论文部分内容阅读
基于一个中等复杂程度模式(ICM)集合预报系统(EPS)产生的海表温度距平(SSTA)预报产品,从误差增长的角度探讨了2015/16超强El Nio事件的“春季可预报性障碍”(SPB)问题.通过分析集合预报成员预报误差的增长倾向,发现了2015/16 El Nio事件的预报误差增长呈现显著的季节依赖性,且在春/夏季具有最大增长率,表明ICM-EPS对2015/16 El Nio事件的预报发生了明显的SPB现象.进一步分析表明,上述SPB现象不是由ICM初始场的不确定性引起,而是由其模式误差导致,而ICM-EPS集合预报成员的平均滤掉了部分模式误差的影响,减弱了SPB现象,从而使得2015/16 El Nio事件的预报产生较小的预报误差.通过探讨由海温方程的倾向误差表征的模式误差,该研究揭示了导致SPB现象发生的倾向误差的主要空间特征,并阐明了ICM-EPS低估2015/16El Nio事件强度的原因.此外,本文也揭示了导致显著SPB现象,尤其是导致最大预报误差的SSTA倾向误差的结构特征.该倾向误差的SSTA分量具有赤道中东太平洋负异常,西太平洋正异常的纬向偶极子结构,与Duan等提出的最敏感非线性强迫奇异向量(NFSV)-倾向误差高度相似,从而表明NFSV-型倾向误差也存在于实际的El Nio预报中.该研究也探讨了其他超强El Nio事件,如1982/83和1997/98事件,得到了类似的结果.因此,如果利用NFSV-型倾向误差校正ICM模式误差,ICM-EPS可大大提高超强El Nio事件的预报技巧.
Based on a SSTA forecasting system based on an intermediate complex pattern (ICM) ensemble forecasting system (EPS), the authors discussed the springtime variability of the 2015/16 super El Niño event Forecasting obstacles "(SPB) problem.By analyzing the increasing trend of forecasting errors of aggregate forecasting members, it is found that the prediction error of 2015/16 El Nio event shows a significant seasonal dependence and has the largest increase in spring / summer Rate, indicating that the ICM-EPS has obvious SPB phenomenon in the prediction of 2015/16 El Nio event.Further analysis shows that the above SPB phenomenon is not caused by the uncertainty of the initial ICM field but due to its model error, However, the average of the ICM-EPS ensemble members filtered out the effects of partial pattern errors, and the SPB phenomenon was weakened, which resulted in smaller forecast errors in the forecast of 2015/16 El Nio events. This study reveals the main spatial characteristics of propensity errors that cause SPB phenomena and clarifies why ICM-EPS underestimates the intensity of the 2015 / 16El Nio event.Furthermore, this paper also reveals the causes of significant SPB phenomena ,especially Is the structural characteristic of the SSTA trend error which leads to the largest forecast error.The SSTA component of this trend error has the equatorial central-eastern Pacific anomalous and the western Pacific positive anomalous zonal dipole structure, which is similar to the most sensitive nonlinear forced singular vector proposed by Duan et al (NFSV) - The propensity errors are highly similar, indicating that the NFSV-type propensity errors also exist in the actual El Nio forecast. Other El Nio events such as 1982/83 and 1997/98 were also explored Event, a similar result is obtained, so ICM-EPS can greatly improve the prediction skills of the super El Nio event if the ICM pattern error is corrected using the NFSV-type propensity error.