Investigating the Role of Ocean-Atmosphere Coupling in the North Pacific Ocean

Air-sea interaction over the North Pacific is diagnosed using a simple, local coupled autoregressive model constructed from observed 7-day running-mean sea surface temperature (SST) and 2-m air temperature T-A anomalies during the extended winter from the 1 degrees x 1 degrees objectively analyzed air-sea fluxes (OAFlux) dataset. Though the model is constructed from 1-week lag statistics, it successfully reproduces the observed anomaly evolution through lead times of 90 days, allowing an estimation of the relative roles of coupling and internal atmospheric and oceanic forcing upon North Pacific SSTs. It is found that east of the date line, SST variability is maintained by, but has little effect on, T-A variability. However, in the Kuroshio-Oyashio confluence and extension region, about half of the SST variability is independent of T-A, driven instead by SST noise forcing internal to the ocean. Including surface zonal winds in the analysis does not alter this conclusion, suggesting T-A adequately represents the atmosphere. Repeating the analysis with the output of two control simulations from a fully coupled global climate model (GCM) differing only in their ocean resolution yields qualitatively similar results. However, for the simulation employing the coarse-resolution (1 degrees) ocean model, all SST variability depends upon T-A, apparently caused by a near absence of ocean-induced noise forcing. Collectively, these results imply that a strong contribution from internal oceanic forcing drives SST variability in the Kuroshio-Oyashio region, which may be used as a justification for atmospheric GCM experiments forced with SST anomalies in that region alone. This conclusion is unaffected by increasing the dimensionality of the model to allow for intrabasin interaction.