Tropical Pacific-North Pacific teleconnection in a coupled GCM: Remote and local effects

Abstract

The connection between Tropical Pacific and North Pacific variability is investigated in a state-of-the art coupled ocean-atmosphere model, comparing two 20th century simulations at T30 and T106 atmospheric horizontal resolutions. Despite a better simulation of the frequency and the spatial distribution of the Tropical Pacific anomalies associated with the El Nino Southern Oscillation (ENSO) in the high-resolution experiment, the response in the North Pacific is scarcely different from the low-resolution experiment where the ENSO variability is weaker and at higher than observed frequency. In the North Pacific, the response of surface atmospheric fields to the variability in the Tropical Pacific appears to be affected by local coupling processes significantly different in the two experiments. The coupling between sea level pressure (SLP) and sea surface temperature (SST) in the North Pacific as well as the influence of the Tropical Pacific SST has been measured here by means of the ‘coupled manifold’ technique. In the low-resolution case the SLP variances linked to the fraction of North Pacific SST not influenced by the Tropical Pacific are weak suggesting that the remote influence is strong, consistently with the observations. On the contrary, in the high-resolution experiment the fractions and the patterns of the SLP variances due to the Tropical Pacific SST and those linked to the North Pacific SST are comparable. In the latter case, model systematic errors in the northwestern Pacific influences the local coupling processes thus triggering the remote response. We conclude that an increased atmospheric horizontal resolution does not reduce the coupled model systematic errors in the representation of the teleconnection between the North and the Tropical Pacific and that the validation of coupled models has to consider both remote and local processes.