摘要
基于ＮＣＥＰ／ＮＣＡＲ再分析资料，利用人工神经网络ＢＰ模型及其优化算法，建立起了赤道太平洋纬向风和滞后的东太平洋海温之间的映射关系和预报模型，成功地模拟和仿真出了１９８２／１９８３、１９８６／１９８７年发生的强ＥＩ Ｎｉｎｏ现象和１９８８年的强Ｌａ Ｎｉｎａ现象。结果表明，这种方法可有效用于辩识和反演复杂的大气、海洋动力系统及其预报模型。
Based on NCEP/NCAR data sets, and by using BP neural network model and its improved calculation method, the mapped relation and prediction model between the equatorial east /west-Pacific zonal wind and the lag sea temperature of east-Pacific are established. Using the SST neural network model, the marked EI Nino of 1982/1983 and noted La Nina of 1988 are successfully simulated and imitated. This method can be effectively used for identifying and retrieving the complex atmosphere/sea dynamic system and their prediction model.
出处
《海洋通报》
CAS
CSCD
北大核心
2000年第4期1-7,共7页
Marine Science Bulletin
基金
国家自然科学基金
总参预报技术课题
