Assessment of responses of North Atlantic winter SST to the NAO in 13 CMIP5 models on the interannual scale

Abstract. This study evaluates the response of winter-averaged sea surface temperature (SST) to the winter North Atlantic Oscillation (NAO) simulated by 13 CMIP5 Earth System Models in the North Atlantic (NA) (0–65° N) on the interannual scale. Only 7 models can reproduce an observed tripolar pattern of the response of SST anomalies to the NAO, and most of the models cannot generate the observed impact of variations of the turbulent heat flux on the response of SST anomalies to the NAO. In the subpolar NA (45–65° N) where the influences of sensible/latent heat fluxes on SST are obvious, most of the models simulate a positive response of SST to the turbulent heat flux in the large area of this region, which is opposite to the observations, and probably generate the incorrect positive response of SST to the NAO in some models. In the subtropical NA (25–45° N), the observations show a significant influence of latent heat flux (LHF) on SST, but the overestimated oceanic role in the interaction of the LHF and SST in most CMIP5 models results in an incorrect positive response of SST anomalies to the LHF anomalies in a large area of the subtropical NA. Besides the turbulent heat flux, the meridional advection is also important to the change of the SST in the NA. The analysis of the simulated and observed results shows that NAO-driven meridional advection can cause the increase/decrease of SST during the positive phase of the NAO in the subtropical/subpolar NA. This is probably one of the main causes why the models can simulate the realistic positive response of SST anomalies to the NAO in the subtropical NA, but the strength of the positive response is relatively weak.