海洋同化系统为年代际预测试验提供初值,其性能可能会影响年代际预测技巧,因此评估其对重要年代际变化现象的模拟能力非常必要。观测发现,东亚夏季风(EASM)和前冬厄尔尼诺—南方涛动(ENSO)的关系在1970s末加强,随后在1990s中期后减弱。基于FGOALS-s2耦合气候模式的海洋同化系统评估了其对这2次年代际变化的模拟能力。结果表明,决定模式能否再现EASM和前冬ENSO关系的年代际变化有2个重要因素:①与前冬ENSO有关的夏季印度洋—太平洋海温型的年代际变化;②模式西北太平洋反气旋对热带海温的响应偏差。模式中西北太平洋反气旋与东北印度洋的暖海温关系稳定,当1970s末前冬ENSO对夏季印度洋海温影响显著增强时,模式能够模拟出北印度洋降水以及赤道东印度洋至海洋大陆上空Kelvin波的增强,从而可再现EASM与前冬ENSO关系的增强;而1990s中期后模式中与前冬ENSO有关的东北印度洋海温异常进一步增强,与观测相反,使得模式未能再现观测中EASM与前冬ENSO关系的减弱。此外,1990s中期后模式对夏季中太平洋冷海温异常的Rossby波响应存在较大偏差,是其未能再现此次年代际变化的另一个原因。研究表明,与ENSO有关的热带印太海温的年代际变化预测水平和模式对海温的响应偏差将在一定程度上制约模式对EASM与ENSO关系的年代际预测能力。

Ocean Assimilation System (OAS) is an important component for decadal prediction experiment, providing initial conditions. Evaluating the atmosphere response in OAS can provide reference for analyzing results from decadal prediction. We analyzed the interdecadal change in relation between the East Asian Summer Monsoon (EASM) and El Niño/Southern Oscillation (ENSO) in the previous winter based on an OAS on the coupled climate model FGOALS-s2. It shows that two factors impact the performance: ① interdecadal change of Ssea Surface Temperature (SST) pattern in the summer Indo-Pacific Basin related with ENSO in previous winter and ② bias in model response of the western North Pacific anticyclone to tropical SST anomalies. The anticyclone shows steady relation with the warm eastern Indian Ocean. When ENSO’s impact on the summer Indian Ocean is strengthened around the end of 1970s, the OAS can reproduce the strengthened EASM-ENSO relation. However, the trend of intensified EASM-ENSO relation in the OAS is still significant after the mid-1990s due to the stronger link between the anticyclone and the northeastern Indian Ocean, differing with the observation which shows a weakened effect of the Indian Ocean on the anticyclone. In addition, the bias in response to the SST anomalies in the central Pacific also partly contributes to the failure in reproducing the weakening EASM-ENSO relation after the mid-1990s. It implies that prediction skill of interdecadal ENSO impact on the tropical Indo-Pacific SST and response bias of model to SST anomalies may to some extent limit the capability to predict the interdecadal change in the EASM-ENSO relation.