The atmospheric response to North Atlantic SST anomalies in seasonal prediction experiments

Seasonal forecasts performed over a 26 yr period as part of the Historical Seasonal Forecasting Project (HFP) are used to analyze the influence of North Atlantic sea surface temperature (SST) anomalies on the atmospheric circulation, its seasonality, and model dependence. The signals related to the El Ni˜no events are first removed from both the SST and the atmospheric data. The North Atlantic SST and the ensemble mean forecast are then correlated over the 26 yr to identify the model response to the SST forcing. The signal-to-noise ratio shows that in spring there is a significant forecast signal that is related to the SST anomaly in the North Atlantic. In that season the two models used in the HFP yield responses to the SST anomaly that are both similar to each other and to the observed response. For the other seasons the agreement between the responses and the observed atmospheric anomalies is poor. In winter the response is very sensitive to the model used.