论文部分内容阅读
采用奇异谱分析（ＳＳＡ）与自回归（ＡＲ）预测模型相结合的方案，对Ｎｉｎｏ海区平均ＳＳＴ逐月距平序列作自适应滤波意义下的超前预报。结果表明，对１９９７－１９９８年这次强ＥＮＳＯ事件的超前预报十分有效；利用相应的历史样本作三次强ＥＮＳＯ事件的回溯预报试验，发现均有较高可信度。可见，该方案预报技巧稳定，独立样本试验和实际预报试验都有很高的准确率。将ＳＳＡＡＲ方案进一步完善，可望作为ＥＮＳＯ业务预报的有效模型。
By using the combination of singular spectrum analysis (SSA) and autoregressive (AR) prediction model, the prediction of mean SST monthly anomalies in Nino sea area is adaptively filtered. The results show that it is very effective to predict the strong ENSO event from 1997 to 1998. Using the corresponding historical samples for the back-forecasting test of the third-strong ENSO event, it is found that there is a high degree of credibility. Can be seen that the program forecasting techniques are stable, independent sample test and the actual forecast test has a high accuracy. Further refinement of the SSA-AR program is expected to be an effective model for ENSO business forecasting.