ENSO预报模式及其改进的进展

近年来，为预报ＥＮＳＯ事件已研制出多种方法、包括海气耦合动力学模式，统计模式，物理海洋／统计大气模式等。其中耦合动力学模式的范围从简单的浅水模式到复杂的大气环流模式。在１９８０－１９９２年的这一时段内，一些动力学攻统计模式预报ＥＮＳＯ指数提前６－１２个月的水平已超过了持续性预报。

REVIEW ON THE ENSO PREDICTION MODEL AND ITS IMPROVEMENT

In recent years, varied models which include coupled atmosphere-ocean moedl and statistical model and physical ocean/statistical atmosphere model have been developed for prediction ofENSO events. During pen6d of 1980-1992, some models (dynamical and statistical)perforrnedconsiderably better than the persistence forecast on predicting typical indices of ENSO on leadtime of 6 to 12 months. However, nearly all models have failed in forecasting SST changes inthe Easter Equatorial Pacific since 1992. It is a challenge not only to the models but also to theunderstanding of the mechanism of ENSO cycle. Thus, scientists are working hard for improving their models. This paper briefly introduces some models and their prediction results, and illustrate the understanding of the mechanism of ENSO cycle and recent work for improving Ze biak-Cane model by author.