Hai Lin
The North Atlantic Oscillation (NAO) has been known as an intrinsic
atmospheric variability, which is excited by the chaotic dynamics of the
atmosphere. From this point of view, it is unlikely that any skillful
seasonal prediction can be achieved for this large-scale phenomenon. In
this talk, an alternative mechanism is proposed that links the NAO to the
diabatic heating anomaly in the tropical Pacific. This link is
demonstrated using the observational data and the ensemble integrations of
the GEM and GCM3. Outputs of wintertime seasonal forecasts from the GEM
and GCM3 as part of the Historical Forecasting Project (HFP2) are
analyzed. A correction scheme is formulated that associates the
atmospheric signal to the SST structures in the tropical Pacific and
reduces the model-dependent biases in the ensemble forecast. After the
correction, both models give a skillful seasonal forecast of the NAO.