North Pacific sea surface temperature (SST) anomalies forced by the El Nino-Southern Oscillation (ENSO) via the atmospheric bridge (e.g., Alexander, Lau) peak a few months after the ENSO maximum in tropical east Pacific SSTs. This immediately suggests a null hypothesis for the Pacific Decadal Oscillation (PDO): variability in North Pacific SST must be distinguished not only from red noise but also from that induced by ENSO. Since the bridge causes the extratropical ocean to act as an integrator of the extratropical atmospheric response to ENSO, we expect that any connection between ENSO and the PDO will be lagged and not simultaneous. This relationship is indeed demonstrated, with simple lagged correlation statistics between three-month running mean indices of ENSO and the PDO. In general, the ENSO index leads the PDO index by a few months, with a correlation of about 0.6. The ENSO-PDO relationship has a pronounced semiannual cycle, however, with maxima in late winter and mid summer and minima in late spring and late fall; the lag of maximum correlation also varies from two months in summer to as much as five months by late winter. These results also show that the PDO, defined as year-to-year persistence of anomalous SST in the central North Pacific, is primarily a winter/early spring phenomenon. In particular, the PDO does not exist during summer.

Wavelet analysis is also used to show a strong relationship between ENSO and the PDO over the past 100 years. This analysis shows a notable interdecadal modulation of the power in different frequency bands for each index. This modulation has a similar pattern for each index, however, and the power spectra are highly coherent, particularly for those years and frequencies of high power. These results also suggest that the PDO cannot be considered to be independent of ENSO, even on decadal timescales. Thus, any analysis of purely extratropical air-sea coupling related to the PDO must first account for the ENSO effects.