlated with one another back to the late nineteenth century. The interdecadal-scale shifts in the prevailing polarity of ENSO that occurred in the 1940s, the 1970s, and around the year 2000 are clearly evident in low-pass-filtered time series of these indices.On the basis of empirical mode decomposition, ENSO-like variability is partitioned into an interannual ?ENSO cycle,? to which equatorial ocean wave dynamics imparts a distinctive equatorial signature, and a red noise background continuum, most prominent on the interdecadal time scale, which resembles the ENSO-like variability in some models in which the atmosphere is coupled to a slab ocean. The background continuum bears the imprint of the Pacific?North American (PNA) pattern, the leading mode of the Northern Hemisphere wintertime variability of the atmospheric circulation over the Pacific sector. The superposition of the ENSO cycle and the background continuum imparts a distinctive frequency dependence to the patterns of ENSO-like climate variability. | |
