运用CSVD和联合CSVD等较新颖的统计方法,在去除/未去除ENSO影响的思路下,探讨了印度洋海温异常和南海夏季风建立迟早的关系,结果表明:在没有去除ENSO信号(外部作用)影响的情况下,全区一致型的海温异常分布对南海夏季风建立迟早起着重要的作用。当全区温度距平为正(负)时,南海夏季风建立较晚(早)。在去除了ENSO信号的影响后,非ENSO全区一致型和SIODM型是影响南海夏季风建立早晚的两个主要的印度洋海温分布型。对于非ENSO全区一致型的海温分布,当前期海温全区为负(正)距平时,南海夏季风建立较早(晚)。而对于SIODM型的海温分布,则当前期海温距平为西负东正(西正东负)的SIODM型时,南海夏季风建立较早(晚)。
The relationship between the Indian Ocean sea surface temperature anomaly (SSTA) and the onset of South China Sea summer monsoon (SCSSM) has been analyzed with 1953 - 1998 Reynolds & Smith monthly sea surface temperature (SST) as well as NCEP/NCAR monthly re-analysis data based on Empirical Orthogonal Function (EOF), the new statistical method CSVD (Conditional Singular Value Decomposition) and CCSVD (Confederate Conditional Singular Value Decomposition). EOF analysis results show that there exist two major patterns of Indian SSTA during all the seasons, one is called the unipole mode which is closely related to ENSO, the other is called the southern Indian Ocean Dipole mode (SIODM). These two modes perform with obvious interdecadal variation. In order to distinguish the impact of ENSO on the Indian Ocean SSTA, Conditional Singular Value Decomposition/ Confederate Conditional Singular Value Decomposition methods are applied in this study. The results show with the influence of ENSO, positive (negative) unipole pattern of the Indian Ocean SSTA is one of the major affecting patterns that will lead to the late (early) onset of SCSSM. Without the influence of ENSO, none ENSO unipole pattern and SIODM pattern are in order the two major affecting patterns: negative (positive) none ENSO unipole pattern may contribute to the early (late) onset of SCSSM, while negative (positive) SIODM pattern plays a role in the early (late) onset of SCSSM.