Diagnoses and Predictions of Time-averaged Circulations
平均环流的诊断与预报

A barotropic linear model is developed for time- averaged circulation forecasts. The relative contribution from each factor is analysed using observation data . It is shown that the accuracy of the persistence scheme in which the initial state alone is included drops rapidly with the increase of the mean time , especially for the planetary waves .The modified persistence scheme in which the dynamical processes of the advection and the spherical effect of the earth are taken into account of is not better than the pure persistence . The longer the mean time , the worse the results . The decrease of pattern correlations occur at almost all wave numbers. When the divergence correction is taken into account, the results are considerably improved for the planetary and long waves .The crucial factor for improving the prediction is the unknown forcing. The possible approaches of improving time- averaged circulation forecasts are also proposed in this paper .