ENSO的春季预报障碍问题是ENSO可预报性研究的一个重要内容。在业务预报中，春季预报障碍问题仍然严重影响着ENSO的预报技巧。本项目拟从初始误差增长的角度探讨两类El Niño事件的春季预报障碍现象。项目将基于模式扰动通过资料同化方法再现两类El Niño事件，在此基础上采用条件非线性最优扰动（CNOP）方法，寻找对两类El Niño事件预报结果具有最大影响的初始误差（即CNOP型初始误差）；考察CNOP型初始误差的季节演变特征，探讨两类El Niño事件春季预报障碍现象的区别和联系。对比最易导致两类El Niño事件显著春季预报障碍的初始误差的异同，揭示两类El Niño事件预测的初值敏感区。考察上述敏感区内初始场质量的改善在提高两类El Niño事件预报技巧中的有效性，从而为优化ENSO的观测提供理论指导。

The spring predictability barrier problem is an important part of ENSO predictability studies, which affects ENSO prediction skills seriously in operational forecast. In this project, the spring predictability barrier problem for two types of El Niño events is investigated from the perspective of initial error evolution. This is enabled by tracing the evolution of a conditional nonlinear optimal perturbation (CNOP), superimposed on an observing system simulation of two types of El Niño events that act as the initial error with the biggest negative effect on the El Niño predictions. The seasonal dependence of CNOP-type error growth will be analyzed related to the spring predictability barrier for two types of El Niño events. Meanwhile, the differences between CNOP-type errors of two types of El Niño events that lead to prominent spring predictability barrier will be discussed, during which, the sensitive area of the predictions for two types of El Niño events demonstrated by CNOP-type errors will be studied. Furthermore, sensitivity experiments are conducted to examine the reliability of results obtained above. These theoretical results could be applied into targeted observation of two types of El Niño events and help improve prediction skills of the events as well.